(Themy 4

The Mycota
Edited by
K. Esser
The Mycota
I Growth, Differentiation and Sexuality

1st edition ed. by J.G.H. Wessels and F. Meinhardt
2nd edition ed. by U. Kües and R. Fischer
II Genetics and Biotechnology

Ed. by U. Kück
III Biochemistry and Molecular Biology

Ed. by R. Brambl and G. Marzluf
IV Environmental and Microbial Relationships

Ed. by D. Wicklow and B. Söderström
V Plant Relationships
Ed. by G. Carroll and P. Tudzynski
VI Human and Animal Relationships

Ed. by D.H. Howard and J.D. Miller
VII Systematics and Evolution

Ed. by D.J. McLaughlin, E.G. McLaughlin, and P.A. Lemke†
VIII Biology of the Fungal Cell

Ed. by R.J. Howard and N.A.R. Gow
IX Fungal Associations
Ed. by B. Hock
X Industrial Applications
Ed. by H.D. Osiewacz
XI Agricultural Applications
Ed. by F. Kempken
XII Human Fungal Pathogens

Ed. by J.E. Domer and G.S. Kobayashi
XIII Fungal Genomics

Ed. by A.J.P. Brown
The Mycota
A Comprehensive Treatise
on Fungi as Experimental Systems
for Basic and Applied Research
Edited by K. Esser
XIII
Volume Editor:
Fungal Genomics
A.J.P. Brown
With 37 Figures, 6 in Color, and 16 Tables
123
Series Editor
Professor Dr. Dr. h.c. mult. Karl Esser
Allgemeine Botanik
Ruhr-Universität
44780 Bochum, Germany
Tel.: +49 (234)32-22211
Fax.: +49 (234)32-14211
e-mail: Karl.Esser@rub.de
Volume Editor
Professor Dr. Alistair J.P. Brown
Aberdeen Fungal Group
School of Medical Sciences
Institute of Medical Sciences
University of Aberdeen
Foresterhill
Aberdeen AB25 2ZD, UK
Tel.: +44 (1224) 555883
Fax: +44 (1224) 555844
e-mail: al.brown@abdn.ac.uk
Library of Congress Control Number: 2005928810
ISBN-10 3-540-25594-X Springer Berlin Heidelberg New York

ISBN-13 978-3-540-25594-9 Springer Berlin Heidelberg New York
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the
rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way,
and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German
Copyright Law of September 9, 1965, in its current version, and permissions for use must always be obtained from Springer-Verlag.
Violations are liable for prosecution under the German Copyright Law.
Springer is a part of Springer Science+Business Media

springeronline.com
© Springer-Verlag Berlin Heidelberg 2006
Printed in Germany
The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of
a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general
use.
Editor: Dr. Dieter Czeschlik, Heidelberg, Germany

Desk editor: Dr. Andrea Schlitzberger, Heidelberg, Germany
Cover design: Erich Kirchner, Heidelberg, Germany
Production: Katja Röser, Leipzig, Germany
Typesetting: LE-TEX Jelonek, Schmidt & Vöckler GbR, Leipzig, Germany
Printed on acid-free paper 31/3152 5 4 3 2 1 0

Karl Esser
(born 1924) is retired Professor of General Botany and Direc-
tor of the Botanical Garden at the Ruhr-Universität Bochum
(Germany). His scientific work focused on basic research in
classical and molecular genetics in relation to practical appli-
cation. His studies were carried out mostly on fungi. Together
with his collaborators he was the first to detect plasmids in
higher fungi. This has led to the integration of fungal genet-
ics in biotechnology. His scientific work was distinguished
by many national and international honors, especially three
honorary doctoral degrees.
Alistair J.P. Brown

(born 1955) is Professor in Molecular and Cell Biology at
the University of Aberdeen (UK). His postdoctoral work at
MIT was followed by faculty positions in the Genetics De-
partment at Glasgow University (UK), and then in the School
of Medical Sciences at Aberdeen University. His work has fo-
cussed on yeast gene regulation and the control of Candida
albicans virulence, morphogenesis and stress responses. His
group routinely combines genomic technologies with molec-
ular and cellular approaches. Along with his colleagues Neil
Gow and Frank Odds, he leads the Aberdeen Fungal Group.
He also leads Proteomics and Transcript Profiling facilities at
Aberdeen University.
Series Preface
Mycology, the study of fungi, originated as a subdiscipline of botany and was a descrip-
tive discipline, largely neglected as an experimental science until the early years of this
century. A seminal paper by Blakeslee in 1904 provided evidence for selfincompatibil-
ity, termed “heterothallism”, and stimulated interest in studies related to the control
of sexual reproduction in fungi by mating-type specificities. Soon to follow was the
demonstration that sexually reproducing fungi exhibit Mendelian inheritance and that
it was possible to conduct formal genetic analysis with fungi. The names Burgeff, Kniep
and Lindegren are all associated with this early period of fungal genetics research.
These studies and the discovery of penicillin by Fleming, who shared a Nobel Prize
in 1945, provided further impetus for experimental research with fungi. Thus began a
period of interest in mutation induction and analysis of mutants for biochemical traits.
Such fundamental research, conducted largely with Neurospora crassa, led to the one
gene: one enzyme hypothesis and to a second Nobel Prize for fungal research awarded to
Beadle and Tatum in 1958. Fundamental research in biochemical genetics was extended
to other fungi, especially to Saccharomyces cerevisiae, and by the mid-1960s fungal
systems were much favored for studies in eukaryotic molecular biology and were soon
able to compete with bacterial systems in the molecular arena.
The experimental achievements in research on the genetics and molecular biology of
fungi have benefited more generally studies in the related fields of fungal biochemistry,
plant pathology, medical mycology, and systematics. Today, there is much interest in the
genetic manipulation of fungi for applied research. This current interest in biotechnical
genetics has been augmented by the development of DNAmediated transformation
systems in fungi and by an understanding of gene expression and regulation at the
molecular level. Applied research initiatives involving fungi extend broadly to areas of
interest not only to industry but to agricultural and environmental sciences as well.
It is this burgeoning interest in fungi as experimental systems for applied as well as
basic research that has prompted publication of this series of books under the title The
Mycota. This title knowingly relegates fungi into a separate realm, distinct from that of
either plants, animals, or protozoa. For consistency throughout this Series of Volumes
the names adopted for major groups of fungi (representative genera in parentheses) are
as follows:
Pseudomycota
Division: Oomycota (Achlya, Phytophthora, Pythium)
Division: Hyphochytriomycota
Eumycota
Division: Chytridiomycota (Allomyces)
Division: Zygomycota (Mucor, Phycomyces, Blakeslea)
Division: Dikaryomycota
VIII Series Preface
Subdivision: Ascomycotina
Class: Saccharomycetes (Saccharomyces, Schizosaccharomyces)
Class: Ascomycetes (Neurospora, Podospora, Aspergillus)
Subdivision: Basidiomycotina
Class: Heterobasidiomycetes (Ustilago, Tremella)
Class: Homobasidiomycetes (Schizophyllum, Coprinus)
We have made the decision to exclude from The Mycota the slime molds which, although
they have traditional and strong ties to mycology, truly represent nonfungal forms
insofar as they ingest nutrients by phagocytosis, lack a cell wall during the assimilative
phase, and clearly show affinities with certain protozoan taxa.
The Series throughout will address three basic questions: what are the fungi, what do
they do, and what is their relevance to human affairs? Such a focused and comprehensive
treatment of the fungi is long overdue in the opinion of the editors.
A volume devoted to systematics would ordinarily have been the first to appear
in this Series. However, the scope of such a volume, coupled with the need to give
serious and sustained consideration to any reclassification of major fungal groups, has
delayed early publication. We wish, however, to provide a preamble on the nature of
fungi, to acquaint readers who are unfamiliar with fungi with certain characteristics
that are representative of these organisms and which make them attractive subjects for
experimentation.
The fungi represent a heterogeneous assemblage of eukaryotic microorganisms.
Fungal metabolism is characteristically heterotrophic or assimilative for organic carbon
and some nonelemental source of nitrogen. Fungal cells characteristically imbibe or
absorb, rather than ingest, nutrients and they have rigid cell walls. The vast majority of
fungi are haploid organisms reproducing either sexually or asexually through spores.
The spore forms and details on their method of production have been used to delineate
most fungal taxa. Although there is a multitude of spore forms, fungal spores are
basically only of two types: (i) asexual spores are formed following mitosis (mitospores)
and culminate vegetative growth, and (ii) sexual spores are formed following meiosis
(meiospores) and are borne in or upon specialized generative structures, the latter
frequently clustered in a fruit body. The vegetative forms of fungi are either unicellular,
yeasts are an example, or hyphal; the latter may be branched to form an extensive
mycelium.
Regardless of these details, it is the accessibility of spores, especially the direct
recovery of meiospores coupled with extended vegetative haploidy, that have made
fungi especially attractive as objects for experimental research.
The ability of fungi, especially the saprobic fungi, to absorb and grow on rather
simple and defined substrates and to convert these substances, not only into essential
metabolites but into important secondary metabolites, is also noteworthy. The metabolic
capacities of fungi have attracted much interest in natural products chemistry and in
the production of antibiotics and other bioactive compounds. Fungi, especially yeasts,
are important in fermentation processes. Other fungi are important in the production
of enzymes, citric acid and other organic compounds as well as in the fermentation of
foods.
Fungi have invaded every conceivable ecological niche. Saprobic forms abound,
especially in the decay of organic debris. Pathogenic forms exist with both plant and
animal hosts. Fungi even grow on other fungi. They are found in aquatic as well as
soil environments, and their spores may pollute the air. Some are edible; others are
poisonous. Many are variously associated with plants as copartners in the formation of
lichens and mycorrhizae, as symbiotic endophytes or as overt pathogens. Association
with animal systems varies; examples include the predaceous fungi that trap nematodes,
Series Preface IX
the microfungi that grow in the anaerobic environment of the rumen, the many insec-
tassociated fungi and the medically important pathogens afflicting humans. Yes, fungi
are ubiquitous and important.
There are many fungi, conservative estimates are in the order of 100,000 species,
and there are many ways to study them, from descriptive accounts of organisms found
in nature to laboratory experimentation at the cellular and molecular level. All such
studies expand our knowledge of fungi and of fungal processes and improve our ability
to utilize and to control fungi for the benefit of humankind.
We have invited leading research specialists in the field of mycology to contribute
to this Series. We are especially indebted and grateful for the initiative and leadership
shown by the Volume Editors in selecting topics and assembling the experts. We have all
been a bit ambitious in producing these Volumes on a timely basis and therein lies the
possibility of mistakes and oversights in this first edition. We encourage the readership
to draw our attention to any error, omission or inconsistency in this Series in order that
improvements can be made in any subsequent edition.
Finally, we wish to acknowledge the willingness of Springer-Verlag to host this
project, which is envisioned to require more than 5 years of effort and the publication
of at least nine Volumes.
Bochum, Germany Karl Esser
Auburn, AL, USA Paul A. Lemke
April 1994 Series Editors
Addendum to the Series Preface
In early 1989, encouraged by Dieter Czeschlik, Springer-Verlag, Paul A. Lemke and

I began to plan The Mycota. The first volume was released in 1994, 12 volumes followed
in the subsequent years. Unfortunately, after a long and serious illness, Paul A. Lemke
died in November 1995. Thus, it was my responsibility to proceed with the continuation
of this series, which was supported by Joan W. Bennett for Volumes X–XII.
The series was evidently accepted by the scientific community, because first volumes
are out of print. Therefore, Springer-Verlag has decided to publish some of the completely
revised and updated new editions of Volumes I, II, III, IV, VI, and VIII. I am glad that
most of the volume editors and authors have agreed to join our project again. I would like
to take this opportunity to thank Dieter Czeschlik, his colleague, Andrea Schlitzberger,
and Springer-Verlag for their help in realizing this enterprise and for their excellent
cooperation for many years.
Bochum, Germany Karl Esser

July 2005
Volume Preface
The Fungal Genomics era is gathering pace, and this is an enormously exciting time
for mycologists. Having trained during an era when whole PhD theses were devoted to
the sequencing of a single gene, it is incredibly exciting to run research programmes
in the current era, when a fungal genome sequence can be generated in a matter
of weeks. Genome sequences for a whole range of fungal species have been made
public in the last few years, including those for Aspergillus fumigatus, Candida glabrata,
Cryptococcus neoformans, Debaryomyces hansenii, Kluyveromyces lactis, Magnaporthe
grisea, Neurospora crassa and Yarrowia lipolytica. Furthermore, many more genome
sequences (for other Candida, Coccidioides, Histoplasma, Mycosphaerella, Pneumocystis
and Saccharomyces species, for example) will be released in the near future and, in all
probability, the generation of additional fungal genome sequences will continue for
some time at a rate of about one every 60 days. These genome sequences are providing
a wealth of invaluable data about the evolution, life cycles, cell biology, and virulence of
fungi.
Genomic technologies have developed to differing extents, depending upon the fun-
gal species one is interested in. The genome sequences for some fungal species are only
just becoming available. For these species, genomic analyses are in the very early stages
of development. The initial process of annotating genes within a genome sequence is
incomplete for many fungi, and the generation of microarrays for transcript profiling
analysis is still some way off. In contrast, a decade has passed since the Saccharomyces
cerevisiae genome sequence was first made public. This was one of the first genome se-
quences available for any cellular organism and, thanks to the sophisticated molecular
toolbox available for bakers’ yeast, this model organism is now leading the way in many
areas of functional genomics and systems biology. (Nearly) complete sets of knockout,
conditional and epitope tagged mutants are available for the genome-wide analyses of
gene function in yeast. Researchers are spoilt for choice with respect to the type and
format of yeast microarray which is available – oligonucleotide or gene arrays for tran-
script profiling, or intergenic arrays for genome-wide chromatin immunoprecipitation
studies. Protein microarrays are now available for S. cerevisiae, and high-throughput
yeast proteomics is now being exploited by numerous research groups. Massively par-
allel analyses of protein–protein interactions in yeast were published several years ago.
Metabolomics is already being applied to system-wide analyses of yeast physiology.
Therefore, for many research groups, the problem has shifted from the generation of
data to the accurate interpretation and incisive exploitation of massive datasets. Thank-
fully, increasingly sophisticated software tools are being developed for the integration of
genetic, transcriptomic, proteomic, interactomic and metabolomic datasets. As a result,
genomics is now revealing important global perspectives of fungal cell biology which
were not obvious using standard reductionist approaches. There were two main chal-
lenges in editing this volume on Fungal Genomics. First, it was impossible to provide
comprehensive coverage of this immense field. Second, the speed of development of
Fungal Genomics precluded the publication of a text which is bang up to date. New
developments do not stop for editors! Therefore, this volume does not attempt to cover
every topic or to include every last-minute development. (I apologise to those whose
XIV Volume Preface
favourite topic is not covered.) Rather, my aims have been twofold. In the 13 chapters of
this 13th volume of The Mycota, my first aim has been to illustrate the current impact
and potential future impact of genomics in different fungal species – in those where
genomics is mature, and others where genomics is relatively immature. Second, I wanted
to show how genomics is being applied to a diverse range of interesting questions in
mycological research. The chapters illustrate fundamental principles of fungal genomics
which are universally applicable. Mycota XIII is divided into three sections, the first of
which addresses fungal systems biology and evolution. The volume starts with a chapter
on Saccharomyces cerevisiae, as befits its status as the pre-eminent model organism in
the genomics era. Also, this chapter addresses the metabolomics and systems biology
of bakers’ yeast, thereby immediately reminding the reader that there is much more
to genomics than genome sequencing and transcript profiling. The other chapters in
this section illustrate the huge impact of genome sequencing upon our understand-
ing of fungal evolution, and the exploitation of novel bioinformatic approaches in the
identification of genes associated with fungal virulence.
The second section, on fungal rhythms and responses, addresses some of the most
topical issues in fungal biology – circadian rhythms, programmed cell death, stress
responses, secretion and polarised growth, and cellular morphogenesis. These chapters
cover a range of fungal species, including filamentous fungi and yeasts. As well as
describing the massive impact of fungal genomics in these species, they highlight areas
where genomics has the potential to significantly accelerate our research efforts.
The last section focuses on fungal pathogenesis, an area of great medical significance.
Chapters in this section discuss how both proteomic and transcriptomic approaches have
provided, and are helping to provide, important new insights into the pathobiology of
some of the major fungal pathogens of humans.
I am very grateful to the authors for their outstanding contributions to Mycota
XIII. All are internationally renowned in their chosen fields, and all are extremely busy
people. I appreciate their willingness to commit valuable time to this rewarding and
interesting project.
Aberdeen, UK Alistair J.P. Brown

July 2005 Volume Editor
Contents
Biocemistry and Molecular Genetics
1 Metabolomics and Systems Biology in Saccharomyces cerevisiae

J.I. Castrillo, S.G. Oliver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 Genome Evolution in Hemiascomycete Yeasts

L.J. Montcalm, K.H. Wolfe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 Investigating the Evolution of Fungal Virulence by Functional Genomics

S. Ahmad, D.M. Soanes, M.C. Barooah, N.J. Talbot . . . . . . . . . . . . . . . . . . 35
Fungal Rythms and Responses
4 Circadian Rhythms, Photobiology and Functional Genomics in Neurospora

J.J. Loros, J.C. Dunlap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5 Genomics of Protein Secretion and Hyphal Growth in Aspergillus

D.B. Archer, G. Turner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6 The Genomics of Stress Response in Fission Yeast

B.T. Wilhelm, J. Bähler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7 Programmed Cell Death and Apoptosis in Fungi

M. Ramsdale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8 Genomic Analysis of Cellular Morphology in Candida albicans

M. Whiteway, A. Nantel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Fungal Pathogenicity
9 Postgenomic Approaches to Analyse Candida albicans Pathogenicity

C.A. Munro, C. Fradin, O. Bader, B. Hube . . . . . . . . . . . . . . . . . . . . . . . . . . 163
10 Integration of Metabolism with Virulence in Candida albicans

A.J.P. Brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
11 Regulators of Candida glabrata Pathogenicity

K. Haynes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
12 Using Genomics to Study the Life Cycle of Histoplasma capsulatum

A. Sil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
XVI Contents
13 Cryptococcus neoformans Pathogenicity

R.T. Nelson, J.K. Lodge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Biosystematic Index
Subject Index
List of Contributors
Sarah Ahmad School of Biological and Chemical Sciences, University of Exeter, Wash-
ington Singer Laboratories, Perry Road, Exeter EX4 4QG, UK
David B. Archer (e-mail: david.archer@nottingham.ac.uk, Tel.: +44-115-9513313)

School of Biology, University of Nottingham, University Park, Nottingham NG7 2RD,
UK
Oliver Bader Robert Koch Institut, NG4, Nordufer 20, 13353 Berlin, Germany
Jürg Bähler (e-mail: jurg@sanger.ac.uk, Tel.: +44-1223-494861) The Wellcome Trust

Sanger Institute, Hinxton, Cambridge CB10 1SA, UK
Madhumita C. Barooah School of Biological and Chemical Sciences, University of

Exeter, Washington Singer Laboratories, Perry Road, Exeter EX4 4QG, UK
Alistair J.P. Brown (e-mail: al.brown@abdn.ac.uk, Tel.: +44-1224-555883,

Fax: +44-1224-555844) Aberdeen Fungal Group, School of Medical Sciences, Institute of
Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UK
Juan I. Castrillo (e-mail: juan.i.castrillo@man.ac.uk) Faculty of Life Sciences, Michael

Smith Building, University of Manchester, Oxford Road, Manchester M13 9PT, UK
Jay C. Dunlap Departments of Genetics and Biochemistry, 704 Remsen, Dartmouth

Medical School, Hanover, NH 03755, USA
Chantal Fradin Robert Koch Institut, NG4, Nordufer 20, 13353 Berlin, Germany
Ken Haynes (e-mail: k.haynes@imperial.ac.uk, Tel.: +44-20-83831245,

Fax: +44-20-83833394) Department of Infectious Diseases, Imperial College London,
Du Cane Road, London W12 0NN, UK
Bernhard Hube (e-mail: HubeB@rki.de, Tel.: +49-1888-7542116,

Fax: +49-30-45472328) Robert Koch Institut, NG4, Nordufer 20, 13353 Berlin, Germany
Jennifer K. Lodge (e-mail: lodgejk@slu.edu, Tel.: +1-314-5778143,

Fax: +1-314-5778156) Edward A. Doisy Department of Biochemistry and Molecular
Biology, Saint Louis University School of Medicine, 1402 S. Grand Blvd., St. Louis, MS
63104, USA
Jennifer J. Loros (e-mail: jennifer.loros@dartmouth.edu, Tel.: +1-603-6501154,

Fax: +1-603-6501128) Departments of Biochemistry and Genetics, 704 Remsen, Dart-
mouth Medical School, Hanover, NH 03755, USA
XVIII List of Contributors
Louis J. Montcalm Department of Genetics, Smurfit Institute, University of Dublin,

Trinity College, Dublin 2, Ireland
Carol A. Munro (e-mail: c.a.munro@abdn.ac.uk, Tel.: +44-1224-555882,

Medical Sciences, Foresterhill, University of Aberdeen, Aberdeen AB25 2ZD, UK
André Nantel (e-mail: andre@bri.nrc.ca) Biotechnology Research Institute, National

Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec H4P 2R2,
Canada
Rex T. Nelson Edward A. Doisy Department of Biochemistry and Molecular Biology,

Saint Louis University School of Medicine, 1402 S. Grand Blvd., St. Louis, MS 63104,
USA; present address: G329 Agronomy Hall, Iowa State University, Ames, IA 50011, USA
Stephen G. Oliver (e-mail: steve.oliver@manchester.ac.uk, Tel.: +44-161-2751578;

Fax: +44-161-2755082) Faculty of Life Sciences, Michael Smith Building, University of
Manchester, Oxford Road, Manchester M13 9PT, UK
Mark Ramsdale (e-mail: m.ramsdale@abdn.ac.uk, Tel.: +44-1224-555882,

Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UK
Anita Sil (e-mail: sil@cgl.ucsf.edu, Tel.: +1-415-5021805, Fax: +1-415-476820) Depart-

ment of Microbiology and Immunology, University of California, San Francisco, 513
Parnassus, P.O. Box 0414, San Francisco, CA 94143-0414, USA
Darren M. Soanes School of Biological and Chemical Sciences, University of Exeter,

Washington Singer Laboratories, Perry Road, Exeter EX4 4QG, UK
Nicholas J. Talbot (e-mail: n.j.talbot@exeter.ac.uk, Tel.: +44-1392-264673) School of

Biological and Chemical Sciences, University of Exeter, Washington Singer Laboratories,
Perry Road, Exeter EX4 4QG, UK
Geoffrey Turner (e-mail: g.turner@shef.ac.uk, Tel.: +44-114-2226211) Department

of Molecular Biology and Biotechnology, University of Sheffield, Firth Court, Western
bank, Sheffield S10 2TN, UK
Malcolm Whiteway (e-mail: Malcolm.Whiteway@cnrc-nrc.gc.ca,

Tel.: +1-514-4966146, Fax: +1-514-4966213) Biotechnology Research Institute, National
Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec H4P 2R2,
Canada
Brian T. Wilhelm The Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10
1SA, UK
Ken Wolfe (e-mail: khwolfe@tcd.ie, Tel.: +353-1-6081253, Fax: +353-1-6798588) De-

partment of Genetics, Smurfit Institute, University of Dublin, Trinity College, Dublin 2,
Ireland
Biocemistry and Molecular Genetics
1 Metabolomics and Systems Biology in Saccharomyces cerevisiae
J.I. Castrillo1 , S.G. Oliver1
CONTENTS more integrative way (Oliver 1997, 2002; Oliver

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 et al. 1998; ter Linde et al. 1999; Delneri et al. 2001;
II. Saccharomyces cerevisiae: A Model Eukaryote Castrillo and Oliver 2004 and references therein).
and a Reference System in Biology . . . . . . . . 3 At the same time that these techniques are be-
III. Functional Genomics of S. cerevisiae: ing applied and refined, the complexity of biolog-
State of the Art . . . . . . . . . . . . . . . . . . . . . . . . 4
A. Functional Genomics: Levels of Regulation 4 ical systems is being rediscovered. Living things
B. S. cerevisiae Functional Genomics: are made up of thousands of components (genes,
State of the Art . . . . . . . . . . . . . . . . . . . . . 5 transcripts, proteins and metabolites), which are
IV. Metabolomics in Comprehensive subject to modification by post-transcriptional and
Post-Genomic Studies: Towards Systems
Biology Using S. cerevisiae as a Model . . . . . 8 post-translational mechanisms. These components
A. Metabolomic Studies: Metabolic Networks participate in anabolic, catabolic and regulatory
and Participation of Metabolites networks in response to environmental and devel-
in Regulation . . . . . . . . . . . . . . . . . . . . . . 8 opmental signals, many of which are still to be elu-
B. Integrative Studies for Global Description cidated (Castrillo and Oliver 2004 and references
of Cellular Complexity: Cellular Networks
and Systems Biology . . . . . . . . . . . . . . . . . 10 therein). In this context, the utilization of well-
C. S. cerevisiae as a Reference Model defined model systems under controlled conditions
in Systems Biology: Advanced Studies, is of central importance in the drive towards an in-
Biological Networks and Genome-Scale tegrative systems biology perspective of the cell.
Models, and Applications . . . . . . . . . . . . . 11
V. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 13 The purpose of this chapter is to present an
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 up-to-date view of the functional genomics of Sac-
charomyces cerevisiae and its growing potential as
a model system for systems biology studies. The
importance of including metabolomics as part of
I. Introduction
this integrative approach to the study of the eu-
karyotic cell as a biological system is emphasised.
The genomic revolution has been characterized by
the generation of huge amounts of information in
the form of genome sequences, and of their direct
application in comprehensive molecular studies II. Saccharomyces cerevisiae: A Model
and comparative genomics strategies (e.g. Goffeau Eukaryote and a Reference System
et al. 1996, 1997; The International Human Genome in Biology
Mapping Consortium 2001a, 2001b; von Mering
et al. 2002; Gavin and Superti-Furga 2003). In the Saccharomyces cerevisiae is a species of budding
post-genomic era, the explosion of new technolo- yeast, a group of unicellular fungi belonging to the
gies, whose exploitation is becoming progressively phylum Ascomycetes. S. cerevisiae is being used as
more refined, is facilitating the study of biological a model eukaryote in biology because the basic
systems on a genome-wide scale. These studies are mechanisms of DNA replication, chromosomal re-
being performed at different functional genomic combination, cell division, gene expression, and
levels, including the genome, transcriptome, metabolism are generally conserved between yeast
proteome and metabolome, and in a progressively and higher eukaryotes (i.e. mammals; Rose and
1Faculty of Life Sciences, Michael Smith Building, The University Harrison 1987–1995; Sherman 1998, 2002; Castrillo
of Manchester, Oxford Road, Manchester M13 9PT, UK and Oliver 2004).
The Mycota XIII
Fungal Genomics
Alistair J.P. Brown (Ed.)
4 J.I. Castrillo and S.G. Oliver
Among the properties which make S. cerevisiae yeast central metabolic pathways and cybernetic
a particularly suitable organism for biological models) could be applied (Bailey and Ollis 1986;
studies are its Generally Regarded As Safe (GRAS) Castrillo and Ugalde 1994 and references therein;
status, rapid growth, well-dispersed cells, simple Cortassa and Aon 1994; Giuseppin and van Riel
methods of cultivation under controlled condi- 2000; Lei et al. 2001). At this point, attempts to
tions, ease of replica plating and mutant isolation. increase the flux through specific pathways by
Moreover, yeasts represent a well-defined genetic metabolic engineering techniques (Bailey 1991;
system with facile techniques of genetic manipula- Stephanopoulos and Vallino 1991) met with
tion (Brown and Tuite 1998; Sherman 1998, 2002; only limited success, revealing our lack of real
Castrillo and Oliver 2004). understanding of the dynamics of metabolism in
The favourable characteristics of this yeast, S. cerevisiae. Such an outcome was anticipated
together with its easy accessibility due to its (at least, by some) as a direct consequence of
economic importance in beer- and bread-making, the application of the Metabolic Control Analysis
make it a cheap source for biochemical studies. (MCA) theory, a fundamental framework in
Thus, many metabolites and enzymes of central quantitative modelling and metabolic control
metabolic processes (i.e. Embden-Meyerhoff (Kacser and Burns 1973; Kacser 1995; Fell 1997;
pathway, tricarboxylic acid cycle, TCA), and the see also last section).
first complete map of central metabolic pathways If S. cerevisiae is to fulfil its potential as a model
were first unveiled in S. cerevisiae, which is eukaryote, and as a reference ‘biological system’
used as a touchstone model for the study of at the cellular level, the progressive incorporation
the eukaryotic cell (Lehninger 1975; Rose and of functional genomic information at the differ-
Harrison 1987–1995; Fell 1997; Alberts et al. 2002; ent levels (i.e. genome, transcriptome, proteome
Castrillo and Oliver 2004 and references therein). and metabolome) into mathematical models rep-
As a result of all this, a wide knowledge of the resentative of the global behaviour of the cell will
genetics, biochemistry and physiology of this be required (Kitano 2002; Ideker 2004a). S. cere-
yeast is presently available (Rose and Harrison visiae was the first eukaryotic organism for which
1987–1995; Brown and Tuite 1998; Sherman 1998, the complete genome was sequenced (Goffeau et al.
2002; Burke et al. 2000; Guthrie and Fink 2002a, 1996, 1997). Hence, it is at the forefront of the post-
2002b, 2004). genomic era (Castrillo and Oliver 2004; see next
At the same time as being considered a plat- section).
form for basic studies, S. cerevisiae has attracted
considerable interest from the early days of
microbiology as a ‘biological system’, capable
of performing specific biotransformations of III. Functional Genomics
interest to the fermentation industry (Pasteur of S. cerevisiae: State of the Art
1857; Rose and Harrison 1993; Olson and Nielsen
2000; Schwartz 2001; Ton and Rao 2004). As A. Functional Genomics: Levels of Regulation
a consequence of this, S. cerevisiae cultures
have been subjected to a number of modelling A schematic representation of the eukaryotic cell
strategies directed towards the representation of as a global system, with the different levels of
different metabolic and cell biological processes. functional genomics (genome, transcriptome,
These were simple models at first, with limited proteome and metabolome; Oliver 1997, 2002;
information on the metabolism and internal Oliver et al. 1998; Delneri et al. 2001), their
regulatory mechanisms, i.e. basic unstructured localization and interactions, main regulatory
models, in which cell growth was modelled as circuits and relationships is presented in Fig. 1.1.
an autocatalytic process, and the specific rates of Our current view of the eukaryotic cell is that of
individual reactions were based on kinetic models a system characterized by a coordinate integration
(e.g. Monod; cf. Bailey and Ollis 1986; Sinclair and of the different functional genomic levels and
Cantero 1990). As a more complete knowledge of individual networks, in direct relation with the
the metabolic pathways and control mechanisms environment. This system is intrinsically com-
became available, more structured models and plex and involves the integration of regulatory
comprehensive modelling strategies (i.e. metabolic mechanisms at the genomic, transcriptional, post-
steady-state flux models including information on transcriptional, post-translational and metabolic
Metabolomics and Systems Biology in Yeast 5
Fig. 1.1. Eukaryotic cell. Functional genomic levels and regulatory mechanisms
levels (Fafournoux et al. 2000; Muratani and man 2002) was the first eukaryotic organism
Tansey 2003; Verger et al. 2003; Castrillo and for which the whole genome sequence was
Oliver 2004; Choudhuri 2004). Among the most completed and made publicly available (Gof-
relevant mechanisms are epigenetic mechanisms feau et al. 1996, 1997). This sequence has been
(e.g. DNA methylation, histone modifications), further certified as the best annotated eukary-
mRNA splicing and small regulatory RNAs, otic genome (Goffeau 2000). The collection of
protein methylation, glycosylation, ubiquitination complete S. cerevisiae chromosome sequences
and sumoylation, protein–protein and protein– and annotations is available at Goffeau et al.
metabolite interactions, and participation of (1997, http://www.nature.com/genomics/papers/
metabolites together with transcription factors s_cerevisiae.html), and can also be obtained at the
and regulatory proteins in signal transduction National Centre for Biotechnological Information
pathways (Day and Tuite 1998; Castrillo and Oliver (NCBI; http://www.ncbi.nlm.nih.gov/mapview/
2004 and references therein; Choudhuri 2004). map_search. cgi? taxid=4932). In this case, the
nucleotide and protein sequences are provided
by the Saccharomyces Genome Database (SGD;
B. S. cerevisiae Functional Genomics: http://www.yeastgenome.org) and are revised as
State of the Art SGD is updated.
The S. cerevisiae genome contains about 12 Mb
S. cerevisiae (laboratory strain S288C; MAT αSUC2 of DNA distributed between 16 chromosomes
mal mel gal2 CUP1 flo1 flo8-1 hap1; Mortimer which contain a total of about 6000 genes, with
and Johnston 1986; Gaisne et al. 1999; Sher- a relatively low proportion containing introns
Table. 1.1. Characteristics of Saccharomyces cerevisiae tre for Biotechnological Information, NCBI, http://www.
(Goffeau et al. 1996; Sherman 1998, 2002; National Cen- ncbi.nlm.nih.gov/mapview/map_search.cgi?taxid=4932)
S. cerevisiae Characteristics
Lineage Eukaryota; Fungi; Ascomycota; Saccharomycotina;
Saccharomycetes; Saccharomycetales; Saccharomycetaceae;
Saccharomyces
Nuclear genome 16 chromosomes (12,052 Kb)
2 µm circle plasmid (6.3 Kb)
Mitochondrial genome Mitochondrial DNA (850 Kb)
ORFs 6183 ORFs encoding for 5773 proteins
tRNA genes 262 tRNA genes
Introns 3.8%
Intracellular dsRNA viruses 0.1% of total nucleic acid content
Cells Haploid Diploid
Volume per cell (µm3 ) 70 120
Global content 1 mg wet weight or 0.25 mg 1 mg wet weight or 0.25 mg
dry weight contains dry weight contains
∼ 0.28 µg of DNA ∼ 0.42 µg of DNA
20 µg of total RNA 24 µg of total RNA
∼ 1 µg of mRNA ∼ 1.2 µg of mRNA
0.10 mg total protein 0.10 mg total protein
(∼ 4%; Goffeau et al. 1996, 1997; Brown and by random transposon insertion, are used
Tuite 1998; Sherman 1998, 2002). From the whole for global analyses, functional profiling and
genome, a total of 5257 protein-coding genes gene characterization (Ross-Macdonald et al.
have been annotated (i.e. reported to code for 1999; Winzeler et al. 1999; Giaever et al. 2002;
a specific protein with a gene ontology category Scherens and Goffeau 2004 and references
curated with experimental evidence) as of the therein). To complement yeast knockout
end of October 2004 (Yeast Proteome Database, collections, the construction of a collection of
http://proteome.incyte.com; Costanzo et al. 2001; yeast strains in which a tetracycline-responsive
Csank et al. 2002). The main characteristics of S. promoter is inserted upstream of individual
cerevisiae are summarized in Table 1.1. essential genes will allow us to explore the
Once the genome sequence is known, post- function of essential genes via conditional and
genomic studies entail, first, the design and imple- titratable promoter alleles (Eisenstein 2004;
mentation of advanced high-throughput methods Mnaimneh et al. 2004).
and genomic strategies to extract the maximum in- 2. At the gene expression (transcriptome) level,
formation at the different functional genomic lev- microarrays have been widely used for the
els and, second, an efficient analysis of the huge global analysis of yeast gene expression
amount of data generated, in order to extract valid patterns (Lashkari et al. 1997; Wodicka et al.
conclusions and new knowledge (e.g. mechanisms 1997; Spellman et al. 1998), with more recent
of regulation and their integration). The most ad- studies revealing the importance of a careful
vanced functional genomics methods and strate- experimental design, controlled conditions
gies investigated in S. cerevisiae have been reviewed and good strategies for data processing and
by Castrillo and Oliver (2004). A summary of the statistical analysis (ter Linde et al. 1999; Hayes
main databases and resources for analysis of yeast et al. 2002; Boer et al. 2003; Tilstone 2003).
genomic data is presented in Table 1.2. The most In addition, new approaches to analyse not
relevant strategies are: only relative changes in gene expression
(mRNA) levels but also net transcription rates
1. At the genome level, application of molecular on a genomic scale are progressively being
genetic techniques on a large scale for the incorporated (Iyer and Struhl 1996; Hirayoshi
generation of comprehensive collections of and Lis 1999; Garcia-Martinez et al. 2004).
yeast mutants. Thus, for example, single 3. At the proteome level: the first whole-proteome
deletion mutants, double mutants and the microarray was developed for yeast, and ad-
‘TRIPLES’ collection of mutants, obtained vanced strategies for the preparation of such
Table. 1.2. Main resources, tools and databases for individual and integrative analysis of yeast genomics data
Databases/resources Reference
Genome databases
S. cerevisiae Genome Database (SGD) http://www.yeastgenome.org
National Centre for Biotechnological Information http://www.ncbi.nlm.nih.gov/mapview/map_search.cgi?taxid=4932
Munich Information centre http://mips.gsf.de/genre/proj/yeast/index.jsp
for Protein Sequences (MIPS)
S. cerevisiae mutants collection http://www.uni-frankfurt.de/fb15/mikro/euroscarf/complete.html
S. cerevisiae resource center http://depts.washington.edu/∼yeastrc
Gene expression (transcriptome) resources
Microarray standards (MIAME) http://www.mged.org/miame
ArrayExpress http://www.ebi.ac.uk/arrayexpress
Stanford Microarray Database http://genome-www5.stanford.edu
Promoter, regulatory sequences and transcription factor databases
S. cerevisiae promoter database http://cgsigma.cshl.org/jian/
Yeast Transcription Factors http://biochemie.web.med.uni-muenchen.de/YTFD/
and related components (YTF)
TRANSFAC http://www.biobase.de/pages/products/transfac.html
Proteome resources
Proteomics Standards Initiative (PSI) http://psidev.sourceforge.net
Proteomics platform (PEDRo) http://pedro.man.ac.uk/
Protein–protein interactions (von Mering et al. 2002)
Yeast proteins localization http://yeastgfp.ucsf.edu/
Yeast protein microarrays (Zhu and Snyder 2003)
Metabolic pathways databases
KEGG http://www.genome.ad.jp/kegg/
BioCyC http://biocyc.org
PathDB http://www.ncgr.org/pathdb/
RIKEN http://genome.gsc.riken.go.jp/DNA-Book/metabolome.shtml
Biomolecular interactions, BIND http://www.blueprint.org/bind/bind.php
Tools for analysis and management of global genomic information
Gene Ontology tools http://www.geneontology.org/GO.tools.shtml
- GoMiner http://discover.nci.nih.gov/gominer/
- GenMAPP http://www.genmapp.org/
Genome Information Management http://www.cs.man.ac.uk/img/gims/ (Cornell et al. 2003)
System (GIMS)
Systems biology resources
Yeast Systems Biology Network http://www.ysbn.org
Systems Biology DataBase (SBDB) http://www.sysbioldb.org
arrays have been developed (Zhu et al. 2001, oped (Gonzalez et al. 1997; Raamsdonk et al.
2003; Zhu and Snyder 2003). In addition, stud- 2001; Allen et al. 2003; Castrillo et al. 2003).
ies on sub-cellular localization of yeast proteins 5. At the bioinformatics level, new machine-
on a proteome-wide scale, phosphoproteome learning methods for the analysis of transcrip-
studies, protein–protein interaction maps and tome, proteome and metabolome data, and
studies on protein turnover have all been un- for the study of yeast regulatory networks
dertaken (Ficarro et al. 2002; Gavin et al. 2002; have been derived. Relevant web resources,
Ho et al. 2002; von Mering et al. 2002; Pratt databases and methods for the global analysis
et al. 2002; Ghaemmaghami et al. 2003; Huh of yeast genomic data, data repositories and
et al. 2003; Wohlschlegel and Yates 2003). data warehouses for storage and rapid access
4. At the metabolome level, new methods for the to yeast genomic raw data have all been gener-
analysis of yeast metabolites, strategies to as- ated. These include the Saccharomyces Genome
cribe function to unknown genes, and the clas- Database (SGD, http://www.yeastgenome.org),
sification of yeast mutants using metabolic fin- microarray databases and repositories (e.g.
gerprinting and footprinting have been devel- Stanford Microarray Database and Array-
Express, http://genome-www5.Stanford.edu; metabolomic studies starting later and receiving

http://www.ebi.ac.uk/arrayexpress), the Yeast less attention. This situation is rapidly being
Protein Database (YPD, http://proteome.incyte. corrected, and the latest studies and strategies are
com), the Proteomics Experiment Data Re- rapidly re-establishing the relevance of metabolites
pository (PEDRo, http://pedro.man.ac.uk), and metabolomics in integrative studies towards
metabolic pathways and metabolic databases a system-level understanding of the cell (Teusink
(KEGG, http://www.genome.ad.jp/kegg/), glo- et al. 1998; Fiehn 2001; Raamsdonk et al. 2001; De
bal management systems of genomic infor- la Fuente et al. 2002; Adams 2003; Allen et al. 2003;
mation (GIMS, Cornell et al. 2003, http:// Harrigan and Goodacre 2003; Weckwerth 2003;
www.cs.man.ac.uk/img/gims/) and gene on- Goodacre et al. 2004).
tology tools (e.g. GoMiner, MAPPFinder and Metabolomics can be defined as the compre-
GeneMAPP, http://www.geneontology.org/GO. hensive analysis of the complete pool of cellular
tools.html). metabolites (the ‘metabolome’) closely interacting
6. Integrated functional analysis strategies incor- with the other functional genomic levels (Fig. 1.1).
porate growth competition experiments and The relevance of metabolomics in the post-
high-throughput methods to quantify the re- genomic era can be exemplified by (1) the scope
sults of the growth competition, for the elu- and importance of metabolic network studies,
cidation of gene function (Baganz et al. 1998; and (2) the rediscovered role of metabolites in
Giaever et al. 2002; Merritt and Edwards 2004). regulation at the different genomic levels.
The central metabolic pathways are the
In summary, the favourable characteristics of S.
biological networks which have been most
cerevisiae, together with the advanced functional
subject to comprehensive analyses. The basic
genomics resources, strategies and methods
information is presently accessible in metabolic
already available, make it an optimal reference
pathways databases (e.g. BioCyc, http://biocyc.org;
organism for studies of integrative systems biology
KEGG, http://www.genome.ad.jp/kegg/; RIKEN,
at the cellular level. In these studies, a number of
http://genome.gsc.riken.go.jp/DNA-Book/metabo-
important factors such as the role of metabolomics
lome.shtml; Metabolic Vision metabolic pathway
and new regulatory mechanisms should not be
database, http://www.ariadnegenomics.com/pro-
overlooked. In order to pursue comprehensive
ducts/meta.html).
holistic approaches to unveil the dynamics of
These metabolic networks have been the first
gene/protein networks and to integrate this
for which conceptual framework theories have
information into progressively more realistic
been defined. Some of these have focused on the
genome-wide models, particular care has to be
description of the networks as flux maps such
taken to incorporate information and mechanisms
as Metabolic Flux Analysis (MFA; Varma and
at all of these different functional levels.
Palsson 1994). In this approach, the dynamic
system is turned into a steady-state model in
which no kinetic information is needed. MFA
IV. Metabolomics in Comprehensive
has the objective of, for example, quantifying all
Post-Genomic Studies: intracellular fluxes for exploitation in metabolic
Towards Systems Biology engineering strategies. This flux balance approach
Using S. cerevisiae as a Model has some limitations, however, since not all
reactions involving major co-factors, such as ATP,
A. Metabolomic Studies: Metabolic Networks NADH and NADPH, are exactly known. Thus, in
and Participation of Metabolites MFA, most recent studies have focused on the
in Regulation incorporation of new analytical strategies such as
isotopomer labelling (e.g. 13C labelling applied to
One of the most relevant breakthroughs in MFA; Wiechert 2001).
functional genomics is the implementation of As soon as better theoretical frameworks be-
progressively more refined high-throughput come available, new strategies can be applied. In
techniques for genome-wide studies of the cell. In spite of this, the continuing limited success of at-
the post-genomic era, primary efforts are being tempts at metabolic engineering has pointed to
invested mainly on analytical strategies at the the existence of overlooked concepts, and the need
genome, gene expression and proteome levels, with for truly comprehensive conceptual approaches in
order to understand the dynamics and control of Apart from these metabolic network stud-
metabolic networks. Among these conceptual ap- ies, a relevant aspect in metabolomics is the
proaches, the most significant is Metabolic Control participation of metabolites in regulation. In
Analysis (MCA) theory (Kacser and Burns 1973; the post-genomic era, considerable efforts are
Kacser 1995; Fell 1997). The three main princi- focused on the role of regulatory mechanisms
ples of MCA theory can be summarized as fol- (at the epigenetic, genetic, transcriptional and
lows. post-transcriptional levels) in the global behaviour
of the cell. With this perspective, metabolites could
1. In a metabolic pathway, control is shared – that be regarded as inert, with minor participation in
is, the control of flux through the pathway is regulation. However, the complete pool of internal
‘distributed’ between the different enzymatic and external cellular metabolites plays a remark-
steps and is not exerted by a single enzyme able role in regulation and control. Firstly, at the
catalysing the so-called rate-limiting step. level of intermediary metabolism, metabolites
2. The control exerted by each enzymatic step is such as fructose-1,6-diphosphate, ATP, ADP and
measured by the ‘flux control coefficient’, CeJ , citrate exert rapid short-term regulation upon cen-
which can be defined as the relative change in tral metabolic fluxes (e.g. by the rapid activation
flux (J) caused by a specific modulation of the or inhibition of enzymes by reversible covalent
activity of an enzyme (e) at steady state. modification or by allosteric effects; Monod et al.
3. For an unbranched pathway, the Summation 1963; Fell 1997; Plaxton 2004). Secondly, the nature
Theorem states that the sum of the flux control and levels of external metabolites (substrates,
coefficients of all steps is equal to unity. products and other external compounds) consti-
tute essential environmental signals detected by
The fundamental difference between the rate-lim- the cell, usually via ligand-membrane receptor
iting step concept and the MCA theory is the no- interactions. These signals are transduced into
tion of distributed control in which each enzy- the cell via specific signal transduction pathways
matic step contributes to the global control of the with the major participation of regulatory proteins
metabolic flux. From this perspective, attempts to (e.g. transcription factors; Hancock 1997; Sprague
increase the flux through a central metabolic path- et al. 2004). However, an increasing number of
way will require the manipulation of several en- studies are reporting an important role for inter-
zymatic steps in order to achieve a measurable nal metabolites (e.g. phosphate, cAMP, inositol
effect on the global flux. Central metabolic path- phosphates, phosphatidic acid, amino acids)
ways are tightly regulated and resist manipula- as regulatory molecules participating in signal
tion, whereas strategies involving the manipula- transduction and complex regulatory mechanisms
tion of peripheral metabolic pathways have been (Zaragoza et al. 1999; Hansen and Johannesen
reported to be more successful (Brown 1997; Fell 2000; Muller et al. 2003; Sellick and Reece 2003;
1998; Stephanopoulos 1999). In modern MCA the- Auesukaree et al. 2004; Loewen et al. 2004; Sprague
ory, other aspects (such as metabolic compartmen- et al. 2004). Moreover, recent studies have reported
tation) are also being considered, and new advances new mechanisms by which metabolites may con-
and conceptual principles are being incorporated trol gene expression (e.g. by direct interaction with
(Peletier et al. 2003). For a good review on the ba- mRNA-riboswitches, without the participation of
sic principles of metabolic control analysis theory, proteins), or which can lead to post-translational
the reader can refer to Fell (1997, 1998) and refer- histone modifications (Cech 2004; Dong and Xu
ences therein. Additional strategies for modelling 2004).
metabolic networks and approaches such as top- The advanced studies on metabolic networks
down analysis are being investigated and incor- and the sometimes overlooked roles of metabolites
porated into the different levels of analysis (Brown in regulation constitute new challenges. More
et al. 1990; Quant 1993; Krauss and Quant 1996). Fi- importantly, they confirm the importance of
nally, among the latest studies on biochemical path- metabolomics, together with other genome-wide
ways are network constraint-based models which high-throughput strategies (e.g. genome, tran-
rely on the definition of minimal functional units scriptome and proteome studies) in the generation
of metabolism (elementary flux modes or extreme of a complete, integrative description of the cell as
pathways; Klamt and Stelling 2003; Papin et al. a global system (Fig. 1.1), which is a prerequisite
2004). for comprehensive systems biology studies.
Fig. 1.2. Systems biology. Integration of global experimental (A) and theoretical studies (B) in the iterative cycle of
knowledge (Kell and Oliver 2004)
B. Integrative Studies for Global Description many of the circuits in the cellular networks are still
of Cellular Complexity: Cellular Networks to be elucidated. These cellular networks, their un-
and Systems Biology derlying mechanisms, interrelationships and flex-
ibility in adapting to environmental changes con-
Genes, RNAs, proteins and metabolites exert their stitute the main central processes controlling the
role in a ‘cell system’ rich in complexity, with the global behaviour of the cell as a ‘biological sys-
participation of mechanisms of regulation at the tem’. Their elucidation constitutes one of the most
genomic, transcriptional, post-transcriptional and daunting challenges in post-genomic biology. As
metabolic level (Fig. 1.1; Day and Tuite 1998; Cas- a result, comprehensive integrative studies are in-
trillo and Oliver 2004; Choudhuri 2004). This fact creasingly being undertaken to advance the under-
has been demonstrated by different authors in sev- standing of the logics of the responsible regulatory
eral studies in which no (simple) direct correlation modules (Fell 2001; Oliver et al. 2002; Phelps et al.
between levels of gene expression (mRNA), pro- 2002; Nurse 2003; Castrillo and Oliver 2004; Stelling
tein content and/or metabolic fluxes could be es- 2004).
tablished (Gygi et al. 1999; Fell 2001; Ideker et al. Systems biology aims at understanding all of
2001b; ter Kuile and Westerhoff 2001; Yoon and the genotype–phenotype relationships brought
Lee 2002; Bro et al. 2003; Glanemann et al. 2003; about by these cellular networks, as well as
Lee et al. 2003; Mehra et al. 2003; Yoon et al. 2003; the principles and mechanisms governing the
Daran-Lapujade et al. 2004). All these results point behaviour of biological systems (Kitano 2002;
to the existence of specific mechanisms of regula- Nurse 2003; Stelling 2004). The main objective
tion at the different ’omic levels, and suggest that of systems biology is the construction of math-
ematical models by which to interrogate and as a touchstone model in the post-genomic era
iteratively refine our knowledge of the system (Castrillo and Oliver 2004) are positioning it at the
(e.g. the eukaryotic cell; Kitano 2002; Nurse 2003; leading edge of advanced systems biology studies.
Ideker 2004a; Kafatos and Eisner 2004; Stelling
2004). Systems biology addresses this task by C. S. cerevisiae as a Reference Model
combining integrative experimental and theo- in Systems Biology: Advanced Studies,
retical approaches (Stelling 2004). This outlook Biological Networks and Genome-Scale
is illustrated in Fig. 1.2, in the context of the Models, and Applications
iterative cycle of knowledge (Kell and Oliver 2004).
First, high-throughput integrative experimental Systems biology entails analysis and comparison of
studies at several functional genomic levels (e.g. results from specifically designed integrative exper-
gene expression together with proteome and/or iments and global theoretical studies (Fig. 1.2). This
metabolome studies; Gygi et al. 1999; Bro et al. goal is still difficult to achieve. This is because the
2003; Lee et al. 2003; Urbanczyk-Wochniak et al. ability to combine two or more functional genomic
2003) provide system-level sets of data (Fig. 1.2A). strategies (e.g. microarrays and metabolome stud-
From this point, comprehensive theoretical studies ies) in a single experiment in conjunction with
(bioinformatics, ‘in silico’ methods), integrative comprehensive ‘in silico’ approaches is limited to
data analysis and/or modelling approaches (e.g. relatively few research groups. As a result, only
Kell and King 2000; Mendes 2002; Yao 2002; a few true systems biology studies, combining ex-
Stelling 2004 and references therein) allow the perimental work with a strong theoretical basis,
generation of new knowledge at a system level have been published. These have concentrated on
and make predictions capable of being tested in well-defined unicellular organisms (e.g. bacteria,
further experimental studies (Fig. 1.2B). Hence, yeast) and sub-cellular systems (Ideker et al. 2001b;
systems biology operates in two main ways: (1) by Covert et al. 2004; Ozbudak et al. 2004).
deduction, by perturbing the biological system in
specifically designed experiments, monitoring the
responses, integrating the data and formulating
mathematical models able to reproduce this be-
haviour; and (2) in an inductive way, by perturbing
the model, making predictions as to what should
happen in response to these perturbations and
testing them through laboratory experiments
(Ideker et al. 2001a; Sage 2004).
The ultimate challenge in systems biology
would be to integrate all genomic information for
a biological system (e.g. mammalian cell line or
tissue) under different conditions, to unveil all
internal regulatory networks responsible for its be-
haviour, and to construct a global dynamic model
which reliably reproduces this behaviour. However,
the limitations of existing high-throughput tech-
niques, the difficulty of obtaining all information at
the different ’omic levels, the intrinsic complexity Fig. 1.3. Principal components analysis (PCA) of electro-
of the biological systems (with many mechanistic spray mass spectrometry (ES-MS) analysis of external
uncertainties), and the predictable limitations at metabolites corresponding to S. cerevisiae strains
growing in exponential phase (OD600 =1. 0). Refer-
the computational level are limiting the scope ence: external medium (R). Parental strain: BY4709
of a functional genomics approach to systems (ATCC 200872) (P). Transactivator mutants: (1) carbon
biology, primarily to studies of single-celled metabolism: BY4709cat8-[Delta] () and BY4709hap4-
[Delta] (); (2) nitrogen assimilation: BY4709nil1-[Delta]
organisms and sub-cellular systems (Stelling 2004
and references therein). In this context, simple ◦ •
( ), BY4709gln3-[Delta] ( ), BY4709arg82-[Delta]
() and BY4709nil1-[Delta], gln3-[Delta] ( ); (3) lipid
reference organisms will constitute the primary metabolism: BY4709opi1-[Delta] (), BY4709ino2-[Delta]
subjects of comprehensive analyses and, at this (), BY4709ino4-[Delta] () and BY4709ino2-[Delta],
point, the optimal characteristics of S. cerevisiae ino4-[Delta] ( )
Reductions in both the costs and operational of transcription factors, and they illustrate the
difficulty of functional genomic technologies, contribution that metabolomics can make to
together with an increasing trend to perform such functional genomic studies.
research in a collaborative manner, is advancing In the field of theoretical systems biology, re-
the systems biology agenda. Thus, systems biology cent years have witnessed the appearance of ad-
is a rapidly emerging field in which novel ‘in vanced bioinformatics strategies, integrative data
silico’ approaches are being used to implement analysis, and network studies towards the imple-
network- and genome-scale modelling strategies mentation of genome-scale models, with S. cere-
which integrate data from the different ’omic visiae (in the majority of cases) being the selected
levels. However, the afore-mentioned difficulties model. Thus, in the development of comprehen-
mean that the majority of these in silico studies sive data analysis tools, efforts have been directed
rely on data extracted from the literature and mainly towards the development of new classifica-
public genomic databases. A word of caution is tion algorithms, clustering, and machine learning
therefore appropriate – particular care should methods (unsupervised or supervised) to unveil
be taken, since the validity of these studies will characteristic patterns and their interrelationships
be critically dependent on the selection of pri- (Kell and King 2000; McInerney 2002; Mendes 2002;
mary datasets (often from different laboratories) Kapetanovic et al. 2004; Wei et al. 2004). The natural
which have been obtained under essentially expansion of these studies is leading to the advent
equivalent conditions (von Mering et al. 2002; of a new conceptual framework in post-genomics
Castrillo and Oliver 2004). This emphasises science – network biology – dedicated to the study
the need for data standards (e.g. MIAME, http:// of the topology, modular components, and the gov-
www.mged.org/Workgroups/MIAME/miame.html; erning principles of cellular networks (Barabási
PEDRo, http://pedro.man.ac.uk; ArtMet; http:// and Oltvai 2004; Papin and Palsson 2004; Stelling
www.armet.org) to ensure there is sufficient 2004). Such work has concentrated on metabolic,
metadata (i.e. well-defined formal descriptions protein–protein interaction, signalling, and regula-
of the experimental conditions and analytical tory networks, with (again) S. cerevisiae as the ref-
procedures) to allow such judgements to be made. erence model (Harbison et al. 2004; Ideker 2004b;
As far as the implementation of new genome- Ihmels et al. 2004; Yeang et al. 2004). For example,
wide techniques is concerned, the vast majority of studies on the dynamics of the transcriptional reg-
the most advanced high-throughput genomic tech- ulatory network of S. cerevisiae have been reported
nologies are being validated in S. cerevisiae, at the recently (Luscombe et al. 2004). These reveal large
genome, transcriptome, proteome, metabolome topological changes depending on environmental
and interaction levels. This makes yeast the most conditions, with transcription factors altering their
advanced biological system for the application interactions in response to stimuli, a few of them
of global experimental strategies, combining serving as permanent hubs but most acting tran-
studies at different ’omic levels (Raamsdonk siently in certain conditions only.
et al. 2001; Gavin et al. 2002; Giaever et al. 2002; The final objective in systems biology is to
Hayes et al. 2002; Ho et al. 2002; Allen et al. 2003; incorporate the experimental data into mathe-
Zhu et al. 2003; Scherens and Goffeau 2004). As matical models which are descriptive of the cell
an example of this, the principal components system, and which are capable of predicting its
analysis of the metabolome profiles of selected S. behaviour. Genome-scale models of metabolic
cerevisiae transactivator mutants, corresponding or other networks may be made at different
to different nutrient assimilation pathways, is levels of granularity, depending on the qual-
presented in Fig. 1.3. This metabolome analysis, itative and quantitative characteristics of the
combined with the use of selected mutants, analytical technologies employed. The most
shows that metabolomic studies are capable of relevant approaches to the mathematical mod-
discriminating between different transcription elling of cellular networks (Stelling 2004) are as
factor mutants. Moreover, when a number of follows.
transactivator mutants tended to cluster together,
these corresponded to mutants of one and the 1. Interaction-based: these result in static mod-
same nutrient assimilation pathway. These results els for the study of individual networks (e.g.
point to the possibility of using metabolomics to gene expression; protein–protein interaction
investigate new transcription factors, or families networks).
2. Constraint-based: network representations while the recently created Yeast System Biology
based on constraints include the stoichiometry Network (http://www.ysbn.org) will promote
and reversibility of metabolic reactions, and collaborations between yeast systems biologists.
allow the calculation of fluxes and steady-state If such joint initiatives prosper, and succeed in
flux distributions. leading the yeast research community in the right
3. Mechanistic representations: these complex direction, we can anticipate a bright future for S.
dynamic models include stoichiometry and cerevisiae as a reference model in systems biology.
kinetic parameters but are more representative
of the real system, and allow us to formulate
precise, experimentally testable hypotheses. V. Conclusions
Among the most remarkable examples of recon-
New advanced studies in the post-genomic era are
structed models reported are an expanded model
unveiling the real picture of the cell as a system
of Escherichia coli K-12 (Reed et al. 2003), the
rich in complexity, with participation of mecha-
first genome-scale reconstructions and models
nisms at the epigenomic, genomic, transcriptomic,
of S. cerevisiae (Förster et al. 2003; US Patent
proteomic and metabolomic levels in network in-
2003228567, US Patent Office 2003), and the con-
teractions which require extensive and integrative
struction and validation of a compartmentalized
investigation. These mechanisms configure the in-
genome-scale model of S. cerevisiae (Duarte et al.
ternal cellular networks responsible for the global
2004).
behaviour of the cell, which include complex reg-
Given its favourable characteristics as a touch-
ulatory circuits and signal transduction pathways
stone model in post-genomic studies (Castrillo and
allowing the cell to respond in an appropriate man-
Oliver 2004) and the success of the network studies
ner to changes in its environment. Many of these
reported above, it is clear that S. cerevisiae is
regulatory mechanisms remain to be elucidated,
emerging as the best model eukaryote for systems
and we urge that the role of the metabolome should
biology studies at the single-cell level. Progressive
not be overlooked in studies of regulatory networks
advances in high-throughput methods, improved
in the post-genomics era.
experimental strategies and more refined models
In order to reach the goal of understanding the
leading to the elucidation of the components
eukaryotic cell in a holistic and integrated manner,
and patterns of regulatory networks are sure to
the utilization of good biological models, which are
result in direct applications in disciplines such
well characterized at the genetic and all functional
as metabolic engineering (e.g. optimisation of
genomics levels, and which can serve as a reference
bioconversion pathways; Stephanopoulos and Gill
platform towards the study of more complex sys-
2001; Gill and Dodge 2004; WO Patent 0107567,
tems (e.g. mammalian cells), is essential. We submit
World Intellectual Property Organisation 2001a),
that the biological characteristics of S. cerevisiae,
biomedical and environmental sciences, and the
together with its advanced position at the forefront
pharmaceutical industry (e.g. biomarkers discov-
of post-genomics studies, will make it the reference
ery, new tools for the identification of targets,
model for systems biology studies of the eukaryotic
and development of new drugs and therapeutic
cell.
strategies; Fernie et al. 2004; Ilyin et al. 2004; Wang
et al. 2004; WO Patent 0178652, World Intellectual Acknowledgements. This work was supported by an
Property Organisation 2001b). In order to realise EC contract to S.G. Oliver within the frame of the
the full potential of yeast as a reference model in Garnish network of FP5 and the BBSRC’s Investigating
Gene Function Initiative within COGEME (Consortium
systems biology, the high level of international for the Functional Genomics of Microbial Eukaryotes;
collaborations between yeast research groups seen http://www.cogeme.man.ac.uk). We thank Dr. M. Bolotin-
in the genome sequencing (Goffeau et al. 1996, Fukuhara (IGM, Université Paris XI, Orsay Cedex, France)
1997; http://www.yeastgenome.org) and functional for providing yeast strains.
genomics projects (Oliver 1997, 2002; Winzeler
et al. 1999; Giaever et al. 2002) must be continued
and even enhanced. To this end, SysBiolDB has References
been established to facilitate the deposition and
exchange of information by systems biology re- Adams A (2003) Metabolomics: small-molecule omics. The
searchers worldwide (http://www.sysbioldb.org), Scientist 17:38–40
Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P nitrogen, or phosphate limitations. Enzyme Microb
(2002) Molecular biology of the cell, 4th edn. Garland Technol 16:761–770
Science, Taylor and Francis Group, New York Costanzo MC, Crawford ME, Hirschman JE, Kranz JE,
Allen J, Davey HM, Broadhurst D, Heald JK, Rowland JJ, Olsen P, Robertson LS, Skrzypek MS, Braun BR,
Oliver SG, Kell DB (2003) High-throughput classifica- Hopkins KL, Kondu P et al. (2001) YPD, PombePD
tion of yeast mutants using metabolic footprinting. Nat and WormPD: model organism volumes of the
Biotechnol 21:692–696 BioKnowledge library, an integrated resource for
Auesukaree C, Homma T, Tochio H, Shirakawa M, protein information. Nucleic Acids Res 29:75–79
Kaneko Y, Harashima S (2004) Intracellular phosphate Covert MW, Knight EM, Reed JL, Herrgard MJ, Palsson BØ
serves as a signal for the regulation of the PHO (2004) Integrating high-throughput and computa-
pathway in Saccharomyces cerevisiae. J Biol Chem tional data elucidates bacterial networks. Nature
279:17289–17294 429:92–96
Baganz F, Hayes A, Farquhar R, Butler PR, Gardner DCJ, Csank C, Costanzo MC, Hirschman J, Hodges P, Kranz JE,
Oliver SG (1998) Quantitative analysis of yeast gene Mangan M, O’Neill K, Robertson LS, Skrzypek MS,
function using competition experiments in continuous Brooks J et al. (2002) Three yeast proteome databases:
culture. Yeast 14:1417–1427 YPD, PombePD, and CalPD (MycoPathPD). Methods
Bailey JE (1991) Toward a science of metabolic engineering. Enzymol 350:347–373
Science 252:1668–1675 Daran-Lapujade P, Jansen ML, Daran JM, van Gulik W,
Bailey JE, Ollis DF (1986) Biochemical engineering funda- de Winde JH, Pronk JT (2004) Role of transcriptional
mentals, 2nd edn. McGraw Hill, New York regulation in controlling fluxes in central carbon
Barabási AL, Oltvai ZN (2004) Network biology: under- metabolism of Saccharomyces cerevisiae. A chemostat
standing the cell’s functional organization. Nat Rev culture study. J Biol Chem 279:9125–9138
Genet 5:101–113 Day DA, Tuite MF (1998) Post-transcriptional gene regu-
Boer VM, de Winde JH, Pronk JT, Piper MDW (2003) latory mechanisms in eukaryotes: an overview. J En-
The genome-wide transcriptional responses of docrinol 157:361–371
Saccharomyces cerevisiae grown on glucose in aerobic De la Fuente A, Snoep JL, Westerhoff HV, Mendes P (2002)
chemostat cultures limited for carbon, nitrogen, Metabolic control in integrated biochemical systems.
phosphorus, or sulfur. J Biol Chem 278:3265–3274 Eur J Biochem 269:4399–4408
Bro C, Regenberg B, Lagniel G, Labarre J, Montero- Delneri D, Brancia FL, Oliver SG (2001) Towards a truly
Lomeli M, Nielsen J (2003) Transcriptional, proteomic, integrative biology through the functional genomics
and metabolic responses to lithium in galactose-grown of yeast. Curr Opin Biotechnol 12:87–91
yeast cells. J Biol Chem 278:32141–323149 Dong L, Xu CW (2004) Carbohydrates induce mono-
Brown AJP (1997) Control of metabolic flux in yeasts and ubiquitination of H2B in yeast. J Biol Chem
fungi. Trends Biotechnol 15:445–447 279:1577–1580
Brown AJP, Tuite MF (1998) Yeast gene analysis. Academic Duarte NC, Herrgard MJ, Palsson BØ (2004) Reconstruction
Press, San Diego, Methods in Microbiology 26 and validation of Saccharomyces cerevisiae iND750,
Brown GC, Hafner RP, Brand MD (1990) A ‘top-down’ ap- a fully compartmentalized genome-scale metabolic
proach to the determination of control coefficients in model. Genome Res 14:1298–1309
metabolic control theory. Eur J Biochem 188:321–325 Eisenstein M (2004) Getting down to the bare essentials. Nat
Burke D, Dawson D, Stearns T (2000) Methods in yeast Methods 20 July 2004. DOI 101038/nmteh030
genetics, 2000 edn. A Cold Spring Harbor Laboratory Fafournoux P, Bruhat A, Jousse C (2000) Amino acid regu-
Course Manual. Cold Spring Harbor Laboratory Press, lation of gene expression. Biochem J 351:1–12
New York Fell DA (1997) Understanding the control of metabolism.
Castrillo JI, Oliver SG (2004) Yeast as a touchstone in post- Portland Press, London
genomic research. Strategies for integrative analysis in Fell DA (1998) Increasing the flux in metabolic pathways:
functional genomics. J Biochem Mol Biol 37:93–106 a metabolic control analysis perspective. Biotechnol
Castrillo JI, Ugalde UO (1994) A general model of yeast Bioeng 58:121–124
energy metabolism in aerobic chemostat culture. Yeast Fell DA (2001) Beyond genomics. Trends Genet 17:680–682
10:185–197 Fernie AR, Trethewey RN, Krotzky AJ, Willmitzer L (2004)
Castrillo JI, Hayes A, Mohammed S, Gaskell SJ, Oliver SG Metabolite profiling: from diagnostics to systems biol-
(2003) An optimised protocol for metabolome analogy. Nat Rev Mol Cell Biol 5:763–769
ysis in yeast using direct infusion electrospray mass Ficarro SB, McCleland ML, Stukenberg PT, Burke DJ,
spectrometry. Phytochemistry 62:929–937 Ross MM, Shabanowitz J, Hunt DF, White FM (2002)
Cech TR (2004) RNA finds a simpler way. Nature 428:263– Phosphoproteome analysis by mass spectrometry
264 and its application to Saccharomyces cerevisiae. Nat
Choudhuri S (2004) The nature of gene regulation. Int Arch Biotechnol 20:301–305
Biosci 2004:1001–1015 Fiehn O (2001) Combining genomics, metabolome analysis
Cornell M, Paton NW, Hedeler C, Kirby P, Delneri D, and biochemical modelling to understand metabolic
Hayes A, Oliver SG (2003) GIMS: an integrated data networks. Comp Funct Genomics 2:155–168
storage and analysis environment for genomic and Förster J, Famili I, Fu P, Palsson BØ, Nielsen J (2003)
functional data. Yeast 20:1291–1306 Genome-scale reconstruction of the Saccharomyces
Cortassa S, Aon MA (1994) Metabolic control analysis of gly- cerevisiae metabolic network. Genome Res 13:244–253
colysis and branching to ethanol production in chemo- Gaisne M, Bécam AM, Verdière J, Herbert CJ (1999) A ‘nat-
stat cultures of Saccharomyces cerevisiae under carbon, ural’ mutation in Saccharomyces cerevisiae strains de-
rived from S288c affects the complex regulatory gene Gygi SP, Rochon Y, Franza BR, Aebersold R (1999) Correla-
HAP1 (CYP1). Curr Genet 36:195–200 tion between protein and mRNA abundance in yeast.
Garcia-Martinez J, Aranda A, Perez-Ortin JE (2004) Ge- Mol Cell Biol 19:1720–1730
nomic run-on evaluates transcription rates for all yeast Hancock JT (1997) Cell signalling. Prentice Hall, Harlow
genes and identifies gene regulatory mechanisms. Mol Hansen J, Johannesen PF (2000) Cysteine is essential for
Cell 15:303–313 transcriptional regulation of the sulfur assimilation
Gavin AC, Superti-Furga G (2003) Protein complexes and genes in Saccharomyces cerevisiae. Mol Gen Genet
proteome organization from yeast to man. Curr Opin 263:535–542
Chem Biol 7:21–27 Harbison CT, Gordon DB, Lee TI, Rinaldi NJ, Macisaac KD,
Gavin AC, Bosche M, Krause R, Grandi P, Marzioch M, Danford TW, Hannett NM, Tagne JB, Reynolds DB,
Bauer A, Schultz J, Rick JM, Michon AM, Cruciat CM Yoo J et al. (2004) Transcriptional regulatory code of
et al. (2002) Functional organization of the yeast pro- a eukaryotic genome. Nature 431:99–104
teome by systematic analysis of protein complexes. Na- Harrigan GG, Goodacre R (2003) Metabolic profiling: its
ture 415:141–147 role in biomarker discovery and gene function analy-
Ghaemmaghami S, Huh WK, Bower K, Howson RW, sis. Kluwer, Boston
Belle A, Dephoure N, O’Shea EK, Weissman JS (2003) Hayes A, Zhang N, Wu J, Butler PR, Hauser NC, Hoheisel JD,
Global analysis of protein expression in yeast. Nature Lim F, Sharrocks AD, Oliver SG (2002) Hybridization
425:737–741 array technology coupled with chemostat culture: tools
Giaever G, Chu AM, Ni L, Connelly C, Riles L, Veronneau S, to interrogate gene expression in Saccharomyces cere-
Dow S, Lucau-Danila A, Anderson K, Andre B et al. visiae. Methods 26:281–290
(2002) Functional profiling of the Saccharomyces cere- Hirayoshi K, Lis JT (1999) Nuclear run-on assays: assessing
visiae genome. Nature 418:387–391 transcription by measuring density of engaged RNA
Gill RT, Dodge T (2004) Special issue on inverse metabolic polymerases. Methods Enzymol 304:351–362
engineering. Metab Eng 6:175–176 Ho Y, Gruhler A, Heilbut A, Bader GD, Moore L, Adams SL,
Giuseppin ML, van Riel NA (2000) Metabolic modelling of Millar A, Taylor P, Bennett K, Boutilier K et al. (2002)
Saccharomyces cerevisiae using the optimal control of Systematic identification of protein complexes in Sac-
homeostasis: a cybernetic model definition. Metab Eng charomyces cerevisiae by mass spectrometry. Nature
2:14–33 415:180–183
Glanemann C, Loos A, Gorret N, Willis LB, O’Brien XM, Huh WK, Falvo JV, Gerke LC, Carroll AS, Howson RW, Weiss-
Lessard PA, Sinskey AJ (2003) Disparity between man JS, O’Shea EK (2003) Global analysis of protein
changes in mRNA abundance and enzyme activity localization in budding yeast. Nature 425:686–691
in Corynebacterium glutamicum and implications for Ideker T (2004a) Systems biology 101 – what you need to
DNA microarray analysis. Appl Microbiol Biotechnol know. Nat Biotechnol 22:473–475
61:61–68 Ideker T (2004b) A systems approach to discovering sig-
Goffeau A (2000) Four years of post-genomic life with 6000 nalling and regulatory pathways – or, how to digest
yeast genes. FEBS Lett 480:37–41 large interaction networks into relevant pieces. Adv
Goffeau A, Barrell BG, Bussey H, Davis RW, Dujon B, Feld- Exp Med Biol 547:21–30
mann H, Galibert F, Hoheisel JD, Jacq C, Johnston M Ideker T, Galitski T, Hood L (2001a) A new approach to
et al. (1996) Life with 6000 genes. Science 274:546– decoding life: systems biology. Annu Rev Genomics
567 Hum Genet 2:343–372
Goffeau A, Aert R, Agostini-Carbone ML, Ahmed A, Ideker T, Thorsson V, Ranish JA, Christmas R, Buhler J,
Aigle M, Alberghina L, Albermann K, Albers M, Eng JK, Bumgarner R, Goodlett DR, Aebersold R,
Aldea M, Alexandraki D et al (1997) The yeast Hood L (2001b) Integrated genomic and proteomic
genome directory. Nature 387 Suppl no 6632 (http:// analyses of a systematically perturbed metabolic
www.nature.com/genomics/papers/s_cerevisiae.html) network. Science 292:929–934
Gonzalez B, François J, Renaud M (1997) A rapid and reliable Ihmels J, Levy R, Barkai N (2004) Principles of transcrip-
method for metabolite extraction in yeast using boiling tional control in the metabolic network of Saccha-
buffered ethanol. Yeast 13:1347–1355 romyces cerevisiae. Nat Biotechnol 22:86–92
Goodacre R, Vaidyanathan S, Dunn WB, Harrigan GG, Ilyin SE, Belkowski SM, Plata-Salaman CR (2004) Biomarker
Kell DB (2004) Metabolomics by numbers: acquiring discovery and validation: technologies and integrative
and understanding global metabolite data. Trends approaches. Trends Biotechnol 22:411–416
Biotechnol 22:245–252 Iyer V, Struhl K (1996) Absolute mRNA levels and transcrip-
Guthrie C, Fink GR (2002a) Guide to yeast genetics and tional initiation rates in Saccharomyces cerevisiae. Proc
molecular and cell biology. Part B. Academic Press, Natl Acad Sci USA 93:5208–5212
Elsevier Science, San Diego, Methods in Enzymology Kacser H (1995) Recent developments beyond metabolic
vol 350 control analysis. Biochem Soc Trans 23:387–391
Guthrie C, Fink GR (2002b) Guide to yeast genetics and Kacser H, Burns JA (1973) The control of flux. Symp Soc
molecular and cell biology. Part C. Academic Press, Exp Biol 27:65–104
Elsevier Science, San Diego, Methods in Enzymology Kafatos FC, Eisner T (2004) Unification in the century of
vol 351 biology. Science 303:1257
Guthrie C, Fink GR (2004) Guide to yeast genetics and Kapetanovic IM, Rosenfeld S, Izmirlian G (2004) Overview
molecular biology. Part A. Academic Press, Elsevier of commonly used bioinformatics methods and their
Science, San Diego, Methods in Enzymology vol 194 applications. Ann N Y Acad Sci 1020:10–21
Kell DB, King RD (2000) On the optimization of classes Muratani M, Tansey WP (2003) How the ubiquitin-
for the assignment of unidentified reading frames in proteasome system controls transcription. Nat Rev
functional genomics programmes: the need for ma- Mol Cell Biol 4:192–201
chine learning. Trends Biotechnol 18:93–98 Nurse P (2003) Systems biology: understanding cells. Nature
Kell DB, Oliver SG (2004) Here is the evidence, now what 424:883
is the hypothesis? The complementary roles of induc- Oliver SG (1997) Yeast as a navigational aid in genome anal-
tive and hypothesis-driven science in the post-genomic ysis. Microbiology 143:1483–1487
era. Bioessays 26:99–105 Oliver SG (2002) Functional genomics: lessons from yeast.
Kitano H (2002) Systems biology: a brief overview. Science Philos Trans R Soc B 357:17–23
295:1662–1664 Oliver SG, Winson MK, Kell DB, Baganz F (1998) System-
Klamt S, Stelling J (2003) Two approaches for metabolic atic functional analysis of the yeast genome. Trends
pathway analysis? Trends Biotechnol 21:64–69 Biotechnol 16:373–378
Krauss S, Quant PA (1996) Regulation and control in com- Oliver DJ, Nikolau B, Wurtele ES (2002) Functional
plex, dynamic metabolic systems: experimental appli- genomics: high-throughput mRNA, protein, and
cation of the top-down approaches of metabolic con- metabolite analyses. Metab Eng 4:98–106
trol analysis to fatty acid oxidation and ketogenesis. Olson OS, Nielsen J (2000) Metabolic engineering of Saccha-
J Theor Biol 182:381–388 romyces cerevisiae. Microbiol Mol Biol Rev 64:34–50
Lashkari DA, DeRisi JL, McCusker JH, Namath AF, Ozbudak EM, Thattai M, Lim HN, Shraiman BI, van Oude-
Gentile C, Hwang SY, Brown PO, Davis RW (1997) naarden A (2004) Multistability in the lactose utiliza-
Yeast microarrays for genome wide parallel genetic tion network of Escherichia coli. Nature 427:737–740
and gene expression analysis. Proc Natl Acad Sci USA Papin JA, Palsson BØ (2004) Topological analysis of mass-
94:13057–13062 balanced signaling networks: a framework to obtain
Lee PS, Shaw LB, Choe LH, Mehra A, Hatzimanikatis V, network properties including crosstalk. J Theor Biol
Lee KH (2003) Insights into the relation between 227:283–297
mRNA and protein expression patterns: II. Experi- Papin JA, Stelling J, Price ND, Klamt S, Schuster S, Pals-
mental observations in Escherichia coli. Biotechnol son BØ (2004) Comparison of network-based pathway
Bioeng 84:834–841 analysis methods. Trends Biotechnol 22:400–405
Lehninger AL (1975) Biochemistry, 2nd edn. Worth Pub- Pasteur L (1857) Mémoire sur la fermentation appelée lac-
lishers, New York tique. In: Mémoires Société Sciences Agriculture Arts
Lei F, Rotboll M, Jorgensen SB (2001) A biochemically struc- Lille, séance 3 août 1857, 2ème Série, vol V, pp 13–26
tured model for Saccharomyces cerevisiae. J Biotechnol Peletier MA, Westerhoff HV, Kholodenko BN (2003) Control
88:205–221 of spatially heterogeneous and time-varying cellular
Loewen CJ, Gaspar ML, Jesch SA, Delon C, Ktistakis NT, reaction networks: a new summation law. J Theor Biol
Henry SA, Levine TP (2004) Phospholipid metabolism 225:477–487
regulated by a transcription factor sensing phospha- Phelps TJ, Palumbo AV, Beliaev AS (2002) Metabolomics
tidic acid. Science 304:1644–1647 and microarrays for improved understanding of phe-
Luscombe NM, Babu MM, Yu H, Snyder M, Teichmann SA, notypic characteristics controlled by both genomics
Gerstein M (2004) Genomic analysis of regulatory net- and environmental constraints. Curr Opin Biotechnol
work dynamics reveals large topological changes. Na- 13:20–24
ture 431:308–312 Plaxton WC (2004) Principles of metabolic control. In:
McInerney JO (2002) Bioinformatics in a post-genomics Storey KB (ed) Functional metabolism of cells: control,
world – the need for an inclusive approach. Pharma- regulation, and adaptation. Wiley, New York, pp 1–23
cogenomics J 2:207–208 Pratt JM, Petty J, Riba-Garcia I, Robertson DHL, Gaskell SJ,
Mehra A, Lee KH, Hatzimanikatis V (2003) Insights into the Oliver SG, Beynon RJ (2002) Dynamics of protein
relation between mRNA and protein expression pat- turnover, a missing dimension in proteomics. Mol
terns. I. Theoretical considerations. Biotechnol Bioeng Cell Proteomics 1:579–591
84:822–833 Quant PA (1993) Experimental application of top-down con-
Mendes P (2002) Emerging bioinformatics for the trol analysis to metabolic systems. Trends Biochem Sci
metabolome. Brief Bioinformatics 3:134–145 18:26–30
Merritt J, Edwards JS (2004) Assaying gene function by Raamsdonk LM, Teusink B, Broadhurst D, Zhang N,
growth competition experiment. Metab Eng 6:212–219 Hayes A, Walsh MC, Berden JA, Brindle KM, Kell DB,
Mnaimneh S, Davierwala AP, Haynes J, Moffat J, Peng WT, Rowland JJ et al. (2001) A functional genomics strategy
Zhang W, Yang X, Pootoolal J, Chua G, Lopez A et al. that uses metabolome data to reveal the phenotype of
(2004) Exploration of essential gene functions via silent mutations. Nat Biotechnol 19:45–50
titratable promoter alleles. Cell 118:31–44 Reed JL, Vo TD, Schilling CH, Palsson BØ (2003) An ex-
Monod J, Changeux J-P, Jacob F (1963) Allosteric proteins panded genome-scale model of Escherichia coli K-12
and cellular control systems. J Mol Biol 6:306–329 (iJR904 GSM/GPR). Genome Biol 4:R54
Mortimer RK, Johnston JR (1986) Genealogy of princi- Rose AH, Harrison JS (1987–1995) The yeasts, vols 1–6.
pal strains of the yeast genetic stock center. Genetics Academic Press, London
113:35–43 Rose AH, Harrison JS (1993) The yeasts, vol 5. Academic
Muller D, Exler S, Aguilera-Vazquez L, Guerrero-Martin E, Press, London
Reuss M (2003) Cyclic AMP mediates the cell cycle dy- Ross-Macdonald P, Coelho PS, Roemer T, Agarwal S, Ku-
namics of energy metabolism in Saccharomyces cere- mar A, Jansen R, Cheung KH, Sheehan A, Symoni-
visiae. Yeast 20:351–367 atis D, Umansky L et al. (1999) Large-scale analysis
of the yeast genome by transposon tagging and gene Ton VK, Rao R (2004) Functional expression of heterolo-
disruption. Nature 402:413–418 gous proteins in yeast: insights into Ca2+ signaling and
Sage L (2004) Genome-scale model predicts gene regulation. Ca2+ -transporting ATPases. Am J Physiol Cell Physiol
The Scientist 18:42 287:580–589
Scherens B, Goffeau A (2004) The uses of genome- Urbanczyk-Wochniak E, Luedemann A, Kopka J, Selbig J,
wide yeast mutant collections. Genome Biol 5:229 Roessner-Tunali U, Willmitzer L, Fernie AR (2003) Par-
(http://genomebiologycom/2004/5/7/229) allel analysis of transcript and metabolic profiles: a new
Schwartz M (2001) The life and works of Louis Pasteur. approach in systems biology. EMBO Rep 4:989–993
J Appl Microbiol 91:597–601 US Patent Office (2003) US Patent 2003228567. Compo-
Sellick CA, Reece RJ (2003) Modulation of transcription sitions and methods for modelling Saccharomyces
factor function by an amino acid: activation of Put3p cerevisiae metabolism. US Patent Office, Alexandria,
by proline. EMBO J 22:5147–5153 VA
Sherman F (1998) An introduction to the genetics and Varma A, Palsson BØ (1994) Metabolic flux balancing: Basic
molecular biology of the yeast Saccharomyces cere- concepts, scientific and practical use. Bio/Technology
visiae (http://dbburmcrochester.edu/labs/sherman_f/ 12:994–998
yeast/) Verger A, Perdomo J, Crossley M (2003) Modification with
Sherman F (2002) Getting started with yeast. Methods En- SUMO. A role in transcriptional regulation. EMBO Rep
zymol 350:3–41 (updated in http://dbburmcrochester. 4:137–142
edu/labs/sherman_f/startedyeast.pdf) von Mering C, Krause R, Snel B, Cornell M, Oliver SG,
Sinclair CG, Cantero D (1990) Fermentation modelling. In: Fields S, Bork P (2002) Comparative assessment of
McNeil B, Harvey LM (eds) Fermentation. A practical large-scale data sets of protein–protein interactions.
approach. IRL Press, Oxford University Press, Oxford, Nature 417:399–403
pp 65–112 Wang S, Sim TB, Kim YS, Chang YT (2004) Tools for target
Spellman PT, Sherlock G, Zhang MQ, Iyer VR, Anders K, identification and validation. Curr Opin Chem Biol
Eisen MB, Brown PO, Botstein D, Futcher B (1998) 8:371–377
Comprehensive identification of cell cycle-regulated Weckwerth W (2003) Metabolomics in systems biology.
genes of the yeast Saccharomyces cerevisiae by microar- Annu Rev Plant Biol 54:669–689
ray hybridization. Mol Biol Cell 9:3273–3297 Wei GH, Liu DP, Liang CC (2004) Charting gene regula-
Sprague GF Jr, Cullen PJ, Goehring AS (2004) Yeast signal tory networks: strategies, challenges and perspectives.
transduction: regulation and interface with cell Biochem J 381:1–12
biology. In: Opresko LK, Gephart JM, Mann MB (eds) Wiechert W (2001) 13C metabolic flux analysis. Metab Eng
Advances in experimental medicine and biology. 3:195–206
Kluwer/Plenum, New York, Advances in Systems Winzeler EA, Shoemaker DD, Astromoff A, Liang H, An-
Biology vol 547, pp 91–105 derson K, Andre B, Bangham R, Benito R, Boeke JD,
Stelling J (2004) Mathematical models in microbial systems Bussey H et al. (1999) Functional characterization of
biology. Curr Opin Microbiol 7:513–518 the S. cerevisiae genome by gene deletion and parallel
Stephanopoulos G (1999) Metabolic fluxes and metabolic analysis. Science 285:901–906
engineering. Metab Eng 1:1–11 Wodicka L, Dong H, Mittmann M, Ho MH, Lockhart DJ
Stephanopoulos G, Gill RT (2001) After a decade of (1997) Genome-wide expression monitoring in Sac-
progress, an expanded role for metabolic engineering. charomyces cerevisiae. Nat Biotechnol 15:1359–1367
Adv Biochem Eng Biotechnol 73:1–8 Wohlschlegel JA, Yates JR (2003) Proteomics: where’s Waldo
Stephanopoulos G, Vallino JJ (1991) Network rigidity and in yeast? Nature 425:671–672
metabolic engineering in metabolite overproduction. World Intellectual Property Organisation (2001a) WO
Science 252:1675–1681 Patent 0107567. Engineering of metabolic control.
ter Kuile BH, Westerhoff HV (2001) Transcriptome meets World Intellectual Property Organisation, Geneva
metabolome: hierarchical and metabolic regulation of World Intellectual Property Organisation (2001b) WO
the glycolytic pathway. FEBS Lett 500:169–171 Patent 0178652. Methods for drug discovery, disease
ter Linde JJ, Liang H, Davis RW, Steensma HY, van Dijken JP, treatment and diagnosis using metabolomics. World
Pronk JT (1999) Genome-wide transcriptional analysis Intellectual Property Organisation, Geneva
of aerobic and anaerobic chemostat cultures of Saccha- Yao T (2002) Bioinformatics for the genomic sciences and
romyces cerevisiae. J Bacteriol 181:7409–7413 towards systems biology. Japanese activities in the
Teusink B, Baganz F, Westerhoff HV, Oliver SG (1998) post-genome era. Prog Biophys Mol Biol 80:23–42
Metabolic control analysis as a tool in the elucidation Yeang CH, Ideker T, Jaakkola T (2004) Physical network
of the function of novel genes. In: Brown AJ, Tuite MF models. J Comput Biol 11:243–262
(eds) Methods in Microbiology, vol 26. Academic Yoon SH, Lee SY (2002) Comparison of transcript levels
Press, London, pp 297–336 by DNA microarray and metabolic flux based on flux
The International Human Genome Mapping Consortium analysis for the production of poly-g-glutamic acid
(2001a) Initial sequencing and analysis of the human in recombinant Escherichia coli. Genome Informatics
genome. Nature 409:860–921 13:587–588
The International Human Genome Mapping Consortium Yoon SH, Han MJ, Lee SY, Jeong KJ, Yoo JS (2003) Combined
(2001b) A physical map of the human genome. Nature transcriptome and proteome analysis of Escherichia
409:934–941 coli during the high cell density culture. Biotechnol
Tilstone C (2003) Vital statistics. Nature 424:610–613 Bioeng 81:753–767
Zaragoza O, Lindley C, Gancedo JM (1999) Cyclic AMP Zhu H, Bilgin M, Bangham R, Hall D, Casamayor A,
can decrease expression of genes subject to catabo- Bertone P, Lan N, Jansen R, Bidlingmaier S, Houfek T
lite repression in Saccharomyces cerevisiae. J Bacteriol et al. (2001) Global analysis of protein activities using
181:2640–2642 proteome chips. Science 293:2101–2105
Zhu H, Snyder M (2003) Protein chip technology. Curr Opin Zhu H, Bilgin M, Snyder M (2003) Proteomics. Annu Rev
Chem Biol 7:55–63 Biochem 72:783–812
2 Genome Evolution in Hemiascomycete Yeasts
L.J. Montcalm1 , K.H. Wolfe1
CONTENTS I. Introduction
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 19
II. Colinearity of Gene Order . . . . . . . . . . . . . . . 20 The genome sequence of Saccharomyces cerevisiae
III. Genomic Evidence
for Whole-Genome Duplication . . . . . . . . . . 21
was determined by an international consortium
IV. Pairs of Centromeres . . . . . . . . . . . . . . . . . . . 23 of more than 600 scientists (Goffeau et al. 1997).
V. A Model for Genome Duplication . . . . . . . . . 24 This landmark achievement took 7 years (1989–
VI. Pseudogenes, Relics and Null Alleles . . . . . . . 26 1996) and resulted in the first publication of the
VII. Interspecies Differences in Gene Content . . . 28 genome sequence of a eukaryote. Technology has
A. Genes Gained . . . . . . . . . . . . . . . . . . . . . . 28
B. Genes Lost . . . . . . . . . . . . . . . . . . . . . . . . 29 advanced since then to the extent that, in 2003,
VIII. Rapidly Evolving Loci . . . . . . . . . . . . . . . . . . 30 the genomes of three other Saccharomyces species
IX. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 31 were sequenced in a few months and published
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 together in a paper with only five authors (Kellis
et al. 2003). This was quickly followed by the
1Department of Genetics, Smurfit Institute, University of Dublin, publications of genome sequences (either draft
Trinity College, Dublin 2, Ireland
Fig. 2.1. Approximate phylogenetic relationships among whole-genome duplication event (WGD), and the thickened
the hemiascomycetes whose genomes have been se- branches show the post-WGD species
quenced. The asterisk indicates the likely position of the
The Mycota XIII

Fungal Genomics
20 L.J. Montcalm and K.H. Wolfe
or complete) for several other yeasts, with the time periods in two independent groups of species:
result that today genome data are available a Saccharomyces group and a Candida group.
for 14 species in the class Hemiascomycetes
(Fig. 2.1).
The phylogenetic relationships among the se-
quenced genomes are not yet fully established. The II. Colinearity of Gene Order
topology shown in Fig. 2.1 is based on the work
of Kurtzman and Robnett (2003) who sequenced One of the most striking features of yeast genome
six genes from every species in the ‘Saccharomyces evolution is the degree to which the order of genes
complex’ of species, combined with data from Cai along chromosomes has been conserved between
et al.’s (1996) 18S rRNA tree for the more deeply species. For the Saccharomyces sensu stricto species
branching lineages, and Rokas et al.’s (2003) 106- (Fig. 2.1), colinearity with S. cerevisiae approaches
gene phylogeny for the Saccharomyces sensu stricto 100%. This conservation of gene order has proved
species. In particular, it is not certain whether the very useful in allowing us to differentiate real genes
three separate lineages leading to Ashbya gossypii, from spurious ORFs, and in facilitating the detec-
Kluyveromyces lactis and K. waltii/S. kluyveri as tion of rapidly evolving loci (Brachat et al. 2003;
shown in Fig. 2.1 really are distinct lineages. The Kellis et al. 2003; Zhang and Dietrich 2003).
key internal branches have low bootstrap values Among the sensu stricto species, there are only
in Kurtzman and Robnett’s (2003) analysis, and in a few places in the genome where colinearity is dis-
a recent analysis involving the same set of 106 genes rupted, and these disruptions fall into two types.
there was a strongly supported clade comprising S. First, there are a few differences in gene content be-
kluyveri, A. gossypii and K. waltii, with K. lactis out- tween sensu stricto species, where a gene is present
side (Hittinger et al. 2004). Comprehensive recon- in one species but either absent or a pseudogene
struction of the phylogenetic relationships among in another species (Bon et al. 2000; Fischer et al.
these species, using the sequences of all the or- 2001; Kellis et al. 2003). An example is the XYZ3
thologous genes from all the sequenced genomes, locus beside the centromere of S. cerevisiae chro-
remains to be done. mosome III, which is a pseudogene homologous
The lack of fossil records for hemiascomycetes to another gene (DOM34, a member of the pelota
makes it difficult to put an absolute timescale onto gene family) on chromosome XIV (Lalo et al. 1993).
the phylogenetic tree in Fig. 2.1. Molecular clock In the other sensu stricto species S. paradoxus and
estimates of the date for the divergence between S. bayanus, the ortholog of XYZ3 is an intact and
Candida albicans and S. cerevisiae have ranged be- apparently functional gene (Kellis et al. 2003; it
tween 140 and 330 million years (Berbee and Taylor is also a pseudogene in S. mikatae). Second, al-
1993; Pesole et al. 1995). It is perhaps more instruc- though all the sensu stricto species have 16 chromo-
tive to compare the average levels of protein se- somes, there have been several interchromosomal
quence divergence between S. cerevisiae and other translocations during the evolution of these species
yeast species to the divergence between human and (Fischer et al. 2000; Kellis et al. 2003), most of
other animals (Dujon et al. 2004). By this yardstick, which are reciprocal translocations, where parts of
C. glabrata is slightly more divergent from S. cere- two arms are reciprocally exchanged between two
visiae than the pufferfish Fugu rubripes is from chromosomes. In addition, many intrachromoso-
humans, and the most divergent species in Fig. 2.1, mal inversions have been mapped among the sensu
Yarrowia lipolytica, is about as divergent as the stricto species. All the endpoints of inversions coin-
sea squirt (tunicate) Ciona intestinalis. Thus, the cide with tRNA genes, and many of the reciprocal
available genome data from the hemiascomycetes translocations have endpoints in Ty elements or
span a complete range of levels of sequence diver- other repeated sequences, suggesting that homol-
gence. This will be further augmented in the years ogous recombination between these elements may
to come by the addition of more genome sequences have catalyzed the translocations.
from species closely related to some of the terminal As one moves out to more distantly related
taxa in Fig. 2.1 other than S. cerevisiae; for exam- species, the colinearity of gene order begins to
ple, two species closely related to C. albicans (C. break down (Keogh et al. 1998; Llorente et al. 2000b;
dubliniensis and C. tropicalis) are currently being Wong et al. 2002). Between S. cerevisiae and C.
sequenced, and so for any gene it will soon be- glabrata (Dujon et al. 2004), only about 58% of
come possible to examine its evolution over short genes have the same neighbors due to an increas-
Genome Evolution in Hemiascomycete Yeasts 21
ing number of chromosomal rearrangements and during the evolution of some species in this
loss of alternative copies of duplicated genes in the group, including S. cerevisiae (Wolfe and Shields
two species. Between S. cerevisiae and A. gossypii, 1997). The most likely phylogenetic position of
there are about 500 breaks in synteny caused by this event is shown by the asterisk in Fig. 2.1,
interchromosomal rearrangements that have oc- and it divides the hemiascomycete species into
curred since those species shared a common an- two groups: ‘pre-WGD’ and ‘post-WGD’. In this
cestor. Between S. cerevisiae and C. albicans, only shorthand nomenclature, ‘post-WGD’ means
about 9% of genes that are immediate neighbors a species that is a descendant of the WGD event
in one species are also immediate neighbors in the (the species marked by thick branches in Fig. 2.1),
other, and gene order is often seen to have become and ‘pre-WGD’ means a species that separated
scrambled by the action of numerous inversions of from the S. cerevisiae lineage before the WGD
segments of DNA containing a few genes (Seoighe event happened (thin branches in Fig. 2.1).
et al. 2000). The WGD resulted in a transient doubling of
Colinearity of gene order among species breaks the number of genes and chromosomes in an an-
down quite noticeably in the subtelomeric regions, cestral yeast, but this was followed quickly by the
which span the terminal 30–50 kilobase pairs of loss of many duplicated genes (Fig. 2.2). Today, the
each chromosome (Dietrich et al. 2004; Dujon S. cerevisiae genome contains more than 500 pairs
et al. 2004; Kellis et al. 2004). This is also apparent of duplicated genes (ohnologs) that were formed
in comparisons among the Saccharomyces sensu simultaneously by this event (Dietrich et al. 2004;
stricto species (Kellis et al. 2003; Maciaszczyk et al. Dujon et al. 2004; Kellis et al. 2004; Wolfe 2004).
2004), and in some cases even among different The functional consequences of the WGD in S. cere-
isolates of S. cerevisiae (Daran-Lapujade et al. visiae have been reviewed recently elsewhere (Kel-
2003; Nomura and Takagi 2004). Subtelomeric lis et al. 2004; Piskur and Langkjaer 2004; Wolfe
regions are notable for the presence of some 2004; Wong and Wolfe 2005) and will not be dis-
groups of co-functional genes – such as the MAL cussed here. Because there have been relatively few
and SUC genes required for growth on particular genomic rearrangements since the WGD, many of
carbon sources, or the BIO genes for synthesis of the ohnologs are still arranged in a pattern that
the enzyme cofactor biotin – and transcriptional betrays their origins: a series of genes on one chro-
regulation of these clusters may involve epigenetic mosome has a series of homologs on another chro-
mechanisms such as chromatin modification mosome, generally with conserved order and ori-
(Robyr et al. 2002; Halme et al. 2004). Subtelomeric entation of the duplicated genes (Wolfe and Shields
regions are also notable for having high sequence 1997). In between the duplicated genes, there are
similarity among different chromosomes, and many singleton genes on each chromosome. These
for frequently containing damaged genes (i. e., are loci that were originally duplicated as part of
recently formed pseudogenes; Harrison et al. 2002; the WGD, but where one of the gene copies was sub-
Lafontaine et al. 2004). These parts of the genome sequently lost. Most likely there was no evolution-
may be ‘evolution’s cauldron’: a sort of combined ary advantage to retaining both copies of the gene,
workshop and junkyard where dynamic DNA and so there was no selection against a mutation
turnover rapidly creates new genes or groups of that inactivated or deleted one of the copies. In C.
genes, epigenetic regulation is used as a way of glabrata, another post-WGD species, the number
testing these gene arrangements in a controlled of genes that have been retained in duplicate is less
manner, and ultimately natural selection either than in S. cerevisiae (about 404 pairs), which means
retains or discards them (Eichler 2001; Kent et al. that when the C. glabrata genome is compared to
2003). itself, the duplicated blocks that can be detected are
shorter and fewer than in S. cerevisiae (Dujon et al.
2004). The patterns of post-WGD gene losses that
are seen in S. cerevisiae, C. glabrata and S. castellii
III. Genomic Evidence are similar, which suggests that these three species
for Whole-Genome Duplication are all descended from the same WGD event (D.R.
Scannell, K.P. Byrne and K.H. Wolfe, unpublished
Comparisons of gene order in the hemias- data).
comycetes are complicated by the presence of Comparing gene order between pre-WGD and
a whole-genome duplication (WGD) that occurred post-WGD species reveals a 1:2 relationship (Keogh
(a) pre-WGD gene order

1 2 3 4 5 6 7 8 9 10
WGD
gene post-WGD order

losses 1 3 6 7 8 10
1 2 4 5 6 9 10
Fig. 2.2.
Fig. 2.3. (a) Illustration of our model of gene order between parts of two sister regions in the post-WGD
evolution following whole-genome duplication (WGD). species S. cerevisiae (Scer: chromosome XIII genes are
The box at the top shows a hypothetical region of darkly shaded and chromosome VII genes lightly shaded),
chromosome containing ten genes numbered 1–10. After S. castellii (Scast) and C. glabrata (Cgla), and the single
WGD, the whole region is briefly present in two copies. homologous genomic region in the pre-WGD species A.
However, many genes subsequently return to single-copy gossypii (Agos), K. lactis (Klac) and K. waltii (Kwal). In
state because there is no evolutionary advantage to this representation, each rectangle represents a gene, and
maintaining both copies. In this example, only genes 1, homologs are arranged as vertical columns. Arrows below
6 and 10 remain duplicated. However, the arrangement the rectangles show transcriptional orientation. Shaded
of these three homolog pairs in the post-WGD species lines connect genes that are adjacent but do not indicate
(bottom) would be sufficient to allow the sister regions the actual gene spacing on the chromosome. The ADH2
to be detected using that genome sequence alone. Also, gene, which is present only in S. cerevisiae (and other
the order of genes in sister regions in post-WGD species sensu stricto species not shown here), is highlighted, as is
have well-defined relationships to the gene order that a tandem duplication shared by the pre-WGD but not the
existed in the pre-WGD genome (top), which will also post-WGD species. This image is a screenshot from a Yeast
be similar to the gene order seen in any species that Gene Order Browser (YGOB) currently under development
diverged from the WGD lineage before the WGD occurred in our laboratory (K.P. Byrne and K.H. Wolfe, unpublished
(from Haber and Wolfe 2005, based on Keogh et al. 1998). data)
(b) (next page) An example of gene order relationships
et al. 1998; Seoighe and Wolfe 1999). For any ge- paired regions that we originally reported are sim-
nomic region in a pre-WGD species such as K. ply the parts of the genome that have retained the
lactis, there are two corresponding genomic re- highest densities of duplicated genes, which made
gions in post-WGD species such as S. cerevisiae it possible to recognize the sisters.
or C. glabrata, each of which contains a subset of
the genes present in the region in the pre-WGD
species (Fig. 2.2). The recent demonstrations that
this 1:2 relationship between pre-WGD and post- IV. Pairs of Centromeres
WGD species extends across essentially the entire
genomes, both of S. cerevisiae and of the pre-WGD Whole-genome duplication should have the effect
species (Wong et al. 2002; Dietrich et al. 2004; Du- of doubling the number of chromosomes, and chro-
jon et al. 2004; Kellis et al. 2004), provides direct mosome number is likely to remain unchanged
proof that the WGD hypothesis is correct, despite during the process of gene loss outlined in Fig. 2.2a.
the earlier controversy (Feldmann 2000; Llorente It has long been known from pulsed-field gel elec-
et al. 2000a, b; Souciet et al. 2000). A similar 1:2 gene trophoresis (PFGE) experiments that some yeasts,
order relationship is seen between the plants maize such as the Saccharomyces sensu stricto and C.
(which is an ancient polyploid) and rice (Lai et al. glabrata, are characterized by having a large num-
2004). The only exceptions in yeasts to this 1:2 rule ber of relatively small chromosomes as compared
are in places where chromosomal rearrangements to other species (Sor and Fukuhara 1989; Vaughan-
in one species make it difficult to track the synteny Martini et al. 1993), and it is now apparent that there
relationships (Fig. 2.2), and in the subtelomeric re- is indeed an approximate doubling of the number
gions, where wholesale movement of genes appears of chromosomes in post-WGD species (Fig. 2.1).
to have taken place in every species. The WGD hy- However, the arithmetic is not precise: the pre-
pothesis has now met with widespread acceptance WGD species have six, seven or eight chromosomes,
in the yeast community (Dujon et al. 2004; Piskur and the post-WGD species have 16 (Saccharomyces
and Langkjaer 2004). sensu stricto), 13 (C. glabrata) or even nine (if
Two misunderstandings of the WGD hypoth- a PFGE estimate for S. castellii is accurate; Petersen
esis are often made and deserve some discussion. et al. 1999). How chromosome number can change
First, it is incorrect to think that the majority of during evolution is discussed below.
duplicated genes in the S. cerevisiae genome were The number of chromosomes in a species
formed by the WGD event. In fact, only a few hun- is determined primarily by the number of cen-
dred pairs of duplicated genes in S. cerevisiae sur- tromeres. For the species whose genomes have
vive from the WGD (we originally identified 376 been sequenced to completion or to high-coverage
pairs in 1997, but by leveraging information from drafts, it is now possible to track the evolutionary
other genome sequences the number is now known history of each centromere (Fig. 2.4). Except in
to be about 551 pairs; Wolfe 2004; K.P. Byrne and S. cerevisiae and K. lactis (Heus et al. 1993), the
K.H. Wolfe, unpublished data). Every gene in the locations of centromeres have been predicted
genome was duplicated transiently, but at about based on matches to the CDE I–CDE II–CDE III
90% of locus pairs, one of the two copies of the consensus in the genome sequence, rather than by
gene was subsequently deleted again. Because the laboratory experiments, but in each case there is
genome that underwent duplication was itself com- only one match to the consensus per chromosome,
prised of many multigene families, like any other and so the predictions are probably accurate.
eukaryotic genome, many S. cerevisiae genes have Based on the identities of the protein-
paralogs that were formed by duplications that hap- coding genes flanking the centromeres, there is
pened independently of the WGD, but none of those a straightforward 2:1 relationship between the
duplications were multigene events involving large 16 centromeres of S. cerevisiae and the eight
segments of chromosome. Second, there is nothing centromeres of K. waltii (Fig. 2.4; Kellis et al.
special about the 50% of the genome that we were 2004), which supports the hypothesis that the
able to arrange into 55 paired regions in our origi- WGD involved an eight-chromosome ancestor
nal analysis (Wolfe and Shields 1997). The hypoth- turning into a 16-chromosome descendant (Keogh
esis is that the whole genome underwent duplica- et al. 1998; Wong et al. 2002). A. gossypii has only
tion, and so in principle every point in the genome seven centromeres, all of which are at positions
should have a sister point somewhere else. The 55 that are also centromeres in K. waltii (Fig. 2.4).
If we assume that the ancestral number of chro- again). An ancestral number of eight seems more
mosomes in the pre-WGD species was eight, plausible than (say) seven, because under a model
then A. gossypii has lost the equivalent of the of a seven-chromosome ancestor undergoing WGD
centromere of K. waltii chromosome 5 (KwCEN5). to produce a 14-chromosome descendant, there
The genes flanking KwCEN5 are found on two must subsequently have been a gain of one cen-
different chromosomes of A. gossypii and are not tromere in K. waltii and a gain of two centromeres
near centromeres. Only three of K. lactis’s six in S. cerevisiae and, eerily, these centromeres must
centromeres (KlCEN2, KlCEN4 and KlCEN5) are in have been gained at equivalent positions on the
colinear relationships with K. waltii centromeres, three chromosomes. At the moment, there are no
both to the left and to the right of the centromere clear examples of centromeres being gained during
(Fig. 2.4). The other three K. lactis centromeres hemiascomycete evolution other than by the WGD
(KlCEN1, KlCEN3 and KlCEN6) have synteny mechanism. In C. albicans there are eight chromo-
with a K. waltii centromere on one chromosome somes and their centromeres have been identified
arm, but there is a rearrangement breakpoint by biochemical methods (Sanyal et al. 2004), but the
immediately on the other side of the centromere. genes flanking these centromeres bear little or no
The most interesting aspect of these centromere relationship to the centromere groups identified in
comparisons among pre-WGD species is that there Fig. 2.4 (data not shown). This may just be a con-
are two places in the K. lactis genome that show sequence of the larger evolutionary distance and
gene order conservation with the regions spanning more numerous rearrangements between C. albi-
centromeres in both K. waltii and A. gossypii, but cans and the S. cerevisiae group of species (Seoighe
which are not centromeres in K. lactis (marked by et al. 2000), though the C. albicans gene order does
X symbols in Fig. 2.4). These ‘non-centromeric’ seem to be even more rearranged than one might
regions in K. lactis are the counterparts of Kw- expect from its evolutionary distance alone.
CEN7/AgCEN3 and KwCEN2/AgCEN5. They occur
on K. lactis chromosomes 4 and 5 at sites hundreds
of kilobases distant from the true centromeres of
those chromosomes. If the ancestral number of V. A Model for Genome Duplication
chromosomes was eight and K. lactis underwent
a reduction to six, then these are the two sites from Although it is clear that the genome became dou-
which centromere function has been lost without bled in an ancestor of S. cerevisiae and the other
further chromosomal rearrangement. post-WGD species, the exact details of what hap-
In the post-WGD species, at least 12 and prob- pened have not yet been worked out. It is convenient
ably all 13 of C. glabrata’s centromeres are directly to refer to this event as a genome duplication, but
orthologous to centromeres in S. cerevisiae (see leg- more accurate to call it a polyploidization. Poly-
end to Fig. 2.4 for details of the one debatable case). ploidy can arise by either of two routes, depending
The difference in chromosome number between C. on whether the two constituent genomes are identi-
glabrata and S. cerevisiae is again accounted for cal (autopolyploidy) or different (allopolyploidy).
by the presence of ‘non-centromeric’ regions: three It seems increasingly likely that the WGD event
places in the C. glabrata genome that show con- was an allopolyploidy, for several reasons. First,
served gene order with the regions flanking three in an autopolyploid the two incoming sequences
centromeres in S. cerevisiae (ScCEN10, ScCEN11 of every gene are identical, and so they cannot
and ScCEN14), but which are not centromeres in C. have diverged at all in function. This means that,
glabrata (Fig. 2.4). in the immediate aftermath of an autopolyploidy
The above comparisons make it very proba- event, the only selective force that can lead to the
ble that the ancestral number of chromosomes was survival of both copies is selection for increased
eight, and the WGD doubled this to 16. The re- dosage of the locus. Although some loci, such as
duced numbers of chromosomes in K. lactis, A. the cytosolic ribosomal protein genes of S. cere-
gossypii and C. glabrata are probably due to chro- visiae, probably have been retained due to selection
mosome fusions that resulted in the death of cen- for increased dosage, this explanation seems appli-
tromeres at the ‘non-centromeric’ sites described cable only to a minority of the ohnologs that have
above (because chromosome fusions are only evo- been retained by S. cerevisiae (Wolfe 2004). Second,
lutionarily viable if one of the two centromeres there are at least two documented instances of in-
becomes inactive, or if the chromosome breaks terspecies hybridizations that occurred during the
Fig. 2.4. Relationships among centromeric regions of S. arm. Hooked lines indicate loss of relatedness. The assign-
cerevisiae (Sc), C. glabrata (Cg), K. waltii (Kw), A. gossypii ment of the centromere of C. glabrata chromosome 9 to
(Ag) and K. lactis (Kl) chromosomes. Each of the eight the group on the bottom right is less certain than for the
panels shows a group of genomic regions that are related other centromeres, and is based on the linkage of CgCEN9
by virtue of their gene contents close to a centromere (or to the gene CAGL0I08107g, which is an ortholog of the genes
a similar non-centromeric region). Closed circles represent KLLA0D01243g (K. lactis) and s56_22439 (K. waltii) that are
centromeres in pre-WGD species, and open circles repre- close to the corresponding regions in those species. Gene
sent centromeres of post-WGD species. X symbols indi- names in K. lactis and C. glabrata have been shortened by
cate the absence of a centromere. Names indicate chro- writing A00627 instead of CAGL0A00627g, etc.
mosome arms, or genes close to the centromere on each
recent evolution of hemiascomycetes, which makes What is also surprising about the evolution of
an older hybridization plausible, too. Hybridiza- the yeast genome is that, despite the common oc-
tion between S. cerevisiae-like and S. bayanus-like currence of aneuploidy in industrial or laboratory
ancestors formed the brewing yeast S. pastorianus strains (Hughes et al. 2000; Hansen and Kielland-
(Hansen and Kielland-Brandt 1994, 2003), and hy- Brandt 2003), and despite the frequency at which
bridization between Zygosaccharomyces rouxii and segmental duplications (single events that dupli-
a related species formed an allopolyploid that is cate large regions of chromosome) can arise in lab-
used in the miso and soy sauce industries (Kinclova oratory experiments that select for increased copy
et al. 2001; J. Gordon and K.H. Wolfe, unpublished number of a gene (Koszul et al. 2003), these types of
data). Third, the pattern of gene loss and retention event seem to have been quite rare during the nat-
in different post-WGD species is suggestive of al- ural evolution of yeasts. In the complete genomes
lopolyploidy (D.R. Scannell, K.P. Byrne and K.H. of four Saccharomyces sensu stricto species, there
Wolfe, unpublished data). is only one example of a segmental duplication
Here we propose a simple model for how an other than in a subtelomeric region (Kellis et al.
allopolyploidy could have resulted in the WGD 2003).
(Fig. 2.5). Much of the genetics literature about
polyploidy has been centered on studies of poly-
ploid plants or animals (Coyne and Orr 2004), and
yeasts differ somewhat from these organisms be- VI. Pseudogenes, Relics and Null Alleles
cause they are able to grow mitotically as either
haploids or diploids. In a yeast, allopolyploidy can It was originally thought that the S. cerevisiae
occur if gametes from two different species fuse genome contains very few pseudogenes, because
to form a hybrid zygote. Unless the two parental the genome is compact and there is relatively little
species are very closely related, the hybrid will be non-coding DNA. The XYZ3 example mentioned
sterile because chromosomal rearrangements will above was one of the first descriptions of an S.
prevent it undergoing meiosis (Delneri et al. 2003), cerevisiae pseudogene (Lalo et al. 1993), and there
and so it will grow only mitotically and may do was so much doubt about its veracity that it was
so for many generations. One way for the hybrid re-sequenced several times by several laboratories
to escape from this life of abstinence and regain before it was finally accepted as real. More recently,
the ability to reproduce sexually is if one allele at careful analysis has revealed the existence of 120
the MAT locus gets deleted spontaneously, which pseudogenes in S. cerevisiae that are so degener-
has the effect of turning a diploid into a haploid ated that they have been called ‘relics’ (Lafontaine
(Fig. 2.5). In other words, the MAT locus sim- et al. 2004), of which about 40% are at internal
ply needs to be among the many duplicated loci (non-telomeric) sites in chromosomes. The severe
that lost one copy in the aftermath of the WGD, truncation and extensive sequence divergence in
during mitotic replication of the hybrid cell lin- the relics had the effect that, in many cases, they
eage. Once the MAT locus becomes single-copy, HO were found only through the painstaking use of
endonuclease will catalyze mating-type switching thousands of dot-matrix plots to further investi-
in mother cells. This will enable mating between gate very short BLASTN hits between known genes
mother and daughter cells, forming diploids with and putatively relic-containing intergenic regions
homogeneous genomes whose chromosome num- (Lafontaine et al. 2004).
ber will be the sum of the numbers of chromosomes Two other types of pseudogenes are notable.
in the parental species (Fig. 2.5). First, there are several examples in S. cerevisiae
It should be noted that the model sketched in where a pair of homologous open reading frames
Fig. 2.5 is also applicable to an autopolyploidization (ORFs) exist, but one of these is much larger than
event, if a deletion of one MAT allele occurs in the other. An early example of this is the CDC4 gene
a normal diploid cell before it undergoes meiosis. on chromosome VI, which codes for a 779-residue
However, for the reasons outlined above, we think protein and has a truncated homolog on chromo-
that an allopolyploidization was more probable. It some V (YER066W, a 186-codon ORF; Yochem and
is also noteworthy that this model is plausible only Byers 1987). In cases such as this, it is often un-
for yeasts that have HO endonuclease, and that HO clear whether the second locus is a pseudogene that
originated in yeasts only shortly before the WGD contains a fortuitous ORF, or whether the shorter
happened (Butler et al. 2004; see Sect. VII). gene has a function. Second, there are several loci
Fig. 2.5. A model for how an allopoly-

ploidization could result in WGD, if
one copy of the MAT locus becomes
deleted after hybridization. HM refers
to the HML and HMR cassettes of silent
mating-type information, which are
located on two different chromosomes
in all species other than S. cerevisiae
(Dujon et al. 2004)
where a gene in the reference S. cerevisiae genome abled S. cerevisiae to adopt a lifestyle whereby it
sequence (from strain S288c) contains an internal uses Adh1 to rapidly ferment carbon sources to
stop codon, but the gene is known to be intact and make ethanol (a product that is toxic to many com-
functional in other laboratory strains of S. cere- peting microorganisms), and then later uses Adh2
visiae such as W303 or Σ1278b. Examples include to respire the ethanol (Ashburner 1998; Benner
FLO8, CRS5, YOL153C and SDC25 (Damak et al. et al. 2002). Without Adh2 it is unlikely that the
1991; Liu et al. 1996; Dujon et al. 1997), and these sensu stricto yeasts would have evolved their vig-
are probably better described as null alleles in S. orous fermentation phenotype.
cerevisiae S288c, rather than pseudogenes. Some of There are a few interesting exceptions to the
these null alleles may have been selected for during general rule of gene conservation across the hemi-
the process of turning S. cerevisiae into a tractable ascomycetes – loci where either a clade of species
laboratory organism (Hall and Linder 1993). has picked up a new gene by horizontal gene trans-
fer, or where a species has lost a gene (or a set of
genes) because they are no longer necessary due
to a change in the lifestyle of the organism. Exam-
VII. Interspecies Differences in Gene ples of these gene content differences are discussed
Content below.
One clear case of horizontal gene transfer into
A. Genes Gained a group of yeasts is URA1, the gene that in S.
cerevisiae codes for dihydroorotate dehydrogenase
One of the most striking differences among the (DHODase), the fourth step in the pathway of
sequenced hemiascomycete genomes is the extent de novo pyrimidine biosynthesis (Gojkovic et al.
to which tandemly duplicated genes are present in 2004). There are two types of DHODase enzyme
the genomes. Tandem repeats are quite rare in S. in eukaryotes, with very different sequences.
cerevisiae (<2% of genes), but in Debaryomyces The ancestral eukaryotic form of DHODase is
hansenii about 10% of genes are part of a tan- mitochondrial and is found in plants, animals
dem array (Dujon et al. 2004). Species-specific tan- and most fungi. In contrast, the URA1 DHODase
dem repeats are one of the most frequent causes present in S. cerevisiae is more closely related to
of non-colinearity in genomic regions aligned be- DHODases of bacteria (specifically Lactococcus
tween species (Fig. 2.2b). In S. cerevisiae one of the lactis) than to the ancestral eukaryotic enzyme,
principal attributes of tandem arrays is that they and is cytosolic in S. cerevisiae. Some yeasts, such
can become further amplified under particular en- as S. kluyveri, K. lactis and K. waltii, have kept
vironmental conditions where there is strong selec- genes for both types of DHODase (Dujon et al.
tion for increased copy number of the gene – e.g., at 2004; Gojkovic et al. 2004; Kellis et al. 2004; Piskur
the CUP1 metallothionein locus (Fogel and Welch and Langkjaer 2004). In S. cerevisiae and S. castel-
1982) and the PMR2 locus coding for a sodium ion lii, the ancestral gene has been lost and only the
pump (Wieland et al. 1995). Hence, it is possible horizontally transferred bacterial URA1 remains
that the tandemly repeated loci in other species – an example of gene displacement. The driving
are also involved in responses to environmental force behind this shift may have been selective
change. pressure on yeasts for increased ability to grow
Although there have been many expansions anaerobically (Gojkovic et al. 2004; Piskur and
and contractions of gene families in different yeast Langkjaer 2004). The mitochondrial DHODase
species (Dujon et al. 2004), the great majority of uses a quinone as an electron acceptor. Hence,
genes in each hemiascomycete have orthologs in pyrimidine synthesis is coupled to mitochondrial
all the other species. One gene family expansion respiration, whereas the cytosolic URA1 enzyme
of great economic importance is the gene duplica- instead uses fumarate as its electron acceptor,
tion that gave rise, specifically in the Saccharomyces and fumarate is made via the glyoxylate cycle in
sensu stricto lineage, to the ADH2 gene by dupli- the absence of respiration (Gojkovic et al. 2004).
cation of the ADH1 alcohol dehydrogenase gene The adoption of URA1 allows S. cerevisiae and S.
(Fig. 2.2b). ADH2 codes for an enzyme that ef- castellii to make pyrimidines even in the absence
ficiently converts ethanol to acetaldehyde and is of oxygen. Rather surprisingly, C. glabrata has
expressed only in aerobically growing cells (Wills done the opposite to S. cerevisiae: it lost the
1976). This duplication is hypothesized to have en- URA1 ortholog and retained only the gene for the
mitochondrial DHODase (CAGL0M12881g) – an HO: it is present in the Saccharomyces sensu stricto

example of failed horizontal gene transfer. species and S. castellii, but not in any of the studied
The gene for HO endonuclease, which cat- pre-WGD species, except for Z. rouxii (Haber and
alyzes mating-type switching in S. cerevisiae by Wolfe 2005). Every studied species has either both
making a double-strand break at the MAT locus, HO and SIR1, or neither, with the sole exception
is found only in species quite closely related to of Candida glabrata. Fabre et al. (2005) recently
S. cerevisiae (it is present in all the post-WGD concluded that HO appeared in yeasts earlier than
species and in one pre-WGD lineage represented SIR1, but we disagree with their interpretation be-
by Zygosaccharomyces rouxii; Butler et al. 2004). cause (1) there is a SIR1 homolog in Z. rouxii (which
The existence of HO has the consequence that implies an older origin of the gene, and secondary
natural isolates of S. cerevisiae grow as diploids, loss in C. glabrata), and (2) the K. lactis HO-like
with only a very brief haploid phase in their sequence they describe does not seem to be a true
life cycle. In contrast, most pre-WGD lineages ortholog of HO (it is an outlier to the HO/VMA1
such as K. lactis grow primarily as haploids; clade in phylogenetic trees, and instead clusters
they switch mating type only rarely and they with the S. castellii intein-like gene Scas_542.2; data
sporulate immediately after mating (Kurtzman not shown).
and Fell 1998; Zonneveld and Steensma 2003).
The evolutionary advantage conferred by HO is
pseudohomothallism: a new, sexually reproducing B. Genes Lost
population of S. cerevisiae can be established by
just a single colonizing cell, whereas in a het- The converse of gene gain is gene loss. One of the
erothallic species such a K. lactis it is necessary benefits of complete genome sequencing is that
for two cells of opposite mating type to be present it allows one to make definitive statements about
if the new population is to reproduce sexually. what is absent from a genome, as well as what
Thus, HO is likely to facilitate the dispersal of is present, and several examples have recently
a species. Sexual reproduction is important for been found of complete biochemical pathways that
the long-term evolutionary success of a species have been lost in one or more hemiascomycete
because without recombination all chromosomes species.
will accumulate some deleterious mutations S. kudriavzevii cannot grow on galactose as
by Muller’s ratchet. Hence, mitotic (asexual) a sole carbon source, and analysis of its genome
reproduction is not a stable long-term strategy showed that seven GAL loci are all pseudogenes in
(Berbee and Taylor 1993). Horizontal gene transfer this species (Hittinger et al. 2004). All the genes that
is implicated in the origin of HO, because it is are exclusively involved in galactose catabolism
derived from an intein – a selfish genetic element have become pseudogenes, but other genes that
found primarily in bacterial genomes – and have additional roles in non-galactose aspects of
because its closest homolog, the intein in the metabolism (such as GAL5) are still intact. Loss
S. cerevisiae gene VMA1, has been horizontally of GAL activity in S. kudriavzevii is quite recent
transferred among yeast species (Gimble and (<10 million years), because the GAL genes are all
Thorner 1992; Gimble 2000; Koufopanou et al. intact in other sensu stricto species. The GAL path-
2002; Okuda et al. 2003; see Haber and Wolfe 2005 way has also been lost in A. gossypii, K. waltii and
for review). C. glabrata, by two or three additional loss events
A third possible candidate for horizontal gene (Hittinger et al. 2004).
transfer is SIR1, whose function is linked to that Similarly, the lactose assimilation pathway has
of HO (Fabre et al. 2005). The Sir1 protein’s role is been lost in many hemiascomycetes. In K. lac-
to mediate the recruitment of the silencer complex tis, the ability to grow on lactose is conferred by
Sir2/Sir3/Sir4 to the silent HM cassettes of mating- LAC4 (beta-galactosidase) and LAC12 (lactose per-
type information. The Sir proteins alter the chro- mease), two genes that are located beside each
matin conformation of the HM cassettes so that other in the genome (Chang and Dickson 1988;
they are both transcriptionally inactive and resis- Poch et al. 1992). LAC12 is a member of a large
tant to cleavage by HO endonuclease. Although family of sugar permeases whose evolutionary his-
Sir2, Sir3 and Sir4 have a broad phylogenetic dis- tory is difficult to trace. LAC4 has no homolog in
tribution (Astrom and Rine 1998), Sir1 is restricted other hemiascomycete species except for Debary-
to almost the same set of species as those that have omyces hansenii, but it also has homologs in several
euascomycetes such as Magnaporthe grisea and VIII. Rapidly Evolving Loci

Neurospora crassa. Given this phylogenetic distri-
bution, it is likely that the LAC4 gene has been
lost from several hemiascomycete lineages, argu- Most genes in hemiascomycete genomes have clear
ing against horizontal gene transfer from a bac- orthologs in other species that show both conserva-
terium into K. lactis as was suggested by Poch et al. tion of protein sequence and conservation of local
(1992). gene order relationships (Fig. 2.2). In a few places
Other examples of pathway losses include the in the genome, however, synteny information sug-
absence in C. glabrata of genes in the DAL (al- gests that some genes in different species might be
lantoin catabolism) pathway, and the SNO and SNZ orthologs, even though there is little or no sequence
genes involved in pyridoxine synthesis (Dujon et al. similarity. These are likely to be loci with very
2004). A three-gene pathway conferring the ability rapid sequence evolution. An example is shown
to grow on pyrimidines as a sole source of nitrogen in Fig. 2.6: in S. cerevisiae a gene of unknown func-
is present in the pre-WGD species S. kluyveri (Goj- tion, YKL023W, lies between URA6 and CDC16. In
kovic et al. 2000, 2001), K. lactis and K. waltii, but every other hemiascomycete genome, there is also
has been lost both in the post-WGD species and a putative gene in the interval between URA6 and
in A. gossypii. These examples serve to illustrate CDC16, and these genes are all transcribed toward
the principle that a gene will be maintained in an CDC16. This suggests that the genes in the interval
organism’s genome only if there exists the natural are orthologs, but they have no detectable sequence
selection to prevent its loss, and that a change in similarity to each other (in a BLASTP search, none
the ecology of the organism can result in a path- of them hits any of the others with an E-value <1)
way becoming unnecessary. This seems to happen and they range in size from 185 codons (A. gossypii
quite frequently in the case of genes involved in ACR169C) to 408 codons (K. lactis KLLA0E08008g).
metabolic pathways that permit growth on partic- S. cerevisiae YKL023W does have orthologs in the
ular substrates. other Saccharomyces sensu stricto species, and it is
One gene loss whose significance is not yet among the fastest 5% of genes in terms of its rate of
fully understood is the loss of MATa2, a gene for protein sequence evolution among the sensu stricto
an HMG-domain DNA-binding protein, from the species (Kellis et al. 2003). This indicates that the
MAT locus in post-WGD yeast species (Tsong et al. genes in the URA6–CDC16 interval in the other
2003; Butler et al. 2004). In C. albicans, MATa2 species are also very probably orthologs, but the
is a positive regulator of expression of a-specific combination of the gene’s inherently fast rate of
genes in cells of mating type a. In S. cerevisiae no sequence change and the longer period of evolu-
such activator is needed, because expression of tionary time involved in comparisons outside the
a-specific genes occurs by default in the absence sensu stricto group has resulted in sequences that
of the repressor molecules Mcm1-α2 (present have diverged beyond the ability of BLAST to detect
in MAT α haploid cells) or a1-α2 (present in their relatedness. Other examples of rapidly evolv-
MATa/MAT α diploids) (Tsong et al. 2003; Scannell ing genes were identified by Kellis et al. (2003) in
and Wolfe 2004). In C. albicans the α2 repressor their comparison of sensu stricto genomes. One
has no effect on the expression of a-specific genes, such locus, YBR184W, has a very high ratio of non-
unlike in S. cerevisiae. The MATa2 gene is present synonymous to synonymous nucleotide substitu-
in all pre-WGD species (except Pichia angusta) but tions (indicating low constraint on the amino acid
absent from all post-WGD species that have been sequence), and has no detectable homologs outside
studied (Butler et al. 2004). the sensu stricto species.
Lastly, it is interesting that the gene for chiti- Gene pairs (ohnologs) that were formed by
nase (CTS1) is absent from the A. gossypii genome the WGD are also sometimes seen to be diverg-
(Dietrich et al. 2004). In mitotic growth of budding ing very rapidly in S. cerevisiae or other post-WGD
yeasts, chitinase degrades the chitin ring between species (Kellis et al. 2004; Wolfe 2004). These pairs
mother and daughter, allowing the cells to separate. are strong candidates for having undergone diver-
A. gossypii is unusual among hemiascomycetes in gence of function following the WGD, and several
growing only filamentously – it forms multinucle- of them show asymmetrical evolution where one
ate hyphae and has no yeast phase (Alberti-Segui member of the pair has apparently retained an an-
et al. 2001). It is the only known hemiascomycete cestral function while the faster-evolving copy has
lacking a CTS1 homolog. taken on a new function (Kellis et al. 2004).
Fig. 2.6. Gene order relationships around the locus K. lactis (Klac) and K. waltii (Kwal). S. castellii also
YKL023W. Although they are shown in separate columns has a large ORF in the URA6–CDC16 interval that is
because they do not have BLASTP hits to one another, missing from this YGOB view (see Fig. 2.2b) because it
the genes in the region marked by the brace may in was overlooked in the genome annotation (Cliften et al.
fact be rapidly evolving orthologs: C. glabrata (Cgla), S. 2003)
castellii (Scast), S. cerevisiae (Scer), A. gossypii (Agos),
IX. Conclusions greatly outstrips our knowledge of their biology

and natural ecosystems. Hence, it is likely that our
appreciation of the genetic causes of the biologi-
S. cerevisiae is so entrenched as a model organism cal differences between these species is still in its
for the hemiascomycetes that it is worth pausing infancy. In a sense, we are at the limits of the reduc-
to consider whether, in the light of the new ge- tionist approach to biology: the cells of all hemi-
nomics data available, it is really a good represen- ascomycetes work in much the same way, and we
tative of this group of yeasts. What the genomics have already obtained a reasonable understand-
data show is extensive conservation of gene content ing of how they work through decades of research
and extensive colinearity of gene order, and so from into the biology of S. cerevisiae and other model
that standpoint S. cerevisiae is highly representa- organisms. That is the beauty of the reduction-
tive, at least of the post-WGD species. However, the ist approach. What is perhaps most interesting,
genome data also show that there are many species- however, and what remains for the future, is what
specific quirks. Overlaid onto the conserved core lies outside this reduced core – the rapidly evolv-
of genes that make up any hemiascomycete are ing genes, the laterally transferred genes, the lost
species-specific attributes such as the presence of genes, and the expanded gene families – that make
particular genes. In some cases – such as LAC4 in the proteome of each species unique and that en-
K. lactis, or ADH2 in S. cerevisiae – these attributes abled each species to evade extinction by carving
confer such a strong phenotype that they almost out a foothold in the biosphere.
define the species, at least from a human perspec-
tive. In other cases – such as the loss of CTS1 in A.
gossypii – they may be correlated with a phenotypic Acknowledgements. We thank Kevin Byrne, Devin Scan-
nell, Jonathan Gordon and Simon Wong for discussion. Re-
change without being the root cause of it. search in our laboratory is supported by Science Foundation
At the moment our knowledge of the genomes Ireland. L.J. Montcalm was supported by the Plains of Abra-
of most hemiascomycetes other than S. cerevisiae ham Trust.
References Dietrich FS, Voegeli S, Brachat S, Lerch A, Gates K, Steiner S,

Mohr C, Pohlmann R, Luedi P, Choi S et al. (2004)
The Ashbya gossypii genome as a tool for mapping
Alberti-Segui C, Dietrich F, Altmann-Johl R, Hoepfner D, the ancient Saccharomyces cerevisiae genome. Science
Philippsen P (2001) Cytoplasmic dynein is required 304:304–307
to oppose the force that moves nuclei towards the hy- Dujon B, Albermann K, Aldea M, Alexandraki D,
phal tip in the filamentous ascomycete Ashbya gossypii. Ansorge W, Arino J, Benes V, Bohn C, Bolotin-
J Cell Sci 114:975–986 Fukuhara M, Bordonné R et al. (1997) The nucleotide
Ashburner M (1998) Speculations on the subject of alcohol sequence of Saccharomyces cerevisiae chromosome
dehydrogenase and its properties in Drosophila and XV. Nature suppl 387:98–102
other flies. Bioessays 20:949–954 Dujon B, Sherman D, Fischer G, Durrens P, Casaregola S,
Astrom SU, Rine J (1998) Theme and variation among Lafontaine I, de Montigny J, Marck C, Neuvéglise C,
silencing proteins in Saccharomyces cerevisiae and Talla E et al. (2004) Genome evolution in yeasts. Nature
Kluyveromyces lactis. Genetics 148:1021–1029 430:35–44
Benner SA, Caraco MD, Thomson JM, Gaucher EA (2002) Eichler EE (2001) Recent duplication, domain accretion and
Planetary biology – paleontological, geological, and the dynamic mutation of the human genome. Trends
molecular histories of life. Science 296:864–868 Genet 17:661–669
Berbee ML, Taylor JW (1993) Ascomycete relationships: Fabre E, Muller H, Therizols P, Lafontaine I, Dujon B,
dating the origin of asexual lineages with 18S ribo- Fairhead C (2005) Comparative genomics in hemi-
somal RNA gene sequence data. In: Reynolds DR, Tay- ascomycete yeasts: evolution of sex, silencing and
lor JW (eds) The fungal holomorph: mitotic, mei- subtelomeres. Mol Biol Evol 22:856–873
otic and pleomorphic speciation in fungal systematics. Feldmann H (2000) Génolevures – a novel approach to ‘evo-
CAB International, Wallingford, UK, pp 67–78 lutionary genomics’. FEBS Lett 487:1–2
Bon E, Neuveglise C, Casaregola S, Artiguenave F, Wincker P, Fischer G, James SA, Roberts IN, Oliver SG, Louis EJ (2000)
Aigle M, Durrens P (2000) Genomic exploration of the Chromosomal evolution in Saccharomyces. Nature
hemiascomycetous yeasts. 5. Saccharomyces bayanus 405:451–454
var. uvarum. FEBS Lett 487:37–41 Fischer G, Neuvéglise C, Durrens P, Gaillardin C, Dujon B
Brachat S, Dietrich FS, Voegeli S, Zhang Z, Stuart L, Lerch A, (2001) Evolution of gene order in the genomes of two
Gates K, Gaffney T, Philippsen P (2003) Reinvestigation related yeast species. Genome Res 11:2009–2019
of the Saccharomyces cerevisiae genome annotation by Fogel S, Welch JW (1982) Tandem gene amplification medi-
comparison to the genome of a related fungus: Ashbya ates copper resistance in yeast. Proc Natl Acad Sci USA
gossypii. Genome Biol 4:R45 79:5342–5346
Butler G, Kenny C, Fagan A, Kurischko C, Gaillardin C, Gimble FS (2000) Invasion of a multitude of genetic niches
Wolfe KH (2004) Evolution of the MAT locus and its by mobile endonuclease genes. FEMS Microbiol Lett
Ho endonuclease in yeast species. Proc Natl Acad Sci 185:99–107
USA 101:1632–1637 Gimble FS, Thorner J (1992) Homing of a DNA endonucle-
Cai J, Roberts IN, Collins MD (1996) Phylogenetic rela- ase gene by meiotic gene conversion in Saccharomyces
tionships among members of the ascomycetous yeast cerevisiae. Nature 357:301–306
genera Brettanomyces, Debaryomyces, Dekkera, and Goffeau A, Aert R, Agostini-Carbone ML, Ahmed A,
Kluyveromyces deduced by small-subunit rRNA gene Aigle M, Alberghina L, Allen E, Alt-Mörbe J, André B,
sequences. Int J Syst Bacteriol 46:542–549 Andrews S et al. (1997) The Yeast Genome Directory.
Chang YD, Dickson RC (1988) Primary structure of the lac- Nature suppl 387:5–105
tose permease gene from the yeast Kluyveromyces lac- Gojkovic Z, Jahnke K, Schnackerz KD, Piskur J (2000) PYD2
tis. Presence of an unusual transcript structure. J Biol encodes 5,6-dihydropyrimidine amidohydrolase,
Chem 263:16696–16703 which participates in a novel fungal catabolic pathway.
Cliften P, Sudarsanam P, Desikan A, Fulton L, Fulton B, J Mol Biol 295:1073–1087
Majors J, Waterston R, Cohen BA, Johnston M (2003) Gojkovic Z, Sandrini MP, Piskur J (2001) Eukaryotic beta-
Finding functional features in Saccharomyces genomes alanine synthases are functionally related but have
by phylogenetic footprinting. Science 301:71–76 a high degree of structural diversity. Genetics 158:999–
Coyne JA, Orr HA (2004) Speciation. Sinauer, Sunderland, 1011
MA Gojkovic Z, Knecht W, Zameitat E, Warneboldt J, Coutelis JB,
Damak F, Boy-Marcotte E, Le-Roscouet D, Guilbaud R, Pynyaha Y, Neuveglise C, Moller K, Loffler M, Piskur J
Jacquet M (1991) SDC25, a CDC25-like gene which (2004) Horizontal gene transfer promoted evolution of
contains a RAS-activating domain and is a dispensable the ability to propagate under anaerobic conditions in
gene of Saccharomyces cerevisiae. Mol Cell Biol yeasts. Mol Genet Genomics 271:387–393
11:202–212 Haber JE, Wolfe KH (2005) Evolution and function of
Daran-Lapujade P, Daran JM, Kotter P, Petit T, Piper MD, HO and VDE endonucleases in fungi. In: Belfort M,
Pronk JT (2003) Comparative genotyping of the Sac- Derbyshire V, Stoddard B, Wood D (eds) Homing en-
charomyces cerevisiae laboratory strains S288C and donucleases and inteins. Springer, Berlin Heidelberg
CEN.PK113-7D using oligonucleotide microarrays. New York, Nucleic Acids and Molecular Biology vol
FEMS Yeast Res 4:259–269 16, pp 161–175
Delneri D, Colson I, Grammenoudi S, Roberts IN, Louis EJ, Hall MN, Linder P (1993) The early days of yeast genetics.
Oliver SG (2003) Engineering evolution to study spe- Cold Spring Harbor Laboratory Press, New York
ciation in yeasts. Nature 422:68–72
Halme A, Bumgarner S, Styles C, Fink GR (2004) Genetic Lalo D, Stettler S, Mariotte S, Slonimski PP, Thuriaux P
and epigenetic regulation of the FLO gene family gen- (1993) Une duplication fossile entre les régions cen-
erates cell-surface variation in yeast. Cell 116:405–415 tromériques de deux chromosomes chez la levure. C R
Hansen J, Kielland-Brandt MC (1994) Saccharomyces carls- Acad Sci Paris 316:367–373
bergensis contains two functional MET2 alleles similar Liu H, Styles CA, Fink GR (1996) Saccharomyces cerevisiae
to homologues from S. cerevisiae and S. monacensis. S288C has a mutation in FLO8, a gene required for
Gene 140:33–40 filamentous growth. Genetics 144:967–978
Hansen J, Kielland-Brandt MC (2003) Brewer’s yeast: genetic Llorente B, Durrens P, Malpertuy A, Aigle M, Artiguenave F,
structure and targets for improvement. Topics Curr Blandin G, Bolotin-Fukuhara M, Bon E, Brottier P,
Genet 2:143–170 Casaregola S et al. (2000a) Genomic exploration of the
Harrison P, Kumar A, Lan N, Echols N, Snyder M, Ger- hemiascomycetous yeasts. 20. Evolution of gene redun-
stein M (2002) A small reservoir of disabled ORFs in dancy compared to Saccharomyces cerevisiae. FEBS
the yeast genome and its implications for the dynamics Lett 487:122–133
of proteome evolution. J Mol Biol 316:409–419 Llorente B, Malpertuy A, Neuveglise C, de Montigny J,
Heus JJ, Zonneveld BJ, Steensma HY, van den Berg JA (1993) Aigle M, Artiguenave F, Blandin G, Bolotin-
The consensus sequence of Kluyveromyces lactis cen- Fukuhara M, Bon E, Brottier P et al. (2000b) Genomic
tromeres shows homology to functional centromeric exploration of the hemiascomycetous yeasts. 18.
DNA from Saccharomyces cerevisiae. Mol Gen Genet Comparative analysis of chromosome maps and
236:355–362 synteny with Saccharomyces cerevisiae. FEBS Lett
Hittinger CT, Rokas A, Carroll SB (2004) Parallel inacti- 487:101–112
vation of multiple GAL pathway genes and ecologi- Maciaszczyk E, Wysocki R, Golik P, Lazowska J, Ulaszewski S
cal diversification in yeasts. Proc Natl Acad Sci USA (2004) Arsenical resistance genes in Saccharomyces
101:14144–14149 douglasii and other yeast species undergo rapid evolu-
Hughes TR, Roberts CJ, Dai H, Jones AR, Meyer MR, Slade D, tion involving genomic rearrangements and duplica-
Burchard J, Dow S, Ward TR, Kidd MJ et al. (2000) tions. FEMS Yeast Res 4:821–832
Widespread aneuploidy revealed by DNA microarray Nomura M, Takagi H (2004) Role of the yeast acetyltrans-
expression profiling. Nat Genet 25:333–337 ferase Mpr1 in oxidative stress: regulation of oxygen
Kellis M, Patterson N, Endrizzi M, Birren B, Lander ES reactive species caused by a toxic proline catabolism
(2003) Sequencing and comparison of yeast species intermediate. Proc Natl Acad Sci USA 101:12616–12621
to identify genes and regulatory elements. Nature Okuda Y, Sasaki D, Nogami S, Kaneko Y, Ohya Y, Anraku Y
423:241–254 (2003) Occurrence, horizontal transfer and degener-
Kellis M, Birren BW, Lander ES (2004) Proof and evolution- ation of VDE intein family in Saccharomycete yeasts.
ary analysis of ancient genome duplication in the yeast Yeast 20:563–573
Saccharomyces cerevisiae. Nature 428:617–624 Pesole G, Lotti M, Alberghina L, Saccone C (1995) Evo-
Kent WJ, Baertsch R, Hinrichs A, Miller W, Haussler D lutionary origin of nonuniversal CUG(Ser) codon in
(2003) Evolution’s cauldron: duplication, deletion, and some Candida species as inferred from a molecular
rearrangement in the mouse and human genomes. phylogeny. Genetics 141:903–907
Proc Natl Acad Sci USA 100:11484–11489 Petersen RF, Nilsson-Tillgren T, Piskur J (1999) Karyotypes
Keogh RS, Seoighe C, Wolfe KH (1998) Evolution of gene of Saccharomyces sensu lato species. Int J Syst Bacteriol
order and chromosome number in Saccharomyces, 49:1925–1931
Kluyveromyces and related fungi. Yeast 14:443–457 Piskur J, Langkjaer RB (2004) Yeast genome sequencing:
Kinclova O, Potier S, Sychrova H (2001) The Zygosaccha- the power of comparative genomics. Mol Microbiol
romyces rouxii strain CBS732 contains only one copy of 53:381–389
the HOG1 and the SOD2 genes. J Biotechnol 88:151–158 Poch O, L’Hote H, Dallery V, Debeaux F, Fleer R, Sodoyer R
Koszul R, Caburet S, Dujon B, Fischer G (2003) Eucaryotic (1992) Sequence of the Kluyveromyces lactis beta-
genome evolution through the spontaneous duplica- galactosidase: comparison with prokaryotic enzymes
tion of large chromosomal segments. EMBO J 23:234– and secondary structure analysis. Gene 118:55–63
243 Robyr D, Suka Y, Xenarios I, Kurdistani SK, Wang A, Suka N,
Koufopanou V, Goddard MR, Burt A (2002) Adaptation for Grunstein M (2002) Microarray deacetylation maps
horizontal transfer in a homing endonuclease. Mol Biol determine genome-wide functions for yeast histone
Evol 19:239–246 deacetylases. Cell 109:437–446
Kurtzman CP, Fell JW (1998) The yeasts, a taxonomic study. Rokas A, Williams BL, King N, Carroll SB (2003) Genome-
Elsevier, Amsterdam scale approaches to resolving incongruence in molec-
Kurtzman CP, Robnett CJ (2003) Phylogenetic relationships ular phylogenies. Nature 425:798–804
among yeasts of the ‘Saccharomyces complex’ deter- Sanyal K, Baum M, Carbon J (2004) Centromeric DNA
mined from multigene sequence analyses. FEMS Yeast sequences in the pathogenic yeast Candida albicans
Res 3:417–432 are all different and unique. Proc Natl Acad Sci USA
Lafontaine I, Fischer G, Talla E, Dujon B (2004) Gene relics 101:11374–11379
in the genome of the yeast Saccharomyces cerevisiae. Scannell DR, Wolfe K (2004) Rewiring the transcriptional
Gene 335:1–17 regulatory circuits of cells. Genome Biol 5:206
Lai J, Ma J, Swigonova Z, Ramakrishna W, Linton E, Llaca V, Seoighe C, Wolfe KH (1999) Yeast genome evolution in the
Tanyolac B, Park Y-J, Jeong O-Y, Bennetzen JL, Mess- post-genome era. Curr Opin Microbiol 2:548–554
ing J (2004) Gene loss and movement in the maize Seoighe C, Federspiel N, Jones T, Hansen N, Bivolarovic V,
genome. Genome Res 14:1924–1931 Surzycki R, Tamse R, Komp C, Huizar L, Davis RW et al.
(2000) Prevalence of small inversions in yeast gene Wolfe K (2004) Evolutionary genomics: yeasts accelerate
order evolution. Proc Natl Acad Sci USA 97:14433– beyond BLAST. Curr Biol 14:R392–R394
14437 Wolfe KH, Shields DC (1997) Molecular evidence for an
Sor F, Fukuhara H (1989) Analysis of chromosomal DNA ancient duplication of the entire yeast genome. Nature
patterns of the genus Kluyveromyces. Yeast 5:1–10 387:708–713
Souciet J, Aigle M, Artiguenave F, Blandin G, Bolotin- Wong S, Wolfe KH (2005) Duplication of genes and
Fukuhara M, Bon E, Brottier P, Casaregola S, de genomes in yeasts. In: Sunnerhagen P, Piskur J (eds)
Montigny J, Dujon B et al. (2000) Genomic exploration Comparative genomics. Springer, Berlin Heidelberg
of the hemiascomycetous yeasts. 1. A set of yeast New York, Topics in Current Genetics (in press) DOI
species for molecular evolution studies. FEBS Lett 10.1007/b105770
487:3–12 Wong S, Butler G, Wolfe KH (2002) Gene order evolution
Tsong AE, Miller MG, Raisner RM, Johnson AD (2003) Evo- and paleopolyploidy in hemiascomycete yeasts. Proc
lution of a combinatorial transcriptional circuit: a case Natl Acad Sci USA 99:9272–9277
study in yeasts. Cell 115:389–399 Yochem J, Byers B (1987) Structural comparison of the yeast
Vaughan-Martini A, Martini A, Cardinali G (1993) cell division cycle gene CDC4 and a related pseudo-
Electrophoretic karyotyping as a taxonomic tool in gene. J Mol Biol 195:233–245
the genus Saccharomyces. Antonie van Leeuwenhoek Zhang Z, Dietrich FS (2003) Verification of a new gene
63:145–156 on Saccharomyces cerevisiae chromosome III. Yeast
Wieland J, Nitsche AM, Strayle J, Steiner H, Rudolph HK 20:731–738
(1995) The PMR2 gene cluster encodes functionally Zonneveld BJM, Steensma HY (2003) Mating, sporulation
distinct isoforms of a putative Na+ pump in the yeast and tetrad analysis in Kluyveromyces lactis. In: Wolf K,
plasma membrane. EMBO J 14:3870–3882 Breunig K, Barth G (eds) Non-conventional yeasts in
Wills C (1976) Production of yeast alcohol dehydrogenase genetics, biochemistry and biotechnology. Springer,
isoenzymes by selection. Nature 261:26–29 Berlin Heidelberg New York, pp 151–154
3 Investigating the Evolution of Fungal Virulence
by Functional Genomics
S. Ahmad1 , D.M. Soanes1 , M.C. Barooah1 , N.J. Talbot1
CONTENTS Fungi infect a multitude of plant species, de-

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 35 stroying harvests and causing immense economic
II. Identifying Virulence Factors damage to agriculture. Approximately 70% of ma-
in Plant Pathogenic Fungi . . . . . . . . . . . . . . . 36 jor crop diseases are caused by fungi, resulting in
III. The Challenge of Comparative Genomics . . . 38 crop losses which amount to 5%–10% of the to-
IV. Orthology and Paralogy . . . . . . . . . . . . . . . . 39
V. Comparative Genomic Analysis tal harvest of developed countries, and 40%–50%
of Pathogens and Saprotrophs . . . . . . . . . . . 40 in developing countries each year. This is in spite
VI. Identifying Pathogen-Conserved Genes of widespread modern farming practices, the use
Using Incomplete Genome Data . . . . . . . . . . 41 of fungicides, and extensive plant breeding pro-
VII. Exploring Gene Expression Patterns Using
EST Datasets . . . . . . . . . . . . . . . . . . . . . . . . . 42 grammes for resistance (Bowyer 1999). Although
VIII. Conclusions and Future Perspectives . . . . . . 44 intensive efforts have been directed towards disease
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 control, crop diseases still continue to threaten food
security. Historically, serious deprivation has re-
sulted directly from fungal and oomycete diseases.
Abbreviations: ATMT, A. tumefaciens-mediated The Irish potato famine of 1845, for example, was
transformation; COGs, clusters of orthologous caused by the oomycete pathogen Phytophthora in-
groups; EST, expressed sequence tag; RAPD, festans. More recently, a rice blast epidemic caused
random amplified polymorphic DNA; RFLP, by the fungus Magnaporthe grisea resulted in the
restriction fragment linked polymorphism; loss of 1090 tonne rice in Bhutan in 1995 (Thinlay
REMI, restriction endonuclease-mediated DNA et al. 2000).
integration; TAG-KO, transposon-arrayed gene The advent of genome sequencing efforts in
knockout filamentous fungi has provided a wealth of new in-
formation by which the characteristics of fungal
species can be investigated and compared. At the
I. Introduction
time of writing in early 2005, there were 26 com-
pleted fungal genome sequences publicly available,
Fungi are a morphologically diverse group of eu- ranging from unicellular yeasts, such as the model
karyotes, most of which have a saprophytic life style eukaryote Saccharomyces cerevisiae and many of
and degrade dead organic matter for their source its closest relatives, to morphologically complex,
of nutrients. The fungal life style is exemplified by multicellular basidiomycete fungi, such as Phane-
an ability to invade and colonise diverse substrates rochaete chrysosporium and Coprinus cinereus (Ta-
using cylindrical hyphal cells, and by the ability of ble 3.1). This vast amount of information has al-
these cells to secrete prodigious amounts of extra- lowed many new types of investigation into the evo-
cellular enzymes, with which complex polymeric lution of fungal virulence to be carried out (Yoder
substrates are degraded into simple sugars and and Turgeon 2001; Tunlid and Talbot 2002). In this
amino acids for subsequent uptake by the growing review, we examine the use of the newly generated
fungus. A small number of fungal species, however, fungal genome sequences and expressed sequence
have acquired the ability to colonise living plant tag collections of filamentous fungi as a resource
tissues and cause disease (Hammond-Kosack and for identifying conserved gene sequences, and as
Jones 1997). a means of learning more about the fascinating bi-
1School of Biological and Chemical Sciences, University of Exeter, ology of this group of organisms, with a particular
Washington Singer Laboratories, Perry Road, Exeter EX4 4QG, UK focus on the evolution of fungal pathogenicity.
The Mycota XIII
Fungal Genomics
36 S. Ahmad et al.
Table. 3.1. Completed genome sequences of fungi which have been publicly released (early 2005)
Fungal species Reference

Saccharomyces cerevisiae Goffeau et al. (1996)
Saccharomyces bayanus Kellis et al. (2003)
Saccharomyces kluyveri http://www.genome.wustl.edu/projects/yeast/
Saccharomyces mikatae Kellis et al. (2003)
Saccharomyces paradoxus Kellis et al. (2003)
Saccharomyces kudriavzevii http://www.genome.wustl.edu/projects/yeast/
Schizosaccharo myces pombe Wood et al. (2002)
Kluyveromyces waltii http://natchaug.labri.u-bordeaux.fr/Genolevures/
Kluyveromyces lactis http://natchaug.labri.u-bordeaux.fr/Genolevures/
Candida albicans http://www-sequence.stanford.edu/group/candida/
Candida glabrata http://genopole.pasteur.fr/glabrata/
Yarrowia lipolytica http://cbi.labri.fr/Genolevures/about.php
Neurospora crassa Galagan et al. (2003), http://www.broad.mit.edu/annotation/fungi/neurospora/
Aspergillus nidulans http://www.broad.mit.edu/annotation/fungi/aspergillus/
Fusarium graminearum http://www.broad.mit.edu/annotation/fungi/fusarium/
Magnaporthe grisea Dean et al. (2005), http://www.broad.mit.edu/annotation/fungi/magnaporthe/
Coccidioides immitis http://www.broad.mit.edu/annotation/fungi/coccidioides_immitis/
Rhizopus oryzae http://www.broad.mit.edu/annotation/fungi/fgi/
Debaromyces hansenii http://natchaug.labri.u-bordeaux.fr/Genolevures/
Encephalitozoon cuniculi http://www.genoscope.cns.fr/externe/English/Projets/Projet_AD/AD.html
Ashbya gossypii Dietrich et al. (2004)
Pichia stipitis http://www.jgi.doe.gov/sequencing/DOEmicrobes.html
Coprinus cinereus http://www.broad.mit.edu/annotation/fungi/coprinus_cinereus/
Ustilago maydis http://www.broad.mit.edu/annotation/fungi/ustilago_maydis/
Phanerochaete chrysosporium Martinez et al. (2004)
Cryptococcus neoformans http://www.broad.mit.edu/annotation/fungi/cryptococcus_neoformans
ever-greater resolution (Gold et al. 2001). The

II. Identifying Virulence Factors development of reporter gene technologies to
in Plant Pathogenic Fungi enable visualisation of spatial patterns of gene
expression during plant infection has also been an
In order to cause plant disease, many species of important advance. Green fluorescent protein from
filamentous fungi undergo a complex sequence of the jellyfish Aequoria victoriae is now widely used
metabolic and developmental events which enable to study gene expression and protein localisation
them to penetrate the host plant cuticle and invade during fungal development (Lorang et al. 2001).
living plant tissue (Tucker and Talbot 2001). These Immunological procedures to target particular
events include attachment of spores to the plant cellular targets have also helped to indicate the
surface, germination of propagules on the plant, role of a given protein in pathogenicity. Antibodies
development of infection structures which mediate have, for example, been used to inhibit cutinase,
penetration of the host and, finally, colonisation of lipases and other cell wall-degrading enzymes
host tissue. in order to assess their roles in plant infection
Recent advances in molecular genetic and ge- (Gold et al. 2001). In Botrytis cinerea, Commenil
nomic techniques have changed the way in which et al. (1998) have used anti-lipase antibodies to
fungal biology is studied and has contributed inhibit fungal colonisation. Immunolocalisation
greatly to our understanding of the strategies of pathogenicity factors is also commonly carried
employed by pathogens to initiate plant disease. At out (Tucker et al. 2004).
the cellular level, improved microscopic techniques Arguably, the most powerful approaches to
have helped to gain a better insight into the nature investigate fungal virulence, however, have been
of fungal infection structures which mediate gene functional studies, involving the generation
entry into plants (Howard 2001). Advances in of targeted deletion mutants, or large-scale for-
immunological labelling procedures, for example, ward genetic screens. Insertional mutagenesis
together with electron microscopy, have enabled has been widely utilised to identify pathogenicity
visualisation of plant–fungal interactions with mutants (for reviews, see Gold et al. 2001; Talbot
Investigating Fungal Virulence by Functional Genomics 37
and Foster 2001; Tucker and Talbot 2001). Restric- quencing templates using a bacterial transposon,
tion enzyme-mediated DNA integration (REMI) such as Tn5, which also serve as gene disruption
is a process whereby DNA is introduced into vectors for subsequent functional analysis of
a fungus in the presence of a restriction endonu- individual genes. The process has been used
clease which cleaves genomic DNA at numerous as a high-throughput means of identifying new
corresponding restriction sites in the genome, targets for fungicide discovery in both M. grisea
and also enhances the rate of DNA-mediated and the wheat blotch pathogen Mycosphaerella
transformation. This procedure has been used graminicola (Adachi et al. 2002).
extensively in the rice blast fungus M. grisea, Map-based cloning strategies have also been
and allowed characterisation of mutants which used to identify pathogenicity genes which were
display various defects such as non-functional previously identified using chemical mutagenesis
appressoria and non-invasive growth (Sweigard and classical genetics. This strategy uses RFLP and
et al. 1998; Balhadère et al. 1999). The PTH11 RAPD markers to identify mutant loci, and direct
gene, which encodes a novel G-protein coupled chromosome ‘walks’ using hybridisation from the
receptor required for appressorium development neighbouring, linked DNA markers (Nitta et al.
(DeZwaan et al. 1999), and the PDE1 gene which 1997). This approach was used, for example, to
encodes a P-type ATPase involved in appressorium clone AVR2-Pi-ta. PWL2, and AVR1-CO39 in M.
function (Balhadère and Talbot 2001) were both grisea (Orbach et al. 2000). The difficulty of carry-
identified in REMI screens. The Tox locus, which ing out conventional mutagenesis studies and the
is responsible for production of T-toxin and inability to use complementation cloning in most
high virulence on T-cytoplasm maize of the corn fungi to identify genes, however, was a principal
pathogen Cochliobolus heterostrophus, was also cause of the widespread use of reverse genetic ap-
tagged using REMI mutagenesis (Yang et al. 1996), proaches to identify pathogenicity genes.
which allowed characterisation of the polyketide Identifying genes in plant pathogenic fungi
synthase involved in T-toxin production (Yang based on their expression pattern has often
et al. 1996). More recently, fungal transformation provided the first stage of such studies. Differ-
mediated through the plant pathogenic bacterium ential cDNA screening involves generating cDNA
Agrobacterium tumefaciens has been used to trans- libraries, either derived from a particular growth
fer foreign genes into several fungal species and to stage of the pathogen which is observed during
carry out gene replacements (Bundock et al. 1995; plant growth, or by exposing the organism to
de Groot et al. 1998). Problems associated with environmental conditions which favour expression
protoplast isolation in Fusarium oxysporum have of particular genes, such as starvation stress
been alleviated using A. tumefaciens-mediated (Talbot et al. 1993). Differential cDNA screening
transformation (ATMT; Mullins et al. 2001), and in M. grisea led, for example, to the identification
the high frequency of transformation achieved of MPG1, a hydrophobin-encoding gene found
by ATMT has allowed rapid generation of large to be important in the development of spores
collections of insertional mutants in a variety and appressoria (Talbot et al. 1993, 1996). The
of fungal pathogens, including most notably GAS1 and GAS2 genes of M. grisea, which are
M. grisea. The fact that disrupted loci can be also highly similar to genes expressed using plant
isolated readily by the inverse polymerase chain infection by the obligate biotrophic pathogen
reaction (PCR) or ‘step-down’ PCR (Zhang and Blumeria graminis, were also identified in this way
Gurr 2000) means that such libraries are an (Xue et al. 2002). A purely deductive approach
extremely valuable resource for pathogenicity has also been used as a strategy for identifying
gene discovery. virulence determinants in plant pathogenic fungi.
Another method for genome-wide mutage- Here, a key process is recognised, based on
nesis studies in filamentous fungi, pioneered in cell physiology or cell biology studies, which is
the private sector, has been reported by Hamer predicted to influence fungal pathogenicity, and
et al. (2001). A transposon-arrayed gene knock- then systematically tested using targeted gene
out (TAG-KO) strategy has been developed to replacement. The prevalence of melanin in the
sequence fungal genomes in a rapid manner and appressorium cell wall of M. grisea, for instance,
to simultaneously create gene disruption cassettes led to targeting of the dihydroxy-naphthalene
for subsequent fungal transformation and mutant (DHN) melanin biosynthetic pathway using
analysis. The process works by generation of se- standard mutagenesis. The application of this
38 S. Ahmad et al.
strategy led ultimately to the isolation of three necessitated the development of new bioinfor-
genes, ALB1, RSY1 and BUF1, which were found matic algorithms to investigate gene function and
to be required for the melanisation of appressoria genome organisation. These include programmes
and essential for pathogenicity (Howard 1997). to search for regions of similarity between two
Similar genes were found to be essential for ap- or more sequences (Altschul et al. 1990, 1997),
pressorium function in the anthracnose pathogen to identify open reading frames (ORFs) and
Colletotrichum lagenarium which also produces regulatory domains within genomic data, or
melanin-pigmented infection structures (Kubo functional motifs and domains in protein-coding
et al. 1991, 1996). Genes involved in virulence- sequences (see http://us.expasy.org/alinks.html).
associated cell signalling have also predominantly To be used effectively, genome sequence data
been identified in pathogenic species based on need to be stored and presented in a form which
homology to genes in S. cerevisiae, and then tested can be readily accessed, queried and compared.
empirically to determine their role in pathogen- The sequence of a single genome is, of course,
esis. For example, the MAP kinase genes UBC3 informative in itself, but its value is greatly
and PMK1 from Ustilago maydis and M. grisea enhanced by comparison with other genomes.
respectively were identified as functional homo- This can assist in the prediction and annotation of
logues of the yeast FUS3/KSS1 MAPK-encoding individual genes, and allow us to relate differences
genes (Xu and Hamer 1996; Mayorga and Gold between species to their gene inventories. Such
1999). information is indicative of how evolutionary
In all of the reverse genetic strategies de- lineages have diverged from a common ancestor,
scribed above, it has been the advancement of and provide clues regarding how adaptation
DNA-mediated transformation techniques for to a particular niche or situation has occurred
phytopathogenic fungi which has permitted the in a given organism (Mushegian and Koonin
testing of putative pathogenicity determinants 1996; Wei et al. 2002; Koonin 2003; Nobrega and
(Mullins and Kang 2001). Transformation fre- Pennacchio 2004).
quencies and, indeed, the rates of homologous The first eukaryotic genome sequence, gen-
recombination which lead to gene disruption vary erated in 1996, was that of the ascomycete yeast
greatly among species, but have so far been used to S. cerevisiae, but it was not until 2002 that large-
identify and validate approximately 70 virulence scale sequencing of fungal genomes was under-
genes in a wide range of phytopathogenic species taken in earnest. As such, until that point, compar-
(Idnurm and Howlett 2001). ative genomics in filamentous fungi was restricted
The availability of full genome sequences for to the analysis of incomplete datasets, including
a variety of phytopathogenic fungi, including Mag- most notably the use of collections of expressed
naporthe grisea, Ustilago maydis and Fusarium sequence tags (ESTs), the uses of which are dis-
graminearum has, however, recently changed the cussed in greater detail in Section VI.. The gen-
research perspective from analysis of the functions eration of DNA sequence information from Neu-
of individual genes in fungal pathogenicity to rospora crassa (Galagan et al. 2003), Phanerochaete
investigation of the orchestrated action of large chrysosporium (Martinez et al. 2004) and Schizosac-
sets of genes during plant infection. charomyces pombe (Wood et al. 2002) and a num-
ber of other species (see Table 3.1) has for the first
time allowed comparative genomic analysis to be
carried out in diverse fungal species. Of particular
III. The Challenge of Comparative importance in these studies has been the analysis
Genomics of gene functions in fungi which are conserved in
other species and yet are specific to the kingdom
The first years of the 21st century have seen Fungi, or those which are conserved among eco-
the generation of a large number of genome logically related groups of fungal species such as
sequences, including those of the major model pathogenic organisms. However, in order to un-
eukaryotic organisms Arabidopsis thaliana, derstand the relationships between such genes, it
Drosophila melanogaster, Mus musculus, Rattus is important first to be able to develop informatic
norvegicus, the mosquito Anopheles gambiae and resources to attempt to predict whether genes fulfil
the human genome (see Brown 2002). The vast similar functions, or have diverged to fulfil distinct
amount of data generated by these projects has biological roles.
IV. Orthology and Paralogy of sequence differences. Protein structural data

and protein engineering, however, indicated that
only a few amino acid substitutions were involved
Homologous genes are described as those which in substrate specificity determination and could
are descended from a common ancestral gene. therefore be considered reliable markers for
Homologues were further classified by Fitch determining orthology or paralogy. In the light
(1970) into orthologous and paralogous genes. of this information, the authors re-examined
Orthologous genes are defined as those resulting sequences within the main databases which were
from a speciation event, such that the gene history annotated as β-decarboxylating dehydrogenases,
reflects the evolutionary history of the species in and reassigned a functional annotation for 26
which it is found. Paralogous genes result from proteins (Chen and Jeong 2000).
a gene duplication event, and so are homologous Notwithstanding these problems, to make
members of the same gene family which have comparisons between the gene inventories of
descended side by side during an organism’s his- different genomes it has become necessary to
tory. These phrases are frequently used (Ouzounis develop means of identifying, in particular,
et al. 2003), although not always with the same orthologous genes, based only on sequence data.
meaning. An important point of contention, for Various methods are now used, most of which
example, is whether the relationship between are based on searching for gene sequences which
genes is indicative of their biological function. It have a certain level of sequence similarity over
is often assumed that orthologous genes carry out a given proportion of their lengths. Tatusov et al.
similar functions and that paralogues differentiate (1996) searched for orthologous gene pairs in
to perform different functions (Tatusov et al. the Haemophilus influenzae bacterial genome
1996). However, this need not be true (Huynen and 75% of the Escherichia coli genome. They
et al. 1998; Gogarten and Olendzenski 1999) and defined an E. coli gene as an orthologue if it
cannot be assumed (Theiben 2002). In spite of had a greater similarity, by at least several per-
these definitions, genes are often referred to as centage points, to a H. influenzae gene than to
orthologues, meaning functionally equivalent any other E. coli gene. This analysis identified
genes, without any reference to speciation. Dif- 1128 pairs of apparent orthologues which had
ferentiating between orthologues and paralogues identities varying in the range 18%–98%, with
in genome sequences is valuable for identifying an average value of 59%. Approximately 50%
equivalent genes in different organisms, and of E. coli genes were identified as belonging to
also provides a means of investigating how the clusters of paralogues. Tatusov and colleagues
evolution of particular genes has occurred. This (1997) recognised that divergence followed by
process is difficult, however, because orthologous duplication can lead to a tree of genes derived
genes exhibit a range of evolutionary rates, and from the original ancestor; which they called
therefore the degree of sequence similarity is clusters of orthologous groups (COGs). Subse-
not sufficient to distinguish orthologues from quently, individual orthologues, or orthologous
paralogues, because orthologous genes do not groups of paralogues, from three or more
always show higher levels of sequence identity phylogenetic lineages were used to construct
than do paralogues (Grishin et al. 2000; Hedges a COGs database of proteins from sequenced
and Kumar 2003). For example, protein sequences genomes, as a tool for functional annotation of
of conserved proteins such as ubiquitins or novel or unknown proteins (Tatusov et al. 2000,
histones in eukaryotes are typically 90%–98% 2001).
identical, whereas the sequences of dihydro- Phylogenomics is another method which uses
orotases (pyrimidine metabolic enzymes) are only orthology detection as a means of predicting con-
typically 20%–30% identical, even when func- served gene function (Eisen 1998). Homologues
tionally equivalent. An example of the difficulty of a gene in question are identified and any re-
inherent in such studies is the report by Chen gions of ambiguous homology are masked. Se-
and Jeong (2000). A family of β-decarboxylating quences are then aligned, and a phylogenetic tree
dehydrogenases was examined. Phylogenetic constructed. The tree is then used to identify possi-
analysis indicated that genes encoding this family ble gene duplication events. In this way, Mushegian
of enzymes diverged from a common ancestral et al. (1998) compared orthologous genes of the
gene and had since accumulated large numbers human, D. melanogaster, C. elegans and S. cere-
40 S. Ahmad et al.
visae genomes. They began by searching for ho- With these more complex and distant gene
mologous genes from each of the genomes. Their relationships, changes in gene function may
criteria for inferring orthology were that the se- occur, or putatively orthologous genes may be lost
quences which showed the highest level of homol- (Jensen 2001). New terms such as co-orthologue
ogy to one another (a ‘reciprocal best hit’ test) and (Gates et al. 1999) and semi-orthologue (Sharman
showed similarity through the entire length of both 1999) have been suggested to describe the relation-
genes were likely to be orthologous. This resulted ships among duplicated gene and their single-copy
in 42 sets of putative orthologous groups of genes. orthologues or pro-orthologues. Using these com-
The sequences thus identified were aligned and putational methods, genome searches have re-
trees constructed to detect phylogenetic relation- vealed extensive gene conservation between phylo-
ships between the species. However, Xie and Ding genetically distinct organisms. For example, 90%
(2000) re-examined these results and commented of rat genes have orthologues in both the mouse
that using such a reciprocal best hit method can de- and human genomes (Rat Genome Sequencing
tect one-to-one orthologous relationships, but can Project Consortium 2004). However, systematic
miss one-to-many or many-to-many relationships differences between whole sets of genes can also
where a gene in one species has more than one or- be apparent. Homologues of S. cereviseae genes
thologue in another species. They also noted that were found to be on average almost 15% shorter
where there are these more distant or complex re- in the microsporidian parasite Encephalitozoon
lationships, orthologues might not retain the same cuniculi, which was proposed to reflect reduced
function. To overcome this limitation, they used protein–protein interactions in this organism
a more rigorous phylogenetic analysis, using a rec- (Katinka et al. 2001).
onciled tree method to identify more complex gene
relationships (Xie and Ding 2000).
A different approach was carried out by Bansal
(1999) who used a bipartite graph matching V. Comparative Genomic Analysis
and fuzzy-logic procedures for orthologue and of Pathogens and Saprotrophs
paralogue differentiation. The process still used
steps similar to those of the other methods, in The availability of full genome sequence informa-
that it involves homology searches to identify tion for a range of saprotrophic and pathogenic
putative gene pairs, then uses the Smith-Waterman fungal species allows general questions about the
algorithm to determine the regions of homology evolution of fungal virulence to be asked for the
between the genes. The bipartite graph matching first time. It is not yet known how genome structure
was then used to group the gene pairs (Bansal and gene inventories have been modified in fungi
1999). during evolution of species to fulfil specific ecolog-
There are currently a number of shortcom- ical niches. There are three mutually non-exclusive
ings in all the available computational methods theories which have been proposed to explain the
described above. Where the species in question evolution of virulence in fungi (Tunlid and Talbot
are only distantly related, orthologous gene pairs 2002).
may have diverged so much, for example, that it
– First, pathogenic fungi may have acquired sets
is not possible to accurately identify those rela-
of specific pathogenicity genes which are con-
tionships where the level of similarity may not be
served among pathogens but are not present in
different to that of unrelated genes. Another rea-
closely related free-living organisms.
son why similarity searches may not reveal func-
– Second, gene inventories in pathogens and
tional relationships is when a particular cellular
non-pathogens may be fairly universal, but
function has been undertaken by the product of
selection has occurred on regulatory genes
a non-orthologous gene. In a comparison of the
such that pathogenicity is a consequence of
E. coli and H. influenzae genomes, Koonin et al.
particular spatio-temporal control of gene
(1996) identified 12 cases where essential cellular
expression.
functions were encoded by non-orthologous genes.
– Finally, pathogenicity may be the result of gene
In addition, there need not be simple one-to-one
loss.
relationships between orthologues (Xie and Ding
2000) – a gene may be orthologous to a number of The study of bacterial pathogens has shown exam-
other genes in another genome. ples of each of these events in particular pathogens
and it is clear, for example, that horizontal transfer

VI. Identifying Pathogen-Conserved
of genes among pathogens has taken place (Wren
2000; Ochman et al. 2000). In eukaryotic organisms Genes Using Incomplete Genome
such as fungi, there is much less evidence of such Data
events (Rosewich and Kistler 2000).
So far, there have been few systematic studies to In the absence of complete genomic sequence, sin-
address whether pathogenic fungi possess unique gle pass, partial sequencing of either 3 or 5 ends
genes which are conserved only in pathogenic of complementary DNA (cDNA) clones to generate
species. To carry out this type of analysis requires a set of expressed sequence tags (ESTs), offers
whole-genome homology searches, followed by a low-cost strategy to identify gene inventories.
a phylogenetic study to determine the evolutionary EST datasets are now available for a large number
relationships between putatively conserved genes, of phytopathogenic fungi, but generally in a flat
in order to infer orthology, or at least functional file format with limited annotation. The largest
relatedness. Once genes have been identified in collection of EST data from phytopathogenic
this way, rigorous experimental analysis using fungi is housed at the consortium for Functional
targeted deletion in different pathogen species Genomics of Microbial Eukaryotes (COGEME)
and cross-species complementation experiments database, which contains 54,083 unique sequences
would be required to confirm the presence of from 18 species of fungi and oomycetes (Soanes
such conserved gene functions. The bioinformatic et al. 2002). In the COGEME database, individual
component of such a project is underway using EST sequences were clustered to produce a set
the M. grisea genome and genome and EST of consensus sequences (unisequences), to which
information from 15 other pathogenic fungal putative functions were assigned based on simi-
species including Mycosphaerella graminicola, larity to sequences of known genes. The BlastX
Botrytis cinerea, F. graminearum, U. maydis and algorithm (Altschul et al. 1990) was used to query
the human pathogens Candida albicans, As- the NCBI non-redundant protein database, and the
pergillus fumigatus and Cryptococcus neoformans. top five most similar sequences (with expectation
When compared to genome information from S. values less than 1 ×10−5 ) were retrieved for each
cerevisiae, N. crassa, Schizosaccharomyces pombe unisequence. The cut-off value selected was more
and Aspergillus nidulans, a set of 65 genes was rigorous than the recommended value of 10−2
found to be unique to the pathogenic species, (Anderson and Brass 1998), below which matches
with some individual genes conserved in up to are considered significant in 98% of cases. On the
five pathogenic species (S. Ahmad and N.J. Talbot, basis of these similarity scores, a putative product
unpublished information). Some of these genes or function was assigned to each unisequence,
are now being functionally analysed. but this remains highly speculative in some cases.
Whole-genome comparisons between two tax- Based on the assignments, the unisequences were
onomically related fungal species such as M. grisea classified by function according to a hierarchical
and N. crassa have also proved revealing in de- scheme used by MIPS (Munich Information
termining the characteristics of pathogenic fungi Centre for Protein Sequences; Mewes et al. 2004).
which may distinguish them from saprotrophic This scheme groups gene products depending
species (Dean et al. 2005). The rice blast fungus on the particular metabolic pathway or cellular
appears to posses approximately 10% more genes process which they are predicted to be involved
than does the saprotrophic M. grisea, including far in. Approximately 44% of the unisequences in
more genes which are predicted to encode secreted the COGEME database had no homology to
proteins. In addition, there have been episodes sequences present in publicly available databases.
of gene family expansion or maintenance in M. Such genes are often termed orphans (ORFs of
grisea. This analysis is constrained, however, by no known function) and are commonly found in
the influence which repeat-induced point mutation the genomes of eukaryotic organisms. Orphan
(RIP) has had on the genome of N. crassa (Galagan genes may represent genes whose phylogenetic
et al. 2003). RIP is a process by which duplicated distribution is restricted to certain evolutionary
genes are removed at the onset of meiosis, result- lineages or genes which rapidly diverge between
ing in little opportunity for paralogous duplication closely related species (Siew and Fischer 2004).
and gene family expansion (Galagan et al. 2003; A comparison of ESTs from N. crassa with the S.
Borkovitch et al. 2004). cerevisiae genome indicated that there is a higher
42 S. Ahmad et al.
proportion of orphans in the N. crassa genome The main limitation in the analysis reported
than in yeast (Braun et al. 2000). This may be by Giles et al. (2003) results from the nature of
due to the increased morphological complexity of the partial EST datasets stored in the COGEME
the multicellular N. crassa when compared with database. These datasets are highly unlikely to
S. cerevisiae, which is also far more intensively meet the assumption required for validity of
studied. However, it will be interesting in due the probability calculations – that the data are
course to investigate more closely related fungal drawn randomly from the entire genome. This
species to determine whether the orphans in N. is because the cDNA libraries from which the
crassa include rapidly evolving genes. ESTs are derived are from different stages of
Although the gene inventories contained fungal development. The unisequence sets created
in the COGEME EST database represent only by clustering EST sequences do, however, have
a fraction of the transcriptome of each fungal less sequence redundancy and may more closely
species, careful analysis of the genes represented meet the assumption the probability calculation
in these collections can give valuable information is based on. Notwithstanding this limitation,
about metabolic pathways which may be present the probability values yield information with
in different species of phytopathogenic fungi. the potential to provide early indications of the
A recent example of such a study examined absence or presence of a particular amino acid
amino acid biosynthesis genes within a group of pathway in an organism of interest. The data, for
fungi, including phytopathogenic species such example, provided evidence for conservation of
as M. grisea, My. graminicola, B. cinerea, F. amino acid biosynthesis pathways in the Barley
graminearum and the barley powdery mildew powdery mildew fungus Blumeria graminis.
fungus Blumeria graminis (Giles et al. 2003). Since this fungus is an obligately pathogenic
A relational database of information regarding species which can grow only on its host plan
amino acid biosynthetic pathways for each fungal (Kobayashi et al. 1991; Giese 1997), a requirement
species was generated and can be viewed at for amino acid biosynthesis might be predicted to
http://cogeme.ex.ac.uk/biosynthesis.html. The da- be dispensable. The data presented by Giles et al.
tabase was developed as a demonstration tool to (2003) suggest that this is not the case and that
show how pathway information can be gained from the organism possesses amino acid biosynthetic
incomplete EST data. Amino acid biosynthesis genes in the same way as a free-living saprotroph
pathways are reasonably well understood and or a facultative pathogen. The pathway prediction
documented (Braus 1991; Fritz 1997; Thomas and tool is adaptable and may of broader utility in
Surdin-Kerjan 1997). A relational database, with determining the presence of other metabolic
a web-interface offering analytic functions, was or signal transduction pathways in organisms
therefore developed to exploit the available fungal for which only partial gene inventories are
genomic data with a view to investigating the con- available.
servation of amino acid biosynthesis pathways of
several pathogenic species. To determine whether
an amino acid biosynthetic pathway was likely to
be conserved within a particular fungal species VII. Exploring Gene Expression
based on the presence of a given number of EST Patterns Using EST Datasets
sequences, Giles et al. (2003) utilised a Bayesian
probability metric, which is given below. When ESTs are generated randomly from non-
normalised cDNA libraries, the frequency of
a given EST is proportional to the abundance of
n G! (N − G)! (N − n)! the mRNA transcript corresponding to this gene,
, in the tissue from which the cDNA library was
g (G − g)! (N − G − n + g)! N!
constructed (Audic and Claverie 1997). If ESTs
are available which have been sequenced from
where N is the number of genes in the genome a number of cDNA libraries, then consequently
(assumed to be 10,000), G is the number of enzymes an expression profile can be constructed for each
in the pathway, n is the number of genes in the gene, showing EST frequency in libraries con-
sample dataset and g is the number of pathway structed from a variety of tissues, or experimental
enzymes in the sample dataset. conditions. The sequencing of ESTs from a cDNA
Fig. 3.1. Screen shots from the web-based front of the CO- library, as well as the percentage of the total ESTs sequenced
GEME fungal EST database (http://cogeme.ex.ac.uk/tran from each library which this represents. B This matrix shows
script.html) showing a transcript profile for the gene which P-values representing the probability that differences in EST
encodes the fungal hydrophobin MPG1 (Talbot et al. 1993, abundance in pair-wise comparisons between each library
1996). A Output from the database shows the number of are due to differential expression of the gene, rather than
ESTs representing this unigene sequenced from each cDNA chance sampling error
44 S. Ahmad et al.
data). A comparison of ESTs sequenced from

two cDNA libraries was then carried out to
identify which putative genes are differentially
regulated between the two conditions or tissue
types from which the libraries were constructed,
as shown in Fig. 3.1. This information is available
at http://cogeme.ex.ac.uk/transcript.html. A sta-
tistical analysis method, developed to identify
whether the difference in EST sequence abundance
between two libraries is significant (Audic and
Claverie 1997), was used to test the robustness
of the results scored. This test is based on the
Fig. 3.2. Screen shots from the web-based front of the CO- probability that the difference in frequency of ESTs
GEME fungal EST database showing a comparison of EST
transcript profiles between cDNA libraries which are de-
between two cDNA libraries is due to random
rived from mRNA extracted from appressoria of a rice sampling, rather than to a difference in expression
pathogenic strain of Magnaporthe grisea, 70-15 (Mag02), levels (Fig. 3.2). The theory links the threshold
or hyphae cultured in standard minimal growth medium of selection of putatively regulated genes to the
(Mag03). A The researcher can use this screen to specify fraction of false positive clones one is willing to
which pair of libraries to compare and to set the minimum
P-value for consideration of a difference in EST frequency risk. This statistical method was used to produce
as significant (0.99 or 0.95). B List of unigenes which are a P-value for each pair of cDNA libraries for each
significantly more highly expressed in one library than in unigene. The P-value indicates the probability
the other, split into those which are more highly expressed that differences in the number of EST transcripts,
in Mag02 (top) and those which are more highly expressed
in Mag03 (bottom)
representing each unigene, sequenced from the
two libraries are due to differential expression
of the gene, rather than chance sampling error
(Audic and Claverie 1997).
library is, however, also a random event and
therefore the frequency of sequencing a particular
cDNA clone is subject to fluctuations due to
chance sampling error. Consequently, the data VIII. Conclusions
produced in this way are inherently noisy, and and Future Perspectives
more reliable results are produced only when large
numbers of ESTs are sequenced from each library The emergence of functional genomics has
(Ewing et al. 1999). The presence of large EST introduced a new era in the study of fungal
collections for a variety of phytopathogenic fungi pathogenicity (Yoder and Turgeon 2001). As
provides an opportunity to carry out this form of illustrated above, genomic technologies allow us to
gene expression analysis, which is of particular acquire vast datasets which have obvious relevance
value where resources such a microarrays are to pathogenesis, but also provide challenges if
not yet available. An example of the strategy they are to be analysed effectively. Molecular
has been carried out with 958 unique genes phylogeny has shown that pathogenic fungi are
(unisequences), which were well-represented by found in many taxonomic groups (including
ESTs, from eight cDNA libraries of M. grisea basidiomycetes and ascomycetes), which suggests
derived from appressoria, conidia and mycelia that the pathogenic life style has evolved multiple
grown either in complete medium, minimal times within the fungal kingdom. There are several
medium, nitrogen starvation medium or with experimental studies which have demonstrated
rice cell walls as the sole carbon source. This that parasitic fungi have unique pathogenicity
was done with wild-type strains, as well as with factors (as reviewed by Idnurm and Howlett 2001).
germinated conidia from a null mutant for the There are fewer examples, however, of large-scale
MAP kinase-encoding gene PMK1 (Xu and Hamer systematic screens for such genes from genome
1996) and a mated culture library constructed sequences, which provides a rich opportunity for
from a mixture of asci, ascospores, perithecia future research. The alternative explanation, in
and mycelium produced by crossing two different which pathogenesis is due to specific differences
strains (D.M. Soanes and N.J. Talbot, unpublished in gene regulation, is exemplified by the presence
B
Compar ison of tr anscr ipt pr ofiles of Mag02 and Mag03
Sequences significantly mor e highly expr essed (P >= 0.99) in Mag02 than Mag03
Unigene ID Putative pr oduct / function Mag021 Mag031 P value2

DNMag0335 homologue of UVI-1 [Bipolaris oryzae], cell wall protein 30 0 0.999
DNMag0371 MAS3 protein 20 0 0.999
DNMag0391 MPG1 class I hydrophobin 35 0 0.999
DNMag0623 unknown 13 0 0.996
Sequences significantly mor e highly expr essed (P >= 0.99) in Mag03 than Mag02
Unigene ID Putative pr oduct / function Mag021 Mag031 P value2

DNMag0052 60S acidic ribosomal protein P2 0 7 0.999
DNMag0054 60S ribosomal protein L10 0 7 0.999
DNMag0166 calcineurin subunit B 0 9 0.999
DNMag0208 coproporphyrinogen oxidase, sixth step in heme biosynthetic pathway 0 7 0.999
DNMag0232 D-hydroxyacid dehydrogenase / D-lactate dehydrogenase 0 10 0.999
DNMag0293 glyceraldehyde 3-phosphate dehydrogenase 3 21 0.999
DNMag0438 pathogenesis related (SnodProt1) 0 8 0.999
DNMag0260 extracellular chitinase 0 6 0.998
DNMag0560 translation elongation factor-1 gamma (EF-1-gamma) 0 6 0.998
DNMag0023 40S ribosomal protein S12 0 5 0.995
DNMag0107 adenosine kinase 0 5 0.995
DNMag0249 enolase (2-phosphoglycerate dehydratase) 0 5 0.995
DNMag0573 ubiquinol-cytochrome c reductase 0 5 0.995
DNMag0036 40S ribosomal protein S25 1 6 0.99
1 Number of ESTs representing each unigene sequenced from each cDNA library.
2 he P value indicates the probability that differences in the number of EST transcripts, representing each unigene, sequenced from the two
libraries are due to differential expression of the gene rather than chance sampling error. (Audic and Claverie, 1997).
Fig. 3.2. (continued)
46 S. Ahmad et al.
of conserved elements of signaling pathways for Acknowledgements. Work on comparative genomics of

infection-related development in diverse phy- fungi in our group is supported by the Biotechnology
topathogenic species. MAP kinase genes similar to and Biological Sciences Research council (BBSRC) grant
BBS/B/17174.
the PMK1-encoded MAP kinase gene of M. grisea
are present in a number of other plant pathogenic
fungi, including Botrytis cinerea (Zheng et al. References
2000), Colletotrichum lagenarium (Takano et al.
2000), Cochliobolus heterostrophus (Lev et al. Adachi K, Nelson GH, Peoples KA, Frank SA, Montenegro-
1999), Fusarium oxysporum (Di Pietro et al. 2001), Chamorro MV, DeZwaan TM, Ramamurthy L,
Pyrenophora teres (Ruiz-Roldan et al. 2001) and Shuster JR, Hamer L, Tanzer MM (2002) Efficient
gene identification and targeted gene disruption in
Ustilago maydis (Mayorga and Gold 1999; Muller the wheat blotch fungus Mycosphaerella graminicola
et al. 1999). In all cases, the corresponding MAPK using TAGKO. Curr Genet 42:123–127
genes have been disrupted and mutant strains Altschul S, Gish W, Miller W, Myers EW, Lipman DJ (1990)
found to be defective in pathogenicity. The extent Basic local alignment search tool. J Mol Biol 215:403–
410
to which conservation of the effector molecules Altschul S, Madden TL, Schaffer AA, Zhang J, Zhang A,
which act downstream of these pathways are Miller W, Lipman DJ (1997) Gapped BLAST and PSI-
conserved in pathogen genomes is, however, not BLAST: a new generation of protein database search
yet clear. programs. Nucleic Acids Res 25:3389–3402
Widespread genome sequencing of filamen- Anderson I, Brass A (1998) Searching DNA databases for
similarities to DNA sequences: when is a match signif-
tous fungi, including phytopathogenic species icant? Bioinformatics 14:349–356
such as M. grisea, U. maydis and F. graminearum, is Audic S, Claverie JM (1997) The significance of digital gene
already providing key insights into the pathogenic expression profiles. Genome Res 7:986–995
life style, including the presence of a complex Balhadère PV, Talbot NJ (2001) PDE1 encodes a P-type AT-
Pase involved in appressorium-mediated plant infec-
secreted proteome, and gene family expansions tion by the rice blast fungus Magnaporthe grisea. Plant
in genes associated with environmental per- Cell 13:1987–2004
ception and cell signalling (Dean et al. 2005). Balhadère PV, Foster AJ, Talbot NJ (1999) Identification of
The contrast and comparison with genomes of pathogenicity mutants of the rice blast fungus Magna-
closely related saprotrophic fungi such as A. porthe grisea by insertional mutagenesis. Mol Plant-
Microbe Interac 12:129–142
nidulans, N. crassa and C. cinereus will prove Bansal AK (1999) An automated comparative analysis of 17
very illuminating. One of the most significant complete microbial genomes. Bioinformatics 15:900–
informatics challenges will be in predicting 908
whether conservation or divergence of particular Borkovich KA, Alex LA, Yarden O, Freitag M, Turner GE,
Read ND, Seiler S, Bell-Pedersen D, Paietta J, Plesof-
biological functions has occurred between these sky N et al. (2004) Lessons from the genome sequence
fungal species. The availability of EST collections of Neurospora crassa: tracing the path from genomic
from a much larger group of phytopathogenic blueprint to multicellular organism. Microbiol Mol
species is also important, not only in carrying out Biol Rev 68:1–108
comparative genomics but also for examination Bowyer P (1999) Plant diseases caused by fungi: phy-
topathogenicity. In: Oliver RP, Schweizer M (eds)
of alternative splicing, gene expression analysis, Molecular fungal biology. Cambridge University
genome annotation, and for determining whether Press, pp 294–321
microRNA-mediated gene regulation occurs Braun EL, Halpern AL, Nelson MA, Natvig DO (2000)
in filamentous fungi as in other multicellular Large-scale comparison of fungal sequence informa-
tion: mechanisms of innovation in Neurospora crassa
eukaryotes. and gene loss in Saccharomyces cerevisiae. Genome
Finally, a large effort is still required to provide Res 10:416–430
better mechanisms for gene functional analysis in Braus GH (1991) Aromatic amino acid biosynthesis in the
filamentous fungi in order that the genes predicted yeast Saccharomyces cerevisiae: a model system for
to serve roles in pathogenesis can be readily tested the regulation of a eukaryotic biosynthetic pathway.
Microbiol Rev 55:349–370
for such a role, and the interplay between gene Brown TA (2002) Genomes 2. Bios Scientific, Ox-
activities then investigated in detail. Drawing to- ford Bundock P, Dendulkras A, Beijersbergen A,
gether these disparate, computational and empiri- Hooykaas PJJ (1995) Trans-kingdom T-DNA transfer
cally derived pieces of information into a cohesive from Agrobacterium-tumefaciens to Saccharomyces
cerevisiae. EMBO J 14:3206–3214
and plausible model for the development of plant Chen R, Jeong SS (2000) Functional prediction: identifi-
disease by fungi will be a key challenge for the next cation of protein orthologs and paralogs. Protein Sci
years. 9:2344–2353
Commenil P, Belingheri L, Dehorter B (1998) Antilipase an- Grishin NV, Wolf YI, Koonin EV (2000) From complete
tibodies prevent infection of tomato leaves by Botrytis genomes to measures of substitution rate variability
cinerea. Physiol Mol Plant Pathol 52:1–14 within and between proteins. Genome Res 10:991–1000
Dean RA, Talbot NJ, Ebbole DJ, Farman ML, Mitchell T, Hamer L, Adachi K, Montenegro-Chamorro MV,
Orbach M, Thon M, Kulkarni R, Xu JR, Pan H et al. Tanzer MM, Mahanty SK, Lo C, Tarpey RW,
(2005) The genome sequence of the rice blast fungus Skalchunes AR, Heiniger RW, Frank SA et al. (2001)
Magnaporthe grisea. Nature 434:980–986 Gene discovery and gene function assignment in fila-
de Groot MJA, Bundock P, Hooykaas PJJ, Beijers- mentous fungi. Proc Natl Acad Sci USA 98:5110–5115
bergen AGM (1998) Agrobacterium tumefaciens- Hammond-Kosack KE, Jones JDG (1997) Plant disease re-
mediated transformation of filamentous fungi. Nature sistance genes. Annu Rev Plant Physiol Plant Mol Biol
Biotechnol 16:839–842 48:575–607
DeZwaan TM, Carroll AM, Valent B and Sweigard JA (1999) Hedges SB, Kumar S (2003) Genomic clocks and evolution-
Magnaporthe grisea pth11p is a novel plasma mem- ary timescales. Trends Genet 19:200–206
brane protein that mediates appressorium differentia- Howard RJ (1997) Breaching the outer barrier – cuticle and
tion in response to inductive substrate cues. Plant Cell cell wall penetration. In: Carroll GC, Tudzynski P (eds)
11:2013–2030 The Mycota, vol V. Plant relationships, part A. Springer,
Dietrich FS, Voegeli S, Brachat S, Lerch A, Gates K, Steiner S, Berlin Heidelberg New York, pp 43–60
Mohr C, Pohlmann R, Luedi P, Choi S et al. (2004) Howard RJ (2001) Cytology of fungal pathogens and plant-
The Ashbya gossypii genome as a tool for mapping host interactions. Curr Opin Microbiol 4:365–373
the ancient Saccharomyces cerevisiae genome. Science Huynen M, Dandekar T, Bork P (1998) Differential genome
304:304–307 analysis applied to the species-specific features of He-
Di Pietro A, Garcia-Maceira FI, Meglecz E, Roncero MIG licobacter pylor. FEBS Lett 426:1–5
(2001) A MAP kinase of the vascular wilt fungus Fusar- Idnurm A, Howlett BJ (2001) Pathogenicity genes of phy-
ium oxysporum is essential for root penetration and topathogenic fungi. Mol Plant Pathol 2:241–255
pathogenesis. Mol Microbiol 39:1140–1152 Jensen RA (2001) Orthologs and paralogs – we need to get
Eisen JA (1998) Phylogenomics: improving functional pre- it right. Genome Biol 1:21–23
dictions for uncharacterized genes by evolutionary Katinka MD, Duprat S, Cornillot E, Metenier G,
analysis. Genome Res 8:163–167 Thomarat F, Prensier G, Barbe V, Peyretaillade E,
Ewing RM, Kahla AB, Poirot O, Lopez F, Audic S, Claverie JM Brottier P, Wincker P et al. (2001) Genome sequence
(1999) Large-scale statistical analyses of rice ESTs re- and gene compaction of the eukaryote parasite
veal correlated patterns of gene expression. Genome Encephalitozoon cuniculi. Nature 414(6862):450–453
Res 9:950–959 Kellis M, Patterson N, Endrizzi M, Birren B, Lander ES
Fitch WM (1970) Distinguishing homologous from analo- (2003) Sequencing and comparison of yeast species
gous proteins. Syst Zool 19:99–113 to identify genes and regulatory elements. Nature
Fritz R, Lanen C, Colas V, Leroux P (1997) Inhibition of 423:241–254
methionine biosynthesis in Botrytis cinerea by the Kobayashi I, Tanaka C, Yamaoka N, Kunoj H (1991) Mor-
anilinopyrimidine fungicide pyrimethanil. Pest Sci phogenesis of Erisyphe graminis conidia on artificial
49:40–46 membranes. Trans Mycol Soc Jpn 187–198
Galagan JE, Calvo SE, Borkovich KA, Selker EU, Read ND, Koonin EV (2003) Comparative genomics, minimal gene-
Jaffe D, Fitzhugh W, Mal J, Smirnov S, Purcell S et al. sets and the last universal common ancestor. Nat Rev
(2003) The genome sequence of the filamentous fungus Microbiol 1:127–136
Neurospora crassa. Nature 422:859–868 Koonin EV, Mushegian AR, Bork P (1996) Non-orthologous
Gates MA, Kim L, Egan ES, Cardozo T, Sirotkin HI, gene displacement. Trends Genet 12:334–336
Dougan ST, Lashkari D, Abagyan R, Schier AF, Kubo Y, Nakamura H, Kobayashi K, Okuno T, Furasawa I
Talbot WS (1999) A genetic linkage map for zebrafish: (1991) Cloning of a melanin biosynthetic gene essen-
comparative analysis and localization of genes and tial for appressorial penetration of Colletotrichum la-
expressed sequences. Genome Res 9:334–347 genarium. Mol Plant-Microbe Interact 4:440–445
Giese H, Hippe-Sanwald S, Somerville S, Weller J (1997) Kubo Y, Takano Y, Endo N, Yasuda N, Tajima S, Furu-
Erisyphe graminis. In: Carroll GC, Tudzynski P (eds) sawa I (1996) Cloning and structural analysis of the
The Mycota, vol V, part B. Springer, Berlin Heidelberg melanin biosynthesis gene SCD1 encoding scytalone
New York, pp 55–78 dehydratase in Colletotrichum lagenarium. Appl Env-
Giles PM, Soanes DM, Talbot NJ (2003) A relational database iron Microbiol 62:4340–4344
for the discovery of genes encoding amino acid biosyn- Lev S, Sharon A, Hadar R, Ma H, Horwitz BA (1999)
thetic enzymes in pathogenic fungi. Comp Funct Ge- A mitogen-activated protein kinase of the corn
nomics 4:4–15 leaf pathogen Cochliobolus heterostrophus is in-
Goffeau A, Barrell BG, Bussey H, Davis RW, Dujon B, Feld- volved in conidiation, appressorium formation, and
mann H, Galibert F, Hoheisel JD, Jacq C, Johnston M pathogenicity: diverse roles for mitogen-activated
et al. (1996) Life with 6000 genes. Science 546:563–567 protein kinase homologs in foliar pathogens. Proc
Gogarten JP, Olendzenski L (1999) Orthologs, paralogs and Natl Acad Sci USA 96:13542–13547
genome comparisons. Curr Opin Genet Dev 9:630–636 Lorang JM, Tuori RP, Martinez JP, Sawyer TL, Redman RS,
Gold SE, García-Pedrajas MD, Martínez-Espinoza AD Rollins JA, Wolpert TJ, Johnson KB, Rodriguez RJ,
(2001) New (and used) approaches to the study Dickman MB et al. (2001) Green fluorescent protein
of fungal pathogenicity. Annu Rev Phytopathol is lighting up fungal biology. Appl Environ Microbiol
39:337–365 67:1987–1994
48 S. Ahmad et al.
Martinez D, Larrondo LF, Putnam N, Gelpke MD, Huang K, Sharman AC (1999) Some new terms for duplicated genes.
Chapman J, Helfenbein KG, Ramaiya P, Detter JC, Semin Cell Dev Biol 5:561–563
Larimer F et al. (2004) Genome sequence of the ligno- Siew N, Fischer D (2004) Structural biology sheds light on
cellulose degrading fungus phanerochaete chrysospo- the puzzle genomic ORFans. J Mol Biol 342:369–373
rium strain RP78. Nat Biotechnol 22:695–700 Soanes DM, Skinner W, Keon J, Hargreaves J, Talbot NJ
Mayorga ME, Gold SE (1999) A MAP kinase encoded by (2002) Genomics of phytopathogenic fungi and the
the ubc3 gene of Ustilago maydis is required for fil- development of bioinformatic resources. Mol Plant-
amentous growth and full virulence. Mol Microbiol Microbe Interact 15:421–427
34:485–497 Sweigard JA, Carroll AM, Farrall L, Chumley FG, Valent B
Mewes HW, Amid C, Arnold R, Frishman D, Guldener U, (1998) Magnaporthe grisea pathogenicity genes ob-
Mannhaupt G, Munsterkotter M, Pagel P, Strack N, tained through insertional mutagenesis. Mol Plant-
Stumpflen V et al. (2004) MIPS: analysis and annota- Microbe Interact 11:404–412
tion of proteins from whole genomes. Nucleic Acids Takano Y, Kikuchi T, Kubo Y, Hamer JE, Mise K, Furu-
Res 32:D41–44 sawa I (2000) The Colletotrichum lagenarium MAP
Muller P, Aichinger C, Fedbrügge M, Kahmann R (1999) The kinase gene CMK1 regulates diverse aspects of fun-
MAP kinase Kpp2 regulates mating and pathogenic de- gal pathogenesis. Mol Plant-Microbe Interact 13:374–
velopment in Ustilago maydis. Mol Microbiol 34:1007– 383
1017 Talbot NJ, Foster AJ (2001) Genetics and genomics of the
Mullins ED, Kang S (2001) Transformation: a tool for rice blast fungus Magnaporthe grisea: developing an
studying fungal pathogens of plants. Cell Mol Life Sci experimental model for understanding fungal diseases
58:2043–2052 of cereals. Adv Bot Res 34:263–287
Mullins ED, Chen X, Romaine P, Raina R, Geiser DM, Kang S Talbot NJ, Ebbole DJ, Hamer JE (1993) Identification
(2001) Agrobacterium-mediated transformation of and characterisation of MPG1, a gene involved in
Fusarium oxysporum: an efficient tool for insertional pathogenicity from the rice blast fungus Magnaporthe
mutagenesis and gene transfer. Phytopathology grisea. Plant Cell 5:1575–1590
91:173–180 Talbot NJ, Kershaw MJ, Wakley GE, DeVries OMH, Wes-
Mushegian AR, Koonin EV (1996) A minimal gene set for sels JGH, Hamer JE (1996) MPG1 encodes a fungal
cellular life derived by comparison of complete bacte- hydrophobin involved in surface interactions during
rial genomes. Proc Natl Acad Sci USA 93:10268–10273 infection-related development of Magnaporthe grise.
Mushegian AR, Garey JR, Martin J, Liu LX (1998) Large- Plant Cell 8:985–999
scale taxonomic profiling of eukaryotic model organ- Tatusov RL, Mushegian AR, Bork P, Brown NP, Hayes WS,
isms: a comparison of orthologous proteins encoded by Borodovsky M, Rudd KE, Koonin EV (1996)
the human, fly, nematode, and yeast genomes. Genome Metabolism and evolution of Haemophilus influenzae
Res 8:590–598 deduced from a whole-genome comparison with
Nitta N, Farman ML, Leong SA (1997) Genome organization Escherichia coli. Curr Biol 6:279–291
of Magnaporthe grisea: integration of genetic maps, Tatusov RL, Koonin EV, Lipman DJ (1997) A genomic per-
clustering of transposable elements and identification spective on protein families. Science 278:631–637
of genome duplications and rearrangements. Theor Tatusov RL, Galperin MY, Natale DA, Koonin EV (2000)
Appl Genet 95:20–32 The COG database: a tool for genome-scale analysis
Nobrega MA, Pennacchio LA (2004) Comparative genomic of protein functions and evolution. Nucleic Acids Res
analysis as a tool for biological discovery. J Physiol 28:33–36
554:31–39 Tatusov RL, Natale DA, Garkavtsev IV, Tatusova TA,
Ochman H, Lawrence JG, Groisman EA (2000) Lateral gene Shankavaram UT, Rao BS, Kiryutin B, Galperin MY,
transfer and the nature of bacterial innovation. Nature Fedorova ND, Koonin EV (2001) The COG database:
405:299–304 new developments in phylogenetic classification of
Orbach MJ, Farrall L, Sweigard JA, Chumley FG, Valent B proteins from complete genomes. Nucleic Acids Res
(2000) A telomeric avirulence gene determines efficacy 29:22–28
for rice blast resistance gene Pi-ta. Plant Cell 12:2019– Theiben G (2002) Secret life of genes. Nature 415:741
2032 Thinlay X, Finckh MR, Bordeos AC, Zeigler RS (2000) Ef-
Ouzounis CA, Coulson RMR, Enright AJ, Kunin V, Pereira- fects and possible causes of an unprecedented rice blast
Leal JB (2003) Classification schemes for protein struc- epidemic on the traditional farming system of Bhutan.
ture and function. Nat Rev Genet 4:508–519 Agric Ecosyst Environ 78:237–248
Rat Genome Sequencing Project Consortium (2004) Thomas D, Surdin-Kerjan Y (1997) Metabolism of sulfur
Genome sequencing of the Brown Norway rat amino acids in Saccharomyces cerevisiae. Microbiol
yields insights into mammalian evolution. Nature Mol Biol Rev 61:503–536
428(6982):475–476 Tucker SL, Talbot NJ (2001) Surface attachment and pre-
Rosewich UL, Kistler HC (2000) Role of horizontal penetration stage development by plant pathogenic
gene transfer in the evolution of fungi. Annu Rev fungi. Annu Rev Phytopathol 39:385–417
Phytopathol 38:325–363 Tucker SL, Thornton CR, Tasker K, Jacob C, Giles G, Egan M,
Ruiz-Roldan MC, Maier FJ, Schafer W (2001) PTK1, Talbot NJ (2004) A fungal metallothionein is required
a mitogen-activated protein kinase gene is required for pathogenicity of Magnaporthe grisea. Plant Cell
for conidiation, appressorium formation, and 16:1575–1588
pathogenicity of Pyenophora teres on barley. Mol Tunlid A, Talbot NJ (2002) Genomics of parasitic and sym-
Plant-Microbe Interact 14:116–125 biotic fungi. Curr Opin Microbiol 5:513–519
Wei L, Liu Y, Dubchak I, Shon J, Park J (2002) Comparative Xue C, Park G, Choi W, Zheng L, Dean RA, Xu JR (2002) Two
genomics approaches to study organism similarities novel fungal virulence genes specifically expressed in
and differences. J Biomed Inform 35:142–150 appressoria of the rice blast fungus. Plant Cell 14:2107–
Wood V, Gwilliam R, Rajandream MA, Lyne M, Lyne R, 2119
Stewart A, Sgouros J, Peat N, Hayles J, Baker S (2002) Yang G, Ross MS, Turgeon BG, Yoder OC (1996) A polyketide
The genome sequence of Schizosaccharomyces pombe. synthase is required for fungal virulence and produc-
Nature 415:871–880 tion of polyketide T-toxin. Plant Cell 8:2139–2150
Wren BW (2000) Microbial genome analysis: Insights into Yoder OC, Turgeon BG (2001) Fungal genomics and
virulence, host adaptation and evolution. Nat Rev pathogenicity. Curr Opin Plant Biol 4:315–321
Genet 1:30–39 Zhang Z, Gurr SJ (2000) Walking in the unknown; a
Xie T, Ding DF (2000) Investigating 42 candidate ortholo- “step-down” PCR-based technique leading to direct
gous protein groups by molecular evolutionary analy- sequence analysis of flanking genomic DNA. Gene
sis on genome scale. Gene 261:305–310 253:145–150
Xu JR, Hamer JE (1996) MAP kinase and cAMP signalling Zheng L, Campbell M, Murphy J, Lam S, Xu J-R (2000)
regulate infection structure formation and pathogenic The BMP1 gene is essential for pathogenicity in the
growth in the rice blast fungus Magnaporthe grisea. gray mold fungus Botrytis cinerea. Mol Plant-Microbe
Genes Dev 10:2696–2706 Interact 13:724–732
Fungal Rythms and Responses
4 Circadian Rhythms, Photobiology
and Functional Genomics in Neurospora
J.J. Loros1 , J.C. Dunlap1
CONTENTS I. Introduction
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 53
A. Fungi as Model Systems . . . . . . . . . . . . . . 54
B. Neurospora as a Research System . . . . . . . 55 Two widespread and tightly interconnected forms
1. Overview . . . . . . . . . . . . . . . . . . . . . . . 55 of cellular, tissue and organismal regulation are
2. Biology of Neurospora . . . . . . . . . . . . . 55
3. Neurospora as a Genetic and Genomic the ability to see environmental light in con-
System . . . . . . . . . . . . . . . . . . . . . . . . . 57 junction with the ability to tell the time of day.
4. Genome Defense Mechanisms . . . . . . . 59 Chromophore-binding photoreceptive molecules
5. Neurospora Genome Projects . . . . . . . . 59 facilitate the harvesting of photons, allowing an
II. Analysis of Circadian Rhythms
in Neurospora . . . . . . . . . . . . . . . . . . . . . . . . 61
organism to respond to changes in light fluence.
A. What is, and is not, a Circadian Rhythm . 61 Biological rhythms provide organisms with the
B. Phylogenetic Conservation of Rhythms capacity to anticipate environmental cycles
and Clock Components . . . . . . . . . . . . . . 61 imposed by the Earth’s rotation. The capability to
C. Assays of Rhythmicity . . . . . . . . . . . . . . . 62 gage time of day is called circadian rhythmicity,
D. Genetic Analysis of Rhythms . . . . . . . . . . 62
E. The Mechanism of the Circadian Clock and the physiological and molecular basis of these
in Neurospora . . . . . . . . . . . . . . . . . . . . . . 63 rhythms has been an object of study for well over
1. The Molecular Clock Cycle . . . . . . . . . . 63 a century. Recent decades have seen the unraveling
2. The Principal Molecular Components of this long mystery, and microbial systems have
of the Clock Cycle . . . . . . . . . . . . . . . . . 65
a) frq Transcripts and FRQ Proteins . . 65
led the way in many regards. In organisms with
b) WC-1 and WC-2, Positive Elements clocks, which include most eukaryotes and the
in the Feedback Loop . . . . . . . . . . . 65 cyanobacteria, most or all cells of the organism
III. Photobiology in Neurospora . . . . . . . . . . . . . 65 have their own molecular clock. The cellular
A. Overview . . . . . . . . . . . . . . . . . . . . . . . . . 65 nature of circadian rhythmicity, now universally
B. Control of the WC-1 Photoreceptor . . . . . 66
IV. Temperature Effects on the Clock . . . . . . . . . 66 recognized as a characteristic of rhythms, was
V. Output from the Clock . . . . . . . . . . . . . . . . . 66 first described in microbes. Phase response curves
A. Clock-Regulated Biology . . . . . . . . . . . . . 66 to light played a central role in delineating the
B. Clock-Controlled Genes – ccgs . . . . . . . . . 67 formal properties of biological oscillators, and
C. Microarray Analysis of Light, Clock
and Temperature Regulation . . . . . . . . . . 68
were first determined in Gonyaulax (Hastings and
D. Cross-Species Arrays . . . . . . . . . . . . . . . . 70 Sweeney 1958), as was the concept of using phase
VI. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 70 response curves to define sensitive and insensitive
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 phases to chemicals that facilitated the probing of
the molecular nature of the oscillator (Hastings
1960). More recently, as studies utilized more
genetic and molecular tools, the microbial system
Abbreviations: EST, expressed sequence tag;
Neurospora crassa has pioneered the way, along
MSUD, meiotic silencing of unpaired DNA; PKC,
with Drosophila, in finally describing molecular
protein kinase C; PTGS, post-transcriptional gene
feedback loops as the basis of rhythms as we
silencing; RIP, repeat-induced point mutations;
understand them today.
SNP, single nucleotide polymorphism
Work on Neurospora describing single gene
rhythm mutants (Feldman and Waser 1971)
1Departments of Biochemistry and Genetics, Dartmouth Medical occurred simultaneously with similar work in
School, Hanover, NH 03755, USA Drosophila (Konopka and Benzer 1971), as did
The Mycota XIII
Fungal Genomics
54 J.J. Loros and J.C. Dunlap
molecular cloning of the clock gene frq (Loros A. Fungi as Model Systems
and Dunlap 2001), which provided the first case of
rescue of a behavioral mutation by DNA-mediated Fungi, plants, and animals represent three main
transformation. Once such molecules were cloned phylogenetic kingdoms within the eukaryotes.
and isolated, work in Neurospora was the first to Fungi play huge roles, both positive and negative,
move beyond a description of the system to the in the economies of both industrialized nations
use of genetic engineering to manipulate the reg- and the world. Although fungi constitute the
ulation of these genes (Aronson et al. 1994a). This kingdom most closely related to Animalia (So-
led the way to establishing their roles in biological gin 1994; Simpson and Roger 2002), many are
oscillators and proving that such feedback loops exceptionally tractable experimentally and are
lay at the core of circadian rhythmicity. Molecular therefore universally used as model organisms
mechanisms for light-induced phase shifting and for understanding all aspects of basic cellular
for temperature entrainment of circadian clocks regulation. These regulatory networks include
were first determined in Neurospora (Crosthwaite cell cycle progression, gene expression, circadian
et al. 1995; Liu et al. 1998), and many of these timing, light sensing, recombination, secretion,
mechanisms show or have foreshadowed direct and multicellular development. Additionally, my-
parallels in mammalian systems (Shigeyoshi corrhizal fungi, those that grow interdependently
et al. 1997). Paradigms used for understanding with the roots of plants, are crucial symbionts
the rhythmic physiology of organisms have also without which most trees and many grasses
been developed in Neurospora, including the view cannot live. Fungi also carry out most biomass
that clock-regulated transcription would provide turnover. Importantly, mycelial fungi rank with
a major means of clock output. This led to the bacteria as the most serious of the human and
first genome-wide screens for clock-controlled plant pathogens. Because they are eukaryotes,
genes and the coining of the term “ccg” that has treatment of opportunistic fungal infections
now become widely used to describe circadianly in humans pose special risks, challenges and
regulated genes and transcripts (Loros et al. problems not posed by bacteria. The use of fungi
1989). in the manufacture of pharmaceutical products is
The ability of fungi to sense light using a multibillion dollar per year industry. Penicillin
photoreceptive molecules is also common to and similar β-lactams, all produced by fungi, are
most or all cells of the organism. The molecular the world’s largest-selling antibiotics. It is esti-
basis of light sensing has also been the focus of mated that 10%–35% of the world’s food supply
intense study, now yielding to the combination is lost each year due to fungal contamination,
of classic genetic and modern molecular, bio- a loss of over US$ 200 billion per year. Within
chemical and genomic techniques. Indeed, the the United States alone, fungicide sales constitute
well-known interplay between daily rhythms more than a US$ 1 billion per year industry. The
and light sensing turned out to have profound commercial production of chemicals by mycelial
interconnections at the molecular level. Many fungi constitutes an industry with a turnover
genes regulated by the clock are additionally of approximately US$ 35 billion per year. For
regulated by light. example, import into the United States of some
The FREQUENCY (FRQ) protein is the central of these chemicals (such as citrate) costs in excess
negative element in the autoregulatory feedback of US$ 1 billion annually. Industrial production
at the core of the circadian clock in Neurospora. of enzymes, largely by mycelial fungi, constitutes
White-collar-1 (WC-1), part of the heterodimeric a US$ 1.5 billion per year industry. About 75%
transcription factor with white-collar-2 (WC-2) of the approximately 250,000 different species
that drives frq expression and is thereby a central of fungi belong to the Ascomycetes, 90% being
component of the clock feedback mechanism, is mycelial fungi and the remainder yeasts. The other
also the photoreceptor for the circadian clock, 25% are Basidiomycetes, also mycelial, and having
as well as all other blue light-regulated genes in fruiting bodies known as “mushrooms”. Much of
Neurospora. This review will summarize much of the above economic impact is due to mycelial fungi
what is presently known about the molecular bases yet, until recently, there has been amazingly little
of rhythms in Neurospora, with a focus on current, research funding and, therefore, effort devoted to
genomics-based approaches. their characterization.
Clocks and Light in Neurospora 55
B. Neurospora as a Research System to generate knockouts of known sequences via

insertion of selectable markers (e.g., Aronson
1. Overview et al. 1994c) at frequencies of up to 90% of
Of the mycelial fungi, N. crassa is neither transformants (depending on the construct and
a pathogen nor the most widely used industrial the locus; 5%–10% is routine). A recent report
fungus. So, why study Neurospora? Neurospora has demonstrates that homologous insertion can be
been studied for decades and is the best understood achieved approaching 100% in Ku 70 and Ku
mycelial fungus. It is a saprophyte that displays 80 knockouts (Ninomiya et al. 2004). A number
both asexual and sexual life cycles, and it is easily of available vectors are routinely used to target
maintained through both cycles in laboratory transformation to specific loci within the genome.
conditions. Neurospora exists vegetatively as an Alternative methods for generating knockouts
incompletely septate syncytium, growing equally or knockdowns are also used with success. Most
well in simple liquid or on solid media of defined commonly used is RIP (repeat-induced-point
composition. It is nonpathogenic, and therefore mutation; Selker 1990), a rapid method whereby
easy and safe to work with, yet it is phylogenetically duplicated genes are detected in a parental strain
related to pathogens. Both asexual development during a sexual cross and mutated prior to meiosis.
and sexual differentiation are highly influenced Knockdowns are also made through quelling,
by environmental factors including nutrient, a form of co-suppression found in Neurospora
light and temperature. Neurospora is typically (Cogoni and Macino 1997). These advances are
haploid, undergoing only a transient diploid described in greater detail below.
stage immediately prior to meiosis. Because of
its interesting and diverse biology, simplicity of 2. Biology of Neurospora
culture, facile genetics and rapid growth rate,
Neurospora remains a widely used model that In Neurospora, as in many organisms, light has two
sustains a large and diverse research community. primary roles. The first and most obvious one is
The genetics of Neurospora is unparalleled to immediately and acutely regulate organismal
within the mycelial fungi. It has the most identified responses such as pigment production. The sec-
genes, and the densest and most accurate genetic ond and more subtle response is to regulate the
map. These attributes are a legacy of 70 years of phase of the biological clock, the internal circadian
effort that began concurrently with Drosophila timer that tells the organism what time of day it is
genetics. Biochemical genetics as a field got its and thereby modulates the organism’s responses to
start from work in Neurospora (Beadle and Tatum many factors, including light. In most organisms,
1945), and the ease of culture, rapid growth light and the clock work together to regulate many
(mass doubling time of 140 min), and ease of aspects of the life cycle, and Neurospora has proven
harvest continue to support research. Sexual to be an ideal model system for understanding the
crosses are technically trivial and the genetic interplay between these two forms of regulation.
generation time (from progeny to progeny) is Figure 4.1 provides an overview of the Neurospora
about 3 weeks. Sexual spores are stable for years life cycle and notes the places in the life cycle that
at 4 ◦ C, and asexual spores or mycelia can be are influenced by light and/or the clock.
stored for decades. Happily, molecular tools are Light signals influence many facets of both the
advanced and being continually improved, as is sexual and asexual (vegetative) stages of life, but
typical in a vibrant research community. The first a good starting place is to consider the basic ge-
mycelial fungus to be transformed, shortly after netics and growth characteristics. In their natu-
yeast (Davis 2000), Neurospora transformation is ral habitat, sexual spores are activated by the heat
routine at frequencies up to many thousands of from fires; Neurospora is classified as a Pyreno-
transformants per microgram of DNA. A variety mycete. Cultures thus emerge after fires, and the
of selectable markers exist, and several regulatable organism spends most of its life growing vegeta-
promoters are regularly used to control expression tively on the burned-over substrate. Neurospora
of transgenes (e.g., Aronson et al. 1994a). As in grows vegetatively as a syncytium with incomplete
animal cells, transformation is typically the result cell walls separating cellular compartments, and
of ectopic insertion through non-homologous comes in two mating types, A and a. Even geneti-
integration, although targeted disruption by cally different strains of the same mating type can
reciprocal homologous integration is widely used often fuse and intermingle their nuclei to form
Fig. 4.1. Regulation by light and the clock in the life cycle endogenous circadian clock and by ambient light. See text
of Neurospora. Many aspects of both the asexual vegetative for detail. Adapted from Davis (2000) with permission
life cycle and the sexual cycle are influenced by both the
heterokaryons, but strains with different mating dergoes meiosis to produce an eight-spored ascus,
types never fuse. Neurospora can exist as a vege- each ascus containing the products of meiosis from
tative culture in two forms, as surface mycelia or a single diploid nucleus.
it can elaborate aerial hyphae. The tips of aerial Acute effects of light during the asexual phase
hyphae form morphologically distinct structures of the life cycle include the acute induction of coni-
that act as asexual spores (called conidia) that are diation, the developmental process leading to the
easily dispersed by wind. When nutrients become production of the asexual spores. More conidia are
scarce, vegetative Neurospora of either mating type produced and they are produced faster (Klemm
is able to induce sexuality by forming a perithe- and Ninneman 1978; Lauter 1996). Although pig-
cium (a fruiting body) that can be fertilized by mentation of the conidia is constitutive, as noted
a nucleus from a piece of mycelium or an asex- above in the historical descriptions of Neurospora,
ual spore of the opposite mating type. Nuclei of carotenogenesis in mycelia is light-induced (Hard-
each parent replicate in tandem and eventually fuse ing and Shropshire 1980) and this response is quite
to make a transient diploid that immediately un- rapid, being observable within the first 30 min after
exposure to light. In a strain defective for light per- cess proceeds better in the dark. Not surprisingly,
ception or transduction of the light signal, the pro- carotenogenesis of perithecial walls is light induced
duction of white mycelia underlying yellow/orange (Perkins 1988). Perhaps the most interesting pho-
conidia is seen; this screen proved to be of cen- tobiology in the sexual phase involves the behavior
tral importance in the genetic identification of the of mature perithecia once they are formed. The
photoreceptors, as described below. Developmen- ejection of spores is induced by light, and the di-
tal responses such as light-induced conidiation are rection in which they are ejected is light regulated –
slower. Aerial hyphae are reported to display pho- that is, the tips (“beaks”) of the perithecia display
totropism (e.g., Siegel et al. 1968), in that they pref- a distinct phototropism (Harding and Melles 1983).
erentially form on the side of a dish near to light, Finally, even in the dark, the number of spores shot
but it is not clear to what extent this response is from perithecia is regulated by the light-phased
distinct from the overall light induction of coni- circadian clock.
diation. Also somewhat controversial are reports The processes described above are all light reg-
of changes in membrane conductivity (hyperpo- ulated at face value, but Neurospora is also capable
larization and an increase in input resistance) in of a more subtle response to light, characteristic
response to light (e.g., Potapova et al. 1984). of further regulatory sophistication: Neurospora
During the asexual cycle, shown in the center can respond to changes in the level of ambient
of the figure, the most global effect of light is to set light, a process known as photoadaptation. This
the phase of the endogenous biological clock that response is manifested in two ways, both tied to
acts to regulate a variety of aspects of the life cycle the response to light at the molecular level. In the
of the organism (Sargent and Briggs 1967; reviewed first, when the organism initially sees light, a re-
in Loros and Dunlap 2001). The clock controls the sponse is triggered that peaks within 15–30 min,
daily timing of a developmental switch that can but this response generally decays within about 2 h.
initiate the morphological changes leading to coni- If the organism is exposed to light within this 2-h
diation. Light is used to set the phase of the clock, in period, no additional response is seen. A related
that a light-to-dark transfer is interpreted as dusk phenomenon is observed if the light remains on –
and a dark-to-light transfer as dawn; the molecular the response decays but, interestingly, after the 2-h
basis of this response will become clear below in latency period the organism can respond again if
the chapter. Continued light also acts to suppress the ambient level of light is increased.
the expression of the clock. Conidiation (differen-
tiation leading to asexual spore production) can 3. Neurospora as a Genetic and Genomic System
be triggered by environmental signals including
blue light, desiccation, and nutrient starvation, as The Neurospora genome is organized among seven
well as by the endogenous circadian clock, in oth- chromosomes corresponding to the seven genetic
erwise constant conditions. This asexual develop- linkage groups, although not as a one-to-one map-
ment involves a major morphological change that ping since the chromosomes were defined cyto-
requires many novel gene products. Although the logically, independently from the genetic accre-
production of asexual spores is the best character- tion of the linkage groups. In any case, the entire
ized light-phased circadian rhythm in Neurospora, genome comprises something approaching 2000
other persisting rhythms at the physiological level map units. General features of the genome are sum-
have been described, which include the production marized in Table 4.1. There are about 11,000 genes,
of CO2 , lipid and diacylglycerol metabolism (e.g., a number that compares favorably with other ge-
Roeder et al. 1982; Lakin-Thomas and Brody 2000; netic model systems – Drosophila (around 14,000
Ramsdale and Lakin-Thomas 2000), a number of genes) and Caenorhabditis elegans (about 19,000
enzymatic activities (e.g., Hochberg and Sargent genes). Genes appear on average every 3.7 kb along
1974; Martens and Sargent 1974), heat shock pro- the chromosome. The average length of a gene is
teins (Rensing et al. 1987), and even growth rate about 1.3 kb, compared to about 1.1 kb for Sac-
(Sargent et al. 1966). charomyces and Schizosaccharomyces. Introns are
During the sexual phase of the life cycle, light found in nearly all genes and are evenly distributed,
again induces many and varied effects. The over- unlike Saccharomyces where most genes do not
all initiation of the sexual process is enhanced by have introns and the few introns that exist tend to
light (Degli Innocenti and Russo 1983). This is par- be clustered in the 5 ends of genes. Due to whole-
ticularly interesting since, once initiated, the pro- genome scanning mechanisms such as RIP that de-
Table. 4.1. General features of the Neurospora crassa genome (updated from Borkovich et al. 2004)
General
Chromosomes 7
%G+C 50%
Protein-coding genes 10,620
Protein-coding genes>100 aa 9,200
Introns 17,118
tRNA genes 424
5S rRNA 74
Percent coding 44%
Average gene size 1673 bp (481 aa)
Median intron size 84 bp
Average intergenic distance 1953 bp
Identified by similarity to known sequences 1,336 (13%)
Conserved hypothetical proteins 4,606 (46%)
Predicted proteins (no similarity to known sequences) 4,140 (41%)
Current genome assembly (as of March 2005)
Average contig length 273 kb
Total length of combined contigs 38,433,854 bp
Remaining sequence gaps 37
tect and eliminate duplications, there is a dearth of twice that found in yeasts, especially because
repeats and gene families relative to Saccharomyces of mechanisms in Neurospora that target and
(Galagan et al. 2003) and other genomes. eliminate duplications (e.g., RIP). For this reason
In general, Neurospora genes appear to be there are few gene families, and hence nearly
typical of those seen in “higher” eukaryotes. As all of the sequence complexity reflects actual
noted above, most genes have several introns, diversity. Neurospora shares gene sequences with
and 5 noncoding regions that may also contain a variety of taxonomic groups, and so information
introns precede coding regions. Promoters can be from Neurospora will inform projects covering
complex, in that multiple promoters can direct the the breadth of biology, not just within the fungi.
expression of a single coding region, and they are Neurospora is definitively not Saccharomyces with
regulated in the combinatorial manner typical of a larger genome. Instead, it is the gateway to an in-
eukaryotes. A number of examples of alternative credibly diverse group of organisms – the mycelial
splicing are becoming known. There is a distinct fungi – which are a cornerstone of our ecosystem
codon bias in favor of C and against A in the and annually contribute over US$ 40 billion to
third position. Additional details can be found the United States economy alone. Although both
in (Borkovich et al. 2004), in the paper de- nonpathogenic and easy to manipulate, Neu-
scribing the genomic sequence (Galagan et al. rospora is phylogenetically very closely allied and
2003), or on the web (http://www.genome.wi.mit. genetically syntenic, both with important animal
edu/annotation/fungi/neurospora/index.html). and plant pathogens and with agriculturally and
Several independent estimates have suggested industrially important production strains (such
a surprising degree of genetic novelty in the as Cochliobolus, Fusarium, Magnaporthe and
Neurospora genome. Repeated estimates suggest Trichoderma). Strong parallels have been noted in
that nearly 50 % of Neurospora genes have no signaling pathways, photobiology, developmental
homologs or orthologs in GenBank (perhaps regulation, and many aspects of metabolism
because of the 13 million sequences in GenBank, including secondary metabolism, to name a few.
only about 270,000 come from ascomycetes). These In a broader sense, there is great potential for
are either novel genes with novel functions, or information synergy, perhaps reflected in the fact
novel genes representing different ways of carrying that the Neurospora Genome site at the White-
out known functions; in either case, they will be of head Institute Center for Genome Research at MIT
great interest. The 11,000 genes in the Neurospora draws over 3500 hits per day. Neurospora is the
crassa genome implies a complexity approaching best understood of the more than 250,000 species
that of Drosophila and Caenorhabditis, and about of mycelial fungi, and work in this system has his-
torically paved the way to developments in other of ascus-dominant mutations (Round spore, Peak,
fungal species. Banana) whose ascus phenotypes constitute the
MSUD-mediated response to loss of the gene, and
4. Genome Defense Mechanisms allowing diploid meiotic nuclei possessing a gene
deletion in a meiosis-required gene (i.e., a single
As a first approximation, we normally expect the unpaired copy) to pass through crosses.
characteristics of an organism, or the characteris- Whereas both RIP and MSUD are active
tics of progeny arising from a cross between mem- only during the sexual cycle, Neurospora exhibits
bers of the same genus, to reflect the basic genetic a third defense mechanism in vegetative cells.
composition of the organism. However, this is not This is quelling, a PTGS (post-transcriptional
always the case. One reason for this, as initially gene silencing) that is in all ways akin to the
described by McClintock (1950), is that mobile ge- double-stranded RNA-induced gene silencing seen
netic elements can gain access to genomes and then in most eukaryotes and that has been studied in
move about causing mutations both by their exci- depth by Macino, Cogoni and colleagues (Cogoni
sion and their reinsertion into the genome. Per- and Macino 1994, 1997). As in most organisms, the
haps because of its syncytial growth habit, fungi process requires an RNA-dependent RNA poly-
like Neurospora would be especially susceptible to merase, an Argonaute-like protein and a RecQ-like
such selfish DNA. As a result, three independent helicase (encoded by qde-1, qde-2, and qde-3),
mechanisms have evolved to protect the genome some of which were first identified as quelling
and to prevent mobile genetic elements from pass- defective mutants of Neurospora.
ing through a cross into the genomes of progeny.
The first of these is RIP, the acronym for 5. Neurospora Genome Projects
repeat-induced point mutations. Initially de-
scribed by Selker and colleagues in 1987 (Selker Currently, there are approximately 70 research
et al. 1987), RIP describes a process in which, laboratories in the US focusing on Neurospora,
in the transient diploid cells that arise during and an additional approximately 80 in the rest of
a sexual cross, both genomes are scanned for the world, comprising a community of around
duplicated sequences, i.e., for sequences present 700 investigators (see http://www.fgsc.net/). In
in more than one copy anywhere in the genome. the US alone this represents more than 70 years
If any are found, both copies are riddled with GC of support and 500 graduate and postdoctoral
to AT transversions with the result that, typically, training positions. Over the past 10 years there has
both copies are inactivated. There are regions of been over US$ 60 million in funding to support
the genome that are protected from RIP, such as Neurospora research. Readers may also appreciate
the ribosomal repeats in the nucleolus organizer that these numbers are likely to underestimate
region, but interestingly these cannot be targeted the actual effort, since they are based largely on
in transformation experiments. self-reporting in surveys prepared for the NIH
A somewhat similar screening mechanism is Non-Mammalian Models Workshop in 1999, and
known as MSUD (meiotic silencing of unpaired for the NSF grant that supported the genomic
DNA; Shiu et al. 2001). Normally, of course, af- sequencing project. Compounding the inexact-
ter karyogamy in a haploid, each gene is present ness is the fact that the Neurospora community
in two copies that pair during the early stages of is also growing rapidly, with at least five new
meiosis. MSUD is a process in Neurospora, akin laboratories starting within the past 3 years. The
to but not the same as transvection, wherein the worldwide Neurospora community is tight-knit.
presence of an unpaired gene results in the meiotic Annual meetings tie the community together, and
silencing of all copies of that gene in the genome. organism/system-specific interests are overseen
This is the basis of barrenness observed in crosses by an elected Neurospora Policy Committee on
involving nonreciprocal translocation strains that which many of the investigators have served.
would, if fruitful, generate segmental aneuploids. If In the early 1990s, interest in the genomics of
the translocated region (present therefore in three Neurospora began to grow. In 1993 the Chair of the
copies) contains genes required for meiosis, then all Neurospora Policy Committee formally initiated,
three copies are silenced and the cross is aborted. in the interests of the community, a Neurospora
Mutations in the sad-1 gene (Shiu et al. 2001) ab- genome focus of effort; with evolving leadership,
rogate this phenomenon, suppressing a number this project is entering its ninth year. Efforts to
physically map the genome (coordinated by the realistic need for approachable genetics recom-
Committee to Order the Neurospora Genome) mended Neurospora for further analysis. Hence,
began in 1993. Independently, the first federally a consortium of universities has come together in
funded EST (expressed sequence tag) project four projects to further work on the Neurospora
began in 1995 at the University of New Mexico, genome. The institutions are Dartmouth Medical
and funds to physically map the genome were School, University of California at Riverside, Uni-
awarded to the University of Georgia in 1998. The versity of California at Los Angeles, University of
initiative to raise funds to sequence the genome Missouri, MIT, the Oregon Health Sciences Univer-
of Neurospora began in 1996 under the direction sity, the University of California at Berkeley, and
of the outgoing chair of the Policy Committee the University of New Mexico, with funding pro-
(http://gene.genetics.uga.edu/white_papers/ncras- vided by the National Institutes of General Medical
sa.html), as did a second, independently funded Sciences.
EST project (a collaboration between the Uni- As further described on the Genome Project
versity of Oklahoma and Dartmouth Medical website (http://www.dartmouth.edu/%7Eneurosp
School). Many members of the international oragenome/), the first project is pursuing the
community stepped forward to work toward systematic disruption of genes through targeted
this goal, the first success coming in 1998 with gene replacements, preliminary phenotyping
the award of DM 7.5 million in Germany for of these strains, and their distribution to the
the complete sequencing of two chromosomes scientific community at large. Project 1 will rely
(http://mips.gsf.de/proj/neurospora/). In the US, on bioinformatic support from Project 2. Through
a Neurospora Genomics Policy Committee (http:// a primary focus on annotation and genomics,
www.unm.edu/∼ngp/WhitePaper.html) was or- Project 2 is producing a platform for electronically
ganized. This led to an NSF grant for US$ capturing community feedback and data about the
5.2 million that was awarded in 2000 for the existing annotation, while building and maintain-
completion of the archival, reference-quality ing a database to capture and display information
(minimum tenfold coverage) genomic sequence about phenotypes that is relying on data from EST
that would serve as the reference genome for analyses (Project 4) to refine the gene structures.
the mycelial fungi. These efforts, based at MIT Oligonucleotide-based microarrays created in
(http://www.broad.mit.edu/annotation/fungi/neur Project 3 are allowing transcriptional profiling
ospora/) , set the stage for further work. of the nearly 11,000 distinguishable transcripts
The availability of whole genomic sequences in Neurospora. This effort will provide a baseline
has vastly accelerated the pace of research in eu- analysis of gene expression under a variety of
karyotic model systems. However, to exploit this growth conditions, and later begin to analyze the
resource, research communities must (1) anno- global effects of loss of novel genes in strains cre-
tate the genome to extract the relevant informa- ated by Project 1. These data will be made available
tion, (2) systematically disrupt the functions of through the web via structures created in Project 2.
the identified genes, (3) examine the regulation of Since alternative splicing, alternative promoters,
the genes in different biological contexts, and fi- and long antisense transcripts contribute widely
nally (4) communicate this information to the sci- to the overall complexity of expressed sequences
entific community at large, particularly to those in Neurospora, in Project 4, cDNA libraries are
studying similar problems in other systems. We being generated from wild-type and related strains
may then integrate all these aspects of phenotype to document this complexity to aid in annotation
and regulation into a comprehensive portrait de- in Project 2. Sequences from related strains have
scribing the biology of organisms. It is a tautol- also driven assembly of an SNP (single nucleotide
ogy to state that the simplest organisms are the polymorphisms) map. Overall, this effort will
easiest to dissect and also reveal the least, and help to anchor genomic exploration within the
that the most complicated organisms, while the largely unexplored phylogenetic kingdom of the
most information-rich, may be beyond the scope fungi.
of current efforts. Yet, it is apparent from the ex- These community-based efforts are providing
tant genomic comparisons that conservation of im- a wonderful context for further work on circadian
portant biological processes is the rule, and that rhythms and photobiology. This remains an
simple models can inform more complex systems. extremely active area of research; a search of
The desirability of rich biology, coupled with the [Neurospora AND circadian] yields over 100
publications since 2000, and genomics tools are character (e.g., Loros 1984; Loros et al. 1986; Aron-
well adapted for building on the general models son 1994b). However, in all cases one or more of
that have derived from the past two decades of the canonical properties of circadian rhythmicity
combined molecular and genetic analyses. As is lost (often temperature compensation), and so
a backdrop for describing these ongoing studies, these rhythms are understood to have a different
the present status of the Neurospora clock and basis, not comparable to that known for circadian
light regulatory machineries will be described clocks.
next.
B. Phylogenetic Conservation of Rhythms
and Clock Components
II. Analysis of Circadian Rhythms All known eukaryotic clocks are based at least in
in Neurospora part upon transcriptional/translational feedback
loops that close within the cell (Dunlap 1999).
A. What is, and is not, a Circadian Rhythm These clocks all share similar components and or-
ganizational logic in their assembly and operation,
Circadian rhythms reflect the output of biologi- although considerable diversity has evolved in the
cal clocks; they can regulate cyclical biochemical, number and types of some of the components ex-
physiological, or behavioral functions and, in na- ecuting the necessary functions. A cartoon of the
ture, have a period length of exactly 24 h. Under general scheme of the core feedback loops in the
constant conditions in the laboratory, they run at clock is seen in Fig. 4.2. In all eukaryotes, the posi-
their own inherent frequency, which is about one tive elements are heterodimers of two proteins that
cycle per day – circadian. At 25 ◦ C the period length interact via PAS domains to make a transcription
of the Neurospora circadian clock is about 22 h. The factor. In Neurospora, the two are WC-1 and WC-
phase of these rhythms is determined by the envi- 2, in Drosophila CYC and CLK, and in mammals
ronmental light/dark cycle and temperature cycle, BMAL1 and CLOCK. WC-1 is a sequence and func-
but they are not simply a reaction to the light/dark tional homolog of CLOCK. There is greater diver-
cycle. Rather, they represent the overt expression sity amongst the negative elements. In Neurospora,
of an endogenous and self-sustaining timekeeping the negative element is FRQ that acts as a dimer, in
mechanism. The period of the cycle varies only a lit- Drosophila it is PER and TIM as a complex or PER
tle, if at all, when the organism is examined on dif- acting alone, and in mammals it is a complex of four
ferent media or at different temperatures (Dunlap
et al. 2003), a characteristic known as temperature,
nutritional, or pH compensation. The ability to
be reset by short light or temperature treatments
(called entrainment), the period length of about
a day under constant conditions, and the compen-
sation capacity are characteristics that distinguish
circadian rhythms from cell cycle-regulated phe-
nomena or from other ca. 24-h metabolic or de-
velopmental rhythms whose period lengths are of-
ten temperature or nutritionally dependent. These
three characteristics of a true circadian rhythm de-
fine it as being circadian, and unite it with simi- Fig. 4.2. A generalized view of the feedback loops associ-
lar rhythms found ubiquitously in most eukaryotes ated with rhythmicity. Positive elements, heterodimers of
and cyanobacteria. PAS domain-containing proteins such as WC-1 and WC-2,
act as transcription factors to drive expression of“negative
In addition to circadian rhythms, organisms in- elements” such as FRQ. Negative elements in turn feedback
cluding Neurospora can express a variety of other to attenuate the activity of their activators, thereby giving
rhythms of varying period lengths. These can of- rise to a negative feedback loop that, with the proper delays,
ten be observed as cycles in morphology (Feldman can oscillate. Negative elements also feed forward to pro-
mote the synthesis of their activators. One aspect of output
and Hoyle 1974) or in growth rate on exotic me- occurs when positive elements activate expression of genes
dia (Lakin-Thomas 1996), but in some cases the that do not participate in the clock loops. Adapted from
rhythm presents at face value a bona fide circadian Dunlap (1999)
proteins, two homologs of the Drosophila PER pro- The CO2 masking effect is alleviated by the band
tein called PER1 and PER2, and two cryptochrome (bd) mutation (Sargent et al. 1966), and for this
homologs, CRY1 and CRY2. reason all laboratory stocks commonly used for
Outside of Neurospora, but within the fungi, circadian rhythm studies carry a mutation in the
relatively little is known about the molecular bases bd gene.
of rhythms. The frq gene has been identified in The Neurospora clock can also be assayed us-
a number of relatives within the Sordariaceae, ing liquid cultures, and this facilitates collection
including several species of Neurospora and of materials for molecular study (Nakashima 1981;
Sordaria. More broadly within the Pyrenomycetes Perlman et al. 1981; Loros et al. 1989; Aronson et al.
and Loculoascomyetes, frq has been identified 1994a; Garceau et al. 1997). The clock runs nor-
in diverse genera including organisms such as mally in liquid cultures, and individual mycelial
Magnaporthe, Fusarium, Leptosphaeria, and samples will retain their endogenous rhythmicity
Chromocrea (Merrow and Dunlap 1994; Lewis and and phase when transferred from liquid to solid
Feldman 1997; Lewis et al. 1997; J. Dunlap and growth media. As an additional alternative, real-
J. Loros, unpublished data). Not surprisingly, the time analysis of rhythmic clock-controlled or core
degree of sequence relatedness drops steeply with clock gene activity can be assayed in vivo using
the degree of phylogenetic distance, but functional firefly luciferase as a reporter for gene expression
conservation of FRQ has been demonstrated by (Mehra et al. 2002; Morgan et al. 2003).
using the Sordaria frq gene to complement a loss-
of-function mutation in Neurospora (Merrow D. Genetic Analysis of Rhythms
and Dunlap 1994). All fungi examined to date
have homologs to WC-1, the core element of the These assays have facilitated the isolation of strains
Neurospora clock showing the most sequence in which the clock runs fast or slow, or not at
conservation with known clock components from all (such as chr and frq; Feldman 1967; Feldman
mammals (Lee et al. 2000). A circadian rhythm and Hoyle 1973; Feldman et al. 1979; Gardner and
has recently been reported in Aspergillus nidulans Feldman 1980), or where both the period length
(Greene et al. 2003), an organism whose genome and temperature compensation are affected. prd-1,
appears to have no close sequence homologs of prd-2, prd-3, prd-4, cys-9, wc-2 and other frq alleles
FRQ. affect both the period of the rhythm and temper-
ature compensation (Feldman and Atkinson 1978;
C. Assays of Rhythmicity Feldman et al. 1979; Onai and Nakashima 1997;
Collett et al. 2002). The frequency (frq) locus, which
To appreciate how genomics have been used to ap- was identified multiple times (Feldman and Hoyle
proach rhythms and photobiology in Neurospora, 1973; Gardner and Feldman 1980), can give rise
background is needed on how the clock is mon- to both long (24, 29 h) and short (16, 19 h) period
itored. Several features have made Neurospora an length mutants as well as mutants lacking circadian
attractive and successful model for the dissection rhythmicity (although they do retain non-circadian
of circadian timing. As noted above, in Neurospora rhythms). Most genes associated with rhythms are
the clock is controlling the potential to develop, identified through a single allele, and most have
rather than the developmental process itself. Once not yet been cloned. Hence their role in the clock,
the switch is thrown, development can take either even to the extent of whether the effect is specific
a long or a short time, hours to days, depending on or pleiotropic, is not known. Occasionally, even for
the nutritional state and temperature of the culture. cloned genes, sequence data have not yielded much
The cycle recurs once a day, such that a character- information as to the molecular mechanism of the
istic conidial banding pattern of developed or de- clock. When frq was first cloned, it was evident that
veloping conidiophores can be readily followed in it did not have extended homology to any gene or
hollow glass culture tubes called race tubes (Ryan protein in the GenBank database. It was only exper-
et al. 1943). Race tubes are glass tubes about a cen- imental manipulation of the gene that identified its
timeter in diameter and 30–100 cm long that are central role in a feedback loop, as described below.
bent up at a 45◦ angle at both ends to hold agar On the other hand, there are a number of genes that
medium for growth (Fig. 4.3). As vegetative growth have been identified in genetic screens that give
proceeds, CO2 levels become elevated and can sup- rise to products of known biochemical function
press conidiation, and therefore mask the rhythm. that can now be fitted into an internally consistent
Fig. 4.3. The Neurospora conidiation rhythm can be viewed under red light. Positions of the easily visualized conidial
on race tubes. Cartoons of the race tube assay for monitor- bands (separated by undifferentiated surface mycelia), rel-
ing the phenotypic expression of the Neurospora clock are ative to the marked growth fronts, permit determination of
shown at top and bottom. Conidia are inoculated at one end both period length and phase of the rhythm. Photographs
of a hollow glass tube about 40 cm long and 16 mm in diam- and cartoon depictions of race tubes representative frq+
eter that is bent upward at both ends to accommodate an and frq mutant strains demonstrating genetic regulation
agar medium. Following growth for a day in constant light, of period length and overall rhythmicity are shown in the
the position of the growth front is marked and the culture is middle. A strain bearing the frq2 allele shows a short period
transferred to constant darkness. This light-to-dark trans- of approximately 19 h relative to the wild-type frq+ period
fer synchronizes the culture and sets the clock running from of 22 h. The frq7 strain has a long period of about 29 h, and
subjective dusk. The growth front can be marked every 24 h the frq10 gene replacement strain is arrhythmic
model for a feedback loop whose function is essen- systems is a general pattern describing how gene
tial for normal operation of the Neurospora circa- products and genes act together in the circadian
dian clock under constant conditions. For exam- system.
ple, phosphorylation has been established to play
an important regulatory role in maintaining the ki-
netics of oscillations of the clock (Kloss et al. 1998; E. The Mechanism of the Circadian Clock
Price et al. 1998; Liu et al. 2000), and several kinases in Neurospora
are found in Table 4.1. One of these, encoding casein
kinase 2, is now known to be universally associated 1. The Molecular Clock Cycle
with all circadian clockworks, but was first iden-
tified by work in Neurospora (Yang et al. 2002). It The white collar-1 (WC-1) and white collar-2
is not surprising that several genes involved in mi- (WC-2) proteins, and both frq mRNA and FRQ are
tochondrial function and energy metabolism have central components of the Neurospora circadian
been identified as clock-affecting genes. What has oscillator (Aronson et al. 1994a, b; Crosthwaite et al.
emerged from the comparisons of clocks in other 1997; Dunlap 1999; Collett et al. 2001; Fig. 4.4). WC-
1 and WC-2 heterodimerize to form a white collar 2001b, 2002). The result is that WC-1 levels begin to
complex, WCC (Ballario et al. 1996, 1998; Linden rise even as phosphorylation-promoted turnover
et al. 1997; Linden and Macino 1997; Talora et al. of FRQ begins. Thus, at close to the same time in
1999; Denault et al. 2001; Cheng et al. 2002), and a daily cycle, FRQ is promoting WC-1 synthesis
the WCC activates expression of the frq gene. Alter- to increase the level of the WCC while blocking
native splicing of frq mRNA (see below) can result activation of the frq promoter by the WCC. This
in the production of two distinct FRQ proteins creates a mass of WCC held inactive by FRQ (Lee
(Garceau et al. 1997) that feed back to block this ac- et al. 2000; Denault et al. 2001; Froehlich et al. 2003).
tivation. In the process of executing this feedback When the phosphorylation of FRQ finally triggers
loop, both frq mRNA and FRQ protein are rhythmi- its turnover (Garceau et al. 1997; Liu et al. 2000),
cally expressed in a daily fashion, and FRQ protein FRQ-promoted synthesis of WC-1 is balanced by
acts to repress the abundance of its own transcript WC-1 degradation, and WC-1 levels peak in the
(Aronson et al. 1994a; Garceau et al. 1997; Merrow night (Lee et al. 2000). The high WCC activity is
et al. 1997; Froehlich et al. 2003). Light and temper- released to initiate the next cycle and to maintain
ature, two of the most important environmental a robust amplitude in the feedback loop (Froehlich
signals, reset the Neurospora clock by changing the et al. 2003).
levels of frq mRNA and FRQ protein (Crosthwaite
et al. 1995; Liu et al. 1998), as further outlined below.
Figure 4.4 serves as a guide for following the
progress of the Neurospora clock cycle through the Light
day. By late night, FRQ protein has recently been
degraded and frq RNA levels are low. WC-1 and
WC-2 in the form of the WCC bind to the pro-
FRQ Nucleus
moter of the frq gene at two sites (LREs or light- P
FRQ
responsive elements) to drive the circadian rhythm P P WCC
ubiquitination & WCC
in transcription of the frq gene (Froehlich et al. turnover
2003). frq primary transcripts are spliced in a com- vvd
plex manner, and gradually by the early morning WCC
FRQ proteins appear (Garceau et al. 1997), dimer- FRQ wc-2
ize (Cheng et al. 2001a), and soon enter the nu- FRQ wc-1
P P
cleus (Garceau et al. 1997; Luo et al. 1998) where frq
processing
they interact with the WCC (Cheng et al. 2001a; kinases
t ranslation ccg's
Denault et al. 2001; Merrow et al. 2001) so as to FRQ
attenuate its activity (Froehlich et al. 2003). By proteins
proteins
midday, WCC activity reaches a nadir as FRQ lev- Clock-controlled processes:
developmental regulation
els rise (Lee et al. 2000), thereby turning down stress responses
intermediary metabolism
the expression of the frq gene. As FRQ appears,
Cell
it is phosphorylated by casein kinase Ia, casein
kinase II, and calcium/calmodulin-dependent ki-
nase, actions that regulate the stability of FRQ (Liu FRQ
frq WC-1
et al. 2000; Gorl et al. 2001; Yang et al. 2001, 2002)
as well as FRQ/WCC interactions (Yang et al. 2002).
As a result of inhibited WCC, frq transcript levels
begin to decline, although continued translation
causes FRQ protein levels to continue to rise. Thus,
frq mRNA levels peak in the midmorning (Aron-
son et al. 1994a; Crosthwaite et al. 1995), about day night
4–6 h before the peak of total FRQ in the afternoon Fig. 4.4. Coupled feedback loops in the Neurospora circadian
(Garceau et al. 1997). system. Known molecular components and their regulatory
FRQ also exerts a second role in the cycle by relationships in the Neurospora circadian clock are shown.
Lines ending in bars connote negative regulation or repres-
promoting, through an unknown mechanism, the sion. Arrows connote positive regulation. P on FRQ denotes
translation of WC-1 from existing wc-1 message phosphorylation. See text for details. Modified from Lee
(Lee et al. 2000; Merrow et al. 2001; Cheng et al. et al. (2000)
As described here, frq mRNA cycles with a peak and protein–protein interactions (reviewed in Bal-
during the subjective day (circadian daytime in lario and Macino 1997; Dunlap 2005; Dunlap and
constant conditions of darkness and temperature) Loros 2005). WC-1 contains a polyglutamine ac-
and FRQ expression lags by about 4 h, peaking tivation tract, a single class 4 DNA-binding Zn-
late during the day, near dusk. In constant con- finger domain, two PAS domains that are important
ditions, therefore, subjective day is defined as the for protein–protein interactions and that are re-
time when these components are at their highest quired for both light and clock functions (Lee et al.
levels, and night corresponds to the troughs in frq 2000), and also an additional, specialized subclass
and FRQ. This information forms the basis of our of PAS domains called LOV (for light, oxygen, and
understanding of how light adjusts the phase of the voltage sensing). WC-1 binds flavin adenine dinu-
rhythm, a process called entrainment. Light acts cleotide (FAD) as a chromophore. WC-2 is 530 aa
through the WCC to rapidly induce frq expression in length and is related to WC-1 at the sequence
(Fig. 4.4), resulting in a large and transient increase level. It has a single PAS domain, as well as a sin-
in frq and FRQ. This increase in frq is mediated by gle activation domain and Zn-finger domain. As
an frq antisense message. Depending on the time in noted above, WC-1 and WC-2 form a white col-
the frq/FRQ cycle during which the light falls, it will lar complex (WCC) via their PAS domains, and
either cause the frq peak to occur sooner or it will bind as a complex to the light-responsive elements
delay its decline, thereby advancing or delaying the (LREs) in promoters of light-regulated genes such
clock. as frq.
Additionally, in Neurospora as in many circa-
dian systems, the response of the system to light
2. The Principal Molecular Components is regulated by the clock, a process referred to as
of the Clock Cycle gating and mediated in part by vivid (VVD), an-
a) frq Transcripts and FRQ Proteins other photoreceptor and member of the PAS pro-
frq primary transcripts arise from multiple up-and tein superfamily (Heintzen et al. 2001; Schwerdt-
downstream promoters (H. Colot, J. Loros and feger and Linden 2003). Photoadaptation, a third
J. Dunlap, unpublished data), and are heavily light-modulating effect, is the temporary insensi-
spliced in a complex manner. Splicing determines tivity of the organism to respond to a second light
whether long or short FRQ proteins are synthe- pulse or to an increase in light intensity. VVD also
sized. A long antisense transcript arises from mediates this in part, perhaps by contributing in
the region of frq. It is light induced, oscillates some way to the post-translational modification
antiphase to sense frq transcripts, and appears of WC-1 that leads to light-induced turnover of
to play a role in ensuring precise entrainment WC-1.
to light/dark cues (Kramer et al. 2003). In any
case, two FRQ proteins are encoded by frq, a long
form of 989 amino acids and a form lacking III. Photobiology in Neurospora
the first 100 amino acids (Garceau et al. 1997).
Both forms can form homo- and heterodimers A. Overview
(Cheng et al. 2001a) and are needed for robust
rhythmicity. Their expression pattern responds Light responses in Neurospora have been appre-
to ambient temperature. At low temperatures ciated for well over 100 years and their analysis
(<22 ◦ C), the short form is preferentially required drove much of the early genetics on the organism:
and less overall FRQ is needed, whereas at high the white collar alleles were among the first to
temperatures (>26 ◦ C) the large form is favored at be generated in the organism. All known light
higher overall levels (Liu et al. 1997). responses in Neurospora are specific to blue light,
and to date no responses to either red or far-red
light have been documented, although many
b) WC-1 and WC-2, Positive Elements people have looked for them. Interestingly, the
in the Feedback Loop Neurospora genomic sequence has exposed a num-
WC-1, a GATA-like transcription factor, is a 1167 ber of possible photoreceptors including several
amino acid protein regulated both transcription- bacteriophytochromes and a cryptochrome
ally and post-transcriptionally at the level of syn- (Borkovich et al. 2004; Dunlap and Loros 2004).
thesis, and additionally through phosphorylation When Macino and colleagues cloned the white
collar genes (Linden et al. 1999), they specu- IV. Temperature Effects
lated that these proteins might act together as on the Clock
the transcription factor that directly regulates
light-inducible genes, with WC-1 being the actual The ambient temperature at which Neurospora
photoreceptor. This foresight was verified when is growing can influence its rhythmicity in three
WC-1 was shown to be the photoreceptor for the ways. (1) Steps up or down in temperature will
circadian clock in Neurospora (Froehlich et al. reset the clock in a manner similar to light pulses.
2002; He et al. 2002), and WC-1 and WC-2 were (2) If it is too hot or too cold, the clock will
shown to interact in vivo at the frequency (frq) not run. (3) The period length of the clock is
promoter (Froehlich et al. 2003). Independent compensated so that it is roughly the same at
of its role in the clock, the WCC is responsible any temperature within the physiological range
for light regulation of the several percent of of operation. This is the phenomenon known as
the genome that responds to light (Lewis et al. “temperature compensation” mentioned above.
2002). Some temperature effects are mediated through
the amount, and perhaps kind of FRQ protein
made. As noted above, temperature influences
B. Control of the WC-1 Photoreceptor frq splicing (H. Colot, J. Loros and J. Dunlap,
unpublished data), and thereby both the total
Neurospora contains many blue light-inducible, amount of FRQ and the ratio of the two FRQ forms.
WCC-regulated genes, but not all are regulated in It was also noted above that the amount of FRQ
the same way. The initial burst of light-induced in the nucleus determines time of day. However,
gene expression is transient whether light is because there is more overall FRQ at higher
delivered as a pulse or is continuous, and the speed temperatures, and FRQ concentrations oscillate
and duration of the response is gene dependent. around higher levels at higher temperatures, the
The response may peak at 15 min, as with the frq “time of day” associated with a given number
gene, or between 90 min and 2 h, as with the eas of molecules of FRQ is different at different
(ccg-2) gene. As noted above, WC-1 is transiently temperatures. At the peak at 25 ◦ C, there are about
hyperphosphorylated in response to light, and 30 molecules of FRQ in the nucleus (Merrow et al.
this corresponds to the brief induction of gene 1997), less than this number at 20 ◦ C, and more
expression in its targets as well as being correlated at 30 ◦ C. Thus, a shift in temperature corresponds
with turnover of the photoreceptor. WC-1 also literally to a step to a different time and a shift
induces its own expression in response to light in the state of the clock, although initially no
(reviewed in Linden et al. 1999; Liu 2003). synthesis or turnover of components occur.
FRQ promotes WC-1 expression, and so there After the step, relative levels of frq and FRQ are
is a rhythm in WC-1 protein levels that is not quite assessed in terms of the new temperature and they
180° out of phase with that of FRQ: WC-1 levels respond rapidly, completing the resetting (Liu
reach a peak in the night as FRQ is reaching its et al. 1998).
trough. WC-1 is also regulated via protein–protein
interactions, and WC-1 and WC-2 have influence
on each other’s levels. Importantly, though, all these
proteins are subject to phosphorylation events that V. Output from the Clock
appear to affect each protein’s activities and in-
teractions. Among the pertinent kinases is PKC A. Clock-Regulated Biology
(Arpaia et al. 1999), which forms a transient in-
teraction with the Zn-finger domain of WC-1 that By allowing anticipation to a changing environ-
is lost immediately after light exposure, only to ment, the functional usefulness of a clock is that
return 2 h later. Over-expression of a catalytically it facilitates an organism’s optimal performance
dead PKC increases WC-1 levels whereas consti- according to the time of day. A major question
tutively active kinase results in decreased levels, in circadian biology, which work in Neurospora
consistent with a role of the kinase in determin- largely defined at the molecular level and to which
ing both activity and stability (Franchi et al. 2005). it has contributed greatly, is how the clock deliv-
Other data, however, suggest that PKC is not the ers information to regulate downstream biologi-
only kinase involved. cal processes. Historically, the macroscopic asexual
developmental process in Neurospora, macroconi- Much can be learned about the specific roles
diation, is the best understood circadian pheno- of the clock in an organism’s life through the iso-
type. Conidiation can be initiated by environmen- lation of rhythmic genes. Many processes in the
tal signals including blue light, desiccation, and biology of Neurospora are under clock regulation,
nutrient starvation as well as by endogenous sig- including both sexual as well as asexual devel-
nals from the circadian clock. It involves major opment, stress responses, and basic intermediary
morphological changes that require many novel metabolism. The eas (ccg-2) gene encodes the Neu-
gene products. Other physiological rhythms have rospora hydrophobin (Bell-Pedersen et al. 1992;
been described including the production of CO2 , Lauter et al. 1992), a class of small, highly expressed
lipid and diacylglycerol metabolism (e.g., Roeder and exported hydrophobic proteins found form-
et al. 1982; Lakin-Thomas and Brody 2000; Rams- ing the bundled rodlet layer on conidia and aerial
dale and Lakin-Thomas 2000), a number of enzy- hyphae. This hydrophobic coating allows fungal
matic activities (e.g., Hochberg and Sargent 1974; structures to emerge from a wet substrate into the
Martens and Sargent 1974), heat shock proteins air (Talbot 1999; Wessels 1999). The pheromone
(Rensing et al. 1987), and even growth rate (Sar- precursor gene ccg-4 (Bobrowicz et al. 2002) is
gent et al. 1966). clock-controlled, as are several other structural
A word of caution is appropriate here. Fungi genes involved in development, such as the ccg-
express a variety of rhythms, many of which are 13, con-6 and con-10 genes (Lauter et al. 1992; Zhu
not circadian according to the definition provided et al. 2001). Trehalose synthase, encoded by the
above. A major difficulty in the analysis of output ccg-9 gene, catalyzes the synthesis of the disaccha-
pathways has been distinguishing circadian clock ride trehalose important in protecting cells from
regulation from other metabolic or developmental environmental stresses. Additionally, inactivation
regulations, many of which may oscillate. Because of ccg-9 results in altered conidiophore morphol-
of the complex and coupled nature of cellular pro- ogy and abolishes circadian control of conidiation,
cesses, assessing the roles of the circadian clock in although the underlying clock continues to oper-
daily physiological changes requires care and it is ate normally (Shinohara et al. 2002). The ccg-12
important to remember that not every rhythm is gene is allelic to the Neurospora cmt gene that en-
a circadian rhythm, nor is every rhythmic process codes copper metallothionein (Bell-Pedersen et al.
controlling conidiation or other events necessarily 1996b). Also, ccg-1 (grg-1), whose mRNA levels can
a part of the circadian clock. reach 10% of the cell’s total mRNA, is induced by
heat shock and repressed by glucose (McNally and
Free 1988; Lindgren 1994; Garceau 1996). This sug-
B. Clock-Controlled Genes – ccgs gests clock control of a stress response. Induction of
heat shock and general stress responses, including
A premise made in the Neurospora system (Loros the induction of heat shock genes, have been as-
et al. 1989) that has proven to be universally true in sociated with conidiation (Hafker et al. 1998), and
all circadian systems examined is that daily clock possibly with resetting of the phase of the clock in
control of gene expression would be a major form Neurospora (Ruoff et al. 1999).
of regulation controlling clock output. The first sys- The first energy-harvesting enzyme in glycoly-
tematic screens for clock-regulated genes were persis, glyceraldehyde-3-phosphate dehydrogenase, is
formed in Neurospora, using subtractive hybridiza- encoded by ccg-7. Its enzymatic activity had been
tion to enrich for morning- and evening-specific shown to be rhythmic in developing cultures on
mRNAs (Loros et al. 1989). The genes identified solid medium (Hochberg and Sargent 1974) and is
were called clock-controlled genes or ccgs, a term now known to cycle, along with transcript abun-
that is now used generically in the circadian lit- dance, in a non-conidiating liquid culture system
erature. Neurospora ccgs have subsequently been (Shinohara et al. 1998). Interestingly, GAPDH has
identified using differential hybridization of time- subsequently been found to be clock regulated in
specific libraries (Bell-Pedersen et al. 1996b), high- other systems, including Gonyaulax (Fagan et al.
throughput cDNA sequencing (Zhu et al. 2001), and 1999) and mammals (Iwasaki et al. 2004). Circa-
cDNA microarrays (Nowrousian et al. 2003). Most dian regulation of central metabolic functions in
exhibit peak expression in the late night to morn- the cell had not been previously documented, in-
ing and, additionally, many are regulated by light dicating a more global role in cell regulation than
and other stimuli that initiate development. had been formerly appreciated.
C. Microarray Analysis of Light, Clock collaboration with Bruce Roe at the Univer-
and Temperature Regulation sity of Oklahoma (Zhu et al. 2001; http://
www.genome.ou.edu/fungal.html). The libraries
Genomic approaches involving microarray analy- were constructed from non-developing mycelia
sis have lent themselves well to examining the reg- grown in dark starvation conditions, and har-
ulation of large subsets of a genome. Using these vested at times equivalent to circadian morning
microarrays to understand clock and photobiolog- and evening. In all, 13,000 cDNA clones were se-
ical regulation has been a natural progression from quenced, yielding approximately 20,000 sequences
past studies. Neurospora microarrays have been that resolved to 1431 unique expressed sequence
used to examine the overall extent of circadian- tags (ESTs). Roughly one-third of the ESTs were
regulated gene expression, light-regulated genes, unique, the other two-thirds appearing in multiple
and temperature-responsive genes in relation to clones. When compared to the National Center for
the clock and even for the identification of genes Biotechnology Information database, nearly half
in a different species, an approach that should lend (710 ESTs) showed significant matches to genes of
itself to the discovery of light-and clock-regulated either known or unknown function. Surprisingly,
genes in other fungal species. the other half of the ESTs did not display significant
In order to examine the extent of light- sequence similarity with any previously identified
regulated gene expression under WC-1 control, gene. Somewhat disappointingly, northern analy-
microarrays containing 1764 clones were prepared sis of a randomly chosen subset of those ESTs that
(Lewis et al. 2002) from ESTs collected from three appeared in one but not both timed libraries found
cDNA libraries representing the mycelial, asexual only a small fraction to be rhythmic (four out of
and sexual stages of the Neurospora life cycle (Nel- 26 ESTs), suggesting that sequencing of 13,000
son et al. 1997). With the release of the Neurospora clones was not of a sufficient depth to uncover
genome sequence, the 1764 ESTs were found to genes whose expression was limited to either the
represent 1343 unique genes. Target cDNAs made morning or the evening (Zhu et al. 2001). The
from dark-grown and light-exposed mycelia har- information from this study yielded a UniGene set
vested between 30 min and 4 h after light exposure of about 1100 genes representing a little over 10%
were hybridized to the probes. Approximately 3% of the genome for array analysis.
of the genes, or 22 probe spots, showed a twofold or Microarrays representing the 1100 genes were
greater increase in expression with light exposure. used to identify novel clock-controlled genes
The microarrays were then hybridized with target on a broad scale, in addition to examining the
cDNAs generated from a wc-1 over-expressing temperature control of genes. Within this set,
strain containing the wc-1 gene under control of about 2%–6% of genes on the arrays (depending
the inducible qa-2 promoter (Cheng et al. 2001b), on the stringency of the data analysis) were
to determine if increased levels of WC-1 was suf- rhythmically expressed when the target RNA was
ficient to induce light-inducible genes. In this case, harvested from cultures grown in constant dark
about 7% or 65 unique genes displayed a twofold and temperature (Nowrousian et al. 2003). These
or greater increase in expression when wc-1 was numbers are similar to that seen in other organ-
highly expressed in the presence of inducer. Strik- isms; 6% in Arabidopsis, 1%–5% in Drosophila,
ingly, there was a very low overlap (four out of 22) and 2%–9% in mammals (Duffield 2003). In
between the light-induced gene set and the genes plants (Harmer et al. 2000), whole subsets of
induced by wc-1 over-expression. This suggested genes corresponding to metabolic pathways are
that most or all of the over-expressed set are direct rhythmically expressed. In Neurospora this was
or indirect targets of the WC-1 transcription factor not the case, with rhythmic genes corresponding
that are independent of light, and that elevated to isolated and, possibly, rate-limiting proteins
levels of WC-1 are not sufficient for inducing such as GAPDH. Clustering of clock-regulated
light-responsive gene expression (Lewis et al. genes within the genome was also not found in this
2002). This is consistent with WC-1’s dark-specific study or previously (Bell-Pedersen et al. 1996b;
role in circadian clock gene regulation. Nowrousian et al. 2003). Using the above study as
In other studies, the genes represented on a baseline, temperature shift-induced genes were
the arrays were derived from a previously stud- examined in clock wild-type and in frq-null strains
ied cDNA library (Bell-Pedersen et al. 1996b) lacking a wild-type circadian clock. As shown
whose inserts were extensively sequenced in with previous individual genes (Arpaia et al. 1993,
1995), light induction of gene expression does ing that up to 20% of genes in rapidly growing or
not require the clock, but unexpectedly and re- developing cultures might be rhythmic under these
markably, all of the temperature-responsive genes parameters. Previously identified ccgs 1, 2, 6, 7 and
identified did require a functional circadian clock 12 were all among these 145. A subset of genes out
for their temperature response (Nowrousian et al. of the 145 were subject to northern analysis and
2003). Cultures grown under a 5 ◦ C temperature all shown to cycle, providing further validation, as
cycle (12 h of 22° alternating with 12 h of 27°) had been done in previous studies. Examination of
revealed 14 genes as cyclically expressed. However, time-of-day expression showed peaks at all phases
none of these were either cyclic or temperature of the cycle. As had been found by previous studies,
inducible in the frq-null strain grown under the most peaked late night to early morning and, as
same conditions. Additionally, all 14 of these expected, all the previously identified ccgs fell
genes belonged to the clock-controlled genes into these classes. The classification of genes from
identified in dark-grown cultures. This suggests this and earlier ccg studies, according to known
an unanticipated role of the circadian circuitry or predicted function based on the Whitehead
as an environmental temperature sensor. There sequencing project (Galagan et al. 2003; http://
are important caveats to these conclusions, in www.genome.wi.mit.edu/annotation/fungi/neuro
that non-abundant transcripts are doubtlessly spora) is shown in Table 4.2. This revealed
under-represented in EST-based libraries and a wide variety of cellular processes to be
more thorough analysis and larger gene sets under clock regulation. The largest classes
should produce additional, as well as possibly included genes involved in metabolism, pro-
a different proportion of, rhythmic genes. It is well tein synthesis, cell signaling, and those of
known that on solid media, temperature-regulated unknown function. BIOPROSPECTOR (http://
conidial banding in Neurospora does not require www.bioprospector.stanford.edu) was used to
frq, suggesting that there must be temperature- look for possible promoter elements, thereby
inducible or repressible genes yet to be identified identifying an eight-nt element (TCTTGGCA)
or, a less likely option, that these developmental that appeared frequently in the late-night and
rhythms do not occur when mycelia are grown in early-day gene classes. This element matches the
liquid media. core sequence previously found to be necessary
An exciting and alternative explanation comes and sufficient for cycling of the morning-expressed
from microarray analyses by Correa et al. (2003) eas (ccg-2) gene (Bell-Pedersen et al. 1996a).
describing the identification of rhythmic genes Of the 145 genes, all but three displayed the
from frq-null strains, whose phase is affected by appropriate period lengths of about 22 versus 29 h,
loss of frq but not the ability to cycle. Using the depending on the strain from which the target
arrays described previously (Lewis et al. 2002), was derived. The other three genes displayed the
clock-controlled genes were sought using target shorter period length in the microarrays, even
cDNA representing mRNAs collected from 22-h when targeted with cDNA derived from the long
frq+ and 29-h frq7 strains. Each slide had two sets period mutant. Surprisingly, when assayed via
of the array and duplicate slides were hybridized, northern analysis, all three continued to cycle in
providing four replicate samples per gene to verify
cycling, half with a 22-h periodicity and half with
a 29-h periodicity. After filtering parameters were Table. 4.2. Functions of Neurospora ccgs (expanded from
used to eliminate poorly expressed genes, 760 Correa et al. 2003 and Nowrousian et al. 2003)
probes remained and three criteria for rhythmicity
Functional category No. of ccgs
were applied: in both the wild-type and frq7
targeted arrays, (1) there must be at least one Cell division 1
peak, as a night-peaking gene will do so only once Signaling/communication 16
during the 1.5-day harvest regime; (2) for those Cell structure/cytoskeleton 8
Cell defense 4
genes with two peaks, the period length must be Development 11
between 18 and 30 h; and (3) the peak-to-trough Gene regulation 5
amplitude must be a minimum of 1.5-fold (smaller Metabolism 42
than the generally twofold amplitude looked for Protein processing 10
Protein synthesis 33
by Nowrousian et al. 2003). This analysis yielded Unclassified 50
145 ccgs out of the 760 expressing probes, suggest-
a loss-of-function frq strain, suggesting control Gene sets as they are developed for various fungal
by an oscillator separate from the frq-based species. Importantly, cross-species microarray
clock. Also intriguing, phasing differences in hybridizations could be useful for examining the
peak expression between the wild-type and photobiology and circadian rhythmicity of other
frq7 -targeted microarrays were seen, indicating fungal species.
that the frq-based mechanism may additionally
influence these genes. The idea that the Neurospora
circadian system might involve multiple oscillators
was not in itself novel (Loros 1984; Dunlap 2004), VI. Conclusions
although molecular components or outputs of
non-frq oscillators have not previously been found. Steady advances in developing molecular technolo-
More recently, the circadianly regulated nit-4 gene, gies and characterizing the genome in the mycelial
encoding nitrate reductase, has been shown to fungus Neurospora crassa have yielded a wealth of
participate in an autoregulatory feedback network information about this important model research
and cycle in a non-temperature-compensated, and system. The sequence shows Neurospora, with ap-
therefore non-circadian manner in the absence proximately 11,000 genes, to be comparable in com-
of frq. However, in a wild-type strain the nit-4 plexity to animal genomes such as those of the fruit
ccg is under clock control (Christensen et al. fly and worm, with most genes having introns. Neu-
2004). rospora’s several genome scanning mechanisms has
resulted in a compact genome with few repeated
sequences or gene families, thereby underlining
D. Cross-Species Arrays the diversity of the coding sequences. A remark-
able 50% of the predicted genes show no signifi-
A successful first effort at cross-species microarray cant homology with other sequences deposited in
analysis has recently been attempted (Nowrousian GenBank, although this may change soon with fur-
et al. 2005). A UniGene set compiled from both ther sequencing of mycelial fungal genomes. A Pro-
EST sets described above was adopted for use in gram Project recently funded by the National In-
microarray analysis of gene expression for devel- stitutes of Health and awarded to a consortium of
opmental mutants in Sordaria macrospora. Closely Neurospora research laboratories is currently un-
related to Neurospora, Sordaria is an excellent dertaking the genome annotation, the global anal-
model for sexual development in the mycelial ysis of gene regulation via full-genome microar-
fungi, as it does not make asexual spores, thus ray chips, and the complete disruption of all cod-
allowing for clean interpretations of the function ing regions of the Neurospora genome in order to
of genes identified as developmental mutations. further the aims of the scientific community at
These species share a high degree of sequence large.
identity, with nucleic acid identity within coding Twenty years ago the circadian oscillatory
regions averaging 89.5% (Nowrousian et al. 2004). system in Neurospora was imagined simply as
Although under high-stringency hybridization three discrete components (input, oscillator, and
conditions the overall signal intensity may be output), and possibly to be based on a single
lower, the possibility of nonspecific hybridization feedback loop with outputs. Several genes in
also appeared to be lower, an important con- Neurospora were known to be light inducible
sideration for cDNA microarray hybridization through a possibly transcriptional mechanism.
analysis. In all, 2880 cDNA clones were spotted Although FRQ, WC-1 and WC-2, as well as their
on the array, with 1420 giving hybridization modifiers and regulators, are now known to be
signals significantly above background in five clock components, the observation that WC-1 also
of six replicate hybridizations. Of these, 172 is the circadian blue-light photoreceptor for all
individual differentially regulated genes were genes has clearly united aspects of input and the
found to be either up- or down-regulated in all clock. Another photoreceptor as well as output
three mutant strains used as targets, with more gene product, VVD, is known to act broadly
genes found to be regulated in a subset of the to influence both the phase and abundance of
mutants, equaling about 12% of genes examined. many of the other clock-controlled genes, and
These results point to the utility of cross-species additionally the frq gene. Non-circadian oscilla-
hybridizations using Neurospora and other Uni- tors such as the nitrate reductase oscillator can
proceed on their own or within the circadian Ballario P, Talora C, Galli D, Linden H, Macino G (1998)
system coordinated with the FRQ/WCC loop and Roles in dimerization and blue light photoresponse
perhaps other, as yet undescribed, feedback loop of the PAS and LOV domains of Neurospora crassa
WHITE COLLAR proteins. Mol Microbiol 29:719–729
oscillators. Beadle GW, Tatum EL (1945) Neurospora II. Methods of
Most of the aspects of Neurospora’s molecular producing and detecting mutations concerned with
and cell biology closely resemble those of animal nutritional requirements. Am J Bot 32:678–686
cells (Sogin 1994; Simpson and Roger 2002). Bell-Pedersen D, Dunlap JC, Loros JJ (1992) The Neurospora
circadian clock-controlled gene, ccg-2, is allelic to eas
Research in the past two decades on circadian and encodes a fungal hydrophobin required for forma-
rhythms has identified a number of parallels tion of the conidial rodlet layer. Genes Dev 6:2382–2394
between clocks in animals models like Drosophila Bell-Pedersen D, Dunlap JC, Loros JJ (1996a) Distinct cis-
and mice, and in many cases has foreshadowed acting elements mediate clock, light, and developmen-
tal regulation of the Neurospora crassa eas (ccg-2) gene.
findings pertinent to understanding clocks in other Mol Cell Biol 16:513–521
systems. Neurospora has many assets as a model Bell-Pedersen D, Shinohara M, Loros J, Dunlap JC (1996b)
system for other mycelial fungi – its excellent Circadian clock-controlled genes isolated from Neu-
genetics and molecular biology, and the ease with rospora crassa are late night to early morning specific.
which it can be cultured and maintained. There Proc Natl Acad Sci USA 93:13096–13101
Bobrowicz P, Pawlak R, Correa A, Bell-Pedersen D,
is every reason to expect work on Neurospora to Ebbole DJ (2002) The Neurospora crassa pheromone
continue to inform research on fungi and animals precursor genes are regulated by the mating type locus
in just the way a good model should. and the circadian clock. Mol Microbiol 45:795–804
Borkovich K, Alex L, Yarden O, Freitag M, Turner G,
Read N, Seiler S, Bell-Pedersen D, Paietta J, Plesofsky
Acknowledgements. We thank laboratory members for N et al. (2004) Lessons from the genome sequence
helpful suggestions on this manuscript. This work was of Neurospora crassa: tracing the path from genomic
supported by grants from the NIH MH44651, R37GM34985 blueprint to multicellular organism. Mol Microbiol
and PO1 GM068087, NSF MCB-0084509 and the Norris Rev 68:1–108
Cotton Cancer Center. Cheng P, Yang Y, Heintzen C, Liu Y (2001a) Coiled coil me-
diated FRQ-FRQ interaction is essential for circadian
clock function in Neurospora. EMBO J 20:101–108
Cheng P, Yang Y, Liu Y (2001b) Interlocked feedback loops
contribute to the robustness of the Neurospora circa-
References dian clock. Proc Natl Acad Sci USA 98:7408–7413
Cheng P, Yang Y, Gardner KH, Liu Y (2002) PAS domain-
mediated WC-1/WC-2 interaction is essential for main-
Aronson B, Johnson K, Loros JJ, Dunlap JC (1994a) Negative taining the steady-state level of WC-1 and the function
feedback defining a circadian clock: autoregulation in of both proteins in clock and light responses of Neu-
the clock gene frequency. Science 263:1578–1584 rospora. Mol Cell Biol 22:517–524
Aronson BD, Johnson KA, Dunlap JC (1994b) The circadian Christensen M, Falkeid G, Hauge I, Loros JJ, Dunlap JC,
clock locus frequency: a single ORF defines period Lillo C, Ruoff P (2004) A frq-independent nitrate re-
length and temperature compensation. Proc Natl Acad ductase rhythm in Neurospora crassa. J Biol Rhythms
Sci USA 91:7683–7687 19:280–286
Aronson BD, Lindgren KM, Dunlap JC, Loros JJ (1994c) Cogoni C, Macino G (1994) Suppression of gene expression
An efficient method of gene disruption in Neurospora by homologous transgenes. Antonie Van Leeuwenhoek
crassa and with potential for other filamentous fungi. 65:205–209
Mol Gen Genet 242:490–494 Cogoni C, Macino G (1997) Isolation of quelling-defective
Arpaia G, Loros JJ, Dunlap JC, Morelli G, Macino G (1993) (qde) mutants impaired in posttranscriptional
The interplay of light and the circadian clock: inde- transgene-induced gene silencing in Neurospora
pendent dual regulation of clock-controlled gene ccg-2 crassa. Proc Natl Acad Sci USA 94:10233–10238
(eas). Plant Physiol 102:1299–1305 Collett MA, Dunlap JC, Loros JJ (2001) Circadian clock-
Arpaia G, Loros JJ, Dunlap JC, Morelli G, Macino G (1995) specific roles for the light response protein WHITE
The circadian clock-controlled gene ccg-1 is induced COLLAR-2. Mol Cell Biol 21:2619-2628
by light. Mol Gen Genet 247:157–163 Collett MA, Garceau N, Dunlap JC, Loros JJ (2002) Light
Arpaia G, Cerri F, Baima S, Macino G (1999) Involvement of and clock expression of the Neurospora clock gene fre-
protein kinase C in the response of Neurospora crassa quency is differentially driven by but dependent on
to blue light. Mol Genet Genomics 262:314–322 WHITE COLLAR-2. Genetics 160:149–158
Ballario P, Macino G (1997) White collar proteins: PASsing Correa A, Lewis ZA, Greene AV, March IJ, Gomer RH, Bell-
the light signal in Neurospora crassa. Trends Microbiol Pedersen D (2003) Multiple oscillators regulate circa-
5:458–462 dian gene expression in Neurospora. Proc Natl Acad Sci
Ballario P, Vittorioso P, Magrelli A, Talora C, Cabibbo A, USA 100:13597–13602
Macino G (1996) White collar-1, a central regulator of Crosthwaite SC, Loros JJ, Dunlap JC (1995) Light-Induced
blue-light responses in Neurospora crassa, is a zinc- resetting of a circadian clock is mediated by a rapid
finger protein. EMBO J 15:1650–1657 increase in frequency transcript. Cell 81:1003–1012
Crosthwaite SC, Dunlap JC, Loros JJ (1997) Neurospora wc-1 Galagan J, Calvo S, Borkovich K, Selker E, Read N,
and wc-2: transcription, photoresponses, and the ori- FitzHugh W, Ma L-J, Smirnov N, Purcell S, Rehman B
gins of circadian rhythmicity. Science 276:763–769 et al. (2003) The genome sequence of the filamentous
Davis RH (2000) Neurospora: contributions of a model fungus Neurospora crassa. Nature 422:859–868
organism. Oxford University Press, Oxford, UK Garceau N (1996) Molecular and genetic studies on the frq
Degli Innocenti F, Russo VEA (1983) Photoinduction and ccg-1 loci of Neurospora. PhD Thesis, Dartmouth
of perithecia in Neurospora crassa by blue light. Medical School, Hanover, NH
Photochem Photobiol 37:49–51 Garceau N, Liu Y, Loros JJ, Dunlap JC (1997) Alternative ini-
Denault DL, Loros JJ, Dunlap JC (2001) WC-2 mediates WC- tiation of translation and time-specific phosphoryla-
1-FRQ interaction within the PAS protein-linked cir- tion yield multiple forms of the essential clock protein
cadian feedback loop of Neurospora crassa. EMBO J FREQUENCY. Cell 89:469–476
20:109–117 Gardner GF, Feldman JF (1980) The frq locus in Neurospora
Duffield GE (2003) DNA microarray analyses of circadian crassa: a key element in circadian clock organization.
timing: the genomic basis of biological time. J Neu- Genetics 96:877–886
roendocrinol 15:991–1002 Gorl M, Merrow M, Huttner B, Johnson J, Roenneberg T,
Dunlap JC (1999) Molecular bases for circadian clocks. Cell Brunner M (2001) A PEST-like element in FRE-
96:271–290 QUENCY determines the length of the circadian
Dunlap JC (2005) Blue light photoreceptors – beyond phy- period in Neurospora crassa. EMBO J 20:7074–7084
tochromes and cryptochromes. In: Schaefer E (ed) Greene AV, Keller N, Haas H, Bell-Pedersen D (2003) A circa-
Photomorphogenesis in plants and bacteria: function dian oscillator in Aspergillus spp. regulates daily devel-
and signal transduction mechanisms. Kluwer, Dor- opment and gene expression. Eukaryot Cell 2:231–237
drecht (in press) Hafker T, Techel D, Steier G, Rensing L (1998) Differential ex-
Dunlap JC, Loros JJ (2004) The Neurospora circadian system. pression of glucose-regulated (grp78) and heat-shock-
J Biol Rhythms 19:414–424 inducible (hsp70) genes during asexual development
Dunlap JC, Loros JJ (2005) Neurospora photoreceptors. In: of Neurospora crassa. Microbiology 144:37–43
Briggs WR, Spudich J (eds) Handbook of photosensory Harding R, Melles S (1983) Genetic analysis of phototropism
receptors. Wiley, Berlin, pp 371–389 of Neurospora crassa perithecial beaks using white col-
Dunlap JC, Loros JJ, Decoursey P (eds) (2003) Chronobiol- lar and albino mutants. Plant Physiol 72:745–749
ogy: biological timekeeping. Sinauer, Sunderland, MA Harding RW, Shropshire WJ (1980) Photocontrol of
Fagan T, Morse D, Hastings JW (1999) Circadian synthesis carotenoid biosynthesis. Annu Rev Plant Physiol
of a nuclear-encoded chloroplast glyceraldehyde- 31:217–238
3-phosphate dehydrogenase in the dinoflagellate Harmer SL, Hogenesch JB, Straume M, Chang H-S, Han B,
Gonyaulax polyedra is translationally controlled. Zhu T, Wang X, Kreps JA, Kay SA (2000) Orchestrated
Biochemistry 38:7689–7695 transcription of key pathways in Arabidopsis by the
Feldman JF (1967) Lengthening the period of a biological circadian clock. Science 290:2110–2113
clock in Euglena by cycloheximide, an inhibitor of pro- Hastings JW (1960) Biochemical aspects of rhythms: phase
tein synthesis. Proc Natl Acad Sci USA 57:1080–1087 shifting by chemicals. Cold Spring Harbor Symp Quant
Feldman JF, Atkinson CA (1978) Genetic and physiologi- Biol 25:131–143
cal characterization of a slow growing circadian clock Hastings JW, Sweeney BM (1958) A persistent diurnal
mutant of Neurospora crassa. Genetics 88:255–265 rhythm of luminescence in Gonyaulax polyedra. Biol
Feldman JF, Hoyle M (1973) Isolation of circadian clock Bull 115:440–448
mutants of Neurospora crassa. Genetics 75:605–613 He Q, Cheng P, Yang Y, Wang L, Gardner K, Liu Y (2002)
Feldman J, Hoyle MN (1974) A direct comparison between WHITE COLLAR-1, a DNA binding transcription fac-
circadian and noncircadian rhythms in Neurospora tor and a light sensor. Science 297:840–842
crassa. Plant Physiol 53:928–930 Heintzen C, Loros JJ, Dunlap JC (2001) VIVID, gating and
Feldman JF, Waser N (1971) New mutations affecting circa- the circadian clock: the PAS protein VVD defines
dian rhythmicity in Neurospora. In: Menaker M (ed) a feedback loop that represses light input pathways
Biochronometry. National Academy of Sciences, Wash- and regulates clock resetting. Cell 104:453–464
ington, DC, pp 652–656 Hochberg ML, Sargent ML (1974) Rhythms of enzyme activ-
Feldman JF, Gardner GF, Dennison RA (1979) Genetic anal- ity associated with circadian conidiation in Neurospora
ysis of the circadian clock of Neurospora. In: Suda M crassa. J Bacteriol 120:1164–1175
(ed) Biological rhythms and their central mechanism. Iwasaki T, Nakahama K, Nagano M, Fujioka A, Ohyanagi H,
Elsevier, Amsterdam, pp 57–66 Shigeyoshi Y (2004) A partial hepatectomy results in
Franchi L, Fulci V, Macino G (2005) Protein kinase C altered expression of clock-related and cyclic glycer-
modulates light responses in Neurospora by regulating aldehyde 3-phosphate dehydrogenase (GAPDH) genes.
the blue light photoreceptor WC-1. Mol Microbiol Life Sci 74:3093–3102
56(2):334–345 Klemm E, Ninneman H (1978) Correlation between ab-
Froehlich AC, Loros JJ, Dunlap JC (2002) WHITE COLLAR-1, sorbance changes and a physiological response in-
a circadian blue light photoreceptor, binding to the duced by blue light in Neurospora crassa. Photochem
frequency promoter. Science 297:815–819 Photobiol 28:227–230
Froehlich AC, Loros JJ, Dunlap JC (2003) Rhythmic binding Kloss B, Price JL, Saez L, Blau J, Rothenfluh A, Wesley CS,
of a WHITE COLLAR containing complex to the fre- Young MW (1998) The Drosophila clock gene double-
quency promoter is inhibited by FREQUENCY. Proc time encodes a protein closely related to human casein
Natl Acad Sci USA 100:5914–5919 kinase Ie. Cell 94:97–107
Konopka RJ, Benzer S (1971) Clock mutants of Drosophila Loros JJ, Denome SA, Dunlap JC (1989) Molecular cloning
melanogaster. Proc Natl Acad Sci USA 68:2112–2116 of genes under the control of the circadian clock in
Kramer C, Loros JJ, Dunlap JC, Crosthwaite SK (2003) Role Neurospora. Science 243:385–388
for antisense RNA in regulating circadian clock func- Luo C, Loros JJ, Dunlap JC (1998) Nuclear localization is
tion in Neurospora crassa. Nature 421:948–952 required for function of the essential clock protein
Lakin-Thomas P (1996) Effects of choline depletion on the Frequency. EMBO J 17:1228–1235
circadian rhythm in Neurospora crassa. Biol Rhythm Martens CL, Sargent ML (1974) Conidiation rhythms of nu-
Res 27:12–30 cleic acid metabolism in Neurospora crassa. J Bacteriol
Lakin-Thomas PL, Brody S (2000) Circadian rhythms in 117:1210–1215
Neurospora crassa. Proc Natl Acad Sci USA 97:256–261 McClintock B (1950) The origin and behavior of mutable
Lauter F-R (1996) Molecular genetics of fungal photobiol- loci in maize. Proc Natl Acad Sci USA 36(6):344–355
ogy. J Genet 75:375–386 McNally M, Free S (1988) Isolation and characterization
Lauter F, Russo V, Yanofsky C (1992) Developmental and of a Neurospora glucose repressible gene. Curr Genet
light regulation of eas, the structural gene for the rodlet 14:545–551
protein of Neurospora. Genes Dev 6:2373–2381 Mehra A, Morgan L, Bell-Pedersen D, Loros J, Dunlap JC
Lee K, Loros JJ, Dunlap JC (2000) Interconnected feed- (2002) Watching the Neurospora clock tick. In: Abstr
back loops in the Neurospora circadian system. Science Vol Conf Society for Research on Biological Rhythms,
289:107–110 22–25 May 2002, Amelia Island, FL, pp 27
Lewis M, Feldman JF (1997) Evolution of the frequency clock Merrow M, Dunlap JC (1994) Intergeneric complementation
locus in ascomycete fungi. Mol Biol Evol 13:1233–1241 of a circadian rhythmicity defect: phylogenetic conser-
Lewis M, Morgan L, Feldman JF (1997) Cloning of (frq) clock vation of the 989 amino acid open reading frame in the
gene homologs from the Neurospora sitophila and Neu- clock gene frequency. EMBO J 13:2257–2266
rospora tetrasperma, Chromocrea spinulosa and Lep- Merrow M, Garceau N, Dunlap JC (1997) Dissection of
tosphaeria australiensis. Mol Gen Genet 253:401–414 a circadian oscillation into discrete domains. Proc Natl
Lewis ZA, Correa A, Schwerdtfeger C, Link KL, Xie X,
Acad Sci USA 94:3877–3882
Gomer RH, Thomas T, Ebbole DJ, Bell-Pedersen D
Merrow M, Franchi L, Dragovic Z, Gorl M, Johnson J, Brun-
(2002) Overexpression of White Collar-1 (WC-1) ac-
ner M, Macino G, Roenneberg T (2001) Circadian reg-
tivates circadian clock-associated genes, but is not suf-
ulation of the light input pathway in Neurospora crassa.
ficient to induce most light-regulated gene expression
EMBO J 20:307–315
in Neurospora crassa. Mol Microbiol 45:917–931
Linden H, Macino G (1997) White collar-2, a partner in Morgan LW, Greene AV, Bell-Pedersen D (2003) Circa-
blue-light signal transduction, controlling expression dian and light-induced expression of luciferase in
of light-regulated genes in Neurospora crassa. EMBO J Neurospora crassa. Fungal Genet Biol 38:327–332
16:98–109 Nakashima H (1981) A liquid culture system for the bio-
Linden H, Ballario P, Macino G (1997) Blue light regulation chemical analysis of the circadian clock of Neurospora.
in Neurospora crassa. Fungal Genet Biol 22:141–150 Plant Cell Physiol 22:231–238
Linden H, Ballario P, Arpaia G, Macino G (1999) Seeing the Nelson MA, Kang S, Braun E, Crawford M, Dolan P,
light: news in Neurospora blue light signal transduc- Leonard P, Mitchell J, Armijo A, Bean L, Blueyes E et al.
tion. Adv Genet 41:35–54 (1997) Expressed sequences form conidial,mycelial,
Lindgren KM (1994) Characterization of ccg-1, a clock- and sexual stages of Neurospora. Fungal Genet Biol
controlled gene of Neurospora crassa. PhD Thesis, 21:348–363
Dartmouth Medical School, Hanover, NH Ninomiya Y, Suzuki K, Ishii C, Inoue H (2004) Highly ef-
Liu Y (2003) Molecular mechanisms of entrainment in the ficient gene replacements in Neurospora strains defi-
Neurospora circadian clock. J Biol Rhythms 18:195–205 cient for nonhomologous end-joining. Proc Natl Acad
Liu Y, Garceau N, Loros JJ, Dunlap JC (1997) Thermally Sci USA 101:12248–12253
regulated translational control mediates an aspect of Nowrousian M, Duffield GE, Loros JJ, Dunlap JC (2003) The
temperature compensation in the Neurospora circa- frequency gene is required for temperature-dependent
dian clock. Cell 89:477–486 regulation of many clock-controlled genes in Neu-
Liu Y, Merrow M, Loros JJ, Dunlap JC (1998) How tem- rospora crassa. Genetics 164:922–933
perature changes reset a circadian oscillator. Science Nowrousian M, Wurtz C, Poggeler S, Kuck U (2004) Com-
281:825–829 parative sequence analysis of Sordaria macrospora and
Liu Y, Loros J, Dunlap JC (2000) Phosphorylation of the Neurospora crassa as a means to improve genome an-
Neurospora clock protein FREQUENCY determines its notation. Fungal Genet Biol 41:285–292
degradation rate and strongly influences the period Nowrousian M, Ringelberg C, Dunlap J, Loros J, Kück U
length of the circadian clock. Proc Natl Acad Sci USA (2005) Cross-species microarray hybridization to
97:234–239 identify developmentally regulated genes in the
Loros JJ (1984) Studies on frq-9, a recessive circadian clock filamentous fungus Sordaria macrospora. Mol Genet
mutant of Neurospora crassa. PhD Thesis, University Genomics 273:137–149
of California Santa Cruz, Santa Cruz, CA Onai K, Nakashima H (1997) Mutation of the cys-9 gene,
Loros JJ, Dunlap JC (2001) Genetic and molecular analysis which encodes thioredoxin reductase, affects the cir-
of circadian rhythms in Neurospora. Annu Rev Physiol cadian conidiation rhythm in Neurospora crassa. Ge-
63:757–794 netics 146:101–110
Loros JJ, Richman A, Feldman JF (1986) A recessive circa- Perkins DD (1988) Photoinduced carotenoid synthesis in
dian clock mutant at the frq locus in Neurospora crassa. perithecial wall tissue of Neurospora crassa. Fungal
Genetics 114:1095–1110 Genet Newslett 35:38–39
Perlman J, Nakashima H, Feldman J (1981) Assay and char- Shigeyoshi Y, Taguchi K, Yamamoto S, Takeida S, Yan L,
acteristics of circadian rhythmicity in liquid cultures Tei H, Moriya S, Shibata S, Loros JJ, Dunlap JC et al.
of Neurospora crassa. Plant Physiol 67:404–407 (1997) Light-induced resetting of a mammalian circa-
Potapova T, Levina N, Belozerskaya T, Kritsky M, dian clock is associated with rapid induction of the
Chailakhian L (1984) Investigation of electrophysio- mPer1 transcript. Cell 91:1043–1053
logical responses of Neurospora crassa to blue light. Shinohara M, Loros JJ, Dunlap JC (1998) Glyceraldehyde-3-
Arch Microbiol 137:262–265 phosphate dehydrogenase is regulated on a daily basis
Price JL, Blau J, Rothenfluh A, Adodeely M, Kloss B, by the circadian clock. J Biol Chem 273:446–452
Young MW (1998) double-time is a new Drosophila Shinohara ML, Correa A, Bell-Pedersen D, Loros JJ, Dun-
clock gene that regulates PERIOD protein accumula- lap JC (2002) The Neurospora crassa clock-controlled
tion. Cell 94:83–95 gene-9 (ccg-9) encodes a novel form of trehalose syn-
Ramsdale M, Lakin-Thomas PL (2000) sn-1,2-Diacyl- thase required for circadian-regulated conidiation. Eu-
glycerol levels in the fungus Neurospora crassa karyot Cell 1:33–43
display circadian rhythmicity. J Biol Chem 275:27541– Shiu PKT, Raju N, Zickler D, Metzenberg R (2001) Meiotic
27550 silencing of unpaired DNA. Cell 107:905–916
Rensing L, Bos A, Kroeger J, Cornelius G (1987) Possible link Siegel RW, Matsuyama S, Urey J (1968) Induced macro-
between circadian rhythm and heat shock response in conidia formation in Neurospora crassa. Experiencia
Neurospora crassa. Chronobiol Int 4:543–549 24:1179–1181
Roeder PE, Sargent ML, Brody S (1982) Circadian rhythms Simpson A, Roger AJ (2002) Eukaryotic evolution: getting
in Neurospora crassa: oscillations in fatty acids. Bio- to the root of the problem. Curr Biol 12:R691–R693
chemistry 21:4909–4916 Sogin ML (1994) The origin of eukaryotes and evolution
Ruoff P, Vinsjevik M, Mohsenzadeh S, Rensing L (1999) The into major kingdoms. In: Bengston S (ed) Early life on
Goodwin Oscillator: on the importance of degradation earth. Nobel Symposium no 84. Columbia University
reaction in the circadian clock. J Biol Rhythms 14:469– Press, New York, pp 181–192
479 Talbot N (1999) Fungal biology. Coming up for air and
Ryan FJ, Beadle GW, Tatum EL (1943) The tube method for sporulation. Nature 398:295–296
measuring the growth rate of Neurospora. Am J Bot Talora C, Franchi L, Linden H, Ballario P, Macino G (1999)
30:784–799 Role of a white collar-1-white collar-2 complex in blue-
Sargent ML, Briggs WR (1967) The effect of light on a circa- light signal transduction. EMBO J 18:4961–4968
dian rhythm of conidiation in Neurospora. Plant Phys- Wessels JG (1999) Fungi in their own right. Fungal Genet
iol 42:1504–1510 Biol 27:134–145
Sargent ML, Briggs WR, Woodward DO (1966) The cir- Yang Y, Cheng P, Zhi G, Liu Y (2001) Identification of
cadian nature of a rhythm expressed by an invertase- a calcium/calmodulin-dependent protein kinase that
less strain of Neurospora crassa. Plant Physiol 41:1343– phosphorylates the Neurospora circadian clock protein
1349 Frequency. J Biol Chem 276:41064–41072
Schwerdtfeger C, Linden H (2003) VIVID is a flavoprotein Yang Y, Cheng P, Liu Y (2002) Regulation of the Neurospora
and serves as a fungal blue light photoreceptor for circadian clock by casein kinase II. Genes Dev 16:994–
photoadaptation. EMBO J 22:4846–4855 1006
Selker EU (1990) Premeiotic instability of repeated se- Zhu H, Nowrousian M, Kupfer D, Colot H, Berrocal-Tito G,
quences in Neurospora crassa. Annu Rev Genet Lai H, Bell-Pedersen D, Roe B, Loros JJ, Dunlap JC
24:579–613 (2001) Analysis of expressed sequence tags from
Selker EU, Cambareri EB, Jensen BC, Haack KR (1987) Re- two starvation, time-of-day-specific libraries of
arrangement of duplicated DNA in specialized cells of Neurospora crassa reveals novel clock-controlled
Neurospora. Cell 51:741–752 genes. Genetics 157:1057–1065
5 Genomics of Protein Secretion and Hyphal Growth in Aspergillus
D.B. Archer1 , G. Turner2
CONTENTS secretion of proteins and hyphal growth, and

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 75 especially where genomic approaches have been
II. Generic Genomic Methodologies . . . . . . . . . 76 applied. Thus, the chapter will emphasise A.
A. Targeted Gene Manipulation . . . . . . . . . . 77 nidulans, A. fumigatus, A. niger and A. oryzae
B. Other Approaches to Gene Disruption . . . 78 (and their close relatives). A. nidulans is often
C. Gene Identification by Complementation 79
D. Expression Cloning . . . . . . . . . . . . . . . . . 79 regarded as the model filamentous fungus for
E. Transcriptomics . . . . . . . . . . . . . . . . . . . . 79 genetic studies and, indeed, it has proven to be an
F. Proteomics . . . . . . . . . . . . . . . . . . . . . . . . 80 exceptionally useful and interesting fungus since
G. Metabolite Profiling . . . . . . . . . . . . . . . . . 81 the pioneering studies on Aspergillus genetics
III. Secretion Genes in Aspergillus spp. . . . . . . . . 81
A. Entry of Proteins into the Secretory (Pontecorvo et al. 1953). Advances in Aspergillus
Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . 81 genetics made with A. nidulans have been facil-
B. Protein Folding in the ER . . . . . . . . . . . . . 82 itated by it having a sexual cycle. A. niger and
C. Post-ER Secretion . . . . . . . . . . . . . . . . . . . 83 A. oryzae are asexual but are species which are
D. ER to Golgi . . . . . . . . . . . . . . . . . . . . . . . . 83 used commercially for the production of enzymes,
E. Golgi to Plasma Membrane . . . . . . . . . . . 84
F. Polarity Establishment . . . . . . . . . . . . . . . 84 metabolites and fermented food products (Archer
G. The Spitzenkörper . . . . . . . . . . . . . . . . . . 85 2000). A. fumigatus is a common fungus in some
H. The Exocyst . . . . . . . . . . . . . . . . . . . . . . . 85 environments (including compost heaps), but it
IV. Secreted Proteins from Aspergillus . . . . . . . . 86 can also be a serious allergen and human pathogen,
A. Native Proteins . . . . . . . . . . . . . . . . . . . . . 86
B. Heterologous Enzymes . . . . . . . . . . . . . . . 87 particularly for immuno-compromised people.
C. The Stress of Protein Secretion . . . . . . . . 88 The genomes from all four of these Aspergillus
D. Polar Growth and Secretion . . . . . . . . . . . 89 species have recently been sequenced (Table 5.1),
E. Hyphal Branching and Polarity . . . . . . . . 90 thus making formal genomics approaches to the
F. Branching Frequency . . . . . . . . . . . . . . . . 90 study of protein secretion and hyphal growth more
V. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 feasible. Selected data on the genomes of each
Aspergillus species are included in Table 5.2. Each
species has a G+C mole percentage of about 50%,
and between 45 and 50% of the sequenced DNA is
Abbreviations: ER, endoplasmic reticulum; EST,
predicted to be coding sequence.
expressed sequence tag; PPI, peptidylprolyl cis-
The Aspergillus genome sequences have been
trans isomerase; SRP, signal recognition particle;
analysed using annotation software, and then an-
UPR, unfolded protein response
notated manually with input from scientists world-
wide. The outputs from the sequencing and an-
I. Introduction notation efforts in Aspergillus follow the detailed
analyses of the genome of Neurospora crassa (Gala-
Aspergillus is a genus which includes over 180 gan et al. 2003; Borkovich et al. 2004). There is
species of filamentous fungi (Pitt et al. 2000). now a growing resource of fungal genome sequence
However, this review will focus on those species data which include yeasts and filamentous fungi.
for which there is relevant information on the The genome sequence of Phanerochaete chrysospo-
1
rium, a white-rot basidiomycete, has also recently
School of Biology, University of Nottingham, University Park,
Nottingham NG7 2RD, UK
been published (Martinez et al. 2004; Teeri 2004).
2 Department of Molecular Biology and Biotechnology, University That species has an extraordinary capacity for the
of Sheffield, Firth Court, Western bank, Sheffield S10 2TN, UK depolymerisation and degradation of wood com-
The Mycota XIII
Fungal Genomics
76 D.B. Archer and G. Turner
Table. 5.1. Websites for As- Species Website URL address

pergillus genome informa-
tion (modified after Archer A. fumigatus http://www.tigr.org/tdb/e2k1/afu1/
and Dyer 2004) http://www.sanger.ac.uk/Projects/A_fumigatus/
A. nidulans http://www.broad.mit.edu/annotation/fungi/aspergillus/
A. niger http://www.dsm.com/dfs/innovation/genomics/a
http://www.integratedgenomics.com/products.htmla
A. oryzae Whole genome not yet publicly availableb
http://www.aist.go.jp/RIODB/ffdb/welcome.html
General Aspergillus sites http://www.aspergillus.man.ac.uk/
http://www.fgsc.net/aspergenome.htm
http://www.aspergillus-genomics.org/
http://www.genome.ou.edu/fungal.html
http://www.cadre.man.ac.uk/
a These sites provide an entry point to the sequence data but access requires agreement
being reached with the respective company
b Genome sequence data for A. oryzae are not yet available. A summary of the A. oryzae
genome sequencing project and related topics has been published (Machida 2002)
Table. 5.2. Summary of information derived from the genome sequences of Aspergillus speciesa
A. fumigatusb A. nidulansc A. nigerd A. oryzaee
Genome size (Mb)f 29 30 36 37

Predicted genesf 10,000 10,000 14,000 14,000
Genes with Pfam hitsf 4,400 4,500 5,300 5,300
a This table was published in extended form elsewhere (Archer and Dyer 2004), and the data were obtained from the
websites and contributions from William Nierman (TIGR, USA), James Galagan (Broad Institute, USA), Masa Machida
(Tsukuba, Japan), and Gert Groot and Noel van Peij (DSM, The Netherlands)
b http://www.tigr.org/tdb/e2k1/afu1/
c http://www.broad.mit.edu/annotation/fungi/aspergillus/
d http://www.dsm.com/dfs/innovation/genomics/
e Not yet publicly available. Data from M. Machida
f Approximately. Current predicted numbers have been rounded
ponents, and provides a resource of secreted fungal believe that post-genomic approaches can be used
enzymes which add to the range of activities avail- to extend our knowledge even further. It is not our
able from the ascomycetous Aspergillus spp. which attention to reiterate a review of protein secretion,
do not degrade wood. and we refer to the cited reviews for those data.
The purpose of this chapter is to discuss the Other chapters in this volume address issues
recent advances in protein secretion and hyphal pertinent to this chapter, as does a recent summary
development in Aspergillus spp. in the context of of the move from genomics to post-genomics in
the newly available genome sequence data and Aspergillus (Archer and Dyer 2004).
their annotation. There have been many recent
reviews published which have discussed, in par-
ticular, the secretion of proteins (and, especially,
heterologous proteins) from Aspergillus spp. II. Generic Genomic Methodologies
(Conesa et al. 2001; Punt et al. 2002; MacKenzie
et al. 2004) but none of them had the advantage of Putative homologues of many of the known
comprehensive annotated genome sequences, even genes of the secretory pathway have been found
if some genome data were available. Our aims in in Aspergillus species and in N. crassa, though
this chapter are therefore to highlight areas where functional analysis is still limited. Although many
we have gained additional understanding from relevant techniques have been developed in the
the genome sequences, and to indicate where we pre-genomics era, improvements will be needed
Aspergillus Protein Secretion and Hyphal Growth 77
to reflect the need to deal with the challenges of the probable secretion pathway of fungi on the as-
functional genomics. sumption of a similar system.
Mutants defective in polar growth, septation, Methods for transformation and gene manip-
or with aberrant hyphal morphology have been se- ulation of Aspergillus species have been fully
lected from temperature-sensitive mutant screens reviewed elsewhere (Turner 1994; Riach and
in both A. nidulans and N. crassa, two of the ma- Kinghorn 1996; Brookman and Denning 2000),
jor genetic models for filamentous fungi (Harris but the availability of several genome sequences
et al. 1994; Seiler and Plamann 2003). Some of the presents new challenges for devising high-
genes affected have been characterised, and they throughput approaches.
often turn out to be genes involved in secretion
and polar growth in other species. In addition,
A. Targeted Gene Manipulation
genes are found for which no function has been
previously described in any species (Gatherar et al. Functional analysis of yeast genes is highly depen-
2004; Pearson et al. 2004). This approach is com- dent on homologous recombination to target suit-
plementary to the search for homologues of known able vector constructs to genes of interest. This can
secretion/polarity genes, and is essential if we are be achieved easily in Saccharomyces cerevisiae us-
to understand how growth and secretion in fila- ing cassette vectors as templates for PCR, since the
mentous fungi differs from the yeast paradigm. addition of short oligonucleotides (about 50 nt) at
We will discuss examples of how genes identi- each end of the PCR product provides sufficient
fied by both approaches may act in secretion and homology for gene targeting (reviewed in Wend-
polar growth, with reference to Fig. 5.1. The Golgi land 2003). Homologous recombination in filamen-
apparatus of fungi is not as well characterised as tous fungi varies with the species, but usually re-
that of mammalian cells, and is often referred to quires longer stretches of homologous sequence
as the “Golgi-equivalent”. However, we will trace to achieve efficient targeting. A modified cassette
template approach which uses recombinant PCR
(fusion PCR; Krawchuck and Wahls 1999) can be
used with Aspergillus species, and provides from
500 bp or more of flanking regions (Yang et al. 2004;
Yu et al. 2004; Zarrin et al. 2005).
While this opens the way to more extensive
genome analysis, it is not easy to detect the
terminal phenotype of a deletion mutant if the
gene is essential. Stable diploid strains can be
constructed from all species of Aspergillus using
the parasexual cycle but, unlike the case in S. cere-
visiae, these are asexual diploids which cannot be
haploidised by meiosis. This precludes the use of
tetrad analysis of heterozygous mutants to observe
the terminal phenotype. Though such diploids can
Fig. 5.1. A representation of the key sub-cellular compart- be haploidised using drugs which cause mitotic
ments which are important in the secretion of proteins by non-disjunction, the recovery of a haploid segre-
a fungal hypha. N Nucleus, ER endoplasmic reticulum, P
proteasome, V vacuole, G Golgi or equivalent. Polypep- gant requires growth of a sector, and so failure to
tides (co-translational) or proteins (post-translational) are recover the mutant haploid is the only indication of
directed to the ER by the N-terminal secretion signal. Pro- lethality/severe growth impairment. Nevertheless,
tein folding, with the formation of disulphide bonds, is such approaches have been adopted to screen for
assisted in the lumen of the ER by chaperones and foldases essential genes in the opportunistic pathogen A.
before vectorial transport to the exterior of the cell. Post-
translational modifications (e.g. glycosylation, protein pro- fumigatus, with a view to the identification of drug
cessing) occur during this process. There are retrieval mech- targets (Firon and d’Enfert 2002).
anisms for the recovery of ER-resident proteins, and pro- A similar approach for essential genes uses
teins which do not satisfy the quality control checks for deletion in heterokaryons, where nuclei carry-
correct fold may be retro-translocated to the cytosol for
degradation by the proteasome. The figure was produced ing both wild-type and deleted alleles can be
by Adrian Watson, and is a modification of similar figures detected by Southern blotting following deletion
produced by several of our colleagues by transformation (Oakley et al. 1990). Failure to
recover a pure, haploid deleted strain from the het- approach could now be extended to Aspergillus
erokaryon is taken as evidence for the gene being species.
essential for viability. In some cases, it is possible to While multiple mutants can be constructed
infer the terminal phenotype of the deletion mutant in A. nidulans using sexual crossing, this is more
by plating out the mixture of uninucleate conidia difficult in the asexual aspergilli. However, this has
which arise from the heterokaryotic mycelium. been partially overcome by using the BLASTER
Essential genes can be investigated in A. nidu- approach to gene disruption, where a two-way
lans and A. fumigatus using the conditional pro- selectable marker such as the pyr-4/pyrG gene
moter of the alcA (alcohol dehydrogenase) gene of (orotidylate decarboxylase) is flanked by sequence
A. nidulans (Waring et al. 1989). This gene is in- repeats, and excision events can be selected on
duced by ethanol, threonine and cyclopentanone, fluoro-orotic acid. This permits reuse of the
and repressed by sucrose and glucose (Felenbok same marker for further disruptions (d’Enfert
et al. 2001). The alcA promoter can be used for rapid 1996).
promoter exchange with any gene using a PCR-
based approach (Fig. 5.2). For protein localisa-
B. Other Approaches to Gene Disruption
tion, GFP and RFP have also been in use for some
time (Fernandez-Abalos et al. 1998; Tavoularis et al. Where disruption of genes is difficult using the di-
2001; Su et al. 2004; Toews et al. 2004). Recently, rect transformation approach, an alternative, two-
the difficulty of visualising actin structures, impor- step method is to disrupt the gene of interest in
tant for cell biology studies, and which cannot be a cosmid clone, and then take advantage of the large
stained by phalloidin as in yeast, was overcome by fungal DNA insert to provide very long stretches
using a GFP-tropomyosin fusion, the gene (tpmA) of homology to improve targeting during fungal
having been found by searching the genome se- transformation (Chaveroche et al. 2000; Langfelder
quence of A. nidulans (Pearson et al. 2004). Thus, et al. 2002). Homologous recombination with the
the tools now exist for construction of a variety clone insert can be achieved using an E. coli strain
of PCR cassettes for rapid analysis of gene func- expressing phage λ Red functions. The E. coli strain
tion and localisation. A valuable application of the carrying a cosmid with a known insert is trans-
PCR cassette approach in yeast has been the TAP- formed with a PCR product carrying selectable
Tag system, used to investigate protein–protein in- markers for E. coli (zeo–zeocin resistance) and the
teractions and to identify protein complexes on fungus (pyrG), and which has 50-bp flanking re-
a genome-wide scale (Gavin et al. 2000), and this gions homologous to the target site within the cos-
mid insert. The recombinant cosmid is then used
to transform the fungus.
Pseudo-random integration can be pro-
vided by transposon mutagenesis or random
non-homologous integration (REMI). REMI has
been attempted for gene function analysis but
has its limitations, since it can lead to major
rearrangements/deletions at the site of integration,
and multiple site integration, complicating inter-
pretation of the phenotype. Examples of the use
of REMI include tagging of multidrug resistance
genes in A. nidulans (de Souza et al. 2000),
where only 36% of the transformants carried the
integrating vector linked to the drug sensitivity
phenotype. REMI approaches have also been used
Fig. 5.2. Use of recombinant PCR for promoter replacement.
a Sequences flanking the promoter region of the gene of in A. niger and A. oryzae (Shuster and Connelley
interest are amplified. b A cassette carrying a selectable 1999; Yaver et al. 2000). It has also been used
marker and conditional promoter is amplified using the to validate signature-tagged mutagenesis in A.
fragments obtained in a, together with additional outer fumigatus, but the mutation frequency achievable
primers. c The linear fragment resulting from b is used
to transform a Pyr− strain of A. nidulans to Pyr+ . d The by transformation was too low to be of practical
homologous integrant is identified by PCR and Southern use for the kind of large-scale screens carried out
analysis (Zarrin et al. 2005) in prokaryotes (Brown et al. 2000).
Transposition within a bacterial host has been enzymes can be rapidly identified. This approach
used by exploiting bacterial transposons for inser- was used to sample secreted enzymes of industrial
tion into a clone, followed by identification of the interest from Aspergillus aculeatus (Dalboge 1997),
clone with the insertion by PCR, and then transfor- and is applicable to species for which no genome
mation of the fungus (Jadoun et al. 2004). Fungal sequence is available.
transposable elements are commonly found in the
genome, but most are inactive. An active trans-
posable element from Fusarium oxysporum, Fot1, E. Transcriptomics
has been used for insertional mutagenesis in A.
nidulans. Most insertions did not occur in cod- Transcriptome studies with Aspergillus nidulans
ing regions, limiting the use of the system (Nicosia are so far limited to glass slide arrays, with PCR
et al. 2001), though a different transposable ele- products designed from partial EST libraries prior
ment from F. oxysporum, impala160, was applied to completion of the genomic sequence. The slides
more successfully to A. fumigatus, where essential have been validated by studying gene expression
genes were identified by insertional mutagenesis of during a shift from ethanol to glucose (Sims et al.
a diploid strain (Firon et al. 2003). 2004a). Subsequently, further ESTs and some PCR
Finally, Agrobacterium has been used success- products (using primers designed from the genome
fully for plasmid transfer and genomic insertion of sequence) were added to the arrays. They were
Ti-based vectors in a number of filamentous fungal used in a comparative and functional genomics
species, including A. awamori and A. niger. Inser- study to predict functions of A. nidulans genes
tion can be by non-homologous or homologous and to enhance the automated annotation (Sims
integration (Groot et al. 1998; Gouka et al. 1999) et al. 2004b). Those arrays have been used further
but this approach has not yet been exploited as in a transcriptomic study of recombinant protein
an efficient tagging method suitable for functional secretion and the unfolded protein response (Sims
genomics. et al. 2005). Although the arrays were predicted to
represent up to ca. 30% of predicted ORFs, they
proved useful in showing that the transcript lev-
C. Gene Identification by Complementation els of a wide range of genes were affected under
secretion stress conditions. The value of express-
Cloning genes by complementation has always
ing a heterologous protein, rather than chemicals,
been much more difficult in A. nidulans than in
to induce stress was highlighted. No data have yet
S. cerevisiae, mostly because of relatively low and
been reported for A. fumigatus, but glass slide ar-
variable transformation frequencies obtained with
rays using 70mer oligonucleotides designed from
fungal protoplasts. The situation has improved
all of the open reading frames (about 10,000) in
recently with the generation of gene libraries made
the completed genome sequence have recently been
in AMA1-based replicating vectors, available from
constructed at TIGR (pfgrc.tigr.org).
the Fungal Genetics Stock Centre (Osherov et al.
The complete A. niger genome is represented
2002). As with S. cerevisiae, mutant screens and
on Affymetrix gene chips which are being used in
subsequent gene identification by complementa-
collaborative studies with DSM (DSM Food Spe-
tion are still essential tools for investigating gene
cialties, The Netherlands; Table 5.1). A generalised
function, even in the post-genomic era. Recent
approach for studying the transcriptomics of the
examples from the genetic models A. nidulans
responses due to stress associated with protein
(Gatherar et al. 2004; Pearson et al. 2004) and N.
secretion is presented in Fig. 5.3. A. niger ESTs have
crassa (Seiler and Plamann 2003) show that we
also been arrayed on slides (in separate studies
cannot expect to find all the components of polar
by T. Goosen, TNO, Netherlands Organisation for
growth and secretion required in filamentous
Applied Research, Zeist, The Netherlands, and
fungi simply by comparative genomics.
M. Gent, University of Manchester, UK, unpub-
lished data). The arrays produced at TNO were
D. Expression Cloning hybridised with fluorescent dye-labelled target
cDNA produced from RNA extracted from control
If a cDNA library derived from a filamentous fun- and dithiothreitol-stressed cells of A. niger. Several
gus is expressed in yeast, provided that a suitable up- and down-regulated transcripts were detected
enzyme assay is available, clones expressing novel in response to dithiothreitol, and the results will be
Fig. 5.3. An approach to the use of gene chips in tran- example, the stress responses due to the reducing agent
scriptomic studies of stress responses relevant to protein dithiothreitol include those which affect redox as well as
secretion by Aspergillus. The images were obtained using those affecting protein secretion and the unfolded protein
Affymetrix chips (from DSM, The Netherlands) of the A. response. The ER stresses due either to strains secreting
niger genome, and we thank DSM for permission to repro- heterologous proteins or those exposed to other chemicals
duce them. We thank Thomas Guillemette for constructing which affect the secretion of proteins (e.g. tunicamycin or
the figure. The data from each oligonucleotide on each test brefeldin) may be separately analysed using gene chips and
chip are compared to that of the control. The compiled compared to those using other stress inducers. In addi-
data are subsequently represented as a ratio of signal intention, strains which are deleted for genes important in stress
sity, as well as a measure of intensity of the control signal responses (e.g. the hacA gene of Aspergillus) may also be
for each oligonucleotide (baseline signal). Different stresses compared using this approach
can then be compared to find overlapping responses. For
published elsewhere. One interesting finding was of A. flavus, a species not yet fully sequenced,
that the transcriptional down-regulation of the has been investigated by two-dimensional elec-
glaA gene in response to dithiothreitol, reported trophoresis and MALDI-TOF mass spectrometry
elsewhere (Al-Sheikh et al. 2004), was confirmed. following induction of secreted proteins by the
flavonoid rutin (Medina et al. 2004). Such an
approach, combined with genome analysis, should
F. Proteomics prove useful for cataloguing the secreted protein
repertoire of filamentous fungi, and monitoring
Proteomic studies on Aspergilli and other filame- changes resulting from mutation and growth
nous fungi are at an early stage. The exoproteome conditions.
G. Metabolite Profiling
A novel approach to the identification of secondary

metabolic pathway genes made use of transcrip-
tional and metabolic profiling (Askenazi et al.
2003). Though a genome sequence was unavailable,
a microarray of DNA fragments was constructed
for transcriptome analysis. Employing a series
of strains which had been previously engineered
to give different yields of lovastatin, association
analysis was used to determine gene expression
patterns which correlated with lovastatin yield.
This led to the identification of DNA fragments
positively associated with lovastatin yield, which Fig. 5.4. Prediction of the number of N-terminal secre-
included the previously isolated lovastatin gene tion signals and secretion anchors, using web-based pro-
grammes, in the genome of A. fumigatus. The analysis was
cluster. Additional genes associated with sec- done by J. Huang (TIGR, USA) and G. Robson (University of
ondary metabolite production were also identified. Manchester, UK). Signal peptide, signal peptide+Pfam,
This approach is especially useful where no signal anchor, signal anchor+Pfam (+Pfam indicates
complete genome sequence is available, which will those proteins with an assigned Pfam description)
remain the case for many of the estimated 100,000
or so known fungal species.
in this chapter) and endocytosis (which, although
not discussed here, is described in detail in N.
III. Secretion Genes in Aspergillus spp. crassa by Borkovich et al. 2004).
The proportion of predicted ORFs in the As-

pergillus genomes which are directly or indirectly A. Entry of Proteins into the Secretory Pathway
relevant to protein secretion and hyphal de-
velopment is not accurately known. Certainly, In the yeast S. cerevisiae, proteins destined for
annotations of the Aspergillus genome sequences entry into the secretory pathway can do so ei-
indicate that there is a wide range of secreted ther co- or post-translationally whereas in mam-
proteins, and that the secretory system itself is mals translocation appears to be, primarily at least,
complex and involves many proteins. The numbers co-translational. The key difference between co-
of genes, predicted to encode proteins containing and post-translational translocation is that only
either signal peptides (cleaved secretion signal) co-translational translocation of proteins across
or signal anchors (un-cleaved secretion signal, re- the ER membrane requires the signal recognition
taining the protein in a membrane) in the genomes particle (SRP) which recognises and binds to the
of A. fumigatus, A. nidulans and A. oryzae is signal (or leader) sequence at the N-terminus of
highly dependent on the confidence limit assigned proteins destined for entry into the secretory sys-
to the prediction. Data for the predicted signal tem. The post-translational translocation system
peptide/anchor proteins encoded in A. fumigatus is independent of the SRP and operates follow-
are presented in Fig. 5.4. At the 95% confidence ing the synthesis of proteins in the cytosol and
level of a correct assignment of a signal peptide, release from the ribosome (Rapoport et al. 1999).
approximately 8% of the proteins encoded in the The signal sequence binds to the so-called Sec com-
A. fumigatus genome are predicted to be secreted plex which includes the protein-conducting chan-
(4.5% for those proteins that have assigned Pfam nel (translocon) across the ER membrane for both
numbers). This analysis indicates the significance the co- and post-translational systems. The com-
of protein secretion to fungi and, in addition, position and functions (Keenan et al. 2001), and
that the secretion process also involves many the structure (Halic et al. 2004) of the eukaryotic
proteins which have functions in maintaining and ribonuclear-protein SRP, together with its role in
regulating the secretion process. The secretion of co-translational translocation of proteins across
proteins is tightly integrated with other processes the ER (Johnson and van Waes 1999), have been
such as hyphal growth and branching (discussed described.
Preliminary data suggested that both translo- is a function of the ER which also operates for
cation systems operate in filamentous fungi non-glycosylated proteins because exit from the
(Conesa et al. 2001). The translocon is comprised ER and continued passage within the secretory
in yeast of the trimeric Sec61p complex (Sec61p, system rely upon those proteins having attained
homologue of mammalian Sec61α; Sss1p, homo- a folded structure which is “stable” (discussed
logue of mammalian Sec61γ; Sbh1p, homologue of below) and does not expose hydrophobic regions.
mammalian Sec61β). The Sec complex of yeast in- Those proteins which are judged by the quality
cludes the translocon and the tetrameric complex: control system to be irretrievably mis-folded (at
Sec62p, Sec63p, Sec71p and Sec72p (Rapoport least within kinetic parameters defined by the
et al. 1999). Recent studies in yeast indicate that cell) are retro-translocated (dislocated) from the
the signal peptide binds to both Sec61p and ER through the translocon and are ubiquitin-
Sec62p in post-translational protein transport tagged for degradation by the proteasome. This
into the ER (Plath et al. 2004). In co-translational is termed ER-associated degradation, ERAD.
translocation, binding of the SRP arrests peptide A closer inspection has revealed that the qual-
elongation until the SRP docks with its receptor ity control system in yeast includes a series of
on the ER membrane, whereupon it is released checkpoints which determines the possible fates
(Keenan et al. 2001). For further detail of the SRP (e.g. proteasomal degradation or targeting to the
and its receptor, including the roles of GTPases, vacuole) of mis-folded proteins (Vashist and Ng
readers are referred to the cited reviews. For this 2004). The core features of protein folding and
chapter, it is sufficient to confirm that mining modifications are found in filamentous fungi
of the Aspergillus genome databases confirms as well as yeasts and higher eukaryotes, albeit
the presence of homologues of genes necessary with minor differences, have been summarised
to support both the co- and post-translational previously in detail (Archer and Peberdy 1997;
translocation systems (D.B. Archer, unpublished Conesa et al. 2001; MacKenzie et al. 2004) and
data; A. Sims, personal communication; Robson will not be repeated here. The availability of the
et al. 2005), including signal peptidase for removal genome sequences for Aspergillus spp. has since
of the signal sequences. We do not know the permitted an in-depth comparison of the genes
functionality of the systems either with different associated with each function.
secretory proteins or under different environ- Protein folding is assisted in the ER lumen
mental conditions, and this will require detailed by chaperones (e.g. BiP) and foldases (e.g. pro-
analysis. tein disulfide isomerase, PDI, and peptidyl-prolyl
cis-trans isomerase, PPI). The PDI family (thiore-
doxin, PF00085) has been annotated for A. fumiga-
B. Protein Folding in the ER tus, A. nidulans and A. oryzae, and each of these
three species contains three genes encoding PDI
Proteins entering the ER lumen may be destined family proteins characterised by thioredoxin do-
for export from the cell, targeted to other cellular mains (CGHC) and a C-terminal ER-retention sig-
destinations or retained within the lumen because nal, (K/H)DEL. Two of the PDI family proteins in
of possession of a tetrameric C-terminal retention each species have two thioredoxin domains, and
signal. Each protein undergoes assisted folding one protein has one thioredoxin domain. These
and modifications which may include structural proteins have high similarity to proteins already
isomerisation at prolines, the formation of disul- described in A. niger (Jeenes et al. 1997; Ngiam
phide bonds and glycosylation (initiated in the et al. 1997, 2000; Wang and Ward 2000). The lat-
ER and continued in subsequent compartments ter authors suggested that PrpA (one of the PDI
of the secretory system). Assisted folding involves family proteins) might be unique to filamentous
ER-resident chaperones whereas ER-resident fungi, and a very close homologue of the encoding
foldases catalyse the isomerisation and formation gene is found in all of the Aspergillus sequences.
of disulphide bonds. Glycosylation is catalysed by BLAST analysis using the sequence without the
several different proteins, and subsequent passage thioredoxin domain showed significant alignment
of glycosylated proteins must pass scrutiny by to other fungal genes (from A. niger, 1e-138, and
a quality control system which employs calnexin, N. crassa, 6e-91) but the next hits were to Dic-
a chaperone bound to the ER membrane but tyostelium discoideum (4e-26), S. pombe (2e-12)
active within the lumen of the ER. Quality control and S. cerevisiae (2e-11). Thus, PrpA and its ho-
mologues may be largely restricted to filamentous C. Post-ER Secretion

fungi. The prpA gene in A. niger is not essential
for viability (Wang and Ward 2000). Transfer of Secretory proteins are packaged in membrane-
electrons during oxidative protein folding (e.g. in bound vesicles from the ER, to the Golgi (or
the formation of disulfide bonds) involves Ero1p functional equivalent in filamentous fungi), and
in S. cerevisiae. Homologues of this ER membrane thence to the plasma membrane (primarily at the
protein are present in the Aspergillus genomes. hyphal tip). This vectorial transport of secretory
Most of the PPI family foldases predicted from proteins is common to all eukaryotes but is partic-
the genome are cytosolic. The published sequence ularly interesting in the filamentous fungi because
of cypB which encodes an ER-resident cyclophilin of the hyphal growth pattern. Secretion-related
type of PPI in A. niger has been described by Derkx GTPases play a key role in this process, and they
and Madrid (2001). The protein has an ER retention have been examined in A. niger before the genome
signal of HEEL, i.e. it diverges somewhat from the sequences were available (Punt et al. 2001) and
more usual (K/H)DEL signal, but it also has a pre- in N. crassa since its genome was sequenced
dicted signal sequence strongly suggesting that the (Borkovich et al. 2004). Five secretion-related
protein is ER-resident. A C-terminal HEEL motif is GTPase-encoding genes were characterised in A.
also found in the predicted CypB homologues in A. niger as members of the Rab/Ypt branch of the Ras
fumigatus and A. oryzae, but the A. nidulans homo- superfamily of GTPases (Punt et al. 2001).
logue has HNEL. This protein also has a predicted
signal sequence, so is likely to be ER-resident.
The A. niger calnexin gene sequence (Wang D. ER to Golgi
et al. 2003) has been described and homologues are
found in the other Aspergillus genomes. No other When ER-resident proteins are moved to the Golgi,
calnexin homologues are present, confirming that their return to the ER (retrograde transport) seems
there is no calreticulin homologue. Searches for to require COPI coated vesicles, and the A. nidu-
the ER-resident Hsp70 protein (PF00012) identify lans COP-α homologue Sodvi C is essential for the
BiP homologues in each genome as well as another establishment and maintenance of polar growth
ER-resident (judged by both the presence of a pre- (Whittaker et al. 1999).
dicted signal sequence and a C-terminal ER reten- Proteins folded correctly in the ER are recog-
tion motif) Hsp70 family member which has not nised as appropriate cargo to be incorporated into
been characterised to date in any filamentous fun- COPII vesicles, which are moved to the cis-Golgi
gus. These Hsp70 proteins may be homologues of (entry point) on microtubules. Small GTPases of
the S. cerevisiae Lhs1p Hsp70 family member and the RAB/Ypt/Sec4 family are involved in many
may be involved in protection against oxygen de- of the fusion steps in vesicle targeting, including
privation as well as in assisting protein folding. The ER to Golgi, and between Golgi, vacuole, plasma
sequences of BiP homologues in the aspergilli are membrane and endosome. Sar1 is involved in
virtually identical to that described first in A. niger budding and docking COPII vesicles (ER to Golgi),
(van Gemeren et al. 1997). BiP is a key molecular and the A. niger homologue SarA is essential
chaperone resident in the ER but it also has im- (Conesa et al. 2001). Docking of the vesicles to
portant roles in the translocation of proteins into the cis-Golgi involves a transport protein particle
the ER (through its association with Sec63 in the (TRAPP I), which is a multisubunit complex.
Sec complex) and in regulating the unfolded pro- Binding of the vesicle to TRAPP I leads to activa-
tein response (UPR) through its association with tion of another small GTPase, Ypt1. Homologues
the trans-membrane kinase IreA (see below). of Ypt1 have been characterised from A. niger
The glycosylation of proteins will not be var awamori and T. reesei, and the latter was
discussed here in detail. The enzymology of able to complement a yeast strain depleted for
N-glycosylation has been discussed previously Ypt1 (Saloheimo et al. 2004). TRAPP II, which
(Maras et al. 1999; Peberdy et al. 2001) and the anal- shares some proteins with TRAPP I, is involved
ysis of the N. crassa genome has revealed many of in transport within the Golgi. Point mutations in
the genes necessary (Borkovich et al. 2004). How- a regulatory subunit of TRAPPII, Trs120 (HypA)
ever, we do not know of a comprehensive mining of result in wide, slow-growing hyphae with thick
the Aspergillus genomes for genes predicted to be walls (Shi et al. 2004), indicating a polarity defect,
involved in the glycosylation of secretory proteins. and enzyme secretion is reduced. Deletion of hypA
is not lethal but results in osmotically sensitive germ tube emergence, though hyphae are morpho-
hyphae with an aberrant morphology. logically aberrant (Xiang et al. 1994).
Some proteins entering the Golgi are targeted The myoA gene of A. nidulans (a type I myosin)
to the vacuoles, and a large number of genes (VPS seems to be essential, and its down-regulation with
for vacuolar protein sorting) are required for this. the conditional alcA promoter led to severe polar-
Mutations in some of these genes also result in hy- ity defects (McGoldrick et al. 1995). MyoA localises
perbranching or other apparent polarity defects, to hyphal tips, septa, and in cortical patches (Ya-
giving abnormal hyphal morphology. These in- mashita et al. 2000). As with hyphal tips, sites of
clude hbrA of A. nidulans (a VPS33 homologue and septation also involve actin polarisation, and the
member of the Sec1 family; Memmott et al. 2001), BNI1 homologue sepA was identified in A. nidu-
and digA (VPS18; Geissenhoner et al. 2001). lans via a mutant defective in septation (Harris
et al. 1997). Interestingly, a GFP fusion with the
secreted protein glucoamylase is also observed at
E. Golgi to Plasma Membrane both tips and septa in A. niger (Gordon et al. 2000).
In yeast, the homologue of MyoA, Myo3, is associ-
For the final stages of secretion, vesicles exit the ated with actin patches and endocytosis, and the
trans-Golgi and are transported to the plasma exact role of different myosins in A. nidulans re-
membrane, where they fuse to deliver their car- quires further study. A putative homologue of the
goes. Some of these secreted proteins are required type V myosin Myo2 is present in A. nidulans, but
for cell wall growth (budding in yeast, or hyphal its function has not been tested. Distinguishing be-
elongation in filamentous fungi), and others are tween actin cables and actin patches has not been
excreted into the surroundings, including the possible until recently, since phalloidin does not
enormous range of hydrolytic enzymes typical of stain Aspergillus actin, and the antibody staining
filamentous fungi (see below). It is not yet known does not permit differentiation. Recently, a GFP fu-
how this differential targeting is achieved, but it sion was made with a tropomyosin homologue, and
presumably involves different types of vesicles. this actin-associated protein revealed some actin
It is generally believed, with some evidence, that cables in the tip region (Pearson et al. 2004).
protein secretion occurs at hyphal tips (Archer
and Perberdy 1997; Gordon et al. 2000).
Vesicle transport in yeast involves actin cables F. Polarity Establishment
and myosin(s) which are in turn directed towards
specific sites in the cell cortex (reviewed in Pruyne Putative genes for much of the machinery of polar-
and Bretscher 2000a, b; Irazoqi and Lew 2004). ity establishment can be detected in the Aspergillus
This is achieved by cortical markers which define genomes, and some of these genes have been inves-
the point at which a bud will develop. The corti- tigated for function, either in Aspergillus species
cal markers result in localised assembly of protein or in related filamentous fungi. The GTPase Cdc42
complexes which nucleate the growth of actin ca- plays a key role in regulating polarity establishment
bles, resulting in transport of vesicles to that point. in all eukaryotes, together with its GEF Cdc24 and
Early cortical markers in yeast, which define the several GAPs. Cdc42 is recruited to the site of polar
bud site, Bud3, Bud8 and Bud9, do not seem to growth in yeast. Unlike S. cerevisiae, Aspergillus
have homologues in Aspergillus species, and alter- species also seem to possess a Rac homologue,
native models have been proposed for how sites of a second Rho-type GTPase (Harris and Momany
polarised growth might be chosen in filamentous 2004). Though no functional analysis has been re-
fungi (Harris and Momany 2004). ported in Aspergillus species, a Rac homologue in
There is evidence from studies on mutants of N. Penicillium marneffei seems to contribute to polar
crassa (Seiler et al. 1999; Riquelme et al. 2000) that growth (Boyce et al. 2003).
in filamentous fungi, cytoplasmic microtubules, to- Other proteins which act with Cdc42 to estab-
gether with associated motor proteins, may play lish polarity include Cla4, Ste20, Gic1, Gic2, and the
a greater role than in yeast in the long-range trans- scaffold proteins Bem1, Boi1 and Boi2. As in yeast,
port of vesicles to the tip region, where actin- the first of two SH3 domains in the Bem1 homo-
myosin make the final delivery. Dynein mutations logue of A. nidulans, BemA, appears to be dispens-
of A. nidulans (nudA), which abolish nuclear mi- able when expressed under the control of the alcA
gration along the microtubules, do not prevent promoter, but down-regulation results in a severe
polarity defect (Zarrin et al. 2005). These polarity H. The Exocyst

establishment proteins are linked to the actin cables
via the polarisome, a key component of which is the Docking of secretory vesicles with the plasma
FH1/2 protein Bni1. The A. nidulans homologue of membrane in yeast and mammalian cells requires
BNI1, sepA, was isolated and identified by comple- a complex of eight subunits known as the exocyst
mentation of a temperature-sensitive mutant (Har- (Fig. 5.5; reviewed in Lipschutz and Mostov 2002;
ris et al. 1997). This mutant is defective in septum Hsu et al. 2004), and putative homologues of these
formation and shows a hyperbranching phenotype components can been found in the Aspergillus
at the restrictive temperature, which may also be genome sequences (Table 5.3). Six of these proteins
indicative of a polar growth defect. were discovered as products of the SEC genes of
S. cerevisiae, and mutations in these genes led to
defects in secretion. While this indicates conserved
G. The Spitzenkörper functions in the secretory pathway, functional
analysis will be necessary to reveal the location
A characteristic feature of filamentous fungi is of this structure in filamentous fungi. It might be
a structure just behind the growing tip known expected that this structure would be located on
as the Spitzenkörper (Reynaga-Pena et al. 1997), the plasma membrane at growing tips. Docking
which can be readily observed by light microscopy, of vesicles involves interaction with the exocyst
more easily in some fungi than in others, and component Sec15, and a vesicle-associated GTPase
which appears in electron micrographs to consist Sec4 (Guo et al. 1999). Sec4 is a member of the
of a concentration of vesicles. It has been suggested Rab family of GTPases, involved in regulation of
that it acts as a vesicle supply centre, and this idea vesicle traffic. While a gene encoding a putative
has been modelled mathematically. Sometimes, the Sec4 homologue, SrgA, has been deleted in A.
body splits into two just before terminal branching niger (Punt et al. 2001), this was not lethal, as
is observed, and new Spitzenkörpers are observed is the case in S. cerevisiae. Deletion resulted in
to form at branch points. The molecular nature
of the Spitzenkörper is not well defined, though
it has been shown that a GFP–SepA (Bni1) fusion
product localises not only to the membrane surface
at the tip, but also to a region just behind the
tip. Hence, this may be useful as a Spitzenkörper
marker (Sharpless and Harris 2002). It further
indicates that an element of the polarisome is
associated with the Spitzenkörper, as well as with
the point of growth. The availability of the genome
sequence, and more rapid methods of GFP/RFP
tagging should help to identify other components
associated with this structure, and clarify its
nature.
There will then remain many interesting ques-
tions about how the Spitzenkörper is assembled,
and how it is positioned. Recent evidence is sug-
gesting that cytoplasmic microtubules may be in-
volved in the long-range movement of vesicles to
the tip region and in positioning the Spitzenkörper, Fig. 5.5. Vesicle docking and the exocyst. S. cerevisiae sec
while short-range transport of vesicles from the mutants are deficient in secretion. Sec3, 5, 6, 8, 10, 15, and
Exo70 and Exo84 are proteins of the exocyst, a 19.5S com-
Spitzenkörper region to the tip may require micro- plex associated with the plasma membrane at sites of se-
filaments (reviewed in Harris and Momany 2004). cretion and cell growth. Rho3, Rho1, Cdc42 and Sec4 are
Immuno-staining of actin in A. nidulans shows its GTPases which interact with components of the exocyst
accumulation at the tip, and at newly forming septa. complex. Docking of a secretory vesicle involves an inter-
action between Sec4 and Sec15. The complex also interacts
Recent use of a GFP–tropomyosin fusion, an actin- with microtubules and actin filaments. Secretory vesicles
associated protein, has also revealed cable-like ele- are delivered via actin cytoskeleton (modified from Guo
ments in the tip region (Pearson et al. 2004). et al. 1999; Lipschutz and Mostov 2002)
Table. 5.3. Percent amino acid identities between exocyst components of A. nidulans and of selected other fungi (courtesy
of Steve Baldwin)
Exocyst Aspergillus Aspergillus Magnaporthe Gibberella Ustilago Neurospora Saccharomyces

subunit nidulans fumigatus grisea zeae maydis crassa cerevisiae
Sec3 100 69 39 39 24 39 19
Sec5 100 70 49 50 25 49 21
Sec6 100 80 51 50 22 50 19
Sec8 100 72 42 42 19 42 20
Sec10 100 74 51 50 31 53 25
Sec15 100 86 61 62 34 63 20
Exo70 100 78 44 45 22 45 23
Exo84 100 76 47 46 24 45 19
aberrant hyphal morphology, indicating some coded by Aspergillus spp. are predicted to have ei-
role in hyphal growth, but this observation also ther secretion signals or secretion anchors. It is not
emphasises that we cannot assume that what is true our intention to list all proteins likely to be secreted
for yeast is true for Aspergillus. Also, it shows the from Aspergillus, but the availability of the genome
need for further investigation of this structure and sequences now provides an unprecedented oppor-
its relationship to hyphal growth and secretion. tunity for data mining and to gain sequences of par-
Studies in yeast have also shown interactions of ticular enzymes or classes of enzymes. Aspergillus
exocyst components with other polarity-associated spp. are noted and exploited commercially for their
genes, including Rho1 and Cdc42 with Sec3. RhoA, capacity to secrete (gluco)amylases, proteases and
the A. nidulans homologue of Rho1, involved in cell pectinases, but there is a far wider range of enzymes
wall integrity control by the MAP kinase pathway, secreted.
also appears to be involved in polarity, branching A. fumigatus is not a commercial organism,
and cell wall deposition (Guest et al. 2004). While but it retains an efficient system for the secretion
comparative genomics is leading to the rapid iden- of enzymes which are key to its ecology, its growth,
tification of components of the secretion pathway its allergenicity and probably to its pathogenic-
in filamentous fungi, it is clear that additional com- ity. A. fumigatus conidia are widely dispersed in
ponents are present where there is no obvious ho- air and their environmental sources are not fully
mologue in S. cerevisiae, and a combination of mu- catalogued. It is likely that A. fumigatus is com-
tant screens and investigation of protein–protein monly distributed in nature and its capacity for
interactions will be required to go beyond the yeast growth on a wide variety of substrates is predicted
paradigm. from the annotated genome sequence. A. fumiga-
tus is known to be a common component of com-
posts and is considered to be the major compost-
IV. Secreted Proteins from Aspergillus associated biohazard (Beffa et al. 1998). Temper-
ature gradients are a feature of composts and the
A. Native Proteins thermotolerance of A. fumigatus, relative to many
other fungi, together with the degradative capacity
The secretion of proteins is a key feature of the due to versatility in the range of enzymes secreted
lifestyle of Aspergillus spp. and many other filamen- by A. fumigatus, probably explains its success in
tous fungi. Enzymes are secreted from hyphal tips composts.
for the mobilization of carbon and energy sources Each Aspergillus sp. contains genes encoding
from polymeric organic materials, and this capac- a wide range of different proteinases, including the
ity underpins the highly successful ecological niche fungal-specific class of aspartyl-endopeptidases
of many filamentous fungi as primary degraders (PF01828; A. Sims, personal communication)
of waste organic matter. This capacity is exploited which are thermostable, active at low pH and
commercially for the production of enzymes with pepstatin-insensitive. Each Aspergillus sp. has
applications in food, detergents and in many other genes encoding a wide range of glycosyl hydro-
industrial sectors (Archer 2000). As already men- lases, (phospho)lipases and pectinases which
tioned, a substantial proportion of all proteins en- ensure that those fungi are capable of degrading
the organic polymers available. Many secreted have been described and many different heterolo-
enzymes have been described from Aspergillus gous proteins expressed, with yields ranging from
spp., either biochemically or in relation to the just detectable to commercial levels (MacKenzie
encoding genes and their regulation. From detailed et al. 2004). The natural hosts from which the het-
analysis of the Aspergillus genomes, A. fumigatus erologous genes were taken have been from many
contains genes encoding more cellulose-binding sources but, in the main, the secreted yields have
domains/modules (fungal type) than do A. oryzae been better when a fungal gene is expressed in an-
and A. nidulans combined (Robson et al. 2005). other fungal host. Yield and authenticity of product
Of the 14 cellulose-binding domains assigned as are the key factors in using fungi as cell factories,
PF00734 in A. fumigatus, only one gene appears and the strategies developed to optimise protein
to encode a cellulose-binding domain attached production have been summarised several times
to a protein without clear annotation while the (Conesa et al. 2001; Punt et al. 2002; MacKenzie
remainder of the cellulose-binding domains are et al. 2004) and will not be recapitulated here.
attached to glycosyl hydrolases (families 5, 6, 7, 45, A summary of the strategies employed most
61) as well as one feruloyl esterase and a cutinase. successfully in the secreted production of heterol-
Not all of the predicted cellulose-binding domain- ogous proteins is:
containing proteins also contain clear signal
– use a strong, native promoter
peptides or signal anchors (G. Robson, personal
– optimise (may involve maximizing, but not
communication) but that may reflect errors in
necessarily) the gene copy number
automated annotations. There are several other
– integrate genes at a transcriptionally active lo-
glycosyl hydrolases without obvious cellulose-
cus
binding domains as well as three cutinases lacking
– use a fungal host of low proteolytic activity, or
a cellulose-binding domain in A. fumigatus.
minimize the induction of proteases by appro-
The degradation of pectins is particularly com-
priate cultural technology
plex due to the heterogeneous nature of the poly-
– express a gene fusion whereby the target pro-
mer. For the complete mineralisation of pectin,
tein is fused downstream of a carrier protein
a wide range of enzymes is required to ensure
(e.g. glucoamylase). Cleavage of the two com-
that both the backbone and side chains are de-
ponents of the fusion protein may be effected in
graded. Pectinolytic enzymes have been well de-
vivo by engineering a dibasic amino acid endo-
scribed from A. niger (de Vries and Visser 2001)
proteolytic cleavage site at the fusion junction.
and, more recently, some expression profiling of
26 pectinolytic genes has been reported when A. There are several pertinent areas which remain
niger was grown on pectin and pectin-derived car- under- investigated. For example, there is known to
bon sources (de Vries et al. 2002). Hybridised dot- be an impact of glycosylation on secreted yields of
blots were used in this study because, at that time, chymosin (Ward 1989), but this effect has not been
genome-wide arrays were not available. The ap- taken very far in other systems. There is a limited
proach did lead to the distinction of subsets of range of promoters in current use but the genome
pectinolytic genes in relation to their regulation. sequences should provide other options quite read-
The available genomes, allied to gene arrays/chips, ily, even if they will require testing. Promoters have
will extend the ability to identify the complete set not, in the main, been modified for optimised ac-
of pectinolytic genes and will facilitate investiga- tivity by, for example, improving core promoter
tions of their regulation in the wider context of activity (Minetoki et al. 1998; Liu et al. 2003). Pro-
co-regulated genes. Similar analyses will now be moters regulated by carbon catabolite repression
possible with genes important in the degradation are in common use and strains defective in carbon
of cellulose and starch by Aspergillus spp. catabolite repression can be useful, although spe-
cific mutation of responsive sites in the promoters
is not yet widely adopted. Knowledge of proteases
B. Heterologous Enzymes from genomic information may prove to be very
useful in strain construction, although it may be
A. niger and A. oryzae (and their close relatives more efficient to examine the regulatory systems
A. awamori and A. sojae respectively) have been so that families of protease genes can be made in-
developed as cell factories for the secreted produc- effective. Where increasing gene copy number be-
tion of heterologous proteins. Host-vector systems yond a critical point becomes ineffective, increased
expression of necessary transcription factors may mediated by the bZIP transcription factor HacA in
overcome a limitation imposed by the titration of Aspergillus (Hac1p in S. cerevisiae). However, some
the native levels of these factors. ER stress responses appear to be independent of
The carrier protein system is highly effective Hac1p and the ER membrane protein Ire1p which
in improving secreted yields of heterologous senses unfolded proteins and initiates the signal
proteins. The strategy depends on the efficient and from the ER (Schröder et al. 2003). Hac1p is the
accurate cleavage of the target protein from the translation product from a spliced HAC1 mRNA
carrier by the furin-type of endoprotease resident in yeast (reviewed in Patil and Walter 2001), and
in the Golgi (Jalving et al. 2000; Punt et al. 2003). analogous sensing by Ire1p and mRNA processing
We know from a growing body of results that the occurs in Aspergillus, albeit with differences in
proteolytic processing is not always faithful or the detail (Saloheimo et al. 2003). Sensing of,
efficient (Spencer et al. 1998; MacKenzie et al. 1998; and the responses to, ER stress in eukaryotes has
Punt et al. 2003; Clemente et al. 2004) and that been summarised recently by Zhang and Kaufman
it probably depends on the context in which the (2004), Rutkowski and Kaufman (2004), and
dibasic amino acid pair is presented to membrane- Schröder and Kaufman (2005). There are alterna-
bound protease. To improve the efficiency of tive possible fates for degradation of mis-folded
cleavage, it is useful to design the context appro- proteins, depending upon which checkpoints
priately (Spencer et al. 1998) and this approach within the secretion pathway have been passed. On
has been effective in the secretion of humanized that basis, properly folded proteins are those which
antibodies from A. niger (Ward et al. 2004). have passed scrutiny by a series of checkpoints
(Vashist and Ng 2004).
Secretion of at least some heterologous
C. The Stress of Protein Secretion proteins induces the UPR in Aspergillus (e.g.
Wiebe et al. 2001; Mulder et al. 2004) and, as
The process of protein secretion is largely con- with studies in yeast, chemicals such as dithio-
served amongst all eukaryotic cells. A key element threitol and tunicamycin also induce the UPR in
of the secretion process is achieving the correct Aspergillus, with the dithiothreitol response being
fold of proteins in the lumen of the endoplasmic very strong. The genome sequences, together
reticulum (ER), so that the cell’s natural quality with gene chips/arrays, provide the opportunity
control systems are satisfied and then, the folded to assess the global transcriptional responses to
proteins can continue the secretion process. The ER stress in Aspergillus. This has been achieved
natural quality control systems within the secre- already with S. cerevisiae where there is a platform
tory pathway of eukaryotic cells ensure that only for comparative studies with Aspergillus. In S.
properly folded protein is secreted (Ellgaard and cerevisiae, nearly 400 genes were transcriptionally
Helenius 2003). How cells discriminate between up-regulated due directly or indirectly to the UPR
properly folded versus mis-folded protein, and (Travers et al. 2000), indicating that the effect
how they selectively retain and target the latter is broad and has an impact on many aspects of
for degradation are challenging questions which the cellular metabolism. Although the UPR leads
remain to be answered, although more informa- to a selective up-regulation of transcription of
tion is now becoming available. Improperly folded genes encoding proteins which combat stress due
proteins (i.e. those which have not attained their to mis-folded proteins, it may also be beneficial
native fold) induce the unfolded protein response to homeostasis if genes which encode secretory
(UPR) and other ER stress responses in yeasts and proteins are selectively transcriptionally down-
filamentous fungi (Travers et al. 2000; Ngiam et al. regulated, thus reducing the protein traffic in the
2000; Patil and Walter 2001; Al-Sheikh et al. 2004; secretory pathway. It has recently been shown
Sims et al. 2005). The unfolded protein response that ER stress may lead to the transcriptional
mediates the transcriptional up-regulation of down-regulation of genes encoding at least some
genes encoding ER-resident chaperones and secreted proteins, but not non-secreted proteins in
foldases, and up-regulates the ER-associated filamentous fungi (Pakula et al. 2003; Al-Sheikh
degradation (ERAD) following retro-translocation et al. 2004) and probably in plants (Martinez and
of proteins from the ER to the cytoplasm, where Chrispeels 2003). This response may be indepen-
those proteins are tagged by ubiquitin and targeted dent of the UPR, but that has to be verified. The
for degradation by the proteasome. The UPR is response may compensate for the apparent lack of
translational attenuation initiated by ER stress. In A link between nitrogen availability and the
mammalian cells, translational attenuation occurs UPR has been reported in yeast by Schröder et al.
and is mediated through the ER trans-membrane (2000). Their study revealed that nitrogen levels,
protein PERK (reviewed recently by Kaufman which themselves regulate the rate of overall trans-
2004). lation, also lead to effects on the UPR. The splicing
The Hac1p transcription factor of yeast (HacA of HAC1 mRNA occurs only in nitrogen-rich con-
in A. niger) is essential for the UPR. Hac1p ditions and, therefore, UPR was suggested to play
binds to the UPR element in sensitive promoters, a role in nitrogen sensing. This study also showed
and leads to the UPR-mediated transcriptional that the range of events emanating from activa-
up-regulation of genes encoding chaperones tion of the UPR was broader than previously ap-
and foldases. Rüegsegger et al. (2001) reported preciated. Indeed, the active form of Hac1p, which
translational regulation of HAC1 mRNA upon mediates the UPR, is a negative regulator of differ-
induction of the UPR in S. cerevisiae. They showed entiation responses to nitrogen starvation, meiosis
that the un-spliced HAC1 mRNA is associated with and pseudohyphal growth (Schröder et al. 2000,
polyribosomes even though it does not lead to the 2004; Kaufman et al. 2002). Repression of the early
synthesis of full-length protein. The expression of meiotic genes was shown to involve an association
the un-spliced HAC1 mRNA was already known of Hac1p with the histone deacetylase complex in
to be post-transcriptionally regulated by an intron S. cerevisiae (Schröder et al. 2004), thus extending
(and this block was removed upon excision of the our knowledge of the roles and mechanisms of ac-
intron by a non-conventional splicing event) but tion of the UPR. However, we do not yet know how
the mechanism was not fully appreciated. Rüegseg- far analogies between yeast and Aspergillus can be
ger et al. (2001) showed that the intron base-paired taken. Global genomic approaches in S. cerevisiae
with a region of the 5 UTR to cause stalling. Hence, have also examined other responses, e.g. to carbon
despite being associated with polyribosomes, no sources (Kuhn et al. 2001) and other nutritional
Hac1p protein was produced. Saloheimo et al. and environmental factors (Gasch et al. 2000), and
(2003) have confirmed the existence of some RNA can now be explored in Aspergillus. The genome
secondary structures in the Hac mRNAs from A. sequences provide a basis not only for global stud-
nidulans and T. reesei which may regulate Hac ies using gene chips/arrays but also for proteomic
mRNA translation in these fungi. It has also been studies, including protein–protein interactions in
confirmed that HacA mediates the UPR in A. Aspergillus, analogous to studies in other systems
niger and that it up-regulates its own transcription which include S. cerevisiae (Uetz et al. 2000; Gavin
(Mulder et al. 2004). It is possible to delete the hacA et al. 2000; Han et al. 2004). Most studies with As-
gene from A. niger (H. Mulder, personal commu- pergillus protein secretion and ER stress responses
nication) to facilitate studies on HacA dependency have involved a limited number of gene targets and
(as was used, for example, in S. cerevisiae by these have been extensively reviewed (Punt et al.
Travers et al. 2000). The constitutive up-regulation 2001; MacKenzie et al. 2004). A recent suggestion
of hacA enhances the secreted yield of at least two that the entire secretory pathway of T. reesei might
heterologous proteins by A. niger (Valkonen et al. be transcriptionally induced by ER stress (Salo-
2003). Other studies indicate that there may be an heimo et al. 2004) begs investigation in Aspergillus
optimal level of individual foldases and chaperones which will be achievable with gene chips/arrays.
for maximized secretion of any given heterologous
protein (Moralejo et al. 2001; Lombrana et al. 2004)
by Aspergillus spp. Genomic-based technologies D. Polar Growth and Secretion
will provide the means to establish the difference
in responses which occur with the secretion of Aspergillus species grow in a highly polarised man-
different heterologous proteins, enabling some ner, typical of all filamentous fungi (Trinci et al.
predictive rules to be established. One factor 1994). The uninucleate, haploid, asexual spore or
determining the likelihood of efficient secretion of conidium germinates to give a growth tube follow-
a protein is its thermodynamic stability (Kowalski ing a short period of isotropic growth. The hypha
et al. 1998a, b; Song et al. 2002; Kjeldsen and then enters a duplication cycle of elongation by tip
Pettersson 2003). Hence, gene chips can be used to growth, mitoses, septation, and lateral branch for-
investigate whether there are changes at the level mation. Though many components and features of
of ER stress with proteins of different stabilities. the mechanism used for budding in yeasts are con-
served in Aspergillus and other filamentous fungi, several nuclei, probably arrested in G1 (Fiddy and
hyphal compartments are multinucleate, septum Trinci 1976). A point of growth for a lateral branch
formation is not followed by cell separation, and is then established, and a new hypha emerges, re-
hyphal elongation does not cease during mitosis sembling events during conidial germination.
(Riquelme et al. 2003). The challenge is to under- One question is how these growth points in
stand how yeast homologues, together with gene conidia and sub-apical compartments are estab-
products specific to filamentous fungi, regulate this lished. S. cerevisiae defines the point of growth with
peculiar form of growth. cortical markers, and some of the BUD gene prod-
A number of genes involved in polar growth ucts required for this give yeast its characteristic
and secretion have been identified prior to the patterns of lateral and apical budding. No clear ho-
completion of the genome sequences using mu- mologues of BUD3, BUD8 and BUD9 were found in
tant screens. In addition, some genes have been the A. nidulans genome, and Harris and Momany
found using degenerate PCR-based approaches to (2004) have proposed three models by which po-
screen for putative homologues of yeast polarity larity might be established, including a stochastic
and secretion genes, and in some cases these have model. In this model, establishment of a random
been tested for function. The completion of the site for polar growth would lead to suppression of
genome sequences and subsequent annotation of other potential sites. Detailed observations on the
open reading frames greatly expands the opportu- site of lateral branch emergence (Fiddy and Trinci
nities for identification of putative homologues of 1976) might favour this model, at least in A. nidu-
known secretion/polar growth genes. However, the lans. Emergence of a lateral branch results in the
application of the gene targeting approaches out- nuclei in the compartment re-entering the mitotic
lined above is essential to confirm function, and cycle. Thus, establishing the signalling networks
such work has not yet been carried out systemat- which link polar growth with the mitotic cycle, al-
ically on a genome-wide scale. At this time, it is ready well studied in S. cerevisiae, should also be
possible to point out some of these putative ho- a fruitful area for investigation aided by the genome
mologues. While yeast is used as the paradigm for data.
polar growth and secretion, and homologues of
many of the relevant yeast genes can be found in
filamentous fungi, this is not always the case. Some F. Branching Frequency
yeast polarity genes have no obvious fungal homo-
logues, and some genes required for polar growth Mutations leading to an increased branching fre-
in fungi have no obvious yeast homologues. This quency are relatively common, and result in more
emphasises the differences as well as the similari- compact colonies on solid media. These mutants
ties in polar growth and secretion mechanisms. can arise during commercial fungal fermentations,
and have consequences for culture viscosity, oxy-
gen limitation, and product formation (Wiebe et al.
E. Hyphal Branching and Polarity 1996; Bocking et al. 1999). Since secretion of en-
zymes is believed to occur at the tips, the question
Since filamentous fungi, unlike S. cerevisiae, are arises as to whether a higher density of tips could
able to maintain more than one axis of polar lead to improved secretion of a biotechnological
growth, many questions remain about how this is product. However, if hyperbranching results from
achieved. While polar growth is maintained at an defects in polar growth and secretion, then the
established tip, polarity has to be established in opposite might be true. Some attempts have been
germinating conidia, and during branch formation made to answer this question, with contradictory
from a sub-apical compartment. This suggests results (Lee et al. 1998; Bocking et al. 1999).
distinct mechanisms of polarity establishment and In a limited number of cases, the genetic basis
maintenance. of hyperbranching mutations has been determined,
During germination, asexual spores (conidia) indicating that they result from mutations in genes
first grow isotropically, and then establish a point associated with polar growth of vesicle traffic. A se-
of growth for germ tube emergence. Later, a sec- ries of temperature-sensitive hyperbranching mu-
ond germ tube emerges from the conidium (Harris tants (Hbr) were isolated in A. nidulans, and two
1999). Within growing hyphae, septation results genes, hbrA and hbrB, isolated by complementa-
in sub-apical compartments which each contain tion cloning. The hbrA gene encodes a homologue
of the vacuolar sorting protein VPS33/SLP1 (Mem- a model organism when investigating the biology
mott et al. 2001), and hbrB encodes a protein of of the less amenable Aspergilli.
unknown function whose sequence is conserved Sequencing the genomes of several fungal
only in filamentous fungi (Gatherar et al. 2004). The species has shown that major, costly projects can be
cot-1 mutant of Neurospora crassa results in a hy- successfully completed in this manner. The annota-
perbranching phenotype, and the gene responsible tion of N. crassa and Aspergillus genomes involved
encodes a serine-threonine protein kinase (Yarden a large part of the fairly small international fungal
et al. 1992). Homologues of this gene have been research community working together. That is
found in many eukaryotes, including CBK1 in S. a strength which should be fostered to make the
cerevisiae. Full down-regulation of hbrB (Gatherar most rapid progress in the post-genomic era in
et al. 2004) and the cot-1 homologue cotA (Turner order to advance the development of tools (e.g.
et al. 2004; Zarrin et al. 2005), using the conditional improved vectors and methodologies for tagging,
alcA promoter, leads to swollen condia (which can- gene regulation, knockouts, better genome-wide
not easily germinate) as well as swollen hyphae. arrays) which can then underpin specific projects
Moderate down-regulation leads to hyperbranch- in individual laboratories. A NIGMS-sponsored
ing. This might indicate that impairment of po- programme project for functional analysis of
lar growth at hyphal tips leads indirectly to more the N. crassa genome has begun recently (http://
frequent lateral branching. Many of these mutants www.dartmouth.edu/∼neurosporagenome/).
appear to have swollen hyphae, which may result Most proteins are secreted from fungi at the
in a critical compartment volume being attained hyphal tips. This form of polar development and
with shorter, broader compartments. This in turn protein secretion is unique, and occurs in filamen-
would lead to an increased tip/length ratio, with- tous fungi which have the other unique property of
out changing the number of branches per compart- being comprised of multinucleate cells. The fungi
ment. are ubiquitous in most environments, cause dis-
It would be interesting to identify mutations ease and are exploited commercially. Fungi there-
which lead to increases in branches per compart- fore provide a system of immense attraction for
ment, since this might help distinguish between the scientific exploration, and the genome sequences
stochastic model and the cortical marker model. provide an unprecedented opportunity for these
studies.
Acknowledgements. Genomic data for A. fumigatus
V. Conclusion was provided by The Institute for Genomic Research
(www.tigr.org/tdb/e2k1/afu1) and The Wellcome Trust
Sanger Institute (www.sanger.ac.uk/Projects/A_fumigatus);
The current availability of genome sequences for genomic data for A. nidulans was provided by The
three Aspergillus spp. (A. nidulans, A. fumigatus Broad Institute (http://www.broad.mit.edu/annota-
tion/fungi/aspergillus/); genomic data for A. oryzae was
and A. oryzae) has undoubtedly opened up new and provided by The National Institute of Advanced Indus-
quicker approaches to explore the molecular basis trial Science and Technology (http://oryzae.cbrc.jp/ and
of protein secretion and hyphal growth. Genomes www.bio.nite.go.jp/dogan/Top). Coordination of analyses
of other Aspergilli have also either been sequenced of these data was enabled by an international collaboration
or this is underway (e.g. A. niger, A. flavus, A. involving more than 50 institutions from 10 countries and
coordinated from Manchester, UK (www.cadre.man.ac.uk
terreus; see Aspergillus Genome Resources at and www.aspergillus.man.ac.uk). The authors thank the
http://www.ncbi.nlm.nih.gov/projects/genome/gui Biotechnology and Biological Sciences Research Council
de/aspergillus/) and, together with genome data for supporting their research.
from other fungal species, there is a major
resource for post-genomic studies with fungi.
Hyphal development and the secretion of proteins
are important facets of the lifestyles of saprophytic References
and pathogenic (A. fumigatus) and commercially
important (A. oryzae, A. niger) Aspergillus spp. Al-Sheikh HM, Watson AJ, Lacey GA, Punt P, MacKen-
but the molecular tools for exploiting the genome zie DA, Jeenes DJ, Pakula T, Penttilä M, Alcocer MJC,
Archer DB (2004) Endoplasmic reticulum stress leads
sequences are best developed in the sexual ‘model’ to the selective transcriptional downregulation of the
species, A. nidulans. We therefore anticipate that glucoamylase gene in Aspergillus niger. Mol Microbiol
there will continue to be a role for A. nidulans as 53:1731–1742
Archer DB (2000) Filamentous fungi as microbial cell fac- de Souza CC, Goldman MH, Goldman GH (2000) Tagging of
tories for food use. Curr Opin Biotechnol 11:478–483 genes involved in multidrug resistance in Aspergillus
Archer DB, Dyer PS (2004) From genomics to post- nidulans. Mol Gen Genet 263:702–711
genomics in Aspergillus niger. Curr Opin Microbiol de Vries RP, Visser J (2001) Aspergillus enzymes involved
7:499–504 in degradation of plant cell wall polysaccharides. Mi-
Archer DB, Peberdy JF (1997) The molecular biology of crobiol Mol Biol Rev 65:497–522
secreted enzyme production by fungi. CRC Crit Rev de Vries RP, Jansen J, Aguilar G, Parenicova L, Joosten V,
Biotechnol 17:273–306 Wulfert F, Benen JAE, Visser J (2002) Expression profil-
Askenazi M, Driggers EM, Holtzman DA, Norman TC, Iver- ing of pectinolytic genes from Aspergillus niger. FEBS
son S, Zimmer DP, Boers ME, Blomquist PR, Mar- Lett 530:41–47
tinez EJ, Monreal AW et al. (2003) Integrating tran- Ellgaard L, Helenius A (2003) Quality control in the endo-
scriptional and metabolite profiles to direct the en- plasmic reticulum. Nat Rev Mol Cell Biol 4:181–191
gineering of lovastatin-producing fungal strains. Nat Felenbok B, Flipphi M, Nikolaev I (2001) Ethanol catabolism
Biotechnol 21:150–156 in Aspergillus nidulans: a model system for study-
Beffa T, Staib F, Fischer JL, Lyon PF, Gumowski P, Marfen- ing gene regulation. Prog Nucleic Acid Res Mol Biol
ina OE, Dunoyer-Geindre S, Georgen F, Roch-Susuki R, 69:149–204
Gallaz L et al. (1998) Mycological control and surveil- Fernandez-Abalos JM, Fox H, Pitt C, Wells B, Doonan JH
lance of biological waste and compost. Med Mycol 36 (1998) Plant-adapted green fluorescent protein is a ver-
suppl I:137–145 satile vital reporter for gene expression, protein lo-
Bocking SP, Wiebe MG, Robson GD, Hansen K, Chris- calization and mitosis in the filamentous fungus, As-
tiansen LH, Trinci AP (1999) Effect of branch pergillus nidulans. Mol Microbiol 27:121–130
frequency in Aspergillus oryzae on protein se- Fiddy C, Trinci APJ (1976) Mitosis, septation, branching
cretion and culture viscosity. Biotechnol Bioeng and the duplication cycle in Aspergillus nidulans. J Gen
65:638–648 Microbiol 97:169–184
Borkovich KA, Alex LA, Yarden O, Freitag M, Turner GE, Firon A, d’Enfert C (2002) Identifying essential genes in
Read ND, Seiler S, Bell-Pedersen D, Paietta J, Plesof- fungal pathogens of humans. Trends Microbiol 10:456–
sky N et al. (2004) Lessons from the genome sequence 462
of Neurospora crassa: tracing the path from genomic Firon A, Villalba F, Beffa R, d’Enfert C (2003) Identifica-
blueprint to multicellular organism. Microbiol Mol tion of essential genes in the human fungal pathogen
Biol Rev 68:1–108 Aspergillus fumigatus by transposon mutagenesis. Eu-
Boyce KJ, Hynes, MJ, Andrianopolous A (2003) Control karyot Cell 2:247–255
of morphogenesis and actin localization by the Peni- Galagan JE, Calvo SE, Borkovich KA, Selker EU, Read ND,
cillium marneffei RAC homolog. J Cell Sci 116:1249– Jaffe D, FitzHugh W, Ma LJ, Smirnov S, Purcell S et al.
1260 (2003) The genome sequence of the filamentous fungus
Brookman JL, Denning DW (2000) Molecular genetics in Neurospora crassa. Nature 422:859–868
Aspergillus fumigatus. Curr Opin Microbiol 3:468–474 Gasch AP, Spellman PT, Kao CM, Carmel-Harel O, Eisen MB,
Brown JS, Aufauvre-Brown A, Brown J, Jennings JM, Arst H Storz G, Botstein D, Brown PO (2000) Genomic expres-
Jr, Holden DW (2000) Signature-tagged and directed sion programs in the response of yeast cells to environ-
mutagenesis identify PABA synthetase as essential for mental changes. Mol Biol Cell 11:4241–4257
Aspergillus fumigatus pathogenicity. Mol Microbiol Gatherar IM, Pollerman S, Dunn-Colemann N, Turner G
36:1371–1380 (2004) Identification of a novel gene hbrB required
Chaveroche MK, Ghigo JM, d’Enfert C (2000) A rapid for polarized growth in Aspergillus nidulans. Fungal
method for efficient gene replacement in the filamen- Genet Biol 41:463–471
tous fungus Aspergillus nidulans. Nucleic Acids Res Gavin A-C, Bösche M, Krause R, Grandi P, Marzioch M,
28:E97 Bauer A, Schultz J, Rick JM, Michon A-M, Cruciat C-
Clemente A, MacKenzie DA, Jeenes DJ, Domoney C (2004) M et al. (2000) Functional organization of the yeast
The effect of variation within inhibitory domains proteome by systematic analysis of protein complexes.
on the activity of pea protease inhibitors from the Nature 415:141–147
Bowman-Birk class. Prot Exp Purif 36:106–114 Geissenhoner A, Sievers N, Brock M, Fischer R (2001)
Conesa A, Punt PJ, van Luijk N, van den Hondel CAMJJ Aspergillus nidulans DigA, a potential homolog of
(2001) The secretion pathway in filamentous fungi: Saccharomyces cerevisiae Pep3 (Vps18), is required
a biotechnological view. Fungal Genet Biol 33:155–171 for nuclear migration, mitochondrial morphol-
Dalboge H (1997) Expression cloning of fungal enzyme ogy and polarized growth. Mol Genet Genomics
genes; a novel approach for efficient isolation of en- 266:672–685
zyme genes of industrial relevance. FEMS Microbiol Gordon CL, Khalaj V, Ram AF, Archer DB, Brookman JL,
Rev 21:29–42 Trinci AP, Jeenes DJ, Doonan JH, Wells B, Punt PJ et al.
d’Enfert C (1996) Selection of multiple disruption events (2000) Glucoamylase: green fluorescent protein fusions
in Aspergillus fumigatus using the orotidine-5 - to monitor protein secretion in Aspergillus niger. Mi-
decarboxylase gene, pyrG, as a unique transformation crobiology 146:415–426
marker. Curr Genet 30:76–82 Gouka RJ, Gerk C, Hooykaas PJ, Bundock P, Musters W,
Derkx PMF, Madrid SM (2001) The foldase CYPB is a com- Verrips CT, Groot MJ de (1999) Transformation of
ponent of the secretory pathway of Aspergillus niger Aspergillus awamori by Agrobacterium tumefaciens-
and contains the endoplasmic reticulum retention sig- mediated homologous recombination. Nat Biotechnol
nal HEEL. Mol Genet Genomics 266:537–545 17:598–601
Groot MJ de, Bundock P, Hooykaas PJ, Beijersbergen AG Kowalski JM, Parekh RN, Wittrup KD (1998a) Secretion ef-
(1998) Agrobacterium tumefaciens-mediated trans- ficiency in Saccharomyces cerevisiae of bovine pancre-
formation of filamentous fungi. Nat Biotechnol atic trypsin inhibitor mutants lacking disulfide bonds
16:839–842 is correlated with thermodynamic stability. Biochem-
Guest GM, Lin X, Momany M (2004) Aspergillus nidulans istry 37:1264–1273
RhoA is involved in polar growth, branching, and cell Kowalski JM, Parekh RN, Mao J, Wittrup KD (1998b) Pro-
wall synthesis. Fungal Genet Biol 41:13–22 tein folding stability can determine the efficiency of
Guo W, Roth D, Walch-Solimena C, Novick P (1999) The escape from endoplasmic reticulum quality control. J
exocyst is an effector for Sec4p, targeting secretory Biol Chem 273:19453–19458
vesicles to sites of exocytosis. EMBO J 18:1071–1080 Krawchuk MD, Wahls WP (1999) High efficiency gene tar-
Halic M, Becker T, Pool MR, Spahn CMT, Grassucci RA, geting in Schizosaccharomyces pombe using a modu-
Frank J, Beckmann R (2004) Structure of the signal lar, PCR-based approach with long tracts of flanking
recognition particle interacting with the elongation- homology. Yeast 15:1419–1427
arrested ribosome. Nature 427:808–814 Kuhn KM, DeRisi JL, Brown PO, Sarnow P (2001) Global
Han J-DJ, Bertin N, Hao T, Goldberg DS, Berriz GF, and specific translational regulation in the genomic
Zhang LV, Dupuy D, Walhout AJM, Cusick ME, Roth response of Saccharomyces cerevisiae to a rapid trans-
FP et al. (2004) Evidence for dynamically organized fer from a fermentable to a nonfermentable carbon
modularity in the yeast protein–protein interaction source. Mol Cell Biol 21:916–927
network. Nature 430:88–93 Langfelder K, Gattung S, Brakhage AA (2002) A novel
Harris SD (1999) Morphogenesis is coordinated with method used to delete a new Aspergillus fumiga-
nuclear division in germinating Aspergillus nidulans tus ABC transporter-encoding gene. Curr Genet
conidiospores. Microbiology 145:2747–2756 41:268–274
Harris SD, Momany M (2004) Polarity in filamentous fungi: Lee IH, Walline RG, Plamann M (1998) Apolar growth of
moving beyond the yeast paradigm. Fungal Genet Biol Neurospora crassa leads to increased secretion of ex-
41:391–400 tracellular proteins. Mol Microbiol 29:209–218
Harris SD, Morrell JL, Hamer JE (1994) Identification and Lipschutz JH, Mostov KE (2002) Exocytosis: the many mas-
characterization of Aspergillus nidulans mutants de- ters of the exocyst. Curr Biol 12:R212–R214
fective in cytokinesis. Genetics 136:517–532 Liu L, Liu J, Qiu RX, Zhu XG, Dong ZY, Tang GM (2003) Im-
Harris SD, Hamer L, Sharpless KE, Hamer JE (1997) The proving heterologous gene expression in Aspergillus
Aspergillus nidulans sepA gene encodes an FH1/2 pro- niger by introducing copies of protein-binding
tein involved in cytokinesis and the maintenance of sequence containing CCAAT to the promoter. Lett
cellular polarity. EMBO J 16:3474–3483 Appl Microbiol 36:358–361
Hsu SC, TerBush D, Abraham M, Guo W (2004) The ex- Lombrana M, Moralejo FJ, Pinto R, Martin JF (2004) Modu-
ocyst complex in polarized exocytosis. Int Rev Cytol lation of thaumatin secretion by Aspergillus awamori
233:243–265 by modification of bipA gene expression. Appl Environ
Irazoqui JE, Lew DJ (2004) Polarity establishment in yeast. Microbiol 70:5145–5152
J Cell Sci 117:2169–2171 Machida M (2002) Progress of Aspergillus oryzae genomics.
Jadoun J, Shadkchan Y, Osherov N (2004) Disruption of the Adv Appl Microbiol 51:81–106
Aspergillus fumigatus argB gene using a novel in vitro MacKenzie DA, Kraunsoe JAE, Chesshyre JA, Lowe G,
transposon-based mutagenesis approach. Curr Genet Komiyama T, Fuller RS, Archer DB (1998) Aberrant
45:235–241 processing of wild-type and mutant bovine pancreatic
Jalving R, van de Vondervoort PJ, Visser J, Schaap PJ trypsin inhibitor secreted by Aspergillus niger. J
(2000) Characterization of the kexin-like mat- Biotechnol 63:137–146
urase of Aspergillus niger. Appl Environ Microbiol MacKenzie DA, Jeenes DJ, Archer DB (2004) Filamentous
66:363–368 fungi as expression systems for heterologous proteins.
Jeenes DJ, Pfaller R, Archer DB (1997) Isolation and charac- In: Kuck U (ed) The Mycota II. Genetics and biotech-
terisation of a novel stress-inducible PDI-family gene nology, chap 15. Springer, Berlin Heidelberg New York,
from Aspergillus niger. Gene 193:151–156 pp 289–315
Johnson AE, van Waes MA (1999) The translocon: a dynamic Maras M, van Die I, Contreras R, van den Hondel CA (1999)
gateway at the ER membrane. Annu Rev Cell Dev Biol Filamentous fungi as production organisms for gly-
15:799–842 coproteins of biomedical interest. Glycoconj J 16:99–
Kaufman RJ (2004) Regulation of mRNA translation by 107
protein folding in the endoplasmic reticulum. Trends Martinez IM, Chrispeels MJ (2003) Genomic analysis of the
Biochem Sci 29:152–158 unfolded protein response in Arabidopsis shows its
Kaufman RJ, Scheuner D, Schröder M, Shen X, Lee K, Liu CY, connection to important cellular processes. Plant Cell
Arnold SM (2002) The unfolded protein response in 15:561–576
nutrient sensing and differentiation. Nat Rev Mol Cell Martinez D, Larrondo LF, Putnam N, Gelpke MDS,
Biol 3:411–421 Huang K, Chapman J, Helfenbein KG, Ramaiya P,
Keenan RJ, Freymann DM, Stroud RM, Walter P (2001) The Detter JC, Larimer F et al. (2004) Genome sequence
signal recognition particle. Annu Rev Biochem 70:755– of the lignocellulose degrading fungus Phane-
775 rochaete chrysosporium strain RP78. Nat Biotechnol
Kjeldsen T, Pettersson AF (2003) Relationship between self- 22:695–700
association of insulin and its secretion efficiency in McGoldrick CA, Gruver C, May GS (1995) myoA of As-
yeast. Prot Exp Purif 27:331–337 pergillus nidulans encodes an essential myosin I re-
quired for secretion and polarized growth. J Cell Biol Aspergillus classification. Harwood, The Netherlands,
128:577–587 pp 9–10
Medina ML, Kiernan UA, Francisco WA (2004) Proteomic Plath K, Wilkinson BM, Stirling CJ, Rapoport TA (2004)
analysis of rutin-induced secreted proteins from As- Interactions between Sec complex and prepro-alpha-
pergillus flavus. Fungal Genet Biol 41:327–335 factor during posttranslational protein transport into
Memmott SD, Pollerman S, Burrow S, Dunn-Colemann N, the endoplasmic reticulum. Mol Biol Cell 15:1–10
Turner G (2001) A vacuolar protein sorting gene af- Pontecorvo G, Roper JA, Hemmons LM, MacDonald KD,
fects hyphal branching and vacuole biogenesis in As- Bufton AWJ (1953) The genetics of Aspergillus nidu-
pergillius nidulans. Genbank Accession AY064215.1 lans. Adv Genet 5:141–238
(http://www.ncbi.nih.gov/) Pruyne D, Bretscher A (2000a) Polarization of cell growth
Minetoki T, Kumagai C, Gomi K, Kitamoto K, Takahashi K in yeast. J Cell Sci 113:365–375
(1998) Improvement of promoter activity by the in- Pruyne D, Bretscher A (2000b) Polarization of cell growth
troduction of multiple copies of the conserved region in yeast. J Cell Sci 113:571–585
III sequence, involved in the efficient expression of Punt PJ, Seiboth B, Weenink XO, van Zeijl C, Lenders M,
Aspergillus oryzae amylase-encoding genes. Appl Mi- Konetschny C, Ram AFJ, Montijn R, Kubicek CP,
crobiol Biotechnol 50:459–467 van den Hondel CAMJJ (2001) Identification and
Moralejo FJ, Watson AJ, Jeenes DJ, Archer DB, Martín JF characterization of a family of secretion-related small
(2001) A defined level of protein disulfide isomerase GTPase-encoding genes from the filamentous fungus
expression is required for optimal secretion of thau- Aspergillus niger: a putative SEC4 homologue is not
matin by Aspergillus awamori. Mol Genet Genomics essential for growth. Mol Microbiol 41:513–525
266:246–253 Punt PJ, van Biezen N, Conesa A, Albers A, Mangnus J, van
Mulder HJ, Saloheimo M, Penttilä M, Madrid S (2004) The den Hondel CAMJJ (2002) Filamentous fungi as cell
transcription factor HACA mediates the unfolded pro- factories for heterologous protein production. Trends
tein response in Aspergillus niger, and up-regulates its Biotechnol 20:200–206
own transcription. Mol Genet Genomics 271:130–140 Punt PJ, Drint-Kuijvenhoven A, Lokman BC, Spencer JA,
Ngiam C, Jeenes DJ, Archer DB (1997) Isolation and van der Kamp EMC, Jeenes D, Archer DB, van den
characterisation of a gene encoding protein disulfide Hondel CAMJJ (2003) The role of the Aspergillus niger
isomerase, pdiA, from Aspergillus niger. Curr Genet furin-type protease gene pclA in processing of fungal
31:133–138 proproteins and fusion proteins. Evidence for alter-
Ngiam C, Jeenes DJ, Punt PJ, van den Hondel CA, Archer native processing of recombinant (fusion-) proteins.
DB (2000) Characterization of a foldase, protein disul- J Biotechnol 106:23–32
fide isomerase A, in the protein secretory pathway of Rapoport TA, Matlack KES, Plath K, Misselwitz B, Staeck O
Aspergillus niger. Appl Environ Microbiol 66:775–782 (1999) Posttranslational protein translocation across
Nicosia MGL, Brocard-Masson C, Demais S, Van AH, the membrane of the endoplasmic reticulum. Biol
Daboussi MJ, Scazzocchio C (2001) Heterologous Chem 380:1143–1150
transposition in Aspergillus nidulans. Mol Microbiol Reynaga-Pena CG, Gierz G, Bartnicki-Garcia S (1997) Anal-
39:1330–1344 ysis of the role of the Spitzenkorper in fungal morpho-
Oakley BR, Oakley CE, Yoon Y, Jung MK (1990) Gamma- genesis by computer simulation of apical branching
tubulin is a component of the spindle pole body that is in Aspergillus niger. Proc Natl Acad Sci USA 94:9096–
essential for microtubule function in Aspergillus nidu- 9101
lans. Cell 61:1289–1301 Riach MBR, Kinghorn JR (1996) Genetic transformation
Osherov M, Mathew NJ, May GS (2002) Polarity-defective and vector developments in filamentous fungi. In: Bos
mutants of Aspergillus nidulans. Fungal Genet Biol CJ (ed) Fungal genetics: principles and practice. Marcel
31:181–188 Dekker, New York, pp 209–233
Pakula TM, Laxell M, Huuskonen A, Uusitalo J, Salo- Riquelme M, Gierz G, Bartnicki-Garcia S (2000) Dynein and
heimo M, Penttilä M (2003) The effects of drugs dynactin deficiencies affect the formation and function
inhibiting protein secretion in the filamentous fungus of the Spitzenkorper and distort hyphal morphogene-
Trichoderma reesei: evidence for down-regulation of sis of Neurospora crassa. Microbiology 146:1743–1752
genes that encode secreted proteins in the stressed Riquelme M, Fischer R, Bartnicki-Garcia S (2003) Apical
cells. J Biol Chem 278:45011–45020 growth and mitosis are independent processes in As-
Patil C, Walter P (2001) Intracellular signaling from the pergillus nidulans. Protoplasma 222:211–215
endoplasmic reticulum to the nucleus: the unfolded Robson GD, Huang J, Wortmann J, Archer DB (2005)
protein response in yeast and mammals. Curr Opin A preliminary analysis of the process of protein
Cell Biol 13:349–356 secretion, and the diversity of putative secreted
Pearson CL, Xu K, Sharpless KE, Harris SD (2004) MesA, hydrolases encoded in Aspergillus fumigatus: insights
a novel fungal protein required for the stabilization from the genome. Med Mycol Supp 1 43:S41–S47
of polarity axes in Aspergillus nidulans. Mol Biol Cell Rüegsegger U, Leber JH, Walter P (2001) Block of HAC1
15:3658–3672 mRNA translation by long-range base pairing is re-
Peberdy JF, Wallis GLF, Archer DB (2001) Protein secretion leased by cytoplasmic splicing upon induction of the
by fungi. Appl Mycol Biotechnol 1:73–114 unfolded protein response. Cell 107:103–114
Pitt JI, Samson RA, Frisvad JC (2000) List of accepted species Rutkowski DT, Kaufman RJ (2004) A trip to the ER: coping
and their synonyms in the family Trichocomaceae. In: with stress. Trends Cell Biol 14:20–28
Samson RA, Pitt JI (eds) Integration of modern taxo- Saloheimo M, Valkonen M, Penttilä M (2003) Activation
nomic methods for Penicillium encoding protein and mechanisms of the HACI-mediated unfolded protein
response in filamentous fungi. Mol Microbiol 47:1149– Tavoularis S, Scazzocchio C, Sophianopoulou V (2001)
1161 Functional expression and cellular localization of
Saloheimo M, Wang H, Valkonen M, Vasara T, Huusko- a green fluorescent protein-tagged proline transporter
nen A, Riikonen M, Pakula T, Ward M, Penttila M in Aspergillus nidulans. Fungal Genet Biol 33:115–125
(2004) Characterization of secretory genes ypt1/yptA Teeri TT (2004) Genome sequence of an omnipotent fungus.
and nsf1/nsfA from two filamentous fungi: induction of Nat Biotechnol 22:679–680
secretory pathway genes of Trichoderma reesei under Toews MW, Warmbold J, Konzack S, Rischitor P, Veith D,
secretion stress. Appl Environ Microbiol 70:459–467 Vienken K, Vinuesa C, Wei H, Fischer R (2004) Es-
Schröder M, Kaufman RJ (2005) ER stress and the unfolded tablishment of mRFP1 as a fluorescent marker in As-
protein response. Mutation Res 569:29–63 pergillus nidulans and construction of expression vec-
Schröder M, Chang JS, Kaufman RJ (2000) The unfolded tors for high-throughput protein tagging using recom-
protein response represses nitrogen-starvation bination in vitro (GATEWAY). Curr Genet 45:383–389
induced developmental differentiation in yeast. Genes Travers KJ, Patil CK, Wodicka L, Lockhart DJ, Weissman JS,
Dev 14:2962–2975 Walter P (2000) Functional and genomic analyses re-
Schröder M, Clark R, Kaufman RJ (2003) IRE1- and HAC1- veal an essential coordination between the unfolded
independent transcriptional regulation in the unfolded protein response and ER-associated degradation. Cell
protein response of yeast. Mol Microbiol 49:591–606 101:249–258
Schröder M, Clark R, Liu CY, Kaufman RJ (2004) The Trinci APJ, Wiebe MG, Robson GD (1994) The mycelium
unfolded protein response represses differentiation as an integrated entity. In: Esser K, Lemke PA (eds)
through the RPD3-SIN3 histone deacetylase. EMBO J The Mycota I. Growth, differentiation and sexuality.
23:2281–2292 Springer, Berlin Heidelberg New York, pp 175–193
Seiler S, Plamann M (2003) The genetic basis of cellular Turner G (1994) Vectors for genetic manipulation. In: Mar-
morphogenesis in the filamentous fungus Neurospora tinelli SD, Kinghorn JR (eds) Aspergillus: 50 years on,
crassa. Mol Biol Cell 14:4352–4364 chap 24. Elsevier, Amsterdam, pp 641–665
Seiler S, Plamann M, Schliwa M (1999) Kinesin and dynein Turner G, Safaie M, John S, Dunn-Coleman N (2004) Control
mutants provide novel insights into the roles of vesicle of polar growth in Aspergillus nidulans by the cotA
traffic during cell morphogenesis in Neurospora. Curr kinase. GenBank Accession AY620243.1
Biol 9:779–785 Uetz P, Giot L, Cagney G, Mansfield TA, Judson RS,
Sharpless KE, Harris SD (2002) Functional characterization Knight JR, Lockshon D, Narayan V, Srinivasan M,
and localization of the Aspergillus nidulans formin Pochart P et al. (2000) A comprehensive analysis
SEPA. Mol Biol Cell 13:469–479 of protein–protein interactions in Saccharomyces
Shi X, Sha Y, Kaminskyj S (2004) Aspergillus nidulans hypA cerevisiae. Nature 403:623–627
regulates morphogenesis through the secretion path- Valkonen M, Ward M, Wang H, Penttilä M, Saloheimo M
way. Fungal Genet Biol 41:75–88 (2003) Improvement of foreign protein production in
Shuster JR, Connelley MB (1999) Promoter-tagged restric- Aspergillus niger var. awamori by constitutive induc-
tion enzyme-mediated insertion (PT-REMI) mutage- tion of the unfolded protein response. Appl Environ
nesis in Aspergillus niger. Mol Gen Genet 262:27–34 Microbiol 69:6979–6986
Sims AH, Robson GD, Hoyle DC, Oliver SG, Turner G, van Gemeren IA, Punt PJ, Drint-Kuyvenhoven A, Broekhui-
Prade RA, Russell HH, Dunn-Coleman NS, Gent ME jsen MP, van Hood A, Beijersbergen A, Verrips CT, van
(2004a) Use of expressed sequence tag analysis den Hondel CAMJJ (1997) The ER chaperone-encoding
and cDNA microarrays of the filamentous fungus bipA gene of black Aspergilli is induced by heat shock
Aspergillus nidulans. Fungal Genet Biol 41:199–212 and unfolded proteins. Gene 198:43–52
Sims AH, Gent ME, Robson GD, Dunn-Coleman NS, Vashist S, Ng DTW (2004) Misfolded proteins are sorted
Oliver SG (2004b) Combining transcriptome data with by a sequential checkpoint mechanism of ER quality
genomic and cDNA sequence alignments to make control. J Cell Biol 165:41–52
confident functional assignments for Aspergillus Wang H, Ward M (2000) Molecular characterisation of
nidulans genes. Mycol Res 108:853–857 a PDI-related gene prpA in Aspergillus niger var
Sims AH, Gent ME, Lanthaler K, Dunn-Coleman NS, awamori. Curr Genet 37:57–64
Oliver SG, Robson GD (2005) Transcriptome analysis Wang H, Lambert J, Morlon E, Archer DB, Peberdy JF,
of recombinant protein secretion by Aspergillus Ward M, Jeenes DJ (2003) Isolation and characteriza-
nidulans and the unfolded-protein response in vivo. tion of a calnexin homologue, clxA, from Aspergillus
Appl Environ Microbiol 71:2737–2747 niger. Mol Genet Genomics 268:684–691
Song Y, Sakai J, Saito A, Usui M, Azakami H, Kato A (2002) Ward M (1989) Production of calf chymosin by Aspergillus
Relationship between the stability of lysozymes mu- awamori. In: Hershberger CL, Queener SW, Hege-
tated at the inside hydrophobic core and secretion in man G (eds) Genetics and molecular biology of
Saccharomyces cerevisiae. Nahrung 46:209–213 industrial microorganisms. American Society for
Spencer JA, Jeenes DJ, MacKenzie DA, Haynie DT, Archer Microbiology, Washington, DC, pp 288–294
DB (1998) Determinants of the fidelity of processing Ward M, Lin C, Victoria DC, Fox BP, Fox JA, Wong DL, Meer-
glucoamylase-lysozyme fusions in Aspergillus niger. man HJ, Pucci JP, Fong RB, Heng MH et al. (2004) Char-
Eur J Biochem 258:107–112 acterization of humanized antibodies secreted by As-
Su W, Li S, Oakley BR, Xiang X (2004) Dual-color imaging pergillus niger. Appl Environ Microbiol 70:2567–2576
of nuclear division and mitotic spindle elongation in Waring RB, May GS, Morris NR (1989) Characterization of
live cells of Aspergillus nidulans. Eukaryot Cell 3:553– an inducible expression system in Aspergillus nidulans
556 using alcA and tubulin-coding genes. Gene 79:119–130
Wendland J (2003) PCR-based methods facilitate targeted Yang L, Ukil L, Osmani A, Nahm F, Davies J, De Souza CP,
gene manipulations and cloning procedures. Curr Dou X, Perez-Balaguer A, Osmani SA (2004) Rapid pro-
Genet 44:115–123 duction of gene replacement constructs and genera-
Whittaker SL, Lunness P, Milward KJ, Doonan JH, Assin- tion of a green fluorescent protein-tagged centromeric
der SJ (1999) sodVIC is an alpha-COP-related gene marker in Aspergillus nidulans. Eukaryot Cell 3:1359–
which is essential for establishing and maintaining po- 1362
larized growth in Aspergillus nidulans. Fungal Genet Yarden O, Plamann M, Ebbole DJ, Yanofsky C (1992) cot-1,
Biol 26:236–252 a gene required for hyphal elongation in Neurospora
Wiebe MG, Blakebrough ML, Craig SH, Robson GD, crassa, encodes a protein kinase. EMBO J 11:2159–
Trinci AP (1996) How do highly branched (colonial) 2166
mutants of Fusarium graminearum A3/5 arise during Yaver DS, Lamsa M, Munds R, Brown SH, Otani S,
Quorn myco-protein fermentations? Microbiology Franssen L, Johnstone JA, Brody H (2000) Using
142:525–532 DNA-tagged mutagenesis to improve heterologous
Wiebe MG, Karandikar A, Robson GD, Trinci APJ, Can- protein production in Aspergillus oryzae. Fungal
dia JLF, Trappe S, Wallis G, Rinas U, Derkx PMF, Madrid Genet Biol 29:28–37
S et al. (2001) Production of tissue plasminogen acti- Yu JH, Hamari Z, Han KH, Seo JA, Reyes-Dominguez Y,
vator (t-PA) in Aspergillus niger. Biotechnol Bioeng Scazzocchio C (2004) Double-joint PCR: a PCR-based
76:164–174 molecular tool for gene manipulations in filamentous
Xiang X, Beckwith SM, Morris NR (1994) Cytoplasmic fungi. Fungal Genet Biol 41:973–981
dynein is involved in nuclear migration in Aspergillus Zarrin M, Leeder A, Turner G (2005) A rapid method for
nidulans. Proc Natl Acad Sci USA 91:2100–2104 promoter exchange in Aspergillus nidulans using re-
Yamashita RA, Osherov N, May GS (2000) Localization combinant PCR. Fungal Genet Biol 42:1–8
of wild type and mutant class I myosin proteins in Zhang K, Kaufman RJ (2004) Signaling the unfolded protein
Aspergillus nidulans using GFP-fusion proteins. Cell response from the endoplasmic reticulum. J Biol Chem
Motility Cytoskeleton 45:163–172 279:25935–25938
6 The Genomics of Stress Response in Fission Yeast
B.T. Wilhelm1 , J. Bähler1
CONTENTS other experiments performed under different con-

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 97 ditions. The ability therefore to combine all avail-
II. Stress Response Pathways in Fission Yeast . . 98 able published data into a coherent and conclusive,
III. Applications of Genomics Techniques overall cellular response is often difficult, if not im-
to Stress Response . . . . . . . . . . . . . . . . . . . . . 100 possible. New experimental approaches developed
A. Identifying Targets of Signalling Pathways 100
B. Identification of Core and Specific Stress within the last decade have already revolutionized
Response Genes . . . . . . . . . . . . . . . . . . . . 100 the way in which molecular biology is performed.
C. Identification of Mediating Factors Most notably, the advent of microarrays has al-
of Global Response . . . . . . . . . . . . . . . . . . 103 lowed a shift in experimental paradigms to a com-
IV. Correlations Between Budding and Fission
Yeast Stress Responses . . . . . . . . . . . . . . . . . . 104 plementary and powerful approach for studying
V. Unanswered Questions not only stress response, but virtually any cellu-
and Future Directions . . . . . . . . . . . . . . . . . . 106 lar behaviour from expression differences to tran-
A. Correlations Between Transcription scription factor binding. The ability to globally and
and Translation . . . . . . . . . . . . . . . . . . . . 106 simultaneously survey the expression level of an en-
B. Chromatin Remodelling in Response
to Stress . . . . . . . . . . . . . . . . . . . . . . . . . . 107 tire genome not only facilitates the study of tran-
C. Mapping Genomic Binding Sites of Stress scriptional regulation but also, through the use of
Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . 108 time-course studies, enables a regulatory network
D. Examining Subtle Changes Associated to be established. Because of their small genomes
with Stress . . . . . . . . . . . . . . . . . . . . . . . . 108
VI. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 109 (relative to other eukaryotes) and well-defined cel-
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 lular behaviours, fungi remain one of the most use-
ful model systems for studying the transcriptional
behaviour at the level of the whole genome. The
Abbreviations: CESR, core environmental stress other obvious advantage of surveying the entire
response; MAPK, MAP kinase; MAPKK, MAP genome using this approach is that, in addition to
kinase kinase; MMS, methyl methanesulfonate; displaying the responses that would be predicted,
SESR, specific environmental stress response whole-genome studies also highlight those changes
which are not expected. Indeed, the ability to dis-
cover novel effects without a priori knowledge of
I. Introduction
the system is a key strength of microarray studies.
The use of microarrays and other genome-wide
The study of the molecular mechanisms behind technologies for experimentation, often generically
stress responses in model organisms has been a rich described as genomics (or functional genomics)
field of research that has defined crucial aspects studies, inherently generates a large amount of
of cellular behaviour. Most traditional studies in data. Typically, the most challenging part in these
this field have relied on a combination of biochem- experiments is to extract biologically meaningful
ical and genetic techniques to highlight the role information from the huge mass of data. In any
of a particular gene or pathway. While such ap- case, such data represent a valuable and reusable
proaches have the advantage of providing a well- resource when new discoveries are made, and the
defined set of behaviours for a particular protein, existing datasets can be reanalysed to create con-
the results are often not directly comparable with nections between novel and established pathways
1Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, and behaviours. However, while the accumulation
UK of data represents a benefit, significant difficulties
The Mycota XIII
Fungal Genomics
98 B.T. Wilhelm and J. Bähler
still exist in making these data available to the cellular functions, etc.) are more similar between
wider scientific community. The current situation S. pombe and higher eukaryotes than is the case for
is similar in many respects to the early days of S. cerevisiae (Forsburg 1999; Sunnerhagen 2002).
genome sequencing before databases and common As such, and although much cellular behaviour has
storage formats for DNA sequences were firmly es- first been studied in S. cerevisiae, there is still sig-
tablished, and also before there was a requirement nificant value in performing similar experiments in
to submit such experimental data to databases as S. pombe. Evidence supporting this argument can
a condition of scientific publication. Although mi- be seen in previous genomic studies performed on
croarray data are clearly more complex than se- S. pombe which clearly demonstrate that, despite
quence data and standardisation remains a signifi- superficial similarities in some characteristics, the
cant challenge, efforts to support a new MIAME re- two yeasts have undergone significant independent
porting standard for microarray experiments, and evolution in the billion years since their last com-
the establishment of microarray databases, should mon ancestor (Mata et al. 2002; Chen et al. 2003;
prove highly beneficial for the field (Brazma et al. Mata and Bähler 2003; Rustici et al. 2004).
2001, 2003; Spellman et al. 2002).
The focus of this chapter will be the genomics
of stress response in the fission yeast Schizosac-
charomyces pombe, highlighting some of the pub- II. Stress Response Pathways in Fission
lished work in this field as well as discussing the Yeast
similarities and differences with the other major
experimental free-living yeast, Saccharomyces cere- Before discussing the role that genomics ap-
visiae. Fission yeast is a single-celled fungus that proaches are beginning to have on the study and
belongs to the phylum Ascomycota and grows as understanding of stress response in fission yeast,
an elongating, rod-shaped cell. Prior to its intro- it is useful to provide a brief overview of what was
duction as a laboratory model organism, it was known before the application of newer techniques
used in the production of millet beer in eastern to the investigation. The study of stress response
Africa. The complete genome of S. pombe has been in fission yeast has benefited from earlier related
sequenced and published (Wood et al. 2002); it is research in other organisms. The discovery of two
comprised of three large chromosomes totalling related stress-activated protein kinases (SAPKs) in
approximately 13.8 Mb and a small (20 kb) mito- budding yeast (Brewster et al. 1993) and mammals
chondrial genome. There are somewhat fewer than (Han et al. 1994) suggested that this pathway
5000 genes (either predicted or experimentally con- represented an evolutionarily conserved way of
firmed) in the genome although, like the genes of signalling stress. The subsequent discovery of
most other organisms whose genomes have been a functional homolog of the Hog1/p38 protein,
sequenced, a majority of these do not yet have de- Sty1/Spc1, in S. pombe confirmed that a similar
fined cellular functions (Wood et al. 2002; Bähler pathway exists in fission yeast (Shiozaki and Rus-
and Wood 2003). sell 1995; Degols et al. 1996). However, the activity
Although often compared to S. cerevisiae, more of Hog1 functions specifically during osmotic
than 1 billion years of evolutionary divergence sep- stress, whereas Sty1 functions more broadly in
arates the two yeasts, according to recent estimates response to many stresses (Toone and Jones 1998).
(Heckman et al. 2001). Fission yeast has similar ex- Similarly, in mammalian cells the SAPK pathway
perimental advantages, including a low-complexity is involved in controlling a variety of processes
genome, as budding yeast, which has been widely (e.g. immune response, development, apoptosis,
used to pioneer functional genomic approaches cell proliferation, and DNA repair; Verheij et al.
(Delneri et al. 2001; Kumar and Snyder 2001). Al- 1996; Wilkinson et al. 1996; Molnar et al. 1997;
though it is impossible to declare either yeast to be Takenaka et al. 1998; Nebreda and Porras 2000).
the “best” experimental system, the case for study- Moreover, in both mammalian cells and fission
ing S. pombe as a complementary model for the yeast, but not in budding yeast, the SAPK pathway
biology of multicellular eukaryotes is strong. Nu- activates AP-1-like transcription factors (Toone
merous aspects of not only the structure of the and Jones 1999). Fission yeast therefore provides
genome (large chromosomes, genes with introns, a valuable model to study this central regulatory
large centromeres) but also the functioning of the pathway. Figure 6.1 shows a summary of the stress
genome (large-number homologs for conserved response pathway based on published work in the
Global Stress Response in S. pombe 99
Fig. 6.1. An overview of the stress signalling pathway in fis- indicate a repressive or inhibitory interaction, and shaded
sion yeast. The stress signalling cascade in S. pombe is shown dotted lines indicate interactions which are unclear. Ques-
with the various proteins being represented by named ovals. tion marks represent either unknown proteins in a pathway
Open arrows indicate activating interactions, closed lines or the possible existence of such a protein
field. It is important to note that in some cases, cade in fission yeast. Despite the role Sty1 plays in
the connections shown have not yet been clearly stress response, it was originally cloned as a sup-
defined and therefore represent targets of further pressor of lethality caused by the loss of protein
investigation, as outlined in Sect. IV below. phosphatase 2C activity (Millar et al. 1995). Sub-
The initial proteins in the cascade involved in sequent research confirmed a connection for this
the sensing of extracellular changes are poorly de- protein in both pathways. Sty1 is inhibited by the
fined, but the Mak proteins (also known as Phk Pyp1/2 proteins (Millar et al. 1995) where Pyp2
proteins) have been shown to be initial targets for is transcribed as a result of Sty1 activity via Atf1
activation during oxidative, temperature and radi- (Wilkinson et al. 1996), thus forming a negative
ation stress (Nguyen et al. 2000; Aoyama et al. 2001; feedback loop.
Buck et al. 2001). A “phosphorelay system” has Many of the proteins involved in the stress re-
been defined involving the Mak proteins, Mpr1 and sponse pathway, either as regulators such as Sty1 or
Mcs4, where the sensor kinase exhibits a histidine- as up-regulated targets such as Srk1 and Ste11, have
kinase activity and the response regulator con- been shown to be involved in normal cell growth
tains a phospho-accepting aspartate in its receiver processes and checkpoints (Shiozaki and Russell
domain. The phosphorylation signal is therefore 1995, 1997; Gachet et al. 2001). The connection be-
passed from sensor to acceptor and then goes on tween stress response and effects on cell growth
to activate either Win 1 or Wak1, depending on the would appear fairly intuitive; cells presumably will
level and type of stress, which results in the phos- have evolved in such a way as to block processes
phorylated activation of Wis1, a MAPKK. The AP- such as mitosis or meiosis if the metabolic environ-
1-like transcription factor Pap1 can also directly ment is unsuitable. However, the evolution of sec-
sense oxidative stress leading to nuclear accumula- ondary roles for these proteins under non-stressed
tion (Toone et al. 1998; Castillo et al. 2003; Vivancos growth is less clear. One simple explanation for
et al. 2004). the existence of proteins with multiple functions is
As illustrated in Fig. 6.1, the MAPK Sty1 is that it is a means of allowing the organism to “do
a central branch point for the stress signalling cas- more with less”, whereby each dual role could have
evolved in either direction (i.e. assume a new role in are major effectors of the stress signal, there were
the stress response for a protein normally involved effects which were specific to the sty1 deletion mu-
in mitotic growth, or the reverse). Regardless of the tant, suggesting that there were other downstream
origins of this complex behaviour, the net effect is targets of Sty1. A microarray screen of putative
that, because the stress response involves proteins regulatory genes expressed after various stresses
with multiple roles under different conditions, it is identified the srk1 (Sty1-regulated kinase 1) gene as
not at all straightforward to unambiguously dissect being up-regulated under several stress conditions,
the role of the proteins in the stress pathway. and these results were confirmed with conventional
northern blot analysis (Smith et al. 2002). Based on
its similarity to the Rck2 kinase in budding yeast,
III. Applications of Genomics which has been shown to be involved in regulating
Techniques to Stress Response stress response through interaction with the Hog1
protein (Bilsland-Marchesan et al. 2000), the func-
tion of Srk1 was further investigated. Srk1 is in the
As discussed above, while the field of stress re- same complex as Sty1 and, upon stress induction,
sponse in S. pombe is quite well established, the the protein translocates to the nucleus (Smith et al.
application of genomic approaches in this regard is 2002).
still just beginning. Initial genomics studies have This study demonstrates the basic utility of mi-
been published in the past 2–3 years and, although croarray experimentation in identifying hundreds
the amount of data currently available is not yet of potential targets for further investigation. While
broadly comparable to that of other organisms, it this approach is often used, the study by Chen et al.
does nevertheless provide some useful insight into (2003) discussed below represents a more compre-
the general effects of stress. The use of microarray- hensive attempt to characterize the global tran-
based studies of stress response could be classi- scriptional response to various stresses in fission
fied into one of three partially overlapping cate- yeast.
gories – broadly descriptive experiments, genome-
wide screening experiments, or hypothesis-driven
experiments – examples of each of these types of
investigations are discussed below. Although the
assignment of an experiment to one of these classes B. Identification of Core and Specific Stress
is rather arbitrary, it is probably nonetheless fair to Response Genes
say that work in the field is now beginning to move In one of the most comprehensive genomics studies
beyond the most straightforward types of investi- to date on the S. pombe stress response, the tran-
gations (broadly descriptive) to those which are, ar- scriptional responses to five major stresses were
guably, more interesting (hypothesis driven). That examined (heat shock, oxidative stress, osmotic
being said, much “obvious” work still needs to be stress, exposure to heavy metals, and DNA dam-
done, as this will provide an invaluable framework age; Chen et al. 2003). This study not only analysed
of reference for all other experiments. the commonalities between genes that were acti-
vated in different stresses, but also looked at the
A. Identifying Targets of Signalling Pathways dynamics of the response in order to characterize
the differences between transient versus sustained
One of the most straightforward applications of stress behaviour.
microarray technology is to conduct a screen for By using five different types of stress, Chen and
expression phenotypes under certain conditions. co-workers were able to identify genes whose ex-
This approach was used by Smith et al. (2002) in pression is altered either in specific stress responses
their investigation of the stress-induced signalling or in all stress types. Genes whose expression was
cascade in fission yeast. The Sty1/Spc1 MAPK had up- or down-regulated in all or most of the five
already been identified as a key component of the conditions tested were called core environmental
S. pombe stress response. As discussed above, this stress response (CESR) genes, while those that were
protein is activated by Wis1 and in turn activates specific to a single stress were called specific envi-
at least two transcription factors, Atf1 and Pap1. ronmental stress response (SESR) genes. The study
Analysis of the phenotypes of atf1 and pap1 dele- of these two groups of genes has provided useful
tion mutants suggested that, although these factors information regarding the “stress pathway”, which
nicely complements what is already known regard- that some pseudogenes were also induced raises
ing stress response in fission yeast. three possibilities: (1) that there are binding sites
Using conservative criteria to determine CESR for stress-induced transcription factors present in
genes, we generated a list of 140 genes which the promoters of these genes, which in some cases
were up-regulated (the entire list can be found at transcribe genes which are evolutionary remnants;
http://www.sanger.ac.uk/PostGenomics/S_pombe; (2) that normal transcriptional regulation or
see also Fig. 6.2) and 106 genes which were maintenance of transcriptional repression (partic-
down-regulated. As mentioned below, it is worth ularly of transposable elements and pseudogenes)
noting that a much higher number of genes are lost during the stress response; or (3) that the
were consistently up-regulated across all stress induction of pseudogenes has some regulatory
conditions tested, but failed to exceed the two-fold function, as is emerging in some cases (e.g.
cut-off used in the study. The size of the subset of Hirotsune et al. 2003). It is important to note that
CESR genes should therefore be seen as a fairly the functions of 80% of the genes in the CESR
conservative estimate. Of the remaining induced group are uncharacterized and, thus, deciding
genes, surprisingly few belonged to classes of whether they directly play a role in stress response
genes traditionally associated with stress such is not possible without further experiments.
as heat shock proteins and proteins involved in The set of genes that did not exhibit univer-
DNA repair. Although antioxidant genes were sal expression changes across all stress conditions
expressed, many others were involved in carbo- tested formed a continuum of partially overlap-
hydrate metabolism, signalling, or transcriptional ping and stress-specific groups. The largest ob-
regulation. The significance of the up-regulation served overlap between SESR genes was in the
of these genes is not clear. Furthermore, the fact heat and oxidative stress conditions where over 200
non-CESR genes were shared. At the other end of
the spectrum, the number of genes specifically in-
duced by only a single stress condition was gener-
ally low, with a high of 56 for oxidative stress and
a low of two for DNA alkylation damage caused
by MMS. Interestingly, significant overlaps exist
between induced non-CESR genes in the case of
MMS treatment and oxidative damage, suggesting
they share a common regulator, as has been pro-
posed by work in S. cerevisiae (Gasch et al. 2001).
Additionally, it is not surprising that the magnitude
of the stress response would depend in some cases
on the strength of the initial stress, as discussed be-
low, leading to variation in the size of gene sets in-
duced. This dataset provides a framework for com-
paring the transcriptional responses in S. pombe
cells to other external factors; for example, cis-
platin induces a gene expression program that par-
tially overlaps with the responses to both cadmium
and MMS (Gatti et al. 2004). Recent microarray
papers also revealed several genetic factors which
lead to a stress response: significant overlaps were
reported between CESR genes and (1) genes dere-
pressed in tup11 mutants (Fagerström-Billai and
Fig. 6.2. CESR gene induction and repression during various
stresses. Data from expression changes during five different Wright 2005), (2) genes derepressed in silencing
stress conditions measured at 0, 15 and 60 min after stress mutants (Hansen et al. 2005), and (3) genes induced
(Chen et al. 2003). The five stresses tested include oxida- in cells with shortening telomeres or circular chro-
tive damage with hydrogen peroxide (H2 O2 ), heavy metal mosomes (Mandell et al. 2005).
exposure with cadmium (Cd), heat, osmotic stress with sor-
bitol (Sorb.) and DNA damage through an alkylating agent Of the CESR genes that were found to be down-
(MMS). The 314 induced CESR genes are depicted in yellow regulated, there was no striking pattern in terms of
and the 420 repressed CESR genes in blue the classification of these genes, although enrich-
ment for transcripts coding for proteins involved poral aspects for gene expression in fission yeast.
in protein synthesis, transcription and cellular sig- By examining global expression profiles at 15 and
nalling was noted. Indeed, only few genes (11) were 60 min after treatment, it was possible to identify
found when using the same requirements for up- differences in transient and sustained changes in
regulated genes being present in four of the five expression induced by different stresses. While sor-
stresses. The requirement had to be lowered to bitol and MMS treatments induced rapid but tran-
three of five conditions to obtain a set similar in sient stress responses, other stresses induced sus-
size to that of the up-regulated genes. This suggests tained stress responses lasting more than 60 min
that genes actively repressed during stress may be (Fig. 6.2). The origin of these differences is not
more specifically tailored to the individual stress. clear, although several possible explanations exist.
This notion is supported by the observation that the The ability of the cells to sense or respond to the
degree of transcriptional repression, but not acti- different stresses may vary, as both sorbitol and
vation, was lower in two of the stress conditions MMS have a relatively weak effect on gene expres-
(MMS and sorbitol), indicating that strong tran- sion relative to the other stresses. Alternatively, the
scriptional repression is not a mandatory feature variation might reflect differences in how the cell
of all stress responses. Alternatively, the strength of deals with the stress or the time required for a stress
the stress in two cases may have been strong enough to become effective. In such a scenario, transcrip-
to induce a positive transcriptional response but tion factors activated by DNA damage stress may
too weak to interrupt normal cell proliferation and remain active until all the DNA damage has been
growth. This influence of response thresholds on repaired, even though the initial stress has been
transcriptional profiles has already been noted in removed.
some signalling pathways, and is also supported by As mentioned above, this study remains the
own recent unpublished work. most comprehensive global analysis of stress re-
To investigate this aspect of the stress response, sponse in S. pombe to date. Despite this fact, the
a follow-up study of this work is currently under- large amount of data generated and the limited
way (Chen et al., unpublished data). This new study space in typical journal articles, combined with
has focussed on oxidative stress, which has been a lack of available data regarding the function of
implicated in the etiology of several human diseases unknown proteins, mean that there is still a wealth
including cancer (Halliwell et al. 1992; Halliwell and of information about stress response to be uncov-
Gutteridge 1999) and is also a major contributor to ered from these experiments. This prospect is fur-
aging (Finkel and Holbrook 2000). Reactive oxygen ther discussed at the end of this chapter in Sect. IV.
species have deleterious effects when present in ex- A second global study of the transcriptional
cess, but they are also critical for cell proliferation stress response of fission yeast to ionising radia-
by modulating the activity of various signalling tion provides additional insight into both general
pathways (Torres 2003). The regulation of gene ex- and specific stress responses (Watson et al. 2004).
pression in response to various oxidants (hydrogen In these experiments, yeast cultures were exposed
peroxide, menadione and t-butylhydroperoxide) to 500 Gy of gamma radiation and mRNA sam-
and to different doses of oxidants has been carefully ples were taken for microarray analysis. Approxi-
dissected. As suggested from previous low-scale mately 200 genes were identified whose expression
analyses (Quinn et al. 2002), strong differences in changed more than two fold. Of these, almost 70%
the responses to low and high doses of oxidative were part of the fission yeast CESR genes identified
stress are evident (adaptive vs. acute response). At in the earlier study by Chen et al. (2003). Of those
least two signalling pathways are used to regulate genes which were not part of the CESR group, only
these distinct responses, with some cross-talk a minority had any defined role in stress whereas
occurring between pathways. A small core set the rest were mostly of metabolic or unknown
of genes induced independently of oxidant and function. While this may be unexpected (although
dose have also been identified, and these genes similar findings have been made in S. cerevisiae;
are all dependent on Pap1 expression. These data see references cited in Watson et al. 2004), it also
will help to understand how cells orchestrate highlights the power of using a microarray-based
and fine-tune their transcriptional programs in approach, as some of the transcripts coding for
response to various degrees of oxidative stress. proteins of unknown function represent conserved
One final aspect examined in the main global proteins which may play a vital role, and which may
study by Chen et al. (2003) was the role of tem- have been overlooked in a “targeted” investigation.
Interestingly, the number of CESR genes C. Identification of Mediating Factors

induced by the stress regulator Sty1 during of Global Response
ionising radiation treatment was much lower
(39%) than that in the other stresses tested In addition to providing direct evidence for the
(70%–90%). In addition, analysis of the dynam- effects of stress on the fission yeast transcriptome,
ics of stress-induced gene expression through genomics approaches for studying stress response
grouping of genes into either a fast or slow have also been able to help characterize events
response group showed that a majority of the which mediate the response and, in some cases,
fast CESR induced genes are dependent on Sty1, tailor the cellular response to the stress. Previous
Rad3 or both, while Chk1 was responsible for reports had demonstrated that the stress-induced
regulating the expression of more than double proteins Atf1 and Pap1 are responsible for a wide
the number of slow CESR genes that were de- variety of stress responses and for low level
pendent on Sty1 or Rad3. The notion that Rad3 H2 O2 /ROS damage respectively. While both of
and Sty1 are the only stress regulators is also these proteins are also known to be regulated by
not supported by the finding that, in both rad3 Sty1, the mechanism by which a distinction is
and sty1 deletion mutants, cta1 is still strongly made between activation of either of these proteins
up-regulated. was unknown. In a screen of genes involved in ox-
The microarray study also is consistent with idative but not osmotic stress, a gene was identified
the fact that Rad3 and the budding yeast protein that encodes the RNA-binding protein Csx1 which
Mec1 are structurally related to the mammalian was essential for the survival of S. pombe under
ATM gene. The model that oxidative damage oxidative stress (Rodriguez-Gabriel et al. 2003).
underlies some aspects of the phenotype of This work further demonstrated that Csx1 was
humans lacking ATM (resulting in the disease capable of binding to atf1 mRNA during oxidative
Ataxlia Telangiectasia) could be supported if ATM, stress, and that this binding strongly increased
through its similarity to the yeast stress regulators, the stability of atf1 mRNA. The stabilizing effect
can also regulate mammalian CESR genes (Watson of Csx1 enables the Atf1-dependent expression of
et al. 2004). Such an intriguing possibility not Pyp2, which in turn dephosphorylates Sty1 (see
only highlights the conservation of regulatory Fig. 6.3). This interesting demonstration of post-
functions in stress responses across an enormous transcriptional regulation was not a general effect,
evolutionary timescale, but also underscores the as the transcription factors Pap1 and Prr1, which
value in using “simple” model organisms to gain are also induced by oxidative stress, were unaf-
insight into more complex eukaryotic organisms. fected by Csx1 levels. The application of microar-
Fig. 6.3. Model of negative feedback regulation of Sty1. The circles represent either phosphorylation or dephosphoryla-
proteins involved in the regulation circuit of Sty1 signalling tion events whereas the other arrows denote either binding
in stress response are represented by named ovals. Curved or other predicted interactions. The atf1 gene and mRNA
open arrows in conjunction with closed arrows and shaded are represented by a rectangle and wavy line respectively
rays to answer the question of the extent of Csx1 sion yeast, S. cerevisiae responds to diverse stresses
regulation was enlightening. Through compar- through a number of different receptor/signalling
isons of mRNA expression levels in various deletion pathways to activate a similar set of genes as does S.
strains, it was shown that virtually all genes whose pombe (Rep et al. 1999; Gasch et al. 2000; Alexandre
expression was regulated by Atf1 are also depen- et al. 2001; Causton et al. 2001). An additional com-
dent on Csx1. In addition, almost 40 genes were mon feature, the fact that the strength of the initial
noted whose expression is dependent on Csx1, but stress is proportional to the magnitude of the ob-
not on Atf1 or its upstream regulator, Sty1. served expression changes, indicates that the ability
Two other examples where microarray infor- to fine-tune the stress response is either a feature
mation led to insight in specific stress regulation which was present in the last common ancestor
were recently published. The transcription factor of the two yeasts or that it is simply under strong
Zip1, which belongs to the same family as Atf1 positive selective pressure. Given the differences
and Pap1, was shown to be essential upon cad- in the regulation of the stress response in the two
mium exposure and to mediate the global response organisms, the latter seems a more likely possibil-
to cadmium including growth arrest; in normally ity. For genes whose expression is influenced only
growing cells, this stress response is suppressed by in specific stress conditions, the use of a variety of
constitutive degradation of the Zip1 protein (Har- stress-specific signalling pathways in budding yeast
rison et al. 2005). In a second recent paper, mi- provides an easy explanation for stress-specific re-
croarray data revealed an unexpected connection sponses. In fission yeast, there is no clear mech-
between the sterol biosynthetic pathway and the anism yet known for the activation of such SESR
fission yeast response to hypoxia: the cells mon- genes, although such a process should exist if the
itor oxygen-dependent sterol synthesis as an in- available experimental data are to be satisfactorily
direct measure of oxygen supply to launch a hy- explained.
poxic response when appropriate (Hughes et al. In fission yeast, the global study of various
2005). stresses discussed in Sect. II highlighted the fact
As demonstrated by these studies, microarrays that the induction of the CESR genes is regulated
represent a powerful tool not only for inves- through the SAPK pathway. This pathway is con-
tigating gene regulation in a direct sense, but served in both budding yeast as well as in mammals
also for rapidly putting new discoveries such (Fig. 6.4), indicating it probably had an ancient ori-
as indirect regulation through mRNA stability gin. In addition, the idea that the SAPK pathway is
into the context of known cellular behaviours,
and for the unbiased discovery of unexpected
connections.
IV. Correlations Between Budding

and Fission Yeast Stress Responses
The large number of global studies of stress re-

sponse as well as other cellular behaviours in bud-
ding yeast has allowed a comparison of the tran-
scriptional programs of these similar yet divergent
yeasts. As with fission yeast, there is a large set of
genes (about 900) whose expression is stereotypi-
cally up-regulated (300 genes) or down-regulated
(600 genes) in response to a broad range of cel- Fig. 6.4. Conservation of the SAPK signalling pathway
lular stresses (Gasch et al. 2000; Causton et al. amongst species. The components of the SAPK signalling
2001). Among the budding yeast core stress genes pathways of two yeasts and mammals are shown. The far
that contain fission yeast orthologs, many are also left-hand column indicates the initial signal which triggers
the cascade while some of the downstream targets are
stereotypically induced or repressed during stress shown in the far right-hand column. For the sake of clarity,
in fission yeast (Table 6.1; Chen et al. 2003). In con- only one simplified form of the signalling pathway for
trast to the single signalling pathway used by fis- mammals is shown
Table. 6.1. Core set of stress-induced genes in budding and fission yeastsa
S. cerevisiae S. pombe Annotation

ortholog ortholog
HSP12 hsp9 Heat shock protein, chaperone activity
HSP104 SPBC16D10.08c Heat shock protein, chaperone activity
HSP26 hsp16 Heat shock protein
HSP78 SPBC4F6.17c AAA family ATPase, mitochondrial chaperone
GLO1 glo1 Glyoxalase I, glutathione metabolism
GLO2 SPCC13B11.03c Hydroxyacylglutathione hydrolase, glyoxalase II family
RIB5 rib5 Riboflavin synthase
YKL151C SPCC61.03 Carbohydrate kinase family
MRF1 mrf1 Zinc-binding dehydrogenase
YJR096W SPAC2F3.05c Aldo/keto reductase, oxidoreductase
BSD2 SPAC328.07c Metal homeostasis protein, putative membrane transporter
PNC1 SPBC365.20c Pyrazinamidase/nicotinamidase
ATG3 SPBC3B9.06c Involved in autophagy
TPS2 SPAC3G6.09c Glycosyl transferase family, trehalose-phosphate synthase
SYM1 SPAC4G9.14 Member of Mpv17 and PMP22 family, peroxisomal protein
SRA1 cgs1 cAMP-dependent protein kinase (regulatory subunit)
YNL274C SPACUNK4.10 Putative hydroxyacid dehydrogenase
YBR056W exg3 Glucan 1,3-beta-glucosidase
GPD1 gpd2 Glycerol-3-phosphate dehydrogenase
GRX1 grx1 Glutaredoxin
GTT1 gst3 Glutathione transferase
TPS1 tps1 Trehalose-phosphate synthase
UBC8 ubc8 Ubiquitin conjugating enzyme
UGA2 SPAC1002.12c Succinate-semialdehyde dehydrogenase
UBP15 ubp22 Protease involved in deubiquitination
YDL124W SPAC19G12.09 Aldo-keto reductase family
YAK1 SPAC1E11.03 Protein kinase, negative regulator of cell growth
AYR1 SPAC23D3.11 1-acyl dihydroxyacetone phosphate reductase
FOX2 SPAC4H3.08 Involved in peroxisomal fatty acid beta-oxidation
PRB1 prb1 Vacuolar protease
MAG1 mag1 DNA-3-methyladenine glycosylase, involved in DNA repair
MOH1 SPAP691.02 Putative zinc-binding protein
SOD1 sod1 Cu, Zn-superoxide dismutase
YNL200C SPAC15A10.05c Conserved protein, unknown function
OXR1 SPAC8C9.16c Conserved protein, unknown function
YJR008W SPAC4H3.04c Conserved protein, unknown function
YMR090W SPBC216.03 Unknown function
YHR087W SPBC21C3.19 Unknown function
YOL032W SPBC21H7.06c Unknown function
MSC1 ish1 Unknown function
YNL115C SPAC23C11.06c Unknown function
YLL023C SPBC1539.04 Unknown function
YLR149C SPCC4G3.03 Unknown function
YNL305C SPCC576.04 Unknown function, similarity to apoptotic inhibitory molecule
a Among the genes which were stereotypically induced under stress conditions in either of two budding yeast studies
(Gasch et al. 2000; Causton et al. 2001), we identified the S. pombe orthologs which were also induced as CESR genes in
fission yeast (Chen et al. 2003)
derived from an ancestral one is also supported by stress response in S. cerevisiae via Hog1 (Cano and
findings that the overlap between genes shared in Mahadevan 1995), in mammals the pathway is part
stress response is much higher than for other cel- of a much more complex signalling network con-
lular processes such as the cell cycle and meiosis trolling normal development and function in addi-
(Mata et al. 2002; Rustici et al. 2004). Interestingly, tion to stress (reviewed in Roux and Blenis 2004).
the functional role of this pathway appears to have The adaptation of the SAPK signalling pathway in
somewhat diverged in the three species. Whereas different species therefore represents an interesting
this pathway now primarily regulates only osmotic example of evolutionary experimentation.
only a single ribosome, supporting the use of

V. Unanswered Questions and Future this methodology to identify transcript targets
Directions that are translationally regulated. In addition, the
study revealed a surprising finding that ribosome
Although the use of microarrays for global studies
density decreases with increasing ORF length,
on stress response in yeast has extended our knowl-
although the significance of these data is not clear.
edge in this field, it has not yet been able to pro-
A more recent study involved a similar frac-
vide a framework for other aspects of yeast biology
tionation approach but also used details of the
which overlap with, and definitely influence, the
mRNA distribution between polysome fractions
stress response. These areas of convergence there-
along with isotope-coded affinity tag (ICAT) to
fore represent potential further avenues of investi-
perform quantitative proteomics (MacKay et al.
gation that would not only add knowledge to how
2004). As in the previous study (Arava et al. 2003),
yeast deal with stress, but also allow for a more
most transcripts (about 80%) are associated with
integrated view of the organism which is the holy
polysomes while those which were not, including
grail of systems biologists.
S. cerevisiae genes such as GCN4 and HAC1, are
known to be translationally regulated. When grow-
A. Correlations Between Transcription ing cells were exposed to mating pheromone and
and Translation compared to untreated cells, 816 genes of 3874 as-
sayed were observed to change at least two fold
Despite their powerful potential for investigation, in their protein levels. Of these, 617 (about 75%)
microarrays have often been faulted as measur- were solely the result of changes in transcript lev-
ing only an intermediary in cell biology, namely els, while the remaining 199 showed alterations in
mRNA, rather than surveying the net effect (i.e. their translational efficiencies. Although the con-
the translated protein). While this criticism is fac- clusions from this study seem to indicate that only
tually valid, the impact of the implied error of this a small number of genes showed regulation at the
approach – that measured mRNA levels do not gen- level of translational regulation in response to mat-
erally correlate with protein levels – has not been ing pheromone, it is important to note the limi-
conclusively ascertained. In any case, it will ulti- tations of this approach. As the authors mention,
mately be necessary to integrate data from differ- small ORFs which show enrichment in monosome
ent levels of regulation to gain a comprehensive fractions may not be translationally regulated, as
picture of gene expression control. The study of their distribution would suggest, but may simply
post-transcriptional and translational regulation be limited in their polysome profile due to their
is an active and emerging field and, despite the short transcript length.
technical difficulties related to this research, sev- Along similar lines, a third paper dealing with
eral groups have used similar techniques to assess S. pombe but bypassing this issue of measuring
the role of translational regulation (Pradet-Balade mRNA intermediates has recently been published,
et al. 2001). looking at the fission yeast response to cadmium
Arava and co-workers, looking at budding exposure through amino acid coded mass tagging
yeast, studied the number of ribosomes associated (AACT), liquid chromatography and tandem mass
with various transcripts through velocity sedimen- spectrometry (LCMS/MS; Bae and Chen 2004). Us-
tation on a sucrose gradient (Arava et al. 2003). ing this approach based on protein measurements,
Although the experiments were not performed in it was demonstrated that 161 proteins were up- or
S. pombe, this analysis has nevertheless validated down-regulated (with threshold values of 1.3- and
an experimental procedure for global investiga- 0.78-fold respectively) in response to cadmium ex-
tions of this phenomenon. The results indicated posure. As approximately 400 genes were found to
that virtually all (>99%) mRNAs examined had be induced (>two fold) under cadmium stress in
a majority of their transcripts associated with one earlier studies (Chen et al. 2003), this would sug-
or more ribosomes and, of the small group for gest that the majority of alterations in gene tran-
which this was not true, neither unusual sequence scription do not strongly influence protein con-
motifs nor overrepresentation by functional class centration. These studies, however, are not directly
were found. Despite this fact, several transcripts comparable, as the experimental stress conditions
that are known to be regulated at the level of used were not identical and additionally, the mRNA
translation were found to be associated with studies indicated that Cd2+ response is transient
and decreases before the 1-h time point but the yeast in particular, histone modifications have been
proteomic study did not begin observations un- shown to be involved in connecting seemingly un-
til 1 h. It is therefore possible that faster changes related pathways, including stress response.
in protein concentrations, due to rapid turnover The silent mating-type locus in fission yeast
of proteins, might have been overlooked. In addi- contains three genes, mat1P/M, mat2P and mat3M,
tion, only 1133 proteins (or ∼25% of all annotated where the latter two genes are transcriptionally
genes) were actually identified in the translational silenced and serve as recombination material when
study, and thus a more comprehensive study would mating-type switching occurs (Beach and Klar
be required for a conclusive comparison. 1984). An investigation of this locus revealed that
These findings, which generally show a lim- the transcriptionally silenced region contained
ited effect of post-transcriptional regulation of histones enriched in histone H3 with methylated
mRNA on protein concentration, should not be lysine residue 9, whereas outside this region the
taken to indicate that translational regulation histones were enriched in H3 with methylated
does not represent an integral facet of overall lysine residue 4 (Noma et al. 2001). Furthermore,
expression regulation, along with mRNA stabil- this study demonstrated that the deletion of the
ity and transcriptional regulation, and protein inverted repeats found in the region allowed the
turnover. Indeed, as so little work has been done histone modification pattern to spread along the
to study the global effects of this phenomenon, chromosome, thereby interfering with the expres-
and even less related to stress response, it is sion of neighbouring genes. Most intriguingly,
quite possible that its importance continues to be a more recent investigation has shown that these
significantly underrated. There are therefore many histone modifications are actually targeted to
potential genomics-based experiments that could the locus by the RNAi machinery as well as by
be performed, using approaches similar to those the direct binding of the two well-characterized,
employed by Arava et al. (2003) and MacKay et al. stress-induced transcription factors Atf1 and Pcr1
(2004) where RNA in polysome preparations is (Jia et al. 2004).
hybridised on a microarray, to assess the extent Two recent publications also provide support
of post-transcriptional regulation under different for the idea that regulation of gene expression at
stress conditions. the level of histone modifications may be a uni-
versal feature of many metabolic pathways. The
first study relates to the Hog1 pathway in budding
B. Chromatin Remodelling in Response yeast. As discussed above, Hog1 is selectively in-
to Stress volved in response to osmotic stress. In studies to
characterize the connection between the transcrip-
There has been an enormous amount of literature tional responses from MAPK signalling and his-
published in the last decade that has highlighted the tone modifications, De Nadal et al. (2004) demon-
role of chromatin structure as it relates to gene ex- strated that Hog1 directly interacts with Rpd3, a hi-
pression across a wide range of eukaryotes, includ- stone deacetylase that acts mainly on histone H3
ing fission yeast. Some of the first research related and H4 tails. The deletion of this histone deacety-
to the acetylation of histone tails created a con- lase (HDAc) resulted in the reduction in expres-
ceptual link between changes in chromatin struc- sion of 80%–90% of the 222 genes previously up-
ture which could be correlated with transcriptional regulated in response to osmotic stress. The inter-
activation (Garcia-Ramirez et al. 1995). Since that action of Hog1 and Rpd3 was shown to be specific
time, a huge variety of histone modifications have for osmoresponsive promoters, and required for
been identified in addition to acetylation, including recruitment of the Rpd3–Sin3 deacetylase complex
phosphorylation (Nowak and Corces 2000), methy- to these promoters.
lation (Strahl et al. 1999), ubiquitination (Sun and The second study involves the well-known fis-
Allis 2002) and sumoylation (Shiio and Eisenman sion yeast stress proteins Atf1 and Pcr1. Through
2003), and functional roles defined for a number of analysis of the chromatin structure surrounding
them (reviewed in Jenuwein and Allis 2001; Morales the area of a CRE-like sequence in ade26-M26 that
et al. 2001; Li 2002; Vermaak et al. 2003). This ex- is bound by the two factors, it was demonstrated
panding body of research has resulted in the theory that osmotic stress, but not other stresses, induces
of a “histone code” which regulates gene expression an increase in the number of micrococcal nucle-
epigenetically in all eukaryotic genomes. In fission ase sites in the region (Hirota et al. 2004; Yamada
et al. 2004). Through additional deletion experi- known transcription factor, but not whether it is
ments, evidence was presented for the involvement immediately upstream. With the development of
of Tup11 and Tup12 in regulating chromatin struc- ChIP, it has been possible to detect the presence
ture in order to tailor transcriptional responses of a specific protein in a specific genomic location
to specific stresses (Greenall et al. 2002). Recent (such as the promoter of a target gene). This ap-
microarray experiments revealed that Tup11 and proach is useful but still has the severe limitations
Tup12 have both overlapping and specialized func- of requiring a priori knowledge of a candidate pro-
tions in stress response regulation (Fagerström- tein to immunoprecipitate, and being able to prac-
Billai and Wright 2005). Another recent microarray tically survey only select regions of the genome.
study points to a potential link between stress re- With the powerful combination of ChIP and mi-
sponse and chromatin remodelling: many of the croarrays, it is now possible, in a series of experi-
genes which were derepressed in silencing mu- ments, to map the binding sites of all stress-related
tants were activated during environmental stresses transcription factors across the genome. Such a col-
(Hansen et al. 2005), although further work is re- lection of information, along with information al-
quired to test whether the absence of silencing pro- ready published, would make it possible to piece to-
teins directly causes this correlation or whether gether a virtually complete transcriptional network
it reflects a secondary effect. These experiments, of stress response. Studies of the sort in budding
as well as others not discussed here, suggest that yeast have already been applied to define a closed
there is an important connection between stress loop of transcription factors that serially regulate
signalling and alterations in chromatin structure the cell cycle (Simon et al. 2001; Lee et al. 2002).
leading to transcriptional reprogramming. A similar approach in fission yeast under stress
While these studies highlight a clear role for would likely provide great insight into the frame-
chromatin modification in stress response in both work which previous research has established.
budding and fission yeast, there are still a vast
number of histone modifications whose role is un-
known. Given the wide variety of roles already de- D. Examining Subtle Changes Associated
fined for known histone modifications, it seems with Stress
extremely likely that others could be found to be
related to stress response. Lastly, all of the studies published to date have,
for practical but arbitrary reasons, used a cut-off
of two-fold change in expression as the level de-
C. Mapping Genomic Binding Sites of Stress fined as unregulated. The use of such a threshold is
Proteins related to the “noise” inherent in microarray exper-
iments. It is clear, however, that many of the more
The recent emergence of a technique that combines subtle changes in expression that are observed will
chromatin immunoprecipitation and microarray probably have an effect in concert with those more
analysis, referred to as ChIP-on-chip, has already dramatic changes. These unanalysed changes also
found many uses (Ren et al. 2000; Iyer et al. 2001; represent a potential wealth of knowledge regard-
Sun et al. 2003). The process involves treating cells ing stress response. However, the effort necessary
with a cross-linking agent (usually formaldehyde) to define with certainty the changes in expression
and then immunoprecipitating this DNA, after increases significantly as one looks at progressively
shearing, with a specific antibody against a DNA smaller changes.
interacting protein. The DNA which is bound to In addition to the actual experimental condi-
the protein by the cross-linking can be recovered tions being examined, gene expression patterns
and used for microarray analysis. This provides are also modulated in response to more subtle
a practical way to globally map the DNA binding physiological variations caused by differences
sites for any protein that either has a suitable in growth media, temperature, and cell density.
antibody available or has been epitope-tagged. Notably, the set of previously identified CESR
Many of the traditional studies of molecular genes (Chen et al. 2003) show adjustments to
pathways, such as intracellular signalling involved growth conditions. For example, CESR-induced
in stress response, often rely on indirect informa- genes are expressed at a slightly higher level
tion. For instance, a deletion mutant may indicate in minimal medium compared to full medium,
that a regulatory factor lies upstream of another and at 25 ◦ C compared to 30 ◦ C, and they show
increasing expression once cells reach densities Aoyama K, Aiba H, Mizuno T (2001) Genetic analysis of
higher than about 107 cells/ml (when they are the His-to-Asp phosphorelay implicated in mitotic cell
still exponentially growing). Thus, cells adapt and cycle control: involvement of histidine-kinase genes
of Schizosaccharomyces pombe. Biosci Biotechnol
fine-tune their gene expression programs to adjust Biochem 65:2347–2352
to even slight differences in the environment. Arava Y, Wang Y, Storey JD, Liu CL, Brown PO, Herschlag
These data indicate that a transcriptional response D (2003) Genome-wide analysis of mRNA translation
to, for example, environmental stress response is profiles in Saccharomyces cerevisiae. Proc Natl Acad
Sci USA 100:3889–3894
not an “all-or-nothing” effect but can be continu- Bae W, Chen X (2004) Proteomic study for the cellular
ously adjusted to different needs. The sensitivity responses to Cd2+ in Schizosaccharomyces pombe
of microarray analysis can therefore be used to through amino acid-coded mass tagging and liquid
reveal such subtle cellular responses and complex chromatography tandem mass spectrometry. Mol Cell
phenotypes which are not apparent with other Proteomics 3:596–607
Bähler J, Wood V (2003) The genome and beyond. In: Egel
methods (Burns et al., unpublished data). R (ed) The molecular biology of Schizosaccharomyces
pombe. Springer, Berlin Heidelberg New York, pp 13–
23
VI. Conclusions Beach DH, Klar AJ (1984) Rearrangements of the trans-
posable mating-type cassettes of fission yeast. EMBO J
3:603–610
The advantages provided by the use of genomics Bilsland-Marchesan E, Arino J, Saito H, Sunnerhagen P,
approaches to the study of fission yeast biology Posas F (2000) Rck2 kinase is a substrate for the os-
and, as discussed in this chapter, of stress response motic stress-activated mitogen-activated protein ki-
are numerous. It is also quite clear that while nase Hog1. Mol Cell Biol 20:3887–3895
Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spell-
such global approaches are useful as a sort of man P, Stoeckert C, Aach J, Ansorge W, Ball CA, Caus-
screening tool by themselves, their maximum ton HC et al. (2001) Minimum information about a mi-
value can be seen when genomics studies are croarray experiment (MIAME) – toward standards for
coupled with a well-designed, hypothesis-driven microarray data. Nat Genet 29:365–371
Brazma A, Parkinson H, Sarkans U, Shojatalab M, Vilo J,
experiment, based on knowledge from tradi- Abeygunawardena N, Holloway E, Kapushesky M,
tional experimental approaches. The difference Kemmeren P, Lara GG et al. (2003) ArrayExpress –
between these two approaches is equivalent to a public repository for microarray gene expression
the difference between having a map that shows data at the EBI. Nucleic Acids Res 31:68–71
cities and one that shows both the cities and the Brewster JL, de Valoir T, Dwyer ND, Winter E, Gustin MC
(1993) An osmosensing signal transduction pathway
roads connecting them. While it might be possible in yeast. Science 259:1760–1763
to identify interesting-looking cities without Buck V, Quinn J, Soto Pino T, Martin H, Saldanha J,
any other knowledge, knowing the roads which Makino K, Morgan BA, Millar JB (2001) Peroxide
connect them all together provides a much deeper sensors for the fission yeast stress-activated mitogen-
activated protein kinase pathway. Mol Biol Cell
understanding of the underlying “biology”. It is 12:407–419
clear that genomics studies in S. pombe and other Cano E, Mahadevan LC (1995) Parallel signal processing
organisms are only at their beginning, and global among mammalian MAPKs. Trends Biochem Sci
approaches will reach their full potential in the 20:117–122
coming years. Castillo EA, Vivancos AP, Jones N, Ayte J, Hidalgo E
(2003) Schizosaccharomyces pombe cells lacking
Acknowledgements. B. Wilhelm is supported by a Sanger the Ran-binding protein Hba1 show a multidrug
post-doctoral fellowship, and core funding for J. Bähler is resistance phenotype due to constitutive nuclear
provided by Cancer Research UK. We wish to express our accumulation of Pap1. J Biol Chem 278:40565–40572
gratitude to Dr. J. Mata and Dr. J.-R. Landry for critical Causton HC, Ren B, Koh SS, Harbison CT, Kanin E, Jen-
reading of this manuscript. nings EG, Lee TI, True HL, Lander ES, Young RA (2001)
Remodeling of yeast genome expression in response to
environmental changes. Mol Biol Cell 12:323–337
Chen D, Toone WM, Mata J, Lyne R, Burns G, Kivinen K,
Brazma A, Jones N, Bähler J (2003) Global transcrip-
References tional responses of fission yeast to environmental
stress. Mol Biol Cell 14:214–229
Alexandre H, Ansanay-Galeote V, Dequin S, Blondin B Degols G, Shiozaki K, Russell P (1996) Activation and reg-
(2001) Global gene expression during short-term ulation of the Spc1 stress-activated protein kinase in
ethanol stress in Saccharomyces cerevisiae. FEBS Lett Schizosaccharomyces pombe. Mol Cell Biol 16:2870–
498:98–103 2877
Delneri D, Brancia FL, Oliver SG (2001) Towards a truly messenger-RNA stability of its homologous coding
integrative biology through the functional genomics gene. Nature 423:91–96
of yeast. Curr Opin Biotechnol 12:87–91 Hughes AL, Todd BL, Espenshade PJ (2005) SREBP pathway
De Nadal E, Zapater M, Alepuz PM, Sumoy L, Mas G, responds to sterols and functions as an oxygen sensor
Posas F (2004) The MAPK Hog1 recruits Rpd3 histone in fission yeast. Cell 120:831–842
deacetylase to activate osmoresponsive genes. Nature Iyer VR, Horak CE, Scafe CS, Botstein D, Snyder M,
427:370–374 Brown PO (2001) Genomic binding sites of the yeast
Fagerström-Billai F, Wright APH (2005) Functional com- cell-cycle transcription factors SBF and MBF. Nature
parison of the Tup11 and Tup12 transcriptional core- 409:533–538
pressors in fission yeast. Mol Cell Biol 25:716–727 Jenuwein T, Allis CD (2001) Translating the histone code.
Finkel T, Holbrook NJ (2000) Oxidants, oxidative stress and Science 293:1074–1080
the biology of ageing. Nature 408:239–247 Jia S, Noma K, Grewal SI (2004) RNAi-independent
Forsburg SL (1999) The best yeast? Trends Genet 15:340–344 heterochromatin nucleation by the stress-activated
Gachet Y, Tournier S, Millar JB, Hyams JS (2001) ATF/CREB family proteins. Science 304:1971–1976
A MAP kinase-dependent actin checkpoint ensures Kumar A, Snyder M (2001) Emerging technologies in yeast
proper spindle orientation in fission yeast. Nature genomics. Nat Rev Genet 2:302–312
412:352–355 Lee TI, Rinaldi NJ, Robert F, Odom DT, Bar-Joseph Z, Ger-
Garcia-Ramirez M, Rocchini C, Ausio J (1995) Modulation of ber GK, Hannett NM, Harbison CT, Thompson CM, Si-
chromatin folding by histone acetylation. J Biol Chem mon I et al. (2002) Transcriptional regulatory networks
270:17923–17928 in Saccharomyces cerevisiae. Science 298:799–804
Gasch AP, Spellman PT, Kao CM, Carmel-Harel O, Eisen MB, Li E (2002) Chromatin modification and epigenetic repro-
Storz G, Botstein D, Brown PO (2000) Genomic expres- gramming in mammalian development. Nat Rev Genet
sion programs in the response of yeast cells to environ- 3:662–673
mental changes. Mol Biol Cell 11:4241–4257 MacKay VL, Li X, Flory MR, Turcott E, Law GL, Serikawa KA,
Gasch AP, Huang M, Metzner S, Botstein D, Elledge SJ, Xu XL, Lee H, Goodlett DR, Aebersold R et al. (2004)
Brown PO (2001) Genomic expression responses to Gene expression analyzed by high-resolution state
DNA-damaging agents and the regulatory role of the array analysis and quantitative proteomics: response
yeast ATR homolog Mec1p. Mol Biol Cell 12:2987–3003 of yeast to mating pheromone. Mol Cell Proteomics
Gatti L, Chen D, Beretta GL, Rustici G, Carenini N, Corna E, 3:478–489
Colangelo D, Zunino F, Bähler J, Perego P (2004) Global Mandell JG, Goodrich KJ, Bähler J, Cech TR (2005) Expres-
gene expression of fission yeast in response to cisplatin. sion of a recQ helicase homolog affects progression
Cell Mol Life Sci 61:2253–2263 through crisis in fission yeast lacking telomerase. J Biol
Greenall A, Hadcroft AP, Malakasi P, Jones N, Morgan BA, Chem 280:5249–5257
Hoffman CS, Whitehall SK (2002) Role of fission yeast Mata J, Bähler J (2003) Correlations between gene expres-
Tup1-like repressors and Prr1 transcription factor in sion and gene conservation in fission yeast. Genome
response to salt stress. Mol Biol Cell 13:2977–2989 Res 13:2686–2690
Halliwell B, Gutteridge JMC (1999) Free radicals in biology Mata J, Lyne R, Burns G, Bähler J (2002) The transcriptional
and medicine. Oxford University Press, Oxford Halli- program of meiosis and sporulation in fission yeast.
well B, Gutteridge JM, Cross CE (1992) Free radicals, Nat Genet 32:143–147
antioxidants, and human disease: where are we now? Millar JB, Buck V, Wilkinson MG (1995) Pyp1 and Pyp2
J Lab Clin Med 119:598–620 PTPases dephosphorylate an osmosensing MAP kinase
Han J, Lee JD, Bibbs L, Ulevitch RJ (1994) A MAP kinase controlling cell size at division in fission yeast. Genes
targeted by endotoxin and hyperosmolarity in mam- Dev 9:2117–2130
malian cells. Science 265:808–811 Molnar A, Theodoras AM, Zon LI, Kyriakis JM (1997)
Hansen KR, Burns G, Mata J, Volpe TA, Martienssen RA, Cdc42Hs, but not Rac1, inhibits serum-stimulated
Bähler J, Thon G (2005) Global effects on gene expres- cell cycle progression at G1/S through a mechanism
sion in fission yeast by silencingand RNA interference requiring p38/RK. J Biol Chem 272:13229–13235
machineries. Mol Cell Biol 25:590–601 Morales V, Giamarchi C, Chailleux C, Moro F, Marsaud V,
Harrison C, Katayama S, Dhut S, Chen D, Jones N, Bäh- Le Ricousse S, Richard-Foy H (2001) Chromatin struc-
ler J, Toda T (2005) SCFPof1-ubiquitin and its target ture and dynamics: functional implications. Biochimie
Zip1 transcription factor mediate cadmium response 83:1029–1039
in fission yeast. EMBO J 24:599–610 Nebreda AR, Porras A (2000) p38 MAP kinases: beyond the
Heckman DS, Geiser DM, Eidell BR, Stauffer RL, Kar- stress response. Trends Biochem Sci 25:257–260
dos NL, Hedges SB (2001) Molecular evidence for the Nguyen AN, Lee A, Place W, Shiozaki K (2000) Multistep
early colonization of land by fungi and plants. Science phosphorelay proteins transmit oxidative stress signals
293:1129–1133 to the fission yeast stress-activated protein kinase. Mol
Hirota K, Hasemi T, Yamada T, Mizuno KI, Hoffman CS, Biol Cell 11:1169–1181
Shibata T, Ohta K (2004) Fission yeast global repres- Noma K, Allis CD, Grewal SI (2001) Transitions in
sors regulate the specificity of chromatin alteration in distinct histone H3 methylation patterns at the
response to distinct environmental stresses. Nucleic heterochromatin domain boundaries. Science
Acids Res 32:855–862 293:1150–1155
Hirotsune S, Yoshida N, Chen A, Garrett L, Sugiyama F, Nowak SJ, Corces VG (2000) Phosphorylation of histone H3
Takahashi S, Yagami K, Wynshaw-Boris A, Yoshiki A correlates with transcriptionally active loci. Genes Dev
(2003) An expressed pseudogene regulates the 14:3003–3013
Pradet-Balade B, Boulme F, Beug H, Mullner EW, Garcia- with transcriptionally active nuclei in Tetrahymena.
Sanz JA (2001) Translation control: bridging the gap Proc Natl Acad Sci USA 96:14967–14972
between genomics and proteomics? Trends Biochem Sun ZW, Allis CD (2002) Ubiquitination of histone H2B
Sci 26:225–229 regulates H3 methylation and gene silencing in yeast.
Quinn J, Findlay VJ, Dawson K, Millar JB, Jones N, Mor- Nature 418:104–108
gan BA, Toone WM (2002) Distinct regulatory pro- Sun LV, Chen L, Greil F, Negre N, Li TR, Cavalli G, Zhao H,
teins control the graded transcriptional response to Van Steensel B, White KP (2003) Protein-DNA inter-
increasing H(2)O(2) levels in fission yeast Schizosac- action mapping using genomic tiling path microar-
charomyces pombe. Mol Biol Cell 13:805–816 rays in Drosophila. Proc Natl Acad Sci USA 100:9428–
Ren B, Robert F, Wyrick JJ, Aparicio O, Jennings EG, Simon I, 9433
Zeitlinger J, Schreiber J, Hannett N, Kanin E et al. (2000) Sunnerhagen P (2002) Prospects for functional genomics in
Genome-wide location and function of DNA binding Schizosaccharomyces pombe. Curr Genet 42:73–84
proteins. Science 290:2306–2309 Takenaka K, Moriguchi T, Nishida E (1998) Activation of the
Rep M, Reiser V, Gartner U, Thevelein JM, Hohmann S, protein kinase p38 in the spindle assembly checkpoint
Ammerer G, Ruis H (1999) Osmotic stress-induced and mitotic arrest. Science 280:599–602
gene expression in Saccharomyces cerevisiae requires Toone WM, Jones N (1998) Stress-activated signalling path-
Msn1p and the novel nuclear factor Hot1p. Mol Cell ways in yeast. Genes Cells 3:485–498
Biol 19:5474–5485 Toone WM, Jones N (1999) AP-1 transcription factors in
Rodriguez-Gabriel MA, Burns G, McDonald WH, Martin V, yeast. Curr Opin Genet Dev 9:55–61
Yates JR III, Bähler J, Russell P (2003) RNA-binding Toone WM, Kuge S, Samuels M, Morgan BA, Toda T, Jones N
protein Csx1 mediates global control of gene expres- (1998) Regulation of the fission yeast transcription fac-
sion in response to oxidative stress. EMBO J 22:6256– tor Pap1 by oxidative stress: requirement for the nu-
6266 clear export factor Crm1 (Exportin) and the stress-
Roux PP, Blenis J (2004) ERK and p38 MAPK-activated pro- activated MAP kinase Sty1/Spc1. Genes Dev 12:1453–
tein kinases: a family of protein kinases with diverse 1463
biological functions. Microbiol Mol Biol Rev 68:320– Torres M (2003) Mitogen-activated protein kinase pathways
344 in redox signaling. Front Biosci 8:d369–391
Rustici G, Mata J, Kivinen K, Lio P, Penkett CJ, Burns G, Verheij M, Bose R, Lin XH, Yao B, Jarvis WD, Grant S,
Hayles J, Brazma A, Nurse P, Bähler J (2004) Periodic Birrer MJ, Szabo E, Zon LI, Kyriakis JM et al. (1996)
gene expression program of the fission yeast cell cycle. Requirement for ceramide-initiated SAPK/JNK sig-
Nat Genet 36:809–817 nalling in stress-induced apoptosis. Nature 380:75–79
Shiio Y, Eisenman RN (2003) Histone sumoylation is asso- Vermaak D, Ahmad K, Henikoff S (2003) Maintenance of
ciated with transcriptional repression. Proc Natl Acad chromatin states: an open-and-shut case. Curr Opin
Sci USA 100:13225–13230 Cell Biol 15:266–274
Shiozaki K, Russell P (1995) Cell-cycle control linked to Vivancos AP, Castillo EA, Jones N, Ayte J, Hidalgo E (2004)
extracellular environment by MAP kinase pathway in Activation of the redox sensor Pap1 by hydrogen per-
fission yeast. Nature 378:739–743 oxide requires modulation of the intracellular oxidant
Shiozaki K, Russell P (1997) Stress-activated protein kinase concentration. Mol Microbiol 52:1427–1435
pathway in cell cycle control of fission yeast. Methods Watson A, Mata J, Bähler J, Carr A, Humphrey T (2004)
Enzymol 283:506–520 Global gene expression responses of fission yeast to
Simon I, Barnett J, Hannett N, Harbison CT, Rinaldi NJ, ionizing radiation. Mol Biol Cell 15:851–860
Volkert TL, Wyrick JJ, Zeitlinger J, Gifford DK, Jaakkola Wilkinson MG, Samuels M, Takeda T, Toone WM, Shieh JC,
TS et al. (2001) Serial regulation of transcriptional reg- Toda T, Millar JB, Jones N (1996) The Atf1 transcription
ulators in the yeast cell cycle. Cell 106:697–708 factor is a target for the Sty1 stress-activated MAP
Smith DA, Toone WM, Chen D, Bähler J, Jones N, Mor- kinase pathway in fission yeast. Genes Dev 10:2289–
gan BA, Quinn J (2002) The Srk1 protein kinase is 2301
a target for the Sty1 stress-activated MAPK in fission Wood V, Gwilliam R, Rajandream MA, Lyne M, Lyne R,
yeast. J Biol Chem 277:33411–33421 Stewart A, Sgouros J, Peat N, Hayles J, Baker S et al.
Spellman PT, Miller M, Stewart J, Troup C, Sarkans U, (2002) The genome sequence of Schizosaccharomyces
Chervitz S, Bernhart D, Sherlock G, Ball C, Lepage pombe. Nature 415:871–880
M et al. (2002) Design and implementation of microar- Yamada T, Mizuno KI, Hirota K, Kon N, Wahls WP, Hart-
ray gene expression markup language (MAGE-ML). suiker E, Murofushi H, Shibata T, Ohta K (2004) Roles of
Genome Biol 3:research0046 histone acetylation and chromatin remodeling factor
Strahl BD, Ohba R, Cook RG, Allis CD (1999) Methylation of in a meiotic recombination hotspot. EMBO J 23:1792–
histone H3 at lysine 4 is highly conserved and correlates 1803
7 Programmed Cell Death and Apoptosis in Fungi
M. Ramsdale1
CONTENTS D. Nucleases . . . . . . . . . . . . . . . . . . . . . . . . . 132

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 113 E. Endogenous BCL-2-Family Proteins . . . . 133
A. Programmed Cell Death in Model VI. Global Responses of Dying Cells . . . . . . . . . . 133
Organisms . . . . . . . . . . . . . . . . . . . . . . . . 114 VII. Fungi as Models for Apoptosis . . . . . . . . . . . 134
II. Fungal Cell Death Responses . . . . . . . . . . . . . 115 A. Genetic Screens . . . . . . . . . . . . . . . . . . . . 134
A. Autophagy . . . . . . . . . . . . . . . . . . . . . . . . 115 1. Overexpression-Induced Lethality . . . . 134
B. Autolysis . . . . . . . . . . . . . . . . . . . . . . . . . . 116 2. Proteasome-Dependent Lethality . . . . . 135
C. Aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 3. Bax Suppressors . . . . . . . . . . . . . . . . . . 136
D. Somatic Incompatibility . . . . . . . . . . . . . . 116 4. Caspase-Family Proteins . . . . . . . . . . . 136
E. Meiotic Death, Spore Discharge 5. Other Non-Yeast Death Proteins . . . . . 137
and Spore Germination . . . . . . . . . . . . . . 117 VIII. Evolutionary Considerations . . . . . . . . . . . . . 137
F. Development and Tissue Formation . . . . 117 IX. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 138
III. Stimulating Fungal Apoptosis . . . . . . . . . . . . 118 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
A. Physical and Environmental Insults . . . . . 118
1. UV Irradiation . . . . . . . . . . . . . . . . . . . 118
2. Sugar . . . . . . . . . . . . . . . . . . . . . . . . . . 118
3. Weak Acids . . . . . . . . . . . . . . . . . . . . . . 119
4. Salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Abbreviations: cAMP, adenosine 3 ,5 cyclic
5. Hydrogen Peroxide . . . . . . . . . . . . . . . . 120 monophophate; CDRE, calcineurin-dependent
6. Mating Pheromone . . . . . . . . . . . . . . . . 120 response element; Cyt c, cytochrome c; dsDNA,
B. Antifungal Agents and Cytotoxic Drugs . . 120 double-stranded DNA; ER, endoplasmic reticu-
1. Amphotericin B . . . . . . . . . . . . . . . . . . 120
2. Pradimicin A . . . . . . . . . . . . . . . . . . . . 121 lum; EM, electron microscopy; FACS, fluorescence
3. Osmotin . . . . . . . . . . . . . . . . . . . . . . . . 121 activated cell sorter; PCD, programmed cell
4. Sphingosines and Ceramides . . . . . . . . 121 death; PI, propidium iodide; PMN, piece-meal
5. Translation Inhibitors . . . . . . . . . . . . . 121 microautophagy of the nucleus; ROS, reactive
6. Histatins . . . . . . . . . . . . . . . . . . . . . . . . 124
C. Lethal Mutations . . . . . . . . . . . . . . . . . . . 124 oxygen species; STRE, general stress response
1. CDC48 . . . . . . . . . . . . . . . . . . . . . . . . . 124 element; TCA, tricarboxylic acid; TUNEL, termi-
2. Cell Division Cycle and DNA Repair nal deoxynucleotidyl transferase-mediated dUTP
Mutants . . . . . . . . . . . . . . . . . . . . . . . . 125 nick-end labelling; VDAC, voltage-dependent
3. ASF1/CIA1 . . . . . . . . . . . . . . . . . . . . . . 125 anion channel; YRE, Yap1 response element
4. mRNA Processing Mutants . . . . . . . . . 126
IV. Death-Related Signalling Pathways . . . . . . . . 126
A. Ras and cAMP . . . . . . . . . . . . . . . . . . . . . 126
B. Phosphoinositide 3-OH Kinase
and Protein Kinase B Signalling . . . . . . . . 127 I. Introduction
C. CDC55, TPD3 (Protein Phosphatase 2A) . 127
D. Calcineurin and Calcium . . . . . . . . . . . . . 127
E. Sphingolipids, Ceramides and Protein Programmed cell death (PCD) has been described
Kinase C . . . . . . . . . . . . . . . . . . . . . . . . . . 128 in nearly all of the major life forms including
F. MAP Kinase Cascades . . . . . . . . . . . . . . . 128 eubacteria (Yarmolinsky 1995; Hochman 1997;
V. Commitment, Effectors and Suppressors . . . 128 Engelberg-Kulka and Glaser 1999; Matsuyama
A. Reactive Oxygen Species . . . . . . . . . . . . . 129
B. Cytochrome c et al. 1999; Lewis 2000), plants (Lam 2004), protists
and Mitochondrial Function . . . . . . . . . . 130 (Ameisen 1996) and animals (Golstein et al.
C. Caspases . . . . . . . . . . . . . . . . . . . . . . . . . . 131 2003). To date, PCD has not been described in the
1Aberdeen Fungal Group, School of Medical Sciences, Institute of
archaebacteria, but this may reflect a failure to
Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen examine this issue, rather than its absence. The
AB25 2ZD, UK ubiquitous nature of PCD hints either at its ancient
The Mycota XIII
Fungal Genomics
114 M. Ramsdale
origin, or that it is an aspect of the life of both heat (Metzstein et al. 1998). Type I mammalian
unicellular and multicellular organisms which is so PCD responses (including apoptosis) are defined
important that it has evolved many times over. In by a set of morphological and biochemical
either case, it is apparent that it plays an important hallmarks which include a rapid loss of mito-
part in the life histories of most organisms. chondrial membrane potential, externalisation
Differences in the nature of death responses are of the phospholipid phosphatidylserine on the
often reflected in basic differences in the cell biol- plasma membrane, DNA fragmentation and cell
ogy of the organisms under consideration. Whilst shrinkage. Later stages of apoptosis include the
many have questioned that PCD as a notion has any blebbing of the plasma membrane and the enclo-
relevance to the biology of unicellular organisms sure of organelles in membrane-bound sacs. In
(Fraser and James 1998), it is now becoming ap- numerous situations, cell suicide involves type II,
parent that at a deep-rooted level, related molecules or autophagic, cell death which is characterized by
are taking part in the cell death decisions of organ- the targeting of organelles (golgi, polyribosomes
isms as diverse as bacteria, yeast, plants, worms, and ER) to the lysosome prior to the onset of
flies and man. The study of these molecules and nuclear pyknosis.
the contributions they make to cell death processes At a molecular level, these events are mediated
in fungi is the key focus of this review. by either internal (intrinsic) or external (extrin-
To date, only a small number of studies have sic) cell suicide programs. In the intrinsic death
utilized genomic technologies such as transcript pathway, death signals induce the release of mito-
profiling in assessments of fungal cell death (e.g. chondrial proteins such as cytochrome c (Cyt c;
Allen et al. 2003; Orlandi et al. 2004). Further- Liu et al. 1996) and Smac/Diablo (Du et al. 2000;
more, no studies have yet been published which Verhagen et al. 2000), through a poorly under-
use proteomic technologies to investigate fungal stood mechanism which may involve the formation
PCD. Consequently, this review largely focuses on of pores by pro-apoptotic BCL-2-family proteins
what we do know about the biology and genetics and/or the opening of the permeability transition
of fungal programmed cell death, in the hope of pore. Once in the cytosol, Smac/Diablo releases
stimulating interest in this field amongst the wider inhibitor-of-apoptosis-protein (IAP)-mediated in-
mycological community. Both genomics and tran- hibition of procaspases, and Cyt c forms a com-
script profiling have been used successfully in the plex with Apaf-1 and procaspase-9. Activation of
dissection of the molecular events which accom- procaspases-9 results in an amplification of the
pany programmed cell death responses in animals, caspase cascade, cleavage of a range of intracel-
and these technologies could be very usefully em- lular substrates and ultimately cell death (Liu et al.
ployed to reveal the molecular basis of such events 1996; Li et al. 1997; Zou et al. 1997). In the extrinsic
in fungi. Identifying the endogenous molecular pathway, signals are mediated by death receptors
switches which trigger and execute PCD responses of the TNF receptor superfamily which can activate
in animals has been of enormous interest to the the caspase cascade more directly. Caspases are
wider medical research community. Knowledge of regarded as the key to apoptotic death responses
the similarities and differences between animal and in mammals because their activation represents
fungal PCD might be conveniently exploited in the a point of no return.
search for novel antifungal agents or targets. In or- Early bioinformatic screens of fungal genomes
der to appreciate what is known about the fungal revealed very few pro- or anti-apoptotic sequences
death response, it is meaningful to start with an which might be expected to participate in a fun-
overview of cell death in animals. gal PCD response (Fraser and James 1998). Whilst
this was initially disappointing, it does increase
the feasibility of developing fungus-specific drug
A. Programmed Cell Death therapies. As a result of more complex bioinfor-
in Model Organisms matic screens and the functional analysis of many
genes in fungi, it appears that some overlap may
In animals, various forms of PCD are used to exist. Indeed, caspase-related, BCL-2-related and
eliminate cells during the course of development, AIF-like molecules have recently been identified
to eradicate cells which have undergone can- (Madeo et al. 2002a). Furthermore, it is now appar-
cerous transformations, or which are damaged ent that suicide pathways exist in animals which
by microbial infection, mutagens, oxidants and are caspase-independent and involve the transloca-
Fungal Cell Death 115
tion of AIF from the mitochondrion to the nucleus, as well as Apg1 (a serine/threonine protein kinase)
particularly in response to reactive oxygen species in complex with Apg13, Apg17, Vac8 and Cvt9
(ROS)-induced stress (Lockshin and Zakeri 2004). (Kamada et al. 2000). Protein conjugation requires
Evidently, there is more than one way to die. Apg5, Apg7, Apg10 and Apg12, which are involved
with ubiquitin-like interactions maintained in
a complex with Apg16 (a coiled-coil protein which
II. Fungal Cell Death Responses binds to Apg5). Autophagosome size is determined
by Aut2 and Aut7, whilst docking of the autophago-
somes requires Vam3, Vam7, Vps11, Vps16, Vps18,
Death events in fungi can be instigated by
Vps33, Vps39 and Vps41, along with a functional
harsh physical and chemical insults, or can be
Rab GTPase encoded by YPT7. Degradation of
developmentally regulated. The organised dis-
autophagosome contents is mediated by a lipase
integration of yeast cells and hyphae is linked
(Cvt18), the protease Prb1 and requires a number
with phenomena such as autophagy and autolysis,
of vacuolar ATPases (VMA-encoded genes; see
aging/senescence (both chronological and replica-
Klionsky and Emr 2000).
tive), somatic and sexual incompatibility, hyphal
Piecemeal microautophagy of the nucleus
interference, spore discharge mechanisms and the
(PMN) has recently been described in Saccha-
sculpturing of complex tissues during fruit-body
romyces cerevisiae by Roberts et al. (2003). Whilst
formation.
the nucleus has not previously been considered
a target for autophagy because it is an essential
A. Autophagy organelle, it appears that nuclear blebs are pinched
from the nucleus into the vacuolar lumen where
Autophagy has been described in numerous physi- they can be degraded. PMN permits selective
ological forms of cell death, including during insect degradation of nonessential nuclear components
development (e.g. insect metamorphosis), mam- whilst preserving chromatin, which is selectively
malian embryogenesis and human neurodegener- excluded from the blebs. PMN occurs at nucleus–
ative disease (Cataldo et al. 1995; Anglade et al. vacuole (NV) junctions, which are formed through
1997; Migheli et al. 1997). Starvation-induced soro- physical interactions between Vac8 in the vacuole
carp formation in Dictyostelium discoideum has membrane and Nvj1 in the nuclear envelope. The
also been linked to autophagic cell death (Cornillon link between PMN and death responses needs
et al. 1994). Autophagy has also been demonstrated further exploration, but it could be important as
in filamentous fungi, both during nutrient starva- a model for autophagy and apoptotic blebbing in
tion (Dementhon et al. 2003) and also as part of higher eukaryotes.
the heterokaryon incompatibility response (Pinan- Both Ras and Tor signals may be involved
Lucarre et al. 2003). Autophagy is distinct from with the regulation of fungal autophagic death
apoptosis but does appear to be required for its responses. Budovskaya et al. (2004) showed that
efficient implementation (Klionsky and Emr 2000; the Ras/protein kinase A signalling pathway
Uchiyama 2001). plays an important role in regulating the onset
Since autophagy appears to be highly con- of autophagy in Saccharomyces cerevisiae. In-
served in animals, plants and fungi (Klionsky creased signalling through Ras leads to a block
and Emr 2000), studies of autophagy in yeast in autophagy, and also inhibits the cytoplasm
have been important in elucidating the molecular to vacuole targeting (CVT) pathway which is
processes involved with the onset and completion responsible for the delivery of a specific subset
of autophagic processes. Autophagy involves the of vacuolar proteins in growing cells. These
sequestration of organelles in a double membrane- findings contrast with the situation in mammals
enclosed vesicle termed the autophagosome where overexpression of Ras results in caspase-
(Klionsky and Ohsumi 1999). The autophago- independent autophagic cell death (Chi et al. 1999),
somes, under the control of the cytoskeleton, fuse emphasizing the differences between mammals
to the vacuolar membrane and then discharge their and fungi. TOR signalling is required for normal
contents into the vacuolar lumen. Several mutants autophagy and stationary-phase entry (Cutler
defective in autophagy have been identified. et al. 2001), and is also linked to the senescence
Signalling of autophagic events requires the phenotypes of Podospora anserina (Dementhon
TOR2 and Ras/protein kinase A/cAMP pathways et al. 2003).
116 M. Ramsdale
B. Autolysis mutases or catalases (Barker et al. 1999; Wawryn

et al. 1999), along with exposure to oxygen at ele-
In Penicillium chrysogenum, carbon-limited cul- vated or reduced partial pressures or the addition
tures undergo autolysis which is associated with of glutathione, all have the ‘expected’ effects on
the production of ROS (Sámi et al. 2001). Autolysis lifespan (Nestelbacher et al. 2000), consistent with
is also a striking feature of foraging mycelia, which the oxygen theory of aging (Harman 1962).
reallocate resources from one part of the mycelium Laun et al. (2001) found that aged mother cells
so that they can invest it in the growth of another of S. cerevisiae contained ROS of mitochondrial
part (Cooke and Rayner 1984). Autolysis is clearly origin, and that older cells displayed markers
linked to nutritional deprivation in fungi, and may of apoptosis including positive TUNEL labeling,
represent one extreme of the autophagic death re- annexin-V binding and chromatin fragmentation.
sponse described above. The accumulation of mini-rDNA circles which has
Mousavi and Robson (2003) demonstrated been described in aged yeasts (Sinclair and Guar-
that autolytic cultures of Aspergillus fumigatus ente 1997) could also be attributed to ROS-induced
displayed several characteristic features of apop- alterations in the activity of DNA damage repair
totic cells, including annexin-V binding to the genes – a hypothesis which should be readily
outer plasma membrane and TUNEL staining of testable with the large bank of DNA damage repair
nuclear material. Moreover, they were able to show mutants available in yeast.
that autolysis was an active process which could Chronological aging sets a limit to the time
be inhibited by the addition of cycloheximide to a culture of cells can remain viable after the onset
cultures, and that the apoptotic phenotype itself of the stationary phase of growth. Terminal cells
could also be abolished by the addition of the in aged yeast populations exhibit phenotypic char-
pan-caspase inhibitor Z-FAD-fmk. Caspase-1- acteristics reminiscent of those seen during mam-
and -8-like activities (but not caspase-3-like malian apoptosis, and chronological longevity of
activity) could be detected in autolysing cultures yeast cultures is temporarily extended by delet-
using p-nitroaniline-tagged peptide substrates, ing the yeast metacaspase YCA1 (Laun et al. 2001;
emphasizing the link to animal apoptosis. Herker et al. 2004).
C. Aging D. Somatic Incompatibility
The biology and genetics of aging/senescence in Many fungi undergo a vegetative or heterokaryon
filamentous fungi have recently been reviewed in incompatibility response (Leslie and Zeller 1996;
this series and elsewhere (Osiewacz 2004; Osiewacz Glass et al. 2000) which has been likened to
and Hamann 2006); it is therefore not necessary to apoptosis in animals (Aylmore and Todd 1986;
duplicate these efforts here. A number of studies on Ainsworth et al. 1990; Jacobson et al. 1998;
yeasts have, however, highlighted the close link be- Ramsdale 1998). PCD in this situation is linked
tween programmed cell death responses and aging to the recognition of non-self (indicated by allelic
and are therefore briefly covered in this review. differences at so-called vegetative or heterokaryon
Two types of aging are recognized in yeasts – incompatibility loci), which may operate to prevent
replicative aging and chronological aging. The non-self takeover reactions (Rayner 1991) or to
death of aged cells is considered ‘useful’ because prevent the spread of harmful genetic elements
it limits the spread of deleterious mutations and such as viruses (Hartl et al 1975; Anagnostakis
other ‘acquired’ characteristics such as oxidized and Day 1979; Debets et al 1994; van Diepeningen
proteins which preferentially segregate in the et al. 1997). Although plasmids, mitochondria
mother cells (Lai et al. 2002). and viruses can be spread between incompatible
Replicative aging is measured as the number of mycelia, this usually occurs at a much reduced
cell divisions which can be undertaken by a mother frequency when compared to compatible reactions
cell before it ceases to divide. Typically, any one (Cortesi et al. 2001; Biella et al. 2002).
mother cell can divide about 25–35 times before During incompatibility responses, hyphae
the onset of senescence (Jazwinski 1993). Genetic progress through a series of stages marked by
alterations and environmental changes which in- cytoplasmic condensation and granulation, vac-
crease levels of ROS lead to a shortening of the uolization, and shrinkage of the plasma membrane
lifespan of mother cells. Deletion of superoxide dis- away from the hyphal wall. Somatic incompat-
ibility is often associated with the generation it may have adaptive value in the elimination of
of ROS (possibly through the action of phenol- arrested basidia, thereby promoting the more effi-
oxidase enzyme systems such as laccase) and an cient utilization of resources (Money 2003). Whilst
increase in proteolytic activity (Esser and Blaich meiotic apoptosis has been observed in spo11 mu-
1994). EM studies of rejection responses reveal tants of mice (Baudat et al. 2000), within fungi it
nuclear degradation (which is often restricted seems only to play a role in the basidiomycetes,
to one partner in an interaction; Aylmore and since homologous sporulation mutants in N. crassa,
Todd 1986) and the accumulation of vesicular S. cerevisiae and Schizosaccharomyces pombe arrest
bodies. Marek et al. (2003) examined the cytology but do not appear to die (Lu et al. 2003).
of the heterokaryon incompatibility responses Raju and Perkins (2000) described PCD as
and senescence in Neurospora crassa and were a part of the normal development of the as-
able to demonstrate TUNEL-positive nuclei in cospores of the homothallic fungus Coniochaete
transformants containing incompatible het-c tetraspora. In this fungus, immature asci contain
alleles. eight uninucleate ascospores. Typically, however,
A significant number of heterokaryon incom- two ascospore pairs survive and two pairs then
patibility loci have now been cloned and identified degenerate. Since the determinant of ascospore
(see Glass et al. 2000 for a review). Data about the death segregates 1:1 in a Mendelian fashion,
molecular events associated with somatic incom- every ascus must originate from a diploid nucleus
patibility have been reviewed by Esser (2006). which is heterozygous for the factor. For this
to occur in self-fertile homokaryons, one of the
two nuclei entering a zygotic partnership must
E. Meiotic Death, Spore Discharge and Spore become epigenetically marked at each generation
Germination to create a new ‘allele’, or a programmed genomic
rearrangement must occur prior to the onset of
Meiotic apoptosis in fungi was first noted in each meiotic event.
spo11 mutants of Coprinus cinereus (Celerin et al. PCD has not been overtly linked with spore
2000) and later studied in more detail by Lu et al. discharge, but it is evident that in some situations
(2003). White-cap mutants of Coprinus cinereus the death of support cells is essential for the disper-
display three classes of meiotic defect which can sal of others (Moore 2004). In the rusts, terminal
be distinguished cytologically. In all three groups, aeciospores are separated by suspensor cells which
nuclei undergo normal meiotic metaphase I, sacrifice their own future reproductive potential,
but later arrest with condensed chromatin and a phenomenon which is potentially regulated by
other apoptotic phenotypes. In some mutants, a PCD mechanism. Autolysis of fruit bodies of
plasmolysis and cytoplasmic shrinkage can also ‘ink-caps’ is another clear example (Umar and van
be observed in basidia, along with condensed Griensven 1997).
chromatin, dispersed fragmented nuclei and PCD may also be an integral part of the me-
TUNEL-positive intracellular bodies. chanics of spore germination and differentiation.
The triggering of the death response appAmut Development of turgor pressure in appressoria of
Bmutears to be linked to cell cycle controls. Exam- Magnaporthe grisea is reliant upon the death of
ination of Amut Bmut strains kept in continuous the germinating spore. Genomic disparity between
light revealed that death was triggered downstream nuclei in spores can also lead to post-germination
of diplotene-metaphase I, and that the checkpoint mortality in several basidiomycetes, e.g. Heteroba-
sensor was only activated once the meiotic nuclei sidion annosum and Stereum hirsutum (Ramsdale
had entered the metaphase I (pseudo-G2/M) check- and Rayner 1996; Ramsdale 1998), which has the
point. In sporulation mutants, death was triggered potential to weed out poor genetic combinations in
at the tetrad stage, probably at the G1/S phase tran- natural populations.
sition. This situation contrasts strikingly with that
in animals, where apoptosis arising from defects
in meiosis usually occurs at the time of the defect, F. Development and Tissue Formation
rather than at a particular checkpoint later on in
the cell cycle (Lu et al. 2003). Higher fungi differentiate a number of tissue
In Coprinus cinereus, meiotic arrest could also types during somatic and sexual development.
be detected in wild-type dikaryons, showing that PCD has been implicated in stipe cavitation
118 M. Ramsdale
and gill formation in a number of mushrooms death in yeasts (see sections below). Indeed, it is
(Lu 1991; Umar and van Giensven 1997, 1998). hard to find any lethal mutations which do not
Examination of developing gills in primordia result in the expression of some or all markers
of Coprinus cinereus reveals extensive cellular of apoptotic cells, suggesting that apoptotic cell
degeneration within the gill cavities, but not death may be a feature common to all ‘slow-killing’
within the gill domains (Lu 1991), which is asso- responses in yeast.
ciated with the activity of numerous hydrolytic Apoptotic markers which have been observed
enzymes. Degenerating hyphae contain multi- in fungi include externalisation of phosphatidylser-
vesicular bodies and residual organelles within ine, enhanced dsDNA accumulation, nuclear con-
the vacuoles. In the final stages of degeneration, densation and fragmentation, cell shrinkage, mem-
cellular debris becomes scattered within the brane blebbing, mitochondrial permeability tran-
intercellular spaces as a result of the lysis of the sition pore opening, release of Cyt c, cleavage of
dead cells. In Agaricus bisporus, morphogenetic caspase substrates, and ROS accumulation (Madeo
cell death is required for the development of gill et al. 2002a).
splits in developing basidiocarps and mycelial
cords. Morphogenetic cell death is associated with
a loss of nuclei, cytoplasmolysis, mitochondrial A. Physical and Environmental Insults
swelling, rupture of the cell membrane and 1. UV Irradiation
dissolution of the hyphal walls. Severely damaged
organelles become enclosed by mucilaginous Del Carratorre et al. (2002) examined the response
material in the inter-hyphal space, along with of S. cerevisiae after exposure to UV light at
clusters of ribosomes, polygonal crystals, and doses of 90–120 J m−2 . Typically, 40%–70% of
glycogen rosettes (Umar and van Griensven 1997). cells were TUNEL-positive. Exposure to higher
Cell death often contributes to the formation doses killed cells, but apoptotic phenotypes were
of cribiforme cavities and gelatinous matrices not detected in these cases. Apoptotic features
which, in some species such as the jelly fungi, observed included nuclear condensation and an
constitute the most significant part of the fruiting accumulation of cells with a sub-G1 DNA content
structure (Moore 1965). Studies performed on (as has been reported in mammalian apoptosis;
a variety of fungi (Stropharia rugoso-annulata, Telford et al. 1992).
Coprinus domesticus, Psathyrella candolleana,
Tremella mesenterica, Otidea onotica and Peziza 2. Sugar
ostracoderma) reveal changes which support the
view that PCD plays a key role in different stages In the absence of nutrients to support growth,
of fungal fruit-body development (Umar and van sugars, including glucose at 2%, can kill yeast
Griensven 1997). cells (Granot et al. 2003). Sugar-induced cell
death is accompanied by nuclear fragmentation,
DNA and RNA degradation, and cytoplasmic
shrinkage. Most cells also stain with PI, showing
III. Stimulating Fungal Apoptosis a necrotic component to the death response after
24-h incubation. Cell death is associated with
The cytological changes associated with mam- the production of ROS and significantly, ROS
malian type I PCD, or apoptosis, have now been scavengers such as ascorbic acid can prevent the
observed in many fungi following a variety of killing. Sugar-induced cell death commonly gives
lethal stimuli. In S. cerevisae, apoptosis can be rise to petite mutants which presumably have
induced by treatment with harsh chemical agents avoided the apoptotic response by virtue of their
or physical insults. A number of physiological inability to produce ROS.
stimuli such as mating pheromone and starvation Soloca et al. (2003) have described killing of
can also induce apoptosis-like changes. Expression yeasts by sugars at very high (60%) concentra-
of heterologous pro-apoptotic regulators from tions which can be accompanied by changes char-
mammals (Bax, Bak) or nematodes (Ced4) can also acteristic of apoptosis – including the activation of
bring about apoptosis-like cell death in a number a caspase-like activity. The relationship of this re-
of yeasts. An increasing number of mutations sponse to that at much lower doses is not clear but
are being identified which result in apoptotic cell it may be osmotic, rather than metabolic in origin.
3. Weak Acids Salt sensitivity was found to be dependent upon

the P-type ATPase (ENA1). A mutant of calcineurin
Weak acids such as acetic, sorbic, pentanoic (a PP2B phosphatase) which regulates Na+ home-
and butyric acid have been shown to kill yeasts, ostasis and mediates salt tolerance (Mendoza et al.
including Zygosaccharomyces bailii and S. cere- 1994) was found to be hypersensitive to NaCl. Cal-
visiae (Fernandes et al. 1997; Ludovico et al. cineurin affects Na+ homeostasis via the transcrip-
2001, 2003). Exposure of S. cerevisiae at pH 3.0 tional activator CRZ1 (which binds to CDREs), and
to acetic acid doses of 20–80 mM produced cells activates a number of genes including ENA1. The
which stained positive by TUNEL, were able to mammalian anti-apoptotic protein Bcl-2 was able
bind annexin and exclude propidium iodide, and to abrogate the increased salt sensitivity of a cnb1
showed nuclear condensation and margination. mutant, but had no effect on the NaCl sensitivity
Cells treated at 120–200 mM displayed hallmark of a hog1 mutant. This suggests that the primary
features of necrosis including PI staining and apoptotic effect is ionic, rather than osmotic.
a complete loss of internal organisation, as Mutants with defects in components of the per-
revealed by transmission electron microscopy. meability transition pore complex (vdac1, vdac2,
Similar observations were made in Z. baillii, vdac3 and aac1, aac2 or aac3) and electron trans-
albeit at higher doses of acetic acid (consistent port chain (atp4) did not affect salt sensitivity. It
with its reputation as an acid-tolerant species). might therefore be concluded that Cyt c release
In Z. bailii, an extensive reorganisation of mi- is not important for the initiation of salt-induced
tochondria was observed which had not been yeast apoptosis, since this requires an intact perme-
seen in S. cerevisiae. FACs analysis, using low ability transition pore. Alternatively, such results
levels of rhodamine 123, revealed a reduction may indicate that salt triggers apoptosis through
in ∆Ψm (membrane depolarization), and EM both Cyt c-dependent and Cyt c-independent path-
studies indicated alterations in the number and ways. Without further data, it will not be possible
structure of mitochondrial cristae, the presence to distinguish between these possibilities.
of myelinic bodies, and some mitochondrial Yeast cells with deletions in the SRO7/SOP1-
swelling (but not lysis). Although a temporal encoded homolog of a Drosophila tumour sup-
analysis was not presented, the proportions of cells pressor gene (l(2)gl) also show enhanced sensi-
displaying apoptotic attributes at different doses tivity to salt stress (Wadskog et al. 2004). Mu-
suggest that chromatin condensation precedes tant sro7 cells exposed to NaCl exhibit a number
phosphatidylserine exposure which precedes or is of markers of apoptosis including ROS accumula-
coincident with dsDNA breakage. tion, DNA breakage, and nuclear fragmentation. In
In S. cerevisiae, death at low (pro-apoptotic) yeast, SRO7 is a multicopy suppressor of rho3 and is
doses, but not at high (pro-necrotic) doses, of acetic cold-sensitive; when combined with a mutation in
acid can be inhibited by cycloheximide, indicat- sni2, it has an exocytic defect, accumulating post-
ing an active aspect to the death response. Fur- Golgi vesicles. Other salt-sensitive mutants includ-
thermore, the failure of cycloheximide to prevent ing gpd1 and ckb1 were not associated with apop-
necrosis in S. cerevisiae indicates that cells can di- totic changes, showing that the effects are specific
rectly enter a necrotic state. This contrasts with the to the sro7-sensitive response. According to Wad-
situation in Candida albicans where necrotic cells skog (2003, cited in Wadskog et al. 2004), the salt
treated with acetic acid are typically derived sec- sensitivity of sro7 is due to the defective targeting
ondarily from prior apoptotic events (Phillips and of the ENA1 ATPase to the plasma membrane.
Ramsdale, unpublished data). Deletion of YCA1 in sro7 and wild-type cells
prevented the apoptotic changes and protected the
4. Salt cells against the primary rapid loss of viability,
although the cells were still salt-sensitive. By
Salt treatment (1.5 M NaCl) causes a time- contrast, sro77 null strains exhibited salt-induced
dependent reduction in viability which is death which was independent of YCA1 but which
accompanied by hallmarks of an apoptotic PCD was still associated with the development of an
response (Huh et al. 2002). Changes are apparent apoptotic phenotype. This supports the idea of
within 40 min of the onset of the treatment, and at least two independent apoptosis pathways in
include nuclear fragmentation, TUNEL staining, yeast. YCA1 activity increased in the presence of
vacuolization and eventually cell lysis. salt in wild-type cells, and was increased further in
120 M. Ramsdale
an sro7 backgound. A direct interaction between in pheromone-treated cells, and FACS analysis the
sro77 and the yeast metacaspase can be inferred accumulation of cells with a sub-G1 DNA content;
from a number of two-hybrid interaction studies phloxin B staining confirmed that the cells were
(Uetz et al. 2000; Drees et al. 2001). However, the dying.
finding that the caspase activity levels were not Deletion of STE20, the mating factor receptor,
enhanced in the sro77 strain does not necessarily abolished pheromone killing of a-type cells and
support the authors’ conclusion that sro7 and sro77 the accumulation of ROS (see section below on
exert antagonistic effects on caspase activation. signalling). Generation of Š− cells using ethidium
bromide abolished pheromone killing, too, indicat-
5. Hydrogen Peroxide ing a requirement for mitochondrial DNA. Ethid-
ium bromide treatment did not, however, prevent
Hydrogen peroxide is a constant threat to micro- schmoo formation, showing that although the cells
organsims, which have developed a number of could still respond to pheromone, they were specif-
antioxidant resistance mechanisms to counteract ically blocked in the PCD response.
its effects. Despite this, exposure of S. cerevisiae, C. Chloroquine treatment of mixed a- and α-type
albicans and A. fumigatus to low doses of hydrogen cells (which prevented cell fusion) induced high
peroxide induces killing with all of the classical levels of killing which could be eliminated by the
attributes of mammalian apoptosis (Madeo et al. addition of cyclosporin A. Thus, mating-induced
1999; Phillips et al. 2003; Mousavi and Robson cell agglutination, in the absence of cell fusion, may
2004). At higher doses, cells die with features lead to PCD under normal conditions. Overall, as
more reminiscent of necrosis. Annexin and Severin and Hyman proposed, the altruistic sui-
TUNEL-positive cells are observed, along with the cide of yeast cells unable to mate after a prolonged
margination of chromatin at the nuclear envelope period of contact might be beneficial for the com-
and/or the fragmentation of nuclei. In S. cerevisiae, munity as a whole.
nuclear condensation is visible after only 30-min Yeasts lacking Cyt c were still able to produce
treatment (Madeo et al. 1999). Some cells show ROS in response to pheromone, but were unable
plasma-membrane vesicles which are similar to to complete their PCD response. Confusingly, in-
the ‘blebs’ linked with apoptosis in mammalian hibition of permeability transition pore formation
cells. A strain lacking GSH1 (therefore, depleted with cyclosporin A was able to prevent both ROS
in glutathione) exhibits a similar death response, generation and cell death. Since cyclosporin A also
which can be exacerbated by the addition of low inhibits the calcineurin/calmodulin signalling
doses of hydrogen peroxide or abrogated by the pathway, CMD1 mutants were also tested. In
exogenous application of glutathione. a cmd1-6 mutant, ROS production and PCD
Co-incubation of cells with cycloheximide was accelerated. Since yeast mutants defective
almost eliminates the TUNEL straining of cells in calmodulin/calcineurin are hypersensitive to
treated with 3 mM H2 O2 and largely prevents pheromone (Cyert and Thorner 1992; Moser et al.
cell death, as revealed by a clonogenic assay, 1996), it seems likely that this system normally
indicating that apoptotic death requires the active inhibits PCD.
participation of the cell machinery.
B. Antifungal Agents and Cytotoxic Drugs
6. Mating Pheromone
1. Amphotericin B
In yeasts, cells of α-mating type produce alpha-
factor which triggers a-type cells to mate. It has The primary mode of action of the antifungal drug
been known for some time that prolonged expo- amphotericin B is well established. Amphotericin
sure to high doses of mating pheromone is toxic B binds to sterols, such as ergosterol, disrupting
(Kurjan 1993). Severin and Hyman (2002) set out the osmotic integrity of the fungal membrane and
to assess if the process could be genetically pro- resulting in the leakage of intracellular ions and
grammed. The addition of alpha-factor pheromone metabolites. Liao et al. (1999) examined physio-
to a-type cells induces the production of ROS in logical changes in C. albicans cells treated with
about 30% of cells after 1.5-h exposure. However, amphotericin B, and described three patterns of
no ROS are produced when alpha-factor is added to death. Death was always accompanied by a drop
α-cells. TUNEL analysis revealed dsDNA damage in ATP level but could be subdivided on the basis
of plasma-membrane integrity and mitochondrial response showed that ROS production preceded
membrane potential. Later, Phillips et al. (2003) dsDNA breaks which, in turn, preceded nuclear
found that nonviable cells which were able to ex- fragmentation.
clude propidium iodide and produce ROS after am-
photericin B treatment corresponded to an apop- 4. Sphingosines and Ceramides
totic subpopulation of cells. Normal treatment of
systemic candidiasis may therefore reduce infec- Phytosphingosine and dihydrospingosine treat-
tion loads by initiating apoptosis of this pathogen. ment of Aspergillus nidulans is followed by the
Protoplasts of A. fumigatus treated with production of ROS, which is in turn followed
0.25 µg ml−1 amphotericin B result in about 15% by nuclear condensation and apoptosis (Cheng
cells staining positive for annexin V, rising to 50% et al. 2003). Pulse field gel electrophoresis also
when treated with 0.5–1 µg ml−1 (Mousavi and revealed extensive DNA degradation and TUNEL-
Robson 2004). TUNEL labelling also increased to positive cells. Exposure to phytosphingosine
65% at 0.5 µg ml−1 amphotericin B, but was lower or dihydrosphingosine for only a few minutes
(at around 20%) when treatment was 1 µg ml−1 . PI was sufficient in this respect, showing that cells
staining (indicative of necrosis) was less than 20% were committed very early on to undergo a PCD
at 0.25 and 0.5 µg ml−1 amphotericin B, but in- response.
creased to 85% at 1 µg ml−1 . Pre-incubation of cells Whilst phytosphingosine-induced nuclear
with cycloheximide prevented the appearance of condensation was not inhibited by ROS scav-
the apoptotic markers, but was not able to prevent engers and was also independent of the fungal
the formation of PI-positive staining at 1 µg ml−1 . metacaspase casA, overexpression of casA on
Taken together, this suggests that low fungicidal its own was sufficient to induce apoptosis. The
doses of amphotericin B are pro-apoptotic and nuclear condensation component of the apoptotic
higher doses are pro-necrotic. response was independent of both condensin (as
revealed by phytosphingosine-induced apoptosis
2. Pradimicin A in a smcBcut14 mutant) and of mitosis per se (since
it still occurred in temperature-sensitive nimA
Pradimicin A, a mannose-binding antifungal an- mutants). Nuclear fragmentation, but not conden-
tibiotic, induced apoptosis-like cell death in S. cere- sation, must require protein synthesis, given that it
visiae (Hiramoto et al. 2003). Nuclear fragmenta- can be inhibited with cycloheximide. Oligomycin,
tion and DNA damage can be observed in yeast which inhibits mitochondrial function, abolished
cells by DAPI and TUNEL staining. Accumulation ROS production and also abolished nuclear
of reactive oxygen species (ROS) was also detected condensation, suggesting that mitochondria play
in dying cells. Moreover, pradimicin-induced cell an essential role in phytosphingosine-mediated
death and the accumulation of ROS could be pre- apoptosis.
vented with a free-radical scavenger, suggesting
some dependency between the two. 5. Translation Inhibitors
3. Osmotin Phenanthroline and other metal ‘phen’ com-

plexes kill both mammalian and C. albicans
The pathogenesis-related protein osmotin in- cells (Coyle et al. 2004). Mammalian cells show
duces apoptosis-like cell death in S. cerevisiae hallmarks of apoptosis, whilst C. albicans cells
(Narasimhan et al. 2001). FACS analysis of DNA accumulate ROS and show elevated oxygen
content revealed a high level of cells with a sub-G1 consumption rates. Nuclear disruption (en-
DNA content in osmotin-treated cells (surpris- largement and crescent-shaped bodies) was
ingly, some sub-G1 contents were also seen in observed after treatment of yeast cells with
untreated cells, albeit at a considerably lower [Ag2 (phen)3 (mal)]. 2H2 O, [Mn2 (phen)3 (mal)].
level). A number of abnormal mitochondrial mor- 2H2 O and [Cu2 (phen)3 (mal)]. 2H2 O, and cy-
phologies were associated with osmotin-induced toplasmic shrinkage after [Cu2 (phen)3 (mal)].
death, but these appear to be consequences, rather 2H2 O treatment. [Ag2 (phen)3 (mal)].2H2 O, [Mn2
than causes of PCD, since oligomycin treatment, (phen)3 (mal)]. 2H2 O, but not [Cu2 (phen)3 (mal)].
ATP4 deletion or Bcl-2 expression have no effect 2H2 O, resulted in a non-specific form of DNA
on killing. A temporal analysis of the death damage. Whilst it may be that such differences
122 M. Ramsdale
Table. 7.1. Genes and proteins with potential link to endogenous apoptotic cell death responses in fungi
Gene or protein Organisma Description Death response References

AIF1/YNR074c Sc Putative apoptosis inducing Mitochondrial protein Wissing et al.
factor (Apaf) with nuclease activity (2004)
which translocates to nucleus
during apoptosis
ASF1/CIA1 Sc Nucleosome assembly factor, Defect induces apoptosis Yamaki et al.
histone chaperone (2001)
ATP10 Sc Mitochondrial inner mebrane Enhanced expression during Yamaki et al.
protein required apoptosis, and required (2001), Ludovico
for F1 F0 ATP synthetase for apoptotic death et al. (2002)
BCY1 Sc Regulatory subunit of PKA Deletion enhances osmotin- Narasimhan et al.
induced apoptosis (2001)
casA An Aspergillus nidulans metacaspase Overexpression accelerates Cheng et al.
apoptosis; deletion inhibits (2003)
apoptosis
CDC13 Sc Single stranded DNA-binding Inactivation induces Qi et al.
protein required caspase activity (2003)
for telomere replication
CDC48 Sc ER-located ATPase required Defect induces apoptosis Madeo et al.
for retrotranslocation (1997)
of ubiquitinated proteins
CDC6 Sc Origin licensing protein Degraded during apoptosis Blanchard et al.
(2002)
CDR1/2 Ca Multidrug resistance transporter; Involved with translocation Smrti et al.
in-to-out floppase of phosphatidylserine (2002)
CMD1 Sc Master regulator of calcium Defect accelerates Severin and Hy-
responses; calmodulin pheromone-induced apoptosis man (2002)
CNB1 Sc Regulatory subunit B Defect accelerates Huh et al.
of calcineurin salt-induced apoptosis (2002)
CRZ1 Sc Stress-linked transcription Defect enhances Edlind et al.
factor regulated by calcineurin sensitivity to antifungals (2002), Huh et al.
and may be involved (2002)
with salt sensitivity
CYC3 Sc Cytochrome c lyase attaches Required for acetic Ludovico et al.
heme to apo-Cyt c acid-induced apoptosis (2002)
CYC8 Sc Transcriptional co-repressor Multicopy suppressor Qi et al.
with Tup1 of cdc13-induced apoptosis (2003)
CYT1/Cyt c Sc Cytochrome c1, mitochondrial Mitochondrion to cytoplasm Yamaki et al.
or cytoplasmic translocation during (2001), Severin
apoptosis, lack of Cyt c and Hyman
blocks pheromone-induced (2002), Ludovico
apoptosis et al. (2002)
DCP1 Sc mRNA processing, decapping Defect induces apoptosis Mazzoni et al.
(2003)
DCP2 Sc mRNA processing, decapping Defect induces apoptosis Mazzoni et al.
(2003)
DGA1 Sp Acyl-CoA:diacylglycerol Mutant accumulates DAG Zhang et al.
acyltransferase which stimulates apoptosis (2003)
ENA1 Sc P-type ATPase required Defect enhances Huh et al.
for salt tolerance salt-induced apoptosis (2002)
GCS1 Ca γ-glutamylcysteine synthetase Defect promotes apoptosis Baek et al.
(2004)
LSM1 Sc mRNA processing, decapping Defect induces apoptosis Mazzoni et al.
and decay (2003)
LSM6 Sc mRNA processing, splicing, Defect induces apoptosis Mazzoni et al.
decapping and decay (2003)
a An, Aspergillus nidulans; Ca, Candida albicans; Cc, Coprinus cinereus; Mr, Mucor racemosus; Sc, Saccharomyces cerevisiae;
Sp, Schizosaccharomyces pombe
Table. 7.1. (continued)

MEC1 Sc Protein kinase required Caspase activation Qi et al.
for DNA damage checkpoints is dependent on Mec1 (2003)
MRas1 Mr Ras-protein; GTP-binding Required for Roze and Linz
protein with homology lovastatin-induced apoptosis (1998)
to mammalian Ras-oncoproteins
MRE11 Sc Component of MRX exonuclease Antagonizes Mec1-dependent Qi et al.
required for double-strand apoptosis (2003)
DNA repair
NMA111 Sc Omi/Htr2A-like serine protease Defect induces apoptosis Fahrenkrog et al.
(YNL123W) (2004)
ORC2 Sp Origin licensing protein Defect induces apoptosis Burhans et al.
(2003)
PARP An Poly-ADP-ribose polymerase Cleaved during PCD Thrane et al.
(2004)
PCA1 Sp S. pombe metacaspase Not required for DAG-mediated Zhang et al.
apoptosis (2003)
PKB/PI3K Sc Protein kinase B signalling Inhibitors of PKB/PI3K Jeon et al (2002)
block apoptosis
PLH1 Sp Phospholipid diacylglycerol Mutant accumulates DAG Zhang et al.
acyltransferase which stimulates apoptosis (2003)
RAD9 Sp DNA damage checkpoint protein Partially suppresses apoptosis Burhans et al.
induced by defect in orc2-1, (2003)
similarity to mammalian
Bcl-2-family proteins
RAS1 Ca Ras-protein; GTP-binding Required for acetic Ramsdale
protein with homology acid-induced apoptosis (unpublished
to mammalian Ras-oncoproteins data)
RAS2 Sc Ras-protein; GTP-binding Required for osmotin-induced Narasimhan et al.
protein with homology apoptosis (2001)
to mammalian Ras-oncoproteins
Rho0 /Š0 Sc Strain lacking mitochondrial DNA Suppresses acetic acid- and Ludovico et al.
cdc13-mediated apoptosis (2002)
RSM23 Sc Mitochondrial ribosomal Null mutation blocks Madeo et al.
protein of small subunit, Yca1-induced apoptosis (2002b)
DAP-3 homolog
SCH9 Sc Serine/thronine kinase Defect blocks aging-dependent Fabrizio et al.
apoptosis (2004)
Spo11 Cc Whitecap mutant Defect produces apoptotic Celerin et al.
basidiospores (2000)
SRO7/SOP1 Sc Yeast homolog of Drosophila Interacts with Yca1; defect Wadskog et al.
tumour suppressor; enhances salt-induced apoptosis (2004)
Suppressor of rho3 defect in a Yca1-dependent fashion
SRO77 Sc Yeast homolog of Drosophila Interacts with Yca1; defect Wadskog et al.
tumour suppressor; enhances salt-induced apoptosis (2004)
Suppressor of rho3 defect in a Yca1-independent fashion
STM1 Sc DNA repair (quadruplex Accumulation promotes apoptosis, Ligr et al.
nucleic acid-binding) protein, deletion reduces H2 O2 -induced (2001)
acts with cdc13 apoptosis
UBP10 Sc Ubiquitin-specific protease Defect induces apoptosis Orlandi et al.
(2004)
WWM1 Sc WW domain containing Growth defect of mutant Szallies et al.
protein involved suppressed by Yca1; (2002)
with desiccation response physical interaction with Yca1
YCA1/MCA1 Sc Putative cysteine protease – Deletion inhibits apoptosis; Madeo et al.
yeast metacaspase overexpression enhances (2002b), Wadskog
H2 O2 -induced apoptosis et al. (2004)
124 M. Ramsdale

YDR333c Sc Unknown function Multicopy suppressor Qi et al.
of cdc13-induced apoptosis (2003)
YME1 Sc AAA-protease located in Activated in response Manon et al.
mitochondrial inner membrane to drop in Cyt c oxidase (1997)
required for degradation
of misfolded proteins
manifest themselves as a result of differences Whilst the results of Wunder et al. (2004) can
in the timing of these events, the production of be taken to suggest that the primary mode of action
ROS in the absence of DNA damage (seen after of histatins is not the direct release of cytochrome c
[Cu2 (phen)3 (mal)].2H2 O treatment) could imply from mitochondria, their failure to detect PCD at
that ROS play a primary role as signals of apoptosis all might be simply due to the fact that they did
but do not act directly as DNA damaging agents. not look for classical markers of apoptosis in intact
Killing by these translational inhibitors may cells. Isolated mitochondria might not be expected
involve a reduction in the levels of cytochrome b to respond to apoptotic stimuli in a natural manner
and c and the associated uncoupling of respiratory when removed from the cellular environment.
function. Such uncoupling may then contribute Several other antifungal drugs also appear to
to the formation of pro-apoptotic ROS. McCann operate independently of apoptotic mechanisms.
et al. (2000) found that phenanthrolines lower re- Treatment of protoplasts of A. fumigatus with itra-
duced:oxidized glutathione ratios, consistent with conazole or of A. nidulans with aureobasidin A,
a pro-oxidant role for the effects of these drugs. amphotericin B and itraconazole does not appear
to induce apoptosis (Cheng et al. 2003; Mousavi
and Robson 2004). Furthermore, fluconazole toxi-
6. Histatins city has been shown to be independent of known
apoptotic mechanisms in yeast, though this was
Histatins are short peptides found in the oral cavity tested only in the context of a failure of Bcl-2 to in-
which have been shown to display both fungistatic hibit killing, and the paper addressed only growth
and fungicidal activity against a range of Candida inhibition and did not look at death characteristics
species. Several workers have independently sug- per se (Kontoyiannis and Murray 2003).
gested that histatins might induce PCD (Helmer-
horst et al. 2001; Phillips et al. 2003), not least
because of the delay after treatment starts before C. Lethal Mutations
killing is observed. However, Wunder et al. (2004)
concluded that histatins did not induce apoptosis in In the sections which follow, an overview is given of
C. albicans, since they could not find any evidence the current state of our knowledge of the molecu-
of DNA laddering or the release of cytochrome c lar events underlying programmed cell death re-
when isolated mitochondria were treated with his- sponses in fungi. A summary of the significant
tatin 5. Although elevated levels of ROS were de- genes which have been implicated in fungal pro-
tected following the exposure of cells to histatin 5 or grammed cell death responses is provided in Ta-
the intracellular expression of an Hst 5 construct, ble 7.1.
the reduction in viability of SOD1/2 mutants of S.
cerevisiae or SOD1 mutants of C. albicans was the 1. CDC48
same as that of wild-type strains. It was therefore
concluded that ROS production merely accompa- CDC48 encodes a protein of the AAA family which
nies the death response, but does not contribute has been implicated in homotypic membrane
significantly to its fungicidal activity. fusion and in a complex with Npl4 and Ufd1. It par-
ticipates in the retro-translocation of ubiquitinated damage bypass and replication fork stability srs2,
proteins from the ER into the cytosol for targeted rqh1 showed very little ROS production.
degradation by the proteasome (Elkabetz et al. Apoptotic phenotypes can be detected in mu-
2004). Expression of a temperature-sensitive mutant yeast strains bearing conditional defects in
tant cdc48S565G at the non-permissive temperature the DNA replication initiation complex. When an
leads to phosphatidylserine exposure, margina- orc2-1 strain is shifted to a non-permissive tem-
tion of chromatin, DNA fragmentation, and the perature, cells produce condensed chromatin and
formation of cell fragments (Madeo et al. 1997). fragmented nuclei, and accumulate ROS (Watan-
The mammalian orthologue of CDC48, abe et al. 2002; Burhans et al. 2003). Significantly,
p97/VCP (Shirogane et al. 1999), and the C. elegans the defect can be partially suppressed by deletion
homologue, Mac-1 (Wu et al. 1999), have now of RAD9 (a yeast homologue of BCL-2-family pro-
also been implicated in apoptosis, making CDC48 teins) or by mutation of MEC1, which indicates that
the first apoptotic regulator identified in yeast the killing is dependent on functional DNA damage
(Frohlich and Madeo 2001). checkpoints.
Burhans et al. (2003) proposed that the orc2-1
2. Cell Division Cycle and DNA Repair Mutants defect induces cell death when the number of repli-
cation forks drops below a threshold required to
Cdc13 is a multifunctional protein involved activate cell cycle checkpoints and/or rescue stalled
with telomere replication and protection and forks. Loss of functional CDC6 can result in an un-
which, when defective, activates the DNA damage foreseen disassembly of pre-replication complexes,
checkpoint. Inactivation of CDC13 has also been and has been implicated in both yeast and mam-
reported to induce metacaspase activity, ROS malian apoptosis (Weinberger et al. 1999; Blan-
production and phosphatidylserine exposure in chard et al. 2002).
yeast (Qi et al. 2003). Elimination of both Chk1 and Cds1 in S.
Multicopy suppressors of these responses were pombe was found to be required to generate ROS
found when a temperature-sensitive cdc13-1 mu- in the presence of hydroxyurea (HU). Both kinases
tant strain was transformed with a yeast genomic activate pathways which converge upon Cdc2,
library and a mammalian HeLa cDNA library. Two a cyclin-dependent kinase. A mutant version of
clones were identified in the yeast screen – a frag- Cdc2 (cdc2Y15F) which is constitutively active
ment of YDR333C and an antisense clone of CYC8. displays extensive ROS production (even in the
Screening of the HeLa cell cDNA library revealed absence of HU), suggesting that the deregulation
a third clone, MTCO3. All three of these genes have of cdc2 and associated premature entry into M
a role to play in mitochondrial functioning, impli- phase is the major source of ROS. In mammals,
cating the mitochondrion in the death response – deregulation of cdk1 (the homologue of S. pombe
a conclusion further supported by the discovery Cdc2) is also a strong apoptotic signal (Castedo
that cdc13-1-driven changes are suppressed in a Š0 et al. 2002).
strain. CDC13-initiated apoptotic signals were also The relationship between apoptosis and DNA
found to be dependent upon the ATM homolog replication is clearly complex. It is apparent that
MEC1 but not TEL1, and could be enhanced in some apoptosis pathways require checkpoint pro-
an mre11 null strain. MRE11 is part of the RMX teins, others are inhibited by them, and yet others
complex which exhibits exonuclease activity. The are independent. The genetic tractability of yeast
enhanced death of mre11 might therefore be due to will help in the elucidation of these pathways and
a failure to activate the appropriate DNA damage their mechanisms.
repair function of the complex.
Induction of apoptosis-like cell death in the 3. ASF1/CIA1
yeasts S. cerevisiae and S. pombe also requires other
functional cell cycle checkpoint pathways (Burhans Disruption of the histone chaperone ASF1/CIA1 in
et al. 2003). In a screen of ROS production by check- S. cerevisiae produces cells with a slow-growth phe-
point mutants, the highest levels were detected notype which is associated with G2/M-phase cell
amongst alleles affecting proteins required for the cycle arrest, and an accumulation of enlarged dead
initiation of replication (e.g. orc2, orp2 orp5, cdc18 cells which contain fragmented nuclei with con-
and dfp1). In general, mutants defective in dsDNA densed chromatin (Yamaki et al. 2001). Using lu-
break repair such as pku70, lig4, rad32 and rad50 or cifer yellow and quinacrine mustard as markers, the
126 M. Ramsdale
acidity of the vacuoles was found to decrease whilst A. Ras and cAMP
that of the cytoplasm increased. Membrane poten-
tials of both vacuoles and mitochondria were found Both pro- and anti-apoptotic functions for Ras have
to be lower in the mutant than in the wild-type. been identified in fungi, mirroring its complexity
Studies on spheroplasts and isolated mitochondria in the determination of cell proliferation, differen-
revealed that death was linked to the release of Cyt c. tiation and apoptosis in animal models.
Microarray analysis revealed that several Lovastatin, an indirect inhibitor of Ras preny-
genes encoding components of the vacuolar lation, can induce an apoptosis-like cell death re-
proton pump (VMA7, VMA8, VMA10, VMA21 and sponse in hyphae but not yeasts of Mucor race-
VMA22) were down-regulated in comparison to mosus (Roze and Linz 1998). Lovastatin treatment
controls, and that expression of the mitochondrial alters the processing of MRas1, blocks the accu-
proton pump genes ATP1 and ATP5 was enhanced. mulation of MRas3 protein, and causes the loss of
The human homolog of ASF1/CIA1 has been an MRas1/p20 protein complex. Exogenous appli-
shown to interact with hCCG1, a key component cation of dibutyryl cAMP (which can initiate mor-
of TFIID which is involved in the regulation of phogenesis from hyphal to yeast-like growth) pre-
mammalian apoptosis (Sekiguchi et al. 1995). vents the lovastatin-induced cell death response,
Holstege et al. (1998) reported that disruption of further implicating the Ras pathway. Some cross-
yeast CCG1 influences VMA gene expression in talk between the Ras and phosphoinositide 3-OH
a similar manner to that seen in ASF1 nulls, and so kinase signalling pathways can be suspected, how-
the link between them certainly deserves further ever, because the addition of wortmannin in com-
attention. bination with dibutyryl cAMP produces a synergis-
tic response, completely blocking cell death (even
4. mRNA Processing Mutants though wortmannin alone has no effect).
In S. cerevisiae, ROS production, the expres-
Saccharomyces cerevisiae cells expressing a trun- sion of antioxidant proteins and the apoptotic re-
cated Lsm4 protein from Kluyveromyces lactis and sponse which follows treatment with osmotin, is
strains with deletions of LSM1, LSM6, DCP1 or partially dependent upon an induced suppression
DCP2 display phenotypic markers of apoptosis in- of the RAS2/cAMP pathway. RAS2G19V , a dominant
cluding chromatin condensation, DNA fragmenta- active allele of RAS2, increased sensitivity to os-
tion, and ROS accumulation (Mazzoni et al. 2003). motin whereas a null mutant decreased sensitivity.
All of these strains have defects in aspects of pre- The response was linked specifically to the Ras-
mRNA splicing, mRNA capping and RNA degra- protein kinase A, rather than to the Ras-MAP ki-
dation. Whilst only a slight induction of apopto- nase pathway, because the effects of the dominant
sis was observed for DHH1 and PAT1, the overall active allele were also seen in a ste20 background.
behaviour of these mutants is indicative of a role Consistent with osmotin-induced Ras-protein ki-
for normal mRNA decapping in the prevention of nase A signalling, a bcy1 null, with a constitutively
apoptosis. It was proposed therefore that stabiliza- active protein kinase A activity, significantly in-
tion of mRNAs could lead to the abnormal accu- creased sensitivity to osmotin.
mulation of pro-apoptotic proteins which, in turn, De-repression of STRE-dependent transcrip-
trigger the apoptotic response. tional responses in ras2 mutants (Stanhill et al.
1999) could account for the elevated resistance
of RAS2 nulls to stress treatments, including
IV. Death-Related Signalling Pathways osmotin. During osmotin-induced PCD, both
STRE-element and YRE-reporter constructs
Understanding of fungal PCD responses will re- are repressed (Narasimhan et al. 2001), and so
quire the elucidation of the signal pathways which it might be argued that the balance between
transmit information about the status of the cell stress and apoptotic signals determines cell fate.
and the environment to the suicide machinery. Although osmotin-induced PCD in wild-type
Both pharmacological approaches and functional cells is associated with a suppression of STRE
genetic analyses have revealed some of the com- signalling, this situation cannot be generalized to
ponents of the death-related signal transduction other PCD responses because, at similar levels of
pathways, although there is still a great deal of re- apoptosis-like cell death, hydrogen peroxide still
search to be done. stimulates STRE-reporter expression (Narasimhan
et al. 2001). It may be concluded, however, that C. CDC55, TPD3 (Protein Phosphatase 2A)
osmotin stimulates pro-apoptotic ROS pro-
duction via the activation of the RAS2/cAMP E4orf4 is a viral regulator protein which down-
pathway which, in turn, inhibits YRE (Yap1- regulates the expression of genes activated by
dependent) and STRE-mediated antioxidant stress cAMP, affects alternative splicing of adenoviral
responses. mRNAs, and induces p53-independent apoptosis
The Ras pathway has been strongly linked to in transformed mammalian cells. Kornitzer et al.
morphogenetic signals in a number of fungi (D’- (2001) were able to show that E4orf4 induced an
Souza and Heitman 2001). One common theme irreversible growth arrest at G2/M in S. cerevisiae,
is the finding of an interrelationship between cell and that this is associated with the production of
death and fungal morphogenesis/differentiation. ROS. Genetic screens revealed that the killing activ-
As highlighted by Narasimhan et al. (2001), overex- ity of E4orf4 required the yeast protein phosphatase
pression of plant defence molecules induces hyper- 2A components CDC55 and TPD3. E4orf4 forms
branching or formation of spiral hyphae (Epple a complex with protein phosphatase 2A which,
et al. 1997), and growth inhibition per se can also be in turn, interacts with the anaphase-promoting
linked to hyphal branching (Garcia-Olmedo et al. complex/cyclosome, leading to its inactivation and
1998; Ali and Reddy 2000). A study of the link concomitant cell cycle arrest/apoptosis.
between death and morphogenesis in C. albicans
showed that it could not be attributed to any of D. Calcineurin and Calcium
the known signalling pathways (EFG1, RIM101,
TEC1, CPH1) which contribute to morphogenesis Several lines of evidence suggest that Ca2+ /calmo-
(Phillips et al. 2003). RAS1 deletion in C. albicans is dulin/calcineurin signals affect the fungal death
able to decelerate, or even prevent PCD induced by response. As discussed above, apoptosis induced by
acetic acid (Phillips et al. 2003; Phillips and Rams- pheromone and salt was shown to be influenced by
dale, unpublished data), just as it is in an S. cere- mutations in calmodulin or calcineurin (Huh et al.
visiae ras2 strain. By contrast, ste12 knockouts in S. 2002; Severin and Hyman 2002). Furthermore,
cerevisiae are defective in their apoptotic response, azole activity against S. cerevisiae can be reduced
but disruption of the homologous downstream ef- by the addition of Ca2+ and enhanced by the
fectors of the Ras pathway in C. albicans (CPH1 or addition of EGTA (Edlind et al. 2002). Inhibitors
EFG1) has no effect. of the Ca2+ -binding regulatory protein calmod-
ulin (fluphenazine, calmidazolin and W-7), and
inhibitors of a Ca2+ calmodulin-regulated phos-
B. Phosphoinositide 3-OH Kinase and Protein phatase, calcineurin (cyclosporin and FK506),
Kinase B Signalling also enhance azole activity. Conversely, mutations
constitutively activating calcineurin demonstrate
The phosphoinositide 3-OH kinase and protein ki- reduced azole susceptibility. Disruption of CRZ1
nase B signalling pathways have been shown to (a transcription factor regulated by calcineurin)
protect a variety of cell types against apoptosis also enhances azole sensitivity – indicating that
(Mathieu et al. 2001; von Gise et al. 2001). Jeon the cell integrity pathway is important for the
et al. (2002) observed that ROS-induced apopto- action of these drugs.
sis in yeast could be stimulated by the addition of Sanglard et al. (2003) reported that FK506
wortmannin or LY294002, or inhibited by the ad- treatment of fluconazole-treated cells induced
dition of PIP2 . Since protein kinase B activity de- a fungicidal, rather than a fungistatic response,
clines during ROS-induced cell death, and protein which could have very important ramifications for
kinase B is activated by phosphorylation, it may future drug therapy regimens. Deletion of CYP1
be that phosphoinositide 3-OH kinase itself effects (cyclophilin) prevented the fungicidal activity,
the primary response. In mammals, phosphoinosi- showing that this component of the signalling
tide 3-OH kinase inhibits caspase-3 (Tessier et al. pathway was essential for fluconzaole toxicity.
2001), delaying the onset of p53-mediated apop- Although the mode of death of the cells has not yet
tosis. Further work needs to be undertaken to ex- been ascertained, it might be speculated that the
amine whether such an interaction is responsible inhibition of the calcium–calmodulin–calcineurin
for the regulation of apoptosis by phosphoinositide signalling pathway could induce apoptosis in the
3-OH kinase/protein kinase B in yeasts. face of external stresses.
128 M. Ramsdale
E. Sphingolipids, Ceramides function in the cell-wall integrity pathway of yeast.

and Protein Kinase C Phosphorylated long-chain base sphingolipids
accumulate in S. cerevisiae ysr2 dpl1 nulls (Kim
The multiallelic HetC locus, which is involved with et al. 2000), and the effect is lethal. Lethality is
vegetative incompatibility in Podospora anserina lost when LCB4 (a long-chain base kinase) is also
(Saupe et al. 1994), is a member of a family of glycol- deleted, showing that the effect is specific to the
ipid transfer proteins. One member of this family, phosphorylated sphingoid bases. The dying cells
ACD1 from Arabidopsis thaliana, has been found show many features of apoptosis (S. Kim, personal
to induce an accelerated cell death phenotype (akin communication).
to apoptosis) during the plant hypersensitivity re- Mutation of the two genes required for the ter-
sponse (Brodersen et al. 2002). ACD1 is involved minal step of triacylglycerol bisosynthesis (DGA1
specifically with the translocation of sphingosine or PLH1) in S. pombe results in an accumulation
(but not glycosylsphingolipids) across the plasma of diacylglycerol which is associated with a loss of
membrane, and may therefore have a direct role viability upon entry into stationary phase (Zhang
in death signalling. Whilst there is no obvious ho- et al. 2003). Death is accompanied by the produc-
molog of this gene in yeast, other lipid translocation of ROS, nuclear fragmentation, dsDNA break-
tors have been identified, such as CDR1 and CDR2, age and phosphatidylserine exposure. Exposure of
which could play a role in lipid-associated death wild-type cells to diacylglycerol is sufficient to in-
signalling. Indeed, Smrti et al. (2002) have been duce apoptosis and expression of a bacterial dia-
able to show that CDR1/2/3 ABC transporters in cylglycerol kinase prevents apoptosis, suggesting
C. albicans can flip phosphatidylserine across the that the effect is specific to the build-up of diacyl-
plasma membrane which is, of course, a conserved glycerol. Construction of a caspase (pca1) null, or
element in all apoptotic death responses. treatment of the cells with a caspase inhibitor, failed
Antifungal defensins produced by plants to block the diacylglycerol-induced death response.
induce a number of changes in fungal membranes, Finding the molecular targets of the death-related
including an increased efflux of K+ ions, increased lipid secondary messengers is clearly an important
Ca2+ uptake, membrane potential changes, and challenge in unravelling their pro-apoptotic func-
membrane permeabilization (Thevissen et al. tions.
1996, 1999). Using a genetic complementation
approach, IPT1 was identified as a gene affecting
sensitivity towards the defensin DmAMP1 in S. F. MAP Kinase Cascades
cerevisiae (Thevissen et al. 2000), whilst GCS was Osmotin activates the mating, invasive growth and
found to reduce sensitivity to RsAFP2 in Pichia pseudohyphal growth MAP kinase cascades of S.
pastoris (Thomma et al. 2002). IPT1 encodes an cerevisiae independently of both the pheromone
enzyme required for the biosynthesis of the sphin- receptor and the associated G protein coupled al-
golipid mannosyldiinositolphosphorylceramide, pha subunit encoded by GPA1 (Yun et al. 1998).
M(IP)2 C (Dickson et al. 1997), whilst GCS encodes The pathway is activated very rapidly; as seen by
a glucosyltransferase required for the biogenesis of the phosphorylation of Ste7, and precedes any of
glucosylceramides. Screening of N. crassa mutants the phenotypic changes associated with PCD. Since
by Ferket et al. (2003) also revealed several mutants exposure to osmotin at lethal or sub-lethal doses is
with altered sphingolipid compositions which were unable to stimulate schmoo formation or to activate
resistant to defensins. Whether such effects can be a FUS1-lacZ reporter, osmotin signalling through
linked to the well-described role of sphingolipids the MAP kinase cascade must target a set of death-
as signals of death, or if the effects were structural, related genes distinct from those involved with
affecting binding of the defensin to the fungal mating or morphogenesis.
cells, is not clear.
The addition of D-erythro-sphingosine to
lovastatin-treated M. racemosus cells accelerated
apoptosis-like cell death (Roze and Linz 1998) – V. Commitment, Effectors
a point not made by the authors. D-erythro- and Suppressors
sphingosine is an inhibitor of protein kinase C,
and so there could be a role for this pathway Identifying the commitment points and effectors
in fungal cell death, particularly in view of its of the programmed death response in fungi re-
mains the key to unlocking its therapeutic poten- ing cell death and functions in part to inhibit STRE-
tial. Whilst this review aims to bring together what mediated stress responses. Further array studies
we do know, our knowledge in the field is strongly will need to be performed under a wider range
biased by hypotheses driven by our understand- of conditions to test the validity of this important
ing of animal PCD. Hence, we now have candidate proposition.
caspases and BCL-2-family proteins. In the future, ROS production is commonly linked to the
we may do well to look for more fungal-specific apoptotic response in animals. Whilst Bcl-2 inter-
pathways. acts with Bax to negate its pro-apoptotic functions,
Bcl-2 may also have a direct antioxidant role to play.
In the short-term, it has been found that Bcl-2 ex-
A. Reactive Oxygen Species pression prevents the death of both sod1 and sod1
sod2 yeast mutants entering into stationary phase
ROS production and the activation of antioxidant (Longo et al. 1997). Such a result is consistent with
defence systems is thought to be involved with an antioxidant role for Bcl-2, acting on basic ele-
a number of differentiation events in fungi, in- ments which must be present in all eukaryotes. Sur-
cluding germination, yeast–hypha transitions and prisingly, in contrast to all other mutations known
conidiation (Hansberg and Aguirre 1990; Hansberg to suppress the growth defects of sod1- and sod1
et al. 1993; Toledo et al. 1995). Significantly, the ma- sod2-deleted strains, Bcl-2 does not have any ef-
jority of the studies outlined in this review have im- fect on the accumulation of Mn2+ or Cu2+ ions.
plicated ROS in the development of a fungal apop- Expression of Bcl-2 is, however, accompanied by
totic phenotype. an increase in catalase activity and an increase in
The fact that intracellular ROS levels increase GSH/GSSG ratios. Both changes suggest a direct
during apoptosis even in the absence of an exter- role for Bcl-2 in promoting antioxidant defence
nal oxidative stress has been used to support a di- mechanisms.
rect role for ROS in the generation of the apoptotic The nature of the active ROS themselves is
phenotype (Madeo et al. 1999; Frohlich and Madeo not certain. Deletion of TSA1, a thiol peroxidase,
2000). ROS production has been seen in every study increases the sensitivity of yeasts to osmotin
which has looked for it, and typically it is an early (Narasimham et al. 2001). Since TSA1 is believed
event preceding the appearance of other apoptotic to specifically detoxify hydrogen peroxide (Lee
markers. ROS must play more than a secondary et al. 1999), it may be surmised that this is one of
role in death since, in the majority of cases, block- the intracellular mediators of the death response.
ing ROS accumulation prevents the appearance of Priault et al. (2002) have also reported that whereas
apoptotic markers and enhancing ROS production a modulation of global ROS did not significantly
stimulates apoptosis. affect Bax-induced killing, the direct modulation
The interaction between stress responses and of lipid oxidation had a strong effect. Superoxide
apoptotic outcomes is particularly important to anions are thought to be a major component of the
consider in this context. Increased levels of ROS active ROS signal in animals (Cai and Jones 1998),
could originate from de novo sources such as an but this has yet to be demonstrated in fungi.
uncoupled electron transport chain or the induc- Currently, we do not know enough to as-
tion of pro-oxidant enzymes such as laccase. Al- certain whether ROS only operate as signals for
ternatively, they could result from a tempering of downstream apoptotic effectors and/or if ROS
oxidative stress responses. Whilst there is consid- can act directly as effectors. Many studies have
erable evidence in favour of the former hypothesis shown that ROS are damaging to DNA and lipids
(e.g. Longo et al. 1996), little has been done to in- (Halliwell and Gutteridge 1989), and so it might
vestigate whether the stress response can be inac- be hypothesized that aspects of PCD, such as
tivated during PCD. Contreras et al. (2002) found the accumulation of dsDNA breaks, could be
in a microarray study of Bax expression in yeast directly attributable to this action. However, the
that at early time points, death is characterized by production of ROS in the absence of measurable
a repression of the major components of the an- DNA damage has been seen after treatment of
tioxidant defence systems, including CTA1, DDR48, C. albicans with [Cu2(phen)3(mal)].2H2 O (Coyle
GRE2, GRX1, HSP12, TRX1 and TTR1. This finding et al. 2004), which could be taken to imply that ROS
fits in well with the earlier report that the Ras– play a primary signalling role during apoptosis
cAMP–protein kinase A pathway is stimulated dur- and do not act directly as DNA damaging effectors.
130 M. Ramsdale
Two cases have been described in which ROS do membrane potential, dysfunction of mitochondrial
not play a central role in the decision to die. The first ATPases, and the release of Cyt c. Bax expression
involves the induction of apoptosis in S. cerevisiae also leads to cytochrome c release in dying yeast
with aspirin (Balzan et al. 2004), and the second cells (Manon et al. 1997; Kluck et al. 1999).
is the induction of apoptosis by phytosphingosine Ludovico et al. (2002) studied Cyt c release in
in A. nidulans (Cheng et al. 2003). It appears that stationary-phase cells (when mitochondria have
growth inhibition by aspirin cannot be reversed by accumulated) and observed that pro-apoptotic
free-radical scavengers such as N-acetyl-cysteine doses of acetic acid decreased the amount of Cyt c
or vitamin E. ROS production is detectable, but it in mitochondria two-to threefold, and that this
only occurs very late on, once apoptotic markers was offset by an increase in cytosolic Cyt c. Other
have already appeared. In this situation it would mitochondrial components such as Cox II and
appear that in the early stages of death, aspirin V could not be detected in the cytosol, arguing
is itself acting as an antioxidant, and that cells be- against a non-specific lytic leakage of proteins.
come committed to die in a ROS-independent fash- Staining of mitochondria with Mitotracker Red
ion. At a later stage, a secondary accumulation of CM-H2 -XRos revealed the production of ROS in
ROS occurs which swamps the intrinsic antioxidant both the mitochondria and cytoplasm. Polari-
properties of aspirin. During phytosphingosine- graphic measurements of oxygen consumption in
induced apoptosis of A. nidulans, ROS production isolated mitochondria revealed a 75% decrease in
precedes nuclear condensation, but the inhibition the activity of NADH oxidase. Levels of a number
of ROS does not prevent cell death. Since death was of respiratory components in the mitochon-
also independent of the metacaspase casA, it ap- drion were also reduced compared to untreated
pears that a second apoptotic pathway may exist controls, suggesting significant changes in the
in fungi which is also independent of ROS produc- organisation of the mitochondrial respiratory
tion. chain. Oligomycin did not inhibit yeast apoptosis,
The fact that ROS production lies at the core arguing that oxidative phosphorylation per se
of most apoptotic responses may be attributable to is not important for yeast apoptosis, but ATP10,
the finding that in yeast, ROS, as a by-product of Š0 and cyc3 mutant cells did not undergo acetic
respiratory activities, are under the control of an ul- acid-induced apoptosis, implicating mitochondria
tradian clock (Lloyd et al. 2003). Senescence, which in the death response.
is associated with apoptosis in yeast, is therefore ex- Linked to Cyt c release is a reduction in the
pected to ensue after a finite number of cycles of level of cytochrome c oxidase which facilitates the
the ‘ticking’ of the ultradian clock which regulates uncoupling of the respiratory chain. Manon et al.
ROS production. Since the clock is driven by in- (1997) linked the decrease in levels of Cyt c oxidase
herent oscillations of central metabolic processes to the activation of Yme1, an AAA-family protease,
such as glycolysis and the electron transport chain, which may be responsible for degrading compo-
it will be a feature of all organisms with oxidative nents such as COX II. Under aerobic conditions, in
metabolism. a yme1-deleted strain, Bax-induced lethality was
only slightly suppressed, indicating that the reduc-
tion in Cyt c oxidase in itself did not contribute
B. Cytochrome c and Mitochondrial Function significantly to the death response (Manon et al.
2001). Since ATP10 deletion, but not oligomycin
The intrinsic apoptosis pathway involves release treatment, was able to inhibit yeast apoptosis, it
of Cyt c from the mitochondrion, which leads to may be surmised that a functional F0 F1 ATPase
the formation of an apoptogenic complex (the mi- complex is required for the release of Cyt c, but that
tochondrial apoptosome) which is able to activate there is no absolute requirement for mitochondrial
caspases in the cytosol. When Cyt c is released from ATP production per se.
the inner mitochondrial membrane, normal elec- Bax expression stimulates release of Cyt c from
tron flow in complex III of the respiratory chain is mitochondria and decreases cytochrome oxidase
interrupted, leading to the generation of superox- (Manon et al. 1997). However, Cyt c release per se
ide radicals (Cai and Jones 1998). was not essential in this case because a Cyt c–GFP
Yamaki et al. (2001), studying yeast apoptosis construct, which remains stuck in the mitochon-
in an ASF1/CIA1 mutant, found that cell death was drial membrane, does not prevent Bax-induced
accompanied by a decrease in the mitochondrial killing (Roucou et al. 2000).
Whilst release of Cyt c has been demonstrated, the outer mitochondrial membrane (Pavlov et al.
its ability to act as a pro-apoptotic signal remains 2001).
less certain. Yeast Cyt c, in contrast to Cyt c from
many other organisms, has been found to be in-
effective in inducing apoptotic changes in a mam- C. Caspases
malian nuclear apoptosis assay (Kluck et al. 1997).
The permeability transition pore is a complex Using nested, iterative PSI-BLAST searches, Uren
upon which many apoptotic signals converge. et al. (2000) identified a family of caspases,
It is composed of a voltage-dependent anion metacaspases and paracaspases in plants, animals
channel (VDAC) and a peripheral benzodiazepine and fungi. In fungi, several metacaspases have
receptor in the outer mitochondrial membrane. subsequently been found. S. cerevisiae and S.
Several reports also suggest that Bax itself is pombe have a single member, YCA1/MCA1 and
also a part of the outer membrane complex PCA1 respectively. C. albicans also has a single
(Marzo et al. 1998; Shimizu et al. 1999). The inner homolog (Ramsdale, unpublished data) whilst A.
mitochondrial membrane components contain fumigatus, A. nidulans and N. crassa may have two
an adenine nucleotide translocator (ANT-1) and each (Cheng et al. 2003; Mousavi and Robson 2004;
the chaperone cyclophilin D. Disintegration of the Thrane et al. 2004).
outer mitochondrial membrane and the release Madeo et al. (2002b) was able to show that
of pro-apoptotic factors such as Smac/Diablo and disruption of YCA1 abrogated hydrogen peroxide-
Cyt c is considered the point of no return in the induced apoptosis and that Yca1 was cleaved in
implementation of a PCD response. a manner typical of mammalian caspases. Yca1
Whilst a number of early studies implicated the also displayed a caspase-like proteolytic activity
ADP/ATP carrier (AAC1, AAC2, AAC3) proteins which was stimulated during apoptosis. Overex-
and a requirement for respiratory function in pression of YCA1 increased caspase-like activity
Bax-induced cell death, Kissova et al. (2000) found and increased the susceptibility of yeast to hydro-
that the cytotoxic effects of Bax expression in gen peroxide-induced killing.
yeast do not require functional oxidative phos- Yca1 is a 52-kDa protein with a 12-kDa cleav-
phorylation, respiration or ADP/ATP carriers. age product which may correspond to the small
Such a major discrepancy could be attributable catalytic subunit found in true caspases. Alanine
to the repression of the GAL promoter of Bax substitution of cysteine (A297C) prevents the
constructs in cells with dysfunctional mito- formation of the small cleavage product and de-
chondria (Kissova et al. 2000). The view that tectable caspase activity. Overexpression of YCA1
Bax effects are independent of mitochondrial and subsequent exposure to hydrogen peroxide
functions is, however, further supported by the revealed a caspase-like activity which was able to
studies made with Bax under the regulation of cleave the fluorescent substrates VEID-AMC and
a tetracycline regulatable promoter (Priault et al. IETD-AMC but not DEVD-AMC. Significantly,
1999a). this proteolytic activity could be abolished with
Using GFP-tagged mitochondrial proteins, Bax the pan-caspase inhibitor zVAD-fmk. Such a sub-
was shown to interfere with the mitochondrial pro- strate preference is indicative of animal initiator
tein import pathway which is specific for the procaspases, such as caspase-8.
teins of the mitochondrial carrier family (Kissova Kang et al. (1999) showed that human cas-
et al. 2000). Bax becomes localised at the outer mi- pase-8, when expressed in yeast, could be activated
tochondrial membrane (Priault et al. 1999a; Pavlov and that it was the only heterologous caspase with
et al. 2001) and induces cell death which is accom- cytotoxic properties. Pre-processing of the procas-
panied by the release of Cyt c (Manon et al. 1997). pase per se was not associated with a significant
This does not require any transition permeability of increase in enzyme activity, suggesting perhaps
the inner mitochondrial membrane, or the VDAC a requirement for an additional trans-activating
proteins, Por1 and Por2 (Priault et al. 1999b; Gross factor. The search for a factor such as Apaf-1 is on-
et al. 2000; Harris et al. 2000; Polcic and Forte 2003) going. However, several proteins have been shown
or the adenine nucleotide carriers (Priault et al. to interact with Yca1, including Sro77 and Wwm1
1999a; Kissova et al. 2000). (Szallies et al. 2002; Wadskog et al. 2004). WWM1
Bax is believed to act upon mitochondria by was found in an overexpression screen to be a very
creating a giant channel, independent of VDAC, in potent inhibitor of cell proliferation, causing cells
132 M. Ramsdale
to arrest in G1 (Stevenson et al. 2001). Significantly, 81-kDa fungal poly-ADP ribose polymerase does
WWM1 overexpression phenotypes are suppressed not contain classical caspase-3 (DEVD) or caspase-
by YCA1 (Szallies et al. 2002). 8 (IETD) cleavage sites, but does contain a KVVDK
In mammals, a mitochondrial mediator site at a location which would give the 60-kDa frag-
of apoptosis, DAP-3, which may act upstream ment observed when it is exposed to extracts with
(Miyazaki and Reed 2001) or downstream (Kissil high endogenous fungal caspase activity.
et al. 1999) of caspase-8, has been identified.
Berger et al. (2000) found that DAP-3 was highly
conserved across many different groups, leading D. Nucleases
Madeo et al. (2002b) to show that the yeast
DAP-3 homolog (RSM23 – a component of the DNA fragmentation, with laddering, is most clearly
mitochondrial small subunit ribosome complex) associated with the PCD response of Mucor racemo-
could completely prevent induction of apoptosis sus (Roze and Linz 1998). Incubation of lovastatin-
by Yca1 overexpression, placing it downstream of treated cells in YPG at pH 4.5 prevented the acti-
caspase activation. vation of the DNA fragmentation response whilst
Characterization of NMA111, an HtrA2-like incubation in a variety of buffers and media at pH
protein, in S. cerevisiae (Fahrenkrog et al. 2004) 7.45 was permissive. It may therefore be that the
reveals that it shares a similar pro-apoptotic endonuclease is pH-dependent, with an optimal
function to that of its mammalian homolog. Yeasts neutral pH. Addition of EDTA, but not EGTA, to
cells lacking Nma111 are able to survive better treated cells was also inhibitory for DNA fragmen-
than wild-type at 50 ◦ C and following exposure tation, perhaps indicating a requirement for Ca2+
to pro-apoptotic doses of hydrogen peroxide. and Mg2+ ions. DNA fragmentation could not be
Furthermore, the nma111 mutant also fails to inhibited by cycloheximide, suggesting that the en-
exhibit hallmark features of apoptosis such as zyme is constitutively produced.
chromatin condensation, nuclear fragmentation DNA laddering per se has not so far been de-
or ROS accumulation. Critically, overexpression of tected in yeasts undergoing apoptosis. It is now
Nma111 can directly stimulate apoptosis-like cell widely recognized that laddering is not a reliable
death. Under stress conditions, Nma111 aggre- feature of apoptosis in mammalian cells (e.g. Ober-
gates in the nucleus and its pro-apoptotic function hammer et al. 1993), where it is often restricted
in yeast is dependent upon its serine-protease to specific cell lines. Madeo et al. (2002a) have
activity. proposed that a lack of laddering may be due to
In mammals, HtrA2 stimulates cell suicide by basic differences in the architecture of yeast and
reducing the levels of XIAP, an inhibitor of apop- mammalian chromatin (Lowary and Widom 1989).
tosis protein (IAP). An IAP-like protein, Bir1, has However, MNase treatment of isolated yeast chro-
also been described in yeast (Uren et al. 1999; Yoon matin does give a strong laddering effect (Kunoh
and Carbon 1999; Li et al. 2000). Bir1 contains et al. 2000), indicating that the architecture may
a baculovirus inhibitor repeat which is conserved not be too dissimilar. Differences in the apoptotic
amongst IAP proteins and is required for the inter- programme itself may therefore be more crucial.
action of the proteins with caspases. This protein During the sugar-induced apoptotic response
has not so far been demonstrated to have a link of yeast, fragments of DNA about 30 kb in size
to yeast apoptosis, but has been implicated in the can be detected by pulse field gel electrophoresis
regulation of cell division and arrest. (Granot et al. 2003). In S. pombe (Ink et al.
The first fungal homologs of poly-ADP ribose 1997), Bak expression has been shown to induce
polymerase, a downstream target of effector cas- chromatin degradation with an accumulation of
pases, have recently been discovered in A. nidu- higher-order structures around 800 kb in size. In
lans and N. crassa (Thrane et al. 2004). During both cases, these fragments may be equivalent to
conidiation of A. nidulans, a time-dependent loss the DFF40 hypersensitive sites in HeLa cells or
of poly-ADP ribose polymerase can be observed, the 30–50 kb DNA fragments which have often
which coincides with the programmed autolysis of been reported in apoptotic animal cells (Brown
the underlying mycelium. A caspase-3 inhibitor de- et al. 1993; Oberhammer et al. 1993) prior to the
layed, but did not abolish the degradation of poly- formation of oligomeric ladders.
ADP ribose polymerase by fungal extracts, with en- Although relatively little information is avail-
dogenously high levels of caspase-like activity. The able concerning RNA degradation during apop-
tosis, specific cleavage of 28S, but not 18S, rRNA stability). The signature stress response accounts
has been observed in several mammalian cell lines for a significant part of this, involving about
(Houge et al. 1995; Houge and Doskeland 1996). 10%–15% of all genes.
Whilst Granot et al. (2003) observed equal amounts Genes induced during the ESR are involved in
of degradation in S. cerevisiae, dying C. albicans glycolysis, antioxidant defence, protein folding and
treated with either acetic acid or hydrogen perox- proteolysis, and vacuolar or mitochondrial func-
ide show a preferential degradation of 28S rRNA tions. Repressed genes are responsible for many
(Ramsdale, personal observation) which is accen- aspects of energy consumption and growth-related
tuated during necrosis. RNA processing activities processes – most notably, ribosome biogenesis. The
are commonly found to be amongst the most sig- response is transient and is proportional to the in-
nificantly altered biological functions in proteome tensity of the stress (Gasch and Werner-Washburne
and array studies of dying mammalian cells (Brock- 2002). When the profiles of S. cerevisiae and S.
stedt et al. 1998; Thiede et al. 2001). Although pombe are compared, a significant degree of over-
there are few comparable studies on fungi, simi- lap is observed, indicating that the core responses
lar changes are seen at the transcriptome level in are highly conserved.
the raw data of a number of studies (Ramsdale, The interrelationships between stress re-
unpublished data; Contreras et al. 2002; Enjalbert sponses and death responses warrant closer inspec-
et al. 2003). tion. In particular, it is important to know the extent
to which responses overlap and to identify the ma-
jor differences. Surprisingly, there are no published
E. Endogenous BCL-2-Family Proteins proteome or transcript profiling studies of stress-
induced PCD in S. cerevisiae or S. pombe. A number
Using rabbit polyclonal antibodies, Bcl-2 and Bax-
of studies have investigated the global responses of
like proteins were detected in Mucor racemosus on
fungi to antifungal treatments (Bammert and Fos-
Western blots, with sizes of 22 and 16–18 kDa re-
tel 2000; de Backer et al. 2001; Zhang et al. 2002a, b;
spectively (though several other proteins of higher
Agarwal et al. 2003) or fungicides (Kitagawa et al.
molecular weight were also detected in each anal-
2003). However, in these cases the agents investi-
ysis, so the evidence is not conclusive; Roze and
gated are either fungistatic or, if they are fungicidal,
Linz 1998). Bioinformatic studies have failed to re-
the array analyses have been performed on RNA
veal convincing homologues of the BCL-2-family
collected from cells which have been treated with
proteins, but Komatsu et al. (2000) have shown
inhibitory, but nonetheless sub-lethal doses (e.g.
that S. pombe Rad9 contains a BH3 domain which
0.5× MIC). Unfortunately, this makes it impossible
is required for its interaction with heterologously
to demarcate death-specific functions.
expressed Bcl-2 proteins. Screens for endogenous
Biella et al. (2002) reported that the response
SpRAD9 interactants in yeasts could be useful in
of Cryphonectria parasitica to infection with
identifying components of an ancestral AIF.
the dsRNA hypovirulence factor resembles PCD.
Using microarrays to look for genes which were
differentially regulated when strains became in-
VI. Global Responses of Dying Cells fected with the virus, 295 unique sequences (out of
2200) were found with changed abundance (Allen
Microarray technologies have allowed an in-depth et al. 2003). Using differential display, Chen et al.
exploration of the responses of fungi to a wide (1996) observed that 65% of the global changes
range of physiological and stress conditions (see initiated by virus infection could be reproduced
Wilhelm and Bähler, Chap. 6, this volume). Several by manipulating G-protein and cAMP signalling
investigators have shown that a common signature pathways – further strong evidence linking the
subset of genes responds to many different stresses; death responses to Ras-protein kinase A activity.
components of the so-called environmental stress Many of the other biological processes identified
response (ESR; Gasch et al. 2000) or common as significantly up-regulated during infection
environmental stress response (Causton et al. (e.g. SAM metabolism, polyamine biosynthesis
2001; Chen et al. 2003). Typically during any given and glutathione metabolism) have also been
stress response, 30%–50% of the RNA encoded detected in an array study of acetic acid-induced
by the genome can alter in abundance (due either apoptosis in C. albicans (Ramsdale, unpublished
to transcriptional regulation or altered mRNA data) Furthermore, such changes have been linked
134 M. Ramsdale
to apoptosis through classical studies of both VII. Fungi as Models for Apoptosis
plants and animals (Kawalleck et al. 1992; Tschopp
et al.1998). The early view that yeast did not undergo
In a pilot study of cell death in C. albicans apoptosis was seized upon by many as an
which has integrated proteomics with tran- opportunity to study mammalian and worm
scriptomics (Ramsdale, unpublished data), cells apoptosis genes in a ‘clean’ genetic environment.
undergoing apoptosis induced by acetic acid We now know that this is not the case, but
have been compared to control cells, stressed nonetheless structure–function relationships and
cells and necrotic cells. Significant numbers of C. genetic screens revealing proteins which interact
albicans proteins and mRNAs appear to change in with components of the plant or animal cell
abundance during the stress and death responses. death machinery have usefully been sought in
Analysis of the proteome reveals that >200 protein the ‘sterile’ environment of the yeast genetics
spots alter specifically in response to stress, 29 laboratory.
spots alter specifically during apoptosis and 50
spots change under necrotic conditions. Further-
more, 29 protein spots can be identified which are A. Genetic Screens
specific to the death response, but are not affected
by stress-inducing conditions. In all, 144 proteins Overexpression of many genes in fungi can pro-
which change in a significant fashion in response duce lethal effects. Often, death is not due to the
to the stress and death stimuli have been identified increased activity of a dedicated pro-death protein
by MALDI-ToF mass spectrometry. Whilst many but rather to an imbalance in some critical process.
of these proteins have no known function, others Nevertheless, such screens could provide a useful
play key roles in cell rescue, core metabolism or de starting point to look for pro-death genes. If the
novo protein synthesis and proteasome-mediated screen is performed under conditions which stim-
degradation. ulate PCD, then antagonists of this process might
A similar picture is emerging from the parallel also be identified with anti-apoptotic properties.
transcript profiling analysis of the same samples. Many of the ‘essential’ genes described in fungi may
C. albicans mRNAs have been identified which re- also have anti-apoptotic roles. It has not yet been
spond specifically to stress (145 transcripts), apop- ascertained in a systematic fashion which of these
tosis (19 transcripts) or necrosis (108 transcripts). induce killing (and, importantly, by what route) and
Many are also observed which change in response which bring about cell cycle arrest. Future studies
to both apoptosis and necrosis, but not stress (109 in this area might be particularly fruitful in reveal-
transcripts). These transcripts encode components ing genes involved with yeast apoptosis.
of the proteasome and ubiquitin-mediated proteol-
ysis pathways, proteases, glycolytic and TCA cycle 1. Overexpression-Induced Lethality
enzymes, ergosterol biosynthetic enzymes and ri-
bosomal proteins as well as a putative C. albicans Lethal outcomes have been observed after over-
metacaspase. One of the most striking findings is expressing genes encoding components of the
that the mRNA abundance of ribosomal proteins cytoskeleton (Rose and Fink 1987; Burke et al.
remains steady during both apoptotic and necrotic 1989; Berlin et al. 1990; Magdolen et al. 1993)
death responses but declines significantly during and signal transduction cascades (Russell and
the stress response, indicating that dying cells are Nurse 1987; Osmani et al. 1988; Whiteway et al.
maintaining their protein biosynthesis machinery. 1990; Millar et al. 1992; Ottilie et al. 1992). In
The data indicate that necrosis, at least in the con- genome-wide library screens using cDNA or
text of acetic acid-induced killing, is probably the genomic clones, lethal effects have been observed
endpoint of prior apoptotic events. for ABP1, ACT1, ARF2, ATE1, AUA1, BIK1, BNI1,
Both proteomics and transcript profiling can BOI1, ERG6, GCL17, HSF1, KAR1, MCM1, NHP6A,
highlight cellular components which might be in- NHP6B, NPS1, NSR1, NTH1, PRK1, PSP1, RBP1,
volved in the regulation or execution of the PCD RHO1, STE4, STE11, STE12, SAC7, SEC17, SIR1,
response. Moreover, such studies emphasize that SNU114, SRP40, TPK1, TPK3, TUB1 and URA2 (Liu
death is an active process, distinct in many ways et al. 1992; Ramer et al. 1992, Espinet et al. 1995;
from the stress response, which requires the full Akada et al. 1997). Intriguingly, BNI1 (Gin2) and
participation of a cell molecular repertoire. BOI1 (Gin7) produce cells with multiple DAPI-
staining bodies – perhaps evidence of nuclear PRE4 strain, but the killing was exacerbated in the
fragmentation and apoptosis (Akada et al. 1997). pre1-1 pre4-1 background. Hence, it was included
This list provides further strong evidence in in the list of potential pro-apoptotic proteasome-
favour of a central role for Ras/protein kinase regulated genes. Using GAL1 promoter shut-off of
A signalling in the death responses of yeast. an STM1:IRS reporter construct, stabilization of
TPK1, TPK3 and NTH1 are obviously associated the Stm1 protein in pre-1 pre4-1 cells was observed,
with Ras-protein kinase A signalling, whilst RPB1 demonstrating that Stm1 is a natural substrate of
expression has been linked to Ras activity (Howard the active proteasome.
et al. 2002). Elevated expression of RAS2 produces Disruption of UBP10, a gene which encodes
defects similar to those of cells expressing mutants a de-ubiquitinating enzyme in S. cerevisiae, re-
of RPB1 with a truncated CTD interacting domain. sults in slow growth and impairment of silencing
Ras2 activation is also synthetically lethal with at telomeres and at HM loci. In addition, mutant
many components of RNA pol II including compo- cells show an accumulation of ROS, DNA fragmen-
nents of the CTD mediator (sin4, gal11, med6, srb4) tation and phosphatidylserine exposure – all hall-
or TFIIH (kin28, ccl1, rig2, tfb1) complexes. Such mark features of apoptosis (Orlandi et al. 2004).
a lethal effect cannot be attributed to a general A genome-wide analysis of a ubp10 null strain re-
defect in translation, since in the sin4 mutant vealed an alteration in the expression of many sub-
background global mRNA accumulation was not telomeric genes, as expected along with transcrip-
affected by Ras2 activation. tional changes indicative of an oxidative stress re-
sponse. Construction of a ubp10 sir4 double null
2. Proteasome-Dependent Lethality strain prevented the large-scale re-modelling of the
transcriptional response and abrogated the apop-
Ligr et al. (2001) developed a screen to search totic phenotypes.
for proteins whose degradation by the ubiquitin- Down-regulated genes identified encoded
proteasome system was required for viability. a variety of transporters (tyrosine, glutamate,
Overexpression of such a protein in a normal isoleucine, ammonium, biotin and peptide carri-
cell should be permissive for growth, but over- ers), ribosomal proteins and translation factors
expression in a strain defective in proteolysis (consistent with many stress-inducing conditions
should be lethal. A screen was performed on – see above discussion), and enzymes involved
a 2 µ-based GAL-regulated cDNA library in with the biosynthesis of arginine, lysine, trypto-
a pre1-1 pre4-1 cyh2 mutant strain containing phan and various nucelotides. Up-regulated genes
a PRE1 CYH2 helper plasmid, pML1 (making the included enzymes involved with carbohydrate
strain phenotypically wild-type with respect to transport, metabolism and energy production,
growth, proteasome-dependent proteolysis and heat-shock proteins and a number of transcription
cycloheximide sensitivity). Growth on glucose factors (notably, HAP2, HAP4, ROX1 and YAP4).
and glucose cycloheximde media (which selected Overall, such changes cluster with many other
for loss of pML1), alongside replica plating of the transcriptional profiles which are associated with
original plates containing clones with a pre1-1 oxidative stresses and the general stress response
pre1-4 background on galactose, revealed 125 mediated by Msn2/4.
library clones which caused growth arrest in The array study indicates that correct
pre1-1 pre4-1 mutants but not in a PRE1 pre4-1 de-ubiquitination is required for normal tran-
background. After sequencing, 62 individual ORFs scriptional balances, and that disruption of
were identified as high-expression lethality genes this can lead to apoptosis associated with the
(HEL). GFP-annexin and TUNEL staining, along production of ROS and an associated oxidative
with other apoptosis indicators such as nuclear stress response. In mammals, levels of the tumour
condensation and fragmentation, were observed suppressor gene p53 are in part controlled by
in six of these clones when overexpressed in the de-ubiquitination activities of the cell (de-
a pre1-1 pre4-1 background (NSR1, PPA1, SAR1, ubiquitination stabilizes p53). Altered patterns
STM1, YNL208w-HEL10 and YOR309c-HEL13). of ubiquitination of pro- or anti-apoptotic pro-
ROS production could not be detected in any of the teins in yeast, in particular those involved with
clones tested. Significantly, TUNEL and annexin chromatin modelling, might therefore account for
staining were not detected in any of the other 62 the death phenotypes observed when UBP10 is
clones. NSR1 caused significant killing in the PRE1 disrupted.
136 M. Ramsdale
3. Bax Suppressors Kuwana et al. (2002) have suggested that car-

diolipin is required for Bax-mediated pore forma-
The pro-apoptotic members of the BCL-2 family of tion in liposome models, but Iverson et al. (2004)
proteins, Bax and Bak, have been shown to kill yeast found that Cyt c release mediated by Bax in yeast
cells, and cell death is typically associated with an was not affected in a crd1 (cardiolipin-deficient)
apoptotic phenotype (Ink et al. 1997; Jurgensmeier strain. Spontaneous release of Cyt c could be de-
et al. 1997; Ligr et al. 1998). Not all fungi respond tected from mitochondria isolated from the crd1
in the same way to these pro-apoptotic proteins. deficient strain (release was not detectable in WT
Expression of murine Bax in Pichia pastoris, for mitochondria), which argues against Cyt c release
example, leads to growth arrest accompanied by the being sufficient to induce apoptosis, as crd1 mu-
condensation of chromatin, and the accumulation tants have normal growth. However, since Cyt c
of autophagic bodies (Martinet et al. 1999), but no release could be observed only under a specific set
other apoptotic features. of ionic conditions which promoted swelling of the
Yeasts have been used to examine structure– CL deficient mitochondrial preparations, this con-
function relationships of mammalian pro- clusion must be viewed cautiously.
apoptotic proteins. Matsuyama et al. (1998a, 1999) A genetic screen in yeast has been able to link
found, using both mammalian and yeast cells Bax expression to both apoptotic and autophagic
expressing Bax, that it has two mechanisms of death. Camougrand et al. (2003) obtained a yeast
killing: putative pore-forming domain-dependent mutant which was resistant to Bax-induced cell
cell death (in both yeast and mammals) and death by virtue of a decrease in the level of UTH1.
dimerization-dependent cell death (only in mam- Null uth1 strains were also found to be resistant
malian cells). Minn et al. (1999) revealed that to Bax which could not be linked to a change in
Bax could induce hyperpolarization of mito- the localisation of Bax, its insertion into the mito-
chondria in yeast, verifying a hypothesis for the chondrial membrane, or the release of Cyt c. UTH1
mode of action of Bax in mammalian cell death disruption confers resistance to rapamycin under
responses. respiratory but not fermentative conditions, link-
Mutagenization of yeast with N-methyl-N - ing the gene and Bax to an autophagic pathway
nitro-N-nitrosoguanidine allowed the identifica- involving mitochondria.
tion of mutants which were resistant to expression A yeast two-hybid screen revealed that Cnx1
of Bax (Matsuyama et al. 1998b). Out of eight (calnexin), an ER integral membrane protein with
clones identified, only one, ATP4, was able to fully a cytosolic C-terminus, interacted with Bak in S.
restore Bax-induced lethality when replaced with pombe. Moreover, expression of a truncated ver-
a wild-type copy. ATP4 (subunit 4 of the F0 F1 - sion of calnexin which lacked the Bak-binding C-
ATPase) is a component of a proton pump located terminus, in a cnx1 knockout strain, could prevent
on the inner membrane of the mitochondrion. Bak-induced cell death (Torgler et al. 1997). This
Pharmacological inhibition of the proton pump study corroborates findings in several assessments
with oligomycin reproduced the effects of deleting of mammalian cells where Bak and other BCL-2-
ATP4 on Bax-induced lethality. Furthermore, by family proteins have been shown to co-localise with
comparing Bax-induced killing in the petites Š− calnexin in the ER.
and ATPδ knockout strains, the authors were
able to conclude that respiration per se was not
4. Caspase-Family Proteins
required for Bax-induced killing, but that an intact
proton pump system was. Mammalian and worm caspases have been ex-
Expression of a soybean cDNA library in pressed in S. cerevisiae and S. pombe. Ced-3
a Bax-expressing yeast strain has been used to is lethal when expressed in S. cerevisiae (Tao
screen for plant genes which prevented Bax- et al. 1999), and lethality can be blocked by
induced killing (Moon et al. 2002). From a total co-expression of ced-9. Ced-3, but not ced-4,
of five separate Bax-inhibiting clones, isolated toxicity was antagonized by co-expression of the
ascorbate peroxidase (sAPX) was studied further. caspase inhibitors CrmA and p35, showing that
The overexpression of sAPX greatly suppressed native apoptotic interactions can be accurately
the generation of Bax-associated ROS which modelled in yeast.
is thought to be responsible for its protective Caspase-3 expression in yeast delays growth
effect. but does not induce killing (Wright et al. 1999), and
expression of active forms of caspases 2,6-9 has no type human p53 has been found to block growth
effect on growth, suggesting that there are no es- of S. pombe and S. cerevisiae (Bischoff et al. 1992;
sential endogenous targets for these proteases in Nigro et al. 1992). In S. pombe, the p53 polypeptide
yeast. In S. pombe, both caspase-1 and caspase-3 localised to the nucleus and became phosphory-
induce growth arrest which is dependent upon lated at both the cdc2 site and the casein kinase II
proteolytic capacity (Ryser et al. 1999). Caspase- site (Bischoff et al. 1992).
mediated death can be rescued by co-expression of In an attempt to identify p53-like proteins
the baculovirus protein p35 but not Bcl-2. Signifi- in yeast, Koerte et al. (1995) isolated a mutant
cantly, the yeast studies revealed that Bcl-2 can be which required wild-type p53 for its viability.
cleaved by caspase-1 and -3. The mutant, rft1-1, was defective in cell cycle
Cascades of caspase activation have been progression and arrested at G1/S when p53
recreated in yeast (Kang et al. 1999). Co-expression was depleted. Genetic and biochemical studies
of caspase-8/-3 resulted in extensive cell lysis and revealed that p53 suppressed the rft1-1 mutation
a punctate DAPI staining of the cytoplasm which by forming a protein–protein complex with the
could be correlated with DNA fragmentation Rft1 protein.
detected on agarose gels. Although oligosomal- A core activity of p53 is an ability to stim-
sized fragments were not detected, a substantial ulate transcription of genes which prevent cells
amount of DNA about 2 kb in size could be with damaged DNA from entering S phase. p53
detected. In addition, RNA was not seen in the was able to strongly stimulate a reporter contain-
cells expressing caspase-8/-3, suggesting some ing multiple repeats of the p53 recognition se-
activation of RNAse activities. Endogenous yeast quence RRRC(A/T)(T/A)GYYY in wild-type yeast,
targets of the expressed caspases must contribute but could only weakly stimulate expression in a trr1
to this terminal phenotype, and so it will be null. The ectopic expression of the thioredoxin re-
of interest to determine their nature. It is often ductase gene, TRR1, could restore p53-dependent
thought that caspase-8 mostly contributes to the reporter gene activity to high levels. Such a study
downstream processing of other caspases, e.g. in yeast led to the proposition that p53 disulphides
caspase-3, but cytotoxicity in yeast could imply in p53 must be reduced in order for the protein
that it has other targets or that it is actually to function as a transcription factor (Pearson and
activating the endogenous yeast metacaspase. Merrill 1998).
In order to identify novel caspases and their Granzyme B and its pro-enzyme form have
regulators, Hawkins et al. (1999) designed a re- both been successfully expressed in the yeast Pichia
porter system for caspase activity in yeast. Co- pastoris (Pham et al. 1998), with no reported ef-
expression of a caspase alongside a transcriptional fects on yeast viability. The constructs described
activator (LexA-B42) fused to a transmembrane were fused in frame with yeast alpha-factor cDNA
protein via a linker which contains multiple (but and were then released into the supernatant by the
distinct) caspase cleavage sites releases the tran- KEX2 signal peptidase, which may explain the ob-
scription factor from its membrane targeted sup- served lack of cytotoxicity.
port, allowing the activation of a lacZ reporter
driven by a lexA UAS. This approach was success-
fully exploited to identify DIAP1, an inhibitor of
the Drosophila caspase DCP-1, although there is no
reason why the screen cannot be used to identify VIII. Evolutionary Considerations
other novel caspases or small molecule inhibitors
of caspases. Regulated cell death has been described in all major
phyla – perhaps with the exception of the archae-
5. Other Non-Yeast Death Proteins bacteria (see introductory references). One over-
riding question yet to be answered by this review
Mutations in the gene encoding p53 have been is why unicellular organisms such as bacteria and
found to be the most common genetic alterations yeasts should have a PCD response at all. Of course,
in human cancer. p53 is thought to exert its tu- it is not possible to look back into the history of
mor suppression functions through an inhibition a yeast such as S. cerevisiae and ask what was the
of cell proliferation and induction of apoptosis in defining event which led to the acquisition of the
response to DNA damage. Overexpression of wild- ancestral death programme. However, we may be
138 M. Ramsdale
sure that it was either a serendipitous event or that responses per se appear to have been sidelined in
its acquisition conferred some adaptive value. favour of pragmatic screens which aim to directly
Taking serendipity as a starting point, it is identify antifungal agents.
clear that some of the key players of the death The failure to focus on death itself has led to
response of ‘higher’ organisms such as mammals the identification of many agents which are merely
have their evolutionary roots in prokaryotic pro- fungistatic, rather than fungicidal, and this has had
teins, e.g. HtrA and Bax. The intrinsic cell suicide a number of important consequences for the suc-
programme in particular may have its evolutionary cess or failure of therapeutic programmes. Mode-
origins at the dawn of the eukaryotic mode of life. of-action studies often appear to follow almost as
We can only imagine the interplay which occurred an after-thought, and these in themselves often do
between ancestral prokaryotes as they formed new not explain why fungal cells die.
alliances but, if the machinery for PCD was already This chapter has examined what little we do
present in prokaryotes, then it may have been re- know about the cell biology of fungal death re-
tained during the subsequent emergence of the eu- sponses, and shows how a combination of cell bi-
karya. This would then help to explain why PCD ological approaches, functional genetic analyses,
has been described in all eukaryotic lineages to genetic screens and global profiling technologies is
date – it is simply built in. Alternatively, PCD in just beginning to unravel this important, but ne-
yeasts may be a hangover from a more recent evo- glected, aspect of fungal growth and development.
lutionary transition. We do not know, for exam- Genomic or proteomic studies have not been widely
ple, if the ancestral relatives of the euascomycetous used to investigate the nature of PCD in fungi to
yeasts were multicellular. If they were, then they date. However, it is hoped that this review will
may have possessed and carried over PCD machin- stimulate such studies, and allow us to highlight
ery for the same reasons which are advocated in both the similarities and differences between the
extant multicellular organisms. events which regulate mammalian PCD and fungal
Since unicellular organisms are often most PCD. I hope it is already apparent that a number
closely associated with their clonal relatives, of discrete endogenous cell suicide pathways exist
explanations for the implementation of a PCD in fungi, which might be usefully exploited in the
response based upon altruism also become plau- search for novel therapies against fungal infections.
sible. Suicide of a single cell which is damaged,
or infected within a clonal colony, can both Acknowledgements. I would like to thank Andrew Phillips,
Jon Crowe and Alistair Brown for their many helpful discus-
release nutrients for use by the colony and prevent sions. Support for this work has been provided by a Lloyd’s
the spread of infectious agents (viruses, selfish Tercentenary Foundation Fellowship, with additional funds
genetic elements) or damaged hereditary material from the Scottish Hospitals Research Trust, the MRC and
throughout a population. Finally, many micro- the BBSRC.
organsims form intimate commensal, mutualistic
and parasitic associations with other organisms
which may require fine control over population References
size. It is not inconceivable, therefore, that PCD can
be used to regulate population density, with the Agarwal AK, Rogers PD, Baerson SR, Jacob MR, Barker KS,
payoff that scarce spatial and nutritional resources Cleary JD, Walker LA, Nagle DG, Clark AM (2003)
are utilized more effectively. Genome-wide expression profiling of the response to
polyene, pyrimidine, azole, and echinocandin antifun-
gal agents in Saccharomyces cerevisiae. J Biol Chem
278:34998–35015
Ainsworth AM, Rayner ADM, Broxholme SJ, Beeching JR
IX. Conclusions (1990) Occurrence of unilateral genetic transfer and
genomic replacement between strains of Stereum hir-
sutum from non-outcrossing and outcrossing popula-
The biology of cell death in fungi is a large, rela- tions. New Phytol 115:119–128
tively unexplored area. This is surprising, given the Akada R, Yamamoto J, Yamashita I (1997) Screening and
obvious need to control fungal infections in plants, identification of yeast sequences that cause growth in-
animals and man. Whilst much effort has been inhibition when overexpressed. Mol Gen Genet 254:267–
274
vested in understanding the growth of fungi, and Ali GS, Reddy ASN (2000) Inhibition of fungal and bac-
studies of their proliferation are at the forefront of terial plant pathogens by synthetic peptides: in vitro
current scientific research, studies of their death growth inhibition, interaction between peptides and
inhibition of disease progression. Mol Plant-Microbe in a human Burkitt lymphoma cell line. J Biol Chem
Interact 13:847–859 273:28057–28064
Allen TD, Dawe AL, Nuss DL (2003) Use of cDNA microar- Brodersen P, Petersen M, Pike HM, Olszak B, Skov S,
rays to monitor transcriptional responses of the chest- Odum N, Jorgensen LB, Brown RE, Mundy J (2002)
nut blight fungus Cryphonectria parasitica to infec- Knockout of Arabidopsis accelerated-cell-death11
tion by virulence-attenuating hypoviruses. Eukaryot encoding a sphingosine transfer protein causes
Cell 2:1253–1265 activation of programmed cell death and defense.
Ameisen JC (1996) The origin of programmed cell death. Genes Dev 16:490–502
Science 272:1278–1279 Brown DG, Sun XM, Cohen GM (1993) Dexamethasone
Anagnostakis SL, Day PR (1979) Hypovirulence conversion induced apoptosis involves cleavage of DNA to large
in Endothia parasitica. Phytopathology 69:1226–1229 fragments prior to internucleosomal fragmentation.
Anglade P, Vyas S, Javoy-Agid F, Herrero MT, Michel PP, J Biol Chem 268:3037–3039
Marquez J, Mouatt-Prigent A, Ruberg M, Hirsch EC, Budovskaya YV, Stephan JS, Reggiori F, Klionsky DJ, Her-
Agid Y (1997) Apoptosis and autophagy in nigral man PK (2004) The Ras/cAMP-dependent protein ki-
neurons of patients with Parkinson’s disease. Histol nase signaling pathway regulates an early step of the
Histopathol 12:25–31 autophagy process in Saccharomyces cerevisiae. J Biol
Aylmore RC, Todd NK (1986) Cytology of non-self hyphal Chem 279:20663–20671
fusions and somatic incompatibility in Phanerochaete Burhans WC, Weinberger M, Marchetti MA, Ramachan-
velutina. J Gen Microbiol 132:581–591 dran L, D’Urso G, Huberman JA (2003) Apoptosis-like
Baek Y-U, Kim Y-R, Yim H-S, Kang S-O (2004) Disrup- yeast cell death in response to DNA damage and repli-
tion of γ-glutamylcysteine synthetase reults in abso- cation defects. Mutat Res 532:227–243
lute glutathione auxotrophy and apoptosis in Candida Burke D, Gasdaska P, Hartwell L (1989) Dominant effects
albicans. FEBS Lett 556:47–52 of tubulin overexpression in Saccharomyces cerevisiae.
Balzan R, Sapienza K, Galea DR, Vassallo N, Frey H, Bannis- Mol Cell Biol 9:1049–1059
ter WH (2004) Aspirin commits yeast cells to apoptosis Cai J, Jones DP (1998) Superoxide in apoptosis. Mitochon-
depending on carbon source. Microbiology 150:109– drial generation triggered by cytochrome c loss. J Biol
115 Chem 273:11401–11404
Bammert GF, Fostel JM (2000) Genome-wide expression Camougrand N, Grelaud-Coq A, Marza E, Priault M,
patterns in Saccharomyces cerevisiae: comparison Bessoule JJ, Manon S (2003) The product of the UTH1
of drug treatments and genetic alterations affect- gene, required for Bax-induced cell death in yeast, is
ing biosynthesis of ergosterol. Antimicrob Agents involved in the response to rapamycin. Mol Microbiol
Chemother 44:1255–1265 47:495–506
Barker MG, Brimage LJ, Smart KA (1999) Effect of Cu, Zn su- Castedo M, Perfettini JL, Roumier T, Kroemer G (2002)
peroxide dismutase disruption mutation on replicative Cyclin-dependent kinase-1: linking apoptosis to cell
senescence in Saccharomyces cerevisiae. FEMS Micro- cycle and mitotic catastrophe. Cell Death Differ
biol Lett 177:199–204 9:1287–1293
Baudat F, Manova K, Yuen JP, Jasin M, Keeney S (2000) Cataldo AM, Barnett JL, Berman SA, Li J, Quarless S,
Chromosome synapsis defects and sexually dimorphic Bursztajn S, Lippa C, Nixon RA (1995) Gene
meiotic progression in mice lacking Spo11. Mol Cell expression and cellular content of cathepsin D
6:989–998 in Alzheimer’s Disease brain: evidence for early
Berger T, Brigl M, Herrmann JM, Vielhauer V, Luckow B, up-regulation of the endosomal-lysosomal system.
Schlondorff D, Kretzler M (2000) The apoptosis me- Neuron 14:671–680
diator mDAP-3 is a novel member of a conserved Causton HC, Ren B, Koh SS, Harbison CT, Kanin E, Jen-
family of mitochondrial proteins. J Cell Sci 113:3603– nings EG, Lee TI, True HL, Lander ES, Young RA (2001)
3612 Remodeling of yeast genome expression in response to
Berlin V, Styles CA, Fink GR (1990) BIK1, a protein required environmental changes. Mol Biol Cell 12:323–337
for microtubule function during mating and mitosis Celerin M, Merino ST, Stone JE, Menzle AM, Zolan ME
in Saccharomyces cerevisiae, colocalizes with tubulin. (2000) Multiple roles of Spo11 in meiotic chromosome
J Cell Biol 111:2573–2586 behavior. EMBO J 19:2739–2750
Biella S, Smith ML, Aist JR, Cortesi P, Milgroom MG (2002) Chen B, Gao S, Choi GH, Nuss DL (1996) Extensive
Programmed cell death correlates with virus trans- alteration of fungal gene transcript accumulation
mission in a filamentous fungus. Proc R Soc Lond B and elevation of G-protein-regulated cAMP levels by
269:2269–2276 a virulence-attenuating hypovirus. Proc Natl Acad Sci
Bischoff JR, Casso D, Beach D (1992) Human p53 inhibits USA 93:7996–8000
growth in Schizosaccharomyces pombe. Mol Cell Biol Chen D, Toone WM, Mata J, Lyne R, Burns G, Kivinen K,
12:1405–1411 Brazma A, Jone N, Bahler J (2003) Global transcrip-
Blanchard F, Rusiniak ME, Sharma K, Sun X, Todorov I, tional responses of fission yeast to environmental
Castellano MM, Gutierrez, Baumann C, Burhans WC stress. Mol Biol Cell 14:214–229
(2002) Targeted destruction of DNA replication protein Cheng J, Park TS, Chio LC, Fischl AS, Ye XS (2003) In-
cdc6 by cell death pathways in mammals and yeast. Mol duction of apoptosis by sphingoid long-chain bases in
Biol Cell 13:1536–1549 Aspergillus nidulans. Mol Cell Biol 23:163–177
Brockstedt E, Rickers A, Kostka S, Laubersheimer A, Chi S, Kitanaka C, Noguchi K, Mochizuki T, Nagashima Y,
Dorken B, Wittmann-Liebold B, Bommert K, Otto A Shirouzu M, Fujita H, Yoshida M, Chen W, Asai A et al.
(1998) Identification of apoptosis-associated proteins (1999) Oncogenic Ras triggers cell suicide through the
140 M. Ramsdale
activation of a caspase-independent cell death pro- Elkabetz Y, Shapira I, Rabinovich E, Bar-Nun S (2004)
gram in human cancer cells. Oncogene 18:2281–2290 Distinct steps in dislocation of luminal endoplasmic
Contreras RH, Eberhardt I, Luyten WHML, Reekman RJ reticulum-associated degradation substrates: roles
(2002) Bax-responsive genes for drug target identifica- of endoplasmic reticulum-bound p97/Cdc48p and
tion in yeast and fungi. Patent International Publica- proteasome. J Biol Chem 279:3980–3989
tion Number WO 02/064766 A2 Engelberg-Kulka H, Glaser G (1999) Addiction modules and
Cooke RC, Rayner ADM (1984) Ecology of saprophytic programmed cell death and antideath in bacterial cul-
fungi. Longman, London, UK tures. Annu Rev Microbiol 53:43–70
Cornillon S, Foa C, Davoust J, Buonavista N, Gross JD, Gol- Enjalbert B, Nantel A, Whiteway M (2003) Stress-induced
stein P (1994) Programmed cell death in Dictyostelium. gene expression in Candida albicans: absence of a gen-
J Cell Sci 107:2691–2704 eral stress response. Mol Biol Cell 14:1460–1467
Cortesi P, McCulloch CE, Song H, Lin H, Milgroom MG Epple P, Apel K, Bohlmann H (1997) Overexpression of an
(2001) Genetic control of horizontal virus transmis- endogenous thionin enhances resistance of Arabidop-
sion in the chestnut blight fungus, Cryphonectria par- sis against Fusarium oxysporum. Plant Cell 9:509–520
asitica. Genetics 159:107–118 Espinet C, de la Torre MA, Aldea M, Herrero E (1995)
Coyle B, Kinsella P, McCann M, Devereux M, O’Connor R, An efficient method to isolate yeast genes caus-
Clynes M, Kavanagh K (2004) Induction of apoptosis ing overexpression-mediated growth arrest. Yeast
in yeast and mammalian cells by exposure to 1,10- 11:25–32
phenanthroline metal complexes. Toxicol In Vitro Esser K (2006) Heterogenic incompatibility. In: Esser K (ed)
18:63–70 The Mycota, vol I, 2nd edn. Kües U, Fischer R (eds)
Cutler NS, Pan X, Heitman J, Cardenas ME (2001) The TOR Growth, differentiation and sexuality. Springer, Berlin
signal transduction cascade controls cellular differen- Heidelberg New York (in press)
tiation in response to nutrients. Mol Biol Cell 12:4103– Esser K, Blaich R (1994) Heterogenic incompatibility in
4113 fungi. In: Wessels JGH, Meinhardt F (eds) The Mycota,
Cyert MS, Thorner J (1992) Regulatory subunit (CNB1 gene vol I. Growth, differentiation and sexuality. Springer,
product) of yeast Ca2+ /calmodulin-dependent phos- Berlin Heidelberg New York, pp 211–228
phoprotein phosphatases is required for adaptation to Fabrizio P, Battistella L, Vardava R, Gattazzo C, Liou L-L,
pheromone. Mol Cell Biol 12:3460–3469 Diaspro A, Dossen JW, Gralla EB, Longo VD (2004)
De Backer MD, Ilyina T, Ma X-J, Vandoninck S, Superoxide is a mediator of an altruistic aging pro-
Luyten WHML, Vanden Bossche H (2001) Genomic gram in Saccharomyces cerevisiae. J Cell Biol 166:1055–
profiling of the response of Candida albicans to 1067
itraconazole treatment using a DNA microarray. Fahrenkrog B, Sauder U, Aebi U (2004) The S. cerevisiae
Antimicrob Agents Chemother 45:1660–1670 HtrA-like protein Nma111p is a nuclear serine pro-
Debets F, Yang X, Griffiths AJF (1994) Vegetative incompat- tease that mediates yeast apoptosis. J Cell Sci 117:115–
ibility in Neurospora: its effect on horizontal transfer 126
of mitochondrial plasmids and senescence in natural Ferket KK, Levery SB, Park C, Cammue BP, Thevissen K
populations. Curr Genet 26:113–119 (2003) Isolation and characterization of Neurospora
Del Carratore R, Della Croce C, Simili M, Taccini E, Scav- crassa mutants resistant to antifungal plant defensins.
uzzo M, Sbrana S (2002) Cell cycle and morphological Fungal Genet Biol 40:176–85
alterations as indicative of apoptosis promoted by UV Fernandes L, Corte-Real M, Loureiro V, Loureiro-Dias MC,
irradiation in S. cerevisiae. Mutat Res 513:183–191 Leao C (1997) Glucose respiration and fermentation
Dementhon K, Paoletti M, Pinan-Lucarre B, Loubradou- in Zygosaccharomyces bailii and Saccharomyces cere-
Bourges N, Sabourin M, Saupe SJ, Clave C (2003) Ra- visiae express different sensitivity patterns to ethanol
pamycin mimics the incompatibility reaction in the and acetic acid. Lett Appl Microbiol 25:249–253
fungus Podospora anserina. Eukaryot Cell 2:238–246 Fraser A, James C (1998) Fermenting debate: do yeast un-
Dickson RC, Nagiec EE, Wells GB, Nagiec MM, Lester RL dergo apoptosis? Trends Cell Biol 8:219–221
(1997) Synthesis of mannose-(inositol-P)2-ceramide, Frohlich KU, Madeo F (2000) Apoptosis in yeast-a mono-
the major sphingolipid in Saccharomyces cerevisiae, cellular organism exhibits altruistic behaviour. FEBS
requires the IPT1 (YDR072c) gene. J Biol Chem Lett 473:6–9
272:29620–29625 Frohlich KU, Madeo F (2001) Apoptosis in yeast: a new
Drees BL, Sundin B, Brazeau E, Caviston JP, Chen GC, Guo W, model for aging research. Exp Gerontol 37:27–31
Kozminski KG, Lau MW, Moskow JJ, Tong A et al. (2001) Garcia-Olmedo F, Molina A, Alamillo JM, Roderiguez-
A protein interaction map for cell polarity develop- Palenzuela P (1998) Plant defense peptides. Biopoly-
ment. J Cell Biol 154:549–571 mers 47:479–491
D’Souza CA, Heitman J (2001) Conserved cAMP signaling Gasch AP, Werner-Washburne M (2002) The genomics of
cascades regulate fungal development and virulence. yeast responses to environmental stress and starvation.
FEMS Microbiol Rev 25:349–364 Funct Integr Genomics 2:181–192
Du C, Fang M, Li Y, Li L, Wang X (2000) Smac, a mitochon- Gasch AP, Spellman PT, Kao CM, Carmel-Harel O, Eisen MB,
drial protein that permits cytochrome c-dependent Storz G, Botstein D, Brown PO (2000) Genomic expres-
caspase activation by eliminating IAP inhibition. Cell sion programs in the response of yeast cells to environ-
102:33–42 mental changes. Mol Biol Cell 11:4241–4257
Edlind T, Smith L, Henry K, Katiyar S, Nickels J (2002) An- Glass NL, Jacobson DJ, Shiu PKT (2000) The genetics of hy-
tifungal activity in Saccharomyces cerevisiae is modu- phal fusion and vegetative incompatibility in filamen-
lated by calcium signalling. Mol Microbiol 46:257–268 tous ascomycete fungi. Annu Rev Genet 34:165–186
Golstein P, Aubry L, Levraud JP (2003) Cell-death alternative grammed cell death in yeast and plants. Plant J
model organisms: why and which? Nat Rev Mol Cell 29:649–659
Biol 4:798–807 Ink B, Zornig M, Baum B, Hajibagheri N, James C, Chitten-
Granot D, Levine A, Dor-Hefetz E (2003) Sugar-induced den T, Evan G (1997) Human Bak induces cell death
apoptosis in yeast cells. FEMS Yeast Res 4:7–13 in Schizosaccharomyces pombe with morphological
Gross A, Pilcher K, Blachly-Dyson E, Basso E, Jockel J, changes similar to those with apoptosis in mammalian
Bassik MC, Korsmeyer SJ, Forte M (2000) Biochemical cells. Mol Cell Biol 17:2468–2474
and genetic analysis of the mitochondrial response of Iverson SL, Enoksson M, Gogvadze V, Ott M, Orrenius S
yeast to BAX and BCL-X(L). Mol Cell Biol 20:3125– (2004) Cardiolipin is not required for Bax-mediated
3136 cytochrome c release from yeast mitochondria. J Biol
Halliwell B, Gutteridge JMC (1989) Free radicals in biology Chem 279:1100–1107
and medicine. Oxford University Press, Oxford Hans- Jacobson DJ, Beurkens K, Klomparens KL (1998) Micro-
berg W, Aguirre J (1990) Hyperoxidant states cause mi- scopic and ultrastructural examination of vegetative
crobial cell differentiation by cell isolation from dioxy- incompatibility in partial diploids heterozygous at het
gen. J Theor Biol 142:201–221 loci in Neurospora crasssa. Fungal Genet Biol 23:45–
Hansberg W, de Groot H, Sies H (1993) Reactive oxy- 56
gen species associated with cell differentiation in Jazwinski SM (1993) The genetics of aging in the yeast Sac-
Neurospora crassa. Free Radic Biol Med 14:287–93 charomyces cerevisiae. Genetica 91:35–51
Harman D (1962) Role of free radicals in mutation, cancer, Jeon BW, Kim KT, Chang SI, Kim HY (2002) Phosphoinosi-
aging and maintenance of life. Radic Res 16:752–763 tide 3-OH kinase/protein kinase B inhibits apoptotic
Harris MH, Vander Heiden MG, Kron SJ, Thompson CB cell death induced by reactive oxygen species in Sac-
(2000) Role of oxidative phosphorylation in Bax toxic- charomyces cerevisiae. J Biochem 131:693–699
ity. Mol Cell Biol 20:3590–3596 Jurgensmeier JM, Krajewski S, Armstrong RC, Wilson GM,
Hartl DL, Dempster ER, Brown SW (1975) Adaptive sig- Oltersdorf T, Fritz LC, Reed JC, Ottilie S (1997) Bax- and
nificance of vegetative incompatibility in Neurospora Bak-induced cell death in the fission yeast Schizosac-
crassa. Genetics 81:553–569 charomyces pombe. Mol Biol Cell 8:325–339
Hawkins CJ, Wang SL, Hay BA (1999) A cloning method to Kamada Y, Funakoshi T, Shintani T, Nagano K, Ohsumi M,
identify caspases and their regulators in yeast: Iden- Ohsumi Y (2000) Tor-mediated induction of au-
tification of Drosophila IAP1 as an inhibitor of the tophagy via an Apg1 protein kinase complex. J Cell
Drosophila caspase DCP-1. Proc Natl Acad Sci USA Biol 150:1507–1513
96:2885–2890 Kang JJ, Schaber MD, Srinivasula SM, Alnemri ES,
Helmerhorst EJ, Troxler RF, Oppenheim FG (2001) The hu- Litwack G, Hall DJ, Bjornsti MA (1999) Cascades
man salivary peptide histatin 5 exerts its antifungal ac- of mammalian caspase activation in the yeast
tivity through the formation of reactive oxygen species Saccharomyces cerevisiae. J Biol Chem 274:3189–
Proc Natl Acad Sci USA 98:14637–14642 3198
Herker E, Jungwirth H, Lehmann KA, Maldener C, Kawalleck P, Plesch G, Hahlbrock K, Somssich IE (1992) In-
Frohlich KU, Wissing S, Buttner S, Fehr M, Sigrist S, duction by fungal elicitor of S-adenosyl-L-methionine
Madeo F (2004) Chronological aging leads to apoptosis synthetase and S-adenosyl-Lhomocysteine hydrolase
in yeast. J Cell Biol 164:501–507 mRNAs in cultured cells and leaves of Petroselinum
Hiramoto F, Nomura N, Furumai T, Oki T, Igarashi Y (2003) crispum. Proc Natl Acad Sci USA 89:4713–4717
Apoptosis-like cell death of Saccharomyces cerevisiae Kim S, Fyrst H, Saba J (2000) Accumulation of phosphory-
induced by a mannose-binding antifungal antibiotic, lated sphingoid long chain bases results in cell growth
pradimicin. J Antibiot (Tokyo) 56:768–772 inhibition in Saccharomyces cerevisiae. Genetics
Hochman A (1997) Programmed cell death in prokaryotes. 156:1519–1529
Crit Rev Microbiol 23:207–214 Kissil JL, Cohen O, Raveh T, Kimchi A (1999) Structure-
Holstege FCP, Jennings EG, Wyrick JJ, Lee TI, Hengart- function analysis of an evolutionary conserved protein,
ner CJ, Green MR, Golub TR, Lander ES, Young RA DAP3, which mediates TNF-alpha- and Fas-induced
et al. (1998) Dissecting the regulatory circuitry of a eu- cell death. EMBO J 18:353–362
karyotic genome. Cell 95:717–728 Kissova I, Polcic P, Kempna P, Zeman I, Sabova L, Kolarov J
Houge G, Doskeland SO (1996) Divergence towards a dead (2000) The cytotoxic action of Bax on yeast cells does
end? Cleavage of the divergent domains of ribosomal not require mitochondrial ADP/ATP carrier but may
RNA in apoptosis. Experientia 52:963–967 be related to its import to the mitochondria. FEBS Lett.
Houge G, Robaye B, Eikhom TS, Golstein J, Mellgren G, 471:113–118
Gjertsen BT, Lanotte M, Doskeland SO (1995) Fine Kitagawa E, Momose Y, Iwahashi H (2003) Correlation of the
mapping of 28S rRNA sites specifically cleaved in cells structure of agricultural fungicides to gene expression
undergoing apoptosis. Mol Cell Biol 15:2051–2062 in Saccharomyces cerevisiae upon exposure to toxic
Howard SC, Budovskaya YV, Chang YW, Herman PK (2002) doses. Environ Sci Technol 37:2788–2793
The C-terminal domain of the largest subunit of RNA Klionsky DJ, Emr SD (2000) Autophagy as a regulated path-
polymerase II is required for stationary phase entry way of cellular degradation. Science 290:1717–1721
and functionally interacts with the Ras/PKA signaling Klionsky DJ, Ohsumi Y (1999) Vacuolar import of proteins
pathway. J Biol Chem 277:19488–19497 and organelles from the cytoplasm. Annu Rev Cell Dev
Huh GH, Damsz B, Matsumoto TK, Reddy MP, Rus AM, Biol 15:1–32
Ibeas JI, Narasimhan ML, Bressan RA, Hasegawa PM Kluck RM, Martin SJ, Hoffman BM, Zhou JS, Green DR,
(2002) Salt causes ion disequilibrium-induced pro- Newmeyer DD (1997) Cytochrome c activation of
142 M. Ramsdale
CPP32-like proteolysis plays a critical role in a Xenopus Ligr M, Madeo F, Frohlich E, Hilt W, Frohlich KU, Wolf DH
cell-free apoptosis system. EMBO J 16:4639–4649 (1998) Mammalian Bax triggers apoptotic changes in
Kluck RM, Esposti MD, Perkins G, Renken C, Kuwana T, yeast. FEBS Lett 438:61–65
Bossy-Wetzel E, Goldberg M, Allen T, Barber MJ, Green Ligr M, Velten I, Frohlich E, Madeo F, Ledig M, Frohlich KU,
DR et al. (1999) The pro-apoptotic proteins, Bid and Wolf DH, Hilt W (2001) The proteasomal substrate
Bax, cause a limited permeabilization of the mitochon- Stm1 participates in apoptosis-like cell death in yeast.
drial outer membrane that is enhanced by cytosol. Mol Biol Cell 12:2422–2432
J Cell Biol 147:809–822 Liu H, Krizek J, Bretscher A (1992) Construction of a GAL1-
Koerte A, Chong T, Li X, Wahane K, Cai M (1995) Suppres- regulated yeast cDNA expression library and its appli-
sion of the yeast mutation rft1-1 by human p53. J Biol cation to the identification of genes whose overexpres-
Chem 270:22556–22564 sion causes lethality in yeast. Genetics 132:665–673
Komatsu K, Hopkins KM, Lieberman HB, Wang H (2000) Liu X, Kim CN, Yang J, Jemmerson R, Wang X (1996) Induc-
Schizosaccharomyces pombe Rad9 contains a BH3-like tion of apoptotic program in cell-free extracts: require-
region and interacts with the anti-apoptotic protein ment for dATP and cytochrome c. Cell 86:147–157
Bcl-2. FEBS Lett 481:122–126 Lloyd D, Lemar KM, Salgado LE, Gould TM, Murray DB
Kontoyiannis DP, Murray PJ (2003) Fluconazole toxicity is (2003) Respiratory oscillations in yeast: mitochondrial
independent of oxidative stress and apoptotic effec- reactive oxygen species, apoptosis and time; a hypoth-
tor mechanisms in Saccharomyces cerevisiae. Mycoses esis. FEMS Yeast Res 3:333–339
46:183–186 Lockshin RA, Zakeri Z (2004) Caspase-independent cell
Kornitzer D, Sharf R, Kleinberger T (2001) Adenovirus death? Oncogene 23:2766–2773
E4orf4 protein induces PP2A-dependent growth ar- Longo VD, Gralla EB, Valentine JS (1996) Superoxide dismu-
rest in Saccharomyces cerevisiae and interacts with the tase activity is essential for stationary phase survival
anaphase-promoting complex/cyclosome. J Cell Biol in Saccharomyces cerevisiae. J Biol Chem 271:12275–
154:331–344 12280
Kunoh T, Sakuno T, Furukawa T, Kaneko Y, Harashima S Longo VD, Ellerby LM, Bredesen DE, Valentine JS, Gralla
(2000) Genetic characterization of rbt mutants that en- EB (1997) Human Bcl-2 reverses survival defects in
hance basal transcription from core promoters in Sac- yeast lacking superoxide dismutase and delays death
charomyces cerevisiae. J Biochem (Tokyo) 128:575–584 of wild-type yeast. J Cell Biol 137:1581–1588
Kurjan J (1993) The pheromone response pathway in Sac- Lowary PT, Widom J (1989) Higher-order structure of Sac-
charomyces cerevisiae. Annu Rev Genet 27:147–179 charomyces cerevisiae chromatin. Proc Natl Acad Sci
Kuwana T, Mackey MR, Perkins G, Ellisman MH, Lat- USA 86:8266–8270
terich M, Schneiter R, Green DR, Newmeyer DD (2002) Lu BC (1991) Cell degeneration and gill remodelling dur-
Bid, Bax, and lipids cooperate to form supramolecular ing basidiocarp development in the fungus Coprinus
openings in the outer mitochondrial membrane. Cell cinereus. Can J Bot 69:1161–1169
111:331–342 Lu BC, Gallo N, Kues U (2003) White-cap mutanats and mei-
Lai CY, Jaruga E, Borghouts C, Jazwinski SM (2002) A mu- otic apoptosis in the basidiomycete Coprinus cinereus.
tation in the ATP2 gene abrogates the age asymmetry Fungal Genet Biol 39:82–93
between mother and daughter cells of the yeast Sac- Ludovico P, Sousa MJ, Silva MT, Leao C, Corte-Real M (2001)
charomyces cerevisiae. Genetics 162:73–87 Saccharomyces cerevisiae commits to a programmed
Lam E (2004) Controlled cell death, plant survival and de- cell death process in response to acetic acid. Microbi-
velopment. Nat Rev Mol Cell Biol 5:305–315 ology 147:2409–2415
Laun P, Pichova A, Madeo F, Fuchs J, Ellinger A, Kohlwein S, Ludovico P, Rodrigues F, Almeida A, Silva MT, Barrien-
Dawes I, Frohlich KU, Breitenbach M (2001) Aged tos A, Corte-Real M (2002) Cytochrome c release and
mother cells of Saccharomyces cerevisiae show mark- mitochondria involvement in programmed cell death
ers of oxidative stress and apoptosis. Mol Microbiol induced by acetic acid in Saccharomyces cerevisiae. Mol
39:1166–1173 Biol Cell 13:2598–2606
Lee J, Spector D, Godon C, Labarre J, Toledano M (1999) Ludovico P, Sansonetty F, Silva MT, Corte-Real M (2003)
A new antioxidant with alkyl hydroperoxide defense Acetic acid induces a programmed cell death process
properties in yeast. J Biol Chem 274:4537–4544 in the food spoilage yeast Zygosaccharomyces bailii.
Leslie JF, Zeller KA (1996) Heterokaryon incompatibility FEMS Yeast Res 3:91–96
in fungi: more than just another way to die. J Genet Madeo F, Frohlich E, Frohlich KU (1997) A yeast mutant
75:415–424 showing diagnostic markers of early and late apoptosis.
Lewis K (2000) Programmed death in bacteria. Microbiol J Cell Biol 139:729–734
Mol Biol Rev 64:503–514 Madeo F, Frohlich E, Ligr M, Grey M, Sigrist SJ, Wolf DH,
Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, Frohlich KU (1999) Oxygen stress: a regulator of apop-
Alnemri ES, Wang X (1997) Cytochrome c and dATP- tosis in yeast. J Cell Biol 145:757–767
dependent formation of Apaf-1/caspase-9 complex ini- Madeo F, Engelhardt S, Herker E, Lehmann N, Maldener C,
tiates an apoptotic protease cascade. Cell 91:479–489 Proksch A, Wissing S, Frohlich KU (2002a) Apoptosis
Li F, Flanary PL, Altieri DC, Dohlman HG (2000) Cell divi- in yeast: a new model system with applications in cell
sion regulation by BIR1, a member of the inhibitor of biology and medicine. Curr Genet 41:208–216
apoptosis family in yeast. J Biol Chem 275:6707–6711 Madeo F, Herker E, Maldener C, Wissing S, Lachelt S, Her-
Liao RS, Rennie RP, Talbot JA (1999) Assessment of the effect lan M, Fehr M, Lauber K, Sigrist SJ, Wesselborg S et al.
of amphotericin B on the vitality of Candida albicans. (2002b) A caspase-related protease regulates apoptosis
Antimicrob Agents Chemother 43:1034–1041 in yeast. Mol Cell 9:911–917
Magdolen V, Drubin DG, Mages G, Bandlow W (1993) High dimerization-dependent and -independent mecha-
levels of profilin suppress the lethality caused by over- nisms. EMBO J 18:632–643
production of actin in yeast cells. FEBS Lett 316:41–47 Miyazaki T, Reed JC (2001) A GTP-binding adapter protein
Manon S, Chaudhuri B, Guerin M (1997) Release of cy- couples TRAIL receptors to apoptosis-inducing pro-
tochrome c and decrease of cytochrome c oxidase in teins. Nat Immunol 2:493–500
Bax-expressing yeast cells, and prevention of these ef- Money NP (2003) Suicidal mushrooms. Nature 423:26
fects by coexpression of Bcl-xL. FEBS Lett 415:29–32 Moon H, Baek D, Lee B, Prasad DT, Lee SY, Cho MJ, Lim CO,
Manon S, Priault M, Camougrand N (2001) Mitochondrial Choi MS, Bahk J, Kim MO et al. (2002) Soybean ascor-
AAA-type protease Yme1p is involved in Bax effects on bate peroxidase suppresses Bax-induced apoptosis in
cytochrome c oxidase. Biochem Biophys Res Commun yeast by inhibiting oxygen radical generation. Biochem
289:1314–1319 Biophys Res Commun 290:457–462
Marek SM, Wu J, Louise Glass N, Gilchrist DG, Bostock RM Moore EJ (1965) Fungal gel tissue ontogenesis. Am J Bot
(2003) Nuclear DNA degradation during heterokaryon 52:389–395
incompatibility in Neurospora crassa. Fungal Genet Moore D (2004) Programmed cell death alive and well in
Biol 40:126–137 fungi. Mycol Res 107:1251–1252
Martinet W, Van den Plas D, Raes H, Reekmans R, Contr- Moser MJ, Geiser JR, Davis TN (1996) Ca2+ -calmodulin
eras R (1999) Bax-induced cell death in Pichia pastoris. promotes survival of pheromone-induced growth ar-
Biotechnol Lett 21:821–829 rest by activation of calcineurin and Ca2+ -calmodulin-
Marzo I, Brenner C, Zamzami N, Jurgensmeier JM, Susin SA, dependent protein kinase. Mol Cell Biol 16:4824–4831
Vieira HL, Prevost MC, Xie Z, Matsuyama S, Reed JC Mousavi SA, Robson GD (2003) Entry into the station-
et al. (1998) Bax and adenine nucleotide translocator ary phase is associated with a rapid loss of viabiliy
cooperate in the mitochondrial control of apoptosis. and an apoptotic-like phenotype in the opportunis-
Science 281:2027–2031 tic pathogen Aspergillus fumigatau. Fungal Genet Biol
Mathieu AL, Gonin S, Leverrier Y, Blanquier B, Thomas J, 39:221–229
Dantin C, Martin G, Baverel G, Marvel J (2001) Activa- Mousavi SA, Robson GD (2004) Oxidative and amphotericin
tion of the phosphatidylinositol 3-kinase/Akt pathway B-mediated cell death in the opportunistic pathogen
protects against interleukin-3 starvation but not DNA Aspergillus fumigatus is associated with an apoptotic-
damage-induced apoptosis. J Biol Chem 276:10935– like phenotype. Microbiology 150:1937–1945
10942 Narasimhan ML, Damsz B, Coca MA, Ibeas JI, Yun DJ,
Matsuyama S, Schendel SL, Xie Z, Reed JC (1998a) Cytopro- Pardo JM, Hasegawa PM, Bressan RA (2001) A plant
tection by Bcl-2 requires the pore-forming alpha5 and defense response effector induces microbial apoptosis.
alpha6 helices. J Biol Chem 273:30995–31001 Mol Cell 8:921–930
Matsuyama S, Xu Q, Velours J, Reed JC (1998b) The mi- Nestelbacher R, Laun P, Vondrakova D, Pichova A,
tochondrial F0 F1 -ATPase proton pump is required for Schuller C, Breitenbach M (2000) The influence of
function of the proapoptotic protein Bax in yeast and oxygen toxicity on yeast mother cell-specific aging.
mammalian cells. Mol Cell 1:327–336 Exp Gerontol 35:63–70
Matsuyama S, Nouraini S, Reed JC (1999) Yeast as a tool for Nigro JM, Sikorski R, Reed SI, Vogelstein B (1992) Human
apoptosis research. Curr Opin Microbiol 2:618–623 p53 and CDC2Hs genes combine to inhibit the pro-
Mazzoni C, Mancini P, Verdone L, Madeo F, Serafini A, liferation of Saccharomyces cerevisiae. Mol Cell Biol
Herker E, Falcone C (2003) A truncated form of 12:1357–1365
KlLsm4p and the absence of factors involved in mRNA Oberhammer F, Wilson JW, Dive C, Morris ID, Hickman JA,
decapping trigger apoptosis in yeast. Mol Biol Cell Wakeling AE, Walker PR, Sikorska M (1993) Apoptotic
14:721–729 death in epithelial cells: cleavage of DNA to 300 and/or
McCann M, Geraghty M, Devereux M, O’Shea D, Mason J, 50 kb fragments prior to or in the absence of internu-
O’Sullivan L (2000) Insights into the mode of action of cleosomal fragmentation. EMBO J 12:3679–3684
the anti-Candida activity of 1,10-phenanthroline and Orlandi I, Bettiga M, Alberghina L, Vai M (2004) Tran-
its metal chelates. Metal-Based Drugs 7:185–193 scriptional profiling of ubp10 null mutant reveals al-
Mendoza I, Rubio F, Rodriguez-Navarro A, Pardo JM (1994) tered subtelomeric gene expression and insurgence
The protein phosphatase calcineurin is essential for of oxidative stress response. J Biol Chem 279:6414–
NaCl tolerance of Saccharomyces cerevisiae. J Biol 6425
Chem 271:8792–8796 Osiewacz HD (2004) Aging and mitochondrial dysfunction
Metzstein MM, Stanfield GM, Horvitz HR (1998) Genetics in the filamentous fungus Podospora anserina. In: Nys-
of programmed cell death in C. elegans: past, present trom T, Osiewacz HD (eds) Model systems in ageing.
and future. Trends Genet 14:410–416 Springer, Berlin Heidelberg New York, pp 17–38
Migheli A, Piva R, Wei J, Attanasio A, Casolino S, Hodes ME, Osiewacz HD, Hamann A (2006) Senescence and longevity.
Dlouhy SR, Bayer SA, Ghetti B (1997) Diverse cell death In: Esser K (ed) The Mycota, vol I, 2nd edn. Kües U,
pathways result from a single missense mutation in Fischer R (eds) Growth, differentiation and sexuality.
weaver mouse. Am J Pathol 151:1629–1638 Springer, Berlin Heidelberg New York (in press) Os-
Millar JB, Russell P, Dixon JE, Guan KL (1992) Negative mani SA, Pu RT, Morris NR (1988) Mitotic induction
regulation of mitosis by two functionally overlapping and maintenance by overexpression of a G2-specfic
PTPases in fission yeast. EMBO J 11:4943–4952 gene that encodes a potential protein kinase. Cell
Minn AJ, Kettlun CS, Liang H, Kelekar A, Vander 53:237–244
Heiden MG, Chang BS, Fesik SW, Fill M, Thomp- Ottilie S, Chernoff J, Hannig G, Hoffman CS, Erikson RL
son CB (1999) Bcl-xL regulates apoptosis by hetero- (1992) The fission yeast genes pyp1+ and pyp2+ encode
144 M. Ramsdale
protein tyrosine phosphatases that negatively regulate tophagy of nucleus in Saccharomyces cerevisiae. Mol
mitosis. Mol Cell Biol 12:5571–5580 Biol Cell 14:129–141
Pavlov EV, Priault M, Pietkiewicz D, Cheng EH, Antons- Rose MD, Fink GR (1987) KAR1, a gene required for
son B, Manon S, Korsmeyer SJ, Mannella CA, Kin- function of both intranuclear and extranuclear
nally KW (2001) A novel, high conductance channel of microtubules in yeast. Cell 48:1047–1060
mitochondria linked to apoptosis in mammalian cells Roucou X, Prescott M, Devenish RJ, Nagley P (2000) A cy-
and Bax expression in yeast. J Cell Biol 155:725–731 tochrome c-GFP fusion is not released from mitochon-
Pearson GD, Merrill GF (1998) Deletion of the Saccha- dria into the cytoplasm upon expression of Bax in yeast
romyces cerevisiae TRR1 gene encoding thioredoxin cells. FEBS Lett 471:235–239
reductase inhibits p53-dependent reporter gene ex- Roze LV, Linz JE (1998) Lovastatin triggers an apoptosis-
pression. J Biol Chem 273:5431–5434 like cell death process in the fungus Mucor racemosus.
Pham CT, Thomas DA, Mercer JD, Ley TJ (1998) Produc- Fungal Genet Biol 25:119–133
tion of fully active recombinant murine granzyme B in Russell P, Nurse P (1987) Negative regulation of mitosis by
yeast. J Biol Chem 273:1629–1633 wee1+ , a gene encoding a protein kinase homolog. Cell
Phillips AJ, Sudbery I, Ramsdale M (2003) Apoptosis 49:559–567
induced by environmental stresses and amphotericin Ryser S, Vial E, Magnenat E, Schlegel W, Maundrell K (1999)
B in Candida albicans. Proc Natl Acad Sci USA Reconstitution of caspase-mediated cell death signal-
100:14327–14332 ing in Schizosaccharomyces pombe. Curr Genet 36:21–
Pinan-Lucarre B, Paoletti M, Dementhon K, Coulary- 28
Salin B, Clave C (2003) Autophagy is induced during Sámi L, Emri T, Pócsi I (2001) Autolysis and ageing of Peni-
cell death by incompatibility and is essential for cillium chrysogenum cultures under carbon starvation:
differentiation in the filamentous fungus Podospora glutathione metabolism and formation of reactive oxy-
anserina. Mol Microbiol 47:321–333 gen species. Mycol Res 105:1246–1250
Polcic P, Forte M (2003) Response of yeast to the regulated Sanglard D, Ischer F, Marchetti O, Entenza J, Bille J (2003)
expression of proteins in the Bcl-2 family. Biochem J Calcineurin A of Candida albicans: involvement in an-
37:393–402 tifungal tolerance, cell morphogenesis and virulence.
Priault M, Camougrand N, Chaudhuri B, Schaeffer J, Mol Microbiol 48:959–976
Manon S (1999a) Comparison of the effects of Saupe S, Descamps C, Turcq B, Bégueret J (1994) Inactiva-
bax-expression in yeast under fermentative and tion of the Podospora anserina vegetative incompati-
respiratory conditions: investigation of the role of bility locus het-c, whose product resembles a glycolipid
adenine nucleotides carrier and cytochrome c. FEBS transfer protein, drastically impairs ascospore produc-
Lett 456:232–238 tion. Proc Natl Acad Sci USA 91:5927–5931
Priault M, Chaudhuri B, Clow A, Camougrand N, Manon S Sekiguchi T, Nakashima T, Hayashida T, Kuraoka A,
(1999b) Investigation of bax-induced release of cy- Hashimoto S, Tsuchida N, Shibata Y, Hunter T, Nishi-
tochrome c from yeast mitochondria permeability of moto T et al. (1995) Apoptosis is induced in BHK cells
mitochondrial membranes, role of VDAC and ATP re- by the tsBN462/13 mutation in the CCG1/TAFII250
quirement. Eur J Biochem 260:684–691 subunit of the TFIID basal transcription factor. Exp
Priault M, Bessoule JJ, Grelaud-Coq A, Camougrand N, Cell Res 218:490–498
Manon S (2002) Bax-induced cell death in yeast de- Severin FF, Hyman AA (2002) Pheromone induces pro-
pends on mitochondrial lipid oxidation. Eur J Biochem grammed cell death in S. cerevisiae. Curr Biol 12:R233–
269:5440–5450 R235
Qi H, Li TK, Kuo D, Nur-E-Kamal A, Liu LF (2003) Inacti- Shimizu S, Narita M, Tsujimoto Y (1999) Bcl-2 family pro-
vation of Cdc13p triggers MEC1-dependent apoptotic teins regulate the release of apoptogenic cytochrome c
signals in yeast. J Biol Chem 278:15136–15141 by the mitochondrial channel VDAC. Nature 399:483–
Raju NB, Perkins DD (2000) Programmed ascospore death 487
in the homothallic ascomycete Coniochaete tetraspora. Shirogane T, Fukada T, Muller JM, Shima DT, Hibi M, Hi-
Fungal Genet Biol 30:213–221 rano T (1999) Synergistic roles for Pim-1 and c-Myc in
Ramer SW, Elledge SJ, Davis RW (1992) Dominant genet- STAT3-mediated cell cycle progression and antiapop-
ics using a yeast genomic library under the control of tosis. Immunity 11:709–719
a strong inducible promoter. Proc Natl Acad Sci USA Sinclair DA, Guarente L (1997) Extrachromosomal rDNA
89:11589–11593 circles – a cause of aging in yeast. Cell 91:1033–1042
Ramsdale M (1998) Genomic conflict in fungal mycelia: Smriti, Krishnamurthy S, Dixit BL, Gupta CM, Milewski S,
a subcellular population biology. In: Worrall JJ (ed) The Prasad R (2002) ABC transporters Cdr1p, Cdr2p and
structure and dynamics of fungal populations. Kluwer, Cdr3p of a human pathogen Candida albicans are gen-
Dordrecht, pp 139–174 eral phospholipid translocators. Yeast 19:303–318
Ramsdale M, Rayner ADM (1996) Phenotype-genotype re- Soloca R, Silva R, Ludovico P, Sansonetti F, Martinez-
lations in allopatrically-derived heterokaryons of Het- Peinado J, Corte-Real M (2003) High sugar con-
erobasidion annosum (Fr.) Bref. New Phytol 133:303– centrations trigger Saccharomyces cerevisiae into
319 a programmed cell death process. In: Abstr Vol 21st Int
Rayner ADM (1991) The challenge of the individualistic Yeast Genetics and Molecular Biology Meet, 7–12 July
mycelium. Mycologia 83:48–71 2003, Goteborg, Sweden, Abstr 18-50. Yeast 20:S306
Roberts P, Moshitch-Moshkovitz S, Kvam E, O’Toole E, Stanhill A, Schick N, Engelberg D (1999) The yeast
Winey M, Goldfarb DS (2003) Piecemeal microau- Ras/cyclic AMP pathway induces invasive growth by
suppressing the cellular stress response. Mol Cell Biol Umar MH, van Griensven LJLD (1997) Morphogenetic cell
19:7529–7538 death in developing primordia of Agaricus bisporus.
Stevenson LF, Kennedy BK, Harlow E (2001) A large-scale Mycologia 89:274–277
overexpression screen in Saccharomyces cerevisiae Umar MH, van Griensven LJLD (1998) The role of morpho-
identifies previously uncharacterized cell cycle genes. genetic cell death in the histogenesis of the mycelial
Proc Natl Acad Sci USA 98:3946–3951 cord of Agaricus bisporus and in the development of
Szallies A, Kubata BK, Duszenko M (2002) A metacaspase macrofungi. Mycol Res 102:719–735
of Trypanosoma brucei causes loss of respiration com- Uren AG, Beilharz T, O’Connell MJ, Bugg SJ, van Driel R,
petence and clonal death in the yeast Saccharomyces Vaux DL, Lithgow T (1999) Role for yeast inhibitor of
cerevisiae. FEBS Lett 517:144–150 apoptosis (IAP)-like proteins in cell division. Proc Natl
Tao W, Walke DW, Morgan JI (1999) Oligomerized Ced- Acad Sci USA 96:10170–10175
4 kills budding yeast through a caspase-independent Uren AG, O’Rourke K, Aravind LA, Pisabarro MT, Sesha-
mechanism. Biochem Biophys Res Commun 260:799– giri S, Koonin EV, Dixit VM (2000) Identification of
805 paracaspases and metacaspases: two ancient families
Telford WG, King LE, Fraker J (1992) Comparative evalua- of caspase-like proteins, one of which plays a key role
tion of several DNA binding dyes in the detection of in MALT lymphoma. Mol Cell 6:961–967
apoptosis associated chromatin degradation by flow Van Diepeningen AD, Debets AJM, Hoekstra RF (1997) Het-
cytometry. Cytometry 13:137–143 erokaryon incompatibility blocks virus transfer among
Tessier C, Prigent-Tessier A, Ferguson-Gottschall S, Gu Y, natural isolates of black Aspergilli. Curr Genet 32:209–
Gibori G (2001) PRL antiapoptotic effect in the rat 217
decidua involves the PI3K/ protein kinase B-mediated Verhagen AM, Ekert PG, Pakusch M, Silke J, Connolly LM,
inhibition of caspase-3 activity. Endocrinology Reid GE, Moritz RL, Simpson RJ, Vaux DL (2000) Iden-
142:4086–4094 tification of DABLO, a mammalian protein that pro-
Thevissen K, Ghazi A, De Samblanx GW, Brownlee C, Os- motes apoptosis by binding to and antagonizing IAP
born RW, Broekaert WF (1996) Fungal membrane re- proteins. Cell 102:43–53
sponses induced by plant defensins and thionins. J Biol von Gise A, Lorenz P, Wellbrock C, Hemmings B, Berberich-
Chem 271:15018–15025 Siebelt F, Rapp UR, Troppmair J (2001) Apoptosis sup-
Thevissen K, Terras FR, Broekaert WF (1999) Permeabiliza- pression by Raf-1 and MEK1 requires MEK- and phos-
tion of fungal membranes by plant defensins inhibits phatidylinositol 3-kinase-dependent signals. Mol Cell
fungal growth. Appl Environ Microbiol 65:5451–5458 Biol 21:2324–2336
Thevissen K, Cammue BP, Lemaire K, Winderickx J, Dick- Wadskog I, Maldener C, Proksch A, Madeo F, Adler L
son RC, Lester RL, Ferket KK, Van Even F, Parret AH, (2004) Yeast lacking the SRO7/SOP1-encoded tumor
Broekaert WF (2000) A gene encoding a sphingolipid suppressor homologue show increased susceptibility
biosynthesis enzyme determines the sensitivity of Sac- to apoptosis-like cell death on exposure to NaCl stress.
charomyces cerevisiae to an antifungal plant defensin Mol Biol Cell 15:1436–1444
from dahlia (Dahlia merckii). Proc Natl Acad Sci USA Watanabe K, Morishita J, Umezu K, Shirahige K, Maki H
97:9531–9536 (2002) Involvement of RAD9-dependent damage
Thiede B, Dimmler C, Siejak F, Rudel T (2001) Predom- checkpoint control in arrest of cell cycle, induction of
inant identification of RNA-binding proteins in Fas- cell death, and chromosome instability caused by de-
induced apoptosis by proteome analysis. J Biol Chem fects in origin recognition complex in Saccharomyces
276:26044–26050 cerevisiae. Eukaryot Cell 1200–1212
Thomma BPHJ, Cammue BPA, Thevissen K (2002) Plant Wawryn J, Krzepilko A, Myszka A, Bilinski T (1999) Defi-
defensins. Planta 216:193–202 ciency in superoxide dismutases shortens life span of
Thrane C, Kaufmann U, Stummann BM, Olsson S (2004) yeast cells. Acta Biochim Pol 46:249–253
Activation of caspase-like activity and poly (ADP- Weinberger M, Trabold PA, Lu M, Sharma K, Huberman JA,
ribose) polymerase degradation during sporulation Burhans WC (1999) Induction by adozelesin and hy-
in Aspergillus nidulans. Fungal Genet Biol 41:361–368 droxyurea of origin recognition complex-dependent
Toledo I, Rangel P, Hansberg W (1995) Redox imbalance DNA damage and DNA replication checkpoints in Sac-
at the start of each morphogenetic step of Neurospora charomyces cerevisiae. J Biol Chem 274:35975–35984
crassa conidiation. Arch Biochem Biophys 319:519–524 Whiteway M, Hougan L, Thomas DY (1990) Overexpression
Torgler CN, de Tiani M, Raven T, Aubry JP, Brown R, Mel- of the STE4 gene leads to mating response in haploid
drum E (1997) Expression of Bak in S. pombe results in Saccharomyces cerevisiae. Mol Cell Biol 10:217–222
a lethality mediated through interaction with the cal- Wissing S, Ludovico P, Herker E, Buttner S, Engelhardt SM,
nexin homologue Cnx1. Cell Death Differ 4:263–271 Decker T, Link A, Proksch A, Rodrigues F, Corte-Real
Tschopp J, Thome M, Hofmann K, Meinl E (1998) The M et al. (2004) An AIF orthologue regulates apoptosis
fight of viruses against apoptosis. Curr Opin Genet in yeast. J Cell Biol 166:969–974
Dev 8:82–87 Wright ME, Han DK, Carter L, Fields S, Schwartz SM, Hock-
Uchiyama Y (2001) Autophagic cell death and its execution enbery DM (1999) Caspase-3 inhibits growth in Sac-
by lysosomal cathepsins. Arch Histol Cytol 64:233–246 charomyces cerevisiae without causing cell death. FEBS
Uetz P, Giot L, Cagney G, Mansfield TA, Judson RS, Lett 446:9–14
Knight JR, Lockshon D, Narayan V, Srinivasan M, Wu D, Chen P-J, Chen S, Hu Y, Nunez G, Ellis RE (1999)
Pochart P et al. (2000) A comprehensive analysis C. elegans MAC-1, an essential member of the AAA
of protein–protein interactions in Saccharomyces family of ATPases, can bind CED-4 and prevent cell
cerevisiae. Nature 403:623–627 death. Development 126:2021–2031
146 M. Ramsdale
Wunder D, Dong J, Baev D, Edgerton M (2004) Human Zhang L, Zhang Y, Zhou Y, An S, Zhou Y, Cheng J
salivary histatin 5 fungicidal action does not induce (2002a) Response of gene expression in Saccha-
programmed cell death pathways in Candida albicans. romyces cerevisiae to amphotericin B and nystatin
Antimicrob Agents Chemother 48:110–115 measured by microarrays. J Antimicrob Chemother
Yamaki M, Umehara T, Chimura T, Horikoshi M (2001) Cell 49:905–915
death with predominant apoptotic features in Saccha- Zhang L, Zhang Y, Zhou Y, Zhao Y, Zhou Y, Cheng J (2002b)
romyces cerevisiae mediated by deletion of the histone Expression profiling of the response of Saccharomyces
chaperone ASF1/CIA1. Genes Cells 6:1043–1054 cerevisiae to 5-fluorocytosine using a DNA microarray.
Yarmolinsky MB (1995) Programmed cell death in bacterial Int J Antimicrob Agents 20:444–450
populations. Science 267:836–837 Zhang Q, Chieu HK, Low CP, Zhang S, Heng CK, Yang H
Yoon HJ, Carbon J (1999) Participation of Bir1p, a mem- (2003) Schizosaccharomyces pombe cells deficient in
ber of the inhibitor of apoptosis family, in yeast chro- triacylglycerol synthesis undergo apoptosis upon en-
mosome segregation events. Proc Natl Acad Sci USA try into the stationary phase. J Biol Chem 278:47145–
96:13208–13213 47155
Yun DJ, Ibeas JI, Lee H, Coca MA, Narasimhan ML, Zou H, Henzel WJ, Liu X, Lutschg A, Wang X (1997) Apaf-1,
Uesono Y, Hasegawa PM, Pardo JM, Bressan RA (1998) a human protein homolog to C. elegans CED-4,
Osmotin, a plant antifungal protein, subverts signal participates in cytochrome c-dependent activation of
transduction to enhance fungal cell susceptibility. Mol caspase-3. Cell 90:405–413
Cell 1:807–817
8 Genomic Analysis of Cellular Morphology in Candida albicans
M. Whiteway1 , A. Nantel1
CONTENTS tantly related organisms and see the processes of

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 147 evolution at work. This ability to address scien-
II. Genomic Technologies Applied tific questions within the framework of the entire
to C. albicans . . . . . . . . . . . . . . . . . . . . . . . . . 148 genome sequence of an organism is revolutioniz-
III. Candida albicans Biology: Morphogenesis ing the approach to the investigation of all living
and Virulence . . . . . . . . . . . . . . . . . . . . . . . . 150
A. Yeast-to-Hyphae Transition . . . . . . . . . . . 151 systems.
B. Biofilms . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Genomic technologies are increasingly being
C. Mating Projections . . . . . . . . . . . . . . . . . . 156 applied to the analyses of fungi. The ascomycete
D. White–Opaque Transition . . . . . . . . . . . . 156 yeast Saccharomyces cerevisiae was the first eukary-
E. Chlamydospore Formation . . . . . . . . . . . 157
IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 157 otic cell to have its genome completely sequenced.
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 This sequencing effort involved an international ef-
fort encompassing many laboratories, each respon-
sible for a part of the genome. However, the de-
Abbreviations: DNA, deoxyribonucleic acid; velopment of high-throughput, industrialized se-
cDNA, copy DNA; mRNA, messenger ribonu- quencing centers has established conditions where
cleic acid; PCR, polymerase chain reaction; cy3, a comprehensive, several-fold redundant sequence
cyanine 3; cy5, cyanine 5; ORF, open reading of a fungal genome can be generated in a few
frame; YPD, yeast extract peptone dextrose; months at a reasonable cost. This technical capacity
cAMP, cyclic adenosine monophosphate; UAU, is putting the sequences of most experimentally or
uracil–arginine–uracil cassette economically important fungi into publicly accessi-
ble databases. These sequences underlie the ability
to address questions of cellular evolution and ge-
I. Introduction
netic function on a comprehensive, genome-wide
level.
Genomics can be defined as the scientific approach The genome of Candida albicans repre-
to biological questions that is based in some way sented the first sequence of a human pathogenic
on the knowledge of the complete sequence of the fungus to be accessible in a public database.
DNA of an organism. Many of the approaches that This tour de force was accomplished by the
make up the science of genomics are not new; they Stanford Genome Technology Center (http://www-
are simply made more efficient through the ap- sequence.stanford.edu/group/candida; Jones et al.
plication of information resulting from this com- 2004). Subsequent efforts from the Candida
plete sequence. Comprehensivity is gained from community, spearheaded by the Biotechnology
this knowledge – one can look at the expression Research Center in Montreal, have led to the devel-
of every gene in the organism through microar- opment of an accurately annotated version of this
ray analysis, one can build collections of simple sequence (http://candida.bri.nrc.ca/), and ongoing
cells that each lack a specific gene, and make that efforts at the Montreal Institute and the Magee
collection encompass all the genes, one can build group at the University of Minnesota are creating
constructs to investigate the cellular location of ev- a final assembly of the genome. Ultimately, all the
ery protein in a cell, and one can compare and genomic information available about C. albicans
contrast the genetic constitution of closely and dis- will be collected and distributed through the
1Biotechnology Research Institute, National Research Council of Candida Genome Database (CGD; http://genome-
Canada, 6100 Royalmount Ave., Montreal, Quebec H4P 2R2, Canada www.stanford.edu/fungi/Candida/) developed a-
The Mycota XIII
Fungal Genomics
148 M. Whiteway and A. Nantel
round the model of the Saccharomyces Genome surface, the two-color array pioneered by the
Database (SGD; http://www.yeastgenome.org). Brown group at Stanford (DeRisi et al. 1997), and
While annotation and assembly provide the the multiple oligonucleotide array technology
genome in a more readily accessible form, the of Affymetrix. The Affymetrix technology was
genomic era of C. albicans started with the release the first to be widely exploited. In this approach,
of the initial sequence information from Stanford. which borrows heavily from the concepts of the
semiconductor industry, short oligonucleotides
of specific sequences are synthesized directly at
a defined location on a solid surface. Each gene to
II. Genomic Technologies Applied be probed can be represented by several oligonu-
to C. albicans cleotides with a perfect match, as well as several
nucleotides with single base mismatches. Thus, the
Candida albicans is a particularly attractive organ- Affymetrix array consists of an enormous number
ism for the application of genome-based experi- of short oligonucleotides representing perfect or
mental strategies, because these approaches can near-perfect matches to each of the genes of the
circumvent some of the technical problems asso- organism to be profiled. When these arrays are
ciated with studies of this fungus. C. albicans is used to determine the transcription profile of
diploid, and prior to the genome sequencing ef- a particular cell type under a particular condition,
fort it was also defined as an asexual fungus. Re- mRNA is isolated, turned into labeled cDNA
cently, however, the recognition that the genome and then hybridized to the array. The strength
contained sequences homologous to the mating- of the signals representing the perfect matches
type loci of S. cerevisiae permitted researchers to are compared to the strengths of the signals
uncover a complex mating regulatory circuitry, and coming from the mismatched oligonucleotides,
to construct strains that exhibit a mating capacity and computer analysis is used to determine the
(Hull et al. 2000; Magee and Magee 2000). Still, even strength and quality of each measured interaction.
if a sexual cycle were to be defined completely, the Repetition of the process is used to establish the
diploid nature of C. albicans precludes the iden- statistical quality of the determined signals.
tification of the recessive mutants in cellular pro- Affymetrix arrays require a sophisticated
cesses that have made S. cerevisiae one of the pre- production capacity that is beyond the scope
eminent model organisms in the analysis of cellular of most research laboratories. Therefore, most
function. The genome sequence, however, provides of the published transcriptional profiles were
access to gene function through alternate routes. produced through the use of either membranes
For example, transcriptional profiling can identify or the two-color array technology developed at
genes whose expression is regulated in a defined Stanford. In general, these arrays are constructed
manner, and such profiles can be used to group by physically spotting the probe nucleotides
genes into functional classes in the same way that at specific locations on a solid surface such as
mutations are able to group genes based on their a membrane or a coated glass microscope slide.
phenotypic consequences. The ability to scan the The spotting strategy can range from the use
genome for genes with important predicted func- of pins to ink-jet printers (Hughes et al. 2001),
tions, or to identify open reading frames that ex- and probes can be PCR products, cDNAs or
hibit interesting evolutionary relationships (such oligonucleotides. If the latter, the oligonucleotides
as being found only in mammals as well as Can- are typically in the range of 70 bases long, and
dida, but not found in other fungi, or being unique therefore are considerably longer than the probes
to C. albicans) also allow researchers to focus their used for the Affymetrix technology. The mem-
efforts on genes with potentially unique and im- brane system involves spotting probes on surfaces
portant characteristics. such as nitrocellulose, and probing typically with
Transcriptional profiling makes use of DNA radioactively labeled samples that are scored
microarrays. Microarrays consist of nucleotide absolutely, not comparatively. Much of the early
sequences affixed to a solid support, which can be transcription profiling work for C. albicans used
used to probe for the presence, and to a certain membrane-based arrays (Lane et al. 2001; Murad
extent for the quantity, of nucleotides in a given et al. 2001a). Unlike the membrane array system,
target. In general, there are three major platforms the two-color system is directly comparative.
for these arrays; DNAs attached to a membrane To use this system to investigate a transcription
Candida albicans Morphogenesis 149
profile, mRNA is extracted from cells under two http://candida.bri.nrc.ca/candida/index.cfm?page

different conditions, such as a mutant compared =CaChrom. Not all chromosomes could be
to a wild-type strain, or a given strain grown completely purified, but substantial enrichment
under two different culture conditions. The RNA was accomplished in each case. This permitted
from one sample is copied into cDNA and labeled each ORF to be attributed to a chromosome with
with a fluorescent dye, typically Cy3 or Cy5, and little ambiguity, and when this assignment was
the RNA from the other sample is labeled with mapped to the currently established contigs, the
the other dye. The two samples are combined, majority of contigs were uniquely connected to
and hybridized to the array. If the two RNAs for a chromosome. However, on some occasions the
a specific probe are equally abundant, then the contigs could map to two chromosomes, and this
laser reads the combined Cy3 and Cy5 signals permits the identification of misassembled contigs
as a yellow signal at that point on the array; if and defines the point at which the misconnection
one or the other sample contains a higher level was made.
of message, then either the Cy3 (green) or the A second genomic technology involves the con-
Cy5 (red) signal predominates. Normalization struction of libraries of mutant strains. The diploid
of signals, signal quality and signal strength are nature of C. albicans precludes the simple produc-
monitored and, as with the Affymetrix technique, tion of collections of recessive mutants. However,
experimental repetitions are used to provide the tremendous utility of the systematic collection
statistical robustness to the data. This approach of S. cerevisiae ORF disruptions has proven the
was used initially to investigate the expression power of comprehensive null mutant collections.
of a subset of the C. albicans genome to look at The effort to create the disruption set in S. cere-
azole sensitive and resistant isolates (Cowen et al. visiae involved a substantial international under-
2002; Rogers and Barker 2002), while the first taking, and only recently, with the construction of
large scale uses of this technology to investigate a set of regulated promoter shut-offs for essential
transcriptional regulation of C. albicans were genes (Mnaimneh et al. 2004), has the collection
studies on the expression profiles of Candida cells truly approached being comprehensive. However,
growing in the presence of the antifungal drug projects are currently underway to apply the sys-
itraconazole (DeBacker et al. 2001) and on cells tematic disruption approach to C. albicans (Bruno
undergoing the yeast-to-hyphal transition (Nantel and Mitchell 2004). The annotated genome is the
et al. 2002). starting point for such studies. A fungal drug dis-
Microarrays are not simply useful for the covery company, Elitra Canada, has developed the
determination of transcription profiles. Because most advanced disruption technology. Researchers
the probes will associate with any homologous, at this company have currently replaced one al-
labeled DNA, an array can be used to determine lele of about 80% of the C. albicans genes with
the presence or absence of a gene in a genome. a sequence coded selectable marker, and for about
This strategy can be used in strain comparisons half these genes have replaced the endogenous pro-
for taxonomic purposes. It can also be useful in moter of the other allele with a regulated promoter.
the assembly of a genome. Purified chromosomes This approach, dubbed the GRACE strategy (for
or chromosome fragments can be hybridized gene replacement and controlled expression), has
to comprehensive gene chips to test the quality created the ability to shut off genes representing
of contig assemblies. All the genes on a specific a major fraction of the entire C. albicans genome
contig should hybridize to a given chromosome (Roemer et al. 2003). Alternative approaches have
or chromosome fragment; this provides a quality also been developed to provide comprehensive mu-
control for the assembly of the contigs, and also tant collections. One strategy has been to examine
connects the contigs to a specific chromosome. the phenotypes of cells that are missing a single
This approach has proven useful in the assem- copy of a given gene. Such an approach depends
bly of the C. albicans genome. Pulse field gel on haplo-insufficiency of the disrupted gene caus-
electrophoresis was used to separate all the chro- ing a detectable phenotype (Uhl et al. 2003). In
mosomes of C. albicans strain SC5314, and these addition, a clever strategy to select homozygosis
separated chromosomes were purified, labeled of an insertion has been developed that obviates
with fluorescent dyes, and hybridized to arrays the need to perform double disruptions of each
consisting of over 6000 open reading frames. The gene to be inactivated (Davis et al. 2002). This ap-
data for these experiments are available online at proach makes use of a disruption cassette that in-
cludes a selectable marker embedded in, and inac- istic flattened appearance. Colonies formed from
tivating a second selectable marker. Excision of the cells with a high proportion of hyphal and pseu-
first marker (ARG4) through recombination will re- dohyphal cells tend to be more wrinkled than are
sult in the activation of the previously inactivated colonies formed from yeast cells. The ability to un-
marker (URA3). This UAU cassette is used to dis- dergo mating is also linked to cell shape. Cells with
rupt a single copy of a chromosomal gene, and the the genetic constitution to be mating competent
disruptant is able to proliferate. If the chromosomal can only mate when they are in the opaque state;
disruptant is not in an essential gene, then the pop- white cells, regardless of the genotype, are unable
ulation arising from the initial event will include to mate (Miller and Johnson 2002). The mating pro-
mitotic recombinants that have homozygosed the cess itself involves changes in cellular morphology;
disruptant allele, creating a rare cell that has two in response to mating pheromones, the cells send
copies of the disruptant allele, and no wild-type out mating projections that serve to link the two
gene copy. This rare cell has the unique ability to mating cells (Soll 2004). Thus, cellular morphol-
excise the ARG4 marker from one copy to create ogy plays an important role in the overall biology
a URA3+ allele, while maintaining the ARG4+ copy of C. albicans.
at the other allele, and thus can give rise to ARG+ These cellular and colony morphologies are
URA+ events. Therefore, the selection of the ARG+ important components of the virulence of this
URA+ cells will identify cells that lack a functional pathogen. The infective properties of white and
copy of the targeted gene. Systematic application of opaque cells are different; white cells are more
this cassette can create a knockout collection of the virulent in systemic infections, while opaque cells
non-essential genes (Bruno and Mitchell 2004). can be more virulent in cutaneous infections
(Kvaal et al. 1997). The ability to shift from a yeast
form to a hyphal cell also appears critical for
virulence. Systemic Candida infections typically
III. Candida albicans Biology: contain yeast, hyphal and pseudohyphal cells.
Morphogenesis and Virulence Mutants trapped in either the yeast form or the
pseudohyphal form are less virulent than wild
C. albicans is an important human pathogen be- type when tested in mouse models of systemic
cause it can opportunistically exploit weaknesses infection, suggesting that the ability to change
in the human defenses against infection. One of the the cellular form plays a fundamental role in
defining characteristics of C. albicans is the mor- the infection process (Braun and Johnson 1997;
phological plasticity of the organism. It can grow as Lo et al. 1997; Rocha et al. 2001). Aspects of
physically separated yeast cells though a budding colony formation are also implicated in virulence.
process, as chains of attached, elongated cells called Biofilms represent a specific colony form capable of
pseudohyphae created by a directed budding pro- strong attachment to a surface (Kumamoto 2002).
cess, or as true hyphae, where a parallel-sided tube Such cellular assemblies can provide a serious
is extended at the apical tip, and separate cellular problem in hospital settings; biofilm formations
compartments are created by septation (Sudbery on, for example, catheters can be difficult to
et al. 2004; Whiteway and Oberholzer 2004). In ad- dislodge, and the cells within biofilms can exhibit
dition, cells can take either a white or an opaque heightened resistance to antifungal drugs (Ramage
form; white cells have a smooth, rounded shape, et al. 2001). Therefore, an understanding of the
while opaque cells are much more elongated and control of cellular morphology is an important
have a pimpled surface (Soll 1992). These differ- goal in experimental analysis of C. albicans, both
ent forms are under genetic control, because mu- in terms of our overall knowledge of basic biology
tations can dramatically perturb the regulation of of the organism as well as for our ability to develop
morphology, but they are also under environmen- new therapies against disease caused by Candida
tal control because changing conditions can effi- infections.
ciently direct the cells toward one morphological Because cellular morphology is under intrinsic
form or another. Intriguingly, changes in the shape genetic control, as well as under control of the envi-
of the individual cells are reflected in changes in the ronment, research efforts have been directed at de-
shape of colonies created by these cells. White cells termining the signaling processes responsible for
form smooth, domed colonies on agar surfaces, the internal control as well as transmission of the
while opaque cells form colonies with a character- external information to the intracellular machin-
ery controlling cell shape. Genomic approaches are others encode components of signal transduction
now providing critical information during these networks. Genomic analyses of these transcription
studies. The following sections provide an overview factors and signaling network members is provid-
of the genomic analyses of several distinct morpho- ing insight into how external conditions serve to
logical states. regulate the yeast-to-hyphal transition.
The observation that pseudohyphal growth in
S. cerevisiae involved transcription factors that had
A. Yeast-to-Hyphae Transition homologs in C. albicans led the Fink laboratory to
investigate whether these C. albicans transcription
Wild-type C. albicans cells can undergo the yeast- factors controlled hyphal development. They con-
to-hyphal transition readily in response to exter- structed strains that lacked the Efg1p and Cph1p
nal conditions, and this process is the most exten- proteins, and showed that these strains were un-
sively studied of the variations in shape exhibited able to enter hyphal development in the presence
by this fungus. Classical inducing conditions in- of serum and high temperature, two conditions that
clude transferring yeast cells to YPD culture me- are usually highly efficient in inducing hyphal de-
dia containing serum and increasing the growth velopment. These mutants were also unable to es-
temperature to 37 ◦ C. These conditions can trig- cape macrophage engulfment, while the wild-type
ger an almost quantitative shift of the yeast cells cells could trigger hyphal development and exit the
to hyphal cells. However, many other conditions macrophage (Lo et al. 1997). Transcription profil-
can also generate an efficient shift from the yeast ing was done to test the hypothesis that loss of
growth pattern to the hyphal mode. These include the transcription factors would disrupt the normal
growth in Lee’s medium at 37 ◦ C, addition of N- expression of hyphal-specific genes. The cph1 efg1
acetylglucosamine, and changes in the pH of the double mutant strain was unable to activate the nor-
growth medium. Many genes have been identified mal pattern of gene expression in response to heat
that are characteristically expressed as part of the and serum; the profile that was detected suggested
transition. These include genes encoding a number that the cells were able to recognize the tempera-
of cell wall proteins, such as Ece1p and Hwp1p, as ture increase but were insensitive to the presence
well as enzymes such as the secreted aspartyl proof serum. Very similar results were observed for
teases encoded by the SAP4, 5 and 6 genes. The use the single efg1 mutant, suggesting that the bulk of
of DNA microarrays has permitted the comprehen- the signaling passes through Efg1p (Nantel et al.
sive analysis of the transcription profile of cells un- 2002). Thus, the morphological consequences of
dergoing the yeast-to-hyphal transition in response the transcription factor mutations are reflected in
to serum (Nantel et al. 2002). This work, and addi- the changes in the transcription profiles of the cells.
tional hybridizations performed in our laboratory The recently identified Efh1p transcription fac-
over the last 2 years, has increased to approximately tor is in the same family as Efg1p (Doedt et al. 2004).
400 the numbers of genes whose transcripts show Deletion of EFH1 did not have a dramatic effect on
a statistically significant change in abundance dur- cellular morphology on its own, but caused hyper-
ing this transition. The most significant of these filamentation in embedded conditions when cou-
modulated transcripts are shown in Table 8.1, and pled to deletion of EFG1. Transcription profiling of
a complete list is available upon request. As evi- the disruption strains were consistent with this ob-
denced in Table 8.2, there are significant differences served relationship. Few genes were influenced by
between the functions of the gene products whose deletion of EFH1 under normal yeast growth con-
transcript are either upregulated or downregulated ditions, while many were either up- or downregu-
during the transition. We have noted that the list lated by loss of EFG1. Many genes were identified
of upregulated transcript shows a much greater re- that were misregulated in the double mutant but
producibility between experimenters. One possible were properly controlled in either single mutant
explanation may be that many of the gene products (Doedt et al. 2004). Transcription profiling has also
in the downregulated list are implicated in growth been used to examine the consequences of Efg1p
rate and metabolism. and Efh1p overexpression. Efg1p acted mainly as
Classical genetic studies have allow us to iden- a repressor, downregulating over 50 genes, but did
tify genes that are functionally required for the activate the expression of several genes, many of
transition from the yeast to the hyphal cell. Several which are also activated during serum-induced hy-
of these genes encode transcription factors, while phal development. Efh1p acted primarily as an ac-
Table. 8.1. The 100 most significantly modulated transcripts observed during the yeast-to-hyphae transition induced by
serum and high temperaturea
Array spot identifier Gene orf19 Fold p-value Product

number change
Upregulated transcripts
Contig4-2242_0004 ECE1 orf19.3374 42.9 0.0000 Secreted cell elongation protein
Contig4-2245_0009 orf19.7304 3.2 0.0000 Hypothetical protein
Contig4-2413_0004 orf.6705 2.1 0.0001 sec7 domain unknown protein
Contig4-2830_0009 SAP5 orf19.5585 11.2 0.0001 Secreted aspartyl proteinase 5
Contig4-2436_0002 HWP1 orf.1321 9.2 0.0001 Hyphal wall protein
Contig4-2682_0003 SNZ1 orf.2947 2.1 0.0001 Stationary-phase protein
Contig4-3065_0005 SOD5 orf.2060 7.2 0.0003 Copper–zinc superoxide dismutase
Contig4-2056_0002 IHD1 orf.5760 5.2 0.0006 Induced in hyphal development; membrane protein
Contig4-2558_0011 SMA2 orf.5644 1.4 0.0007 Conserved protein; spore cell wall biogenesis
Contig4-2734_0005 SAP6 orf.5542 4.1 0.0007 Candidapepsin 6 precursor
Contig4-2876_0008 CHA2 orf.1996 3.1 0.0008 Catabolic serine/threonine dehydratase
Contig4-2999_0011 orf.4924 1.2 0.0011 Splicing factor 3b
Contig4-3048_0002 orf.7531 2.0 0.0011 Conserved hypothetical protein
Contig4-2584_0004 CTA2 orf.631 1.4 0.0012 Transcription factor
Contig4-2187_0001 PHR1 orf.3829 2.7 0.0017 pH-regulated GPI-anchored membrane protein
that is required for morphogenesis
Contig4-2714_0002 orf.815 2.2 0.0017 DOCK180 protein
Contig4-3060_0001 PHR1 orf.3829 3.3 0.0017 pH-regulated GPI-anchored membrane protein
that is required for morphogenesis
Contig4-2790_0004 orf.6604 1.4 0.0021 Hypothetical protein
Contig4-2038_0003 SAP4 orf.5716 6.6 0.0022 Candidapepsin 4 precursor
Contig4-2567_0004 orf.4182 1.3 0.0022 Weak similarity to 3-oxoacyl-[acyl-carrier-protein]
reductase
Contig4-3059_0006 HGT2 orf.3668 4.6 0.0027 Hexose transporter
Contig4-2831_0003 GRP4 orf.3150 1.8 0.0029 Similar to plant dihydroflavonol 4-reductase
Contig4-3108_0039 orf.7602 1.3 0.0031 Activator of HSP90 ATPase
Contig4-1814_0003 PFY1 orf.5076 1.5 0.0033 Profilin
Contig4-2034_0002 ALR1 orf.1607 1.6 0.0033 Putative divalent cation transporter
Contig4-2810_0012 IHD2 orf.6021 1.8 0.0033 Induced in hyphal development
Contig4-2461_0005 CBR1 orf.1801 1.3 0.0040 Cytochrome-b5 reductase
Contig4-2578_0004 IRO1 orf.1715 1.8 0.0040 Transcription factor
Contig4-2876_0017 PRY4 orf.6202 2.9 0.0044 Pathogenesis-related protein, repressed by TUP1 protein 4
Contig4-2859_0007 MCM6 orf.2611 1.2 0.0045 Involved in replication
Contig4-2768_0006 RBT12 orf.3384 3.5 0.0050 Repressed by TUP1 protein 1
Contig4-2948_0008 RIS1 orf.5675 1.2 0.0050 SNF2 family DNA-dependent ATPase
Contig4-2779_0003 SPT14 orf.487 1.2 0.0053 N-acetylglucosaminyl-phosphatidylinositol (GPI)
biosynthetic protein
Contig4-2911_0012 PRC2 orf.4135 1.6 0.0053 Carboxypeptidase Y precursor
Contig4-2972_0005 TUB2 orf.6034 1.6 0.0053 beta-tubulin
Contig4-2696_0012 EXG1 orf.2952 1.5 0.0053 exo-1,3-beta-glucanase
Contig4-2815_0003 orf.1105.2 1.8 0.0053 Hypothetical protein
Contig4-2151_0007 orf.4749 4.0 0.0059 Hypothetical membrane protein
Contig4-2189_0002 APR1 orf.1891 1.8 0.0059 Vacuolar aspartic proteinase precursor
Contig4-3053_0003 MOB2 orf.6044 1.3 0.0059 Protein kinase activator involved in cell cycle
and shape regulation
Contig4-2283_0006 orf.3889 1.4 0.0059 Component of the COPII-coated vesicles
Contig4-2801_0007 MSB2 orf.1490 1.5 0.0059 Integral membrane protein; cell surface flocculin
Contig4-3102_0012 NoHit 1.4 0.0062 No homolog in orf
Contig4-3002_0007 RNH1 orf.5564 1.5 0.0065 Ribonuclease H, exon 2
Contig4-3058_0013 RHO2 orf.2204.2 1.5 0.0065 GTP-binding protein of the rho subfamily
Contig4-2963_0002 NoHit 1.7 0.0065 No homolog in orf
Array spot identifier Gene orf19 Fold p-value Product

number change
Contig4-2537_0005 SEC17 orf.2518 1.2 0.0066 Transport vesicle fusion protein
Downregulated transcripts
Contig4-2520_0008 HYP2 orf.3426 1.8 0.0000 Translation initiation factor eIF5A.1
Contig4-2843_0005 RCK2 orf.2268 1.4 0.000 Ca2+ /calmodulin-dependent serine/threonine
protein kinase
Contig4-2262_0002 MRPS5 orf.989 1.3 0.0007 Mitochondrial ribosomal protein S5
Contig4-3020_0010 RPL13 orf.2994 1.6 0.0007 Ribosomal protein
Contig4-2882_0010 APE2 orf.5197 1.4 0.0008 Alanine/arginine aminopeptidase
Contig4-2864_0006 ATP5 orf.5419 1.3 0.0011 F1 F0 -ATPase subunit
Contig4-3035_0011 HMO1 orf.6645 1.6 0.0017 High-mobility group-like protein
Contig4-2154_0002 ADK1 orf.683 1.5 0.0017 Adenylate kinase
Contig4-2826_0005 CBF1 orf.2876 2.0 0.0021 Centromere-binding kinetochore protein
Contig4-2921_0012 ALD5 orf.5806 1.8 0.0021 Aldehyde dehydrogenase
Contig4-2469_0001 RHD1 orf.54 3.1 0.0022 Repressed in hyphal development;
family of conserved protein of unknown function
Contig4-2944_0009 MRP20 orf.3350 1.7 0.0022 Mitochondrial ribosomal protein
Contig4-2970_0009 orf.2582 1.4 0.0023 Highly conserved hypothetical protein
Contig4-2812_0004 GDH3 orf.4716 1.9 0.0024 NADP-glutamate dehydrogenase
Contig4-1966_0003 ADK12 orf.3391 1.7 0.0029 Adenylate kinase
Contig4-3022_0004 STM1 orf.5943.1 1.7 0.0029 Purine motif triplex-binding protein;
specific affinity for guanine-rich quadruplex nucleic acids
Contig4-2973_0009 RPS0A orf.6975 1.5 0.0030 Ribosomal protein S0A
Contig4-3094_0036 COR1 orf.4016 1.4 0.0031 Part of ubiquinol cyt-c reductase complex
Contig4-3104_0059 CLC1 orf.4594 2.0 0.0031 Clathrin light chain
Contig4-2977_0010 orf.6062.3 1.4 0.0031 Conserved hypothetical protein
Contig4-3037_0004 STP3 orf.5917 1.6 0.0033 Zinc finger protein involved in pre-tRNA splicing
Contig4-2863_0003 VSP24 orf.2031 1.4 0.0037 Endosomal Vps protein complex subunit involved
in secretion
Contig4-2426_0003 FCY1 orf.4195.1 1.7 0.0044 Cytosine deaminase
Contig4-2933_0007 CSP37 orf.2531 1.9 0.0044 Cell surface protein
Contig4-3006_0011 orf.6748 1.3 0.0044 Translation initiation factor
Contig4-3083_0027 orf.2668 2.3 0.0044 gag-pol polyprotein
Contig4-1953_0001 TOM22 orf.3696 1.6 0.0046 Mitochondrial import receptor protein
Contig4-3087_0001 RPL35 orf.5964.2 2.2 0.0048 Ribosomal protein L35A
Contig4-2863_0021 MSF1 orf.2039 1.2 0.0050 Phenylalanyl-tRNA synthetase alpha subunit,
mitochondrial
Contig4-3099_0016 orf.897 1.2 0.0052 Conserved hypothetical protein; coiled coils
Contig4-2521_0007 orf.4517 1.3 0.0053 Conserved hypothetical membrane protein
Contig4-2748_0005 RPS23 orf.6253 1.4 0.0053 Ribosomal protein S23A
Contig4-2763_0003 RPL5 orf.6541 1.5 0.0053 Ribosomal protein L5
Contig4-2813_0019 CVB1 orf.1970 1.3 0.0053 Complements S. cerevisiae vacuole biogenesis mutant,
unpublished data
Contig4-2613_0002 COX20 orf.6461 1.3 0.0054 Cytochrome oxidase
Contig4-2997_0025 DYS1 orf.1626 1.3 0.0054 Deoxyhypusine synthase
Contig4-2732_0018 RPL24A orf.3789 1.5 0.0062 Ribosomal protein L24A (rp29) (YL21) (L30A)
Contig4-2970_0018 OPT9 orf.2584 3.5 0.0062 Oligopeptide transporter specific for tetra-
and pentapeptides; possible pseudogene
Contig4-2904_0001 NAM9 orf.498 1.3 0.0062 Mitochondrial S4 ribosomal protein
Contig4-3097_0058 MRP7 orf.7203 1.4 0.0065 Mitochondrial large ribosomal subunit
Contig4-2649_0003 MRPL32 orf.549 1.6 0.0067 Mitochondrial ribosomal large subunit protein
Contig4-2622_0004 SFU1 orf.4869 1.2 0.0068 GATA-type transcriptional activator of
nitrogen-regulated genes
Contig4-3001_0005 ATP3 orf.3223 1.2 0.0068 ATP synthase gamma subunit
a These were identified from a Student t-test analysis (multiple testing correction: Benjamini and Hochberg false discovery
rate) of the combined results of 14 independent RNA preparations and 20 microarray hybridizations prepared by Nantel
et al. (2002), Lee et al. (2004) and Oberholzer (unpublished data). Transcripts are sorted by statistical significance
Table. 8.2. Selected GO terms associated with gene products tivator, upregulating over 50 genes, and downreg-
whose transcripts are significantly modulated during the
yeast-to-hyphae transition
ulating less than 30. Only about 10% of the genes
were co-regulated by overproduction of the two
GO term # of genes transcription modulators (Doedt et al. 2004).
Upregulated transcripts (172 genes)
Other transcription factors play a role in hyphal
C:COPII-coated vesicle 4 development, and the global effects of some of these
C:bud neck 4 genes have been investigated by transcriptional
C:bud tip 2 profiling. Nrg1p and Tup1p are transcriptional reg-
C:cytoplasmic microtubule 5 ulators that play a role in hyphal development.
C:plasma membrane 18
Deletion of either TUP1 (Braun and Johnson 1997)
C:polar microtubule 4
C:spindle pole body 3 or NRG1 (Murad et al. 2001b) leads to activation of
F:1,3-beta-glucanosyltransferase 2 filamentous development and results in the consti-
F:histone deacetylase 4 tutive production of pseudohyphae under standard
F:Rho small monomeric GTPase 3 yeast growth conditions. The two mutants do not,
F:aspartic-type endopeptidase 5 however, create identical phenotypes; in response
F:structural constituent of cytoskeleton 5 to serum stimulation, the nrg1 mutant is still able to
F:thioredoxin peroxidase 6
form true hyphae, while the tup1 mutant remains
P:ER to Golgi transport 5 locked in the pseudohyphal growth pattern (Mu-
P:beta-1,6 glucan biosynthesis 4 rad et al. 2001b). Transcriptional profiling has been
P:bud growth 2
P:bud site selection 3 used to assess the consequences of these deletions.
P:cell wall organization and biogenesis 14 Partial genome arrays spotted on filters were used
P:establishment of cell polarity 6 to profile both the nrg1 mutant and the tup1 mutant
P:homologous chromosome segregation 5 (Murad et al. 2001a). Among the genes that were
P:inactivation of MAPK 4 derepressed in both mutants were hyphal-specific
P:nuclear migration 4
P:regulation of redox homeostasis 8
genes such as HYR1, HWP1 and ECE1. Overall,
P:small GTPase-mediated signal transduction 3 Tup1p and Nrg1p were needed for the proper ex-
pression of many genes, and together with Mig1p
Downregulated transcripts (228 genes)
C:19S proteasome regulatory particle 8 formed overlapping clusters of co-regulated genes.
C:actin cortical patch 2 This profiling work is consistent with Nrg1p/Tup1p
C:centromere 5 controlling the expression of genes involved in hy-
C:hydrogen-transporting ATP synthase, stator stalk 7 phal development, while Mig1p/Tup1p control as-
C:nucleus 16 pects of carbon metabolism. The profiling data also
C:ribosome 28
C:respiratory chain complex III 4 suggest that Mig1p and Nrg1p do more than act as
C:spindle pole body 2 co-regulators of the Tup1p repressor; removal of
F:RNA binding 5
either Nrg1p or Mig1p affects expression of genes
F:cytoskeletal adaptor 2 that are not affected by Tup1p depletion (Murad
F:general RNA polymerase II transcription factor 3 et al. 2001a).
F:hydrogen-transporting two-sector ATPase 7 Other cellular components have been im-
F:proteasome endopeptidase 8 plicated in the transition to hyphal growth. In
F:protein transporter 7 particular, the loss of components of the cAMP
F:structural constituent of ribosome 39
F:translation initiation factor 7 regulatory circuit affects hyphal development.
Mutation of CDC35, which encodes adenylyl
P:35S primary transcript processing 7
P:ATP synthesis coupled proton transport 7
cyclase, totally abolishes the ability to trigger
P:UDP-N-acetylglucosamine biosynthesis 2 hyphal development under all conditions tested
P:actin filament organization 2 (Rocha et al. 2001). Deletion of RAS1, a regulator of
P:aerobic respiration 8 adenylyl cyclase, blocks true hyphal development
P:establishment of cell polarity 3 in response to serum and temperature induction,
P:mRNA splicing 5 but the mutant cells still can form pseudohyphae
P:mitochondrial translocation 7
P:protein biosynthesis 62 (Feng et al. 1999; Leberer et al. 2001). Mutations in
P:regulation of transcription from Pol II promoter 4 the cAMP-dependent protein kinase subunits also
P:secretory pathway 4 influence hyphal development (Bockmuhl et al.
P:ubiquitin-dependent protein catabolism 8 2001; Cloutier et al. 2003). Transcription profiling
P:vesicle-mediated transport 3 has been applied to investigate the consequences
of deletion of three of the elements of the cAMP both serum and high temperature, which resulted
regulon – CDC35, encoding the adenylyl cyclase instead in a mild activation of their heat shock re-
(Harcus et al. 2004), as well as RAS1 and EFG1 sponse (Lee et al. 2004).
(Nantel et al. 2002; Harcus et al. 2004). In addition, In addition to transcription profiling, large-
profiling has investigated the function of an Efg1p scale genomic screens have been applied to the
homolog, Efh1p (Doedt et al. 2004). The loss of question of hyphal development. A modification of
either CDC35 or RAS1 has related consequences the transposon mutagenesis strategy that has been
on gene expression in C. albicans. In addition very successful in S. cerevisiae has been used to
to disrupting the transcriptional response to create disruptions that occur on average once every
hyphal development triggered by serum and high 2.5 kb in the C. albicans genome (Uhl et al. 2003).
temperature, the cdc35 and ras1 mutants affect Screening for insertions that cause a modification
the expression of many genes involved in general in the yeast–hyphal transition identified over 140
metabolism and stress responses. The loss of genes that affect the process when one copy of the
adenylyl cyclase function changes the expression gene is disrupted. These mutations affected a num-
of hundreds of genes; loss of RAS1 influences ber of known genes implicated in hyphal develop-
expression levels of a subset of those genes, and ment, such as TUP1, NRG1, CBK1 and CZF1, and
little else. The transcription profiles are very also identified a large number of genes that were
different for cells deleted for the EFG1 gene. The unique to C. albicans. Therefore, both expression
genes requiring EFG1 overlap with those requiring level measurements and functional studies are be-
CDC35 and RAS1 for only the hyphal-induced ing used on a genome-wide level to provide insights
set; other cellular processes have the EFG1 and into this medically important cellular transition in
CDC35/RAS1 genes acting in distinct ways. These C. albicans.
differences in gene expression profiles are reflected
in the physiology of the cells; for example, ras1
and cdc35 mutants are more resistant to chemicals B. Biofilms
or enzymes that attack the cell wall, while efg1
mutants are more sensitive. Finally, this work A second morphological state that has been in-
demonstrated that significant regulatory connec- vestigated with microarrays is that of the biofilm.
tions remain to be identified, especially those Biofilms are structurally complex collections of
that regulate cAMP-dependent changes in gene cells of various morphologies that are associated
expression that are independent of Efg1p (Harcus with solid surfaces, and are significantly more re-
et al. 2004). sistant to antifungal drugs, thus providing a source
Another enzyme with an important role in hy- for re-infections. García-Sánchez et al. (2004) used
phal signaling is the Sit4p phosphatase. Morpho- macro- and microarrays to study the transcrip-
logical and virulence studies of homozygous sit4 tional profiles in biofilms that arise under different
mutants have indicated that this type 2A protein culture conditions. The Gcn4p transcription factor
phosphatase is required for proliferation, virulence was shown to be necessary for biofilm formation, as
and hyphal growth. Although serum-treated sit4 it is required for the activation of numerous genes
null cells have a pseudohyphal phenotype and fail involved in protein synthesis. One possible role for
to elongate into true hyphae, transcriptional pro- excess protein synthesis lies in the production of
filing has further indicated that the effects of the the extracellular matrix that provides the structural
sit4 mutation are quite distinct from what was ob- basis of the biofilm. Furthermore, it was suggested
served with other hyphal-impaired mutants (Lee that activation of the genes involved in the sulfur
et al. 2004). Most of the canonical hyphal-specific amino acid biosynthesis/salvage pathway may be
genes, such as ECE1, RBT1, SAP4 and HWP1, show involved in cell wall rearrangement and the pro-
a normal increase in transcript abundance. The duction of quorum-sensing molecules. Compari-
pseudohyphal phenotype might be attributed to son with biofilm composed of yeast-only efg1/cph1
problems in cell wall reorganization, as the sit4 mutants allowed Garcia-Sanchez et al. (2004) to
mutants fail to upregulate transcripts encoding the discriminate between genes that are involved with
exo-beta-glucanase Xog1p and the endo-1-3- beta- biofilm formation and those implicated in morpho-
glucanase Acf1p. Furthermore, as was observed logical switching. Although transcripts for CDR1,
with the efg1 and efg1/cph1 null mutants, sit4 mu- ERG25 and ERG16 were more abundant under cer-
tants fail to properly respond to the presence of tain biofilm-inducing conditions, the correlations
between these higher-order structures and antifun- both mating pheromone and the yeast- to-hyphal
gal resistance remains unclear. To further elucidate transition.
this important question, it might be interesting to Among the pheromone-induced gene products
use transcriptional profiling to determine whether that are also hyphal specific is the cell wall protein
cells in these various morphological states respond Hwp1p. Hwp1p is a substrate for mammalian
differently to azole treatment. transglutaminase, and this characteristic of the
protein is exploited by C. albicans cells to facilitate
their adherence to epithelial cells. Intriguingly,
C. Mating Projections in the mating process Hwp1p is localized to the
conjugation tips of the a-mating cells responding
One of the initial consequences of the sequencing of to α-factor (Daniels et al. 2003). This establishes
the C. albicans genome was the identification of the that the mating projection is a discrete cellular
MTL loci. These regions encode proteins similar to state; it is not just an elongated bud because Hwp1p
the transcription factors that control the definition is not expressed in buds. HWP1 is expressed only
of cell type in S. cerevisiae. Further investigations in mating-type homozygous cells responding to
established that the typical C. albicans strain was α-factor, a direct test of the mating factor from
heterozygous for the MTL locus, a genotype that a cells is not possible because this compound has
in S. cerevisiae would preclude mating (Soll 2004). not been identified.
Therefore, researchers tested whether strains ma- Overall then, the mating process creates novel
nipulated to be homozygous for each of the MTL cellular morphologies. Cells responding to mating
loci could be induced to mate with each other. pheromones can initiate a cellular elongation that is
Inefficient mating was detected (Hull et al. 2000; superficially like a germ tube, but functionally dis-
Magee and Magee 2000), and the subsequent recog- tinct. Transcription profiling has identified a series
nition that this mating took place significantly only of genes that are induced specifically during this
when the two mating partners were in the opaque process, and the involvement of these proteins in
state (see next section) allowed Miller and Johnson the morphological transition is now open to anal-
(2002) to make a detailed characterization of the ysis.
mating process.
Analysis of the genome sequence allowed
researches to identify a putative gene encoding the D. White–Opaque Transition
MTLα-specific mating pheromone (Bennett et al.
2003; Lockhart et al. 2003; Panwar et al. 2003). The white–opaque transition leads to dramatic
A peptide based on the amino acid sequence of the changes in both the shape of the individual cells
predicted pheromone has been synthesized, and and the shape of the colony. Initial work defining
several groups have examined the consequences the characteristics of the white and opaque cells
of pheromone treatment on cell cycle progression identified genes that were expressed uniquely in
and cellular morphology. Depending on the strain one cell type or the other. For example, SAP1
and the assay conditions, pheromone treatment (White and Agabian 1995), CDR3 (Balan et al.
can cause significant cell cycle arrest (Panwar 1997) and OP4 (Morrow et al. 1993) were found
et al. 2003). Perhaps the most striking phenotypic to be opaque specific, while WHI1 and EFG1 were
modulation is that of the formation of mating white specific (Sonneborn et al. 1999b). Recent
projections (Lockhart et al. 2003). Analysis of analysis of the two cell types using transcription
the expression of genes in response to mating profiling has made possible the identification of
pheromone confirms that the pheromone can the entire spectrum of white- and opaque-specific
modulate transcription patterns (Bennett et al. genes (Lan et al. 2002). These profiles establish
2003). Over 60 genes are induced by the mating that over 150 genes are more expressed in the
pheromone. This induction is slow relative to white state, and over 200 genes are more expressed
the induction of pheromone-responsive genes in the opaque state. There are major differences
in S. cerevisiae, although several of these genes in the expression of metabolic genes in the
represent the homologs of induced yeast genes. two cell types, consistent with the white cells
However, there are many pheromone-induced exhibiting a fermentative profile, and opaque cells
genes that are not found in S. cerevisiae, and there an oxidative metabolic state. These distinctions,
is also a group of transcripts that are induced by as well as the fact the fatty acid β-oxidation
pathway is upregulated in opaque cells, can help E. Chlamydospore Formation

explain the observation that the opaque cells
are more effective pathogens on skin, while the Chlamydospores are a unique cellular form of C. al-
white cells are more pathogenic during systemic bicans that consist of large, thick-walled cells that
infections. form on the ends of suspensor cells. These suspen-
A second cellular process that is different be- sor cells are typically attached to hyphae. This cell
tween the white and opaque cells is mating. Mat- type is induced under oxygen-limited, nutrient-
ing is exclusive to opaque cells, and genes such poor conditions; growth under glass cover slips
as those encoding the pheromone receptor STE3 on cornmeal agar, or growth directly embedded
and the mating factor α-factor are upregulated in agar can be used to induce chlamydospore for-
in opaque cells, relative to white cells (Lan et al. mation. Both Efg1p and Hog1p are required for
2002). The ability to switch to the opaque phase chlamydospore formation, but not for hyphal for-
is limited to cells that are homozygous for the mation, in embedded conditions (Sonneborn et al.
mating type; this observation clarified why only 1999a; Alonso-Monge et al. 2003).
a minority of strains are capable of undergoing the A collection of over 200 defined homozygosed
white–opaque switch. This implies that the mating- insertions was created by the UAU insertion cas-
type heterozygous state represses the opaque mor- sette strategy (Davis et al. 2002), and screened for
phological phase (Miller and Johnson 2002), and defects in chlamydospore formation. Three genes
therefore cells require two infrequent events, first (IWS2, SCH9 and SUV3) are essential for chlamy-
a switch to mating-type homozygosity and then dospore formation, while loss of three other genes
a switch to the opaque state, to create a mating (RIM101, RIM13 and MDS3) delays their formation
competent state. Because the formation of a sin- (Nobile et al. 2003). This large-scale approach to
gle mating competent cell appears infrequent, the functional analysis provides a connection between
chance of generating two cells of opposite mat- the pH regulatory pathway that influences hyphal
ing competencies in the same vicinity is extremely development, and chlamydospore formation. This
small. This would make mating a highly improb- work shows how systematic analysis of disruption
able event in nature, and as such would make it constructs can provide new insights into morpho-
apparently unreasonable for C. albicans to main- logical plasticity in C. albicans.
tain this capacity.
Other regulatory circuits may interact with the
white–opaque switching machinery to promote IV. Conclusions
mating competence at a greater frequency. Inac-
tivation of Efg1p transcription factor creates cells After only a few years, genomic technologies have
that are morphologically similar to opaque cells, yielded an abundance of novel concepts and may
and EFG1 is one of the genes that is considered already be changing the ways by which researchers
white specific. In addition, the metabolic state of are classifying morphological states. In addition
opaque cells and efg1 mutant cells are both shifted to the simple identification of modulated genes,
to a non-fermentative, oxidative mode (Lan et al. the detailed transcriptional profiles that can be
2002; Doedt et al. 2004). Therefore, it is possible obtained with DNA microarrays facilitate a more
that some combination of environmental cues precise classification of the various types of mor-
may facilitate the creation of a mating competent phologies that are observed in response to envi-
state at frequencies higher than that observed in ronmental and genetic cues. This is especially use-
laboratory conditions. Recent evidence shows that ful, as cells with a similar morphology may in fact
HBR1, a hemoglobin response gene, is implicated arise from completely different mechanisms. For
in the mating process (Pendrak et al. 2004). example, the elongated cell phenotype that is ob-
Hbr1p functions as a repressor of white–opaque served when efg1 or sit4 mutants are treated with
switching; haplo-insufficiency at HBR1 results in serum and high temperature may appear similar
white–opaque switching without the need to create under a microscope but have, in fact, very differ-
homozygosity at MTL. Therefore, environmental ent transcriptional profiles, especially in relation
regulation of gene expression may promote to the canonical hyphal-specific genes. Neverthe-
the formation of mating competent cells that less, correlations between these profiles and the
have not undergone the genomic rearrangements responses of Candida albicans to environmental
characteristic of experimentally controlled mating. stresses, such as heat shock, indicate that both fac-
tor are involved in coordinating signals from two dent protein kinase catalytic subunit are involved
external stimuli (serum and high temperature) into in the control of dimorphism in the human fungal
a single morphological response. Similar correla- pathogen Candida albicans. Fungal Genet Biol 38:133–
141
tions between different datasets are becoming more Cowen LE, Nantel A, Whiteway MS, Thomas DY, Tessier DC,
common and help us to understand the intercon- Kohn LM, Anderson JB (2002) Population genomics of
nections between the various signaling cascades drug resistance in Candida albicans. Proc Natl Acad
that modulate morphogenesis. Furthermore, pro- Sci USA 99:9284–9289
Daniels KJ, Lockhart SR, Staab JF, Sundstrom P, Soll DR
filing of the components of a signal transduction (2003) The adhesin Hwp1 and the first daughter cell lo-
chain will help us determine whether novel com- calize to the a/a portion of the conjugation bridge dur-
ponents remain to be discovered, as was the case ing Candida albicans mating. Mol Biol Cell 14:4920–
with our recent study of cAMP signaling (Harcus 4930
et al. 2004). Finally, novel screening strategies are Davis DA, Bruno VM, Loza L, Filler SG, Mitchell AP (2002)
Candida albicans Mds3p, a conserved regulator of pH
taking advantage of our knowledge of the Can- responses and virulence identified through insertional
dida genome to identify novel modulators. The mutagenesis. Genetics 162:1573–1581
genes encoding these modulators can then be mu- De Backer MD, Ilyina T, Ma XJ, Vandoninck S, Luyten WH,
tated and the consequences of these mutations on Vanden Bossche H (2001) Genomic profiling of the
response of Candida albicans to itraconazole treat-
transcriptional profiles compared to available data. ment using a DNA microarray. Antimicrob Agents
Thus, the increasing amount and quality of these Chemother 45:1660–1670
large datasets, coupled with advances in the qual- DeRisi JL, Iyer VR, Brown PO (1997) Exploring the
ity of the Candida genome annotation, is bound metabolic and genetic control of gene expression on
to produce additional discoveries and, hopefully, a genomic scale. Science 278:680–686
Doedt T, Krishnamurthy S, Bockmuhl DP, Tebarth B, Stem-
unexpected but informative correlations. pel C, Russell CL, Brown AJ, Ernst JF (2004) APSES
proteins regulate morphogenesis and metabolism in
Acknowledgements. This work was supported by the NRC Candida albicans. Mol Biol Cell 15:3167–3180
Genomics and Health Research Initiative, as well as by CIHR Feng Q, Summers E, Guo B, Fink G (1999) Ras signaling is
grant number HOP-67260 (to A. Nantel) and CIHR grant required for serum-induced hyphal differentiation in
MOP42516 (to M. Whiteway). We would like to thank all Candida albicans. J Bacteriol 181:6339–6346
members of the BRI Genetics group for discussions. This is Garcia-Sanchez S, Aubert S, Iraqui I, Janbon G, Ghigo JM,
NRC publication # 46234. d’Enfert C (2004) Candida albicans biofilms: a devel-
opmental state associated with specific and stable gene
expression patterns. Eukaryot Cell 3:536–545
References Harcus D, Nantel A, Marcil A, Rigby T, Whiteway M (2004)
Transcription Profiling of cyclic AMP signaling in Can-
dida albicans. Mol Biol Cell 15:4490–4499
Alonso-Monge R, Navarro-Garcia F, Roman E, Negredo AI, Hughes TR, Mao M, Jones AR, Burchard J, Marton MJ, Shan-
Eisman B, Nombela C, Pla J (2003) The Hog1 mitogen- non KW, Lefkowitz SM, Ziman M, Schelter JM, Meyer
activated protein kinase is essential in the oxidative MR et al. (2001) Expression profiling using microar-
stress response and chlamydospore formation in Can- rays fabricated by an ink-jet oligonucleotide synthe-
dida albicans. Eukaryot Cell 2:351–361 sizer. Nat Biotechnol 19:342–347
Balan I, Alarco AM, Raymond M (1997) The Candida al- Hull CM, Raisner RM, Johnson AD (2000) Evidence for mat-
bicans CDR3 gene codes for an opaque-phase ABC ing of the “asexual” yeast Candida albicans in a mam-
transporter. J Bacteriol 179:7210–7218 malian host. Science 289:307–310
Bennett RJ, Uhl MA, Miller MG, Johnson AD (2003) Iden- Jones T, Federspiel NA, Chibana H, Dungan J, Kalman S,
tification and characterization of a Candida albicans Magee BB, Newport G, Thorstenson YR, Agabian N,
mating pheromone. Mol Cell Biol 23:8189–8201 Magee PT et al. (2004) The diploid genome sequence
Bockmuhl DP, Krishnamurthy S, Gerads M, Sonneborn A, of Candida albicans. Proc Natl Acad Sci USA 101:7329–
Ernst JF (2001) Distinct and redundant roles of the 7334
two protein kinase A isoforms Tpk1p and Tpk2p in Kumamoto CA (2002) Candida biofilms. Curr Opin Micro-
morphogenesis and growth of Candida albicans. Mol biol 5:608–611
Microbiol 42:1243–1257 Kvaal CA, Srikantha T, Soll DR (1997) Misexpression of the
Braun BR, Johnson AD (1997) Control of filament formation white-phase-specific gene WH11 in the opaque phase
in Candida albicans by the transcriptional repressor of Candida albicans affects switching and virulence.
TUP1. Science 277:105–109 Infect Immun 65:4468–4475
Bruno VM, Mitchell AP (2004) Large-scale gene function Lan CY, Newport G, Murillo LA, Jones T, Scherer S,
analysis in Candida albicans. Trends Microbiol Davis RW, Agabian N (2002) Metabolic specialization
12:157–161 associated with phenotypic switching in Candida
Cloutier M, Castilla R, Bolduc N, Zelada A, Martineau P, albicans. Proc Natl Acad Sci USA 99:14907–14912
Bouillon M, Magee BB, Passeron S, Giasson L, Can- Lane S, Birse C, Zhou S, Matson R, Liu H (2001) DNA ar-
tore ML (2003) The two isoforms of the cAMP-depen- ray studies demonstrate convergent regulation of vir-
ulence factors by Cph1, Cph2, and Efg1 in Candida Panwar SL, Legrand M, Dignard D, Whiteway M, Magee PT
albicans. J Biol Chem 276:48988–48996 (2003) MFalpha1, the gene encoding the alpha mating
Leberer E, Harcus D, Dignard D, Ushinsky S, Thomas DY, pheromone of Candida albicans. Eukaryot Cell
Schroppel K (2001) Ras links cellular morphogenesis to 2:1350–1360
virulence by regulation of the MAP kinase and cAMP Pendrak ML, Yan SS, Roberts DD (2004) Hemoglobin reg-
signalling pathways in the pathogenic fungus Candida ulates expression of an activator of mating-type locus
albicans. Mol Microbiol 42:673–687 alpha genes in Candida albicans. Eukaryot Cell 3:764–
Lee CM, Nantel A, Jiang L, Whiteway M, Shen SH (2004) 775
The serine/threonine protein phosphatase SIT4 mod- Ramage G, Wickes BL, Lopez-Ribot JL (2001) Biofilms of
ulates yeast-to-hypha morphogenesis and virulence in Candida albicans and their associated resistance to
Candida albicans. Mol Microbiol 51:691–709 antifungal agents. Am Clin Lab 20:42–44
Lo HJ, Kohler JR, DiDomenico B, Loebenberg D, Caccia- Rocha CRC, Schroppel K, Harcus D, Marcil A, Dignard D,
puoti A, Fink GR (1997) Nonfilamentous C. albicans Taylor BN, Thomas DY, Whiteway M, Leberer E (2001)
mutants are avirulent. Cell 90:939–949 Signalling through adenylyl cyclase is essential for hy-
Lockhart SR, Zhao R, Daniels KJ, Soll DR (2003) Alpha- phal growth and virulence in the pathogenic fungus
pheromone-induced “shmooing” and gene regulation Candida albicans. Mol Biol Cell 12:3631–3643
require white–opaque switching during Candida albi- Roemer T, Jiang B, Davison J, Ketela T, Veillette K, Breton A,
cans mating. Eukaryot Cell 2:847–855 Tandia F, Linteau A, Sillaots S, Marta C et al. (2003)
Magee BB, Magee PT (2000) Induction of mating in Can- Large-scale essential gene identification in Candida
dida albicans by construction of MTLa and MTLalpha albicans and applications to antifungal drug discovery.
strains. Science 289:310–313 Mol Microbiol 50:167–181
Miller MG, Johnson AD (2002) White-opaque switching in Rogers PD, Barker KS (2002) Evaluation of differential gene
Candida albicans is controlled by mating-type locus expression in fluconazole-susceptible and – resistant
homeodomain proteins and allows efficient mating. isolates of Candida albicans by cDNA microarray anal-
Cell 110:293–302 ysis. Antimicrob Agents Chemother 46:3412–3417
Mnaimneh S, Davierwala AP, Haynes J, Moffat J, Peng WT, Soll DR (1992) High-frequency switching in Candida albi-
Zhang W, Yang X, Pootoolal J, Chua G, Lopez A et al. cans. Clin Microbiol Rev 5:183–203
(2004) Exploration of essential gene functions via Soll DR (2004) Mating-type locus homozygosis, phenotypic
titratable promoter alleles. Cell 118:31–44 switching and mating: a unique sequence of dependen-
Morrow B, Srikantha T, Anderson J, Soll DR (1993) Coor- cies in Candida albicans. Bioessays 26:10–20
dinate regulation of two opaque-phase-specific genes Sonneborn A, Bockmuhl DP, Ernst JF (1999a) Chlamy-
during white–opaque switching in Candida albicans. dospore formation in Candida albicans requires
Infect Immun 61:1823–1828 the Efg1p morphogenetic regulator. Infect Immun
Murad AM, d’Enfert C, Gaillardin C, Tournu H, Tekaia F, 67:5514–5517
Talibi D, Marechal D, Marchais V, Cottin J, Brown AJ Sonneborn A, Tebarth B, Ernst JF (1999b) Control of
(2001a) Transcript profiling in Candida albicans re- white–opaque phenotypic switching in Candida
veals new cellular functions for the transcriptional re- albicans by the Efg1p morphogenetic regulator. Infect
pressors CaTup1, CaMig1 and CaNrg1. Mol Microbiol Immun 67:4655–4660
42:981–993 Sudbery P, Gow N, Berman J (2004) The distinct mor-
Murad AM, Leng P, Straffon M, Wishart J, Macaskill S, Mac- phogenic states of Candida albicans. Trends Microbiol
Callum D, Schnell N, Talibi D, Marechal D, Tekaia F 12:317–324
et al. (2001b) NRG1 represses yeast-hypha morpho- Uhl MA, Biery M, Craig N, Johnson AD (2003)
genesis and hypha-specific gene expression in Candida Haploinsufficiency-based large-scale forward ge-
albicans. EMBO J 20:4742–4752 netic analysis of filamentous growth in the diploid
Nantel A, Dignard D, Bachewich C, Harcus D, Marcil A, human fungal pathogen C. albicans. EMBO J
Bouin A-P, Sensen CW, Hogues H, Van het Hoog M, 22:2668–2678
Gordon P et al. (2002) Transcription profiling of Can- White TC, Agabian N (1995) Candida albicans secreted as-
dida albicans cells undergoing the yeast to hyphal tran- partyl proteinases: isoenzyme pattern is determined
sition. Mol Biol Cell 13:3452–3465 by cell type, and levels are determined by environ-
Nobile CJ, Bruno VM, Richard ML, Davis DA, Mitchell AP mental factors. J Bacteriol 177:5215–5221
(2003) Genetic control of chlamydospore formation in Whiteway M, Oberholzer U (2004) Candida morphogenesis
Candida albicans. Microbiology 149:3629–3637 and host–pathogen interactions. Curr Opin Microbiol
7:350–357
Fungal Pathogenicity
9 Postgenomic Approaches to Analyse Candida albicans Pathogenicity
C.A. Munro1 , C. Fradin2 , O. Bader2 , B. Hube2
CONTENTS antibacterial treatment or when physical barriers

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 163 of the human body are damaged due to injury or
II. The Hunt for Pathogenicity-Related Genes surgical procedures (Odds 1988; Calderone 2002).
Using Comparative Genomics and Predictive C. albicans has probably co-evolved with hu-
Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . 164 man beings for millions of years and, through per-
A. Comparing the C. albicans Translated ORF
Set to the Outside World . . . . . . . . . . . . . 164 manent competition with the bacterial flora and
B. Genes Encoding Proteins continuous interactions with human cells, has be-
with Unknown Function . . . . . . . . . . . . . 166 come highly adapted to survival on epithelial sur-
III. Genes with Possible Virulence Functions . . . 167 faces (Odds et al. 2004). As a pathogen, C. albicans
IV. Gene Families . . . . . . . . . . . . . . . . . . . . . . . . 168
V. Role of Metabolism During Infection . . . . . . 169 can not only colonise a multiplicity of body sites
VI. Role of Morphology . . . . . . . . . . . . . . . . . . . . 171 but can also invade tissue, survive in the blood-
VII. Role of the Cell Surface During stream and cause life-threatening systemic infec-
Host/Pathogen Interactions . . . . . . . . . . . . . 171 tions when host defence mechanisms are weakened
VIII. Global Approaches to Identify Virulence (De Repentigny 2004). Therefore, C. albicans is able
Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
IX. Transcriptional Profiling During Infection . . 174 to exist and proliferate in radically changing envi-
X. Gene Expression During Host/Pathogen ronments with strong changes in oxygen and car-
Interaction: Examples . . . . . . . . . . . . . . . . . . 175 bon dioxide levels, pH, osmolarity, availability of
A. C. albicans Facing Epithelial and nutrients, and temperature (Hube 2004). Further-
Endothelial Barriers . . . . . . . . . . . . . . . . . 175
B. C. albicans Facing Phagocytes . . . . . . . . . 177 more, the fungus needs to counteract the attack of
XI. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 179 immune cells, amongst others by phagocytosis and
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 the secretion of antimicrobial agents. Such a high
degree of flexibility requires the expression of spe-
cial sets of attributes during commensal growth
Abbreviations: ALS, agglutinin-like sequence; and at each stage of infection in response to the
GPI, glycosylphosphatidylinositol; ORF, open specific demands of the many microenvironments
reading frame; RHE, reconstituted human ep- which the fungus encounters. In this context, it is
ithelium; SAP, secreted aspartyl proteinase genes; rather surprising that C. albicans, although highly
TCA, tricarboxylic acid flexible in source of nutrients and robust to rela-
tively harsh conditions, is rarely found outside the
body in the global environment (e.g. soil), in con-
trast to most other human pathogenic fungi such
as Histoplasma capsulatum, Cryptococcus neofor-
I. Introduction mans and Aspergillus fumigatus (Hube 2004).
Growth, morphology, metabolism, responses
The yeast Candida albicans can act as both to the environment and expression of attributes es-
a harmless commensal in healthy individuals and pecially required for pathogenicity during the dif-
an aggressive pathogen in immunocompromised ferent stages of infection are ultimately regulated by
patients, for example, when the normal microbial genes and their products, and adaptation to chang-
flora is removed or unbalanced by extensive ing environment occurs predominately on the tran-
1 School of Medical Sciences, Institute of Medical Sciences,
scriptional level. With the complete genome se-
Foresterhill, University of Aberdeen, Aberdeen AB25 2ZD, UK quence of C. albicans available, we are now in a po-
2 Robert Koch Institut, NG4, Nordufer 20, 13353 Berlin, Germany sition to analyse the pathogenesis of infections with
The Mycota XIII
Fungal Genomics
164 C.A. Munro et al.
this fungus not only on a traditional gene-by-gene major rearrangements, genome duplication events
basis but also by using much broader approaches. and gene losses. The diversity of changes taken
By means of in silico methods, such as when com- by each branch reflects the long evolutionary time
paring the genome of C. albicans with the genomes span separating these organisms – indeed, this is
of less virulent related species or when compar- as long as that for the phylum of chordates (a fact
ing the genomes of strains of C. albicans differing which should be kept in mind when we discuss the
in pathogenic potential, we may identify genes or experiments described below).
genetic properties which are crucial for infection, The most straightforward way to find at-
which exist only in pathogenic fungi or which are tributes unique to pathogenic species is to
even unique to this fungus. By searching for se- generate two groups of genomes, one containing
quences homologous to genes known to be associ- non-pathogenic and the other pathogenic species,
ated with disease in other pathogenic microorgan- and then to simply extract all those genes and
isms, we may identify as yet unknown putative vir- features which are unique to or possibly even
ulence factors of C. albicans. The discovery of new conserved within the latter. This approach has
gene families may hint at functions crucial for sur- been used successfully in the analysis of bacterial
vival on or within a human host. Global experimen- pathogenicity (reviewed in Brosch et al. 2001;
tal approaches based on the availability of complete Whittam and Bumbaugh 2002).
genome sequences will identify genes directly asso- If the aim is to identify genes unique to
ciated with the infection process. This includes the a pathogenic organism, it is of utmost importance
use of microarrays representing all identified open to use well-annotated open reading frame (ORF)
reading frames of C. albicans, which will identify sets with high coding probability. Comparison of
not only infection-associated genes but also global several closely related genomes can yield an im-
transcription patterns, the interplay of gene net- proved set of protein-coding open reading frames.
works, and regulators of these networks. Finally, As shown by Kellis et al. (2003), the comparison
we can apply postgenomic approaches to analyse of four Saccharomyces species (S. cerevisiae, S.
the interactions between C. albicans and its host in paradoxus, S. mikatae and S. bayanus) led to the
vivo or in models which mimic the in vivo situa- elimination of 500 ORFs with very low coding
tion. Thus, postgenomic approaches will enable us probability, the inclusion of 43 new genes, and
to further shed light on the complex process of the the correction of several splice sites, yielding an
pathogenesis of C. albicans infections. improved catalogue of 5538 protein-encoding
ORFs of a size larger than 100 amino acids. For C.
albicans, genome sequences of such closely related
species are not yet available, but the ongoing
II. The Hunt for Pathogenicity-Related
sequencing of the Candida dubliniensis genome,
Genes Using Comparative Genomics the closest known but less virulent relative, and the
and Predictive Algorithms genome of the closely related apathogenic yeast
Debaryomyces hansenii (teleomorph of Candida
The multitude of sequencing projects completed famata), commonly used for cheese production,
or in progress has brought with them a power- will provide a first step in this direction.
ful opportunity to search for genes which may be
involved in pathogenicity and virulence – compar-
ative genomics. Here, the full genome sequences of A. Comparing the C. albicans Translated ORF
two or more species are compared to one another, Set to the Outside World
making it possible to analyse them for similari-
ties as well as distinct differences between species Strain SC5314 was chosen for the C. albicans se-
or groups. A recent example is the study of yeast quencing project (Table 9.1). This strain is a clin-
genome evolution by the Genolevures project (Du- ical isolate belonging to the predominant clade of
jon et al. 2004). Here, the authors were able to de- closely related C. albicans strains which represents
scribe the major evolutionary events which led to almost 40% of all isolates worldwide (Odds et al.
the specification of an entire phylum, by analysing 2004). SC5314 and its derivatives are now the most
the genomes of five yeasts (Saccharomcyes cere- commonly utilised laboratory strains. C. albicans
visiae, Kluveromyces lactis, Candida glabrata, De- lacks a known haploid or homozygous stage and
baryomyces hansenii and Yarrowia lipolytica) for unfortunately, SC5314 displays, as all other known
Candida albicans Pathogenesis 165
Table. 9.1. Websites of C. albicans genome and annotation projects
Project Website
Candida genome sequencing www.sequence.stanford.edu/group/candida/
at Stanford Genome Technology Center
Candida Genome Database candidagenome.org
CandidaDB at the Insitut Pasteur genolist.pasteur.fr/CandidaDB
Candida information at the University of Minnesota alces.med.umn.edu/Candida.html
CanoDB at the University of San Francisco agabian.ucsf.edu/canoDB/
Candida albicans at national research council of Canada candida.bri.nrc.ca/candida/index.cfm
C. albicans strains, a high degree of heterozygosity C. albicans ORFs (3027, the “housekeeping genes”)
throughout its diploid genome which made assem- had putative orthologues in all three species, 1119
bling the entire sequence a major challenge (Jones had a match in S. cerevisiae and in either one of
et al. 2004). the other species, 613 had matches only in S. cere-
With the race to create DNA microarrays for visiae, and 1426 had no match at all. The remaining
genome-wide transcriptional profiling, annotation 234 ORFs were not found in S. cerevisiae but either
of the C. albicans genome sequence was carried only in S. pombe (91) or in human (83) or in both
out in parallel to the assembly and completion of (60). The high number with matches only in the
the genome sequencing project by several differ- human genome is very intriguing and raises the
ent groups (Table 9.1). This has led to the devel- question of their origin. Other yeast species have
opment of several different annotation databases genes with close homologs found only in bacteria
whose datasets are presently being incorporated – Y. lipolytica contains eight such genes, K. lactis
into the Candida Genome Database (Table 9.1). five and D. hansenii one (Dujon et al. 2004). One
Currently, the only human pathogenic yeast se- possible explanation is that these genes may have
quenced completely in addition to C. albicans is been acquired from other distinct species (such
C. glabrata, a fungus more closely related to S. as humans or bacteria). However, horizontal gene
cerevisiae than to C. albicans. Many other fungal transfer, which is very important in prokaryotes,
sequencing projects are underway (an up-to-date has not yet been demonstrated in fungi. The evo-
list can be found at the Genomes OnLine Database; lution of these genes in C. albicans requires further
Bernal et al. 2001), also for several fungal pathogens investigation.
of humans and plants, from different phylae. Sim- The functional classification of the genes
ply assembling evolutionary distant organisms into unique to C. albicans within this comparison
pathogen/non-pathogen groups might yield a dis- (Jones et al. 2004) shows numerous distinct
tinct set of “universal” pathogenicity genes. This differences in oxidative metabolism. C. albicans
approach is discussed by Ahmad et al. (Chap. 3, possesses several additional components for respi-
this volume). However, most of these organisms ration, such as an extra complex I of the electron
have evolved into specific niches, and have prob- transport chain, a pyruvate dehydrogenase kinase
ably developed different approaches towards sur- for possible flow regulation from glycolysis to the
vival in those environments. Hence, a comparison TCA cycle, a complete family of secreted lipases
on this level would most likely yield a reliable and (Hube et al. 2000; see Sect. IV) and other enzymes
conserved set of housekeeping genes, and a multi- in fatty acid metabolism and, last but not least,
tude of overlapping subgroups which would reflect additional amino acid catabolic pathways, includ-
different nutritional and ecological adaptations as ing one for cysteine. This difference in sulphur
well as phylogeny. metabolism is also reflected by an additional
In the original paper describing the C. albicans direct cysteine biosynthesis pathway, which may
genome sequence (Jones et al. 2004), the authors indicate an increased significance for glutathione
presented a BLASTp-based classification of the pre- metabolism in C. albicans.
dicted protein sequences from 6419 ORFs of the Finally, C. albicans has a number of genes re-
reduced haploid genome of C. albicans, against S. lated to the pH regulatory genes of Aspergillus
cerevisiae, Schizosaccharomyces pombe and the hu- (Davis et al. 2000), possibly reflecting an adaptation
man databases. They found that almost half of the to the environment of the digestive tract. Also en-
coded are a small family of chloride channels, with of S. cerevisiae, but have adapted to promote
members resembling types expressed in mammals adhesion to host cells in C. albicans (see Sect. IV).
(Jones et al. 2004). Kellis et al. (2003) found evidence for rapidly
In CandidaDB (d’Enfert et al. 2005), the hap- evolving proteins, mainly due to changes in the
loid genome of C. albicans was found to contain copy number of trinucleotide repeats. Such a fea-
5917 protein-coding ORFs, including some shorter ture may well be missed in microarray studies if
than 100 amino acids. A comparison of these pre- it lies outside the region selected for probe design,
dicted protein sequences to all publically available but it may be very significant in terms of func-
sequences revealed 548 proteins specific to C. al- tion. The same is true for (point) mutations which
bicans and D. hansenii, and around 600 sequences change substrate specificity or regulatory features
unique to C. albicans (C. d’Enfert and F. Tekaia, of a protein and, of course, the regulatory elements
personal communication). These latter sequences of the promoter region.
presumably contribute to pathogenicity and other
traits which separate these two otherwise closely
related organisms. B. Genes Encoding Proteins
Assuming that genes responsible for with Unknown Function
pathogenicity are not found in closely related
but rather in apathogenic species or strains, there Researchers are faced with another problem –
is another commonly used method for comparative currently, about 25% of predicted proteins in C.
genomics using cross-hybridisation of genomic albicans have no known function and contain no
DNA to a DNA microarray of the species of interest identifiable motif signatures. Some of these have
(Murray et al. 2001; Daran-Lapujade et al. 2003; similarities with proteins of unknown function
Moran et al. 2004). Here the spots of interest, in other organisms. Most proteins of at least
representing a gene not present or very divergent partially known function are involved in processes
at the nucleotide level, will give only a weak signal which have been extensively studied, such as
or no signal at all. A drawback of this method metabolic enzymes and structural or regulatory
is that one misses all of those genes absent from proteins. Within groups of genes identified in
the array but present in the organism used for studies of host/pathogen interaction in C. albicans,
comparison. the proportion of proteins of unknown function
Within the framework of the various ap- can be much larger, up to 35% (Fradin et al.
proaches presented above, one important issue 2005), which reflects the current status of detailed
remains to be discussed. Genes which are present biological understanding of pathogenesis. Here,
in a pathogen and absent in non-pathogens clearly the so-called guilt by association principle can
have a higher probability of being involved in be of great help – genes involved in one and the
pathogenicity, this probability increasing the same cellular process are likely to be co-regulated
smaller the genetic distance becomes between (Allocco et al. 2004). Hence, comparative genomics
the organisms compared. A good example of this offers another promising approach, namely, the
is the family of secreted aspartic proteinases, identification of cis-regulatory elements in closely
which are unique to C. albicans (Naglik et al. related species by promoter sequence comparison.
2003a, 2004; see Sect. IV). Originating from several It is feasible that cis-acting elements are more
gene duplications, the members of this family conserved than are the rest of the upstream or
have subsequently diversified with respect to downstream regulatory region of a protein-coding
their expression patterns and their functions in ORF. Thus, it should be possible to identify motifs
pathogenesis. However, the inference that proteins which are conserved across orthologous genes.
present in both organisms cannot be involved in Following this strategy, Kellis et al. (2003) have
pathogenesis is not always correct. For example, identified 36 known cis-regulatory motifs in S.
the PLD1 gene in S. cerevisiae is essential for cerevisiae, which corresponds to 85% of the full
sporulation, while PLD1 in C. albicans (which set with highly conserved patterns. In addition,
has a deduced amino acid sequence very similar they found 42 new motifs of which 25 correlated
to that of Pld1p in S. cerevisiae) plays a role in well with functional classification based on gene
dimorphism and invasion, attributes involved ontology categories and expression patterns of the
in pathogenesis (Hube et al. 2001). Similarly, adjacent protein-coding regions. Taken together
agglutinin-like sequences play a role in mating with the data produced by cluster analysis of
transcriptional profiling experiments, this leads III. Genes with Possible Virulence
to the large-scale functional classification of gene Functions
products of no known function.
Sequence-based prediction algorithms are
progressively becoming more sophisticated, Different microbial pathogens of humans face
gaining very high importance in whole-genome similar conditions during colonisation and prolif-
annotation projects. Today, the main focus lies eration on or within a host. Microbial cells attach
on large-scale protein functional classification to human cells and surfaces, use the nutrients
(e.g. InterProScan, www.ebi.ac.uk/InterProScan/) available, adapt to changing pH and oxygen levels,
and prediction of subcellular localisation. In develop stress tolerance, develop mechanisms to
the latter, the prediction of amino acid motifs evade the immune system, and damage host tissue
which target proteins to the secretory pathway in order to promote invasion. Thus, pathogenic
(signal peptides) is the furthest developed, and microorganisms share a similar repertoire of
several studies have aimed to computationally strategies for growth and adaptation in association
annotate the “secretome”, or subsets thereof, from with their hosts and pathogenesis. Orthologues
the Candida genome sequence (Lee et al. 2003; of known fitness and virulence genes in bacteria
De Groot et al. 2003; Alberti-Segui et al. 2004; or other fungi have already been identified in the
Eisenhaber et al. 2004; d’Enfert et al. 2005). For the sequenced genome of C. albicans, and it is likely
analysis of the host–pathogen interface, the subset that additional genes will be found. Therefore, we
of these proteins localised to the cell surface is of can identify fitness and virulence genes which are
special interest, as this enables them to directly novel to C. albicans by searching for sequences
and physically interact with the host. Research with similarity to genes known to be associated
has mainly focused on the prediction of proteins with these strategies in other pathogenic mi-
which are attached to the cell wall via remnants croorganisms. For example, in order to survive
of glycosylphosphatidylinositol (GPI) anchors (De phagocytosis by immune cells, genes involved
Groot et al. 2003), revealing over 100 potential in an antioxidative response must be expressed
candidates in C. albicans, at least one-third more to avoid or survive the oxidative burst. This is
than in all of the other species (all non-pathogenic) a reaction displayed by numerous pathogenic
analysed. C. albicans GPI-proteins not found in bacteria (Ruckdeschel 2002; Samuel et al. 2003; Ng
S. cerevisiae include the Als, Hyr1, Csa1/Rbt5 et al. 2004), and similar genes have been discovered
and Hwp1/Rbt1 protein families (De Groot et al. in C. albicans (Fradin et al. 2003; Rubin-Bejerano
2003). Compared to S. cerevisiae, the dataset in C. et al. 2003; Martchenko et al. 2004; Fradin et al.
albicans also has a higher number of hypothetical 2005; see Sect. X). Other bacterial gene products
proteins of unknown function and sugar hydro- are essential for growth in carbohydrate-poor
lases (Eisenhaber et al. 2004). C. albicans uniquely conditions of the phagosome (McKinney et al.
contains GPI-anchored superoxide dismutases 2000), and also seem to play a critical role for
Sod4p (Pga2p), Sod5p (Pga3p) and Sod6p (Pga9p), survival of C. albicans in macrophages (Lorenz
as reported by De Groot et al. (2003), Martchenko and Fink 2001; see Sect. V).
et al. (2004) and Fradin et al. (2005). Extracellular hydrolytic enzymes contribute
In summary, computer-based methodologies to virulence of pathogenic microorganisms, and
are already vital in identifying and analysing genes genes encoding secreted proteolytic and lipolytic
possibly involved in microbial pathogenesis in enzymes with known or putative virulence func-
postgenomic analysis, and the impact of their tions have been identified by pre- and postgenomic
contributions will certainly increase in the next approaches (Ghannoum 2000; Hube et al. 2000;
few years. Nevertheless, the definite elucidation of Naglik et al. 2003a). In particular, C. albicans genes
the function and role of identified genes will always encoding extracellular lipases have been identified
require experimental proof. This may be done on by a combination of screening a genomic fosmid
a gene-by-gene basis. However, promising global library (Magee and Scherer 1998) and the first
experimental approaches to elucidate gene func- sequence data released by the Stanford C. albicans
tions and to identify virulence genes have already genome project (Hube et al. 2000).
been used (see Sect. VIII) and will, in combination One essential component for growth which
with computer-based methodologies, boost our has limited availability within the human host is
knowledge about the process of infection. iron. Therefore, pathogenic microorganisms have
Table. 9.2. Genes involved in iron metabolism identified in the genome of C. albicans
Function Gene names References

Siderophore transporter ARN1/SIT1, CA3740 Ardon et al. (2001), Heymann et al. (2002),
Hu et al. (2002), CandidaDB (see Table 9.1)
Ferric permeases FTR1, FTR2, CA5345, CA5354 Ramanan and Wang (2000), CandidaDB
Iron transporter FTH1 (CA1521), FTH2 (CA1493) CandidaDB
Reductases CFL1, CFL95, CA3460, CA1701, Hammacott et al. (2000), Knight et al. (2002),
CA2397, CA3461, CA0947 CandidaDB
Ferroxidases FET3, CA2922, CA2923, CA2924, Eck et al. (1999), CandidaDB
CA1431, CA5401, CA2920
gained numerous ways for iron uptake and main- sation by gene conversion (Dujon et al. 2004).
tenance of intracellular iron homeostasis (Crosa In C. albicans, several gene families have been
1997; Rodriguez and Smith 2003; Perkins-Balding identified whose functions may be associated
et al. 2004). This includes siderophore production with fitness, adaptation to the host environment,
and uptake of siderophores, transferrin or lactofer- and virulence (Jones et al. 2004). These include
rin receptors, enzymes which reduce and oxidise genes encoding proteinases (Naglik et al. 2003a),
iron, iron permeases and iron transporters. Iron lipases (Hube et al. 2000), adhesins (Hoyer 2001),
uptake mechanisms are also crucial for growth superoxide dismutases (Martchenko et al. 2004;
and survival of environmental microorganisms in Fradin et al. 2005), oligopeptide transporters
competition with other microorganisms. Not sur- (Lubkowitz et al. 1997), iron transporter and
prisingly, a large number of genes involved in iron permeases (Ramanan and Wang 2000; CandidaDB;
metabolism have been identified in the genome of Table 9.2), multicopper oxidases (Eck et al. 1999;
C. albicans by screening for sequences similar to CandidaDB; Table 9.1), iron reductases (Hamma-
those of S. cerevisiae genes encoding siderophore cott et al. 2000; Knight et al. 2002; CandidaDB;
transporters (Ardon et al. 2001; Heymann et al. Table 9.2), mannosyltransferases (Hobson et al.
2002; Hu et al. 2002), iron reductases (Hammacott 2004, Munro et al. 2005), peroxisomal carnitine
et al. 2000; Knight et al. 2002), ferroxidases (Eck acetyl transferases (Prigneau et al. 2004), and
et al. 1999), or ferric permeases (Ramanan and putative permeases with transmembrane regions
Wang 2000; Lan et al. 2004; Table 9.2). It should be (CandidaDB). A few of these families, such as
pointed out that it is very likely that C. albicans secreted aspartic proteinases (SAPs; Naglik et al.
has genes necessary not only for adaptation to and 2003a, 2004), agglutinin-like adhesins (ALSs;
interaction with host cells but also for survival in Hoyer 2001) and extracellular lipases (LIPs; Hube
competition with other members of the normal et al. 2000), which all have at least nine members,
microbial flora. The search for genes involved have been investigated in more detail (Fradin and
in iron metabolism in C. albicans led to the Hube 2003).
identification of large gene families (see Sect. IV), The extracellular proteinase activity of C. albi-
reflecting the importance of iron as a vital nutrient cans is due to the secretion of aspartic proteinases
for growth and survival of C. albicans as both encoded by ten SAP genes (reviewed by Naglik
a commensal and a pathogen. et al. 2003a). Although genes encoding aspartic
proteinases exist in S. cerevisiae (e.g. PRA1, BAR1,
and the yapsines YPS1, 3, 5, 6, 7), the majority
of the SAP gene family seem to be unique to C.
IV. Gene Families albicans. Even the most related Candida species,
C. dubliniensis, does not contain such a variety
Gene families in ascomycetes seem to be relatively of genes encoding these enzymes (Moran et al.
frequent (Dujon et al. 2004). Some families with 2004). Saps seem to be involved in a number of
orthologous proteins in distinct species show ei- processes relevant to the pathogenesis of C. al-
ther relatively low sequence identities (25%–50%), bicans. For example, Saps may contribute to cell
probably representing ancient gene duplications, damage and tissue invasion by the hydrolysis of
or highly similar proteins (90%–100%), reflecting host proteins such as keratin and collagen. Al-
recent duplications and/or sequence homogeni- most all investigated proteins of the human im-
mune system, including immunoglobulins IgG, IgA the activity of other proteins such as the secreted
and sIgA, alpha-macroglobulin, proteins associ- proteinases. As has been shown for other microbial
ated with leukocytes and salivary proteins, are hy- lipases, Candida lipases may also directly affect the
drolysed by purified Sap2p proteinase. Since ad- host immune system by inhibiting cell-mediated
hesion of C. albicans to epithelial and endothelial chemotaxis, by damaging phagocytic cells or by
cells was shown to be inhibited by the classical as- triggering local inflammatory responses.
partic proteinase inhibitor pepstatin A, Saps may Fradin and Hube (2003) have recently specu-
also function in attachment processes, possibly by lated why C. albicans (and perhaps other eukary-
modifying surface proteins. otic pathogenic microorganisms) may possess such
The ALS gene family was identified by the large gene families. They concluded that the pos-
similarity of their products to the S. cerevisiae session of a variety of gene families by pathogenic
mating-associated adhesion glycoprotein α- microorganisms may confer at least five possible
agglutinin (Sag1) encoded by AGα1 (SAG1; Hoyer advantages.
2001). Despite a relatively high sequence similarity
1. Gene families may have evolved to facilitate
between Als1p and Sag1p, Als1p is not believed
the coordinated regulation of members of this
to be involved in C. albicans mating (Magee et al.
family, together with other virulence attributes.
2002). Als1p has been demonstrated to mediate
the adherence of C. albicans to epithelial and en-
2. The different gene products may have adapted
dothelial cells via the amino terminal domains of
and co-evolved to function in different tissues
these proteins (Gaur and Klotz 1997; Fu et al. 1998;
and/or environments during colonisation and
Gaur et al. 1999), and Als5p was shown to bind to
infection.
host extracellular matrix proteins (Gaur and Klotz
3. Members of a gene family may be functionally
1997). Sheppard et al. (2004) have extended these
redundant and act by providing a second pro-
studies and demonstrated that different members
tein when another member of the family fails,
of the Als family have different adhesion profiles to
is removed or is otherwise lost.
a variety of host substrates, this specificity being
4. They may encode proteins with similar activi-
mediated through the N-terminus. Furthermore,
ties but distinct and different functions.
immunohistological analysis has shown that Als
5. Finally, the concomitant expression of a num-
antigens are produced in murine disseminated
ber of similar, but functionally distinct, mem-
candidosis (Hoyer et al. 1999).
bers of a gene family may result in synergistic
Extracellular lipase activity encoded by ten
effects to promote colonisation or infection.
members of the LIP gene family has no counterpart
Thus, several gene products may act in uni-
in S. cerevisiae or any other ascomycetes (Fu et al.
son to carry out a series of tasks to possibly
1997; Hube et al. 2000; Stehr et al. 2004). The
provide the microorganism with a biological
most closely related gene identified in databases
advantage.
was from Mycobacterium tuberculosis, which
probably encodes a lipase (Hube et al. 2000).
The open reading frames of all ten lipase genes
encode highly similar proteins, with up to 80% V. Role of Metabolism During Infection
identical amino acid sequences (Hube et al. 2000).
These lipases may have evolved to adapt to the Some of the most notable differences which ex-
permanent association of C. albicans with the ist between the genomes of the pathogen C. al-
human body, and may have important functions in bicans and the model yeast S. cerevisiae are con-
colonisation and infection processes. One obvious cerned with metabolism. C. albicans appears to
role of lipases may be the utilisation of lipid have the ability to utilise a larger range of nutrients
substrates on human tissue, such as the skin or for growth, and this may be one of the attributes
the intestinal tract. The high number of LIP genes which has enabled C. albicans to exist as a commen-
may provide an adaptive advantage to persist on sal organism in healthy individuals and to evolve
these surfaces, in the absence of carbohydrates, into a successful pathogen in the immunocompro-
and may assist C. albicans in competing with the mised. In order to colonise and cause disease in
normal microbial flora. Furthermore, release of a wide range of host sites, C. albicans must be able to
fatty acids due to lipolytic activity may modify the survive and grow in vastly different environments.
surrounding pH of fungal cells, thereby optimising As C. albicans spreads from mucosal surfaces to
deep-seated organs via the bloodstream, it must with phagocytosis is nutrient starvation, and that
cope with changes in pH, oxygen availability, and C. albicans is induced to break down lipids as an
variations in carbon and nitrogen sources. Even as alternative carbon source.
a commensal it can adapt to colonise the oral cav- Like many other pathogens, C. albicans has the
ity, vaginal tract and gastrointestinal tract, which capability to switch phenotypes. In some species,
are different environments where the fungus must for example, the African trypanosome, this
compete with other microbes which make up the involves altering their surface proteins, and thus
natural flora in these niches. they can evade the host’s immune system (Barry
During an infection, C. albicans will enounter and McCulloch 2001). This phenomenon, termed
cells of the immune response. When engulfed by antigenic variation, is reversible and stochastic.
macrophages, C. albicans yeast cells switch to hy- In C. albicans, phenotypic switching results in
phal growth. Not only does C. albicans survive the altered physiology, morphology and pathogenicity
hostile environment within these phagocytes, but in murine models. Many clinical isolates exhibit
also hyphal growth enables the fungus to grow out a high frequency of switching (Hellstein et al.
and escape these host defences. In a key paper, 1993). The best characterised phenotypic switch-
Lorenz and Fink (2001) described for the first time ing in C. albicans involves the transition from
changes in the transcriptional profile of a fungus white to opaque cells in strains homozygous for
(in this case, S. cerevisiae) in response to phago- the MTL (mating type-like) locus. Accompanying
cytosis by mammalian macrophages. They showed the change from white to opaque colonies are al-
that S. cerevisiae responded to conditions within terations in cell size and shape, adhesion, ability to
the phagolysosome by up-regulating genes of the form hyphae, hydrophobicity, drug susceptibility,
glyoxylate cycle. The glyoxylate cycle enables sim- sensitivity to neutrophils and pathogenicity (Soll
ple two-carbon compounds to be assimilated by the 2002). White cells are more virulent in a systemic
TCA cycle. The authors inferred from their results murine model, and opaque cells are more success-
that the phagosome was low in complex carbon ful in a cutaneous infection experimental model
compounds causing nutrient deprivation, a major (Kvaal et al. 1999). Therefore, both switch pheno-
stress to the fungal cells. No conventional stress types appear to have an advantage at different host
responses were observed. Lorenz and Fink (2001) sites. In addition, the white-to-opaque transition
went on to show the importance of the glyoxylate is critical for C. albicans mating, with opaque cells
cycle in the virulence of C. albicans, by generating mating at a million-fold higher frequency than do
a mutant in the isocitrate lyase gene (ICL1) which white cells (Miller and Johnson 2002). A genome-
was avirulent when tested in a murine model of wide comparison of the expression profile of white
systemic candidosis. So, here we have an example and opaque cells of the classical switching strain
of the genome-wide response of a non-pathogen WO-1 revealed the differential expression of 373
(S. cerevisiae) providing valuable leads to factors genes (Lan et al. 2002). Approximately one-third
which contribute to virulence in a pathogen (C. of these genes were involved in metabolism. The
albicans). transcriptional profile of white cells reflected
Prigneau et al. (2003) identified seven genes in fermentative metabolism and, in contrast opaque
C. albicans induced by macrophage phagocytosis, cells, gave patterns consistent with oxidative
using differential display coupled to reverse tran- metabolism. White cells had higher levels of three
scription (DDRT)-PCR. Four of the seven genes glucose transporters (HXT3, HXT4 and HXT7) and
were predicted to encode peroxisomal proteins in- genes encoding glycolytic enzymes (hexokinase,
volved in β-oxidation (acyl CoA oxidase, Pxp2p, phosphofructokinase and pyruvate kinase). By
carnitine acetyl transferase, Cat3p) and the gly- contrast, opaque cells had higher levels of genes
oxylate cycle (isocitrate lyase, Icl1p, malate perme- encoding TCA cycle and β-oxidation enzymes.
ase, Mae1p). A fifth gene encoded cytosolic for- The abundance of other genes involved in amino
mate dehydrogenase, which is also associated with acid metabolism, amino acid, phosphate and
the glyoxylate cycle because formate is an indirect sulphate transport also differed between the two
product. The remaining two genes encoded pu- switch types. These differences in metabolic status
tative plasma membrane sensors for extracellular between the two switch types may provide the
nutrients: Snf3p, a glucose sensor/transporter, and advantage to survive and colonise vastly different
Gpr1p, a G-protein coupled receptor. These find- host sites. The up-regulation of the β-oxidation
ings again suggest that the major stress associated pathway in the opaque phase correlates with its
success in the cutaneous model, the skin being rich 3. Hyphal cells, but not yeast cells, have been
in lipids but relatively poor in sugars. Therefore, shown to be endocytosed by endothelial cells,
we can add metabolic specialisation to the list a mechanism which is discussed as a poten-
of properties which are modified as a result of tial strategy of C. albicans to escape from the
phenotypic switching, and which may contribute bloodstream (Zink et al. 1996; Phan et al. 2000).
to C. albicans success as a pathogen with the ability 4. Hyphal cells have stronger adherence proper-
to colonise a diverse range of anatomical sites. This ties (e.g. due to the expression of the Als ad-
issue is discussed further by Brown (Chap. 10, this hesins; Hoyer 2001), which may help the fungus
volume). to adhere to endothelial cells.
5. The higher expression level of superoxide dis-
mutases may counteract the oxidative burst of
VI. Role of Morphology phagocytic cells (Nantel et al. 2002; Fradin et al.
2005).
6. Finally, hyphal cells are known to have greater
One of the most intriguing properties of C. albi-
invasive properties in tissue, which assists the
cans is the ability to switch between yeast and hy-
fungus in penetrating into the surrounding,
phal forms of growth (dimorphism). The signalling
deeper tissue of blood vessels (Calderone and
pathways regulating this transition are discussed
Fonzi 2001; Gow et al. 2003).
by Whiteway and Nantel (Chap. 8, this volume).
In this chapter, we will focus on particular aspects Although it has long been proposed that the
of the role of dimorphism in pathogenesis in the yeast form may be the preferred cell shape for
postgenomic era. haematogenous dissemination (Rooney and Klein
It has long been postulated that the transition 2002), C. albicans normally switches to hyphal
to filamentous growth is associated with invasive growth when exposed to blood plasma. However,
growth of C. albicans. However, during infection hyphal morphology per se is not sufficient for
both morphological forms are found, and both pathogenesis. Instead, the entire transcriptional
forms may play their individual roles during infec- programme associated with the transition (Odds
tion. Furthermore, many pathogenic fungi such as et al. 2003) is likely to account for any hyphae-
H. capsulatum, Penicillium marneffei and Paracoc- associated pathogenesis. It should be noted that
cidioides brasiliensis are also dimorphic, but here cells which are locked in the yeast phase still
the hyphal form is the environmental growth form have the ability to cross the endothelial barrier of
which switches to a yeast form in the human body. blood vessels, but clearly have reduced virulence
Thus, the ability to produce filaments is not gen- potential (Saville et al. 2003). Furthermore, as
erally associated with fungal virulence (Gow et al mentioned above, the yeast form and genes
2002). associated with the yeast form are likely to play
Clearly, dimorphism in C. albicans is regulated important roles during other types of infection at
by a complex transcriptional programme which not different body sites.
only modulates the morphology but also includes
the expression of a number of hyphae-associated
genes, encoding surface proteins (HWP1, HYR1, VII. Role of the Cell Surface During
ALS3/8), secreted proteinases (SAP4-6), and detox- Host/Pathogen Interactions
ification enzymes, such as superoxide dismutases
SOD5 (Hube et al. 1994; Brown and Gow 1999; Nan-
The cell wall of C. albicans is in intimate contact
tel et al. 2002). During systemic infections, the mor-
with host cells and tissues during an infection.
phological transition and expression of hyphal-
Exposed on the C. albicans outer surface are highly
associated factors may benefit C. albicans in several
glycosylated mannoproteins which act as adhesins
ways (Hube 2004; Fradin et al. 2005).
(Hoyer 2001; Sundstrom 2002), antigenicity
1. Hyphal cells secrete factors which inhibit factors and immunomodulators (Suzuki 2002;
killing by neutrophils (Smail et al. 1992). Lopez-Ribot et al. 2004), and contribute to biofilm
2. C. albicans cells phagocytosed by macrophages formation (Douglas 2003; Garcia-Sanchez et al.
produce hyphae which penetrate the sur- 2004). These properties involve both the protein
rounding membranes, causing the death of the and carbohydrate moieties of mannoproteins
macrophage (Borg-von Zeppelin et al. 1998). (Buurman et al. 1998; Sheppard et al. 2004; Munro
et al. 2005). C. albicans can adhere to and colonise tions, expressed 14 proteins which were covalently
a wide range of niches in the host, reflecting a ver- attached to the cell wall. These included the ad-
satile cell wall composition. Indeed, experiments in hesins Als1p, Als4p and flocculin-like Pga24p, the
vitro have shown that the fungal cell wall is highly superoxide dismutase Sod4p, carbohydrate active
dynamic, and alters its structure and architecture enzymes Cht2p, Crh11p, Pga4p, Phr1p, Scw1p, and
in response to changing environments, different others with no known function, Ecm33.3p, Pir1p,
stresses and in response to defects in cell wall Pga29p, Rbt5p and Ssr1p. The next challenge will
components (reviewed by Klis et al. 2002). Many be to examine the cell wall proteome when cells
of the environmental conditions which lead to cell are grown under conditions which mimic in vivo
wall modifications would be encountered during environments.
an infection as the fungus colonises new sites in The cell wall proteome of C. albicans has greatly
the body, for example, changes in pH, oxygen diverged from that of S. cerevisiae, including the ad-
and nutrient availability. The ability of the cell dition of the Als family of adhesins. Comparison of
to respond to a wide variety of environmental the genomes of the closely related C. albicans and C.
conditions reflects the complex and intricate dubliniensis has also revealed significant sequence
network regulating cell wall biogenesis. The cell divergence amongst the GPI-proteins, and the ab-
wall proteome in particular seems adaptable, re- sence of several members of the Als family (Moran
sponding to fluctuating pH and oxygen levels, for et al. 2004). The structure and number of cell sur-
example (Klis et al. 2002). Two classes of proteins face proteins appear to be evolutionary less con-
are covalently attached to the C. albicans cell wall: strained than is the case for metabolic enzymes –
(1) the Pir proteins (proteins with internal repeats), for example, enabling the cell wall proteome to be-
attached to β(1,3) glucan by an alkali-sensitive come diversified between species, leading to varia-
bond and (2) GPI-anchored proteins, attached tions in the complement of adhesins giving speci-
via a GPI-anchor remnant to β(1,6) glucan. The ficity for a variety of host tissues. Many of these cell
GPI-cell wall proteins share a conserved structure wall proteins are antigenic and elicit an antibody
which includes a serine/threonine-rich domain. response in vivo (reviewed by Lopez-Ripot et al.
This class includes some of the best studied cell 2004). Such proteins may be suitable candidates
wall proteins of C. albicans, including the Als for the development of serodiagnosis of systemic
family of adhesins (Hoyer 2001), the pH-regulated candidiasis or as vaccines for novel therapies.
glycosyltransferases Phr1p and Phr2p (Saporito- Analysis of the C. albicans genome sequence
Irwin et al. 1995; Muhlschlegel and Fonzi 1997; has revealed that many of the cell wall synthetic
Fonzi 1999) and the hyphal-specific proteins genes in C. albicans are similar to their S. cerevisiae
Hyr1p (Bailey et al. 1996) and Hwp1p, a substrate counterparts. There are, however, some notable dif-
for mammalian transglutaminase (Staab et al. ferences. For example, C. albicans has an additional
1999). chitin synthase gene CHS8, giving it two Class I en-
The availability of an annotated genome se- zymes homologous to ScCHS1 (Munro and Gow
quence has greatly facilitated the analysis of the cell 2001; Munro et al. 2003). Also, S. cerevisiae has two
wall proteome, as discussed in Sect. II. Cell wall β(1,3) glucan synthase subunits FKS1 and FKS2
proteins can be fractionated depending on their whereas C. albicans has only one essential enzyme
mechanism of attachment to the cell wall, reduc- Gsl21p (Mio et al. 1997). There are a number of ex-
ing the enormous complexity of cell wall analysis. amples where gene families of functionally redun-
A comparison has been made by two-dimensional dant proteins have been expanded in C. albicans,
gel electrophoresis of cell wall proteins in yeast and compared to S. cerevisiae. These include MNT-like,
hyphal cells (Pitarch et al. 2002), revealing a hetero- MNN1-like, MNN2-like and MNN4-like gene fam-
geneous set of proteins associated with the cell wall. ilies (Hobson et al. 2004; Bates, Munro and Gow,
Along with the known hyphal-specific proteins, personal communication). The significance of this
a large number of cell surface-associated proteins is not yet known. The presence of gene families
were up-regulated under hyphal-inducing condi- complicates gene function analysis, as the true ho-
tions. More recently, De Groot et al. (2004) have mologues of the S. cerevisiae genes are hard to pre-
developed a method to analyse covalently linked dict by amino acid sequence alone (see Sect. II).
cell wall proteins using tandem liquid chromatog- Mutants lacking single members of the MNN1-like
raphy and mass spectrometry. Exponential growth and MNN2-like families have no distinguishable
phase yeast cells, grown under laboratory condi- phenotype (Bates, Hughes, Munro and Gow, per-
sonal communication). Therefore, further analysis recombination to create null mutants are time-
will require multiple gene knockouts, which are consuming and laborious. Recent advances using
difficult in C. albicans, or complementation tests PCR-based methods, flipper cassettes and a posi-
with S. cerevisiae mutants. It will be interesting to tive selectable marker have improved this process
determine whether members of these gene families (Morschhäuser et al. 1999; Wilson et al. 2000; Gola
exhibit differential expression patterns which may et al. 2003; Shen et al. 2004; Reuß et al. 2004).
give clues to the reasons why these families have Over the last 3 years, several strategies have been
become expanded. employed to develop methods for genome-wide
The fungal cell wall is under complex regula- or large-scale analysis. Although some have been
tion at the transcriptional and post-transcriptional pioneered by companies in drug discovery pro-
level. A large cohort of cell wall-related genes are grammes aimed to identify essential genes, they
differentially regulated during the dimorphic tran- may provide useful insights also into genes in-
sition, and some are hyphal specific (HYR1, ALS3, volved in pathogenicity and growth in vivo. In
HWP1). Therefore, the pathways which regulate cell 2001, De Backer et al. described an approach that
wall biosynthesis will also play an important role used antisense RNA and promoter interference to
in pathogenesis. Using cell wall-specific microar- analyse gene function. Transformation with an an-
rays comprising 117 probes, Efg1p, a basic helix- tisense cDNA library yielded over 2000 transfor-
loop-helix transcription factor, has been identified mants – 87 genes were identified which were re-
as a major regulator of cell wall biosynthesis con- quired for growth in vivo, and 45 of these were of
trolling both hyphal-specific genes such as HWP1 unknown function. The role of these genes in in
and yeast-specific genes (Sohn et al. 2003). EFG1 vivo growth has not been published. Roemer et al.
had already been implicated in regulating mor- (2003) developed the GRACE (Gene Replacement
phogenesis (Stoldt et al. 1997). Using the same And Conditional Expression) method primarily to
cell wall arrays, Lotz et al. (2004) have examined screen for novel antifungal drug targets. In this
the role of Rim101p, the transcription factor in- approach, a single copy of each gene is disrupted
volved in pH-dependent gene regulation. Rim101p with a PCR-derived cassette comprising a HIS3
was already known to activate PHR1 and PRA1 in selectable marker and unique bar codes for tag-
response to alkaline pH, and PHR2 in response ging. The second copy of each gene is placed under
to acid pH. Out of the 117 genes on the array, the control of a tetracycline regulatable promoter,
32 were differentially regulated by Rim101p, with with the SAT-1 gene conferring resistance to noure-
nine activated including the hyphal-specific genes seothricin used as a dominant selectable marker.
HWP1, ALS1, ALS5 and RBT1. RIM101 negatively This results in the systematic generation of a bank
regulated 23 genes, including RBR1 which encodes of conditional mutants with expression controlled
a GPI-protein. The disruption of the RBR1 gene re- by the absence or presence of the repressor tetra-
sulted in defective pH-induced hyphal growth. Ex- cycline.
pression of RBR1 was dependent on the transcrip- A total of 567 C. albicans genes were found to be
tional repressor Nrg1p, and Rim101p was shown essential for growth in vitro using this technique,
to repress Nrg1p. Therefore, a complex regulatory with only 61% of S. cerevisiae essential genes be-
circuit exists which links morphogenesis, pH re- ing also essential in C. albicans. This highlights the
sponses and cell wall dynamics, allowing C. albi- important differences between these two species.
cans to respond to external cues. GRACE technology can also be applied to look
at gene function in vivo. An important part of
drug target validation is that the gene selected
plays a role in pathogenesis and growth in vivo.
VIII. Global Approaches A subset of 140 GRACE strains have been tested
to Identify Virulence Genes in vivo using a murine model of systemic candi-
dosis, feeding mice tetracycline or doxycycline to
The availability of the genome sequence has been switch off the expression of the target gene. Tetra-
invaluable in providing the C. albicans research cycline may be administered prior to, during or
community with easily accessible gene sequence after challenge with C. albicans (Kauffman et al.
data. However, the analysis of gene function in 2004). Data for six genes which were identified as
C. albicans has been hampered by its diploid na- essential for growth in vitro, and which included
ture. Current methods which rely on homologous two conditional mutants for BRX1 and YJL109C,
showed 80% or greater survival in the presence of formation, as judged by alterations in colony mor-
doxycycline. Upon removal of the repressor, mice phology. Identification of the transposition point in
infected with the YJL109C strain still had a high each strain narrowed down the number of unique
fungal burden in the kidneys. The BRX1 strain was genes affected in hyphal formation to a total of 146.
completely cleared from the animals, even in the Among these genes were several already implicated
absence of doxycycline. The remaining four genes, in filamentous growth (TUP1, RFG1, PDE2, CZF1,
TRP5, PDC1, MRPL6 and EXG1, were not essential NRG1, NOT1), but 27% lacked significant similarity
for growth in vivo, emphasising the need for more to any genes of S. cerevisiae or in the databases.
detailed, large-scale analysis of the requirements Genetic screens provide valuable insights into
for the growth of C. albicans in vivo. cellular processes in C. albicans, and highlight im-
Large-scale gene analysis performed in aca- portant differences between this pathogen and the
demic laboratories has involved the generation model yeast S. cerevisiae. As the ability to switch
of mutant libraries by transposon mutagenesis. between yeast and hyphae is required for virulence,
Davis et al. (2002) modified a Tn7 transposon the genes identified in this screen may have a po-
with a UAU1 marker cassette comprised of the tential role in pathogenicity. These approaches may
ARG4 gene flanked by sections of the URA3 gene not be so amenable for in vivo studies, due to the
which can recombine to excise ARG4. Addition of large number of animals required to screen such
the UAU1 cassette to the transposon allows one libraries. However, preliminary screens could be
to select for homozygous insertion mutations. performed in vitro using a number of model sys-
The Tn7–UAU1 was transposed into a C. albicans tems, and only those strains exhibiting phenotypes
genomic library, and 253 insertion cassettes were where defects in pathogenicity may be inferred
isolated and transformed into C. albicans to gen- would then be tested in vivo.
erate heterozygous insertion mutants. The UAU1
cassette was then employed to identify homozy-
gous mutants which had arisen through mitotic
recombination or gene conversion, and which IX. Transcriptional Profiling
now expressed both ARG4 and URA3 markers. During Infection
A total of 217 homozygous mutants were isolated,
with insertions in 197 ORFs. These homozygous A number of technologies such as reverse
mutants were screened for defects in alkaline transcriptase-PCR (RT-PCR) and in vivo expres-
pH-induced hyphal growth, and insertions in sion technology (IVET) have been used to study
three genes gave the desired phenotype – SLA2, the expression of selected genes from C. albicans
RIM13 and MSD3. This was the first time that during infection (reviewed in Hube 2004). Much
MSD3 had been implicated in pH response and broader, genome-wide approaches including
filamentous growth. differential display (Prigneau et al. 2003), cDNA
Uhl et al. (2003) introduced a modified Tn7 subtractive hybridisation (Fradin et al. 2003),
transposon into genomic DNA fragments to cre- antibody-based strategies (Cheng et al. 2003), and
ate a library which was then transformed back into DNA microarrays (Rubin-Bejerano et al. 2003;
C. albicans. A library of 18,000 C. albicans strains Fradin et al. 2003, 2005) have been used to identify
was generated which was estimated to have a trans- infection-associated genes. Microarray analysis
poson insertion every 2.5 kb of haploid sequence. provides a fascinating tool to unravel the complex
Each strain harboured one heterozygous insertion genetic processes underlying the interaction
mutation, and was screened for phenotypes aris- between microorganisms and the host, and will
ing from loss of function in one copy of the target prove invaluable to our understanding of these
gene, a phenomenon termed haploinsufficiency. diseases (Bryant et al. 2004). The ultimate goal of
Such a gene dosage effect is commonly observed whole-genome expression studies of pathogenic
in C. albicans (Braun and Johnson 1997; Munro microorganisms is the identification of microbial
et al. 1998; Gale et al. 1998). The mutant library was genes which are differentially regulated in the host
first screened for transformants with modifications (Schoolnik 2002). However, there are a number
in filamentous growth under two inducing condi- of major technical challenges for in vivo tran-
tions, Spider medium and in the presence of 1% scriptional profiling, including (1) contamination
foetal calf serum. Approximately 2% (340) of the with host tissue/RNA, (2) mRNA half-life and
strains tested had reproducibly modified hyphal instability, (3) small tissue samples, necessitating
amplification of RNA/cDNA probes, (4) heteroge- about the behaviour of C. albicans in the presence
neous population of cells which colonise or invade of host cells, as well as the host response against
tissue, (5) heterogeneity of biological samples, this fungus, but it is very important to complement
and (6) heterogeneity of microbial strains in these studies with models more representative of
patient samples (Hube 2004). These technical the complexity of the different host environments
problems may be reduced by simply using culture (Fig. 9.1).
conditions which mimic certain features of the
host (Enjalbert et al. 2003). Expression profiles
obtained from microorganisms grown in media A. C. albicans Facing Epithelial
simulating host microenvironments may yield and Endothelial Barriers
a portrait of interacting metabolic pathways
and multistage developmental programmes, and Adhesion of C. albicans to epithelial cells is the first
disclose regulatory networks (Schoolnik 2002). important step of colonisation. Molecular meth-
More sophisticated ex vivo or in vitro infection ods have been used to identify genes involved in
models may mimic mucosal (Schaller et al. 1998) adhesion to epithelial cells. Non-adherent S. cere-
or systemic infections (Fradin et al. 2003, 2005), or visiae were transformed with a C. albicans ge-
may focus on the interaction with a particular type nomic library, and adherent transformants were
of host cell (Lorenz and Fink 2001; Rubin-Bejerano found to contain AAF1, the first C. albicans adher-
et al. 2003). ence gene described (Barki et al. 1993), ALS5 (for-
merly named ALA1; Gaur and Klotz 1997), which
is a member of the ALS gene family mentioned
above encoding cell wall glycoproteins, and EAP1,
X. Gene Expression
which encodes another putative GPI-anchored pro-
During Host/Pathogen Interaction: tein (Li and Palecek 2003). These proteins were
Examples further characterised and shown to promote ad-
hesion of C. albicans to oral epithelial or kidney
The balance between commensalism and cells. In order to mimic more closely the complex-
pathogenicity is determined by both the host ity of mucosal infections, models based on recon-
and C. albicans. To determine how this balance is stituted human oral, oesophageal or vaginal ep-
regulated, it is crucial to analyse in more detail ithelia (RHE) have been used as in vitro models
the host–fungus interface. The availability of the of mucosal candidosis (Schaller et al. 1998, 2003;
entire human and C. albicans genome sequences Li et al. 2002; Green et al. 2004). The use of these
facilitates the high-throughput analysis of human models has brought new insights into the colonisa-
and fungal genes regulated in response to different tion and invasion mechanisms of C. albicans. For
infections or infection-like situations. example, several members of the ALS gene family
Several studies have been performed in order were shown to be expressed in the RHE model, and
to answer different questions which are essential to are possibly involved in the formation of a pseu-
understand the mechanisms of pathogenicity of C. domembranous structure on top of the epithelium,
albicans. The following points need to be clarified. resembling a biofilm (Green et al. 2004). The abil-
ity of C. albicans to grow in close associations with
1. Which virulence factors are important for
surfaces under biofilm-like conditions may have
which type of infection?
several benefits for the fungus (Baillie and Dou-
2. Which environments represent specific host
glas 2000). Therefore, Garcia-Sanchez et al. (2004)
cells or sites for C. albicans?
studied C. albicans during biofilm formation by
3. Which host cells can efficiently prevent C. albi-
analysing the genome-wide transcriptional profile.
cans infection, and how is this achieved?
In this study, ALS1 was found to be expressed, sup-
4. Which host responses can be modulated by C.
porting the view that the ALS gene family is asso-
albicans?
ciated with biofilm formation. Expression of two
Incubation of C. albicans with different mam- other genes (CHK1 and CSSK1), encoding a pu-
malian cell lineages representing different cell tative two-component histidine kinase and a re-
types has been widely used as a simple model sponse regulator, was gradually induced by C. albi-
to study C. albicans-host interaction. This type cans during adherence and colonisation to epithe-
of model provides some important information lial tissue (Li et al. 2002).
Fig. 9.1. Host/pathogen interactions during surface be associated with these different stages of infection
infections (epithelium), deep-seated infections (tissue) and (numbered and listed on the right-hand side) and (2)
bloodstream infections (endothelium and bloodstream). factors and processes characteristic of the host response to
Shown are (1) C. albicans genes which are known to C. albicans infections
The identification of genes which were in- (HWP1), adhesion to host cells, nutrient up-
duced in vivo during oral thrush confirmed take, phospholipid biosynthesis and amino acid
the importance of genes involved in diverse catabolism (LPD1) during the interaction of C.
functions in this process. These included the albicans with the epithelial layers (Cheng et al.
regulation of yeast-to-hyphal morphogenesis 2003).
Genes encoding hydrolytic enzymes were also inflammatory cytokines (Wollina et al. 2004), and
expressed during colonisation and penetration of neutrophils were attracted by physiologically ac-
mucosal models (Hube and Naglik 2002). The SAP tive Candida cells after RHE infection (Schaller
genes have been the most studied gene family in et al. 2004). This pro-inflammatory response can
C. albicans (see Sect. IV), and the order of their also induce stimulation of T-cells, which are cru-
temporal expression was dependent upon the type cial for host defence against C. albicans.
of RHE infection used (Schaller et al. 1998, 2000, After penetration through the epithelium,
2003). In addition, the expression of these genes C. albicans may reach the endothelial barrier
was analysed in vivo in samples from patients suf- and again must display a programme to combat
fering from oral or vaginal infections (Naglik et al. weakened host defences. As for epithelial cells, the
2003b), and compared to the expression pattern interactions between C. albicans and endothelial
of samples from Candida carriers (Naglik et al. cells have not yet been widely studied using
1999). These studies showed that SAP1 and SAP3 functional genomics. Cultured endothelial cells
were more commonly expressed in patients than harvested from human umbilical vein with colla-
in carriers, suggesting that these genes may be genase were used to study genes expressed by these
involved in mucosal infections but not in com- cells after stimulation with C. albicans. Endothelial
mensal growth on mucosal surfaces. Transcripts cells responded to in vitro C. albicans infection by
from lipase and phospholipase genes were also de- expressing genes involved in the pro-inflammatory
tected during RHE infection, during experimental response (Orozco et al. 2000). This response is
mucosal infection of mice, and in patient samples important, as it can determine the success of the
(Naglik et al. 2003b; Schofield et al. 2003; Stehr et al. infection by (1) initiating acute-phase response,
2004). (2) recruiting leukocytes from the bloodstream
As C. albicans is a commensal microorganism, to the site of infection, (3) initiating the adaptive
epithelial cells should have attributes to maintain immune response and (4) increasing specific
the homeostasis between resident C. albicans and immune responses. The recruitment of leukocytes,
the epithelial barrier. Antimicrobial agents such as especially neutrophils, by endothelial cells was
lactoferrin or IgA, which are present in the saliva, confirmed by the expression of genes encoding
as well as other components secreted by epithelial leukocyte adhesion molecules such as E-selectin,
cells, like the defensins, are probably essential to and ICAM-1 by endothelial cells stimulated with C.
control the microbial colonisation of the epithe- albicans (Filler et al. 1996). Taken together, these
lium. RT-PCR data showed that the β-defensins data show that endothelial cells have the capacity
HBD2 and HBD3 are induced in oral tissues infected to enhance the host response against C. albicans in
with C. albicans (Dunsche et al. 2001, 2002). Inter- vivo.
estingly, hyphal cells seem to be the main growth
form able to induce this epithelial response. The
yeast-to-hyphal transition is commonly associated B. C. albicans Facing Phagocytes
with better adhesion and penetration properties
(see Sect. VI), which might explain why epithe- Neutrophils play a key role in the host defences
lial cells respond more strongly and efficiently to against C. albicans, as neutropenic patients are
hyphae than do yeast cells. Furthermore, a non- more susceptible to disseminated candidiasis.
specific defence mechanism, represented by epi- These cells can also play an important role during
dermal cell proliferation, was enhanced during an superficial infections, as they rapidly migrate
acute experimental C. albicans infection (Korting from the bloodstream to the site of infection.
et al. 1998; Bowers et al. 1999). Their recruitment is mediated by chemotactic
In vitro infection models of mucosal surfaces factors which are produced by epithelial cells,
lack non-epithelial cell factors such as dendritic such as Interleukin-8 (Wollina et al. 2004). Ex
cells, macrophages and neutrophils, which might vivo, after contact with neutrophils, the majority
have a dramatic influence on the outcome of C. of C. albicans cells are phagocytosed (Peltroche-
albicans infections in vivo. Phagocytes are usu- Llacsahuanga et al. 2000). Phagocytosis and killing
ally found at the site of mucosal infection, where of an ingested microorganism by neutrophils
they are probably recruited by chemoattractants is a complex mechanism involving oxidative
secreted by epithelial cells. Keratinocytes can in- and non-oxidative agents acting in concert or
deed be stimulated by C. albicans to produce pro- independently of each other. The phagosome
of neutrophils constitutes an extremely hostile burst, and C. albicans responded by expressing

environment with an acidic or alkaline pH, a large set of genes in order to survive this
antimicrobial peptides, reactive oxidative species, stress (Fradin et al. 2005). Effectively, genes
and relatively poor nutrient conditions (Hampton encoding enzymes involved in the dismutation
et al. 1998; Faurschou and Borregaard 2003). of the superoxide anion in H2 O2 , as well as those
Neutrophils efficiently inhibit the yeast-to-hyphal encoding enzymes involved in the transforma-
transition of C. albicans, either by phagocytosis tion of H2 O2 (CTA1) into water, were induced.
or via soluble factors. For example, antimicrobial Interestingly, the expression of the SOD5 gene,
peptides such as lactoferrin, which is contained which is normally hyphae regulated, was strongly
in the neutrophil’s granules and released in expressed in yeast cells, even though neutrophils
the phagosome and extracellular compartment, block the yeast-to-hyphal-transition (Fradin et al.
are very toxic to C. albicans and were shown 2005). The extracellular enzyme Sod5p seems
to inhibit Candida hypha formation (Okutomi particularly important for the fungus when facing
et al. 1997; Wakabayashi et al. 1998). By contrast, extracellular reactive oxygen species. Conse-
C. albicans cells which were phagocytosed by quently, a mutant lacking this gene was more
monocyte-derived macrophages were shown susceptible to killing by neutrophils (Fradin et al.
to have the potential to produce hyphal cells, 2005).
which may assist escaping and killing of the Clearly, neutrophils dominate the host re-
macrophages. Transcript profiling of C. albicans sponse against C. albicans in blood. However,
incubated with purified polymorphonuclear cells, the effect of neutrophils on the transcript profile,
consisting mainly of neutrophils, showed that growth and survival of C. albicans in blood is
genes induced in response to neutrophils reflect attenuated when compared with the response of
the aggressive environment C. albicans is exposed C. albicans to isolated neutrophils (Fradin et al.
to (Rubin-Bejerano et al. 2003; Fradin et al. 2005). 2005). One obvious explanation is that C. albicans
The fungus must overcome the nutrient deficiency can interact with other cells in the blood such
of the phagosome, and it does this by inducing as monocytes, allowing the fungus to escape and
methionine and arginine biosynthetic pathways survive. This suggests that it is important to work
(genes such as MET1 and ARG1; Rubin-Bejerano with a model which is as close as possible to
et al. 2003). The neutrophil phagosome seems to the in vivo situation. Even with such a model,
be a poor source of nitrogen and carbohydrates, it can be difficult to discriminate between the
as genes involved in other amino acid pathways, respective roles of each component in C. albicans
nitrogen metabolism and the glyoxylate cycle behaviour, and between the different patterns of
were also induced by neutrophils (Fradin et al. fungal gene expression in response to the different
2005). Nitrogen deprivation seemed reduced or components.
completely absent in monocyte phagosomes. By As neutrophils are efficient phagocytic cells
contrast, C. albicans cells also appear to be de- which do not have a real role in the initiation
prived of sources of carbohydrate in macrophages, of a specific immune response, it would be of
as the genes involved in glyoxylate metabolism interest to determine whether these cells display
were induced after phagocytosis of C. albicans a transcriptional response upon incubation with
by macrophages, and genes such as ICL1 were microorganisms, including C. albicans. Incubation
crucial for survival of macrophages (see Sect. V). of C. albicans with cells of a granulocytoid lineage,
Furthermore, transcript profiling showed that HL60, showed that a number of neutrophil genes
C. albicans cells were less physiologically active were in fact regulated in response to interaction
after incubation with neutrophils than were cells with the fungus. As expected, several genes
phagocytosed by macrophages (Fradin et al. involved in the pro-inflammatory response were
2005). Several genes encoding components of induced by C. albicans in a dose-dependent
the protein synthesis machinery (RPS10) and manner (Mullick et al. 2004). Interestingly, some
glycolytic enzymes (HXK2, PMI40) were down- genes involved in innate defence, such as the
regulated in the presence of these phagocytes. defensin gene HNP1, were repressed, showing
The transcriptome also revealed that C. albicans a modulation of the neutrophils’ response by C.
faced a strong oxidative environment in the albicans. C. albicans also induced the neutrophils
presence of neutrophils. The NADPH oxidase to undergo apoptosis. Evidently, C. albicans not
of neutrophils initiated the so-called oxidative only expresses different factors to overcome the
neutrophil defence mechanism but the fungus can mans and its mechanisms of virulence (Berman
also modulate neutrophil functions. However, it and Sudbery 2002). Gaps in our knowledge will be
should be noted that neutrophils are known to be filled by a combination of comparative and func-
apoptotic cells and it has to be clarified whether tional genomics, including techniques such as pro-
the apoptosis is due to C. albicans itself blocking teomics, bioinformatics, structural biology and mi-
the neutrophil defence mechanism or whether the croarrays (Brosch et al. 2001).
apoptosis is a normal neutrophil programme to Comparative genomics will uncover novel vir-
stem inflammation at the site of infection. ulence determinants and hidden aspects of patho-
Monocyte-derived macrophages and dendritic genesis (Whittam and Bumbaugh 2002). The role
cells have a definitive role in the initiation of of metabolic and biosynthetic pathways which have
the adaptive immune response. C. albicans, but been neglected in the past as elements of fungal re-
not S. cerevisiae, has been shown to induce sponse to host environments have now come into
apoptosis of macrophages, reducing their defence focus (see Brown, Chap. 10, this volume). The role
activities against the fungus (Ibata-Ombetta et al. of morphology, a more traditional field in analysing
2003). Microarray analysis of the response of virulence attributes of C. albicans, will be further
dendritic cells to C. albicans revealed induction elucidated and pathways with novel key regulatory
of a different set of genes (Huang et al. 2001). factors will be identified. More molecules associ-
Genes involved in the innate response (e.g. ated with the cell surface and involved in direct
phagocytosis and pathogen recognition) were host/fungus interactions will be discovered and
rapidly down-regulated, suggesting that a more their function(s) analysed. New global approaches
specific immune programme was induced during to identify virulence genes will be developed. Fi-
the incubation time. Factors contributing to nally, genome-wide transcriptional profiling will
the recruitment of immune cells to the site of continue to be used to investigate the response of
infection, as well as mediators of shape change C. albicans cells to their natural environments us-
and migratory behaviour of activated dendritic ing infection models. Together with the knowledge
cells, were induced by C. albicans. Furthermore, of the genome of the human host and the use of
several antigen processing and presentation genes host microarrays, these approaches will allow us to
were highly expressed, suggesting preparation look simultaneously at both the host and the fun-
for a more specific immune response against C. gus, providing fascinating insights into the complex
albicans. Nevertheless, it should be noted that process of pathogenesis, and the transition from
overall, dendritic cells are less stimulated by C. commensalism to parasitism of C. albicans. This
albicans than by Escherichia coli. Also, C. albicans knowledge will ultimately enable us to develop new
modulates only a subset of the genes regulated treatments or strategies in the fight against fungal
by E. coli. These observations might highlight the infections.
specific modulation of dendritic cell functions by
C. albicans. Acknowledgements. Donna MacCallum, Ricardo Almeida,
Sascha Thewes and Antje Albrecht for help in preparing
the manuscript. Our own investigations were supported by
the Wellcome Trust (063204), the Robert Koch-Institut, the
Deutsche Forschungsgemeinschaft (Hu 528/7/8/10), and the
XI. Conclusions European Commission (Galar Fungail Consortium QLK-
2000-00795 and MRTN-CT-2003-504148).
The enormous influx of information from genome
sequencing projects is revolutionising the science
of microbial pathogenesis. This ranges from under- References
standing the most basic aspects of gene content and
pathogen genome organisation, to elucidating the
Alberti-Segui C, Morales AJ, Xing H, Kessler MM, Will-
regulatory networks of virulence gene expression, ins DA, Weinstock KG, Cottarel G, Fechtel K, Rogers B
and the investigation of the global patterns of host (2004) Identification of potential cell-surface proteins
response to infection (Whittam and Bumbough in Candida albicans and investigation of the role of
2002). Since the release of the complete C. albicans a putative cell-surface glycosidase in adhesion and vir-
ulence. Yeast 21:285–302
genome sequence in 2000, significant progress has Allocco DJ, Kohane IS, Butte AJ (2004) Quantifying the re-
been made in unravelling the intriguing biology of lationship between co-expression, co-regulation and
this medically important fungal pathogen of hu- gene function. BMC Bioinformatics 5:18
Ardon O, Bussey H, Philpott C, Ward DM, Davis-Kaplan S, Daran-Lapujade P, Daran JM, Kotter P, Petit T, Piper MD,
Verroneau S, Jiang B, Kaplan J (2001) Identification Pronk JT (2003) Comparative genotyping of the Sac-
of a Candida albicans ferrichrome transporter and its charomyces cerevisiae laboratory strains S288C and
characterization by expression in Saccharomyces cere- CEN.PK113-7D using oligonucleotide microarrays.
visiae. J Biol Chem 276:43049–43055 FEMS Yeast Res 4:259–269
Bailey DA, Feldmann PJF, Bovey M, Gow NAR, Brown AJP Davis D, Wilson RB, Mitchell AP (2000) RIM101-dependent
(1996) The Candida albicans HYR1 gene, which is ac- and-independent pathways govern pH responses in
tivated in response to hyphal development, belongs to Candida albicans. Mol Cell Biol 20:971–978
a gene family encoding yeast cell wall proteins. J Bac- Davis DA, Bruno VM, Loza L, Filler SG, Mitchell AP (2002)
teriol 178:5353–5360 Candida albicans Mds3p, a conserved regulator of pH
Baillie GS, Douglas LJ (2000) Matrix polymers of Candida responses and virulence identified through insertional
biofilms and their possible role in biofilm resistance to mutagenesis. Genetics 162:1573–1581
antifungal agents. J Antimicrob Chemother 46:397–403 De Backer MD, Nelissen B, Logghe M, Viaene J, Loonen I,
Barki M, Koltin Y, Yanko M, Tamarkin A, Rosenberg M Vandoninck S, de Hoogt R, Dewaele S, Simons FA,
(1993) Isolation of a Candida albicans DNA se- Verhasselt P et al. (2001) An antisense-based functional
quence conferring adhesion and aggregation on genomics approach for identification of genes critical
Saccharomyces cerevisiae. J Bacteriol 175:5683–5689 for growth of Candida albicans. Nat Biotechnol 19:235–
Barry JD, McCulloch R (2001) Antigenic variation in try- 241
panosomes: enhanced phenotypic variation in a eu- De Groot PW, Hellingwerf KJ, Klis FM (2003) Genome-wide
karyotic parasite. Adv Parasitol 49:1–70 identification of fungal GPI proteins. Yeast 20:781–796
Berman J, Sudbery PE (2002) Candida albicans: a molecular De Groot PWJ, De Boer AD, Cunningham J, Dekker HL,
revolution built on lessons from budding yeast. Nat Rev De Jong L, Hellingwerf KJ, De Koster C, Klis F (2004)
Genet 3:918–930 Proteomic analysis of Candida albicans cell wall re-
Bernal A, Ear U, Kyrpides N (2001) Genomes OnLine veals covalently bound carbohydrate-active enzymes
Database (GOLD): a monitor of genome projects and adhesins. Eukaryot Cell 3:955–965
world-wide. Nucleic Acids Res 29:126–127 d’Enfert C, Goyard S, Rodriguez-Arnaveilhe S, Frangeul L,
Borg-von Zepelin M, Beggah S, Boggian K, Sanglard D, Jones L, Tekaia F, Bader O, Albrecht A, Castillo L,
Monod M (1998) The expression of the secreted as- Dominguez A et al. (2005) CandidaDB: a genome
partyl proteinases Sap4 to Sap6 from Candida albicans database for Candida albicans pathogenomics. Nucleic
in murine macrophages. Mol Microbiol 28:543–554 Acids Res 33:D353–D357
Bowers W, Blaha M, Alkhyyat A, Sankovich J, Kohl J, Wong G, De Repentigny L (2004) Animal models in the analysis of
Patterson D (1999) Artificial human skin: cytokine, Candida host–pathogen interactions. Curr Opin Mi-
prostaglandin, Hsp70 and histological responses to crobiol 7:24–329
heat exposure. J Dermatol Sci 20:172–182 Douglas LJ (2003) Candida biofilms and their role in infec-
Braun BR, Johnson AD (1997) Control of filament formation tion. Trends Microbiol 11(1):30–36
in Candida albicans by the transcriptional repressor Dujon B, Sherman D, Fischer G, Durrens P, Casaregola S,
TUP1. Science 277:105–109 Lafontaine I, De Montigny J, Marck C, Neuveglise C,
Brosch R, Pym AS, Gordon SV, Cole ST (2001) The evolution Talla E et al. (2004) Genome evolution in yeasts. Nature
of mycobacterial pathogenicity: clues from compara- 430:35–44
tive genomics. Trends Microbiol 9:452–458 Dunsche A, Acil Y, Siebert R, Harder J, Schroder JM, Jepsen S
Brown AJP, Gow NAR (1999) Regulatory networks control- (2001) Expression profile of human defensins and an-
ling Candida albicans morphogenesis. Trends Micro- timicrobial proteins in oral tissues. J Oral Pathol Med
biol 7:333–338 30:154–158
Bryant PA, Venter D, Robins-Browne R, Curtis N (2004) Dunsche A, Acil Y, Dommisch H, Siebert R, Schroder JM,
Chips with everything: DNA microarrays in infectious Jepsen S (2002) The novel human beta-defensin-3 is
diseases. Lancet Infect Dis 4:100–111 widely expressed in oral tissues. Eur J Oral Sci 110:121–
Buurman ET, Westwater C, Hube B, Brown AJP, Odds FC, 124
Gow NAR (1998) Molecular analysis of CaMnt1p, Eck R, Hundt S, Hartl A, Roemer E, Kunkel W (1999)
a mannosyl transferase important for adhesion and A multicopper oxidase gene from Candida albicans:
virulence of Candida albicans. Proc Natl Acad Sci USA cloning, characterization and disruption. Microbiol-
95:7670–7675 ogy 145:2415–2422
Calderone RA (2002) Introduction and historical perspec- Eisenhaber B, Schneider G, Wildpaner M, Eisenhaber F
tive. In: Calderone RA (ed) Candida and candidiasis. (2004) A sensitive predictor for potential GPI lipid
ASM Press, Washington, DC, pp 3–13 modification sites in fungal protein sequences and its
Calderone RA, Fonzi WA (2001) Virulence factors of Can- application to genome-wide studies for Aspergillus
dida albicans. Trends Microbiol 9:327–335 nidulans, Candida albicans, Neurospora crassa,
Cheng S, Clancy CJ, Checkley MA, Handfield M, Hillman JD, Saccharomyces cerevisiae and Schizosaccharomyces
Progulske-Fox A, Lewin AS, Fidel PL, Nguyen MH pombe. J Mol Biol 337:243–253
(2003) Identification of Candida albicans genes in- Enjalbert B, Nantel A, Whiteway M (2003) Stress-induced
duced during thrush offers insight into pathogenesis. gene expression in Candida albicans: absence of a gen-
Mol Microbiol 48:1275–1288 eral stress response. Mol Biol Cell 14:1460–1467
Crosa JH (1997) Signal transduction and transcriptional Faurschou M, Borregaard N (2003) Neutrophil granules and
and posttranscriptional control of iron-regulated secretory vesicles in inflammation. Microbes Infect
genes in bacteria. Microbiol Mol Biol Rev 61:319–336 5:1317–1327
Filler SG, Pfunder AS, Spellberg BJ, Spellberg JP, Edwards human epithelium (RHE) model of oral candidiasis
JE Jr (1996) Candida albicans stimulates cytokine pro- and in model biofilms. Microbiology 150:267–275
duction and leukocyte adhesion molecule expression Hammacott JE, Williams PH, Cashmore AM (2000) Candida
by endothelial cells. Infect Immun 64:2609–2617 albicans CFL1 encodes a functional ferric reductase
Fonzi WA (1999) PHR1 and PHR2 of Candida albicans en- activity that can rescue a Saccharomyces cerevisiae fre1
code putative glycosidases required for proper cross- mutant. Microbiology 146:869–876
linking of beta-1,3- and beta-1,6-glucans. J Bacteriol Hampton MB, Kettle AJ, Winterbourn CC (1998) Inside
181:7070–7079 the neutrophil phagosome: oxidants, myeloperoxidase,
Fradin C, Hube B (2003) Tissue infection and site-specific and bacterial killing. Blood 92:3007–3017
gene expression in Candida albicans. Adv Appl Micro- Hellstein J, Vawter-Hugart H, Fotos P, Schmid J, Soll DR
biol 53:271–290 (1993) Genetic similarity and phenotypic diversity of
Fradin C, Kretschmar M, Nichterlein T, Gaillardin C, d’En- commensal and pathogenic strains of Candida albi-
fert C, Hube B (2003) Stage-specific gene expression cans isolated from the oral cavity. J Clin Microbiol
of Candida albicans in human blood. Mol Microbiol 31:3190–3199
47:1523–1543 Heymann P, Gerads M, Schaller M, Dromer F, Winkel-
Fradin C, De Groot P, MacCallum D, Schaller M, Klis F, mann G, Ernst JF (2002) The siderophore iron trans-
Odds FC, Hube B (2005) Granulocytes govern the tran- porter of Candida albicans (Sit1p/Arn1p) mediates up-
scriptional response, morphology and proliferation take of ferrichrome-type siderophores and is required
of Candida albicans in human blood. Mol Microbiol for epithelial invasion. Infect Immun 70:5246–5255
56(2):397–415 Hobson RP, Munro CA, Bates S, MacCallum DM, Cutler JE,
Fu Y, Ibrahim AS, Fonzi W, Zhou X, Ramos CF, Ghannoum Heinsbroek SE, Brown GD, Odds FC, Gow NA (2004)
MA (1997) Cloning and characterization of a gene Loss of cell wall mannosylphosphate in Candida albi-
(LIP1) which encodes a lipase from the pathogenic cans does not influence macrophage recognition. J Biol
yeast Candida albicans. Microbiology 143:331–340 Chem 279:39628–39635
Fu Y, Rieg G, Fonzi WA, Belanger PH, Edwards JE Jr, Filler SG Hoyer LL (2001) The ALS gene family of Candida albicans.
(1998) Expression of the Candida albicans gene ALS1 Trends Microbiol 9:176–180
in Saccharomyces cerevisiae induces adherence to en- Hoyer LL, Clevenger J, Hecht JE, Ehrhart EJ, Poulet FM
dothelial and epithelial cells. Infect Immun 66:1783– (1999) Detection of Als proteins on the cell wall of Can-
1786 dida albicans in murine tissues. Infect Immun 67:4251–
Gale CA, Bendel CM, McClellan M, Hauser M, Becker JM, 4255
Berman J, Hostetter MK (1998) Linkage of adhesion, Hu CJ, Bai C, Zheng XD, Wang YM, Wang Y (2002) Charac-
filamentous growth, and virulence in Candida albicans terization and functional analysis of the siderophore-
to a single gene, INT1. Science 279:1355–1358 iron transporter CaArn1p in Candida albicans. J Biol
Garcia-Sanchez S, Aubert S, Iraqui I, Janbon G, Ghigo JM, Chem 277:30598–30605
d’Enfert C (2004) Candida albicans biofilms: a devel- Huang Q, Liu D, Majewski P, Schulte LC, Korn JM, Young RA,
opmental state associated with specific and stable gene Lander ES, Hacohen N (2001) The plasticity of den-
expression patterns. Eukaryot Cell 3:536–545 dritic cell responses to pathogens and their compo-
Gaur NK, Klotz SA (1997) Expression, cloning, and charac- nents. Science 294:870–875
terization of a Candida albicans gene, ALA1, that con- Hube B (2004) From commensal to pathogen: stage- and
fers adherence properties upon Saccharomyces cere- tissue-specific gene expression of Candida albicans.
visiae for extracellular matrix proteins. Infect Immun Curr Opin Microbiol 7:336–341
65:5289–5294 Hube B, Naglik J (2002) Extracellular Hydrolases In:
Gaur NK, Klotz SA, Henderson RL (1999) Overexpression Calderone RA (ed) Candida and candidiasis. ASM
of the Candida albicans ALA1 gene in Saccharomyces Press, Washington, DC, pp 3–13
cerevisiae results in aggregation following attachment Hube B, Monod M, Schofield DA, Brown AJP, Gow NAR
of yeast cells to extracellular matrix proteins, adher- (1994) Expression of seven members of the gene family
ence properties similar to those of Candida albicans. encoding secretory aspartyl proteinases in Candida
Infect Immun 67:6040–6047 albicans. Mol Microbiol 14:87–99
Ghannoum MA (2000) Potential role of phospholipases in Hube B, Stehr F, Bossenz M, Mazur A, Kretschmar M,
virulence and fungal pathogenesis. Clin Microbiol Rev Schafer W (2000) Secreted lipases of Candida albicans:
13:122–143 cloning, characterisation and expression analysis of
Gola S, Martin R, Walther A, Dunkler A, Wendland J (2003) a new gene family with at least ten members. Arch
New modules for PCR-based gene targeting in Candida Microbiol 174:362–374
albicans: rapid and efficient gene targeting using 100 Hube B, Hess D, Baker CA, Schaller M, Schafer W, Dolan
bp of flanking homology region. Yeast 20:1339–1347 JW (2001) The role and relevance of phospholipase D1
Gow NA, Brown AJ, Odds FC (2002) Fungal morphogenesis during growth and dimorphism of Candida albicans.
and host invasion. Curr Opin Microbiol 5:366–371 Microbiology 147:879–889
Gow NA, Knox Y, Munro CA, Thompson WD (2003) In- Ibata-Ombetta S, Idziorek T, Trinel PA, Poulain D, Jouault T
fection of chick chorioallantoic membrane (CAM) as (2003) Candida albicans phospholipomannan pro-
a model for invasive hyphal growth and pathogenesis motes survival of phagocytozed yeasts through
of Candida albicans. Med Mycol 41:331–338 modulation of Bad phosphorylation and macrophage
Green CB, Cheng G, Chandra J, Mukherjee P, Ghan- apoptosis. J Biol Chem 278(15):13086–13093
noum MA, Hoyer LL (2004) RT-PCR detection of Can- Jones T, Federspiel NA, Chibana H, Dungan J, Kalman S,
dida albicans ALS gene expression in the reconstituted Magee BB, Newport G, Thorstenson YR, Agabian N,
Magee PT et al. (2004) The diploid genome sequence Magee BB, Legrand M, Alarco AM, Raymond M, Magee PT
of Candida albicans. Proc Natl Acad Sci USA 101:7329– (2002) Many of the genes required for mating in Sac-
7334 charomyces cerevisiae are also required for mating in
Kauffman S, Roemer T, Breton A, Veillette K, Sherrill TP, Candida albicans. Mol Microbiol 46:1345–1351
Becker JM (2004) Mouse model utilizing conditional Martchenko M, Alarco AM, Harcus D, Whiteway M (2004)
gene expression in Candida albicans. In: Abstr Vol 7th Superoxide dismutases in Candida albicans: transcrip-
American Society for Microbiology Conf Candida and tional regulation and functional characterization of
Candidiasis, 18–22 March 2004, Austin, TX, Abstr S7:2 the hyphal-induced SOD5 gene. Mol Biol Cell 15:456–
Kellis M, Patterson N, Endrizzi M, Birren B, Lander ES 467
(2003) Sequencing and comparison of yeast species McKinney JD, Honer zu BK, Munoz-Elias EJ, Miczak A,
to identify genes and regulatory elements. Nature Chen B, Chan WT, Swenson D, Sacchettini JC, Jacobs
423:241–254 WR Jr, Russell DG (2000) Persistence of Mycobacterium
Klis FM, Mol P, Hellingwerf K, Brul S (2002) Dynamics of tuberculosis in macrophages and mice requires the gly-
cell wall structure in Saccharomyces cerevisiae. FEMS oxylate shunt enzyme isocitrate lyase. Nature 406:735–
Microbiol Rev 26:239–256 738
Knight SA, Lesuisse E, Stearman R, Klausner RD, Dancis A Miller MG, Johnson AD (2002) White-opaque switching in
(2002) Reductive iron uptake by Candida albicans: role Candida albicans is controlled by mating-type locus
of copper, iron and the TUP1 regulator. Microbiology homeodomain proteins and allows efficient mating.
148:29–40 Cell 110:293–302
Korting HC, Patzak U, Schaller M, Maibach HI (1998) Mio T, AdachiShimizu M, Tachibana Y, Tabuchi H, Inoue SB,
A model of human cutaneous candidosis based on Yabe T, YamadaOkabe T, Arisawa M, Watanabe T, Ya-
reconstructed human epidermis for the light and madaOkabe H (1997) Cloning of the Candida albicans
electron microscopic study of pathogenesis and homolog of Saccharomyces cerevisiae GSC1/FKS1 and
treatment. J Infect Dis 36:259–267 its involvement in beta-1,3-glucan synthesis. J Bacte-
Kvaal C, Lachke SA, Srikantha T, Daniels K, McCoy J, riol 179:4096–4105
Soll DR (1999) Misexpression of the opaque-phase- Moran G, Stokes C, Thewes S, Hube B, Coleman D, Sulli-
specific gene PEP1 (SAP1) in the white phase of van D (2004) Comparative genomics using Candida
Candida albicans confers increased virulence in albicans DNA microarrays reveals absence and diver-
a mouse model of cutaneous infection. Infect Immun gence of virulence associated genes in Candida dublin-
67:6652–6662 iensis. Microbiology 150:3363–3382
Lan CY, Newport G, Murillo LA, Jones T, Scherer S, Morschhäuser J, Michel S, Staib P (1999) Sequential gene
Davis RW, Agabian N (2002) Metabolic specialization disruption in Candida albicans by FLP-mediated site-
associated with phenotypic switching in Candida specific recombination. Mol Microbiol 32:547–556
albicans. Proc Natl Acad Sci USA 99:14907–14912 Muhlschlegel FA, Fonzi WA (1997) PHR2 of Candida albi-
Lan CY, Rodarte G, Murillo LA, Jones T, Davis RW, Dungan J, cans encodes a functional homolog of the pH-regulated
Newport G, Agabian N (2004) Regulatory networks gene PHR1 with an inverted pattern of pH-dependent
affected by iron availability in Candida albicans. Mol expression. Mol Cell Biol 17:5960–5967
Microbiol 53:1451–1469 Mullick A, Elias M, Harakidas P, Marcil A, Whiteway M,
Lee SA, Wormsley S, Kamoun S, Lee AF, Joiner K, Wong B Ge B, Hudson TJ, Caron AW, Bourget L, Picard S et al.
(2003) An analysis of the Candida albicans genome (2004) Gene expression in HL60 granulocytoids and
database for soluble secreted proteins using computer- human polymorphonuclear leukocytes exposed to
based prediction algorithms. Yeast 20:595–610 Candida albicans. Infect Immun 72:414–429
Li F, Palecek SP (2003) EAP1, a Candida albicans gene in- Munro CA, Gow NAR (2001) Chitin synthesis in human
volved in binding human epithelial cells. Eukaryot Cell pathogenic fungi. Med Mycol 39 suppl 1:41–53
2:1266–1273 Munro CA, Schofield DA, Gooday GW, Gow NAR (1998)
Li D, Bernhardt J, Calderone R (2002) Temporal expres- Regulation of chitin synthesis during dimorphic
sion of the Candida albicans genes CHK1 and CSSK1, growth of Candida albicans. Microbiology 144:391–
adherence, and morphogenesis in a model of reconsti- 401
tuted human esophageal epithelial candidiasis. Infect Munro CA, Whitton RK, Hughes HB, Rella M, Selvaggini S,
Immun 70:1558–1565 Gow NAR (2003) CHS8-a fourth chitin synthase gene
Lopez-Ribot JL, Casanova M, Murgui A, Martinez JP (2004) of Candida albicans contributes to in vitro chitin syn-
Antibody response to Candida albicans cell wall anti- thase activity, but is dispensable for growth. Fungal
gens. FEMS Immunol Med Microbiol 41:187–196 Genet Biol 40:146–158
Lorenz MC, Fink GR (2001) The glyoxylate cycle is required Munro CA, Bates S, Buurman ET, Hughes HB, Mac-
for fungal virulence. Nature 412:83–86 Callum DM, Bertram G, Atrih A, Ferguson MA,
Lotz H, Sohn K, Brunner H, Muhlschlegel FA, Rupp S (2004) Bain JM, Brand A et al. (2005) Mnt1p and Mnt2p
RBR1, a novel pH-regulated cell wall gene of Can- of Candida albicans are partially redundant α-1,2-
dida albicans, is repressed by RIM101 and activated mannosyltransferases that participate in O-linked
by NRG1. Eukaryot Cell 3:776–784 mannosylation and are required for adhesion and
Lubkowitz MA, Hauser L, Breslav M, Naider F, Becker JM virulence. J Biol Chem 280:1051–1060
(1997) An oligopeptide transport gene from Candida Murray AE, Lies D, Li G, Nealson K, Zhou J, Tiedje JM (2001)
albicans. Microbiology 143:387–396 DNA/DNA hybridization to microarrays reveals gene-
Magee PT, Scherer S (1998) Genome mapping and gene specific differences between closely related microbial
discovery in Candida albicans. ASM News 64:505–511 genomes. Proc Natl Acad Sci USA 98:9853–9858
Naglik JR, Newport G, White TC, Fernandes-Naglik LL, tion: homology, disruption and phenotypic analysis of
Greenspan JS, Greenspan D, Sweet SP, Challacombe SJ, CTN3 gene. Fungal Genet Biol 41:783–793
Agabian N (1999) In vivo analysis of secreted aspartyl Ramanan N, Wang Y (2000) A high-affinity iron perme-
proteinase expression in human oral candidiasis. In- ase essential for Candida albicans virulence. Science
fect Immun 67:2482–2490 288:1062–1064
Naglik JR, Challacombe SJ, Hube B (2003a) Candida al- Reuß O, Vik Å, Kolter R, Morschhäuser J (2004) The SAT1
bicans secreted aspartyl proteinases in virulence and flipper, an optimized tool for gene disruption in Can-
pathogenesis. Microbiol Mol Biol Rev 67:400–428 dida albicans. In: Abstr Vol 7th American Society for
Naglik JR, Rodgers CA, Shirlaw PJ, Dobbie JL, Fernandes- Microbiology Conf Candida and Candidiasis, 18–22
Naglik LL, Greenspan D, Agabian N, Challacombe SJ March 2004, Austin, TX Rodriguez GM, Smith I (2003)
(2003b) Differential expression of Candida albicans se- Mechanisms of iron regulation in mycobacteria: role
creted aspartyl proteinase and phospholipase B genes in physiology and virulence. Mol Microbiol 47:1485–
in humans correlates with active oral and vaginal in- 1494
fections. J Infect Dis 188:469–479 Roemer T, Jiang B, Davison J, Ketela T, Veillette K, Breton A,
Naglik J, Albrecht A, Bader O, Hube B (2004) Candida pro- Tandia F, Linteau A, Sillaots S, Marta C et al. (2003)
teinases and virulence. Cell Microbiol 6:915–926 Large-scale essential gene identification in Candida
Nantel A, Dignard D, Bachewich C, Harcus D, Marcil A, albicans and applications to antifungal drug discovery.
Bouin AP, Sensen CW, Hogues H, van het Hoog M, Gor- Mol Microbiol 50:167–181
don P et al. (2002) Transcription profiling of Candida Rooney PJ, Klein BS (2002) Linking fungal morphogenesis
albicans cells undergoing the yeast-to-hyphal transi- with virulence. Cell Microbiol 4:127–137
tion. Mol Biol Cell 13:3452–3465 Rubin-Bejerano I, Fraser I, Grisafi P, Fink GR (2003) Phago-
Ng VH, Cox JS, Sousa AO, MacMicking JD, McKinney JD cytosis by neutrophils induces an amino acid depriva-
(2004) Role of KatG catalase-peroxidase in mycobacte- tion response in Saccharomyces cerevisiae and Can-
rial pathogenesis: countering the phagocyte oxidative dida albicans. Proc Natl Acad Sci USA 100:11007–
burst. Mol Microbiol 52:1291–1302 11012
Odds FC (1988) Candida and candidosis, 2nd edn. Bailliere Ruckdeschel K (2002) Immunomodulation of macrophages
Tindal, London Odds FC, Calderone RA, Hube B, by pathogenic Yersinia species. Arch Immunol Ther
Nombela C (2003) Virulence in Candida albicans: Exp (Warszawa) 50:131–137
views and suggestions from a peer-group workshop. Samuel JE, Kiss K, Varghees S (2003) Molecular Pathogen-
ASM News 69:54–55 esis of Coxiella burnetii in a Genomics Era. Ann N Y
Odds FC, Brown AJ, Gow NA (2004) Candida albicans Acad Sci 990:653–663
genome sequence: a platform for genomics in the Saporito-Irwin SM, Birse CE, Sypherd PS, Fonzi WA (1995)
absence of genetics. Genome Biol 5:230–232 PHR1, a pH regulated gene of Candida albicans, is
Okutomi T, Abe S, Tansho S, Wakabayashi H, Kawase K, required for morphogenesis. Mol Cell Biol 15:601–613
Yamaguchi H (1997) Augmented inhibition of growth Saville SP, Lazzell AL, Monteagudo C, Lopez-Ribot JL (2003)
of Candida albicans by neutrophils in the presence Engineered control of cell morphology in vivo reveals
of lactoferrin. FEMS Immunol Med Microbiol 18:105– distinct roles for yeast and filamentous forms of Can-
112 dida albicans during infection. Eukaryot Cell 2:1053–
Orozco AS, Zhou X, Filler SG (2000) Mechanisms of the 1060
proinflammatory response of endothelial cells to Can- Schaller M, Schafer W, Korting HC, Hube B (1998) Differ-
dida albicans infection. Infect Immun 68:1134–1141 ential expression of secreted aspartyl proteinases in
Peltroche-Llacsahuanga H, Schnitzler N, Schmidt S, Tin- a model of human oral candidosis and in patient sam-
telnot K, Lutticken R, Haase G (2000) Phagocytosis, ples from the oral cavity. Mol Microbiol 29:605–615
oxidative burst, and killing of Candida dubliniensis Schaller M, Schackert C, Korting HC, Januschke E, Hube
and Candida albicans by human neutrophils. FEMS B (2000) Invasion of Candida albicans correlates with
Microbiol Lett 191:151–155 expression of secreted aspartic proteinases during ex-
Perkins-Balding D, Ratliff-Griffin M, Stojiljkovic I (2004) perimental infection of human epidermis. J Invest Der-
Iron transport systems in Neisseria meningitidis. Mi- matol 114:712–717
crobiol Mol Biol Rev 68:154–171 Schaller M, Bein M, Korting HC, Baur S, Hamm G,
Phan QT, Belanger PH, Filler SG (2000) Role of hyphal for- Monod M, Beinhauer S, Hube B (2003) The secreted
mation in interactions of Candida albicans with en- aspartyl proteinases Sap1 and Sap2 cause tissue
dothelial cells. Infect Immun 68:3485–3490 damage in an in vitro model of vaginal candidiasis
Pitarch A, Sanchez M, Nombela C, Gil C (2002) Sequential based on reconstituted human vaginal epithelium.
fractionation and two-dimensional gel analysis unrav- Infect Immun 71:3227–3234
els the complexity of the dimorphic fungus Candida Schaller M, Boeld U, Oberbauer S, Hamm G, Hube B,
albicans cell wall proteome. Mol Cell Proteomics 1:967– Korting HC (2004) Polymorphonuclear leukocytes
982 (PMNs) induce protective Th1-type cytokine epithe-
Prigneau O, Porta A, Poudrier JA, Colonna-Romano S, lial responses in an in vitro model of oral candidosis.
Noel T, Maresca B (2003) Genes involved in beta- Microbiology 150:2807–2813
oxidation, energy metabolism and glyoxylate cycle Schofield DA, Westwater C, Warner T, Nicholas PJ,
are induced by Candida albicans during macrophage Paulling EE, Balish E (2003) Hydrolytic gene expres-
infection. Yeast 20:723–730 sion during oroesophageal and gastric candidiasis in
Prigneau O, Porta A, Maresca B (2004) Candida albicans immunocompetent and immunodeficient gnotobiotic
CTN gene family is induced during macrophage infec- mice. J Infect Dis 188:591–599
Schoolnik GK (2002) Functional and comparative genomics Stoldt VR, Sonneborn A, Leuker CE, Ernst JF (1997) Efg1,
of pathogenic bacteria. Curr Opin Microbiol. 5:20– an essential regulator of morphogenesis of the human
26 pathogen Candida albicans, is a member of a conserved
Shen J, Guo W, Kohler JR (2004) A heterologous dominant class of bHLH proteins regulating morphogenetic pro-
selectable marker for C. albicans. In: Abstr Vol 7th cesses in fungi. EMBO J 16:1982–1991
American Society for Microbiology Conf Candida and Sundstrom P (2002) Adhesion in Candida spp. Cell Micro-
Candidiasis, 18–22 March 2004, Austin, TX biol 4:461–469
Sheppard DC, Yeaman MR, Welch WH, Phan QT, Fu Y, Suzuki S (2002) Serological differences among pathogenic
Ibrahim AS, Filler SG, Zhang M, Waring AJ, Edwards Candida spp. In: Calderone RA (eds) Candida and can-
JE Jr (2004) Functional and structural diversity in the didosis. ASM Press, Washington DC, pp 29–36
Als protein family of Candida albicans. J Biol Chem Uhl MA, Biery M, Craig N, Johnson AD (2003) Haplo-
279:30480–30489 insufficiency-based large-scale forward genetic analy-
Smail EH, Cronstein BN, Meshulam T, Esposito AL, Rug- sis of filamentous growth in the diploid human fungal
geri RW, Diamond RD (1992) In vitro, Candida al- pathogen C. albicans. EMBO J 22:2668–2678
bicans releases the immune modulator adenosine and Wakabayashi H, Abe S, Teraguchi S, Hayasawa H, Ya-
a second, high-molecular weight agent that blocks neu- maguchi H (1998) Inhibition of hyphal growth of
trophil killing. J Immunol 148:3588–3595 azole-resistant strains of Candida albicans by triazole
Sohn K, Urban C, Brunner H, Rupp S (2003) EFG1 is a major antifungal agents in the presence of lactoferrin-
regulator of cell wall dynamics in Candida albicans related compounds. Antimicrob Agents Chemother
as revealed by DNA microarrays. Mol Microbiol 42:1587–1591
47:89–102 Whittam TS, Bumbaugh AC (2002) Inferences from whole-
Soll DR (2002) Phenotypic switching. In: Calderone RA genome sequences of bacterial pathogens. Curr Opin
(eds) Candida and candidosis. ASM Press, Washing- Genet Dev 12:719–725
ton DC, pp 123–142 Wilson RB, Davis D, Enloe BM, Mitchell AP (2000)
Staab JF, Bradway SD, Fidel PL, Sundstrom P (1999) A recyclable Candida albicans URA3 cassette for PCR
Adhesive and mammalian transglutaminase substrate product-directed gene disruptions. Yeast 16:65–70
properties of Candida albicans Hwp1. Science Wollina U, Kunkel W, Bulling L, Funfstuck C, Knoll B, Ven-
283:1535–1538 newald I, Hipler UC (2004) Candida albicans-induced
Stehr F, Felk A, Gacser A, Kretschmar M, Mahnss B, Neu- inflammatory response in human keratinocytes. My-
ber K, Hube B, Schafer W (2004) Expression analysis of coses 47:193–199
the Candida albicans lipase gene family during exper- Zink S, Nass T, Rosen P, Ernst JF (1996) Migration of the
imental infections and in patient samples. FEMS Yeast fungal pathogen Candida albicans across endothelial
Res 4:401–408 monolayers. Infect Immun 64:5085–5091
10 Integration of Metabolism with Virulence in Candida albicans
A.J.P. Brown1
CONTENTS sites in the host, including the skin, mouth, gas-

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 185 trointestinal and urogenital tracts, bloodstream,
II. The Experimental Dissection of Candida kidneys, liver and other internal organs (Ruhnke
albicans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 2002; Filler and Kullberg 2002; Kullberg and Filler
III. Overview of Metabolism in Candida albicans 187 2002).
IV. Conservation of a Metabolic Response
in Candida albicans . . . . . . . . . . . . . . . . . . . . 188 Risk factors which are thought to predispose
V. Integration of Metabolism with Virulence . . 192 patients to candidemia include protracted neu-
A. Amino Acid Metabolism . . . . . . . . . . . . . 192 tropenia (reflecting the importance of neutrophils
B. Carbon Metabolism . . . . . . . . . . . . . . . . . 193 in host defences), the use of broad-spectrum
VI. Co-Evolution of Virulence and Fitness . . . . . 197
VII. Conclusions and Future Perspectives . . . . . . 198 antibiotics (which eliminate the endogenous
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 bacterial flora) and the application of indwelling
intravenous devices such as catheters (Wey et al.
1989; Kullberg and Filler 2002). The significance
Abbreviations: 3AT, 3-aminotriazole; bHLH, β- of catheters relates to the ability of C. albicans
helix loop helix domain; bZIP, leucine zipper do- to form biofilms on the surfaces of these devices
main; GCN response, general amino acid control (Kumamoto 2002; Douglas 2003), which can
become a source of recurrent infection. Further-
more, C. albicans cells growing within biofilms are
I. Introduction relatively resistant to antifungal agents (Baillie and
Douglas 1999). Hence, biofilms represent another
Candida albicans is the major systemic pathogen of medically relevant microenvironment in which C.
humans (Odds 1988; Calderone 2002). This fungus albicans can thrive. Of particular relevance to this
exists as a relatively harmless commensal organ- review is the observation that the fungus adjusts
ism in the oral cavity and gastrointestinal tracts of its metabolic activities during growth in biofilms
at least half of all individuals. However, when the (Garcia-Sanchez et al. 2004).
defences of the host become compromised, C. albi- Several factors are thought to promote the
cans can cause mucocutaneous infections such as virulence of C. albicans (Odds 1994; Haynes 2001;
oral or vaginal candidiasis. About 75% of women Calderone and Fonzi 2001; van Burik and Magee
suffer at least one vaginal Candida infection in their 2001). Yeast-(pseudo)hyphal morphogenesis is
lifetime, and about 5% of these infections are re- thought to promote fungal dissemination and
current. In severely immunocompromised individ- invasion (Gow et al. 2002; Saville et al. 2003).
uals, C. albicans can establish deep-seated systemic Adhesins are thought to promote adherence
infections. For example, in patients undergoing to host tissue and colonisation (Hoyer 2001;
chemotherapy or organ transplant surgery, the fun- Sundstrum 2002). Secreted aspartyl proteinases
gus is able to disseminate via the bloodstream and and lipases may promote invasion, counteract
colonise internal organs. In some patient groups, host defences and provide nutrients (Hube and
one-third to one-half of these systemic infections Naglik 2002). High-frequency switching between
are fatal. Hence, depending upon the immunocom- different phenotypic forms might help the fungus
petence of the host, C. albicans can infect numerous to evade host defences (Soll 2002). These virulence
1Aberdeen Fungal Group, School of Medical Sciences, Institute of
attributes are thought to be required to differing
Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen extents during disease establishment and pro-
AB25 2ZD, UK gression (Odds 1994; Brown and Gow 1999). The
The Mycota XIII
Fungal Genomics
186 A.J.P. Brown
impact of genomics upon our understanding of C. classical genetic approaches in this pathogen. In
albicans pathogenicity is discussed by Munro and the absence of genetic screens, the development of
colleagues in Chap. 9 (this volume). genomic screens has proved especially important
Some years ago we predicted that the regulation for the dissection of C. albicans pathobiology.
of metabolism and virulence might be mechanis- A second problem hindered the molecular
tically linked in C. albicans (Brown and Gow 1999; dissection of C. albicans until the mid-1990s.
Brown et al. 2000). We reasoned that these links The CUG codon, which is normally decoded
might allow this pathogen to adjust its metabolic as leucine, is translated as serine in C. albicans
programme in parallel with its portfolio of viru- (Santos et al. 1993). As a result, most standard
lence attributes in response to the new microen- reporter genes such as Escherichia coli lacZ, firefly
vironments it encounters during disease establish- luciferase or jellyfish GFP are not expressed in
ment and progression. In this chapter I review re- a functional form in C. albicans. This problem was
cent studies which address this prediction and the circumvented by the development of specialised
significant impact which C. albicans genomics has reporter genes for C. albicans, such as Renilla
had in this field. Genomics has revealed unexpected reniformis luciferase, codon-optimised GFP and
links between metabolism and virulence in C. albi- Streptococcus thermophilus lacZ (Srikantha et al.
cans which were not exposed by pre-genomic ex- 1996; Cormack et al. 1997; Uhl and Johnson 2001).
perimentation. Other essential components of the molecular
toolbox were developed in the 1990s (De Backer
et al. 2000; Berman and Sudbery 2002). These
included improved plasmid vectors, ectopic ex-
II. The Experimental Dissection pression vectors and regulatable promoter systems
of Candida albicans (Bailey et al. 1996; Leuker et al. 1997; Care et al.
1999). Significantly, robust procedures were estab-
C. albicans has been the focus of a large number lished for the generation of C. albicans mutants
of research studies because of its clinical impor- (Fonzi and Irwin 1993), which were improved
tance, and the number of publications on Candida further by the development of PCR-based gene
species has increased steadily over the previous disruption methods (Wilson et al. 1999, 2000;
two decades (Calderone 2002). This increase has Enloe et al. 2000). These tools were exploited
been due, at least in part, to the development of an widely for gene analysis before the C. albicans
improved molecular toolbox for what used to be genomics era, and they remain important for the
an experimentally intransigent fungus. The aim of functional analysis of genes identified by genome
this section is to provide a brief historical perspec- sequencing.
tive on the development of this molecular toolbox, C. albicans genome sequencing began in the
which has led to the establishment of C. albicans private sector with Incyte in the mid-1990s, but
genomics. this was followed shortly thereafter by a public
C. albicans is a diploid fungus, and it was sequencing programme at the Stanford DNA
thought to be asexual (Magee 1998). However, Sequencing and Technology Center, which was
C. albicans is indeed able to mate (Magee and funded by the US National Institute for Dental
Magee 2000; Hull et al. 2000; Miller and Johnson and Cranofacial Research and the Burroughs
2002), and the tetraploid products of these matings Wellcome Fund. The latest genome sequence
can be induced to shed chromosomes, such that assembly from the Stanford Group represents the
diploid recombinants can be isolated (Bennett diploid genome sequence of C. albicans (Assembly
and Johnson 2003). Hence, a parasexual cycle 19; Jones et al. 2004). Assembly 19 includes
has now been defined. However, meiosis has not a haploid set of sequence contigs and a set of allelic
been demonstrated in C. albicans. C. albicans contigs which describe the heterozygosities in the
genome sequencing has revealed homologues of genome of C. albicans strain SC5314 (a clinical
many Saccharomyces cerevisiae genes involved in isolate from Bristol-Myers Squibb). In addition
the sexual cycle, but C. albicans appears to lack to the Stanford sequencing website (http://www-
critical factors which are required for meiosis in sequence.stanford.edu/group/candida/), two
budding yeast (Tzung et al. 2001). Hence, no true main C. albicans genome databases have been
sexual cycle has been established for C. albicans, established. CandidaDB was first released in
and this has prevented the efficient application of 2001 (http://genolist.pasteur.fr/CandidaDB; d’En-
Metabolism and Virulence in C. albicans 187
fert et al. 2005), and the Candida Genome the pathways involved in ergosterol biosynthesis
Database was released in 2004 (http://genome- and cell wall biogenesis, both of which are anti-
www.stanford.edu/fungi/Candida/). fungal targets. Even now, C. albicans metabolism
The Stanford Group generously released their is not addressed directly in the latest definitive
C. albicans genome sequence data at various text on Candida and candidiasis (Calderone 2002).
stages of the sequencing process (http://www. However, pathways of central and amino acid
sequence.stanford.edu/group/candida/). These metabolism have been highlighted as potential
preliminary sequence data provided a strong antifungal targets by large-scale gene deletion
platform for many groups to initiate C. albicans studies in C. albicans (Roemer et al. 2003) and
transcript profiling, proteomics and large-scale by transcript profiling studies (Lorenz and Fink
gene deletion studies even before sequence Assem- 2001, 2002). With respect to antifungal therapy,
bly 19 was released (Lane et al. 2001; Murad et al. the priority is to kill the fungus, rather than to
2001b; Cowen et al. 2002; Lan et al. 2002; Nantel inhibit its virulence. Hence, metabolic functions
et al. 2002; Pitarch et al. 2002; Bruneau et al. 2003; which are required for growth and survival in
Chauhan et al. 2003; Fradin et al. 2003; Roemer the host represent potential antifungal targets. Of
et al. 2003; Rubin-Bejerano et al. 2003; Tsong et al. course, the validity of such targets and the efficacy
2003; Garcia-Sanchez et al. 2004; Hernandez et al. of drugs which hit these targets will depend upon
2004; Yin et al. 2004). the degree of cross-reactivity with the host. Never-
Initially, the experimental dissection of C. albi- theless, an increased understanding of C. albicans
cans was confined to biochemical and immunolog- metabolism and its regulation should facilitate the
ical approaches. The molecular era, which began development of new antifungal therapies in the
in earnest in the 1990s, saw the isolation and char- long term.
acterisation of many genes involved in C. albicans To date, S. cerevisiae has provided a reason-
pathobiology. However, in the absence of classi- able metabolic paradigm for C. albicans. The path-
cal genetics, these genes were often targeted for ways of central carbon metabolism are conserved in
study on the basis of their sequence relatedness to, this fungus, including the glycolytic, gluconeogenic
or functional interactions with S. cerevisiae genes, and pentose phosphate pathways, and the tricar-
and their analysis was mostly restricted to their pre- boxylic and glyoxylate cycles (Odds 1988; Jones
dicted functions. As a result, unexpected roles for et al. 2004). Pathways for the generation of stor-
these genes often remained undetected. Arguably age and cell wall carbohydrates are conserved. Fur-
one of the greatest impacts of the C. albicans ge- thermore, pathways of amino acid, lipid and nu-
nomics era, which took off at the beginning of the cleotide assimilation and anabolism appear to be
new millennium, was the ability to examine the conserved. However, metabolic differences do ex-
global roles of specific genes in a relatively unbi- ist between C. albicans and S. cerevisiae, the most
ased fashion. As described below, this led to the obvious of which relates to their patterns of sugar
finding that the regulation of metabolic and vir- utilisation. Indeed, differences in the patterns of
ulence functions appears to be integrated in this carbohydrate assimilation are used routinely to dis-
pathogen. tinguish C. albicans from other microbes in the
clinical setting (Wickerham 1951; Williamson et al.
1986).
With regard to the regulation of metabolism,
III. Overview of Metabolism in Candida the S. cerevisiae paradigm is less robust. This is
albicans hardly surprising, given that the pathogen C. al-
bicans and the relatively benign S. cerevisiae have
Investigations of C. albicans metabolism began evolved in fundamentally different niches. C. al-
over half a century ago (e.g. Van Neil and Cohen bicans ferments glucose to ethanol and displays
1942), and enzymes of central carbon metabolism glucose repression. Also, at least some glucose-
were characterised as long ago as 1960 by Rao repressible functions appear to be regulated by the
and co-workers. However, in general, studies of C. transcriptional repressor Mig1 in a fashion anal-
albicans metabolism have focused either directly ogous to that of S. cerevisiae (Murad et al. 2001a,
or indirectly upon virulence attributes or the mode b). Furthermore, C. albicans contains a homologue
of action of antifungal drugs (Odds 1988). For of S. cerevisiae Snf1 (Petter et al. 1997), which is
example, considerable attention has been paid to required for the derepression of glucose-repressed
188 A.J.P. Brown
genes in budding yeast. However, C. albicans Mig1 and this signal stimulates the induction of almost
displays several differences vis-à-vis S. cerevisiae all amino acid biosynthetic pathways. This broad
Mig1, including the lack of a putative Snf1 phos- metabolic response can be activated by starvation
phorylation site (Zaragoza et al. 2000). Also, S. for even a single amino acid.
cerevisiae snf1 mutants are viable, whereas Snf1 The GCN response has been well characterised
appears to be essential for viability in C. albicans at the molecular level in S. cerevisiae (Hinnebusch
(Petter et al. 1997), suggesting that Snf1 might ex- 1988; Natarajan et al. 2001; Hinnebusch and Natara-
ecute additional functions in this yeast. Therefore, jan 2002). Briefly, amino acid starvation leads to
significant differences do appear to exist between the intracellular accumulation of uncharged tR-
C. albicans and S. cerevisiae with respect to glucose NAs, which bind to the histidyl tRNA synthetase-
repression mechanisms. like domain of Gcn2 (Wek et al. 1995). This ac-
The differences extend to the regulation of hy- tivates the protein kinase activity of Gcn2, which
poxic gene expression in C. albicans and S. cere- then phosphorylates the α-subunit of the transla-
visiae. For example, the transcriptional regulator tion initiation factor eIF2 (Dever et al. 1992). eIF2a
Rox1 controls hypoxic gene expression in S. cere- phosphorylation increases the affinity of eIF2 for
visiae in a haem-dependent fashion (Kastaniotis eIF2B, its guanine nucleotide exchange factor, and
and Zitomer 2000). In contrast, its orthologue in C. inhibits the regeneration of recharged eIF2-GTP re-
albicans, Rfg1, regulates yeast-hypha morphogen- quired for translation initiation (Krishnamoorthy
esis, but not hypoxic gene expression (Kadosh and et al. 2001). This leads to a decrease in the overall
Johnson 2001). rate of translation initiation in the yeast cell, but to
Further significant differences between C. albi- an increase in the translation of the GCN4 mRNA.
cans and S. cerevisiae metabolism have been sug- The translation of the GCN4 mRNA is regu-
gested by the genome sequence data. Jones et al. lated by four short upstream open reading frames
(2004) indicate that unlike S. cerevisiae, C. albi- (uORFs), which lie in its unusually long 5 -leader
cans possesses both mitochondrial- and nuclear- region (Mueller and Hinnebusch 1994). Essentially,
encoded subunits of electron transport complex this 5 -leader normally acts to repress the trans-
I, a pyruvate dehydrogenase kinase, a large lipase lation of the main Gcn4-encoding open reading
gene family and additional enzymes involved in frame on the GCN4 mRNA under non-starvation
fatty acid and amino acid catabolism. These dif- conditions. However, in response to amino acid
ferences might reflect an increased emphasis on starvation, this repression is released when eIF2-
respiratory and oxidative metabolism in C. albi- GTP levels are reduced (Mueller and Hinnebusch
cans (Jones et al. 2004). This view is consistent with 1994). The resultant increase in GCN4 mRNA trans-
the observation that C. albicans converts a lower lation leads to an elevation in the abundance of the
proportion of glucose to ethanol in comparison to Gcn4 protein. Gcn4 is a bZIP transcription fac-
S. cerevisiae (Bertram et al. 1996). tor which binds as a dimer to GCRE elements (5 -
To summarise, whilst many metabolic path- TGA(C/G)TCA) located in the promoters of target
ways are conserved between S. cerevisiae and C. genes, and activates their transcription (Oliphant
albicans, these organisms display significant differ- et al. 1989; Ellenberger et al. 1992). These target
ences in metabolic regulation. Nevertheless, many genes include amino acid biosynthetic enzymes on
investigators still extrapolate from S. cerevisiae to all amino acid biosynthetic pathways, with the ex-
C. albicans when interpreting their experimental ception of the cysteine pathway (Natarajan et al.
data. 2001). Therefore, amino acid biosynthesis is in-
duced in response to amino acid starvation via
a signalling pathway involving the key regulators
Gcn2 and Gcn4.
IV. Conservation of a Metabolic Transcript profiling of amino acid starvation
Response in Candida albicans in S. cerevisiae has revealed that the GCN response
is much broader than was initially expected
To date, only one metabolic response has been com- (Natarajan et al. 2001). In these experiments the
pared directly in C. albicans and S. cerevisiae at the histidine analogue, 3-aminotriazole (3AT), was
genomic level: General Amino Acid Control, or the used to activate the GCN response. Exposure to
GCN response. The environmental stimulus which 3AT, which causes histidine starvation, is a classic
triggers the GCN response is amino acid starvation, means of activating the GCN response (Hinneb-
usch 1988). The transcript profiling experiments which was roughly analogous to the corresponding
showed that, in addition to inducing the expression S. cerevisiae transcriptome (Natarajan et al. 2001).
of amino acid biosynthetic genes, Gcn4 induces Forty-five percent of the 52 Gcn4-dependent
the expression of aminoacyl tRNA synthetases, changes identified corresponded to amino acid
amino acid transporters, vitamin biosynthetic biosynthetic enzymes on the arginine, asparagine,
pathways and glycogen biosynthetic functions, cysteine, chorismate, glutamine, glycine, histidine,
and represses the expression of ribosomal proteins isoleucine–valine, leucine, lysine, methionine, ser-
(Natarajan et al. 2001). The transcript profiling ine, threonine and tryptophan pathways (Yin et al.
also confirmed the earlier observation that Gcn4 2004). Clearly, this proteomic study confirmed
regulates purine biosynthetic functions (ADE that histidine starvation (using 3AT) invokes
genes; Rolfes and Hinnebusch 1993). Furthermore, a broad response across most (if not all) amino
the transcript profiling revealed that Gcn4 controls acid biosynthetic pathways. In addition, 9% of
the expression of numerous transcription factors the Gcn4-dependent changes represented purine
including Arg80, Leu3, Lys14, Met4, Met28, Bas1, biosynthetic enzymes (Ade proteins), and 8%
Gln3, Rtg3, Pip2, Gat1, Uga3, Mal13, Cup9 and were involved in the biosynthesis of vitamins and
Rim101. This led Natarajan et al. (2001) to propose cofactors. Proteins involved in carbon metabolism
that Gcn4 is a master regulator of gene expression and energy generation were also elevated by 3AT
in S. cerevisiae. in a Gcn4-dependent fashion. Hence, many of the
Gcn4-like proteins are conserved in other fungi changes observed by transcript profiling (Natara-
(Paluh et al. 1988; Wanke et al. 1997; Hoffmann jan et al. 2001) were reflected in the proteome (Yin
et al. 2001; Tripathi et al. 2002; Tournu et al. 2005). et al. 2004). Indeed, there was a positive correlation
Furthermore, the mRNAs encoding these proteins between the S. cerevisiae GCN transcriptome and
carry unusually long 5 -leader regions with multi- proteome (Spearman rank correlation coefficient
ple short uORFs. This suggested that these mRNAs = 0.59, p<1.89 ×10−06 ).
might be regulated at the translational level, like the An analogous proteomic experiment has
S. cerevisiae GCN4 mRNA. S. cerevisiae Gcn4 lev- been performed for C. albicans (Fig. 10.1; Yin
els are also regulated at the transcriptional level et al. 2004). Of 942 spots identified reproducibly
and via accelerated protein turnover (Kornitzer on the two-dimensional gels, 218 (23%) were
et al. 1994; Albrecht et al. 1998), but most control reproducibly elevated by 3AT, and 147 of these
appears to be executed at the translational level changes were Gcn4-dependent. Thirty percent of
(Hinnebusch and Natarajan 2002). In contrast, the the 31 Gcn4-dependent protein changes identified
expression of GCN4-like genes in C. albicans and by peptide mass fingerprinting corresponded to
Aspergillus nidulans appears to be regulated pri- amino acid biosynthetic enzymes on the arginine,
marily at the transcriptional level (Hoffmann et al. chorismate, isoleucine–valine, leucine, lysine,
2001; Tournu et al., unpublished data). Therefore, methionine and serine pathways. In addition,
subtle differences do exist between fungi with re- a significant number of C. albicans proteins
spect to their GCN responses. Hence, we have com- involved in carbon metabolism and energy gener-
pared the global GCN responses in S. cerevisiae and ation were elevated by 3AT in a Gcn4-dependent
C. albicans (Yin et al. 2004; Tournu et al., unpub- fashion. To this extent, the C. albicans GCN
lished data). Our aims were firstly to test whether proteome closely mirrored the S. cerevisiae GCN
the transcriptional responses observed by Natara- proteome. However, one interesting difference was
jan et al. (2001) are reflected in the S. cerevisiae observed between C. albicans and S. cerevisiae. All
proteome, and secondly to determine the extent of four Ade proteins identified in our S. cerevisiae
conservation of this GCN response in C. albicans. two-dimensional map were induced by 3AT. In
Initially, the S. cerevisiae Gcn4 proteome contrast, none of the seven Ade proteins identified
was examined. Protein extracts were prepared in the C. albicans two-dimensional map were
from wild-type and gcn4 cells exposed to 3AT or elevated in response to 3AT.
carrier alone. Of 956 spots identified reproducibly More recently, transcript profiling has
on the resultant two-dimensional gels, 107 were been performed on the C. albicans GCN
reproducibly elevated by 3AT, and 80 of these response (Tournu et al., unpublished data;
changes were Gcn4-dependent (Fig. 10.1; Yin et al. http://www.pasteur.fr/recherche/unites/Galar_Fun
2004). Therefore, a significant proportion (11%) gail/). About 12% of the ca. 6200 C. albicans genes
of the S. cerevisiae proteome responded to 3AT, analysed were regulated by 3AT in a Gcn4-
190 A.J.P. Brown
Fig. 10.1. Proteomic identification of S. cerevisiae and C. albicans proteins which are up-regulated by 3AT in a Gcn4-
dependent fashion
dependent manner. This was significantly less than S. cerevisiae, has been conserved in C. albicans.
the 27% of C. albicans proteins which seemed to be Second, in contrast to the situation in S. cerevisiae,
regulated in a Gcn4-dependent manner (Yin et al. purine biosynthetic genes are not regulated in
2004). Indeed, there was no positive correlation response to amino acid starvation in C. albicans.
between the C. albicans GCN transcriptome and Molecular studies have revealed another sig-
proteome (Spearman rank correlation coefficient nificant difference between the GCN responses of
= 0.09). Hence, the levels of many proteins might S. cerevisiae and C. albicans. Amino acid starvation
be post-transcriptionally regulated during the induces cellular morphogenesis in the pathogenic
GCN response in C. albicans. Consistent with this fungus, but does not do so in the benign yeast (Tri-
idea, transcript profiling revealed the induction pathi et al. 2002). The formation of pseudohyphae
of many putative peptidase genes by 3AT (APE3, by C. albicans in response to amino acid starvation
PRC1, PRC3, PRE3, PRE8, PUP2, RPN1, RPN2, is dependent upon Gcn4. Furthermore, ectopic ex-
RPN6, RPN7, RPN8, RPN10, RPT1, RPT4, RPT5, pression of Gcn4 in C. albicans stimulates pseudo-
RPT6, IPF4866), suggesting that protein turnover hyphal development in the absence of an environ-
rates increase following amino acid starvation mental stimulus. Hence, Gcn4 co-ordinates mor-
in C. albicans. Nevertheless, there were strong phogenetic and metabolic responses in this organ-
similarities between the C. albicans and S. cere- ism (Fig. 10.2A). Collectively, the data suggest sub-
visiae transcriptomes with respect to the types of tle but potentially significant differences between
cellular function which were regulated by Gcn4. the GCN responses in S. cerevisiae and C. albicans.
In both cases, amino acid biosynthetic genes were Does the GCN response contribute to the viru-
significantly elevated, and ribosomal protein genes lence of C. albicans? The inactivation of Gcn4 does
were significantly repressed (Natarajan et al. 2001; not attenuate the virulence of this fungus in the
Tournu et al., unpublished data). However, unlike mouse model of systemic candidiasis (Brand et al.
S. cerevisiae, C. albicans ADE mRNAs were not 2004). In this model, C. albicans cells are injected
induced by 3AT in a Gcn4-dependent fashion. directly into the bloodstream of the host. Hence,
Hence, our recent transcript profiling data are these experiments indicate that the GCN response
entirely consistent with the proteomic data with is not required for C. albicans to establish a systemic
respect to two important conclusions. First, in infection once it has entered the bloodstream. How-
general the global GCN response, as defined in ever, Gcn4 is required for C. albicans to form nor-
Fig. 10.2. Co-ordinated regulation of metabolism and vir- the relevance of these responses to specific medically rel-
ulence attributes in C. albicans. Genomic analyses have evant environments (below). See text for details. A Gcn4;
revealed the influence of specific regulators (box) upon B Efg1; C Tup1; D general model
metabolism (left) and virulence (right) and, in some cases,
192 A.J.P. Brown
mal biofilms (Garcia-Sanchez et al. 2004). Hence, morphogenesis remain to be defined. It is attractive
the GCN response does appear to be required for to speculate that, like Gcn4, Csy1 might mediate its
survival in at least one medically important niche. morphogenetic effects via the Ras-cAMP pathway.
It remains to be seen whether C. albicans depends Recent expression profiling of C. albicans
upon the GCN response in other infection sites, cells growing in biofilms has revealed another
for example, during oral candidiasis, vaginitis or unexpected link between amino acid metabolism
during commensalism. and virulence (Garcia-Sanchez et al. 2004). In this
study, cells growing in biofilms were compared
with those in planktonic cultures under a variety
of different growth conditions. The rationale was
V. Integration of Metabolism to identify sets of genes which were regulated
with Virulence specifically in response to growth in biofilms,
rather than in response to the conditions used to
A. Amino Acid Metabolism generate the biofilms. Amino acid biosynthetic
genes were amongst those which were consis-
As described above (Sect. IV), a detailed examina- tently up-regulated in biofilms. These included
tion of the GCN response in C. albicans has revealed ARO, CYS, HIS, ILV, MET, SER and TRP genes,
a well-defined molecular link between a morpho- suggesting that many amino acid biosynthetic
genetic and metabolic response. Amino acid star- pathways are induced during biofilm formation.
vation in C. albicans stimulates pseudohyphal de- Garcia-Sanchez et al. (2004) reasoned that this
velopment as well as the induction of amino acid induction might be mediated by the GCN re-
biosynthetic genes and other metabolic functions. sponse (Tripathi et al. 2002; Yin et al. 2004), and
Both of these responses depend upon the tran- hence they tested the effects of inactivating Gcn4
scription factor Gcn4 (Tripathi et al. 2002). Hence, upon biofilm formation. Biofilm formation was
through Gcn4, the regulation of a virulence at- significantly inhibited in C. albicans gcn4 cells,
tribute is integrated with the control of amino acid compared with control cells in which the GCN4
metabolism in C. albicans (Fig. 10.2A). This sec- gene had been reintroduced (Garcia-Sanchez et al.
tion will review further experimental observations 2004). Therefore, the GCN response appears to be
which lend weight to the view that the control required for efficient biofilm formation.
of virulence and metabolism is integrated in this As described above (Sect. I), C. albicans
pathogen. biofilms which form on the catheters of hospital
The exact molecular mechanism by which patients appear to be a relatively frequent source
Gcn4 stimulates morphogenesis in C. albicans is of bloodstream infections in these patients. The
not known. However, Gcn4 is known to activate regulation of amino acid metabolism still appears
morphogenesis specifically via the Ras-cAMP to be important when C. albicans cells enter the
signalling pathway (Tripathi et al. 2004). Mutations bloodstream. This view is supported by a global
on the MAP kinase pathway do not prevent the study of fungal gene expression during phagocy-
ability of ectopically expressed Gcn4 to stimulate tosis by human neutrophils (Rubin-Bejerano et al.
morphogenesis, but the inactivation of Efg1 blocks 2003). Neutrophils represent an important weapon
this response. Hence, Gcn4 appears to co-ordinate in the host’s defences against disseminated can-
morphogenesis with amino acid metabolism by didiasis. Following phagocytosis by neutrophils,
connecting specific signalling pathways. C. albicans cells remain in the yeast form and are
A further mechanistic link between morpho- killed, whereas the fungus can undergo morpho-
genesis and amino acid metabolism has been de- genesis inside cultured macrophages and escape
scribed recently. Csy1 is thought to be the main from these cells (Lo et al. 1997; Rubin-Bejerano
amino acid sensor in C. albicans. The inactiva- et al. 2003). The transcript profiling study of
tion of Csy1 blocks the transcriptional induction of Rubin-Bejerano et al. (2003) focussed mainly
amino acid permease genes in response to the pres- upon S. cerevisiae. However, these authors did
ence of amino acids in the growth medium (Brega show that, to a great extent, C. albicans behaves
et al. 2004). C. albicans csy1 mutants display mor- analogously to S. cerevisiae following exposure to
phogenetic defects in response to serum and pH neutrophils. Genes on the methionine and argi-
induction. However, the molecular mechanisms by nine biosynthetic pathways were induced when
which Csy1 links amino acid sensing with cellular C. albicans was exposed to human neutrophils.
The corresponding S. cerevisiae response required host must depend upon the efficient assimilation
fresh serum, because it was lost when the serum of available carbon sources in vivo. Furthermore,
was inactivated. Also, this metabolic response as described above (Sect. V.A), the fungus must ad-
required direct contact between the fungal and just its metabolic programme as it encounters new
human cells, because it was lost if a membrane microenvironments in the host. A number of stud-
separated the cells. Furthermore, the response was ies have indicated that the regulation of carbon
specific to neutrophils, and was not observed in metabolism, like amino acid metabolism, is inti-
monocytes. C. albicans genes on other amino acid mately linked to the control of virulence in C. albi-
biosynthetic pathways were generally not affected cans. This section will focus mainly on the pathways
by exposure to neutrophils (Rubin-Bejerano et al. of central carbon metabolism: glycolysis, gluconeo-
2003), suggesting that the neutrophil response is genesis and the tricarboxylic (TCA) and glyoxylate
distinct from the GCN response described above. cycles.
This might explain why blocking the GCN response Over 10 years ago it was reported that the lev-
does not inhibit disseminated candidiasis (Brand els of glycolytic mRNAs change during yeast-hypha
et al. 2004). Instead, the neutrophil response morphogenesis in C. albicans (Swoboda et al. 1994).
seems more reminiscent of stress responses, It was concluded that these changes, which were ob-
which also induce the expression of genes on the served by Northern blotting of PYK1, ADH1, PGK1
methionine and arginine biosynthetic pathways and GPM1 mRNAs, reflected the underlying phys-
(supplementary data in Enjalbert et al. 2003). Of iological changes which accompany morphogene-
course, this might reflect the potent cidal effect of sis, rather than bone fide morphogenetic regulation
neutrophils upon C. albicans cells. of these genes. A strict definition of morphogenetic
Whilst MET and ARG genes were induced when regulation, which excluded genes which are regu-
C. albicans cells were exposed to neutrophils in lated by medium composition alone, was helpful in
RPMI medium containing fresh serum, the expres- terms of identifying hypha-specific genes (Bailey
sion of ADE genes was decreased (Rubin-Bejerano et al. 1996). However, this shifted attention away
et al. 2003). This observation reinforces the idea from potential links between C. albicans physiol-
that purine and amino acid biosynthesis is regu- ogy and development. Attention has only returned
lated by distinct mechanisms in C. albicans (Yin to these potential links following the publication
et al. 2004). However, in the context of this review, of several transcript profiling studies which have
the main conclusion which can be drawn from this reinforced and extended this early observation.
study of Rubin-Bejerano et al. (2003) is that C. al- Nantel et al. (2002) published a transcript
bicans regulates amino acid biosynthetic pathways profiling study of C. albicans morphogenesis in
as part of its response to attack by neutrophils. which the behaviour of wild-type and mutant cells
A recent study has suggested that C. albicans was compared during serum-induced morpho-
responds slightly differently from S. cerevisiae genesis. Naturally, the discussion in this paper
during phagocytosis by mammalian macrophages focussed to a large extent on the behaviour of
(Lorenz et al. 2004). Under the conditions anal- hypha-specific and signalling genes and, as ex-
ysed in this study, the C. albicans cells survived pected, hypha-specific genes such as ECE1, HWP1,
phagocytosis, formed germ tubes and escaped RBT1 and SAP4, 5, 6 were induced during hyphal
from the macrophage. Transcript profiling of the development. However, of particular relevance
fungal response suggested that major changes in to this review is the fact that numerous genes
carbon metabolism accompanied phagocytosis encoding glycolytic enzymes were also shown to be
by the macrophages (discussed below), and that regulated during hyphal development, including
arginine biosynthesis was significantly induced. ENO1, FBA1, PYK1 (CDC19), TPI1 and PGI1 (Nan-
The response of MET genes was not highlighted in tel et al. 2002). Clearly, the regulation of these genes
this study (Lorenz et al. 2004). might be indirect, as implied in the early study by
Swoboda et al. (1994), and these genes do appear to
respond to physiological stimuli such as ambient
B. Carbon Metabolism growth temperature (Nantel et al. 2002). However,
Nantel and co-workers also showed that these
Carbon assimilation is essential for the generation glycolytic genes respond to the key morphogenetic
of new biomass, i.e. growth. Therefore, the rapid regulator Efg1, and this observation has been
growth of C. albicans in the immunocompromised confirmed by an independent transcript profiling
194 A.J.P. Brown
study (Doedt et al. 2004). Hence, the link between Although Efg1 has been shown to bind DNA in
morphogenetic and glycolytic regulation might be a sequence-specific fashion in vitro (Leng et al.
more direct than was originally anticipated. 2001), a consensus sequence for its normal binding
Ernst and co-workers were the first to identify site in vivo has not been defined unambiguously.
the bHLH transcription factor, Efg1, as an impor- Hence, this issue cannot be addressed by screen-
tant regulator of yeast-hypha morphogenesis in C. ing the promoters of these metabolic genes for
albicans (Stoldt et al. 1997). The inactivation or a consensus Efg1 binding site. Nevertheless, this
depletion of Efg1 severely attenuates hyphal de- metabolic regulation by Efg1 is physiologically
velopment (Lo et al. 1997; Stoldt et al. 1997), and significant, because efg1 cells are more sensitive
hence Efg1 has been described as an activator of to an inhibitor of respiration, antimycin A (Doedt
morphogenesis. Efg1 has now been shown to con- et al. 2004). Hence, Efg1 regulates metabolism as
trol chlamydospore formation as well as hyphal well as virulence attributes (Figs. 10.2B and 10.3).
development, and the inactivation of Efg1 also af- The link between central carbon metabolism
fects phenotypic switching (Stoldt et al. 1997; Son- and virulence attributes in C. albicans is further
neborn et al. 1999; Srikantha et al. 2000). Efg1 has strengthened by another transcript profiling study
also been shown to regulate the expression of ad- (Lan et al. 2002). The focus of this study was phe-
hesins such as Als3 (formally known as Als8) and notypic switching in C. albicans. C. albicans strain
Hwp1 (Sharkey et al. 1999; Lane et al. 2001; Leng WO1 switches at high frequency between white and
et al. 2001; Nantel et al. 2002). Hence, Efg1 is a key opaque forms. White cells are slightly ellipsoidal
developmental regulator which controls the expres- and their cell walls have a relatively smooth surface
sion of several virulence attributes in C. albicans when examined by scanning electron microscopy.
including yeast-hypha morphogenesis, phenotypic In contrast, opaque cells are larger than are white
switching and adhesins. Not surprisingly, the inac- cells, are more elongated, contain a large vacuole
tivation of Efg1 (albeit in combination with a cph1 and display pimples on their cell surface (Soll 2002).
mutation) attenuates the virulence of C. albicans in Significantly, clinical isolates of C. albicans from
the mouse model of systemic infection, reduces in- diseased patients undergo phenotypic switching at
vasion in epithelial infection models, and inhibits higher rates than do commensal strains (Hellstein
biofilm formation (Lo et al. 1997; Gow et al. 2003; et al. 1993). Furthermore, white cells are more vir-
Korting et al. 2003; Garcia-Sanchez et al. 2004). ulent than are opaque cells in the mouse model
Efg1 acts in concert with the structurally and of systemic candidiasis, whereas opaque cells are
functionally related bHLH factor, Efh1 (Doedt more virulent than are white cells in a cutaneous
et al. 2004). Whereas Efh1 is a transcriptional mouse model (Kvaal et al. 1997, 1999). Therefore,
activator, Efg1 appears to act mechanistically as phenotypic switching is closely associated with C.
a transcriptional repressor in C. albicans (Tebarth albicans virulence.
et al. 2003; Doedt et al. 2004). Recently, Ernst’s Lan et al. (2002) compared the global expres-
group published a detailed analysis of the Efg1 and sion patterns of the two main switch phenotypes
Efh1 regulons in C. albicans (Doedt et al. 2004). of C. albicans strain WO1 – white and opaque cells.
In this experiment, transformants which over- Well-defined sets of genes were found to be up- or
expressed Efg1 or Efh1 were included alongside down-regulated in opaque cells, compared with
efg1 and efh1 mutants. A clear picture emerged white cells during growth on rich or defined me-
in which the central metabolic pathways are dia. Genes involved in adhesion, stress responses,
regulated by Efg1 and, to a lesser extent, by Efh1. drug resistance and signalling were found to be
The glycolytic genes GLK1, HXK2, PGI1, PFK1,2, regulated. Interestingly, about one-third of the
FBA1, TPI1, GAP1, PGK1, GPM1 and ENO1 were all regulated genes encoded metabolic functions.
up-regulated by Efg1. In contrast, gluconeogenic Glycolytic genes were down-regulated in opaque
(FBP1), TCA cycle (CIT1, MDH1, FUM12, SDH12, cells (HXT3, HXT4, HXK1, PFK2 and PYK1
KGD1) and respiratory functions (ATP1, ATP17, (CDC19)), whereas TCA genes were up-regulated
PET9) were down- regulated. Therefore, Efg1 (IDP2, MDH1, MLS1). Also, genes involved in fatty
stimulates fermentative metabolism and represses acid β-oxidation were up-regulated in opaque cells
respiratory metabolism. It is not clear whether (FAA2, POX1, ECI1, FOX2, FOX3). This led Lan
Efg1 regulates these metabolic genes by acting et al. (2002) to suggest that white and opaque cells
directly at their promoters, or indirectly by con- express different metabolic programmes which
trolling the activities of other regulatory factors. might be associated with their preference for
different anatomical niches in the host. White cells Global analyses of another regulator, Tup1,
favour fermentative metabolism, whereas opaque have reinforced the view that the regulation of
cells favour respiratory metabolism (Fig. 10.3). carbon metabolism and virulence attributes are
The findings of Lan et al. (2002) are entirely mechanistically linked in C. albicans. C. albi-
consistent with those of Doedt et al. (2004). The cans Tup1 was first identified as a repressor of
inactivation of Efg1 leads to the generation of hyphal development (Braun and Johnson 1997).
opaque cells, and EFG1 is expressed at low levels C. albicans tup1 mutants grow constitutively in
in opaque cells. As described above, fermentative an elongated pseudohyphal form, even in the
metabolism is down-regulated and oxidative absence of a morphogenetic stimulus. Ectopic
metabolism is up-regulated in efg1 and opaque Tup1 expression affects white–opaque phenotypic
cells. The opposite is true in EFG1 over-expressing switching, and tup1 mutants display attenuated
cells and in white cells. Hence, Efg1 appears to link virulence (Murad et al. 2001a; Zhao et al. 2002). By
mechanistically central carbon metabolism with analogy with its homologue in S. cerevisiae (Smith
phenotypic switching, as well as with yeast-hypha and Johnson 2000), C. albicans Tup1 is thought
morphogenesis (Fig. 10.2B). Efg1 activity is to operate as a global transcriptional repressor
regulated by the Ras-cAMP signalling pathway which is targeted to specific promoters through
(Bockmuhl and Ernst 2001). In S. cerevisiae, the protein–protein interactions with specific DNA
Ras-cAMP pathway is known to regulate growth binding proteins. Nrg1 is thought to be one such
and carbon metabolism (Thevelein 1994; Rolland protein. Like Tup1, Nrg1 is a repressor of hyphal
et al. 2001). development in C. albicans (Braun et al. 2001;
Fig. 10.3. Regulation of central carbon metabolism in C. (nucleotide, fatty acid and amino acid metabolism) is in-
albicans. Fermentative metabolism (left) is up-regulated by ferred, but will depend in actuality upon the available nutri-
Efg1, observed in white cells, and appears to be activated in ents in each microenvironment. Flux towards the cell wall
the bloodstream and possibly in organs. Inright contrast, precursors mannose and N-acetylglucosamine (GlcNAc) is
respiratory metabolism (right) is down-regulated by Efg1 inferred from the observation that the cell wall comprises
and Tup1, observed in opaque cells, and appears to be ac- about 30% of C. albicans biomass, and therefore that these
tivated in during phagocytosis and possibly skin infections precursors must be synthesised under all growth condi-
(see text for details). Flux to or from peripheral pathways tions
196 A.J.P. Brown
Murad et al. 2001a). Nrg1 is a sequence-specific morphogenesis, and (2) that efg1 and tup1 mutants
DNA binding protein which interacts with Nrg1 re- display attenuated virulence do not necessarily
sponse elements in the promoters of hyphal genes imply that the regulation of carbon metabolism
to repress their transcription. This repression is is essential for pathogenicity. Is this the case? So
dependent upon Tup1, which is consistent with far, no papers have been published which describe
the idea that Nrg1 targets Tup1 to the promoters of the global expression profiles of C. albicans cells
hypha-specific genes to repress their expression. isolated from diseased tissue. No doubt these types
Preliminary transcript profiling, using arrays of experiment will be published in the near future,
containing about one-third of C. albicans genes, as soon as the technical challenges posed by the
confirmed that Tup1 represses hypha-specific limited availability of fungal biomass and con-
genes in concert with Nrg1 (Murad et al. 2001b). tamination with host tissue have been overcome.
These genes included ECE1 and the adhesin Meanwhile, evidence highlighting the importance
genes ALS3 and HWP1. This has been further of carbon metabolism for fungal pathogenicity is
confirmed by more recent transcript profiling ex- based on two main observations, both of which
periments with whole-genome arrays (Mavor 2004; arose through the expression profiling of fungal
http://www.pasteur.fr/recherche/unites/Galar_Fun cells using ex vivo infection models.
gail/). However, these experiments have also re- The first observation arose through the
vealed that Tup1 represses numerous other transcript profiling of S. cerevisiae cells following
functions in addition to morphogenesis, includ- phagocytosis by macrophages (Lorenz and Fink
ing pathways of central and peripheral carbon 2001). Glyoxylate cycle genes (CIT2, ICL1, MDH2,
metabolism (http://www.pasteur.fr/recherche/unit MLS1) were shown to be amongst the most strongly
es/Galar_Fungail/). Gluconeogenic genes (PCK1, induced genes in these phagocytosed yeast cells,
FBP1) and TCA and glyoxylate cycle genes (IDP2, and these glyoxylate cycle genes were induced
MDH11, ICL1) were derepressed in tup1 cells, much more strongly than were TCA cycle-specific
indicating that Tup1 repressed non-fermentative genes (FUM1, IDH1, KGD1, SDH1,2,3,4). Northern
metabolism. Consistent with this idea, genes blotting was then used to demonstrate that ICL1 is
required for the catabolism of non-fermentative induced in phagocytosed C. albicans cells (Lorenz
carbon sources were derepressed in tup1 cells and Fink 2001), and this has been confirmed
(GAL1,10, ADH3,4,5, POX4). Therefore, like Efg1, by an independent DDRT-PCR-based screen for
Tup1 appears to provide another mechanistic link C. albicans genes which are induced following
between carbon metabolism and morphogenesis ingestion by macrophages (Prigneau et al. 2003).
and phenotypic switching (Figs. 2C and 3). Lorenz and Fink (2001, 2002) noted that glyoxylate
There is some evidence that the regulation of cycle genes are required for the virulence of
carbon metabolism by Tup1 might be relatively di- other microbial pathogens, and hence they tested
rect, operating via Tup1-targeting proteins such as whether this was the case for C. albicans. Their
Nrg1 and Mig1. Several glycolytic genes contain prediction was confirmed by demonstrating that
Nrg1 response elements in their promoters (PGI1, the inactivation of ICL1 attenuates the virulence
FBA1, GPM1), and recent proteomic data from our of C. albicans. This study suggests that C. albicans
laboratory have indicated that Fba1 and Gpa1 pro- cells are deprived of fermentative carbon sources
tein levels are regulated by Nrg1 (Yin et al., unpub- following ingestion by a macrophage (Fig. 10.3).
lished data). Hence, these genes would appear to be Most significantly, these authors highlighted the
regulated by Nrg1-Tup1. Mig1 and Tup1 have been importance of central metabolic pathways for the
shown by Northern blotting to regulate the gluco- survival of this pathogenic fungus in vivo.
neogenic gene PCK1, and transcript profiling has These observations have been reinforced
revealed that Mig1 and Tup1 regulate additional by a recent analysis of the transcriptional re-
genes involved in carbon metabolism (Murad et al. sponse of C. albicans to phagocytosis by murine
2001a, b). macrophages (Lorenz et al. 2004). This study
Collectively, these data reinforce the links confirmed that glycoxylate cycle genes are induced
between carbon metabolism and morphogenesis following phagocytosis, and revealed that lipid
(and other virulence attributes). However, the β-oxidation genes are also induced (FOX2, POX1,
relevance of these links to pathogenesis is only POX2). This led the authors to suggest that, fol-
inferred. For example, the observations (1) that lowing phagocytosis by macrophages, C. albicans
Efg1 and Tup1 control carbon metabolism and reprograms its metabolism to generate glucose
via lipid catabolism, the glycoxylate cycle and might have been directed more towards its sur-
gluconeogenesis (Lorenz et al. 2004). vival as a commensal or mucocutaneous pathogen,
The second observation was generated by the rather than as a systemic pathogen.
transcript profiling of C. albicans cells following Many C. albicans genes, including those en-
exposure to human blood (Fradin et al. 2003). This coding metabolic enzymes, have been reported to
study, which has provided important clues about be required for virulence (reviewed by Navarro-
the regulation of numerous virulence attributes Garcia et al. 2001). This has sparked debate in
during the early stages of disseminated candidiasis, the field about the definition of a virulence gene.
is discussed in some detail in the chapter by Munro “Virulence genes” have now been defined as those
and colleagues (Chap. 9, this volume). Hence, my encoding factors which interact directly with host
discussion of their data will be restricted to obser- components, such as adhesins or secreted aspartyl
vations relating to carbon metabolism. Both gly- proteinases (Odds et al. 2001, 2003). Factors such
colytic genes (PGI1, PFK2, FBA1, GAP1, PGK1, as Efg1 and Tup1 which control the expression of
ENO1, PYK1 (CDC19)) and glyoxylate cycle genes virulence genes should be termed “virulence reg-
(ICL1, MLS1, MDH1, ACS1) were induced in C. al- ulators”. Odds et al. (2003) also suggest that fac-
bicans cells within 20 min of exposure to human tors which are required for microbial function but
blood. Fradin et al. (2003) then used RT-PCR to which do not interact directly with the host should
examine the expression of a subset of these fungal not be termed virulence attributes. Nevertheless,
genes following the intravenous injection of C. al- although position effects which influence the ex-
bicans cells into a murine host. This confirmed, for pression of the URA3 marker potentially compro-
the first time, the induction of the glyoxylate cycle mise the interpretation of some virulence stud-
genes in vivo. ies (Brand et al. 2004), it is clear that some non-
The simultaneous induction of fermentative virulence attributes, such as metabolic genes, are
and respiratory metabolism might seem counter- required for virulence (Navarro-Garcia et al. 2001;
intuitive (Fradin et al. 2003). However, the expla- Roemer et al. 2003). These include genes involved
nation is probably that the population of fungal in lipid and nucleotide biosynthesis. Clearly, these
cells examined contained a mixture of cells in dif- attributes are required for the growth and survival
ferent microenvironments. Presumably, glyoxylate of the fungus in the host. Hence, these functions
cycle genes were induced in phagocytosed C. albi- will be referred to here as “fitness attributes”. The
cans cells, whereas glycolytic genes were induced inactivation of fitness attributes would be expected
in those fungal cells which remained in the plasma to attenuate the ability of the fungus to assimi-
and, hence, retained access to glucose (Fig. 10.3; late nutrients or combat environmental stress, for
Fradin et al. 2003). This highlights the importance example, and hence would attenuate its ability to
of developing parallel technologies allowing one establish an infection (Fig. 10.2D).
to analyse individual fungal cells within complex There is mounting evidence for the evolution
populations which occupy different microenviron- of fitness attributes in C. albicans. For example,
ments in the mammalian host (Barelle et al. 2004). this pathogen appears to have evolved specialised
Nevertheless, in the context of this review, the work stress responses which differ significantly from
of Fradin et al. (2003) clearly confirms that C. albi- stress responses in S. cerevisiae and Schizosac-
cans undergoes significant metabolic reprogram- charomyces pombe (Enjalbert et al. 2003; Nicholls
ming when it comes in contact with the host. et al. 2004; Smith et al. 2004). These specialised
stress responses appear to reflect the evolutionary
adaptation of C. albicans to its host. Unlike
budding and fission yeast, C. albicans does not
VI. Co-Evolution of Virulence appear to recognise an ambient temperature of
and Fitness 37 ◦ C as a stress, and this fungus is relatively
resistant to the oxidative stresses it encounters
C. albicans is an opportunistic pathogen. It causes following phagocytosis (Smith et al. 2004). The
systemic infections relatively rarely, and the devel- inactivation of Hog1, which is required to activate
opment of such infections is probably more depen- the oxidative stress responses, has been reported
dent upon the immune status of the host than on the to attenuate the virulence of C. albicans (Alonso-
virulence status of the fungus. Hence, it could be Monge et al. 1999). Similarly, genome sequencing
argued that evolutionary pressure upon C. albicans has provided evidence of metabolic specialisation
198 A.J.P. Brown
in C. albicans (Jones et al. 2004), as discussed in How is metabolism integrated with virulence?
Sect. III. Metabolic genes and pathways appear As described above, initial studies indicate that
to have evolved to allow this fungus to assimilate Efg1, Tup1 and Gcn4 are involved, but how do they
those nutrients which are available to it within the link metabolism and virulence mechanistically?
various microenvironments it encounters during Tup1 might achieve this by regulating metabolic
the initial colonisation of the host and subsequent and virulence genes via specific DNA binding
disease progression. Hence, fitness attributes ap- proteins such as Nrg1, Mig1 and Rfg1 (Kadosh and
pear to have evolved alongside virulence attributes Johnson 2001; Murad et al. 2001a, b). However,
in C. albicans. this prediction needs to be tested experimentally.
Not only have fitness and virulence attributes Gcn4 appears to regulate morphogenesis via the
themselves been evolving, but also the mechanisms Ras-cAMP pathway (Tripathi et al. 2002), but
which regulate their activity appear to have been the specific link between Gcn4 and Ras-cAMP
evolving in C. albicans. There is clear evidence for signalling has not been defined. Efg1 is known to
the co-ordinated regulation of virulence attributes regulate the expression of both metabolic and viru-
in a niche-specific fashion (Lane et al. 2001; Murad lence genes (Doedt et al. 2004), but the mechanisms
et al. 2001a; Nantel et al. 2002). This presumably re- by which this is achieved remain obscure.
flects the differential contributions of the various Other factors are probably involved in the co-
virulence attributes within different microenviron- ordination of virulence and fitness attributes, and
ments in the host (Brown and Gow 1999; Brown these remain to be identified. However, the increas-
2002). Similarly, as described here (Sect. V), there is ing availability of large mutant collections will fa-
mounting evidence for differential metabolic reg- cilitate the design of global screens for mutations
ulation in a niche-specific fashion in C. albicans which affect both virulence attributes and fitness.
(Lorenz and Fink 2001; Fradin et al. 2003; Rubin- These include the doxycycline-conditional mutants
Bejerano et al. 2003). In addition, some data suggest in the GRACE collection (Roemer et al. 2003), and
that stress responses might be regulated in a niche- the collections of transposon insertion mutants
specific fashion in C. albicans (Fradin et al. 2003). generated by the Mitchell and Johnson groups (No-
Therefore, it is attractive to speculate that virulence bile et al. 2003; Uhl et al. 2003). In addition, the po-
and fitness attributes are regulated in such a man- tential of an antisense-based screen (de Backer et al.
ner as to ensure that both are expressed appropri- 2001) for the isolation of this type of pleiotropic
ately in particular microenvironments. This review factor should be tested.
has highlighted emerging evidence which suggests How does C. albicans regulate its fitness and
that metabolic reprogramming is mechanistically virulence attributes in response to specific host
linked to the control of virulence through the regu- environments? The limited data addressing this
lators Efg1, Tup1 and Gcn4 (Sect. V; Lan et al. 2002; important issue have been derived largely from
Tripathi et al. 2002; Doedt et al. 2004). Hence, C. al- elegant analyses of a limited number of genes or
bicans appears to have evolved specific signalling proteins in infection models, and in patient sam-
mechanisms which facilitate the co-ordinated reg- ples using RT-PCR and immunological techniques
ulation of the virulence and fitness attributes which (Glee et al. 1995; Staab et al. 1996; Hoyer et al. 1999;
it requires for survival in the host (Fig. 10.2D). Naglik et al. 1999). Whilst transcript profiling
has been performed on ex vivo infection models
(Fradin et al. 2003; Rubin-Bejerano et al. 2003),
further technological developments are required
VII. Conclusions before genomic technologies can be applied to in
and Future Perspectives vivo samples, for example, in combination with
laser capture dissection microscopy. Even with
In this chapter I have argued that metabolism is in- such developments, problems will remain. For
timately linked with virulence in C. albicans. I have example, most of these technologies average the
also highlighted the critical role of C. albicans ge- responses of a fungal population. However, in vivo
nomics in revealing these links. Unexpected links samples are non-homogenous and complex, with
between metabolism and virulence were first re- individual fungal cells being exposed to differing
vealed by the global perspectives provided by tran- microenvironments. Hence, to gain a true un-
script profiling. However, these findings have raised derstanding of fungal behaviour during disease
major new questions which need to be addressed. progression, technologies such as Single Cell
Profiling must be used in parallel with transcript Alonso-Monge R, Navarro-Garcia F, Molero G, Diez-Ore-
profiling and proteomics. Single cell profiling al- jas R, Gustin M, Pla J, Sanchez M, Nombela C (1999)
lows the molecular responses of individual fungal Role of the mitogen-activated protein kinase Hog1p
in morphogenesis and virulence of Candida albicans.
cells to be examined in vivo (Barelle et al. 2004). J Bacteriol 181:3058–3068
Finally, what happens to metabolic reprogram- Bailey DA, Feldmann PJF, Bovey M, Gow NAR, Brown AJP
ming when fungal cells respond and adapt to an- (1996) The Candida albicans HYR1 gene, which is ac-
tifungal drugs? Transcript profiling and proteomic tivated in response to hyphal development, belongs to
a gene family encoding yeast cell wall proteins. J Bac-
studies of C. albicans cells as they acquire drug re- teriol 178:5353–5360
sistance have suggested that metabolism is affected Baillie GS, Douglas LJ (1999) Candida biofilms and their
(Cowen et al. 2002; Bruneau et al. 2003; Rogers and susceptibility to antifungal agents. Methods Enzymol
Barker 2003). These changes in metabolism do not 310:644–656
appear to be restricted to the antifungal targets Barelle CJ, Manson C, MacCallum D, Odds FC, Gow NAR,
Brown AJP (2004) GFP as a quantitative reporter of
themselves. For example, they are not limited to gene regulation in Candida albicans. Yeast 21:333–
ergosterol biosynthesis following exposure to an 340
azole antifungal. Is this metabolic reprogramming Bennett RJ, Johnson AD (2003) Completion of a parasexual
required for drug adaptation, or is it an indirect cycle in Candida albicans by induced chromosome loss
in tetraploid strains. EMBO J 22:2505–2515
by-product of exposure to the antifungal agent? Berman J, Sudbery PE (2002) Candida albicans: a molecular
What relevance has the issue of C. albicans fit- revolution built on lessons from budding yeast. Nat Rev
ness and virulence to the patient suffering from Genet 3:918–930
a Candida infection? To date, the academic commu- Bertram G, Swoboda RK, Gow NAR, Gooday GW,
nity has focussed most attention upon C. albicans Brown AJP (1996) Structure and regulation of the
Candida albicans ADH1 gene encoding an immuno-
virulence attributes and has paid relatively little genic alcohol dehydrogenase. Yeast 12:115–128
attention to metabolism. Ironically, most virulence Bockmuhl DP, Ernst JF (2001) A potential phosphorylation
attributes are likely to be poor targets for antifungal site for an A-type kinase in the Efg1 regulator pro-
therapy because most are polygenic and not essen- tein contributes to hyphal morphogenesis of Candida
albicans. Genetics 157:1523–1530
tial for survival in vivo. In contrast, most successful Brand A, MacCallum DM, Brown AJP, Gow NAR, Odds FC
antifungal drug screens have focussed on the fitness (2004) Ectopic expression of URA3 can influence the
of C. albicans. Also, central metabolic functions do virulence phenotypes and proteome of Candida albi-
appear to be essential for fitness in vivo (Lorenz cans but can be overcome by targeted reintegration of
and Fink 2001; Roemer et al. 2003). Clearly, a com- URA3 at the RPS10 locus. Eukaryot Cell 3:900–909
Braun BR, Johnson AD (1997) Control of filament formation
plete understanding of C. albicans pathogenicity in Candida albicans by the transcriptional repressor
is dependent upon a characterisation of fitness as TUP1. Science 277:105–109
well as virulence attributes. Braun BR, Kadosh D, Johnson AD (2001) NRG1, a repres-
sor of filamentous growth in Candida albicans, is
Acknowledgements. I am grateful to many colleagues for down-regulated during filament induction. EMBO J
stimulating debates about Candida genomics, especially 20:4753–4761
to Frank Odds, Neil Gow, other members of the Aberdeen Brega E, Zufferey R, Mamoun CB (2004) Candida albicans
Fungal Group, my partners in the European Galar Fungail Csy1 is a nutrient sensor important for activation of
Consortium, and my collaborators with the COGEME amino acid uptake and hyphal morphogenesis. Eu-
Consortium, particularly Ken Haynes and Jan Quinn. karyot Cell 3:135–143
My work is supported by grants from the Biotechnology Brown AJP (2002) Morphogenetic signalling pathways in
and Biological Sciences Research Council (1/P17124, Candida albicans. In: Calderone R (ed) Candida and
1/G188883, BBS/B/06679, BBS/B/13764, BB/C501176/1), candidiasis. ASM Press, Washington, DC, pp 95-106
the British Society for Antimicrobial Chemotherapy Brown AJP, Gow NAR (1999) Regulatory networks control-
(PG/01), the British Council and National Research ling Candida albicans morphogenesis. Trends Micro-
Council of Canada (00CRP04), the European Commission biol 7:333–338
(MRTN-CT-2003-504148) and the Wellcome Trust (063204, Brown AJP, Barelle CJ, Budge S, Duncan J, Harris S,
068143). Lee PR, Leng P, Macaskill S, Murad AMA, Ramsdale M
et al. (2000) Gene regulation during morphogenesis
in Candida albicans. In: Ernst JF, Schmidt A (eds)
Contributions to microbiology: dimorphism in
References human pathogenic and apathogenic yeasts. Karger,
Basel, vol 5, pp 112–125
Albrecht G, Mosch H-U, Hoffman B, Reusser U, Braus GH Bruneau JM, Maillet I, Tagat E, Legrand R, Supatto F, Fu-
(1998) Monitoring the Gcn4 protein-mediated re- dali C, Caer JP, Labas V, Lecaque D, Hodgson J (2003)
sponse in the yeast Saccharomyces cerevisiae. J Biol Drug induced proteome changes in Candida albicans:
Chem 273:12696–12702 comparison of the effect of beta(1,3) glucan synthase
200 A.J.P. Brown
inhibitors and two triazoles, fluconazole and itracona- Fradin C, Kretschmar M, Nichterlein T, Gaillardin C, d’En-
zole. Proteomics 3:325–336 fert C, Hube B (2003) Stage-specific gene expression
Calderone R (2002) Candida and candidiasis. ASM Press, of Candida albicans in human blood. Mol Microbiol
Washington, DC Calderone R, Fonzi WA (2001) Vir- 472:1523–1543
ulence factors of Candida albicans. Trends Microbiol Garcia-Sanchez S, Aubert S, Iraqui I, Janbon G, Ghigo J-M,
9:327–335 d’Enfert C (2004) Candida albicans biofilms: a devel-
Care RS, Trevethick J, Binley KM, Sudbery PE (1999) The opmental state associated with specific and stable gene
MET3 promoter: a new tool for Candida albicans expression patterns. Eukaryot Cell 3:536–545
molecular genetics. Mol Microbiol 34:792–798 Glee PM, Sundstrom P, Hazen KC (1995) Expression of sur-
Chauhan N, Inglis D, Roman E, Pla J, Li D, Calera JA, face hydrophobic proteins by Candida albicans in vivo.
Calderone R (2003) Candida albicans response regula- Infect Immun 63:1373–1379
tor gene SSK1 regulates a subset of genes whose func- Gow NAR, Brown AJP, Odds FC (2002) Fungal morpho-
tions are associated with cell wall biosynthesis and genesis and host invasion. Curr Opin Microbiol 5:366–
adaptation to oxidative stress. Eukaryot Cell 2:1018– 371
1024 Gow NAR, Knox Y, Munro CA, Thompson WD (2003) In-
Cormack B, Bertram G, Egerton M, Gow NAR, Falkow S, fection of chick chorioallantoic membrane (CAM) as
Brown AJP (1997) Yeast Enhanced Green Fluorescent a model for invasive hyphal growth and pathogenesis
Protein (yEGFP): a reporter of gene expression in Can- of Candida albicans. Med Mycol 41:331–338
dida albicans. Microbiology 143:303–311 Haynes K (2001) Virulence in Candida species. Trends Mi-
Cowen LE, Nantel A, Whiteway MS, Thomas DY, Tessier DC, crobiol 9:591–596
Kohn LM, Anderson JB (2002) Population genomics of Hellstein J, Vawter-Hugart H, Fotos P, Schmid J, Soll DR
drug resistance in Candida albicans. Proc Natl Acad (1993) Genetic similarity and phenotypic diversity of
Sci USA 99:9284–9289 commensal and pathogenic strains of Candida albicans
De Backer MD, Magee PT, Pla J (2000) Recent developments isolated from the oral cavity. J Clin Microbiol 31:3190–
in molecular genetics of Candida albicans. Annu Rev 3199
Microbiol 54:463–498 Hernandez R, Nombela C, Diez-Orejas, Gil C (2004) Two-
De Backer MD, Nelissen B, Logghe M, Viaene J, Loonen I, dimensional reference map of Candida albicans hyphal
Vandoninck S, de Hoogt R, Dewaele S, Simons FA, forms. Proteomics 4:374–382
Verhasselt P et al. (2001) An antisense-based functional Hinnebusch AG (1988) Mechanisms of gene regulation in
genomics approach for identification of genes critical the general control of amino acid biosynthesis in Sac-
for growth of Candida albicans. Nat Biotechnol 19:235– charomyces cerevisiae. Microbiol Rev 52:248–273
241 Hinnebusch AG, Natarajan K (2002) Gcn4p, a master regu-
d’Enfert C, Goyard S, Rodriguez-Arnaveilhe S, Frangeul L, lator of gene expression is controlled at multiple levels
Jones L, Tekaia F, Bader O, Albrecht A, Castillo L, by diverse signals of starvation and stress. Eukaryot
Dominguez A et al. (2005) CandidaDB: a genome Cell 1:22–32
database for Candida albicans pathogenomics. Nucleic Hoffmann B, Valerius O, Andermann M, Braus GH (2001)
Acids Res 33:D353–D357 Transcriptional autoregulation and inhibition of
Dever TE, Feng L, Wek RC, Cigan AM, Donahue TD, Hin- mRNA translation of amino acid regulator gene cpcA
nebusch AG (1992) Phosphorylation of initiation fac- of filamentous fungus Aspergillus nidulans. Mol Biol
tor 2α by protein kinase GCN2 mediates gene-specific Cell 12:2846–2857
translational control of GCN4 in yeast. Cell 68:585– Hoyer LL (2001) The ALS gene family of Candida albicans.
596 Trends Microbiol 9:176–180
Doedt T, Krishnamurthy S, Bockmühl DP, Tebarth B, Stem- Hoyer LL, Clevenger J, Hecht JE, Ehrhart EJ, Poulet FM
pel C, Russell CL, Brown AJP, Ernst JF (2004) APSES (1999) Detection of als proteins on the cell wall of Can-
proteins regulate morphogenesis and metabolism in dida albicans in murine tissues. Infect Immun 67:4251–
Candida albicans. Mol Biol Cell 15:3167–3180 4255
Douglas LJ (2003) Candida biofilms and their role in infec- Hube B, Naglik J (2002) Extracellular hydolases. In:
tion. Trends Microbiol 11:30–36 Calderone R (ed) Candida and candidiasis. ASM Press,
Ellenberger TE, Brandl CJ, Struhl K, Harrison SC (1992) Washington, DC, pp 107–122
The GCN4 basic region leucine zipper binds DNA as Hull CM, Raisner RM, Johnson AD (2000) Evidence for mat-
a dimer of uninterrupted a helices: crystal structure of ing of the “asexual” yeast Candida albicans in a mam-
the protein-DNA complex. Cell 71:1223–1237 malian host. Science 289:307–310
Enjalbert B, Nantel A, Whiteway M (2003) Stress-induced Jones T, Federspiel NA, Chibana H, Dungan J, Kalman S,
gene expression in Candida albicans: absence of a gen- Magee BB, Newport G, Thorstenson YR, Agabian N,
eral stress response. Mol Biol Cell 14:1460–1467 Magee PT et al. (2004) The diploid genome sequence
Enloe B, Diamond A, Mitchell AP (2000) A single- of Candida albicans. Proc Natl Acad Sci USA 101:7329–
transformation gene function test in diploid Candida 7334
albicans. J Bacteriol 182:5730–5736 Kadosh D, Johnson AD (2001) Rfg1, a protein related to the
Filler, SG, Kullberg BJ (2002) Deep-seated candidal infec- S. cerevisiae hypoxic regulator Rox1, controls filamen-
tions. In: Calderone R (ed) Candida and candidiasis. tous growth and virulence in C albicans. Mol Cell Biol
ASM Press, Washington, DC, pp 341–348 21:2496–2505
Fonzi WA, Irwin MY (1993) Isogenic strain construction Kastaniotis AJ, Zitomer RS (2000) Rox1 mediated repres-
and gene mapping in Candida albicans. Genetics sion. Oxygen dependent repression in yeast. Adv Exp
134:717–728 Med Biol 475:185–95
Kornitzer D, Raboy B, Kulka RG, Fink GR (1994) Regulated erdeen Miller MG, Johnson AD (2002) White-opaque
degradation of the transcription factor Gcn4. EMBO J switching in Candida albicans is controlled by mating-
13:6021–6030 type locus homeodomain proteins and allows efficient
Korting HC, Hube B, Oberbauer S, Januschke E, Hamm G, mating. Cell 110:293–302
Albrecht A, Borelli C, Schaller M (2003) Reduced ex- Mueller PP, Hinnebusch AG (1994) Multiple upstream AUG
pression of the hyphal-independent Candida albicans codons mediate translational control of GCN4. Cell
proteinase genes SAP1 and SAP3 in the efg1 mutant is 45:201–207
associated with attenuated virulence during infection Murad AMA, Leng P, Straffon M, Wishart J, Macaskill S,
of oral epithelium. J Med Microbiol 52:623–632 MacCallum D, Schnell N, Talibi D, Marechal D, Tekaia
Krishnamoorthy T, Pavitt GD, Zhang F, Dever TE, Hinneb- F et al. (2001a) NRG1 represses yeast-hypha morpho-
usch AG (2001) Tight binding of the phosphorylated genesis and hypha-specific gene expression in Candida
a subunit of initiation factor 2 (eIF2α) to the regulatory albicans. EMBO J 20:4742–4752
subunits of guanine nucleotide exchange factor eIF2B Murad AMA, d’Enfert C, Gaillardin C, Tournu H, Tekaia F,
is required for inhibition of translation initiation. Mol Talibi D, Marechal D, Marchais V, Cottin J, Brown AJP
Cell Biol 21:5018–5030 (2001b) Transcript profiling in Candida albicans re-
Kullberg BJ, Filler SG (2002) Candidemia. In: Calderone R veals new cellular functions for the transcriptional re-
(ed) Candida and candidiasis. ASM Press, Washington, pressors, CaTup1, CaMig1 and CaNrg1. Mol Microbiol
DC, pp 327–340 42:981–993
Kumamoto CA (2002) Candida biofilms. Curr Opin Micro- Naglik JR, Newport G, White TC, Fernandes-Naglik LL,
biol 5:608–611 Greenspan JS, Greenspan D, Sweet SP, Challacombe SJ,
Kvaal CA, Srikantha T, Soll DR (1997) Misexpression of the Agabian N (1999) In vivo analysis of secreted aspartyl
white-phase-specific gene WH11 in the opaque phase proteinase expression in human oral candidiasis. In-
of Candida albicans affects switching and virulence. fect Immun 67:2482–2490
Infect Immun 65:4468–4475 Nantel A, Dignard D, Bachewich C, Harcus D, Marcil A,
Kvaal C, Lachke SA, Srikantha T, Daniels K, McCoy J, Bouin A-P, Sensen CW, Hogues H, van het Hoog M,
Soll DR (1999) Misexpression of the opaque-phase- Gordon P et al. (2002) Transcript profiling of Candida
specific gene PEP1 (SAP1) in the white phase of albicans cells undergoing the yeast-to-hyphal transi-
Candida albicans confers increased virulence in tion. Mol Biol Cell 13:3452–2365
a mouse model of cutaneous infection. Infect Immun Natarajan K, Meyer MR, Jackson BM, Slade D, Roberts C,
67:6652–6662 Hinnebusch AG, Marton MJ (2001) Transcriptional
Lan C-Y, Newport G, Murillo LA, Jones T, Scherer S, profiling shows that Gcn4p is a master regulator of
Davis RW, Agabian N (2002) Metabolic specialization gene expression during amino acid starvation in yeast.
associated with phenotypic switching in Candida Mol Cell Biol 21:4347–4368
albicans. Proc Natl Acad Sci USA 99:14907–14912 Navarro-Garcia F, Sanchez M, Nombela C, Pla J (2001) Vir-
Lane S, Birse C, Zhou S, Matson R, Liu H (2001) DNA ar- ulence genes in the pathogenic yeast Candida albicans.
ray studies demonstrate convergent regulation of vir- FEMS Microbiol Rev 25:245–268
ulence factors by Cph1, Cph2, and Efg1 in Candida Nicholls S, Straffon M, Enjalbert B, Nantel A, Macaskill S,
albicans. J Biol Chem 276:48988–48996 Whiteway M, Brown AJP (2004) Msn2/4-like transcrip-
Leng P, Lee PR, Wu H, Brown AJP (2001) Efg1, a morpho- tion factors play no obvious roles in the stress re-
genetic regulator in Candida albicans, is a sequence- sponses of the fungal pathogen, Candida albicans. Eu-
specific DNA binding protein. J Bacteriol 183:4090– karyot Cell 3:1111–1123
4093 Nobile CJ, Bruno VM, Richard ML, Davis DA, Mitchell AP
Leuker CE, Sonneborn A, Delbruck S, Ernst JF (1997) Se- (2003) Genetic control of chlamydospore formation in
quence and promoter regulation of the PCK1 gene Candida albicans. Microbiology 149:3629–3637
encoding phosphoenolpyruvate carboxykinase of the Odds FC (1988) Candida and candidosis, 2nd edn. Bailliere
fungal pathogen Candida albicans. Gene 192:235–240 Tindall, London Odds FC (1994) Candida species and
Lo HJ, Kohler JR, DiDomenico B, Loebenberg D, Caccia- virulence. ASM News 60:313–318
puoti A, Fink GR (1997) Nonfilamentous C. albicans Odds FC, Gow NAR, Brown AJP (2001) Fungal virulence
mutants are avirulent. Cell 90:939–949 studies come of age. Genome Biol 2:1009.1–1009.4
Lorenz MC, Fink GR (2001) The glyoxylate cycle is required Odds FC, Calderone RA, Hube B, Nombela C (2003) Viru-
for fungal virulence. Nature 412:83–86 lence in Candida albicans: views and suggestions from
Lorenz MC, Fink GR (2002) Life and death in a macrophage: a peer-group workshop. ASM News 69:54–55
role of the glycoxylate cycle in virulence. Eukaryot Cell Oliphant AR, Brandl CJ, Struhl K (1989) Defining the se-
1:657–662 quence specificity of DNA binding proteins by selecting
Lorenz MC, Bender JA, Fink GR (2004) Transcriptional re- binding sites from random sequence oligonucleotides:
sponse of Candida albicans upon internalization by analysis of yeast GCN4 protein. Mol Cell Biol 9:2944–
macrophages. Eukaryot Cell 3:1076–1087 2949
Magee PT (1998) Analysis of the Candida albicans genome. Paluh JL, Orbach MJ, Legerton TL, Yanofsky C (1988) The
Methods Microbiol 26:395–415 cross-pathway control gene of Neurospora crassa, cpc-
Magee BB, Magee PT (2000) Induction of mating in Can- 1, encodes a protein similar to GCN4 of yeast and the
dida albicans by construction of MTLa and MTLalpha DNA-binding domain of the oncogene v-jun-encoded
strains. Science 289:310–313 protein. Proc Natl Acad Sci USA 85:3728–3732
Mavor A (2004) Transcriptional regulation of morphogene- Petter R, Chang YC, Kwon-Chung KJ (1997) A gene ho-
sis in Candida albicans. PhD Thesis, University of Ab- mologues to Saccharomyces cerevisiae SNF1 appears
202 A.J.P. Brown
to be essential for the viability of Candida albicans. by the Efg1p morphogenetic regulator. Infect Immun
Infect Immun 65:4909–4917 67:4655–4660
Pitarch A, Sanchez M, Nombela C, Gil C (2002) Sequential Srikantha T, Klapach A, Lorenz WW, Tsai LK, Laughlin LA,
fractionation and two-dimensional gel analysis un- Gorman JA, Soll DR (1996) The sea pansy Renilla reni-
ravels the complexity of the dimorphic fungus Can- formis luciferase serves as a sensitive bioluminescent
dida albicans cell wall proteome. Mol Cell Proteomics reporter for differential gene expression in Candida
112:967–982 albicans. J Bacteriol 178:121–129
Prigneau O, Porta A, Poudrier JA, Colonna-Romano S, Srikantha T, Tsai LK, Daniels K, Soll DR (2000) EFG1 null
Noel T, Maresca B (2003) Genes involved in β- mutants of Candida albicans switch but cannot express
oxidation, energy metabolism and glyoxylate cycle the complete phenotype of white-phase budding cells.
are induced by Candida albicans during macrophage J Bacteriol 182:1580–1591
infection. Yeast 20:723–730 Staab JF, Ferrer CA, Sundstrom P (1996) Developmental
Rao GR, Ramakrishnan T, Sirsi M (1960) Enzymes in expression of a tandemly repeated, proline and
Candida albicans: I pathways of glucose dissimilation. glutamine-rich amino acid motif on hyphal surfaces
J Bacteriol 80:654–658 of Candida albicans. J Biol Chem 271:6298–6305
Roemer T, Jiang B, Davison J, Ketela T, Veillette K, Breton A, Stoldt VR, Sonneborn A, Leuker CE, Ernst J (1997) Efg1p,
Tandia F, Linteau A, Sillaots S, Marta C et al. (2003) an essential regulator of morphogenesis of the human
Large-scale essential gene identification in Candida al- pathogen Candida albicans, is a member of a conserved
bicans and applications to antifungal drug discovery. class of bHLH proteins regulating morphogenetic pro-
Mol Microbiol 50:167–181 cesses in fungi. EMBO J 16:1982–1991
Rogers PD, Barker KS (2003) Genome-wide expression pro- Sundstrum P (2002) Adhesion in Candida spp. Cell Micro-
file analysis reveals coordinately regulated genes asso- biol 4:461–469
ciated with stepwise acquisition of azole resistance in Swoboda RK, Bertram G, Delbruck S, Ernst JF, Gow NAR,
Candida albicans clinical isolates. Antimicrob Agents Gooday GW, Brown AJP (1994) Fluctuations in gly-
Chemother 47:1220–1227 colytic mRNA levels during the yeast-to-hyphal tran-
Rolfes RJ, Hinnebusch AG (1993) Translation of the yeast sition in Candida albicans reflect underlying changes
transcriptional activator GCN4 is stimulated by purine in growth and are not a response to cellular dimor-
limitation: implications for activation of the protein phism. Mol Microbiol 13:663–672
kinase GCN2. Mol Cell Biol 13:5099–5111 Tebarth B, Doedt T, Krishnamurthy S, Weide M, Monterola F,
Rolland F, Winderickx J, Thevelein JM (2001) Glucose- Dominguez A, Ernst JF (2003) Adaptation of the Efg1p
sensing mechanisms in eukaryotic cells. Trends morphogenetic pathway in Candida albicans by nega-
Biochem Sci 26:310–317 tive autoregulation and PKA-dependent repression of
Rubin-Bejerano I, Fraser I, Grisafi P, Fink GR (2003) Phago- the EFG1 gene. J Mol Biol 329:949–962
cytosis by neutrophils induces an amino acid depri- Thevelein JM (1994) Signal-transduction in yeast. Yeast
vation response in Saccharomyces cerevisiae and Can- 10:1753–1790
dida albicans. PNAS USA 100:11007–11012 Tripathi G, Wiltshire C, Macaskill S, Tournu H, Budge S,
Ruhnke M (2002) Skin and mucous membrane infections. Brown AJP (2002) CaGcn4 co-ordinates morpho-
In: Calderone R (ed) Candida and candidiasis. ASM genetic and metabolic responses to amino acid
Press, Washington, DC, pp 307–325 starvation in Candida albicans. EMBO J 21:5448–5456
Santos MAS, Keith G, Tuite MF (1993) Non-standard trans- Tsong AE, Miller MG, Raisner RM, Johnson AD (2003) Evo-
lational events in Candida albicans mediated by an lution of a combinatorial transcriptional circuit: a case
unusual seryl-tRNA with a 5 -CAG-3 (leucine) anti- study in yeasts. Cell 115:389–399
codon. EMBO J 12:607–616 Tournu H, Tripathi G, Gertram G, Macaskill S, Mavor A,
Saville SP, Lazzell AL, Monteagudo C, Lopez-Ribot JL (2003) Walker L, Odds FC, Gow NAR, Brown AJP (2005)
Engineered control of cell morphology in vivo reveals Blobal role of the protein kinase, Gcn2, in the human
distinct roles for yeast and filamentous forms of Can- pathogen, Candida albicans. Eukaryotic Cell (in
dida albicans during infection. Eukaryot Cell 2:1053– press)
1060 Tzung K-W, Williams RM, Scherer S, Federspiel N, Jones T,
Sharkey LL, McNemar MD, Saporito-Irwin SM, Sypherd PS, Hansen N, Bivolarevic V, Huizar L, Komp C, Surzycki
Fonzi WA (1999) HWP1 functions in the morpholog- R et al. (2001) Genomic evidence for a complete sexual
ical development of Candida albicans downstream of cycle in Candida albicans. Proc Natl Acad Sci USA
EFG1, TUP1 and RBF1. J Bacteriol 181:5273–5279 98:3249–3253
Smith RL, Johnson AD (2000) Turning off genes by Ssn6- Uhl MA, Johnson AD (2001) Development of Streptococ-
Tup1: a conserved system of transcriptional repression cus thermophilus lacZ as a reporter gene for Candida
in eukaryotes. Trends Biochem Sci 25:325–330 albicans. Microbiology 147:1189–1195
Smith DA, Nicholls S, Morgan BA, Brown AJP, Quinn J (2004) Uhl MA, Biery M, Craig N, Johnson AD (2003)
A conserved stress-activated protein kinase regulates Haploinsufficiency-based large-scale forward ge-
a core stress response in the human pathogen Candida netic analysis of filamentous growth in the diploid
albicans. Mol Biol Cell 15:4179–4190 human fungal pathogen C. albicans. EMBO J
Soll DR (2002) Phenotypic switching. In: Calderone R (ed) 22:2668–2678
Candida and candidiasis. ASM Press, Washington, DC, van Burik J-AH, Magee PT (2001) Aspects of fungal patho-
pp 123–142 genesis in humans. Annu Rev Microbiol 55:743–772
Sonneborn A, Tebarth B, Ernst J (1999) Control of White- Van Neil CB, Cohen AL (1942) The metabolism of Candida
Opaque phenotypic switching in Candida albicans albicans. J Cell Comp Physiol 20:95–112
Wanke C, Eckert S, Albrecht G, van Hartingsveldt W, Punt PJ, Wilson RB, Davis D, Mitchell AP (1999) Rapid hypothesis
van den Hondel CA, Braus GH (1997) The Aspergillus testing with Candida albicans through gene disruption
niger GCN4 homologue, cpcA, is transcriptionally reg- with short homology regions. J Bacteriol 181:1868–
ulated and encodes an unusual leucine zipper. Mol Mi- 1874
crobiol 23:23–33 Wilson RB, Davis D, Enloe BM, Mitchell AP (2000)
Wek SA, Zhu S, Wek RC (1995) The histidyl-tRNA A recyclable Candida albicans URA3 cassette for PCR
synthetase-related sequence in the eIF-2a protein product-directed gene disruptions. Yeast 16:65–70
kinase GCN2 interacts with tRNA and is required Yin Z, Stead D, Selway L, Walker J, Riba-Garcia I, Mclner-
for activation in response to starvation for different ney T, Gaskell S, Oliver SG, Cash P, Brown AJP
amino acids. Mol Cell Biol 15:4497–4506 (2004) Divergence between Candida albicans and
Wey SB, Mori M, Pfaller MA, Woolson RF, Wenzel P Saccharomyces cerevisiae in their global responses to
(1989) Risk factors for hospital-acquired candidemia. amino acid starvation. Proteomics 4:2425–2436
A matched case-control study. Arch Intern Med Zaragoza O, Rodriguez C, Gancedo C (2000) Isolation of
149:2349–2353 the MIG1 gene from Candida albicans and effects of
Wickerham LJ (1951) Taxonomy of yeast. Dept Agric Tech its disruption on catabolite repression. J Bacteriol
Bull 1029:1–19 182:320–326
Williamson MI, Samaranayake LP, MacFarlane TW (1986) Zhao R, Lockhart SR, Daniels K, Soll DR (2002) Roles of
Biotypes of Candida albicans using the API 20C system. TUP1 in switching, phase maintenance, and phase-
FEMS Microbiol Lett 37:27–29 specific gene expression in Candida albicans. Eukaryot
Cell 1:353–365
11 Regulators of Candida glabrata Pathogenicity
K. Haynes1
CONTENTS 2003). Additionally, patients contracting Candida

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 205 infection stay in hospital for up to 30 days longer
II. Regulators of C. albicans Virulence . . . . . . . . 205 than do similar uninfected individuals (Wey et al.
A. C. albicans Fkh2 and Ace2 . . . . . . . . . . . . 206 1988). In one Spanish study the increase in direct
B. C. albicans Mds3 and Rim101 . . . . . . . . . 207 costs associated with Candida infection was esti-
III. C. glabrata Transcriptional Regulators . . . . . 208
A. C. glabrata Amt1 . . . . . . . . . . . . . . . . . . . 208 mated at e 16,000 (Olaechea et al. 2004). Thus, Can-
B. C. glabrata Rap1 . . . . . . . . . . . . . . . . . . . . 208 dida species exert a significant clinical and eco-
C. C. glabrata Ste12 . . . . . . . . . . . . . . . . . . . . 209 nomic impact.
D. C. glabrata Ace2 . . . . . . . . . . . . . . . . . . . . 210 Candida glabrata is now responsible for about
IV. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 15%–20% of candidiasis cases (Pfaller et al. 2001)
and its incidence is still increasing (Pfaller and
Diekema 2002; Tortorano et al. 2004). This is a cause
Abbreviations: FOA, 5-fluoro-orotic acid; GLA, for concern, as this species may be more resis-
genome-wide location analysis tant to antifungal agents than are other Candida
species, including Candida albicans (Pfaller et al.
2004). A recent European study suggests that C.
I. Introduction glabrata is more often associated with infection in
surgical and solid tumour patients, and this is even
The incidence of life-threatening fungal infections more pronounced in older individuals (Tortorano
increased substantially world-wide from the 1980s et al. 2004).
to the 1990s (Beck-Sague and Jarvis 1993; Chen This clinical problem clearly needs address-
et al. 1997). The current reported incidence rates ing, and a better understanding of how C. glabrata
for candidiasis range from 0.2 to 0.55 per 1000 causes disease is imperative. Unfortunately, the in-
hospital admissions (Asmundsdottir et al. 2002; crease in fungal disease caused by C. glabrata has
Tortorano et al. 2004). Candida species now cause not been matched by a concomitant rise in our un-
more bloodstream infections than do all individ- derstanding of the basic biology of the pathogen.
ual Gram-negative bacterial species, including Es- Here I will attempt to review what is known about
cherichia coli (Wisplinghoff et al. 2003). Indeed, the regulation of C. glabrata virulence, within the
candidaemia now accounts for 10%–15% of all context of C. albicans virulence and protein func-
bloodstream infections (Jarvis 1995), one of the tion in Saccharomyces cerevisiae.
top 15 causes of mortality in the United States,
resulting in over 200,000 deaths per year (Pittet
et al. 1997). Given the continuing use of broad-
spectrum antibiotics and indwelling intravenous II. Regulators of C. albicans Virulence
catheters, two major risk factors for the develop-
ment of candidiaemia, plus the inexorable rise of A large number of attributes have been shown to be
at-risk patients, it seems unlikely that this threat important in C. albicans virulence (Navarro-Garcia
will diminish (Calderone 2002). This increased in- et al. 2001). These include the ability to undertake
cidence of candidiasis is matched by high direct morphogenetic transitions, to respond correctly to
and attributable mortality rates (Gudlaugsson et al. changes in the host environment, to secrete hy-
1Department of Infectious Diseases, Imperial College London, Du drolytic enzymes and to adhere to host surfaces.
Cane Road, London W12 0NN, UK For these processes to occur in an appropriate lo-
The Mycota XIII
Fungal Genomics
206 K. Haynes
cation at the correct time, they must be regulated. tor of Drosophila melanogaster (Weigel et al. 1989).
Transcription factors are likely to play a large role Fkh2 regulates, in partnership with Fkh1, expres-
in this regulation and have thus received a high sion of the CLB2 cluster, a group of 33 genes, the
amount of interest as potential virulence regula- transcription of which peaks in early mitosis (Zhu
tors in C. albicans (Table 11.1). Other reviews in et al. 2000). These include the transcription factor-
this volume will concentrate on many of these pro- encoding genes SWI5 and ACE2 whose products
teins. Here I will discuss only two pairs of C. al- regulate separately, and in tandem, genes expressed
bicans transcription factors: Fkh2 and Ace2 plus at the M/G1 boundary of the cell cycle. In C. albi-
Mds3 and Rim101. These have been linked to pro- cans there is only a single forkhead homologue,
cesses which are probably involved in C. glabrata FKH2 (Bensen et al. 2002). The ability of C. al-
virulence. bicans fkh2/fkh2 mutants to cause disease in vivo
has not been explored, as these cells have a slow
growth phenotype (Bensen et al. 2002). However,
A. C. albicans Fkh2 and Ace2 their ability to damage both oral epithelial and vas-
cular endothelial cells was much reduced in com-
S. cerevisiae FKH2 encodes a DNA-binding protein parison to wild type C. albicans, suggesting a role
which is similar to the forkhead transcription fac- in virulence (Bensen et al. 2002). In addition, the
Table. 11.1. Known Candida albicans transcription factors
Regulator Regulatory domaina TF typeb Mutant Reference

attenuatedc
Ace2 Cell separation, adherence, biofilms Zinc finger Yes (Kelly et al. 2004)
Ash1 Filamentous growth GATA Yes (Inglis and Johnson 2002)
Brf1 Transcription from Pol III promoters RNA Pol III n.d.d (Khoo et al. 1994)
Cap1 Response to drug and oxidative stress bZIP n.d. (Alarco and Raymond 1999)
Cbf1 Chromosome segregation, Helix-loop n.d. (Eck et al. 2001)
methionine biosynthesis helix-zipper
Cph1 Conjugation, filamentous growth Homeodomain Yese (Liu et al. 1994; Lo et al. 1997)
Cph2 Hyphal growth bHLH family n.d. (Lane et al. 2001)
Cwt1 Cell wall Zinc finger n.d. (Moreno et al. 2003)
Czf1 Filamentous growth Zinc finger n.d. (Whiteway et al. 1992)
Efg1 Cell adhesion, filamentous growth bHLH family Yese (Stoldt et al. 1997)
Fcr1 Response to drug Zinc finger n.d. (Talibi and Raymond 1999)
Fcr3 Response to drug bZIP n.d. (Yang et al. 2001)
Fkh2 Hyphal growth Forkhead n.d. (Bensen et al. 2002)
Gat1 Nitrogen utilization Zinc finger Yes (Limjindaporn et al. 2003)
Gcn4 Filamentous growth bZIP No (Tripathi et al. 2002;
Brand et al. 2004)
Mds3 Chlamydospore formation, No homology Yes (Davis et al. 2002)
hyphal growth, response to pH
Mig1 Carbon utilization Zinc finger No (Zaragoza et al. 2000)
Nrg1 Cellular morphogenesis, Zinc finger Yes (Murad et al. 2001;
stress response Saville et al. 2003)
Pdc2 Glucose metabolism DNA bindingf n.d. (Kaiser et al. 1999)
Rap1 Binds telomeres RPG-box binding n.d. (Biswas et al. 2003)
Rbf1 Filamentous growth RPG-box binding n.d. (Ishii et al. 1997)
Rfg1 Filamentous growth HMG domain Yes (Kadosh and Johnson 2001;
Khalaf and Zitomer 2001)
Rim101 Chlamydospore formation, Zinc finger Yes (Ramon et al. 1999;
filamentous growth, response to pH Davis et al. 2000)
Suc1 Maltose and sucrose metabolism Zinc finger n.d. (Kelly and Kwon-Chung 1992)
Tec1 Hyphal growth TEA/ATTS family Yes (Schweizer et al. 2000)
Tup1 Filamentous growth WD repeat Yes (Braun and Johnson 1997)
a As
defined in CGD (www.candidagenome.org) b Family of transcriptional regulator c In murine infection experiments
d Not
determined e cph1/cph1 efg1/efg1 double mutants are attenuated f Centromere protein B-like DNA binding and DDE
endonuclease domains
Candida glabrata Pathogenicity Regulators 207
deletion of the Fkh2-regulated gene ACE2 in C. albi- (Davis et al. 2002). Mds3 has two homologues in
cans results in complete attenuation of virulence in S. cerevisiae, Mds3 and Pmd1, which function as
a murine model of candidiasis (Kelly et al. 2004). In inhibitors of meiosis and have no known role in
complete contrast, C. glabrata ace2 cells are almost pH adaptation (Benni and Neigeborn 1997). Dou-
200-fold increased in their ability to cause disease ble C. albicans mds3/mds3 rim101/rim101 mutants
(Kamran et al. 2004). have more extreme phenotypes (growth inhibition
Why does the inactivation of homologous tran- at pH 9 and sensitivity to lithium chloride) than
scription factors in two Candida species result in does each single mutant (Davis et al. 2002). This
such different effects on virulence? The deletion of suggests that the two pathways act in parallel, a hy-
FKH2 in C. albicans results in a severe cell sep- pothesis which is further supported by the fact that
aration defect (Bensen et al. 2002). The cells un- a constitutive RIM101-405 allele, which encodes
dergo cytokinesis forming intact septa, but mother a gain-of-function Rim101 protein, cannot restore
and daughter cells remain joined, resulting in large alkaline-induced filamentous growth to a C. albi-
clumps of cells. The use of a C. albicans partial cans mds3/mds3 strain (Davis et al. 2002). Addi-
microarray demonstrated that the expression of tionally, the double mutants were more attenuated
CHT2, one of at least three chitinase genes in the in a murine model of candidiasis than was each of
genome, is down-regulated in fkh2/fkh2 cells, sug- the single mutants (Davis et al. 2000, 2002).
gesting a possible mechanistic explanation for the It is not surprising that the ability to adapt to
cell separation defect (Bensen et al. 2002). Similarly, changes in environmental pH plays a role in Can-
C. albicans ace2/ace2 mutants also have a cell sepa- dida virulence. PHR1 and PHR2 encode glycosi-
ration defect and they exhibit defective expression dases and have a pH-dependent expression pattern.
of genes encoding proteins required for cell separa- PHR1 is expressed in neutral and alkaline pH envi-
tion, including the chitinase Cht3 (Kelly et al. 2004). ronments and suppressed at acidic pH. PHR2 has an
Expression of CHT2 is not changed in C. albicans inverse pattern of expression. The two proteins are
ace2/ace2 mutants, but the cells are defective in functional homologues of each other, as forced ex-
separation. These data tentatively suggest that both pression of PHR1 in a phr2 mutant at acid pH, and of
functional Cht2 and Cht3 are required for C. albi- PHR2 at alkaline pH in a phr1 background rescues
cans cells to separate efficiently, and they support the conditional growth and morphogenetic defects
the observation that there is differential regulation of each mutant (Muhlschlegel and Fonzi 1997). In
of each chitinase gene (McCreath et al. 1995; Nantel addition, C. albicans phr1/phr1 mutants are com-
et al. 2002). C. glabrata has only a single chitinase pletely avirulent in a murine model of systemic can-
gene, CTS1, which is not expressed in the absence didiasis; murine blood has a pH of about 7.3 (De
of Ace2 (Kamran et al. 2004). It is possible that this Bernardis et al. 1998). They have wild-type viru-
differential regulation of chitinase gene expression lence in a rat vaginal model; the pH of the rat vagina
results in cell wall modifications which interact in is about 4.5. Conversely, C. albicans phr2/phr2 cells
different ways with the host. These possibilities will exhibit wild-type virulence in the systemic mouse
be discussed in more detail below. model but are attenuated in the rat vaginal model
(De Bernardis et al. 1998). pH-dependent gene reg-
ulation is also required for virulence of other fun-
B. C. albicans Mds3 and Rim101 gal pathogens. For example, Sclerotinia sclerotio-
rum pac1 (pacC homologue) loss-of-function mu-
The S. cerevisiae zinc finger transcription factor tants show aberrant sclerocial development and re-
Rim101 was first described as a positive regula- duced virulence on Arabidopsis thaliana and toma-
tor of IME1, which encodes a protein required for toes (Rollins 2003). Pal1 regulates the expression of
the expression of many meiotic genes (Smith et al. the virulence factor-encoding pectate lyase gene
1990; Su and Mitchell 1993). Subsequently, Rim101 pelB in Colletotrichum gloeosporioides (Drori et al.
has been shown to function in a S. cerevisiae path- 2003). Also, in Fusarium oxysporum, loss of func-
way analogous to the PacC pH response pathway tion pacC−|− mutants are more virulent than are
of Aspergillus nidulans (Penalva and Arst 2004). wild-type cells in a tomato root infection assay,
In C. albicans, Rim101 is also involved in regu- whereas gain-of-function pacCC mutants are at-
lating the response to changes in environmental tenuated (Caracuel et al. 2003). Recently we have
pH (Davis et al. 2000), although this is done in shown that elimination of PacC, prevention of PacC
tandem with at least one other regulator, Mds3 processing or blocking passage of the ambient pH
208 K. Haynes
signal to PacC via inactivation of the pal genes metallothionein genes independently of external
result in attenuation of A. nidulans virulence in copper levels (Lachke et al. 2000). High-frequency
a murine intranasal infection model (Bignell et al. switching in C. albicans regulates the expression
2005). In stark contrast, constitutive activation of of many phase-specific genes, resulting in a re-
PacC results in increased virulence, even when the markable level of phenotypic plasticity, including
pH signal is blocked. several phenotypes which facilitate pathogenesis
Taken together, these observations suggest (White and Agabian 1995). In C. glabrata, pheno-
that the ability to successfully adapt to variations typic switching has been demonstrated in hosts,
in ambient pH may be a “universal” virulence both orally and vaginally (Brockert et al. 2003),
trait in fungi. Homologues of RIM101 (CAG and it is possible that this ability plays a role in
L0E3762g), MDS3 (CAGL0D01782g) and PMD1 virulence, although this remains to be tested.
(CAGL0K02101g) are present in the C. glabrata
genome (cbi.labri.fr/Genolevures/elt/CAGL), and
it is probable that the encoded proteins will play B. C. glabrata Rap1
a significant role in virulence.
The ability to adhere to host cells is an accepted vir-
ulence attribute of C. albicans. The fungus is able
III. C. glabrata Transcriptional to express a large number of adhesins which can
Regulators interact with disparate host cell ligands including
the integrin Int1 (Gale et al. 1998), the host trans-
glutaminase substrate Hwp1 (Staab et al. 1999) and
The analysis of transcriptional regulation in C. al- the ALS (agglutinin-like sequence) family of pro-
bicans virulence is relatively well developed (see teins (Hoyer 2001). In a search for C. glabrata ad-
M. Ramsdale, M. Whiteway and A. Nantel, C. Munro hesins, Brendan Cormack and colleagues at Johns
et al. and A.J.P. Brown, Chaps. 7, 8, 9 and 10 respec- Hopkins compared the ability of 4800 signature
tively, this volume; Table 11.1). However, to date tagged mutants (Hensel et al. 1995) to adhere to
only four transcription factors have been analysed cultured human (HEp2) epithelial cells (Cormack
in C. glabrata. In this section I will deal with each et al. 1999). They identified 16 non-adherent mu-
of these in turn. tants, 14 of which had disruptions in the promoter
region upstream of the same large open reading
A. C. glabrata Amt1 frame, EPA1 (epithelial adhesin 1). C. glabrata epa1
null strains have 95% reduced adherence to ep-
Initial work on C. glabrata transcription factors ithelial cells, but they are not attenuated in ani-
was not focused on understanding virulence, but mal models (Cormack and Falkow 1999). It is pos-
rather on the regulation of metallothionein genes sible that Epa1 is a member of a family of pro-
(MT-I, MT-IIa and MT-IIb) which encode small teins which can compensate for its loss. Indeed,
proteins chelating copper (Kagi and Schaffer 1988; EPA1 is a member of a family of sub-telomerically
Zhou and Thiele 1991; Thorvaldsen et al. 1993). located genes which encode related glycosylphos-
The metal-responsive transcription factor Amt1 fa- phatidylinositol (GPI) anchored cell wall proteins
cilitates de-toxification following exposure to high with a consistent domain structure (Frieman et al.
external levels of copper (Zhou and Thiele 1993). 2002; De Las Penas et al. 2003). Of the EPA genes,
Such exposure results in rapid auto-activation of only EPA1 is expressed to any extent in vitro. How-
AMT1 expression which is greatly facilitated by ever, inactivation of Sir3 or truncation of Rap1
a homopolymeric (dA–dT) nucleosomal element (rap1-21) results in de-repression of EPA2-5 ex-
(Zhu and Thiele 1996). This analysis revealed that pression (De Las Penas et al. 2003). RAP1 can-
such nucleosomal elements may generally func- not be deleted in C. glabrata, as it appears to be
tion in eukaryotes to enable rapid access of tran- an essential gene (Haw et al. 2001). In S. cere-
scription factors, and may provide a mechanism visiae, Sir3 is one of four proteins which main-
for rapid transcriptional activation in response to tain repressed chromatin structures at telomeres
specific signals (Koch and Thiele 1999). and silent mating-type loci (Rine and Herskowitz
The role of Amt1 in virulence has not been anal- 1987; Aparicio et al. 1991). They do this not by
ysed in murine models. However, C. glabrata switch binding directly to DNA, but via interactions with
phenotypes express differing levels of the MT-II DNA-binding proteins such as the repressor ac-
tivator protein Rap1 (Shore 1994). Rap1 binds to Ste12 does not have zinc fingers but, unusually
[C(1−−3) A]n repeats in telomeric DNA, and inter- for the homeodomain group of Ste12 proteins, it
acts with Sir2, Sir3 and Sir4 to facilitate gene si- lacks an apparent consensus binding motif for the
lencing (Shore and Nasmyth 1987) and with Rif1 regulatory proteins Dig1 and Dig2 (Pi et al. 1997).
and Rif2 to maintain telomere ends (Shore 1997). It In haploid S. cerevisiae strains, Ste12 regulates
is therefore likely that the de-repression of EPA2-5 the response to mating pheromone and the inva-
expression in C. glabrata sir3 and rap1-21 cells is sive growth phenotype, whereas in diploids it plays
due to release of sub-telomeric silencing. This was a role in filamentous growth (Gustin et al. 1998;
shown to be the case in an elegant series of exper- Roberts et al. 2000; Gancedo 2001). The C. albicans
iments by De Las Penas et al. (2003). Transposon homologue Cph1 also plays a role in filamentous
mutagenesis was used to place the C. glabrata URA3 growth and, in combination with another tran-
gene into intergenic regions of the EPA clusters at scription factor Efg1, a crucial role in virulence
various distances from the telomere. Gene silenc- (Lo et al. 1997). In C. glabrata, Ste12 plays an es-
ing would result in a strain capable of growth in sential role in filamentation and also regulates the
the presence of 5-fluro-orotic acid. Wild-type cell expression of genes encoding proteins with func-
strains containing the URA3 gene up to about 24 kb tions in cell wall maintenance and biosynthesis, in-
from the telomeres could still grow in the presence cluding two structural components, Cis3 and Tip1
of FOA, demonstrating active gene silencing. In (Calcagno et al. 2003). Furthermore, the inactiva-
C. glabrata sir3 and rap1-21 cells, this growth was tion of C. glabrata Ste12 results in the partial atten-
completely ablated, suggesting that gene silencing uation of virulence (Calcagno et al. 2003). Median
was not active (De Las Penas et al. 2003). What survival times for mice infected with C. glabrata
role does gene silencing have in virulence? The ste12 cells is 5 days, compared with 3 days for an-
deletion of EPA1-5 and HYR1 resulted in a mod- imals infected with wild-type cells (p < 0.01), sug-
est decline in C. glabrata virulence as measured gesting that Ste12-regulated processes play some
by organ burden (De Las Penas et al. 2003). How- role in mediating virulence. How can these pro-
ever, an analysis of the C. glabrata genome se- cesses be identified? The genomic sequence of C.
quence (cbi.labri.fr/Genolevures/elt/CAGL) by the glabrata has recently been reported and, although
Cormack group suggests the existence of at least whole-genome oligonucleotide arrays have been
11 other Epa-like proteins, of which six are sub- designed, no transcript profiling experiments have
telomeric (De Las Penas et al. 2003). These could appeared in the literature so far.
compensate for loss of Epa1-5 and Hyr1. No re- C. glabrata Ste12 can complement both the
ports of the virulence of sir3 and rap1-21 mutants mating and filamentation defects of haploid or
have been described. However, it is intriguing that diploid S. cerevisiae ste12 mutants respectively.
the usual synteny between C. glabrata and S. cere- This demonstrates that C. glabrata Ste12 can
visiae breaks down in sub-telomeric regions (see activate transcription from native S. cerevisiae
L.J. Montcalm and K.H. Wolfe, Chap. 2, this vol- Ste12 responsive promoters. This suggests that
ume). the identification of all S. cerevisiae promoters
which bind Ste12 would be useful in an analysis
of C. glabrata Ste12 regulation. Genome-Wide
C. C. glabrata Ste12 Location Analysis (GLA) combines chromatin
immunoprecipitation and DNA microarray tech-
Ste12 belongs to a large family of fungal tran- nology to identify protein–DNA interactions on
scription factors, regulating processes involved a genome-wide scale (Ren et al. 2000). Briefly,
in mating, filamentation, substrate invasion, cell cells are fixed with formaldehyde and disrupted by
wall integrity and virulence (Liu et al. 1994; Singh sonication to yield DNA fragments of 500–1000 bp.
et al. 1994; Lo et al. 1997; Wickes et al. 1997; Vallim Transcription factor-bound DNA sequences are
et al. 2000; Young et al. 2000; Chang et al. 2000, enriched by immunoprecipitation, the cross-links
2001; Borneman et al. 2001). There are two major reversed, and the enriched DNA labelled with
subclasses within the Ste12 family of proteins: Cy-5 using ligation-mediated PCR. Un-enriched
those with both homeodomain and C2 H2 zinc DNA is labelled in the same way with Cy-3. The
finger DNA-binding domains such as Penicillium extracts are then combined and used to probe
marneffei SteA, and those which only have the an intergenic array (Ren et al. 2000). Initially,
homeodomain like, C. albicans Cph1. C. glabrata a total of 29 pheromone-induced genes which were
210 K. Haynes
directly regulated by Ste12 were reported (Ren D. C. glabrata Ace2

et al. 2000). Subsequently, 55 intergenic regions
were shown to be bound by Ste12 when cells were In S. cerevisiae, the zinc finger transcription fac-
cultured in YPD (Lee et al. 2002). Most recently, tor Ace2 activates expression of early G1-specific
Ste12-GLA was performed with cells cultured in genes (McBride et al. 1999; Simon et al. 2001).
one of three conditions: YPD; YPD with 5 mg/ml Of a total of 31 Ace2 cell cycle targets defined
α-factor for 30 min to induce mating; and YPD by Ace2-GLA, 17 also bound the paralogous tran-
plus 1% (v/v) butanol to encourage filamentation scription factor Swi5 (Simon et al. 2001). Subse-
(Harbison et al. 2004). This revealed binding to quent GLA analysis demonstrated that promoter
87 intergenic regions, potentially regulating the elements binding both Ace2 and Swi5 occur sta-
expression of 107 protein-encoding genes. In tistically more often (p = 0.005) than expected by
total, 143 genes which are potentially regulated chance (Harbison et al. 2004). The pair of genes
by Ste12 have been identified by GLA. Of these, encoding these transcription factors arose as a re-
five are dubious ORFs and a further four are sult of an ancient genome duplication event (Wolfe
hypothetical ORFs. The remaining 134 genes (Ta- and Shields 1997). C. glabrata has both ACE2 and
ble 11.2) encode functions mapped by GO Mapper SWI5 homologues, but C. albicans contains only
(db.yeastgenome.org/cgi-bin/GO/goTermMapper) a single ACE2 gene (Kamran et al. 2004; Kelly et al.
to morphogenesis, cytoskeletal organisation and 2004).
biogenesis, cell wall organisation and biogene- The inactivation of Ace2 in S. cerevisiae results
sis, signal transduction, response to stress and in a severe cell separation phenotype, the basis
cytokinesis. All of these processes have been for which is partially understood. Ace2 is local-
implicated in Candida virulence (Navarro-Garcia ized to the daughter cell nucleus after cytokinesis,
et al. 2001). as a result of Cbk1-Mob2-dependent phosphoryla-
Remarkably, only the promoters of seven S. tion, where it regulates the expression of at least
cerevisiae genes bound Ste12 in all three GLA eight daughter cell-specific genes including AMN1,
experiments: CIK1, ERG24, FUS1, HYM1, PCL2, CTS1, DSE1-4, PRY3 and SCW11 (Colman-Lerner
PRM1 and STE12. Cik1 has microtubule motor et al. 2001; Weiss et al. 2002). The chitinase Cts1
activity and is important for the establishment and the three glucan 1,3-β-glucosidases Dse2, Dse4
of the spindle and chromosome segregation and Scw11 are required to break down the trilam-
(Manning et al. 1999). Erg24 is a C-14 sterol inar septum which links the mother and daughter
reductase which acts in ergosterol biosynthesis cells (Kuranda and Robbins 1991; Cappellaro et al.
(Lorenz and Parks 1992). The function of Fus1 is 1998). Amn1 binds directly to the GTPase Tem1,
unknown, but it is located at the shmoo tip in the preventing it from interacting with its target ki-
membrane, and plays an integral role in cell fusion, nase Cdc15 and blocking activation of the mitotic
perhaps as a substrate for Cdc28 (McCaffrey exit network (Wang et al. 2003), a process which is
et al. 1987; Bagnat and Simons 2002). Hym1 is co-ordinately regulated with cell separation (Yeong
a transcriptional repressor which localizes to et al. 2002). Dse1, Dse3 and Pry3 are cell wall pro-
sites of polarized growth during budding, and teins of unknown function.
acts to mediate cell separation and regulation of Thus, it appears that one of the primary roles of
cell shape, as a component of the RAM network Ace2 is to regulate cell separation. However, Ace2
(Nelson et al. 2003). The G1 cyclin Pcl2 forms also appears to be involved in other processes. Ev-
a complex with the cyclin-dependent kinase idence for this hypothesis comes from a number
Pho85, which localizes to sites of polarized growth, of observations. Only 31 of the 119 genes, the pro-
regulates entry into the mitotic cell cycle, and is moters of which bind Ace2, are cell cycle-regulated.
essential for morphogenesis (Wang et al. 2001). Others include genes which encode proteins in-
The membrane protein Prm1 localizes to the volved in response to oxidative stress (Yap1, Snq2
shmoo tip and plays a role in cell fusion (Heiman and Mcr1) and cell wall organisation and biogenesis
and Walter 2000). Ste12 also binds to its own (Dse1, Exg1, Hsp150, Psa1, Rot2 and Wsc4). Fur-
promoter. Homologues of all seven genes are thermore, recent proteomics work in C. glabrata
present in the C. glabrata genome. The disruption has demonstrated a role for Ace2 in the regulation
of these genes and subsequent phenotypic analysis of metabolism, protein synthesis, folding and tar-
of the C. glabrata mutants will allow their role in geting, and aspects of cell growth and polarization
virulence to be determined. (Stead et al. 2005).
Table. 11.2. Saccharomyces cerevisiae promoters identified by genome location analysis which bind Ste12
ORF Gene Biological process Molecular function

YFL039C ACT1 Cell wall organisation and biogenesisa Structural constituent of cytoskeletona
YAR015W ADE1 Purine nucleotide biosynthesisa Phosphoribosylaminoimidazole-
succinocarboxamide synthase activity
YDR085C AFR1 Signal transduction during conjugation Receptor signalling protein activity
with cellular fusiona
YNR044W AGA1 Agglutination during conjugation Cell adhesion molecule binding
with cellular fusion
YGL032C AGA2 Agglutination during conjugation Cell adhesion molecule binding
YDL192W ARF1 ER to Golgi transporta GTPase activity
YPR122W AXL1 Bud site selectiona Metalloendopeptidase activity
YIL015W BAR1 Protein catabolism Aspartic-type endopeptidase activity
YER155C BEM2 Cell wall organisation and biogenesisa Signal transducer activitya
YPR171W BSP1 Actin cytoskeleton organisation and biogenesisa Molecular function unknown
YAR014C BUD14 Cellular morphogenesis during vegetative growth Molecular function unknown
YLR353W BUD8 Pseudohyphal growtha Molecular function unknown
YPL048W CAM1 Regulation of translational elongation Translation elongation factor activity
YAL038W CDC19 Glycolysisa Pyruvate kinase activity
YNL192W CHS1 Budding Chitin synthase activity
YBR023C CHS3 Response to osmotic stressa Chitin synthase activity
YMR198W CIK1 Meiosisa Microtubule motor activity
YNL298W CLA4 Protein amino acid phosphorylationa Protein serine/threonine kinase activity
YPR119W CLB2 G2/M transition of mitotic cell cyclea Cyclin-dependent protein kinase
regulator activity
YMR199W CLN1 Regulation of cyclin-dependent protein Cyclin-dependent protein kinase
kinase activity regulator activity
YKL179C COY1 Golgi vesicle transport Molecular function unknown
YLR216C CPR6 Protein folding Unfolded protein bindinga
YGR189C CRH1 Biological process unknown Molecular function unknown
YGR218W CRM1 mRNA-nucleus exporta Protein carrier activity
YDR179C CSN9 Adaptation to pheromone during conjugation Molecular function unknown
YBR158W CST13 Negative regulation of exit from mitosisa Protein binding
YIL036W CST6 Transcription initiation from Pol II promotera Specific RNA polymerase II
transcription factor activity
YPL177C CUP9 Transcription initiation from Pol II promotera Specific RNA polymerase II
YOL082W CVT19 Protein-vacuolar targeting Protein binding
YCR017C CWH43 Cell wall organisation and biogenesisa Molecular function unknown
YKL096W CWP1 Cell wall organisation and biogenesis Structural constituent of cell wall
YAL039C CYC3 Cytochrome c-heme linkage Holocytochrome-c synthase activity
YMR135C DCR1 Negative regulation of gluconeogenesis Molecular function unknown
YPL049C DIG1 Invasive growth (sensu Saccharomyces) Transcription factor binding
YOR092W ECM3 Cell wall organisation and biogenesis ATPase activity
YNL280C ERG24 Ergosterol biosynthesis C-14 sterol reductase activity
YJL157C FAR1 Signal transduction during conjugation Cyclin-dependent protein kinase
with cellular fusiona inhibitor activity
YBR040W FIG1 Cellular morphogenesis during conjugation Molecular function unknown
YCR089W FIG2 Cellular morphogenesis during conjugation Molecular function unknown
YMR306W FKS3 Biological process unknown 1,3-beta-glucan synthase activity
YCL026C-A FRM2 Negative regulation of fatty acid metabolism Molecular function unknown
YCL027W FUS1 Conjugation with cellular fusion Molecular function unknown
YMR232W FUS2 Plasma membrane fusion Molecular function unknown
YBL016W FUS3 Protein amino acid phosphorylationa MAP kinase activity
YOL030W GAS5 Biological process unknown 1,3-beta-glucanosyltransferase activity
YMR136W GAT2 Transcription Transcription factor activity
a Annotated to more than one biological process or molecular function in SGD, only the first is shown
212 K. Haynes

YKL104C GFA1 Cell wall chitin biosynthesis Glutamine-fructose-6-phosphate
transaminase (isomerizing) activity
YDR309C GIC2 Establishment of cell polarity (sensu Fungi)a Small GTPase regulatory/interacting
protein activity
YHR005C GPA1 Signal transduction during conjugation GTPase activity
YFL027C GYP8 Vesicle-mediated transport Rab GTPase activator activity
YPR005C HAL1 Positive regulation of transcription Molecular function unknown
from Pol II promotera
YKL189W HYM1 Cytokinesis, completion of separationa Transcriptional repressor activity
YBR157C ICS2 Biological process unknown Molecular function unknown
YNL106C INP52 Cell wall organisation and biogenesisa Inositol-polyphosphate 5-phosphatase
activity
YER019W ISC1 Response to salt stressa Phospholipase C activity
YCL055W KAR4 Meiosisa Transcription regulator activity
YMR065W KAR5 Karyogamy during conjugation with cellular fusion Molecular function unknown
YHR082C KSP1 Protein amino acid phosphorylation Protein serine/threonine kinase activity
YKR061W KTR2 N-linked glycosylationa Mannosyltransferase activity
YCR019W MAK32 Host–pathogen interaction Molecular function unknown
YNL145W MFA2 Signal transduction during conjugation Pheromone activity
YLR332W MID2 Cell wall organisation and biogenesisa Transmembrane receptor activity
YJL186W MNN5 Protein amino acid glycosylation Alpha-1,2-mannosyltransferase
activity
YGR014W MSB2 Establishment of cell polarity (sensu Fungi)a Osmosensor activity
YNL053W MSG5 Protein amino acid dephosphorylationa Prenylated protein tyrosine
phosphatase activity
YHR086W NAM8 Nuclear mRNA splicing, via spliceosomea RNA bindinga
YPR155C NCA2 Aerobic respirationa Molecular function unknown
YGL067W NPY1 NADH metabolism NAD+diphosphatase activity
YBR066C NRG2 Invasive growth (sensu Saccharomyces) Transcriptional repressor activity
YBR060C ORC2 DNA replication initiationa DNA replication origin binding
YDL127W PCL2 Cell cycle Cyclin-dependent protein kinase
regulator activity
YBL017C PEP1 Protein-vacuolar targetinga Signal sequence binding
YKL127W PGM1 Glucose 1-phosphate utilizationa Phosphoglucomutase activity
YGR233C PHO81 Phosphate metabolism Cyclin-dependent protein kinase
inhibitor activity
YNL279W PRM1 Plasma membrane fusion Molecular function unknown
YIL037C PRM2 Conjugation with cellular fusion Molecular function unknown
YPL192C PRM3 Karyogamy Molecular function unknown
YPL156C PRM4 Conjugation with cellular fusion Molecular function unknown
YML047C PRM6 Conjugation with cellular fusion Molecular function unknown
YML046W PRP39 Nuclear mRNA splicing, via spliceosome RNA binding
YGL062W PYC1 Gluconeogenesisa Pyruvate carboxylase activity
YDL135C RDI1 Actin filament organisationa Signal transducer activitya
YPR165W RHO1 Cell wall organisation and biogenesisa GTPase activitya
YNL180C RHO5 Rho protein signal transduction GTPase activity
YNL178W RPS3 Protein biosynthesisa Structural constituent of ribosome
YHR083W SAM35 Mitochondrial outer membrane protein import Protein binding
YHR205W SCH9 Protein amino acid phosphorylationa Protein serine/threonine kinase activity
YBR024W SCO2 Copper ion transport Molecular function unknown
YMR305C SCW10 Conjugation with cellular fusion Glucosidase activity
YGL028C SCW11 Cytokinesis, completion of separation Glucan 1,3-beta-glucosidase activity
YNL272C SEC2 Exocytosis Guanyl-nucleotide exchange factor
activity
YLR105C SEN2 tRNA splicing tRNA-intron endonuclease activity
YPR198W SGE1 Response to druga Xenobiotic-transporting ATPase
activity

YOL031C SIL1 SRP-dependent cotranslational membrane Molecular function unknown
targeting, translocation
YIL123W SIM1 Microtubule cytoskeleton organisation Molecular function unknown
and biogenesis
YBL061C SKT5 Response to osmotic stressa Enzyme activator activity
YBR156C SLI15 Protein amino acid phosphorylationa Protein kinase activator activity
YDR240C SNU56 Nuclear mRNA splicing, via spliceosome mRNA binding
YGR013W SNU71 Nuclear mRNA splicing, via spliceosome RNA binding
YER018C SPC25 Chromosome segregationa Structural constituent of cytoskeleton
YHR152W SPO12 Regulation of exit from mitosisa Molecular function unknown
YOR247W SRL1 Nucleobase, nucleoside, nucleotide Molecular function unknown
and nucleic acid metabolism
YKR091W SRL3 Nucleobase, nucleoside, nucleotide Molecular function unknown
and nucleic acid metabolism
YDR312W SSF2 Ribosomal large subunit assembly rRNA binding
and maintenancea
YDR086C SSS1 Cotranslational membrane targeting Protein transporter activity
YLR452C SST2 Signal transductiona GTPase activator activity
YHR084W STE12 Pseudohyphal growtha Transcription factor activity
YFL026W STE2 Response to pheromone during conjugation Mating-type alpha-factor pheromone
with cellular fusiona receptor activity
YKL209C STE6 Peptide pheromone export ATPase activity, coupled
to transmembrane movement
of substances
YDR310C SUM1 Chromatin silencing at telomerea Transcriptional repressor activity
YPL163C SVS1 Response to chemical substance Molecular function unknown
YJL187C SWE1 G2/M transition of mitotic cell cyclea Protein kinase activity
YNL087W TCB2 Biological process unknown Molecular function unknown
YBR083W TEC1 Positive regulation of transcription Specific RNA polymerase II
from Pol II promotera transcription factor activity
YDR311W TFB1 Transcription initiation from Pol II promotera General RNA polymerase II
YBR162C TOS1 Biological process unknown Molecular function unknown
YBR082C UBC4 Response to stressa Ubiquitin-conjugating enzyme activity
YKR042W UTH1 Mitochondrion organisation and biogenesis Molecular function unknown
YMR197C VTI1 Intra-Golgi transporta v-SNARE activity
YKL095W YJU2 Nuclear mRNA splicing, via spliceosome Molecular function unknown
YBL060W Biological process unknown Molecular function unknown
YCL056C Biological process unknown Molecular function unknown
YGR038C-A Ty element transposition RNA bindinga
YGR038C-B Ty element transposition DNA-directed DNA polymerase
activity
YIL083C Coenzyme A biosynthesis Phosphopantothenate-cysteine
ligase activity
YJR054W Biological process unknown Molecular function unknown
YLR412C-A Biological process unknown Molecular function unknown
YLR413W Biological process unknown Molecular function unknown
YNL024C-A Biological process unknown Molecular function unknown
YOL155C Cell wall organisation and biogenesis Glucosidase activity
YOR129C Response to druga Structural constituent of cytoskeleton
YOR246C Biological process unknown Oxidoreductase activity
YOR343W-A Ty element transposition RNA bindinga
YOR343w-B Ty element transposition DNA-directed DNA polymerase
activity
214 K. Haynes
Inactivation of the single Ace2/Swi5-encoding whereas C. glabrata Ace2 regulates expression of

gene in C. albicans results in defective cell sepa- its sole chitinase-encoding gene (Kamran et al.
ration accompanied by pseudohyhal growth, even 2004; Kelly et al. 2004). It is possible that significant
in the absence of specific inducers. In addition, C. differences in the regulatory domains of Ace2 in the
albicans ace2/ace2 cells have a reduced ability to two species contribute to the observed virulence
adhere to plastic surfaces, cannot form confluent differences. Comparison of the Ace2 regulons in the
biofilms, and are less able to invade solid agar (Kelly two species has not yet been done, but a recent pro-
et al. 2004). The same authors showed that C. albi- teomics investigation identified reproducible and
cans Ace2 is localized to the daughter cell nucleus statistically significant alterations in the levels of
in yeast phase cells and regulates the expression of 61 proteins in wild-type and C. glabrata ace2 cells.
SCW11, DSE1 and one of the C. albicans chitinase- C. glabrata Ace2 regulates proteins involved in pro-
encoding genes CHT3. These data suggest that the tein synthesis/turnover (Cdc33, Cof1, Efb1, Mnp1,
function of C. albicans Ace2 resembles that of S. Rpn10, Rps12, Sec28, Tif34, Ykl056c), transport
cerevisiae Ace2. C. glabrata Ace2 also regulates the (Dss4, Nce103) and interactions with the cellular
expression of CTS1, DSE1 and SCW11 (unpublished environment (Atp2, Atp16). Notably, homologues
data) and has a strong cell separation phenotype of a number of proteins which are down-regulated
(Kamran et al. 2004). In addition, C. glabrata ACE2 in C. glabrata ace2 cells play roles in S. cerevisiae
can functionally complement the cell separation stress responses. Fes1 encodes a Hsp70 nucleotide
defect of S. cerevisiae ace2 cells (unpublished data). exchange factor, and the protein kinase inhibitor
These data strongly suggest that the C. glabrata Lsp1 acts along with Pil1 to down-regulate heat
ACE2 gene also encodes a functional homologue of stress resistance (Zhang et al. 2004) whereas Ahp1
S. cerevisiae Ace2. and Tsa1 have thioredoxin peroxidase activities.
The role of the functionally related Ace2 pro- Can these observations be related to the vir-
teins in fungal virulence has been explored in both ulence phenotype of C. glabrata ace2 cells? In S.
C. albicans and C. glabrata (Kamran et al. 2004; cerevisiae the RAM network regulates Ace2 func-
Kelly et al. 2004). C. albicans ace2/ace2 mutants tion (Nelson et al. 2003). At least two RAM compo-
are completely attenuated for virulence. Tail vein nents, Ssd1 and Kic1, have been annotated to cell
inoculation of equivalent doses of wild-type or re- wall organisation and biogenesis. S. cerevisiae ssd1
constituted ace2/ACE2 cells into DBA/2 mice results mutants are more virulent than are wild-type fungi
in 100% mortality within 5 days. This is accompa- in DBA2 mice, and can induce over-stimulation
nied by significant tissue burden. Conversely, when of pro-inflammatory cytokines from macrophages
C. albicans ace2/ace2 cells are inoculated, all mice (Wheeler et al. 2003). The striking similarity in the
survive to the end of the experiment (28 days) and virulence phenotypes of S. cerevisiae ssd1 and C.
only about 40% have C. albicans recoverable from glabrata ace2 cells suggests that they might share
kidneys (Kelly et al. 2004). a common basis. Interestingly, S. cerevisiae ssd1
In stark and unexpected contrast, inactivation cells have significant alterations in cell wall com-
of ACE2 in C. glabrata results in increased viru- position. The question therefore arises – do alter-
lence. C. glabrata ace2 cells are about 200-fold more ations in the C. glabrata proteome suggest a role
virulent than is the wild-type strain (Kamran et al. for Ace2 in cell wall organisation and biogene-
2004). Infection is characterized by fungal escape sis? It is known that some proteins which were
from the vasculature, tissue penetration, prolifer- down-regulated in ace2 cells (Stead et al. 2005)
ation in vivo and massive over-stimulation of the have roles in cell growth, polarity and morphogen-
pro-inflammatory arm of the innate immune sys- esis, and cell wall remodelling (Act1, Ahp1, Cdc33,
tem. Systemic levels of IL-6 and TNF-α, 18 h post- Cof1, Dss4, Fsa1, Sec28, Tub2). By analogy with
infection, were 16- and 38-fold higher respectively S. cerevisiae, altered levels of many of these pro-
in ace2-infected mice compared to mice infected teins would result in significant changes in the C.
with an ACE2-reconstituted strain (Kamran et al. glabrata cell wall and secretome (Roberts et al.
2004). The obvious question arises – why does in- 1997). Ace2 also regulates the expression of the
activation of proteins with apparently similar func- endochitinase-encoding gene CTS1 in C. glabrata,
tions result in such different virulence phenotypes? which plays a major role in cell wall remodelling
Unfortunately, the answer still remains elusive. at the site of cell separation. Additionally, proteins
However, C. albicans Ace2 regulates the expression identified in this study (YJL123c) have been local-
of only one of three chitinase genes in this species ized to COPI-coated vesicles. Other Ace2-regulated
proteins are involved in signalling (e.g. Bmh1). Elaine Bignell and Helen Findon, and Herb Arst for helpful
We have previously demonstrated that inactivation discussions. This work was supported by grants from the
of transcription factors activated by well-defined BBSRC (BBS/B/10331) and the Wellcome Trust (072420).
signalling pathways can result in alterations to C.
glabrata cell wall integrity (Calcagno et al. 2003).
The 14-3-3 proteins Bmh1 and Bmh2 are essen-
tial for pseudohyphal production in S. cerevisiae, References
a morphological event which has major effects on
cell wall structure (Roberts et al. 1997). Further- Alarco AM, Raymond M (1999) The bZip transcription fac-
more, the central carbon metabolic functions regu- tor Cap1p is involved in multidrug resistance and ox-
idative stress response in Candida albicans. J Bacteriol
lated by Ace2 are involved in the provision of sugars 181:700–708
essential for the biosynthesis of cell wall carbohy- Aparicio OM, Billington BL, Gottschling DE (1991)
drates and the post-translational modification of Modifiers of position effect are shared between
cell surface proteins. telomeric and silent mating-type loci in S. cerevisiae.
Cell 66:1279–1287
Is there any other evidence to support the view Asmundsdottir LR, Erlendsdottir H, Gottfredsson M (2002)
that Ace2 regulates cell wall biogenesis? Systems Increasing incidence of candidemia: results from a 20-
analysis of S. cerevisiae Ace2 functional connec- year nationwide study in Iceland. J Clin Microbiol
tions using the STRING tool (string.embl.de) sug- 40:3489–3492
gests important links to a number of genes encod- Bagnat M, Simons K (2002) Cell surface polarization during
yeast mating. Proc Natl Acad Sci USA 99:14183–14188
ing functions which significantly affect cell wall bi- Beck-Sague C, Jarvis WR (1993) Secular trends in the
ology. These include ubiquitin-dependent protein epidemiology of nosocomial fungal infections in
turnover (e.g. Ubp15, Rsp5), glycoprotein degrada- the United States, 1980–1990. National Nosoco-
tion (Mnl1), maintenance of the cell integrity path- mial Infections Surveillance System. J Infect Dis
167:1247–1251
way (Bck1), and transcriptional regulation (Fkh1, Benni ML, Neigeborn L (1997) Identification of a new class
Fkh2). Taken together, these observations begin to of negative regulators affecting sporulation-specific
suggest a mechanistic explanation for the increased gene expression in yeast. Genetics 147:1351–1366
virulence of C. glabrata ace2 mutants. Analysis of Bensen ES, Filler SG, Berman J (2002) A forkhead tran-
the C. albicans Ace2 regulon and proteome should scription factor is important for true hyphal as well
as yeast morphogenesis in Candida albicans. Eukaryot
reveal differences which may explain the variation Cell 1:787–798
in virulence seen in the two species as a result of Bignell E, Negrete-Urtasun S, Calcagno AM, Haynes K, Arst
inactivation of Ace2 (Kamran et al. 2004; Kelly et al. HN Jr, Rogers T (2005) The Aspergillus pH-responsive
2004). transcription factor PacC regulates virulence. Mol Mi-
crobiol 55(4):1072–1084
Biswas K, Rieger KJ, Morschhauser J (2003) Functional anal-
ysis of CaRAP1, encoding the Repressor/activator pro-
IV. Conclusion tein 1 of Candida albicans. Gene 307:151–158
Borneman AR, Hynes MJ, Andrianopoulos A (2001) An
STE12 homolog from the asexual, dimorphic fungus
The analysis of fungal transcriptional regulation Penicillium marneffei complements the defect in sex-
has revealed many insights into the pathobiology of ual development of an Aspergillus nidulans steA mu-
tant. Genetics 157:1003–1014
C. albicans and is beginning to shed light on disease Brand A, MacCallum DM, Brown AJ, Gow NA, Odds FC
causation by C. glabrata. This analysis is essential (2004) Ectopic expression of URA3 can influence the
to understand the host–Candida relationship. This virulence phenotypes and proteome of Candida albi-
relationship can be meaningfully assessed also by cans but can be overcome by targeted reintegration of
using information available for other species, espe- URA3 at the RPS10 locus. Eukaryot Cell 3:900–909
Braun BR, Johnson AD (1997) Control of filament formation
cially S. cerevisiae, and no doubt many attributes in Candida albicans by the transcriptional repressor
will be shared. However, the observation that C. TUP1. Science 277:105–109
albicans and C. glabrata ace2 mutants display dra- Brockert PJ, Lachke SA, Srikantha T, Pujol C, Galask R,
matically different virulence phenotypes suggests, Soll DR (2003) Phenotypic switching and mating type
switching of Candida glabrata at sites of colonization.
not surprisingly, that species-specific processes are Infect Immun 71:7109–7118
also at work in maintaining host–pathogen inter- Calcagno A, Bignell E, Warn P, Jones MD, Denning DW,
actions. Mühlschlegel FA, Rogers TR, Haynes K (2003) Candida
glabrata STE12 is required for wild-type virulence and
Acknowledgements. I would like to thank all members of the nitrogen starvation induced filamentation. Mol Micro-
Imperial College Molecular Mycology Group, in particular biol 50:1309–1318
216 K. Haynes
Calderone RA (2002) Candida and Candidiasis. ASM Press, Gale CA, Bendel CM, McClellan M, Hauser M, Becker JM,
Washington, DC Berman J, Hostetter MK (1998) Linkage of adhesion,
Cappellaro C, Mrsa V, Tanner W (1998) New potential cell filamentous growth, and virulence in Candida albicans
wall glucanases of Saccharomyces cerevisiae and their to a single gene, INT1. Science 279:1355–1358
involvement in mating. J Bacteriol 180:5030–5037 Gancedo JM (2001) Control of pseudohyphae formation in
Caracuel Z, Roncero MI, Espeso EA, Gonzalez-Verdejo CI, Saccharomyces cerevisiae. FEMS Microbiol Rev 25:107–
Garcia-Maceira FI, Di Pietro A (2003) The pH sig- 123
nalling transcription factor PacC controls virulence in Gudlaugsson O, Gillespie S, Lee K, Vande Berg J, Hu J,
the plant pathogen Fusarium oxysporum. Mol Micro- Messer S, Herwaldt L, Pfaller M, Diekema D (2003)
biol 48:765–779 Attributable mortality of nosocomial candidemia, re-
Chang YC, Wickes BL, Miller GF, Penoyer LA, Kwon-Chung visited. Clin Infect Dis 37:1172–1177
KJ (2000) Cryptococcus neoformans STE12alpha regu- Gustin MC, Albertyn J, Alexander M, Davenport K (1998)
lates virulence but is not essential for mating. J Exp MAP kinase pathways in the yeast Saccharomyces cere-
Med 191:871–882 visiae. Microbiol Mol Biol Rev 62:1264–1300
Chang YC, Penoyer LA, Kwon-Chung KJ (2001) The sec- Harbison CT, Gordon DB, Lee TI, Rinaldi NJ, Macisaac KD,
ond STE12 homologue of Cryptococcus neoformans is Danford TW, Hannett NM, Tagne JB, Reynolds DB,
MATa-specific and plays an important role in viru- Yoo J et al. (2004) Transcriptional regulatory code of
lence. Proc Natl Acad Sci USA 98:3258–3263 a eukaryotic genome. Nature 431:99–104
Chen YC, Chang SC, Sun CC, Yang LS, Hsieh WC, Luh KT Haw R, Yarragudi AD, Uemura H (2001) Isolation of a Can-
(1997) Secular trends in the epidemiology of noso- dida glabrata homologue of RAP1, a regulator of tran-
comial fungal infections at a teaching hospital in scription and telomere function in Saccharomyces cere-
Taiwan, 1981 to 1993. Infect Control Hosp Epidemiol visiae. Yeast 18:1277–1284
18:369–375 Heiman MG, Walter P (2000) Prm1p, a pheromone-
Colman-Lerner A, Chin TE, Brent R (2001) Yeast Cbk1 and regulated multispanning membrane protein,
Mob2 activate daughter-specific genetic programs to facilitates plasma membrane fusion during yeast
induce asymmetric cell fates. Cell 107:739–750 mating. J Cell Biol 151:719–730
Cormack BP, Falkow S (1999) Efficient homologous and Hensel M, Shea JE, Gleeson C, Jones MD, Dalton E,
illegitimate recombination in the opportunistic yeast Holden DW (1995) Simultaneous identification of
pathogen Candida glabrata. Genetics 151:979–987 bacterial virulence genes by negative selection.
Cormack BP, Ghori N, Falkow S (1999) An adhesin of the Science 269:400–403
yeast pathogen Candida glabrata mediating adherence Hoyer LL (2001) The ALS gene family of Candida albicans.
to human epithelial cells. Science 285:578–582 Trends Microbiol 9:176–180
Davis D, Edwards JE, Mitchell AP, Ibrahim AS (2000) Can- Inglis DO, Johnson AD (2002) Ash1 protein, an asymmet-
dida albicans RIM101 pH response pathway is required rically localized transcriptional regulator, controls fil-
for host-pathogen interactions. Infect Immun 68:5953– amentous growth and virulence of Candida albicans.
5959 Mol Cell Biol 22:8669–8680
Davis DA, Bruno VM, Loza L, Filler SG, Mitchell AP (2002) Ishii N, Yamamoto M, Yoshihara F, Arisawa M, Aoki Y (1997)
Candida albicans Mds3p, a conserved regulator of pH Biochemical and genetic characterization of Rbf1p,
responses and virulence identified through insertional a putative transcription factor of Candida albicans.
mutagenesis. Genetics 162:1573–1581 Microbiology 143:429–435
De Bernardis F, Muhlschlegel FA, Cassone A, Fonzi WA Jarvis WR (1995) Epidemiology of nosocomial fungal infec-
(1998) The pH of the host niche controls gene expres- tions, with emphasis on Candida species. Clin Infect
sion in and virulence of Candida albicans. Infect Im- Dis 20:1526–1530
mun 66:3317–3325 Kadosh D, Johnson AD (2001) Rfg1, a protein related to
De Las Penas A, Pan SJ, Castano I, Alder J, Cregg R, Cor- the Saccharomyces cerevisiae hypoxic regulator Rox1,
mack BP (2003) Virulence-related surface glycopro- controls filamentous growth and virulence in Candida
teins in the yeast pathogen Candida glabrata are en- albicans. Mol Cell Biol 21:2496–2505
coded in subtelomeric clusters and subject to RAP1- Kagi JH, Schaffer A (1988) Biochemistry of metallothionein.
and SIR-dependent transcriptional silencing. Genes Biochemistry 27:8509–8515
Dev 17:2245–2258 Kaiser B, Munder T, Saluz HP, Kunkel W, Eck R (1999) Identi-
Drori N, Kramer-Haimovich H, Rollins J, Dinoor A, fication of a gene encoding the pyruvate decarboxylase
Okon Y, Pines O, Prusky D (2003) External pH and gene regulator CaPdc2p from Candida albicans. Yeast
nitrogen source affect secretion of pectate lyase 15:585–591
by Colletotrichum gloeosporioides. Appl Environ Kamran M, Calcagno AM, Findon H, Bignell E, Jones MD,
Microbiol 69:3258–3262 Warn P, Hopkins P, Denning DW, Butler G, Rogers T
Eck R, Stoyan, T, Kunkel W (2001) The centromere-binding et al. (2004) Inactivation of transcription factor gene
factor Cbf1p from Candida albicans complements the ACE2 in the fungal pathogen Candida glabrata results
methionine auxotrophic phenotype of Saccharomyces in hypervirulence. Eukaryot Cell 3:546–552
cerevisiae. Yeast 18:1047–1052 Kelly R, Kwon-Chung KJ (1992) A zinc finger protein from
Frieman MB, McCaffery JM, Cormack BP (2002) Modu- Candida albicans is involved in sucrose utilization.
lar domain structure in the Candida glabrata adhesin J Bacteriol 174:222–232
Epa1p, a beta1,6 glucan-cross-linked cell wall protein. Kelly MT, MacCallum DM, Clancy SD, Odds FC, Brown AJ,
Mol Microbiol 46:479–492 Butler G (2004) The Candida albicans CaACE2 gene
affects morphogenesis, adherence and virulence. Mol gene PHR1 with an inverted pattern of pH-dependent
Microbiol 53:969–983 expression. Mol Cell Biol 17:5960–5967
Khalaf RA, Zitomer RS (2001) The DNA binding protein Murad AM, Leng P, Wishart JA, Straffon M, Schnell NF,
Rfg1 is a repressor of filamentation in Candida albi- Talibi D, Marechal D, d’Enfert C, Gaillardin C, Brown
cans. Genetics 157:1503–1512 AJ (2001) NRG1 represses yeast-hypha morphogenesis
Khoo B, Brophy B, Jackson SP (1994) Conserved functional and hypha-specific gene expression in Candida albi-
domains of the RNA polymerase III general transcrip- cans. EMBO J 20:4742–4752
tion factor BRF. Genes Dev 8:2879–2890 Nantel A, Dignard D, Bachewich C, Harcus D, Marcil A,
Koch KA, Thiele DJ (1999) Functional analysis of a ho- Bouin AP, Sensen CW, Hogues H, Van Het Hoog M,
mopolymeric (dA-dT) element that provides nucleo- Gordon P et al. (2002) Transcription profiling of
somal access to yeast and mammalian transcription Candida albicans cells undergoing the yeast-to-hyphal
factors. J Biol Chem 274:23752–23760 transition. Mol Biol Cell 13:3452–3465
Kuranda MJ, Robbins PW (1991) Chitinase is required for Navarro-Garcia F, Sanchez M, Nombela C, Pla J (2001) Vir-
cell separation during growth of Saccharomyces cere- ulence genes in the pathogenic yeast Candida albicans.
visiae. J Biol Chem 266:19758–19767 FEMS Microbiol Rev 25:245–268
Lachke SA, Srikantha T, Tsai LK, Daniels K, Soll DR (2000) Nelson B, Kurischko C, Horecka J, Mody M, Nair P, Pratt L,
Phenotypic switching in Candida glabrata involves Zougman A, McBroom LD, Hughes TR, Boone C et al.
phase-specific regulation of the metallothionein gene (2003) RAM: a conserved signaling network that regu-
MT-II and the newly discovered hemolysin gene HLP. lates Ace2p transcriptional activity and polarized mor-
Infect Immun 68:884–895 phogenesis. Mol Biol Cell 14:3782–3803
Lane S, Birse C, Zhou S, Matson R, Liu H (2001) DNA ar- Olaechea PM, Palomar M, Leon-Gil C, Alvarez-Lerma F,
ray studies demonstrate convergent regulation of vir- Jorda R, Nolla-Salas J, Leon-Regidor MA (2004) Eco-
ulence factors by Cph1, Cph2, and Efg1 in Candida nomic impact of Candida colonization and Candida
albicans. J Biol Chem 276:48988–48996 infection in the critically ill patient. Eur J Clin Micro-
Lee TI, Rinaldi NJ, Robert F, Odom D, Bar-Joseph Z, Ger- biol Infect Dis 23:323–330
ber GK, Hannett NM, Harbison CT, Thompson CM, Si- Penalva MA, Arst HN Jr (2004) Recent advances in the char-
mon I et al. (2002) Transcriptional regulatory networks acterisation of ambient pH regulation of gene expres-
in Saccharomyces cerevisiae. Science 298:799–804 sion in filamentous fungi and yeasts. Ann Rev Micro-
Limjindaporn T, Khalaf RA, Fonzi WA (2003) Nitrogen biol 58:425–451
metabolism and virulence of Candida albicans require Pfaller MA, Diekema DJ (2002) Role of sentinel surveillance
the GATA-type transcriptional activator encoded by of candidemia: trends in species distribution and an-
GAT1. Mol Microbiol 50:993–1004 tifungal susceptibility. J Clin Microbiol 40:3551–3557
Liu H, Kohler J, Fink GR (1994) Suppression of hyphal for- Pfaller MA, Diekema DJ, Jones RN, Sader HS, Fluit AC, Hol-
mation in Candida albicans by mutation of a STE12 lis RJ, Messer SA (2001) International surveillance of
homolog. Science 266:1723–1726 bloodstream infections due to Candida species: fre-
Lo HJ, Kohler JR, DiDomenico B, Loebenberg D, Caccia- quency of occurrence and in vitro susceptibilities to
puoti A, Fink GR (1997) Nonfilamentous C. albicans fluconazole, ravuconazole, and voriconazole of iso-
mutants are avirulent. Cell 90:939–949 lates collected from 1997 through 1999 in the SENTRY
Lorenz RT, Parks LW (1992) Cloning, sequencing, and dis- Antimicrobial Surveillance Program. J Clin Microbiol
ruption of the gene encoding sterol C-14 reductase in 39:3254–3259
Saccharomyces cerevisiae. DNA Cell Biol 11:685–692 Pfaller MA, Messer SA, Boyken L, Hollis RJ, Rice C,
Manning BD, Barrett JG, Wallace JA, Granok H, Snyder M Tendolkar S, Diekema DJ (2004) In vitro activities of
(1999) Differential regulation of the Kar3p kinesin- voriconazole, posaconazole, and fluconazole against
related protein by two associated proteins, Cik1p and 4169 clinical isolates of Candida spp. and Cryptococcus
Vik1p. J Cell Biol 144:1219–1233 neoformans collected during 2001 and 2002 in the
McBride HJ, Yu Y, Stillman DJ (1999) Distinct regions of ARTEMIS global antifungal surveillance program.
the Swi5 and Ace2 transcription factors are required Diagn Microbiol Infect Dis 48:201–205
for specific gene activation. J Biol Chem 274:21029– Pi H, Chien CT, Fields S (1997) Transcriptional activation
21036 upon pheromone stimulation mediated by a small do-
McCaffrey G, Clay FJ, Kelsay K, Sprague GF Jr (1987) Identi- main of Saccharomyces cerevisiae Ste12p. Mol Cell Biol
fication and regulation of a gene required for cell fusion 17:6410–6418
during mating of the yeast Saccharomyces cerevisiae. Pittet D, Li N, Woolson RF, Wenzel RP (1997) Mi-
Mol Cell Biol 7:2680–2690 crobiological factors influencing the outcome of
McCreath KJ, Specht CA, Robbins PW (1995) Molecular nosocomial bloodstream infections: a 6-year val-
cloning and characterization of chitinase genes from idated, population-based model. Clin Infect Dis
Candida albicans. Proc Natl Acad Sci USA 92:2544– 24:1068–1078
2548 Ramon AM, Porta A, Fonzi WA (1999) Effect of environ-
Moreno I, Pedreno Y, Maicas S, Sentandreu R, Herrero E, mental pH on morphological development of Candida
Valentin E (2003) Characterization of a Candida albi- albicans is mediated via the PacC-related transcription
cans gene encoding a putative transcriptional factor factor encoded by PRR2. J Bacteriol 181:7524–7530
required for cell wall integrity. FEMS Microbiol Lett Ren B, Robert F, Wyrick JJ, Aparicio O, Jennings EG, Simon I,
226:159–167 Zeitlinger J, Schreiber J, Hannett N, Kanin E et al. (2000)
Muhlschlegel FA, Fonzi WA (1997) PHR2 of Candida albi- Genome-wide location and function of DNA binding
cans encodes a functional homolog of the pH-regulated proteins. Science 290:2306–2309
218 K. Haynes
Rine J, Herskowitz I (1987) Four genes responsible for a po- mentation of a pdr1 pdr3 mutation in Saccharomyces
sition effect on expression from HML and HMR in cerevisiae. J Bacteriol 181:231–240
Saccharomyces cerevisiae. Genetics 116:9–22 Thorvaldsen JL, Sewell AK, McCowen CL, Winge DR (1993)
Roberts RL, Mosch HU, Fink GR (1997) 14-3-3 proteins Regulation of metallothionein genes by the ACE1 and
are essential for RAS/MAPK cascade signaling dur- AMT1 transcription factors. J Biol Chem 268:12512–
ing pseudohyphal development in S. cerevisiae. Cell 12518
89:1055–1065 Tortorano AM, Peman J, Bernhardt H, Klingspor L, Kib-
Roberts CJ, Nelson B, Marton MJ, Stoughton R, Meyer MR, bler CC, Faure O, Biraghi E, Canton E, Zimmermann K,
Bennett HA, He YD, Dai H, Walker WL, Hughes TR Seaton S et al. (2004) Epidemiology of candidaemia in
et al. (2000) Signaling and circuitry of multiple MAPK Europe: results of 28-month European Confederation
pathways revealed by a matrix of global gene expres- of Medical Mycology (ECMM) hospital-based surveil-
sion profiles. Science 287:873–880 lance study. Eur J Clin Microbiol Infect Dis 23:317–
Rollins JA (2003) The Sclerotinia sclerotiorum pac1 gene is 322
required for sclerotial development and virulence. Mol Tripathi G, Wiltshire C, Macaskill S, Tournu H, Budge S,
Plant Microbe Interact 16:785–795 Brown AJ (2002) Gcn4 co-ordinates morphogenetic
Saville SP, Lazzell AL, Monteagudo C, Lopez-Ribot JL (2003) and metabolic responses to amino acid starvation in
Engineered control of cell morphology in vivo reveals Candida albicans. EMBO J 21:5448–5456
distinct roles for yeast and filamentous forms of Can- Vallim MA, Miller KY, Miller BL (2000) Aspergillus SteA
dida albicans during infection. Eukaryot Cell 2:1053– (sterile12-like) is a homeodomain-C2 /H2 -Zn+2 finger
1060 transcription factor required for sexual reproduction.
Schweizer A, Rupp S, Taylor BN, Rollinghoff M, Schroppel K Mol Microbiol 36:290–301
(2000) The TEA/ATTS transcription factor CaTec1p Wang Z, Wilson WA, Fujino MA, Roach PJ (2001) The yeast
regulates hyphal development and virulence in Can- cyclins Pc16p and Pc17p are involved in the control
dida albicans. Mol Microbiol 38:435–445 of glycogen storage by the cyclin-dependent protein
Shore D (1994) RAP1: a protean regulator in yeast. Trends kinase Pho85p. FEBS Lett 506:277–280
Genet 10:408–412 Wang Y, Shirogane T, Liu D, Harper JW, Elledge SJ (2003)
Shore D (1997) Telomere length regulation: getting the mea- Exit from exit: resetting the cell cycle through Amn1
sure of chromosome ends. Biol Chem 378:591–597 inhibition of G protein signaling. Cell 112:697–709
Shore D, Nasmyth K (1987) Purification and cloning of Weigel D, Jurgens G, Kuttner F, Seifert E, Jackle H (1989) The
a DNA binding protein from yeast that binds to both homeotic gene fork head encodes a nuclear protein and
silencer and activator elements. Cell 51:721–732 is expressed in the terminal regions of the Drosophila
Simon I, Barnett J, Hannett N, Harbison CT, Rinaldi NJ, embryo. Cell 57:645–658
Volkert TL, Wyrick JJ, Zeitlinger J, Gifford DK, Jaakkola Weiss EL, Kurischko C, Zhang C, Shokat K, Drubin DG,
TS et al. (2001) Serial regulation of transcriptional reg- Luca FC (2002) The Saccharomyces cerevisiae Mob2p-
ulators in the yeast cell cycle. Cell 106:697–708 Cbk1p kinase complex promotes polarized growth and
Singh P, Ganesan K, Malathi K, Ghosh D, Datta A (1994) acts with the mitotic exit network to facilitate daughter
ACPR, a STE12 homologue from Candida albicans, is cell-specific localization of Ace2p transcription factor.
a strong inducer of pseudohyphae in Saccharomyces J Cell Biol 158:885–900
cerevisiae haploids and diploids. Biochem Biophys Res Wey SB, Mori M, Pfaller MA, Woolson RF, Wenzel RP (1988)
Commun 205:1079–1085 Hospital-acquired candidemia. The attributable mor-
Smith HE, Su SS, Neigeborn L, Driscoll SE, Mitchell AP tality and excess length of stay. Arch Intern Med
(1990) Role of IME1 expression in regulation of meiosis 148:2642–2645
in Saccharomyces cerevisiae. Mol Cell Biol 10:6103– Wheeler RT, Kupiec M, Magnelli P, Abeijon C, Fink GR
6113 (2003) A Saccharomyces cerevisiae mutant with in-
Staab JF, Bradway SD, Fidel PL, Sundstrom P (1999) Adhe- creased virulence. Proc Natl Acad Sci USA 100:2766–
sive and mammalian transglutaminase substrate prop- 2770
erties of Candida albicans Hwp1. Science 283:1535– White TC, Agabian N (1995) Candida albicans secreted as-
1538 partyl proteinases: isoenzyme pattern is determined
Stead D, Findon H, Yin Z, Walker J, Selway L, Cash P, Du- by cell type, and levels are determined by environ-
jon BA, Hennequin C, Brown AJP, Haynes K (2005) mental factors. J Bacteriol 177:5215–5221
Proteomic changes associated with inactivation of the Whiteway M, Dignard D, Thomas DY (1992) Dominant neg-
Candida glabrata ACE2 virulence-moderating gene. ative selection of heterologous genes: isolation of Can-
Proteomics 5:1838–1848 dida albicans genes that interfere with Saccharomyces
Stoldt VR, Sonneborn A, Leuker CE, Ernst JF (1997) Efg1p, cerevisiae mating factor-induced cell cycle arrest. Proc
an essential regulator of morphogenesis of the human Natl Acad Sci USA 89:9410–9414
pathogen Candida albicans, is a member of a conserved Wickes BL, Edman U, Edman JC (1997) The Cryptococ-
class of bHLH proteins regulating morphogenetic pro- cus neoformans STE12alpha gene: a putative Saccha-
cesses in fungi. EMBO J 16:1982–1991 romyces cerevisiae STE12 homologue that is mating
Su SS, Mitchell AP (1993) Molecular characterization of the type specific. Mol Microbiol 26:951–960
yeast meiotic regulatory gene RIM1. Nucleic Acids Res Wisplinghoff H, Seifert H, Tallent SM, Bischoff T, Wenzel RP,
21:3789–3797 Edmond MB (2003) Nosocomial bloodstream infec-
Talibi D, Raymond M (1999) Isolation of a putative tions in pediatric patients in United States hospitals:
Candida albicans transcriptional regulator involved epidemiology, clinical features and susceptibilities. Pe-
in pleiotropic drug resistance by functional comple- diatr Infect Dis J 22:686–691
Wolfe KH, Shields DC (1997) Molecular evidence for an protein kinase-like kinase Pkh1p and downstream
ancient duplication of the entire yeast genome. Nature signaling pathways Pkc1p and Ypk1p. J Biol Chem
387:708–713 279:22030–22038
Yang X, Talibi D, Weber S, Poisson G, Raymond M (2001) Zhou PB, Thiele DJ (1991) Isolation of a metal-activated
Functional isolation of the Candida albicans FCR3 gene transcription factor gene from Candida glabrata by
encoding a bZip transcription factor homologous to complementation in Saccharomyces cerevisiae. Proc
Saccharomyces cerevisiae Yap3p. Yeast 18:1217–1225 Natl Acad Sci USA 88:6112–6116
Yeong FM, Lim HH, Surana U (2002) MEN, destruction and Zhou P, Thiele DJ (1993) Rapid transcriptional autoregula-
separation: mechanistic links between mitotic exit and tion of a yeast metalloregulatory transcription factor
cytokinesis in budding yeast. Bioessays 24:659–666 is essential for high-level copper detoxification. Genes
Young LY, Lorenz MC, Heitman J (2000) A STE12 homolog is Dev 7:1824–1835
required for mating but dispensable for filamentation Zhu Z, Thiele DJ (1996) A specialized nucleosome modu-
in Candida lusitaniae. Genetics 155:17–29 lates transcription factor access to a C. glabrata metal
Zaragoza O, Rodriguez C, Gancedo C (2000) Isolation of responsive promoter. Cell 87:459–470
the MIG1 gene from Candida albicans and effects of Zhu G, Spellman PT, Volpe T, Brown PO, Botstein D,
its disruption on catabolite repression. J Bacteriol Davis TN, Futcher B (2000) Two yeast forkhead genes
182:320–326 regulate the cell cycle and pseudohyphal growth.
Zhang X, Lester RL, Dickson RC (2004) Pil1p and Lsp1p Nature 406:90–94
negatively regulate the 3-phosphoinositide-dependent
12 Using Genomics to Study the Life Cycle of Histoplasma capsulatum
A. Sil1
CONTENTS individuals. Approximately 500,000 infections

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 221 are thought to occur every year in the US alone
II. Previous Use of Functional Genomics (Eissenberg and Goldman 1991; Marques et al.
in H. capsulatum . . . . . . . . . . . . . . . . . . . . . . 222 2000; Woods 2003; Wheat and Kauffman 2003).
A. Generation of a Shotgun Genomic Immunocompromised individuals tend to develop
Microarray . . . . . . . . . . . . . . . . . . . . . . . . 222
B. Studying Differences Between Yeast progressive, disseminated disease that can be
and Mycelia . . . . . . . . . . . . . . . . . . . . . . . 224 fatal. Specifically, H. capsulatum has been docu-
III. Current Genome-Sequencing Projects mented to cause severe disease in patients with
for H. capsulatum . . . . . . . . . . . . . . . . . . . . . 225 lymphoreticular neoplasms and those undergoing
A. Choice of Genomes to be Sequenced . . . . 225
B. From Genome Sequence to Functional immunosuppressant chemotherapy (Kauffman
Genomics . . . . . . . . . . . . . . . . . . . . . . . . . 227 et al. 1978; Bryan and DiSalvo 1979; Brown 1990;
IV. Other Applications of Functional Genomics Samonis and Bafaloukos 1992; Bradsher 1996). H.
in H. capsulatum . . . . . . . . . . . . . . . . . . . . . . 227 capsulatum is endemic in the Ohio River Valley
A. Macrophages . . . . . . . . . . . . . . . . . . . . . . 227 through the mid-western United States into Texas,
B. Understanding the Development
and Germination of Conidia . . . . . . . . . . 229 and is a leading pathogen affecting AIDS patients
C. Clinical Isolates . . . . . . . . . . . . . . . . . . . . 230 in these endemic areas (Sternberg 1994).
D. Study of Mating Type . . . . . . . . . . . . . . . . 231 A description of the H. capsulatum life cycle
V. Technologies that Interface with Genomics . 231 reveals that this organism responds to its envi-
A. Gene Disruption . . . . . . . . . . . . . . . . . . . . 231
B. RNA Interference . . . . . . . . . . . . . . . . . . . 231 ronment (either the soil or a mammalian host)
C. Insertional Mutagenesis . . . . . . . . . . . . . . 232 in a distinct manner (Fig. 12.1). H. capsulatum is
VI. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 233 a haploid dimorphic fungus that exists in a hyphal
References . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 form in soil. This mycelial form of H. capsulatum
is competent for mating if it encounters a strain
of opposite mating type (see Sect. IV.D), and the
resultant diploid undergoes meiosis to produce
I. Introduction eight meiotic spores per ascus (Kwon-Chung
1973). Additionally, the mycelial form of the
Histoplasma capsulatum is a fungal pathogen organism produces vegetative spores, or conidia.
on the brink of a molecular revolution. Years Conidia and hyphal fragments become airborne
of tool-building (Goldman 1995; Magrini and when the soil is disrupted, and are then inhaled
Goldman 2001; Woods 2002) and ongoing by the host and taken up by macrophages and
genome-sequencing projects have set the stage other phagocytic cells (Eissenberg and Goldman
for rich molecular experimentation with this 1991; Bullock 1993; Woods 2003). Once inside
organism. The goal of this chapter is to review the the host, conversion of the mycelial form to
past, present, and future of functional genomics in a budding yeast form is triggered within hours.
H. capsulatum. Yeast cells evade phagocytic killing and multiply
H. capsulatum, the etiologic agent of histo- within alveolar macrophages. Subsequently, yeast
plasmosis, is a primary fungal pathogen that cells use phagocytic cells as vehicles to spread
infects healthy as well as immunocompromised to multiple organs of the reticuloendothelial
1 Department of Microbiology and Immunology, University of Cal-
system such as the spleen, liver, lymph nodes,
ifornia, San Francisco, 513 Parnassus, P.O. Box 0414, San Francisco, and bone marrow. In patients with disseminated
CA 94143-0414, USA disease, other organs such as the skin, heart,
The Mycota XIII
Fungal Genomics
222 A. Sil
the systemic dimorphic fungi, little is understood

about how any of these organisms sense tempera-
ture and trigger the appropriate response. For H.
capsulatum, the ability to grow in both forms is key
to establishing and maintaining disease in the host.
The mycelial form facilitates the initiation of infec-
tion because it becomes airborne and produces in-
fectious conidiospores. Transformation to the yeast
form is thought to be essential for the fungus to
survive and proliferate in the host, since mycelial
cells that are artificially blocked from shifting to
the yeast form are unable to cause disease (Maresca
et al. 1977; Medoff et al. 1986b). This type of dimor-
phic switch is a key virulence determinant for many
Fig. 12.1. Life cycle of Histoplasma capsulatum. The figure
depicts different stages of the H. capsulatum life cycle in the pathogenic fungi (Borges-Walmsley and Walms-
soil and in the host. The mycelial form grows in the soil (or at ley 2000). Once inside the host cell, the H. capsu-
25 ◦ C in the laboratory), and can mate and undergo meiosis latum yeast cell subverts the normal response of
to form ascospores. The mycelia can also generate vegetative macrophages to microbes by unknown means, as
conidiospores by a process known as conidiation. There are
at least two types of conidiospores, macro- or microconidia.
described in Sect. IV. A below. Thus, H. capsulatum
These cells can germinate and adopt either the mycelial form can be thought of as a model system for studying
or the yeast form. The natural route of infection occurs when temperature-regulated morphology as well as in-
the host inhales conidia or hyphal fragments that aerosolize teractions between macrophages and intracellular
when the soil is disrupted. Once inside the host, these cells eukaryotic pathogens.
convert to the yeast form. The yeast form is infectious if
introduced into an animal in the laboratory, but there is With the advent of a genome sequence for H.
normally no host-to-host transmission of the yeast form. capsulatum, we are at the dawn of the heyday of
Schematic courtesy of Davina Hocking Murray functional genomics in this organism. Especially
because previous genetic studies have been limited,
functional genomics is a powerful methodology for
brain, adrenal glands, and gastrointestinal tract identifying candidate genes that are implicated in
can be colonized. In the majority of hosts, the a particular process by virtue of regulated gene ex-
immune system curtails the infection by means pression. Because it is still early days for genome-
that are not yet clearly defined, but that depend wide analyses in H. capsulatum, this chapter will
on the development of cell-mediated immunity cover (1) previous uses of functional genomics, (2)
and the corresponding activation of macrophages the status of the genome projects for this organism,
(Eissenberg and Goldman 1991; Newman 1999; (3) future uses of functional genomics for stud-
Huffnagle and Deepe 2003). ies of interesting aspects of H. capsulatum biology,
One of the most interesting aspects of H. cap- and (4) molecular tools that interface with genomic
sulatum biology is that it changes morphology and analyses.
associated characteristics in response to its envi-
ronment, as described above. All of the systemic
dimorphic fungal pathogens, including H. capsula-
tum, Coccidioides immitis, Blastomyces dermatidi- II. Previous Use of Functional
tis, and Paracoccidioides brasiliensis, use tempera- Genomics in H. capsulatum
ture as a key signal to determine morphology and
cell specialization (Maresca and Kobayashi 1989). A. Generation of a Shotgun
The transformation of H. capsulatum mycelial cells Genomic Microarray
to yeast cells, or vice versa, can be recapitulated in
culture simply by manipulating the growth temper- H. capsulatum is one of the few organisms for which
ature (Maresca and Kobayashi 1989; Maresca et al. functional genomic analysis preceded a genome
1994). When H. capsulatum cells are grown at 25 ◦ C, sequence. Hence, this organism serves as a tes-
they grow in the mycelial form. When these cells timonial for fungal genomics in both sequenced
are shifted to 37 ◦ C, they shift to the budding yeast and unsequenced organisms. Hwang et al. (2003)
form. Though this behaviour is characteristic of all compared the gene expression profiles of yeast and
Histoplasma capsulatum Genomics 223
to a single “spot” on the microarray, and will be

referred to hereafter as an array element. Based
on the estimated genome size at that time, it was
thought that the array covered approximately 1|3 of
the genome. Hwang et al. (2003) isolated RNA from
yeast and mycelial cells, generated differentially la-
beled cDNA, and subjected these labeled probes to
a competitive hybridization on the microarray to
identify a set of yeast-specific and mycelial-specific
genes (discussed further below).
Hwang et al. (2003) chose to generate a mi-
croarray from a genomic library, rather than
a cDNA library, for several reasons. Most impor-
tantly, a random genomic microarray provides
better opportunity for equivalent representation
of genes. Because cDNAs are generated from cells
grown under a discrete number of conditions,
genes that are not transcribed under these condi-
tions will not be represented in the corresponding
cDNA library. Additionally, unless one makes
a normalized cDNA library, highly expressed
genes will be over-represented in the cDNA library
and, hence, on the microarray. In a random
genomic library, most genes have an equivalent
chance of being represented, with the exception of
the minority of genes that fall in genomic repeat
Fig. 12.2. Construction of a shotgun genomic microarray. H. sequences, which will be over-represented.
capsulatum genomic DNA is shown at the top of the figure. The random genomic library approach (or
This DNA was subjected to a partial restriction enzyme di-
gestion, size selected, cloned into a vector, and transformed “shotgun approach”) is unlikely to work if (1) an
into E. coli. Approximately 10,000 independent colonies organism has very large intron sequences that
were inoculated into 96-well plates, and colony PCR was would interfere with the ability of a cDNA probe to
performed on each culture using common primers in the bind to its cognate genomic fragment in an array
vector. The resultant PCR products were spotted on glass
slides to generate shotgun genomic microarrays. Reprinted
experiment, and/or (2) the genome contains large
with permission from Hwang et al. (2003) intergenic regions that will clutter the microarray
with non-coding sequence. Because the average
intron size in H. capsulatum is small (in the order
of 100 nucleotides), the shotgun approach was
mycelial cells by building a shotgun genomic mi- likely to work. To validate that the microarray
croarray (Fig. 12.2) at a time when only a handful of was mainly comprised of coding sequence rather
genes had been cloned from H. capsulatum. To con- than intergenic sequence, Hwang et al. (2003)
struct the array, 1–2 kb genomic DNA fragments carried out a competitive hybridization on a trial
were generated by partial restriction enzyme di- random genomic microarray with genomic DNA,
gestion of genomic DNA from the virulent G217B and with a cDNA probe generated from RNA from
strain. This pool of fragments was ligated into the H. capsulatum cells growing in the yeast form.
E. coli vector pBLUESCRIPT to generate a library Genomic DNA will give a signal for all elements
of genomic fragments. The genomic inserts from on the random genomic microarray, whereas
9600 individual library clones were subjected to cDNA will only give a signal for elements that
amplification by polymerase chain reaction (PCR) correspond to coding sequence. They observed
using a common pair of primers that flanked the that 75% of the spots lit up with the cDNA probe,
site of genomic DNA insertion in the vector. Each indicating that a minimum of 75% of the array
PCR product was precipitated and then printed on elements contained coding sequence. Because the
glass slides to generate the first H. capsulatum mi- cDNAs represent genes expressed under a single
croarray. Thus, each PCR product corresponded growth condition, it is likely that even more
224 A. Sil
than 75% of the array elements contained coding

sequence.
B. Studying Differences Between Yeast

and Mycelia
Hwang et al. (2003) used their shotgun genomic

microarray to compare the transcriptional profile
of yeast and mycelial cells. Since morphology is
regulated by temperature, they reasoned that tem-
perature triggers a signal transduction pathway re-
sulting in either a yeast transcriptional program
at 37 ◦ C, or a mycelial transcriptional program at
25 ◦ C. They were interested in identifying two main
classes of genes: (1) master transcriptional reg-
ulators that govern morphology-related gene ex-
pression, and (2) yeast and mycelial-specific genes
that might shed light on phenotypic differences be-
tween the two morphologic phases.
Previous work from a variety of laboratories
had identified a number of phase-specific genes
(Harris et al. 1989a, b; Keath et al. 1989; Keath
and Abidi 1994; Di Lallo et al. 1994; Gargano et al. Fig. 12.3. A, B Identification of yeast- or mycelial-specific
1995; Patel et al. 1998; Tian and Shearer 2001, 2002; genes. A The histogram depicts the number of spots on the
Johnson et al. 2002), but no large-scale analysis of microarray vs. the log2 of the ratio of the mycelial signal
to the yeast signal. Spots that are equivalently expressed
differential gene expression had been attempted. have a log2 ratio of zero. The numbers of clones that show
Gene expression profiling is a powerful way to iden- differential expression from five- to 100-fold in one mor-
tify both regulatory and target genes, and thus was phologic form compared to the other are shown in the fig-
a good choice to probe differences between yeast ure. B A Northern blot analysis of gene expression in yeast
and mycelia. Each blot has yeast total RNA (Y) in the left
and mycelial cells. lane and mycelial total RNA (M) in the right lane. ACT1
To perform this experiment, Hwang et al. (actin) is equivalently expressed between the two samples,
(2003) grew cells at 25 and 37 ◦ C, and isolated CBP1 (calcium-binding protein), 63G8 (unknown microar-
RNA from the resultant mycelial and yeast-form ray clone), ABC1, and ASY1 are yeast-specific. 94B7 (un-
cells. Differentially labeled cDNAs were generated known microarray clone) and TYR1 are mycelial-specific.
Reprinted with permission from Hwang et al. (2003)
and subjected to a competitive hybridization on
the shotgun genomic microarray. Approximately
500 clones from the microarray were expressed at
3. elements that overlapped with putative genes
least fivefold higher in one morphology compared
were annotated. Approximately half of the dif-
to the other, and a small subset of these clones
ferentially expressed genes were annotated by
were validated as being differentially expressed by
this method.
Northern blot analysis (Fig. 12.3). These clones
were annotated as follows:
A variety of genes were identified as specifically
1. sequence information from each array clone expressed in one morphology or the other, and
was used to map its location on contigs from the identity of many of these genes generated in-
an ongoing genome-sequencing project (de- teresting hypotheses about phenotypic differences
scribed below); between yeast and mycelia (Fig. 12.4). For exam-
2. each contig was subjected to Basic Local Align- ple, only the mycelial form of H. capsulatum is able
ment Search Tool (BLAST) analysis against the to produce vegetative spores, or conidia. Several
non-redundant (nr) database of genes at the orthologs of genes that regulate conidiation in As-
National Center for Bioinformatics to identify pergillus species were identified as mycelial-specific
the location of putative genes based on ho- in H. capsulatum. Hence, these orthologs are can-
mologs in other organisms; and didates for regulators of conidiation in H. capsu-
It should be noted, of course, that every ge-

nomics experiment has its caveats. Because tem-
perature is a key regulator of morphology in H. cap-
sulatum, the yeast phase sample used for microar-
ray comparison was grown at 37 ◦ C, whereas the
mycelial phase sample was grown at 25 ◦ C. Some
genes identified as differentially expressed may be
temperature-regulated, rather than morphology-
regulated. Future studies with cells that are trapped
in one morphologic phase independent of tempera-
ture would facilitate further categorization of these
Fig. 12.4. Putative functions for differentially expressed differentially expressed genes.
genes identified by functional genomics. The figure shows
H. capsulatum cells growing in either the yeast form at
37 ◦ C or the mycelial form at 25 ◦ C. A number of yeast-
specific genes were annotated as potentially being involved
in sulfur metabolism and growth rate/host survival based III. Current Genome-Sequencing
on the function of their ortholog in a different organism. Projects for H. capsulatum
Similarly, mycelial-specific genes included those implicated
in polarized cell growth, melanin production, soil survival,
and conidiation. Adapted with permission from Hwang A. Choice of Genomes to be Sequenced
et al. (2003). Cell images, Anita Sil
Although the shotgun approach described above
was useful for identifying interesting genes, the
genome project for H. capsulatum is making a ge-
latum, which can be tested further by molecular nomics approach much more straightforward. At
genetic experiments. Since no molecular informa- the time of writing, two separate projects to se-
tion was previously available about conidiation in quence three different H. capsulatum strains are
H. capsulatum, these genes provide the first entrée ongoing. To appreciate the choice of strains for se-
into understanding how H. capsulatum might reg- quencing, it is first necessary to understand the
ulate spore production (see Sect. IV.B below). differences between these strains.
Other mycelial-specific genes included or- Molecular studies of H. capsulatum biology and
thologs of genes involved in polarized cell growth pathogenesis have largely taken place in two dis-
and melanin production, which under normal tinct strains: G217B, a North American clinical iso-
media conditions is confined to mycelia. Other late, and G186A, a clinical isolate from Panama.
yeast-specific genes included orthologs of genes These strains were originally differentiated on the
involved in cell cycle regulation and nutrient basis of polysaccharide composition of their cell
metabolism. Since the yeast form of the organism walls (Reiss 1977; Davis et al. 1977; Reiss et al.
is the pathogenic form in the host, all of the 1977). G217B, designated “chemotype 1”, lacks α-
yeast-specific genes are candidates for genes that (1,3)-glucan in its cell wall, whereas G186A, des-
could be important for host survival. Although ignated “chemotype II”, has copious amounts of
no conclusions can be made about the function α-(1,3)-glucan in its cell wall, much like the other
of any gene based simply on sequence homology systemic dimorphic fungal pathogens Blastomyces
and gene expression data, the rapid identification dermatitidis and Paracoccidioides brasiliensis. In-
of morphology-specific genes provided a num- terestingly, the G217B strain (which lacks α-(1,3)-
ber of valuable hypotheses that can be tested glucan) is virulent, but variants of G186A that lack
in future experiments. In a system where few α-(1,3)-glucan are avirulent (Klimpel and Goldman
genes had been identified prior to this work, 1987, 1988).
these data highlight the value of genomics for A second North American isolate, the Downs
opening up a molecular analysis of biological strain, has also been the subject of many research
problems. As discussed below, the current con- studies. The Downs strain, which was originally
struction of a whole-genome microarray for a clinical isolate from an octogenarian who was
H. capsulatum will continue to enhance the probably immunocompromised, is less thermotol-
value of genomics in dissecting H. capsulatum erant than G217B or G186A (Medoff et al. 1986a;
biology. Keath et al. 1989). Furthermore, unlike G217B
226 A. Sil
and G186A, the Downs strain is not virulent Brazil, Argentina, Colombia, and Panama) mainly
in standard animal models (Lambowitz et al. fell into two clades (LAm A and LAm B), but also
1983). The literature indicates that the genomic included seven lone lineages. Since G186A appears
structure of the three strains may be different. to be only distantly related to G217B and Downs,
DNA renaturation kinetics suggested that the and yet both G186A and G217B are virulent strains,
genome of the Downs strain is considerably larger it is likely that determining similarities and differ-
than the genome of G186A (Carr and Shearer ences between the genomes will be quite informa-
1998). Additionally, though it was difficult to tive.
visualize chromosomes with high resolution, The Fungal Genome Initiative at the Broad
contour-clamped homogenous electric field gels Institute (http://www.broad.mit.edu/annotation/
and field-inversion gel electrophoresis suggested fungi/fgi/) is sequencing WU24, a virulent
that the Downs strain and G217B may have NAm 1 strain that is likely to be much more
chromosomes of different size and number (Steele closely related to the Downs strain than G217B
et al. 1989). Restriction fragment length polymor- and G186A. Although no information about
phism analysis of genomic and mitochondrial the sequence was available at the time of this
DNA also confirms that Downs and G217B are writing, genome information is predicted to be
distinct strains (Vincent et al. 1986; Spitzer et al. publicly available by early 2005. The Genome
1989). Ideally, because G217B, G186A, and the Sequencing Center (GSC) at Washington Uni-
Downs strain have markedly different phenotypic versity in St. Louis was funded by the National
characteristics, a comparison of the three strains Institute for Allergy and Infectious Disease to
at the sequence level will probably generate some sequence both G217B and G186A (http://www.
testable hypotheses about how genetic variation genome.wustl.edu/projects/hcapsulatum/), using
results in different biological traits. a combination of whole-genome shotgun sequenc-
The most recent and extensive phylogenetic ing, fosmid end sequencing (and physical mapping;
analyses of 137 different H. capsulatum isolates Magrini et al. 2004), and cDNA sequencing (the
confirm that G217B, G186A, and Downs are latter to contribute to gene-finding and annotation
members of different clades (distinct phylogenetic of the genome). Interestingly, the genomes of
groups of organisms that include the most recent these two strains appear to be quite different, as
common ancestor of all of its members, and all predicted from the phylogenetic analysis described
of the common ancestor’s descendants; Kasuga above. The genome size of G217B appears to be
et al. 1999, 2003). In the most recent analysis, about 41 Megabases (Mb), whereas the genome size
the authors used sequence comparison of four of G186A is approximately 30 Mb. A comparison
unlinked loci to identify 80 multilocus genotypes. of the two genomes is currently in progress. It
Seventy-three of the 80 fell into eight clades: North appears that they have a strikingly different repeat
American class 1 (NAm 1), North American class content: 34% of the G217B genome is composed
2 (NAm 2), Latin American group A (LAm A), of repeat sequence (including transposon se-
Latin American group B (LAm B), Australian, quences), whereas only 7.5% of the G186A genome
Dutch, Eurasian, and African. All clades except the is made up of repeats. The exact nature of the
Eurasian clade were genetically isolated, and thus repeat sequence is currently being analyzed. Final
can be defined as different phylogenetic species. annotation of both genomes is now underway. It
The Eurasian clade, in contrast, was derived from will be of great interest to determine whether there
Latin American group A. The seven isolates that are genes that are unique to each of these strains.
did not belong to any of the clades were also Although the sequencing of G217B, G186A,
distinct from each other, and were designated as and WU24 will make for interesting and infor-
lone lineages. mative comparisons, there are other isolates of H.
G217B and Downs are members of the NAm 2 capsulatum that will be fascinating subjects for
and NAm 1 clades, respectively. G186A, which was future sequence analysis. For example, several of
originally isolated from Panama as stated above, the clades described above included Histoplasma
did not fall into one of the eight clades, and instead isolates that cause disease in horses (H. capsulatum
is a member of a lone lineage. Interestingly, the iso- var. farciminosum), but there is little understand-
lates of H. capsulatum from Latin America, including of what allows these isolates, but not others,
ing G186A, showed the most phylogenetic diversity. to cause equine disease. Additionally, the African
These isolates (from Mexico, Guatemala, Surinam, clade includes isolates of H. capsulatum var.
dubosii, which cause a disease known as African specific transcription factors, and study their
histoplasmosis. These infections are typified by function using molecular genetic approaches.
lesions in the skin and bone, rather than the Additionally, by determining the gene expression
pulmonary infection caused by H. capsulatum profile as cells transit from the yeast form to the
var. capsulatum (Rippon 1988). Sequence analysis mycelial form, and vice versa, it will be possible
and functional studies of these isolates would to use cluster analysis (Eisen et al. 1998) to
shed some light on Histoplasma characteristics identify classes of co-regulated genes (by virtue
that influence the clinical manifestations of of having similar transcriptional profiles during
infection. the morphologic transition). Some of these classes
of genes will be targets of common transcription
factors. Furthermore, it may be possible to use
B. From Genome Sequence known groups of co-regulated genes in other fungi
to Functional Genomics to identify regulatory circuits in H. capsulatum
(Gasch et al. 2004). Finally, by comparing the
A whole-genome oligonucleotide array will be built morphology-specific gene expression profiles in
for the G217B and G186AR strains as part of the G217B and G186, it will be possible to determine
GSC genome-sequencing project, thereby easing similarities and differences in how these strains
the transition from genome sequence to functional regulate their morphology. Ultimately, this type of
genomics. This oligonucleotide array will represent comparative analysis will be extended to include
each putative gene in these two strains as a unique WU24.
70-mer. As part of the genome project, this mi-
croarray will be used to determine the global gene
expression profile of H. capsulatum cells grown IV. Other Applications of Functional
in the yeast form and the mycelial form, as de- Genomics in H. capsulatum
scribed above (Sect. II.B). Because this array will be
a whole-genome array, rather than the previously
used shotgun genomic array, this analysis will sig- Whole-genome microarrays, and other related
nificantly extend the identification of genes that are technologies described below, will open up
differentially expressed between the two forms (or a number of areas of molecular exploration in H.
regulated by temperature). One goal of this anal- capsulatum.
ysis is to contribute to genome annotation. It will
be possible to confirm that putative open reading A. Macrophages
frames truly represent genes because they encode
an expressed transcript. In addition, because the Macrophages are the main host cell for H. capsula-
mycelial form is specialized for growth in the soil tum, and functional genomics is an excellent tool to
and the yeast form is specialized for growth in the probe the interaction between H. capsulatum and
host, a truly global understanding of the gene ex- the macrophage. To understand how functional ge-
pression profile of each of these forms will shed nomics can be used to study the relationship be-
light on H. capsulatum biology and gene function. tween microbe and host cell, it is necessary to first
These experiments will determine whether describe the different steps of interaction between
each gene is more highly expressed in yeast H. capsulatum and the macrophage.
or mycelia, but this analysis is only the start- H. capsulatum is phagocytosed by macro-
ing point. The ultimate goal is to determine phages; once inside the macrophage, the microbe
the regulatory networks that are required to continues to divide (Fig. 12.5). Phagocytosis
establish and maintain the yeast and mycelial of H. capsulatum by human monocyte-derived
forms. It may be possible to use pattern-finding macrophages is mediated by the CD18 family
algorithms such as MEME to identify putative of adhesion-promoting glycoproteins, including
regulatory elements that subsets of yeast- CD11b/CD18 (otherwise known as CR3; Newman
specific or mycelial-specific genes have in common et al. 1990; Newman 1999). Recent work suggests
(http://meme.sdsc.edu/meme/website/intro.html). that HSP60 on the cell surface of H. capsulatum
By combining sequence homology with expression mediates recognition by these receptors (Long
profile, it will be possible to pick out putative et al. 2003). Once phagocytosis has occurred, the
yeast-specific transcription factors, or mycelial- intracellular fate of H. capsulatum has been studied
228 A. Sil
H. capsulatum resides in a phagosome that does not

fuse with the lysosome, and displays reduced levels
of the host vacuolar membrane ATPase/proton
pump that is responsible for acidification of the
phagosome during maturation (Newman et al.
1997; Strasser et al. 1999). Thus, H. capsulatum
interferes with normal phagosome maturation in
these cells. In other types of macrophage host cells
such as the murine P388D1 cell line, H. capsulatum
resides in the phagolysosome (Eissenberg et al.
1988). In either case, H. capsulatum prevents
acidification of either the phagosome or the
phagolysosome, instead maintaining the pH at
approximately 6.0–6.5 (Eissenberg et al. 1993;
Newman 1999; Strasser et al. 1999). Inhibition
of acidification is thought to interfere with the
function of lysosomal hydrolases. Furthermore,
Fig. 12.5. Infection of RAW264.7 cells with H. capsulatum. H. capsulatum is thought to maintain the pH at
Monolayers of macrophages were infected with H. capsula- 6–6.5 to maximize acquisition of iron from host
tum. At different time points, the monolayer was fixed and transferrin (Newman 1999). This hypothesis is
fungal cells were stained with periodic acid/Schiff base. The supported by the fact that H. capsulatum cannot
time point is indicated in the lower left of each quadrant.
At later time points, the Histoplasma fills the macrophage.
survive in macrophages treated with chloroquine,
Cell images, Anita Sil, and courtesy of Dervla Isaac a weak base that raises the endocytic pH such
that iron cannot be released by transferrin. This
detrimental effect of chloroquine on intracellular
growth of H. capsulatum is reversed by addition
in a variety of macrophages, including human of iron compounds that are soluble at neutral or
monocyte-derived macrophages, human alveolar alkaline pH, but not by iron donors that require
macrophages, murine peritoneal or alveolar acidic pH to release iron (Newman et al. 1994).
macrophages, and murine macrophage-like cell Thus, H. capsulatum is thought to survive and
lines such as RAW264.7 and P388D1 (Eissenberg grow in macrophages by virtue of its ability to
and Goldman 1991; Newman 1999; Woods 2003; prevent acidification of the phagosome while
Fig. 12.5). In some but not all of these host cells, maintaining an optimum pH for iron acquisition.
a respiratory burst is triggered after phagocytosis Interestingly, H. capsulatum is known to neutralize
of H. capsulatum, but the viability of H. capsula- the pH of both acidic and basic culture media,
tum seems to be unaffected by reactive oxygen suggesting that the organism can sense and modify
intermediates. the ambient pH of its environment (Berliner 1973).
Once internalized, the intracellular fate of Functional genomics could be used in a num-
H. capsulatum seems to differ depending on the ber of different ways to probe the interaction be-
type of macrophage studied. For the average tween the host cell and H. capsulatum. One obvious
nonpathogenic microbe, the nascent phagosome experiment is to perform a time course of infection
undergoes a series of events known as phago- of macrophages with H. capsulatum. At each time
some maturation, whereby the contents of the point, it is possible to lyse host cells in a guani-
phagosome and phagosomal membrane are dinium thiocyanate buffer, pellet the H. capsula-
altered to promote destruction of the microbe tum cells in the lysate by centrifugation, prepare
(Duclos and Desjardins 2000; Vieira et al. 2002). RNA from these cells, and use microarray analy-
Maturation ultimately results in the formation sis to determine how the gene expression profile
of a phagolysosome, which displays several evolves with infection (M. Paige Nittler and Mar-
degradative properties including acidic pH, the gareta Andersson, personal communication; Anita
presence of acid-activated lysosomal hydrolases, Sil, personal observation). These experiments are
anti-microbial peptides, and toxic oxidative best done coupled with an analysis of the cell bi-
compounds. In the case of infection of human ology of infection, so that it is possible to corre-
macrophages and the murine RAW264.7 cell line, late the timing of gene expression with traffick-
ing through and replication within the host cell.

The goal is to identify putative virulence factors by
identifying genes whose expression changes dur-
ing infection. For example, genes that are induced
when H. capsulatum is present in early phago-
somes might be important for blocking phago-
some maturation. Because the fate of H. capsulatum
seems to be somewhat different in different types
of macrophages (as described above), gene expres-
sion profiling of H. capsulatum in these different
types of host macrophages may provide a sensitive
assay to determine whether the microbe is actually
responding differently in each case.
This experimental setup also allows the purifi-
cation of host cell RNA from the lysate supernatant
after pelleting the H. capsulatum. Transcriptional
profiling of host cells (using mouse or human mi- Fig. 12.6. G186AR macroconidia. The G186AR strain was
croarrays) has been employed to understand how grown under conidiating conditions. The image shows
macrophages are manipulated by pathogens (Det- masses of tuberculate macroconidia. Image, Anita Sil, and
weiler et al. 2001; Boldrick et al. 2002; Nau et al. courtesy of Diane Inglis
2002, 2003; McCaffrey et al. 2004). By monitoring
changes in the expression profiles of both the host
and Histoplasma, it will be possible to monitor the 6 µm, whereas macroconidia (Fig. 12.6) have been
communication between host and pathogen. reported to range in size from 8 to 14 µm or 10 to
Functional genomics can also be used to iden- 25 µm, depending on the strain and growth con-
tify potential virulence factors by subjecting H. cap- ditions. It is thought that the microconidia are the
sulatum to environmental conditions that mimic appropriate size to lodge in the alveoli of the lungs.
those experienced during infection. This type of Therefore, microconidia are thought to be the most
strategy has been used successfully to identify vir- common infectious Histoplasma particle. Interest-
ulence factors in Salmonella, by performing gene ingly, even a single viable spore is thought to be
expression profiling after exposing the microbe capable of causing disease in the mouse (Ajello and
to conditions that mimic the environment of the Runyon 1953), and infection of mice with spores
macrophage phagosome (Detweiler et al. 2003). In results in high mortality (Procknow et al. 1960).
the case of H. capsulatum, it could be informa- A number of studies published prior to 1973 dis-
tive to determine the response to exogenous anti- cuss conditions that promote conidiation, as well
microbial stimuli that mimic the anti-microbial ac- as the morphology of the resultant spores (Pine
tivities of the macrophage, such as reactive oxygen 1960; Artis and Baum 1963; Goos 1964; Smith 1964;
and nitrogen species (Nittler et al. 2005). Addition- Neilsen and Evans 1964; Smith and Furcolow 1964;
ally, stimuli such as low iron levels or low pH could Anderson and Marcus 1968; Garrison and Lane
trigger gene expression programs that are impor- 1973). Additionally, there have been cell biological
tant for survival in host cells. These experiments studies monitoring the germination of micro- and
may identify candidate genes whose roles in infec- macroconidia (Howard 1959; Procknow et al. 1960;
tion must be explored by further experimentation. Goos 1964; Garrison and Boyd 1977, 1978). How-
ever, nothing is known about molecules that are
required for conidial development or germination.
B. Understanding the Development Gene expression profiling would be an ideal
and Germination of Conidia method to identify genes that influence spore for-
mation and germination in H. capsulatum. Coni-
The mycelial form of Histoplasma undergoes diation has been studied extensively in the fungus
asexual sporulation to give rise to at least two Aspergillus nidulans, and several steps are known
types of conidia, macro- and microconidia, which to be regulated at the level of transcription (Tim-
are distinguished mainly on the basis of size berlake 1991). By determining the gene expression
(Pine 1960). Microconidia range in size from 2 to profile of conidiating mycelia, it will be possible to
230 A. Sil
identify genes required for conidial development. mine whether a particular chromosome from one
Indeed, the comparison of yeast and conidiating isolate contains the same complement of genes as
mycelia using functional genomics as described the analogous chromosome from a different isolate.
above (Sect. II.B) revealed that several orthologs These experiments are of particular interest, since
of Aspergillus conidiation genes are specifically ex- the precedent from bacterial pathogens suggests
pressed in H. capsulatum mycelia, as compared to that genome rearrangement is found more fre-
yeast (Hwang et al. 2003). Extending this transcrip- quently in clinical strains (Hughes 2000). Compari-
tional analysis to the whole H. capsulatum genome son of different isolates at the RNA level will also be
will be very valuable. Gene expression profiling of of interest. Gene expression profiling of individual
pure populations of micro- and macroconidia (as isolates grown under similar conditions (e.g., in the
compared to yeast and mycelia) will allow the iden- yeast form, the mycelial form, within macrophages,
tification of conidial-specific transcripts. These etc.) will reveal whether a given isolate expresses
data will shed light on the molecular properties of a similar set of genes as the reference strain on
conidia. Of particular interest is a greater molecular which the array is based. It would be interesting to
understanding of the pathogenicity of microconi- extend this analysis to fresh soil isolates of H. cap-
dia, which could be revealed by studying genes that sulatum to see if these strains have a different ex-
are specifically expressed in these cells. Other func- pression pattern than that of fresh clinical isolates.
tional genomics experiments should include de- These types of experiments may contribute
termination of gene expression of conidia that are to an understanding of resistance to anti-fungal
germinating at 25 ◦ C, at 37 ◦ C, and within host cells. treatments. Previous studies have documented the
evolution of drug resistance in a clinical setting
(Wheat et al. 2001). Acquired Immunodeficiency
C. Clinical Isolates Syndrome (AIDS) patients were treated with
fluconazole for histoplasmosis in sequential
A wide variety of H. capsulatum strains have been clinical trials, and the authors were able to monitor
isolated as clinical samples from all over the world, the emergence of fluconazole-resistant strains.
as evidenced in the phylogenetic study described By comparing the gene expression profile of the
above (Kasuga et al. 2003). One goal of functional resistant strains to that of the parent strain, it may
genomics will be to analyze molecular differences be possible to understand some of the molecular
between these strains, and correlate these data to changes that contribute to drug resistance.
phenotypic differences. Individual H. capsulatum Finally, this type of approach could assist in
isolates will exhibit sequence differences, and so it a molecular dissection of H. capsulatum virulence.
will be optimal to employ a microarray that will tol- For example, a comparison of the gene expres-
erate mismatches between the sequence displayed sion profile of isogenic avirulent and virulent
on the array and the sequence of the given isolate. strains could shed light on changes in transcript
Thus, these studies may be better accomplished accumulation that influence virulence. An ideal
with a microarray that represents the entire ORF opportunity to perform this type of experiment
sequence, rather than a unique 70-mer oligonu- arises because extended laboratory passaging of
cleotide. It will be of interest to compare strains H. capsulatum strain G217B results in reduction
at both the DNA and the RNA level. In the for- of virulence in the mouse model of histoplasmo-
mer case, for example, a comparison of genomic sis, but passaging through an animal results in
DNA from different clinical isolates might be used restoration of virulence (George Deepe, personal
to determine whether different isolates have the communication). One could use microarray
same or different complement of genes as the ref- technology to compare the gene expression profile
erence strain used to design the microarray. This of yeast cells grown in culture from an avirulent
type of analysis can be particularly valuable when and virulent pair to determine if the cells have
combined with electrophoretic analysis of chro- significant differences in gene expression outside
mosomes to study whether particular chromoso- of the host. Similarly, since it is feasible to monitor
mal rearrangements are manifested in different iso- the gene expression profile of microbes in a host
lates. For example, pulsed-field gel electrophoresis animal (Talaat et al. 2004), it should be possible
can be used to separate individual chromosomes. to compare the gene expression profile of the two
The DNA from these chromosomes can then be la- strains once they are in the animal, to analyze
beled and subjected to microarray analysis to deter- differences in how each strain responds to the host.
D. Study of Mating Type V. Technologies that Interface

with Genomics
H. capsulatum has two mating types: – and +.
Though the two mating types are thought to be
The point of functional genomics is to implicate
equivalently represented in the soil, the – mat-
genes in a particular process. Analogous to a large-
ing type is more prevalent in immunocompetent
scale genetic screen, one ends up with candidates
hosts (Kwon-Chung et al. 1974). Both mating types
that might directly or indirectly be involved in a bi-
are equally represented in immunocompromised
ological function. It is extremely useful to use func-
hosts, however, suggesting that each has equiva-
tional genomics to identify candidate genes, but
lent access to the host (Kwon-Chung et al. 1984).
of course it is essential to examine the function
Taken as a whole, these data suggest that the – mat-
of these candidates by other means. A number of
ing type is more virulent than the + mating type in
technical advances now make it possible to make
immunocompetent hosts. However, next to noth-
the most of functional genomics data in H. capsu-
ing is known about how these mating types differ,
latum.
and the mating-type locus itself has not been iden-
tified. Additionally, since strains seem to lose the
ability to mate through serial passaging in the lab- A. Gene Disruption
oratory, many of the laboratory strains (such as
G217B and G186A) are of unknown mating type. To evaluate the role of candidate genes, it is de-
However, + and – mating-type strains are avail- sirable to disrupt the gene and examine the re-
able at the American Type Culture Collection; these sultant phenotype. Only a small number of genes
strains are a wonderful starting point for genomic have been disrupted in H. capsulatum. The first
analysis of mating type. gene disrupted by homologous recombination was
By comparing the gene expression profile of + the URA5 gene (Woods et al. 1998). A URA5 dis-
and – strains by microarray, it will now be possible ruption construct was generated by replacing an
to judge which genes are differentially expressed internal portion of the URA5 gene with the hy-
between the two mating types. (Again, as described gromycin resistance gene. H. capsulatum cells were
in Sect. IV.C, these studies are better accomplished transformed with a linear fragment of DNA con-
with a microarray that represents the entire ORF taining the URA5 disruption, and rare homologous
sequence, so that sequence mismatches between recombination at the genomic URA5 locus was ob-
the mating strains and the reference strain will be served. More recently, the laboratory of William
better tolerated.) Mating-type-specific genes are Goldman improved this disruption technology by
possible candidates for genes in the mating-type placing the gene disruption construct on a telom-
locus, or for targets of the mating-type locus. eric vector (Sebghati et al. 2000). Because linear
One might expect to identify genes such as plasmids with telomere ends can be maintained ex-
pheromones, pheromone receptors, transcription trachromosomally in H. capsulatum, this strategy
factors, and signaling components. If putative allowed cells to be grown for multiple generations
pheromone-encoding genes are identified, it could in the presence of the disruption construct. Essen-
by very informative to synthesize pheromone tially, the telomeres reduced ectopic integration of
in vitro, apply it to the opposite mating type, the disruption construct at non- homologous loci,
and determine the gene expression profile. Such and thereby increased the relative frequency of the
experiments would make it possible to identify homologous targeting event. This technology was
pheromone response genes (Bennett et al. 2003). used to show that the H. capsulatum CBP1 gene
Additionally, since the – mating type seems to is required for virulence (Sebghati et al. 2000). De-
cause a higher frequency of clinical cases than the spite this improvement in the gene disruption tech-
+ mating type, any differentially expressed genes nology, however, it remains fairly arduous to dis-
are candidates that might influence virulence in rupt genes in H. capsulatum, and only a handful of
the host. It may be informative to examine the gene disruption strains have been generated.
gene expression profile of each mating type in
macrophages to see if – strains induce a different B. RNA Interference
set of genes than + strains. This information could
initiate an understanding of how mating type RNA interference (RNAi) technology is a relatively
might influence virulence. fast and powerful alternative to gene disruption in
232 A. Sil
other organisms (Meister and Tuschl 2004). The ex- Agrobacterium-mediated insertions in H. capsula-
pression of a double-stranded RNA causes specific tum, Sullivan et al. (2002) used a specialized system
gene silencing in a number of systems. The ability where the functions of the Ti plasmid have been
to “knock down” gene function with RNAi in H. split into two binary vectors (allowing easy manip-
capsulatum would provide a relatively quick way to ulation of these plasmids). One plasmid is a partial
screen for phenotypes for candidate genes identi- Ti plasmid that retains only the virulence genes
fied in microarray experiments. The laboratory of necessary for transfer of T-DNA and integration of
William Goldman recently developed RNAi tech- this DNA into the host genome, and the other plas-
nology for H. capsulatum (Rappleye et al. 2004). In mid contains the T-DNA and a selectable marker
these experiments, RNA hairpins were expressed (such as the hygromycin resistance marker) in the
from a strong promoter in wild-type cells. These region of DNA that is transferred to the host. The
hairpins consisted of a large piece of coding se- latter allows H. capsulatum recipients of the T-DNA
quence separated from its reverse complement by to be selected as hygromycin-resistant strains. Sul-
a spacer region, such that a stem-loop structure livan et al. (2002) successfully used either T-DNA
is formed by the expressed RNA. These constructs carrying this hygromycin resistance marker or T-
were introduced into H. capsulatum on telomeric DNA carrying a URA5 marker (for transformation
vectors, which allowed them to be maintained ex- of ura5-mutant strains).
trachromosomally. The authors were able to reduce The exact nature of the T-DNA boundaries that
expression of both exogenous (green fluorescent are transferred to the host genome can vary from
protein) and endogenous genes in a specific and transformant to transformant, although the ma-
stable fashion. In other systems, it is clear that ex- jority of insertions span the region between the
pression of short interfering RNAs (as compared to left border and right border of the T-DNA (Zhu
the longer sequences used by Rappleye et al. 2004) et al. 2000; Christie 2001; Sullivan et al. 2002). On
is quite effective at triggering silencing (McManus occasion, the whole T-DNA plasmid can be trans-
et al. 2002). These constructs are easier to gener- ferred and integrated, or concatemers of T-DNA
ate than are the long hairpin constructs. If they are can be inserted in the host genome. Sullivan et al.
sufficient to trigger RNAi in H. capsulatum, they (2002) examined 15 H. capsulatum insertions by
will make going from gene sequence to phenotype Southern blot. They determined that 12 transfor-
even more expeditious. mants showed single sites of integration, six had
integrated concatemerized T-DNA, and seven had
integrated the entire T-DNA plasmid. The sizes of
C. Insertional Mutagenesis bands on the Southern blot were consistent with
different sites of integration in the genome, im-
Until recently, it has been difficult to generate in- plying that integration of the T-DNA occurs at
sertion mutants in H. capsulatum, since the intro- random sites in the genome. These results have
duction of DNA on non-telomeric vectors often since been expanded to an analysis of over 100
results in complex integration events and multiple independent insertion strains (Van Nguyen, per-
integrations per genome (Worsham and Goldman sonal communication; Anita Sil, personal observa-
1990). However, as described below, it has now been tion).
possible to generate simpler insertions in H. cap- This insertion technology could interface
sulatum. I will first describe how this is done, and with functional genomics in a couple of different
then discuss the potential interface between this ways. First, the introduction of gene disruption
technology and functional genomics. constructs via A. tumefaciens instead of telomeric
Sullivan et al. (2002) used Agrobacterium tume- vectors may increase the frequency of targeted
faciens, a bacterial pathogen of plants, to generate gene knockouts, as has been shown for other
insertions in H. capsulatum and the related sys- fungi (Das 1998; Gouka et al. 1999). As mentioned
temic dimorphic fungus Blastomyces dermatiditis above, facilitating gene disruption will speed up
(Sullivan et al. 2002). A. tumefaciens can transfer the analysis of candidate genes that are identified
DNA from a 200-kb-virulence plasmid (the Ti plas- by expression analyses. Second, it is possible
mid) to a variety of hosts (Zhu et al. 2000; Christie that Agrobacterium-mediated insertion could
2001; Sullivan et al. 2002). The transferred DNA, or mimic transposon site hybridization (TraSH),
T-DNA, integrates into the host genome in a ran- which has been an extremely productive means
dom fashion, most often at a single site. To generate of using genomics to identify avirulent insertion
mutants (Sassetti and Rubin 2002, 2003; Sassetti latum, coupled with the current state of molecular
et al. 2003). To perform TraSH analysis, a library genetic tools in this system, bode well for the fu-
of transposon-generated insertion mutants is ture of functional genomic analyses in this fungal
pooled and subjected to growth under two con- pathogen.
ditions – for example, in vitro culture versus
infection of an animal. The transposon contains Acknowledgements. I would like to thank Joseph DeRisi and
all members of the Sil laboratory for interesting discussions.
an outward-facing RNA polymerase promoter I especially thank Davina Hocking Murray for tireless and
(such as T7 polymerase) that will transcribe invaluable assistance with all the figures.
RNA that is complementary to the chromosomal
DNA adjacent to the transposon insertion. By
making genomic DNA from the pooled strains,
performing an in vitro transcription reaction References
to make probes that represent the transposon
insertion site, generating differentially labeled Ajello L, Runyon LC (1953) Infection of mice with single
spores of Histoplasma capsulatum. J Bacteriol 66:34–
cDNA from each condition, and hybridizing the 40
resultant probes to a microarray, one identifies Anderson KL, Marcus S (1968) Sporulation characteristics
insertion strains that are not present in the pool of Histoplasma capsulatum. Mycopathol Mycol Appl
of interest (for example, mutant strains that 36:179–187
fail to colonize an animal). The sites of these Artis D, Baum GL (1963) Tuberculate spore formation by
thirty-two strains of Histoplasma capsulatum. Myco-
insertions will implicate genes that are important pathol Mycol Appl 21:29–35
for virulence. The hope for H. capsulatum is that Bennett RJ, Uhl MA, Miller MG, Johnson AD (2003) Iden-
one could use the T-DNA in an analogous fashion tification and characterization of a Candida albicans
to the TraSH transposon. To do so, one would mating pheromone. Mol Cell Biol 23:8189–8201
Berliner MD (1973) Histoplasma capsulatum: effects of pH
introduce an RNA polymerase promoter adjacent on the yeast and mycelial phases in vitro. Sabouraudia
to a boundary of the T-DNA. One difficulty 11:267–270
is that a fraction of Agrobacterium-mediated Boldrick JC, Alizadeh AA, Diehn M, Dudoit S, Liu CL,
insertions contain not only the T-DNA, but other Belcher CE, Botstein D, Staudt LM, Brown PO, Rel-
plasmid sequences as well, which means that man DA (2002) Stereotyped and specific gene expres-
sion programs in human innate immune responses to
the RNA polymerase promoter may no longer bacteria. Proc Natl Acad Sci USA 99:972–977
flank H. capsulatum genomic sequence in the Borges-Walmsley MI, Walmsley AR (2000) cAMP signalling
insertion. Thus, these insertion strains would in pathogenic fungi: control of dimorphic switching
not be detectable by generating T7 RNA probes. and pathogenicity. Trends Microbiol 8:133–141
Bradsher RW (1996) Histoplasmosis and blastomycosis.
Nonetheless, because Agrobacterium-mediated Clin Infect Dis 22 Suppl 2:102–111
insertion is an easy method for generating large Brown AE (1990) Overview of fungal infections in cancer
numbers of insertion strains in H. capsulatum, patients. Semin Oncol 17:2–5
a small number of undetectable strains may Bryan CS, DiSalvo AF (1979) Overwhelming opportunistic
not interfere with adapting the technology for histoplasmosis. Sabouraudia 17:209–212
Bullock WE (1993) Interactions between human phago-
a TraSH-like purpose. cytic cells and Histoplasma capsulatum. Arch Med Res
24:219–223
Carr J, Shearer G Jr (1998) Genome size, complexity, and
ploidy of the pathogenic fungus Histoplasma capsula-
VI. Conclusions tum. J Bacteriol 180:6697–6703
Christie PJ (2001) Type IV secretion: intercellular transfer
of macromolecules by systems ancestrally related to
As discussed above, it is a propitious time to study conjugation machines. Mol Microbiol 40:294–305
H. capsulatum biology. Previous experiments have Das A (1998) DNA transfer from Agrobacterium to plant
validated the power of functional genomic studies cells in crown gall tumor disease. Subcell Biochem
29:343–363
in this organism: a comparison of gene expression Davis TE Jr, Domer JE, Li YT (1977) Cell wall studies of
profiles in two major stages of the H. capsulatum Histoplasma capsulatum and Blastomyces dermati-
life cycle yielded a large number of differentially tidis using autologous and heterologous enzymes.
expressed genes of varied putative function. Cur- Infect Immun 15:978–987
Detweiler CS, Cunanan DB, Falkow S (2001) Host microar-
rently, three diverse H. capsulatum isolates are be- ray analysis reveals a role for the Salmonella response
ing sequenced, providing ample fodder for compar- regulator phoP in human macrophage cell death. Proc
ative genomics. Finally, the rich biology of H. capsu- Natl Acad Sci USA 98:5850–5855
234 A. Sil
Detweiler CS, Monack DM, Brodsky IE, Mathew H, Falkow S Huffnagle GB, Deepe GS (2003) Innate and adaptive deter-
(2003) virK, somA and rcsC are important for sys- minants of host susceptibility to medically important
temic Salmonella enterica serovar Typhimurium in- fungi. Curr Opin Microbiol 6:344–350
fection and cationic peptide resistance. Mol Microbiol Hughes D (2000) Evaluating genome dynamics: the con-
48:385–400 straints on rearrangements within bacterial genomes.
Di Lallo G, Gargano S, Maresca B (1994) The Histoplasma Genome Biol 1 REVIEWS0006:1–8
capsulatum cdc2 gene is transcriptionally regulated Hwang L, Hocking-Murray D, Bahrami AK, Andersson M,
during the morphologic transition. Gene 140:51–57 Rine J, Sil A (2003) Identifying phase-specific genes
Duclos S, Desjardins M (2000) Subversion of a young phago- in the fungal pathogen Histoplasma capsulatum using
some: the survival strategies of intracellular pathogens. a genomic shotgun microarray. Mol Biol Cell 14:2314–
Cell Microbiol 2:365–377 2326
Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Clus- Johnson CH, Klotz MG, York JL, Kruft V, McEwen JE (2002)
ter analysis and display of genome-wide expression Redundancy, phylogeny and differential expression
patterns. Proc Natl Acad Sci USA 95:14863–14868 of Histoplasma capsulatum catalases. Microbiology
Eissenberg LG, Goldman WE (1991) Histoplasma variation 148:1129–1142
and adaptive strategies for parasitism: new perspec- Kasuga T, Taylor JW, White TJ (1999) Phylogenetic relation-
tives on histoplasmosis. Clin Microbiol Rev 4:411– ships of varieties and geographical groups of the hu-
421 man pathogenic fungus Histoplasma capsulatum Dar-
Eissenberg LG, Schlesinger PH, Goldman WE (1988) ling. J Clin Microbiol 37:653–663
Phagosome-lysosome fusion in P388D1 macrophages Kasuga T, White TJ, Koenig G, McEwen J, Restrepo A, Cas-
infected with Histoplasma capsulatum. J Leukoc Biol taneda E, Da Silva Lacaz C, Heins-Vaccari EM, De Fre-
43:483–491 itas RS, Zancope-Oliveira RM et al. (2003) Phylogeog-
Eissenberg LG, Goldman WE, Schlesinger PH (1993) raphy of the fungal pathogen Histoplasma capsulatum.
Histoplasma capsulatum modulates the acidification Mol Ecol 12:3383–3401
of phagolysosomes. J Exp Med 177:1605–1611 Kauffman CA, Israel KS, Smith JW, White AC, Schwarz J,
Gargano S, Di Lallo G, Kobayashi GS, Maresca B (1995) Brooks GF (1978) Histoplasmosis in immunosup-
A temperature-sensitive strain of Histoplasma capsu- pressed patients. Am J Med 64:923–932
latum has an altered delta 9-fatty acid desaturase gene. Keath EJ, Abidi FE (1994) Molecular cloning and sequence
Lipids 30:899–906 analysis of yps-3, a yeast-phase-specific gene in the
Garrison RG, Boyd KS (1977) The fine structure of mi- dimorphic fungal pathogen Histoplasma capsulatum.
croconidial germination and vegetative cells of Histo- Microbiology 140(4):759–767
plasma capsulatum. Ann Microbiol (Paris) 128:135– Keath EJ, Painter AA, Kobayashi GS, Medoff G (1989) Vari-
149 able expression of a yeast-phase-specific gene in Histo-
Garrison RG, Boyd KS (1978) Electron microscopy of plasma capsulatum strains differing in thermotoler-
yeastlike cell development from the microconidium of ance and virulence. Infect Immun 57:1384–1390
Histoplasma capsulatum. J Bacteriol 133:345–353 Klimpel KR, Goldman WE (1987) Isolation and charac-
Garrison RG, Lane JW (1973) Scanning-beam electron mi- terization of spontaneous avirulent variants of Histo-
croscopy of the conidia of the brown and albino fila- plasma capsulatum. Infect Immun 55:528–533
mentous varieties of Histoplasma capsulatum. Myco- Klimpel KR, Goldman WE (1988) Cell walls from avirulent
pathol Mycol Appl 49:185–191 variants of Histoplasma capsulatum lack alpha-(1,3)-
Gasch AP, Moses AM, Chiang DY, Fraser HB, Berardini M, glucan. Infect Immun 56:2997–3000
Eisen MB (2004) Conservation and evolution of Kwon-Chung KJ (1973) Studies on Emmonsiella capsulata.
cis-regulatory systems in ascomycete fungi. PLoS Biol I. Heterothallism and development of the ascocarp.
2(e398):1–18 Mycologia 65:109–121
Goldman WE (1995) Molecular genetic technology transfer Kwon-Chung KJ, Weeks RJ, Larsh HW (1974) Studies on
to pathogenic fungi. Arch Med Res 26:437–440 Emmonsiella capsulata (Histoplasma capsulatum). II.
Goos RD (1964) Germination of the macroconidia of Histo- Distribution of the two mating types in 13 endemic
plasma capsulatum. Mycologia 56:662–671 states of the United States. Am J Epidemiol 99:44–49
Gouka RJ, Gerk C, Hooykaas PJ, Bundock P, Musters W, Kwon-Chung KJ, Bartlett MS, Wheat LJ (1984) Distribution
Verrips CT, de Groot MJ (1999) Transformation of of the two mating types among Histoplasma capsula-
Aspergillus awamori by Agrobacterium tumefaciens- tum isolates obtained from an urban histoplasmosis
mediated homologous recombination. Nat Biotechnol outbreak. Sabouraudia 22:155–157
17:598–601 Lambowitz AM, Kobayashi GS, Painter A, Medoff G (1983)
Harris GS, Keath EJ, Medoff J (1989a) Characterization Possible relationship of morphogenesis in pathogenic
of alpha and beta tubulin genes in the dimorphic fungus, Histoplasma capsulatum, to heat shock re-
fungus Histoplasma capsulatum. J Gen Microbiol sponse. Nature 303:806–808
135(7):1817–1832 Long KH, Gomez FJ, Morris RE, Newman SL (2003) Identi-
Harris GS, Keath EJ, Medoff J (1989b) Expression of alpha- fication of heat shock protein 60 as the ligand on Histo-
and beta-tubulin genes during dimorphic-phase plasma capsulatum that mediates binding to CD18 re-
transitions of Histoplasma capsulatum. Mol Cell Biol ceptors on human macrophages. J Immunol 170:487–
9:2042–2049 494
Howard DH (1959) Observations on tissue cultures Magrini V, Goldman WE (2001) Molecular mycology: a ge-
of mouse peritoneal exudates inoculated with netic toolbox for Histoplasma capsulatum. Trends Mi-
Histoplasma capsulatum. J Bacteriol 78:69–78 crobiol 9:541–546
Magrini V, Warren WC, Wallis J, Goldman WE, Xu J, in response to reactive nitrogen species. Mol Biol Cell
Mardis ER, McPherson JD (2004) Fosmid-based (in press)
physical mapping of the Histoplasma capsulatum Patel JB, Batanghari JW, Goldman WE (1998) Probing the
genome. Genome Res 14:1603–1609 yeast phase-specific expression of the CBP1 gene in
Maresca B, Kobayashi GS (1989) Dimorphism in Histo- Histoplasma capsulatum. J Bacteriol 180:1786–1792
plasma capsulatum: a model for the study of cell differ- Pine L (ed) (1960) Morphological and physiological
entiation in pathogenic fungi. Microbiol Rev 53:186– characteristics of Histoplasma capsulatum. Thomas,
209 Springfield, IL Procknow JJ, Page MI, Loosli CG (1960)
Maresca B, Medoff G, Schlessinger D, Kobayashi GS (1977) Early pathogenesis of experimental histoplasmosis.
Regulation of dimorphism in the pathogenic fungus Arch Pathol 69:413–426
Histoplasma capsulatum. Nature 266:447–448 Rappleye CA, Engle JT, Goldman WE (2004) RNA interfer-
Maresca B, Carratu L, Kobayashi GS (1994) Morphological ence in Histoplasma capsulatum demonstrates a role
transition in the human fungal pathogen Histoplasma for alpha-(1,3)-glucan in virulence. Mol Microbiol
capsulatum. Trends Microbiol 2:110–114 53:153–165
Marques SA, Robles AM, Tortorano AM, Tuculet MA, Ne- Reiss E (1977) Serial enzymatic hydrolysis of cell walls
groni R, Mendes RP (2000) Mycoses associated with of two serotypes of yeast-form Histoplasma capsula-
AIDS in the Third World. Med Mycol 38 Suppl 1:269– tum with alpha(1 leads to 3)-glucanase, beta(1 leads
279 to 3)-glucanase, pronase, and chitinase. Infect Immun
McCaffrey RL, Fawcett P, O’Riordan M, Lee KD, Havell EA, 16:181–188
Brown PO, Portnoy DA (2004) A specific gene expres- Reiss E, Miller SE, Kaplan W, Kaufman L (1977) Anti-
sion program triggered by Gram-positive bacteria in genic, chemical, and structural properties of cell walls
the cytosol. Proc Natl Acad Sci USA 101:11386–11391 of Histoplasma capsulatum yeast-form chemotypes 1
McManus MT, Petersen CP, Haines BB, Chen J, Sharp PA and 2 after serial enzymatic hydrolysis. Infect Immun
(2002) Gene silencing using micro-RNA designed hair- 16:690–700
pins. RNA 8:842–850 Rippon JW (1988) Medical mycology. W.B. Saunders,
Medoff G, Maresca B, Lambowitz AM, Kobayashi G, Philadelphia, PA Samonis G, Bafaloukos D (1992)
Painter A, Sacco M, Carratu L (1986a) Correlation Fungal infections in cancer patients: an escalating
between pathogenicity and temperature sensitivity in problem. In Vivo 6:183–193
different strains of Histoplasma capsulatum. J Clin Sassetti C, Rubin EJ (2002) Genomic analyses of microbial
Invest 78:1638–1647 virulence. Curr Opin Microbiol 5:27–32
Medoff G, Sacco M, Maresca B, Schlessinger D, Painter A, Sassetti CM, Rubin EJ (2003) Genetic requirements for my-
Kobayashi GS, Carratu L (1986b) Irreversible block of cobacterial survival during infection. Proc Natl Acad
the mycelial-to-yeast phase transition of Histoplasma Sci USA 100:12989–12994
capsulatum. Science 231:476–479 Sassetti CM, Boyd DH, Rubin EJ (2003) Genes required for
Meister G, Tuschl T (2004) Mechanisms of gene silencing by mycobacterial growth defined by high density muta-
double-stranded RNA. Nature 431:343–349 genesis. Mol Microbiol 48:77–84
Nau GJ, Richmond JF, Schlesinger A, Jennings EG, Lan- Sebghati TS, Engle JT, Goldman WE (2000) Intracellular
der ES, Young RA (2002) Human macrophage activa- parasitism by Histoplasma capsulatum: fungal viru-
tion programs induced by bacterial pathogens. Proc lence and calcium dependence. Science 290:1368–1372
Natl Acad Sci USA 99:1503–1508 Smith CD (1964) Ii. Evidence of the presence in yeast extract
Nau GJ, Schlesinger A, Richmond JF, Young RA (2003) of substances which stimulate the growth of Histo-
Cumulative Toll-like receptor activation in human plasma capsulatum and Blastomyces dermatitidis sim-
macrophages treated with whole bacteria. J Immunol ilarly to that found in starling manure extract. Myco-
170:5203–5209 pathol Mycol Appl 22:99–105
Neilsen GE, Evans RE (1964) A study of the sporulation of Smith CD, Furcolow ML (1964) The demonstration of
Histoplasma capsulatum. J Bacteriol 68:261–264 growth stimulating substances for Histoplasma
Newman SL (1999) Macrophages in host defense against capsulatum and Blastomyces dermatitidis in infusions
Histoplasma capsulatum. Trends Microbiol 7:67–71 of starling (Sturnis vulgaris) manure. Mycopathol
Newman SL, Bucher C, Rhodes J, Bullock WE (1990) Phago- Mycol Appl 22:73–80
cytosis of Histoplasma capsulatum yeasts and micro- Spitzer ED, Lasker BA, Travis SJ, Kobayashi GS, Medoff
conidia by human cultured macrophages and alveolar G (1989) Use of mitochondrial and ribosomal DNA
macrophages. Cellular cytoskeleton requirement for polymorphisms to classify clinical and soil isolates of
attachment and ingestion. J Clin Invest 85:223–230 Histoplasma capsulatum. Infect Immun 57:1409–1412
Newman SL, Gootee L, Brunner G, Deepe GS Jr (1994) Steele PE, Carle GF, Kobayashi GS, Medoff G (1989) Elec-
Chloroquine induces human macrophage killing of trophoretic analysis of Histoplasma capsulatum chro-
Histoplasma capsulatum by limiting the availability of mosomal DNA. Mol Cell Biol 9:983–987
intracellular iron and is therapeutic in a murine model Sternberg S (1994) The emerging fungal threat. Science
of histoplasmosis. J Clin Invest 93:1422–1429 266:1632–1634
Newman SL, Gootee L, Kidd C, Ciraolo GM, Morris R (1997) Strasser JE, Newman SL, Ciraolo GM, Morris RE, How-
Activation of human macrophage fungistatic activity ell ML, Dean GE (1999) Regulation of the macrophage
against Histoplasma capsulatum upon adherence to vacuolar ATPase and phagosome-lysosome fusion by
type 1 collagen matrices. J Immunol 158:1779–1786 Histoplasma capsulatum. J Immunol 162:6148–6154
Nittler MP, Hocking-Murray D, Foo CK, Sil A (2005) Identifi- Sullivan TD, Rooney PJ, Klein BS (2002) Agrobacterium
cation of Histoplasma capsulatum transcripts. Induced tumefaciens integrates transfer DNA into single
236 A. Sil
chromosomal sites of dimorphic fungi and yields Wheat LJ, Kauffman CA (2003) Histoplasmosis. Infect Dis
homokaryotic progeny from multinucleate yeast. Clin N Am 17:1–19
Eukaryot Cell 1:895–905 Wheat LJ, Connolly P, Smedema M, Brizendine E, Hafner R
Talaat AM, Lyons R, Howard ST, Johnston SA (2004) The (2001) Emergence of resistance to fluconazole as
temporal expression profile of Mycobacterium tuber- a cause of failure during treatment of histoplasmosis
culosis infection in mice. Proc Natl Acad Sci USA in patients with acquired immunodeficiency disease
101:4602–4607 syndrome. Clin Infect Dis 33:1910–1913
Tian X, Shearer G Jr (2001) Cloning and analysis of mold- Woods JP (2002) Histoplasma capsulatum molecular ge-
specific genes in the dimorphic fungus Histoplasma netics, pathogenesis, and responsiveness to its envi-
capsulatum. Gene 275:107–114 ronment. Fungal Genet Biol 35:81–97
Tian X, Shearer G Jr (2002) The mold-specific MS8 gene is Woods JP (2003) Knocking on the right door and making
required for normal hypha formation in the dimorphic a comfortable home: Histoplasma capsulatum intra-
pathogenic fungus Histoplasma capsulatum. Eukaryot cellular pathogenesis. Curr Opin Microbiol 6:327–331
Cell 1:249–256 Woods JP, Retallack DM, Heinecke EL, Goldman WE (1998)
Timberlake WE (1991) Temporal and spatial controls of Rare homologous gene targeting in Histoplasma cap-
Aspergillus development. Curr Opin Genet Dev 1:351– sulatum: disruption of the URA5Hc gene by allelic re-
357 placement. J Bacteriol 180:5135–5143
Vieira OV, Botelho RJ, Grinstein S (2002) Phagosome mat- Worsham PL, Goldman WE (1990) Development of a genetic
uration: aging gracefully. Biochem J 366:689–704 transformation system for Histoplasma capsulatum:
Vincent RD, Goewert R, Goldman WE, Kobayashi GS, Lam- complementation of uracil auxotrophy. Mol Gen Genet
bowitz AM, Medoff G (1986) Classification of Histo- 221:358–362
plasma capsulatum isolates by restriction fragment Zhu J, Oger PM, Schrammeijer B, Hooykaas PJ, Farrand SK,
polymorphisms. J Bacteriol 165:813–818 Winans SC (2000) The bases of crown gall tumorigen-
esis. J Bacteriol 182:3885–3895
13 Cryptococcus neoformans Pathogenicity
R.T. Nelson1,2 , J.K. Lodge1
CONTENTS a) Manipulation of mRNA Levels . . . . 255

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 237 b) Large-Scale Gene Deletion . . . . . . . 255
II. Human Disease . . . . . . . . . . . . . . . . . . . . . . . 238 c) Genome-Wide Insertional
A. Patient Population . . . . . . . . . . . . . . . . . . 238 Mutagenesis . . . . . . . . . . . . . . . . . . 256
B. Disease Presentation . . . . . . . . . . . . . . . . 238 B. Proteomics . . . . . . . . . . . . . . . . . . . . . . . . 258
C. Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . 240 VIII. Conclusions and Future Impact . . . . . . . . . . 259
III. Models for Studying Pathogenesis . . . . . . . . 241 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
A. Animal Models . . . . . . . . . . . . . . . . . . . . . 241
B. Alternative Models . . . . . . . . . . . . . . . . . . 242
IV. Biology of Cryptococcus neoformans . . . . . . 242
A. Taxonomy . . . . . . . . . . . . . . . . . . . . . . . . . 242
B. Life Cycle . . . . . . . . . . . . . . . . . . . . . . . . . 243
Abbreviations: BAC, bacterial artificial chro-
C. Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . 243 mosome; CHEF, contour-clamped homogeneous
D. Molecular Genetics . . . . . . . . . . . . . . . . . . 243 electric field; CSF, cerebral spinal fluid; EST, ex-
V. Genome Project . . . . . . . . . . . . . . . . . . . . . . . 244 pressed sequence tag; GalXM, galactoxylomannan;
A. Description of the Genome . . . . . . . . . . . 244 CRAG, cryptococcal capsular polysaccharide anti-
B. Strain Choices . . . . . . . . . . . . . . . . . . . . . 245
C. Physical Mapping . . . . . . . . . . . . . . . . . . . 245 gen; GXM, glucuronoxylomannan; HAART, highly
D. Meiotic Map . . . . . . . . . . . . . . . . . . . . . . . 245 active antiretroviral therapy; HR, homologous
E. Shotgun Sequencing . . . . . . . . . . . . . . . . . 245 recombination; MP, mannoprotein; SAGE, serial
F. EST Sequencing . . . . . . . . . . . . . . . . . . . . 246 analysis of gene expression; STM, signature tagged
G. Annotation . . . . . . . . . . . . . . . . . . . . . . . . 247
VI. Analysis of Genes that Affect Pathogenicity
mutagenesis
of C. neoformans and the Prospects
for Genomic Impact . . . . . . . . . . . . . . . . . . . 247
A. Capsule Biosynthesis . . . . . . . . . . . . . . . . 248
B. Melanin Biosynthesis . . . . . . . . . . . . . . . . 249 I. Introduction
C. Phospholipase . . . . . . . . . . . . . . . . . . . . . 249
D. Urease . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
E. Signal Transduction Pathways . . . . . . . . . 250 Cryptococcosis has become a major problem since
F. Mating Type . . . . . . . . . . . . . . . . . . . . . . . 250 the advent of AIDS (for review, see Casadevall and
G. Taxonomy . . . . . . . . . . . . . . . . . . . . . . . . . 251 Perfect 1998). In the Western world, the impact of
H. Survival in a Phagocytic Environment
and Resistance to Oxidative this systemic fungal disease has been mitigated by
and Nitrosative Stress . . . . . . . . . . . . . . . . 251 antifungal agents and therapies that reduce HIV
VII. Post-Genomic Approaches for Identification and prevent the severe reduction of T-cells that is
of Virulence Genes associated with late stages of AIDS and the onset of
and Their Use in C. neoformans . . . . . . . . . . 252 opportunistic infections. However, in other parts of
A. Functional Genomic Techniques . . . . . . . 252
1. Transcriptome Analysis . . . . . . . . . . . . 252 the world, cryptococcosis is still a major problem
a) Microarrays . . . . . . . . . . . . . . . . . . 252 and it has been estimated that 30% of HIV-infected
b) SAGE . . . . . . . . . . . . . . . . . . . . . . . . 253 individuals in sub-Saharan Africa will succumb to
c) Other Methods of Global Gene cryptococcal meningitis. Recent outbreaks of Cryp-
Expression Profiling . . . . . . . . . . . . 254
2. Gene Expression Modulation . . . . . . . . 254 tococcus neoformans among immunocompetent in-
dividuals highlight the need for vigilance and con-
1 Edward A. Doisy Department of Biochemistry and Molecular tinued investigation. Treatments are not completely
Biology, Saint Louis University School of Medicine, 1402 S. Grand
Blvd., St. Louis, MS 63104, USA
effective or are highly toxic, and so novel drug tar-
2 Current address: G329 Agronomy Hall, Iowa State University, gets are needed. Genomics has accelerated this field
Ames, IA 50011, USA of research, facilitating rapid identification of genes
The Mycota XIII
Fungal Genomics
238 R.T. Nelson and J.K. Lodge
important for virulence, and permitting identifica- Immune suppression due to anti-neoplastic
tion of new fungal-specific genes that would be therapy is also a risk factor for the disease, with
excellent candidates for antifungal therapies. a projected incidence rate of approximately 18 of
every 100,000 cancer patients becoming infected
(Kontoyiannis et al. 2001). In a retrospective study
of HIV-negative patients with malignancy at the
II. Human Disease M.D. Anderson Cancer Center, 65% of patients
with cryptococcosis also had a hematologic
A. Patient Population malignancy, indicating the close association of
the disease with the status of the host’s cellular
Disseminated cryptococcosis is a disease primar- immunity (Kontoyiannis et al. 2001). Long-term
ily associated with individuals whose cellular im- heavy corticosteroid use is also a risk factor for the
munity has been compromised by viral infection, disease (Cunha 2001a).
suppression due to tissue transplantation or anti-
neoplastic chemotherapy. An estimated 6%–10%
of AIDS patients acquire cryptococcosis during the B. Disease Presentation
course of their HIV disease (Eng et al. 1986; Currie
et al. 1994). A more recent figure for the incidence In the HIV-compromised host, patients with
of cryptococcosis in AIDS patients indicates that cryptococcosis generally present with symptoms
the annual incidence rate for HIV-infected persons of meningitis such as fever, headache, and malaise
ranges from 17 to 66 in every 1000 (Hajjeh et al. with or without a stiff neck (reviewed in Casadevall
1999). This same study concluded that the inci- and Perfect 1998). Nausea or altered mentation
dence rate for non-HIV-infected people in the same may or may not be present. Meningioencephalitis
metropolitan areas was between 0.2 and 0.9 per caused by Cryptococcus presents as an indolent
100,000. This estimate, if applied to the US popula- infection with insidious onset, fever, and often
tion as a whole (about 250 million), indicates that a headache. It is the most frequent presenting
as many as 2250 HIV-negative people will acquire syndrome of AIDS patients, representing approx-
the disease each year. imately 60%–85% of all cryptococcosis cases.
Fungal infection is a significant threat to pa- Neurologic symptoms are present in approxi-
tients undergoing tissue transplantation, with as mately 50% of patients, with cranial nerve palsies
many as 59% acquiring a fungal disease at some being the most frequent manifestation (Moosa and
centers. However, infection with Cryptococcus is Coovadia 1997; Cunha 2001a). However, computed
a relatively rare event, with a mean incidence of tomographic analyses of these patients are normal
about 2.8% (Husain et al. 2001). Susceptibility to in approximately 60% of cases (Mitchell et al.
cryptococcosis does not appear to be influenced by 1995). The most common abnormalities visualized
the tissue being transplanted, with 2% of thoracic by computed tomographic analysis are single mass
organ transplant patients (Grossi et al. 2000), 3% lesions (22%) and cerebral atrophy (19%; Graybill
of liver transplant recipients (Rabkin et al. 2000), et al. 2000). Once formed, these lesions may persist
and 3.9% of renal transplant patients acquiring many years after apparent disease eradication
the disease (Bach et al. 1973). The susceptibility (Hospenthal and Bennett 2000). The persistence
of transplantation patients to cryptococcosis can of these lesions will undoubtedly complicate the
apparently be influenced by the type and dura- diagnosis of a possible relapse, and may thus be of
tion of immune suppression, most infections oc- limited use as indication of therapeutic response.
curring more than 2 months after transplantation Definitive diagnosis of cryptococcal meningitis
or as a result of increased suppression due to rejec- is made by isolating organisms from the CSF. There-
tion (Snydman 2001). Variation in the incidence of fore, lumbar puncture of these patients is critical to
cryptococcosis associated with different transplan- the proper diagnosis and appraisal of their infec-
tation groups may be a product of variation in en- tion. A number of CSF parameters are of particular
vironmental exposure to this pathogen during the diagnostic and prognostic value, namely, opening
post-transplantation period (Snydman 2001), and pressure, mycological culture, cryptococcal anti-
not a function of the type of tissue transplanted, gen titer, and India ink staining (Fig. 13.1).
as has been seen in outbreaks of histoplasmosis in Opening pressure of the CFS has been shown to
renal transplant patients (Wheat et al. 1983). be a prognostic indictor of cryptococcal meningi-
Cryptococcus neoformans Pathogenicity 239
Cerebrospinal CRAG titers are also subject to false

positive reactions. HIV-positive patients have been
identified as having CSF CRAG titers without evi-
dence of the growth of the organism from serum
or urine, and without evidence of serum CRAG
titers (Manfredi et al. 1996). This may be a result
of the penetration of polysaccharide antigen in the
brain tissue of infected patients (Lee and Casadevall
1996). These deposits could be the source of anti-
gen detected in the CSF long after apparent clinical
cure has occurred.
India ink staining of the CSF from these pa-
tients is also a fairly reliable diagnostic test, with
76%–95% of meningitis patients having a positive
test (Graybill et al. 2000) when the test was per-
Fig. 13.1. India ink stain of Cryptococcus neoformans that
illustrates the large polysaccharide capsule formed on the sedimented cell pellet (Saito et al.
1999). One of the limitations of this test is that it
would not readily allow one to differentiate between
live and dead cells, and thus is of limited value in
tis, with a high opening pressure (>250 mm H2 O) determining treatment progress.
an indication of a poor outcome (Graybill et al. These tests are good indicators of the presence
2000). Management of these patients may require of Cryptococcus in the brain. However, they are
repeated lumbar punctures to reduce intercranial of lesser value as indicators of the amount of the
pressure (Graybill et al. 2000; Saag et al. 2000). fungus in the brain. In addition, only CSF culture
High intercranial pressure may also be treated by provides indications of the metabolic state of the
shunting excess CSF from the brain. In a study fungus, i.e., whether fungal cells are alive or dead.
of four patients that had intracranial hypertension Thus, a test that is more sensitive to the burden of
that was uncontrollable by pharmacologic therapy live or quiescent cryptococcal cells in the brain is
alone, Liliang et al. (2002) demonstrated that the needed in order to properly assess the effectiveness
installation of a ventriculoperitoneal shunt could of anti-cryptococcal treatment in these patients.
reduce intercranial pressure in these patients, and Cryptococcal pneumonia is the next most fre-
presumably increase their chances of survival. quent manifestation of cryptococcosis in AIDS pa-
Cerebral spinal fluid parameters may be nor- tients. It occurs as a primary infection in approx-
mal or near normal due to an absence of an inflam- imately 4% of cases, and is associated with gen-
matory response in patients with advanced HIV eral symptoms of pneumonia such as fever, cough,
disease (Graybill et al. 2000), and are thus of lesser pleuritic chest pain and/or dyspnea (Wasser and
diagnostic value. However, CRAG titers in infected Talavera 1987; White and Armstrong 1994; Cunha
CSF can also be very high, ranging up to 1:8000 or 2001b). However, approximately 50% of all crypto-
higher (Eng et al. 1986; Graybill et al. 2000). Be- coccosis patients have lung involvement. Lung in-
cause of this, testing CSF for the presence of CRAG fection is associated with the presence of subacute
in these patients is reliable, with a sensitivity of or chronic infiltrates on chest X-ray examination
approximately 90%–100% in some studies (Chuck (Cunha 2001b; Sax 2001). The findings can include
and Sande 1989; Tanner et al. 1994). Monitoring multifocal nodular or perihilar infiltrates with con-
CRAG titers in the CSF as an indication of treat- solidation. Lung function, measured as pO2 , is gen-
ment response may be of some predictive value. erally normal, which may explain why as many as
In a study of 73 cryptococcal meningitis patents, 25% of patients with pulmonary involvement may
83% of patients who ultimately responded to treat- not be symptomatic (Pappas et al. 2001). The defini-
ment had a reduction in CRAG titer (Powderly et al. tive diagnosis of cryptococcal pneumonia requires
1994). However, only 57% of those who did not re- the isolation of the organism from bronchoalveolar
spond to treatment had no change or an increase in lavage or pleural fluid (White and Armstrong 1994).
their CRAG titer. Therefore, CSF CRAG titer reduc- This test is fairly sensitive, with a positive culture
tion is a better indicator of treatment success than obtained in approximately 85% of cases (Cameron
an increase in antigen titer is of treatment failure. et al. 1991).
In patients with disseminated disease, measur- Cryptococcosis in children is relatively rare,

able serum cryptococcal antigen (CRAG) is present even in HIV-infected patients (Gonzalez et al.
60% or more of the time. Graybill et al. (2000) found 1996). On the basis of a retrospective study of
CRAG titers in the patients examined could be very 1478 pediatric AIDS cases, Abadi et al. (1999)
high, ranging from 1:1037 to 1:16,384. Testing of inferred that the 10-year point prevalence of
serum from patients with disseminated disease for cryptococcosis in these patients was 1.4% (Abadi
CRAG has a similar sensitivity as that of testing CSF et al. 1999), compared to an estimated 6%–10%
(Tanner et al. 1994). To reduce the interpretation of rate for adults. This disparity may be influenced
false positive reactions, a titer of less than1:8 has by exposure to this pathogen. In a study of urban
been proposed as a useful cutoff point (White and children (Goldman et al. 2001), less than 50%
Armstrong 1994). of children under 2 years had serum reactive to
Serum CRAG titers can also be misleading. In cryptococcal proteins. This figure increased to 70%
animal models, circulating polysaccharide and pu- in children 5 or more years old. In another study
rified glucuronoxylomannan are both quickly re- of HIV-negative adults from the same urban area,
moved from the circulation by macrophages and 100% (n = 26) had serum reactive to cryptococcal
concentrated in the tissues of the reticuloendothe- antigens, indicating that most adults in this urban
lial system (Kappe and Muller 1991; Lendvai et al. setting have been immunologically exposed to
1998; Grinsell et al. 2001). If this were the case with C. neoformans (Chen et al. 1999). The symptoms
human patients, then there is a potential source associated with infection appear to be similar to
of antigen present in the body possibly long af- those seen with adults, with headache and fever
ter the elimination of the organism. If this store of being the most frequent (Abadi et al. 1999).
antigen were transiently released by the death of Cryptococcal infection in HIV-negative adults
the macrophage, then this could explain the iden- is similar to that seen in HIV-positive patients,
tification of patients with antigenemia yet with- with the exception of a slightly reduced frequency
out culture-proven disease. Conversely, a newly es- of CNS infection (51%) and elevated frequency of
tablished infection may go undiagnosed by serum pulmonary infection (36%; Pappas et al. 2001). The
testing because the reticuloendothelial system can symptoms of both pulmonary and CNS infection
efficiently remove circulating antigens and yeast also appear to be similar to that seen in HIV-
cells. This notion is bolstered by the observation positive patients, with the exception of a higher
that Cryptococcus cells may survive in macrophages frequency of nausea (72.3%) and altered mental
that engulf them (Feldmesser et al. 2000). Eventu- status (52%) associated with CNS infection in HIV-
ally, the macrophage fills with polysaccharide and negative compared to HIV-positive patients.
dies, liberating the organism (Tucker and Casade-
vall 2002). Thus, a test that does not rely on anti-
genic molecules of the capsule and yet is very sen- C. Therapy
sitive to the presence of the organism is obviously
needed. Practice guidelines for the management of cryp-
DNA- or RNA-based detection methods have tococcal disease have been issued periodically. For
been demonstrated to be both sensitive and accu- a complete treatment of the current recommen-
rate tests for the presence of a number of pathogens. dations, the reader should see Saag et al. (2000).
The use of DNA-based detection methods has been The recommendations for management of cryp-
demonstrated using medically important yeasts in- tococcosis vary depending on the immune sta-
cluding Cryptococcus (Rappelli et al. 1998; Aoki tus of the patient, and the site and extent of dis-
et al. 1999; Posteraro et al. 2000; Lindsley et al. ease. Treatment options include amphotericin B,
2001). DNA-based detection methods, while sensi- a combination of amphotericin B and flucytocine,
tive and specific, currently only provide evidence fluconazole or itraconazole. For patients with re-
of the presence of the organism. They do not pro- nal disease, a lipid formulation of amphotericin
vide information regarding the metabolic state of B can be substituted in the induction regimen.
the organism. Polymerase chain reaction amplifi- Cryptococcal meningitis in any patient is poten-
cation of the cDNA of genes associated with the tially life threatening and should be treated aggres-
vegetative or pathogenic growth of the organism sively.
would provide an indication both of the presence In the immune-competent host, asymptomatic
of the organism as well as its metabolic state. patients with pulmonary disease may be closely
monitored for disease progression or treated In HIV-negative patients (cancer and trans-
with fluconazole for 3–6 months. Patients with plant) the relapse rate is variable. In an early study,
symptomatic mild pulmonary disease should be the relapse or treatment failure rate approached
treated with fluconazole orally for 6–12 months. 100% (Kaplan et al. 1977). In this study of 46 crypto-
Meningeoenchephalits of HIV-negative patients coccosis patients, all patients died within 600 days
can be treated with a combination of amphotericin of diagnosis. This may be due to the rather ad-
and flucytocine, possibly followed by oral flucona- vanced disseminated disease these patients had at
zole for at least 8–10 weeks. To assess whether the time of diagnosis. In more contemporary stud-
the chosen therapy has achieved sterilization of ies, the relapse rate appears to approach 0% (White
the CSF, lumbar puncture should be performed. et al. 1992; Kontoyiannis et al. 2001), probably due
If CSF sterility has not been achieved, additional to earlier diagnosis and better maintenance ther-
therapies for 6–12 months could be instituted. If apy.
oral fluconazole is not tolerated, itraconazole can
be substituted.
The treatment regimen for HIV-positive III. Models for Studying Pathogenesis
individuals with CNS involvement is different than
that of HIV-negative patients, in that treatment A. Animal Models
durations are much longer owing to the profound
and lifelong immune suppression experienced The pathogenesis of Cryptococcus has been studied
by these patients. Even with effective, highly using various animal models of pulmonary, dis-
active antiretroviral therapy (HAART), lifelong seminated and CNS disease. Disseminated and pul-
maintenance therapy is recommended. However, monary models of disease have been established in
in a study of six patients who responded to HAART mice and rats (Casadevall and Perfect 1998). The
of more than 12-month duration with an increase disseminated mouse model is based on the injec-
in total CD4+ count (>150 cells/µL), Aberg et al. tion of live cells into the circulation via the lateral
(2002) demonstrated that anti-cryptococcal treat- tail vein. This leads to the hemotogenous dissemi-
ment could be discontinued without relapse. If this nation of the cells to all parts of the body.
observation is supported by more studies, then the A mouse model of pulmonary infection has
recommendation to continue anti-cryptococcal been established using two different inoculation
treatment for the life of the patient may be revised routes. The first involves the surgical resection of
in the future. the trachea of a mouse. The inoculum is delivered
Due to the severity of the disease, it is rec- to the lungs in a 30-GA needle inserted into the
ommended that HIV-compromised individuals trachea between adjacent cartilage rings.
with pneumonia and CD4 lymphocyte counts of A second inoculation route is through inhala-
less than 200 cells/ml always be tested for dissem- tion. This is thought to more closely mimic the
inated cryptococcosis. For HIV-positive patients natural course of infection. In this procedure, the
with asymptomatic culture-proven pulmonary mice are anesthetized and the animals suspended
disease or those with mild to moderate symptoms, by their superior incisors on a string. An inoculum
treatment of fluconazole for life is recommended. is then delivered to each animal by applying a cryp-
Severe or progressive disease should be treated by tococcal cell suspension to one nare of the animal.
the use of amphotericin B, followed by fluconazole The inoculum flows through the nasal cavity down
or itraconazole therapy. These patients should also the back of the animal’s throat and is aspirated into
be evaluated for extra-neural disease, which is the lungs. This type of inoculation is a very efficient:
present in up to 50% of cases (Kovacs et al. 1985). greater than 90% of the inoculum can be recovered
Maintenance therapy in HIV-positive patients from the lung 2 hours after inoculation, and the
currently consists of a lifelong course of oral flu- procedure is less technically demanding than are
conazole. Amphotericin B can be substituted for the surgical or tail vein techniques.
patients who cannot tolerate fluconazole or have A cerebral model of cryptococcal infection has
had multiple relapses while on azole maintenance also been established in corticosteroid-treated rab-
therapy. Maintenance therapy is necessary because bits (Perfect et al. 1980). The rabbits are immuno-
of a 20%–60% relapse rate following initial anti- suppressed by methyl-prednisone treatments. The
cryptococcal therapy (Bozzette et al. 1991; Pow- rabbits are then infected by inter-cisternal inoc-
derly et al. 1992; Powderly 1996). ulation. The type and extent of infection in each
model of cryptococcosis is subject to variation re- is believed to grow in a yeast form in the environ-
lated to strain differences between Cryptococcus ment, similar to the growth form in the tissues. It
strains and the strain of animals used in the exper- has a bipolar mating system with α(Mat α) and
iment. a (Mat a) mating type cells. This fungus is saprobic
and has a strong association with pigeon droppings
and some plants. In part because of its association
B. Alternative Models with the excreta of the cosmopolitan pigeon and
other birds, C. neoformans has a worldwide distri-
Non-mammalian models of pathogenesis have bution.
been developed for C. neoformans and include Based on the antigens of the polysaccharide
amoeba, slime mold, worms and flies. Although capsule, Cryptococcus neoformans has been sepa-
these models will probably be useful, none will rated into four serotypes (A, B, C, D; Wilson et al.
completely mimic human disease, and it is unlikely 1968). There are three varieties of C. neoformans
that they will replace mammalian models. Because currently recognized – var. neoformans, var. grubii
C. neoformans is commonly found in soil and does (Franzot et al. 1999) and var. gattii. Variety grubii is
not have an obligate life-cycle stage that requires composed of C. neoformans of serotype A whereas
passage through a mammalian host, it has been variety neoformans is composed of C. neoformans
puzzling why environmental isolates generally of serotype D. Variety gattii is composed of C. neo-
retain the capacity to cause disease. It has been formans serotypes B and C, and has been found as-
hypothesized that many C. neoformans virulence sociated with eucalyptus trees. Varieties grubii and
mechanisms have been maintained by constant neoformans (serotypes A and D) are responsible
selective pressure from various soil organisms for the majority of human infections in the United
(Steenbergen et al. 2001). The interactions with States and Europe whereas variety gattii is more
amoeba and slime mold are very similar to the commonly seen in infections of immunocompe-
interactions of C. neoformans with macrophages tent individuals from Australia, Southeast Asia and
(Steenbergen et al. 2001, 2003). C. neoformans has other tropical areas of the world. Infection of im-
the ability to survive phagocytosis by Acantham- munocompromised individuals is predominantly
boeba castillanii and Dictyostelium discoideum, due to variety grubii, regardless of geographic ori-
replicate in the phagocytic vacuole, and cause gin. A recent outbreak of var. gattii on Vancouver
killing of these phagocytic hosts. The presence of Island has been puzzling (Stephen et al. 2002). At
the polysaccharide capsule on C. neoformans was least 66 individuals have been diagnosed with Cryp-
shown to inhibit phagocytosis and be important tococcus since 1999, a tenfold increase in the normal
for fungal survival in the presence of the hosts. rate, and many of these individuals have been oth-
Other alternative models include Caenorhabdi- erwise healthy. The organism responsible has been
tis elegans (Mylonakis et al. 2002) and Drosophila identified as a serotype B strain, and rather than be-
melanogaster (Apidianakis et al. 2004). Both of ing found in conjunction with eucalyptus trees, it
these organisms are killed when they ingest C. ne- has been found in the soil or associated with other
oformans, whereas related fungal species do not trees. This expansion of habitat by this primary
have this effect. Although some similarities with pathogen and the associated increase of exposure
mammalian models systems are evident, there are of humans makes it imperative to understand the
some clear distinctions, for example, the capsule similarities and differences between the varieties.
does not appear to be important for virulence in The clinical presentation and outcome differ
either C. elegans or D. melanogaster. between infection with var. gattii and var. grubii.
Infection with var. gattii tends to occur in immune
a competent host, which separates it from infection
IV. Biology of Cryptococcus neoformans with var. neoformans. The clinical presentation of
infections with variety gattii in these patients tends
A. Taxonomy to differ somewhat from that of variety grubii. In-
dividuals infected with var. gattii have a higher
Cryptococcus neoformans is a member of the Ba- frequency of abnormal CT findings (78%), with
sidiomycetes. Its telomorph state was recognized more multiple ring enhancing lesions (40%), fo-
in 1975 and assigned to the genus Filobasidiella cal CNS involvement (29%), pulmonary infection
by Kwon-Chung (1975). Cryptococcus neoformans (65%), papilledema (50%), decreased mentation
(31%) and seizure (38%) than those infected with ease (Yue et al. 1999). However, STE12 does appear
var. grubii (Mitchell et al. 1995; Speed and Dunt to affect the virulence phenotype in var. neofor-
1995). Conversely, they have a lower frequency of mans (serotype D) strains (Chang et al. 2000). The
urinary tract infection (≤ 8%) and cryptococcemia only alpha mating type-specific gene is SXI1, which
(≤ 8%). The outcome from infection with var. gat- has been identified as a protein that is essential for
tii is significantly poorer than that for infections cell identity and sexual development, but not for
with var. grubii. Patients with var. gattii infections virulence (Hull et al. 2002, 2004).
require more days of treatment, have more neuro- In a recent study of 358 clinical isolates from
logical sequelae, require more surgery, and suffer the United States, no Mat a cells of var. grubii were
a higher frequency of relapse than those infected isolated (Yan et al. 2002). These data support the
with var. grubii. notion that Mat a cells of var. grubii must be ex-
ceedingly rare (Kwon-Chung and Bennett 1978),
but it is not extinct. In a survey of clinical isolates
B. Life Cycle from Tanzania, a single isolate of C. neoformans var.
grubii was identified as being Mat a (Lengeler et al.
This fungus is normally haploid and reproduce
2000b), indicating that this mating type, while rare,
asexually by budding or haploid fruiting. Hap-
still exists in the environment. Other serotype A
loid fruiting was first discovered in Mat α cells
Mat a strains were identified from the environment
that were subjected to severe nitrogen and mois-
(Viviani et al. 2001) and from a clinical isolate (Vi-
ture limitation (Wickes et al. 1996), but it can also
viani et al. 2003). Additional Mat a strains were
occur in Mat a cells (Tscharke et al. 2003). In re-
identified from Botswana and shown to have en-
sponse to this stress, cells form monokaryotic hy-
gaged in sexual recombination in the environment
phae with pseudo-clamp connections. These hy-
(Litvintseva et al. 2003).
phae then produce basidiospores on which haploid
Recent work has taken the rare Mat a strains
basidiospores form. Some of these basidiospores
and crossed them to a hyper-responsive serotype A
are less than 3 µm and are thus capable of deep
Mat α strain that was particularly prolific at mat-
alveolar penetration. Under similar conditions, C.
ing (Neilson et al. 2003). The resulting strains were
neoformans is capable of sexual reproduction. Un-
crossed with the commonly used clinical isolate
der nitrogen starvation, if hyphae from opposite
H99, and then backcrossed ten times to produce
mating types encounter each other, the hyphae fuse
a congenic mating pair for var. grubii. This mat-
and dikaryotic hyphae are produced. Sexual devel-
ing pair is proving highly useful for many genetic
opment continues and haploid basidiospores are
applications, including crossing strains with differ-
produced.
ent mutations and for linking virulence traits with
mutations.
C. Genetics
The mating type locus in C. neoformans is found D. Molecular Genetics

on a chromosomal segment of about 100 kb (Karos
et al. 2000; Lengeler et al. 2002), which is much C. neoformans is normally haploid, which facili-
larger than previously thought. It contains at least tates certain reverse genetic manipulations such as
20 genes, both specific for mating as well as of gene deletion and replacement experiments. Over
other functions. In this region, homologs of the the last 15 years, tools have been developed, such
mating-related genes STE20, STE11, STE12, MF1 as the selectable markers, transformation systems,
(pheromone precursor) and CPR1 (pheromone re- reporter genes and regulatable promoters, which
ceptor) are located. The MF1 and CPR1 homologs make genetic manipulation possible.
are not mating type-specific whereas STE20, STE11, A number of several selectable markers have
and STE12 have mating type-specific alleles for been developed for use with C. neoformans. These
Mat and Mat a cells (Lengeler et al. 2000a, 2002; include hygromycin B resistance conferred by the
Clarke et al. 2001). The STE12 homolog is impor- hygromycin B phosphotransferase gene (hpt) from
tant for haploid fruiting but does not appear to sig- Escherichia coli (Cox et al. 1996), G418 resistance
nificantly affect mating or virulence of var. grubii conferred by the kanamycin gene (kan) from
(serotype A) strains using the rabbit model of CNS the transposable element Tn5 (Hua et al. 2000),
disease and the mouse model of disseminated dis- resistance to the antibiotic phleomycin from the
bleomycin resistance gene (ble) also from the and episomal forms of the transforming DNA. The
transposable element Tn5 (Hua et al. 2000), and relative frequency of each form also varied among
resistance to the aminoglycoside nourseothricin the transformation procedures.
conferred by the nourseothricin acetyltransferase The second transformation procedure for C.
gene (nat1) from Streptomyces noursei (McDade neoformans is biolistic transformation (Toffaletti
and Cox 2001). In addition to these dominant et al. 1993). In this procedure, transforming DNA
selectable markers, the auxotrophic markers is coated onto 0.6-µm gold beads. These beads are
of phosphoribosylaminoimidazole carboxylase then shot out of a device using high-pressure he-
activity conferred by the ADE2 gene isolated lium gas. The beads are accelerated by the gas and
from var. grubii (Toffaletti et al. 1993) and var. strike recipient cells on agar plates, transferring the
neoformans (Sudarshan et al. 1999), and oroti- coated beads to the nucleus of the cell and liberat-
dine monophosphate pyrophosphorylase activity ing the attached DNA. The efficiency of this form
conferred by the URA5 gene isolated from var. of transformation is subject to the same sources
neoformans (Edman and Kwon-Chung 1990) have of variation as that of the electroporation pro-
been developed. In order to use these auxotrophic cedure. Published transformation efficiencies also
markers, appropriate recipient strains have been vary from 0 to >1000 transformants per µg of trans-
created. forming DNA due to those factors.
To drive the expression of these genes as well Recently, Agrobacterium tumafaciens-mediat-
as the expression of other genes, the promoter ed transformation has been demonstrated for C. ne-
and terminator region of a number of several oformans (Idnurm et al. 2004). Transforming DNA
genes have been isolated and incorporated into is inserted between the flanking T-DNA ends in
numerous vectors. These include the promoter the A. tumafaciens plasmid. The bacteria are co-
and terminator of the inducible C. neoformans var. cultured with Cryptococcus, and then selection is
neoformans GAL7 gene (Wickes and Edman 1995), applied to eliminate the bacteria and identify fun-
the constitutive promoter of the C. neoformans var. gal transformants. The advantages of this system is
grubii actin gene (Cox et al. 1995), the terminator that it is relatively inexpensive to set up, it can be
region of the N-myristoyltransferase (NMT1) gene used to make random insertions, it does not seem
of var. grubii (Hua et al. 2000) or the TRP1 gene to generate unstable extrachromosomal transfor-
(McDade and Cox 2001), the constitutive promoter mants, tandem repeats of the inserted DNA do not
of the glyceraldehyde-3-phosphate dehydrogenase seem to occur, and because the ends of the inserted
gene from var. neoformans (Varma and Kwon- DNA are defined, cloning the flanking sequences
Chung 1999), and the inducible promoter from the is easier. The disadvantages are that constructing
copper-regulated CTR4 gene (Ory et al. 2004). the Agrobacterium plasmid is still relatively time-
In addition to these selectable markers, green consuming, multiple insertion events in a single
fluorescent protein (GFP) technology has also been cell can still occur, resulting phenotypes are not
developed for use with C. neoformans. A yeast- always linked to the inserted DNA, and it has not
optimized green fluorescent protein (yGFP) was been demonstrated to integrate by homologous re-
fused to the GAL7 and the MFα promoter to combination.
demonstrate differential expression in the rabbit
model of CNS infection (Del Poeta et al. 1999a).
Along with the development of vectors for gene V. Genome Project
expression and disruption, technologies for the
transfer of the genes to recipient cells are required. A. Description of the Genome
Three transformation procedures have been devel-
oped to transfer recombinant DNA into the genome The genome project for C. neoformans began in
of C. neoformans. The first is the use of electropo- 1999 (Heitman et al. 1999), and by mid-2004, five
ration to transform yeast cells (Edman and Kwon- Cryptococcus genomes have been sequenced. As
Chung 1990). The efficiency of this procedure varies preliminary studies for the genome sequencing,
depending on the locus interrupted, the marker several features of the genome were determined
used, the vector construction, and differences in to help anticipate timeframe and potential pitfalls.
the recipient strain. Published transformation ef- To estimate the size of the genome, C. neofor-
ficiencies vary between 0 and >10,000 per µg of mans chromosomes were separated by pulsed field
transforming DNA, resulting in both integrated electrophoresis using contour-clamped homo-
geneous electric field (CHEF) gels (Perfect et al. assays. Finally, R265 is the var. gattii strain that
1989; Wickes et al. 1994). Examination of 20 is causing the outbreak on Vancouver Island and
strains by pulsed field electrophoresis suggested being sequenced at lower coverage.
that there are between nine and 15 chromosomes,
depending on the strain (Wickes et al. 1994).
C. Physical Mapping
Most bands on the CHEF gels were counted as
single chromosomes, but bands with more intense The genomes of JEC21, H99, B3501, and WM276
ethidium bromide staining were counted as two or have been physically mapped using fingerprinting
three chromosomes, depending on the intensity of of bacterial artificial chromosome (BAC) libraries
the staining. Within varieties, the number of chro- at the University of British Columbia (Schein et al.
mosomes was not constant. Five var. grubii strains 2002; James Kronstad, personal communication).
varied between 10 and 13 chromosomes, five var. BAC libraries were constructed for each strain. The
neoformans strains had 12–15 chromosomes, and inserts for the JEC21 library averaged 108 kb, and
ten var. gattii strains had 9–14 chromosomes. the inserts for the H99 library averaged 107 kb. The
Careful size analysis of chromosome bands from BACs were restricted with HindIII and the frag-
five strains representing all serotypes suggested ment sizes carefully analyzed. Overlapping BACs
that the genome is between 21 and 25 megabases were identified by having overlapping HindIII frag-
(Wickes et al. 1994), although the sizes of the ments. The BACs have been assembled into 20 con-
larger chromosomes had to be estimated because tigs for each strain. The contigs are long enough to
the largest available marker was smaller than the be chromosomes, and they cover 15.79 Mb of JEC21
larger C. neoformans chromosomes. genomic DNA and 15.55 Mb of H99 genomic DNA.
Analysis of 24 genomic clones of C. neofor- These sizes are smaller than the original predicted
mans genes deposited in Genbank suggested that sizes of the genomes based on migration of chro-
the A+T content was close to 50%, indicating that mosomes on CHEF gels. One likely explanation for
there should be little difficulty with sequencing. some of the discrepancy is that there are a few gaps
Comparison of genomic clones to cDNA clones sug- in the assembly, since there are more contigs than
gested that typical C. neoformans genes contained chromosome bands on the CHEF gels. These gaps
an average of six introns per gene, and that the may represent centromers or other repetitive ele-
introns were generally short, with an average size ments that are difficult to clone. The physical maps
of 50–60 nt. This relatively high number of introns now provide an excellent and useful framework for
suggested that a concerted effort be made to de- the assembly and finishing of JEC21 and H99.
lineate the C. neoformans intron/exon boundaries,
and to develop gene prediction software specifically
for C. neoformans. D. Meiotic Map
A genetic linkage map for the MAT α strain B3501

and its sibling MAT a strain B3502 was constructed
B. Strain Choices
(Marra et al. 2004). Sixty-three polymorphic mi-
Five strains are currently the subject of genome crosatellite loci, 228 restriction fragment length
projects: var. neoformans, strains JEC21 and polymorphisms (RFLPs), two indel markers, and
B3501, var. grubii, strain H99, and var. gattii, six mating type-specific genes were identified and
strains WM276 and R265. JEC21 is a MAT α used to create the linkage map. The 301 genetic
strain that has a congenic mating partner, and markers were placed into 20 linkage groups that
B3501 shares 50% of the JEC21 genome, but is were subsequently mapped to the 14 chromosomes
significantly more thermotolerant and virulent. in both B3501 and B3502 on CHEF gels. Six chro-
From var. grubii, the strain H99 was the unanimous mosomes contained two linkage groups. Relating
choice and is a clinical isolate that has been used the linkage groups to the separated chromosomes
extensively for reverse genetics and virulence revealed two reciprocal translocations.
assays. It is highly virulent in all animal models,
particularly the immunocompromised rabbit E. Shotgun Sequencing
model (Perfect et al. 1980). WM276 was selected
as the type strain for var. gattii since it was used Shotgun sequencing has been completed for the
extensively in the laboratory and in virulence five Cryptococcus neoformans strains, and the data
are available for BLAST analysis and download- to genes’ important for virulence. Sequencing and
ing from the different genome center websites (see annotation of both strains revealed that there are
Table 13.1). The coverage is greater than 10× for only a few strain-specific genes.
each of the strains, except WM272 that has 6× The var. grubii H99 genome was cloned by
coverage. The genome of var. neoformans strain random shotgun cloning into plasmids. This
JEC21 genome is considered the “type” strain for has been a joint effort by Duke University and
C. neoformans, and has been finished by The In- the Broad Institute, whereas sequencing var.
stitute for Genome Research (TIGR; Loftus et al. gattii strain WM276 has been performed at the
2005). TIGR made JEC21 libraries using small and University of British Columbia. Completion of
medium inserts in plasmid vectors, so that paired these projects will provide valuable comparative
reads could be obtained. Paired reads are assem- information about strain differences. Because the
bled by sequencing each end of an inserted genomic varieties have different epidemiologies, it may be
fragment. The advantage of paired reads is that if possible to dissect parameters that are responsible
there is a gap in the assembled sequence, some of for the different host and environmental ranges
the paired reads may fall into two different contigs by comparison of the genomes and by functional
and thus can be used to link the contigs together characterization of gene expression.
and provide a template for closing the gap. Contigs
from the assembled genome sequence have been
mapped to the physical BAC map (Schein et al. F. EST Sequencing
2002) and to the meiotic map (Marra et al. 2004)
described above. C. neoformans genes have an average of about six
The genome of the other var. neoformans introns each. The identification of intron splice
strain, B3501, has been completed by Stanford signals is an important prelude to efficient and
University (Loftus et al. 2005). The data for accurate gene finding. To this end, an expressed
this project have been generated by sequencing sequence tag (EST) sequencing project has been
a combination of plasmid clones and M13 clones. completed at the University of Oklahoma (Kupfer
M13 is a single-stranded vector, so only one side et al. 2004), and the data from that project can be
of the insert can be sequenced. The clones are viewed and downloaded from their website (see
rapidly prepared, but lack the advantage of the Table 13.1). cDNA libraries have been constructed
paired end reads. Because the virulence of JEC21 for H99 and for B3501. Both ends of several
and B3501 are very different, with JEC21 being thousand clones have been sequenced. Analyses of
weakly pathogenic and B3501 being much more the cDNA sequences have resulted in the assembly
virulent, comparison of the genomes of these two of complete coding sequences. Comparison of
closely related strains could provide some clues as the cDNA sequences with the genomic sequence
Table. 13.1. Websites of C. neoformans genome databases
Website Genome center Strain Task

www.tigr.org/tdb/e2k1/cna1/ The Institute JEC21, var. neoformans Sequencing, assembly,
for Genome Research (TIGR) annotation
www.sequence.stanford.edu/ Stanford Genome B3501, var. neoformans Sequencing, assembly,
group/C.neoformans/ and Technology Center annotation
cgt.genetics.duke.edu/data/ Duke University Center H99, var. grubii Sequencing, assembly,
index.html for Genome Technology annotation
www.broad.mit.edu/annotation/ Broad Institute H99, var. grubii Sequencing, assembly,
fungi/cryptococcus_neoformans/ annotation
rcweb.bcgsc.bc.ca/cgi-bin/ University H99, var. grubii; Physical mapping
cryptococcus/cn.pl of British Columbia B3501, var. neoformans;
JEC21, var. neoformans;
WM276, var. gattii
www.bcgsc.ca/about/news/ WM276, var. gattii Sequencing, assembly,
crypto_public annotation
www.genome.ou.edu/cneo.html University of Oklahoma H99, var. grubii; EST sequencing
B3501, var. neoformans
has pinpointed the location of introns within the Bank. Sequencing showed that when the manually
genomic sequence, and analysis of these introns annotated genes followed the GA/UG or GC/UG
have identified likely splice site signals, including splice junctions, the manual annotations were cor-
a 5 splice junction of GU(A/G)(A/C)G(U/C), a 3 rect, but that when they did not, TwinScan identi-
splice junction of (C/U)AG, and a branch point fied the correct junction, suggesting that the splice
signal of (C/U)U(G/A)A(C/U). junction sequence should be an important consid-
Importantly, 45,000 clones from a normalized eration of manual annotation (Tenney et al. 2004).
cDNA library from JEC21 were sequenced (Lof- Tests of over 100 genes predicted by TwinScan sug-
tus et al. 2005). The RNA was isolated from C. gested that it could correctly identify introns about
neoformans that was grown under eight different 90% of the time. This direct testing of gene pre-
conditions including different mating types, nutri- diction is valuable for several reasons: it is comple-
ents, stresses, pHs and temperatures. Sequencing mentary to EST datasets, it facilitates assessment
of a large number of cDNAs provided a substantial of the quality of the annotation, and it permits re-
framework for annotation of the genome. EST finement of the gene prediction programs.
coverage is estimated to be about 80% of the C. Another valuable approach to gene prediction
neoformans gene space. Sequencing this large has been to use multiple gene-finder programs
number of ESTs also provided the interesting and to compile the results (Loftus et al. 2005).
observation that C. neoformans has a significant By compiling the results of multiple gene-finding
level of alternative slicing and antisense RNA. Over programs with the EST sequencing data, a rela-
270 genes expressed differentially spliced RNAs, tively accurate set of predicted proteins has been
using the stringent criteria that three independent generated. It is estimated that there are about 6500
clones were found to contain the alternatively protein-coding genes in the JEC21 genome, 80% of
spliced form. Spliced, antisense transcripts were which are well supported by EST data. The number
detected for over 50 genes. The function of of genes is likely to change as gene prediction
alternative splicing or the potential for regulation software is improved, and as the presence or
of gene expression by these two mechanisms is yet absence of genes are experimentally determined.
to be determined.
VI. Analysis of Genes that Affect

G. Annotation Pathogenicity of C. neoformans
Genomic and cDNA sequences from the genome and the Prospects
projects have been subjected to BLAST analysis us- for Genomic Impact
ing the non-redundant databases at NCBI, and the
results are available on their respective websites. Many genes that influence virulence have been
The data on gene splicing signals derived from the identified in Cryptococcus using a variety of
EST projects have been used to train gene-finding techniques, including predictions based on other
software, including TwinScan, specifically for C. ne- pathogenic organisms and the analysis of mutants
oformans (Tenney et al. 2004). TwinScan utilizes with obvious in vitro phenotypes. These factors
ab initio signals like splice junctions, but incorpo- include the ability to produce a polysaccharide
rates comparative genomic data from closely re- capsule, melanin, urease, and phospholipase.
lated species. The availability of multiple, but di- Signal transduction pathways that regulate the
verged strains of Cryptococcus has enhanced the production of these virulence factors have been
gene-finding capabilities. Exons and splice junc- studied extensively. The ability to produce melanin,
tions are more likely to be conserved between re- urease and phospholipase are single-gene traits
lated species than are introns and other intergenic whereas capsule production requires a multi-step
regions. Incorporating the comparison of the H99 biosynthetic pathway requiring many genes in
genome to the JEC21 gene predictions improved the its biosynthesis. The fungi must also be able
gene predictions by 10%. The TwinScan predictions to survive in the phagolysosome and resist the
were empirically tested by cloning and sequencing reactive oxygen, nitrogen, and chlorinating species
predicted transcripts. RT-PCR was used to amplify that are produced as microbicidal agents. The
transcripts that were predicted by TwinScan that availability of genomic sequences is already having
differed from manually annotated genes in Gen- a major impact on facilitating the analysis of the
contribution of many more genes to virulence. munization (Kozel et al. 1977). This tolerance is as-
Having a complete set of predicted proteins fa- sociated with T-cell-dependent and -independent
cilitates rapid identification of paralogs of known pathways (Sundstrom and Cherniak 1993).
virulence genes, and homologs of virulence genes Polysaccharide can also modulate the innate
from other organisms. Comparison of gene sets of immune system. Whole cells or purified polysac-
fungal pathogens to non-pathogens may allow us charide induce the production of the complement
to identify pathogen-specific gene sets. Transcrip- factor C3 in mouse peritoneal cells (Blackstock and
tome or proteomic analysis of virulence mutants Murphy 1997) as well as C3 binding in both mouse
will probably lead to the identification of other and human serum (Wilson and Kozel 1992; Pfrom-
genes that impact the expression or actively of the mer et al. 1993). Factor C3 is converted to C3b by
virulence traits. The availability of mutants, either contact with the capsule (Pfrommer et al. 1993).
by insertional mutagenesis or by targeted deletion, The active C3b fragment is then quickly converted
will enable us to screen for genes that affect the to the inactive iC3b by contact with the capsule.
expression or activity of these virulence factors. Complement factor C5 and C5a are also involved in
the reaction with whole cells or polysaccharide of
Cryptococcus (Vecchiarelli et al. 1998). In addition,
A. Capsule Biosynthesis capsular polysaccharides may be capable of inter-
fering with the C5a receptor on neutrophils, which
Capsule production is a major virulence deter- also reduces the cell-mediated immune response
minant of Cryptococcus (for review, see Bose (Monari et al. 2002).
et al. 2003). The ability of C. neoformans isolates Mannoprotein is a relatively minor constituent
to produce a thick polysaccharide capsule is of the capsule. However, it is capable of producing
unique among the pathogenic fungi. Production immune-modulating effects on its own (Pietrella
of the capsule is regulated by environmental cues et al. 2001). Mannoprotein induces a strong
such as CO2 concentration, iron availability, and protective response when injected into mice,
the presence of asparagine and glucose in the leading to the production of the pro-inflammatory
culture medium (Granger et al. 1985; Vartivarian cytokines IL-12 and INF-γγ by macrophages.
et al. 1993). This capsule is composed of the Recently, two mannoprotein antigens have been
polysaccharides glucuronoxylomannan (GXM), isolated and identified. Both were deacetylases that
galactoxylomannan (GalXM) and mannoprotein were capable of producing a protective response
(MP; Cherniak et al. 1980, 1988, 1992; Bhattachar- in animal models when used as an immunogen
jee et al. 1984). The GXM is composed of mannose, (Levitz et al. 2001; Biondo et al. 2002).
xylose and glucuronic acid residues, and the minor There are at least eight genes known to affect
polysaccharide antigens GalXM and MP. The ma- capsule biosynthesis. CAP59, CAP64, CAP60 and
jority of the soluble cryptococcal polysaccharide CAP10 were discovered on the basis of their ability
consist of GXM and a lesser amount of GalXM and to complement capsule-deficient mutants (Chang
MP antigens (Cherniak and Sundstrom 1994). Be- and Kwon-Chung 1994, 1998, 1999; Chang et al.
sides being an impediment to phagocytosis (Kozel 1996). The function of these genes is unknown,
and Gotschlich 1982; Richardson et al. 1993) and but recent work suggests that cap59 mutants
subsequent vacuolar killing (Goldman et al. 2000; cannot transport the capsular material to the cell
Tucker and Casadevall 2002), the capsule of Crypto- surface (Garcia-Rivera et al. 2004). By predicting
coccus has immune-modulating properties as well. the biochemical reactions that would be required
Capsular polysaccharide has been demon- for capsule synthesis, Doering and colleagues
strated to affect the immune system in a number have identified another gene important for capsule
of different ways. Soluble polysaccharide interferes synthesis, a UDP-xylose synthase encoded by UXS1
with the cell-mediated immunity by disturbing (Bar-Peled et al. 2001). uxs1∆ strains are com-
the chemotactic and proliferative response of pletely avirulent (Moyrand et al. 2002). Perfect and
phagocytic cells (Mody and Syme 1993; Fujihara colleagues (Wills et al. 2001) cloned and character-
et al. 1997; Syme et al. 1999; Coenjaerts et al. 2001). ized a phosphomannose isomerase gene (MAN1)
It also reduces the antigen-presenting capacity of that affects capsule biosynthesis and virulence,
these cells (Retini et al. 1998). Injections of puri- since mannose is a major constituent of the C. neo-
fied polysaccharide can also reduce the immune formans capsule. Janbon and colleagues have iden-
response to subsequent cryptococcal antigen im- tified mutants that make an altered capsule, and
have identified the corresponding genes by com- gene (CNLAC1) of Cryptococcus (Williamson
plementation. One of these, CAS1, is required for 1994). This enzyme is found in the cell wall
GXM O-acetylation (Janbon et al. 2001; Kozel et al. of the fungus as well as the plasma membrane
2003), and the deletion of this gene results in a more (Polacheck et al. 1982; Zhu et al. 2001). Specific
virulent phenotype. This is the first example of the inactivation of this gene has been demonstrated
capsule structure playing a role in the pathobiology to confer a significantly less virulent phenotype to
of disease progression. Using predictions of bio- the recipient strain (Salas et al. 1996).
chemical pathways important for capsule synthesis Melanin production has been demonstrated
and the identification of potential paralogs of genes to protect C. neoformans from oxidative stresses
known to affect capsule synthesis, genomic anal- (Jacobson and Emory 1991; Jacobson and Tin-
ysis reveals at least 30 additional genes that could nell 1993). This property has been suggested to
be important for capsule biosynthesis. Two gene protect Cryptococcus from oxidative attack by
families related to CAP64 and CAP10 are found in phagocytic cells (Liu et al. 1999a). In addition,
the C. neoformans genome, and these can now be melanin may also protect the yeast from the effect
tested for their impact on capsule biosynthesis. of microbiocidal proteins (Doering et al. 1999),
It was shown recently that the capsule requires and may interfere with the host’s cell mediated
alpha-1,3-glucan in the cell wall in order to bind to immune reactions (Huffnagle et al. 1995). The
the cell surface of C. neoformans (Reese and Do- phenoloxidase activity of the laccase enzyme is
ering 2003). Glucanase treatments that included responsible for converting neurochemical com-
alpha glucanases reduced binding of capsular ma- pounds such as the catecholamines epinephrine,
terial. Availability of the genome sequence made it norepinephrine and dopamine to melanin (Shaw
possible to rapidly define the single gene respon- and Kapica 1972; Polacheck et al. 1982). This
sible for alpha-glucan synthesis, and to show by activity has led some to postulate that the presence
RNAi that reduction of expression of the alpha- of melanogenic substrates in the brain contributes
glucan synthase reduced the binding of capsule to to the cerebral tropism of Cryptococcus. However,
C. neoformans. although Cryptococcus can produce melanin in the
brain (Nosanchuk et al. 1999b, 2000; Rosas et al.
2000), there is evidence for the production of other
B. Melanin Biosynthesis oxidative compounds being produced (Liu et al.
1999b).
Cryptococcus neoformans is the only member of Expression of the laccase enzyme is negatively
the genus Cryptococcus that produces significant regulated by temperature and glucose concentra-
quantities of melanin, a brown to black pigment, tion. Cnlac1 production is repressed by the pres-
when grown on medium containing diphenolic ence of glucose in the culture medium (Polacheck
compounds (for review, see Zhu and Williamson et al. 1982). The derepression of laccase activity in
2004). Other species of Cryptococcus are apparently medium deficient in glucose was sensitive to cy-
capable of producing lesser amounts of the pig- cloheximide, indicating that protein synthesis is
ment (Ikeda et al. 2002). This trait has been used involved in the increase in Cnlac1 activity, and
to differentiate Cryptococcus neoformans from thus glucose does not interfere with the activity of
other Cryptococcus species as well as other yeasts the enzyme. Similarly, laccase activity is reduced
of medical importance such as Candida albicans in cells incubated at 37 and 40 ◦ C (Jacobson and
(Korth and Pulverer 1971; Shaw and Kapica 1972). Emory 1991), but the catalytic activity of the en-
Melanin is produced by Cryptococcus in the envi- zyme is not significantly affected at these temper-
ronment, is assumed to protect the cells against atures. Genomic analysis of C. neoformans has re-
oxidative stress and UV light damage (Wang and vealed a second laccase gene that contributes to
Casadevall 1994; Nosanchuk et al. 1999a), and may melanin production (Missall et al. 2005a; Pukkila-
help protect against antifungal agents (van Duin Worley et al. 2005).
et al. 2002).
The ability to produce melanin has been
demonstrated to be a virulence factor in experi- C. Phospholipase
mental models of cryptococcosis (Kwon-Chung
et al. 1982; Rhodes et al. 1982; Kwon-Chung and The ability to produce extracellular phospholipase
Rhodes 1986). Melanin is produced by the laccase has been identified as a virulence factor in both
bacteria and fungi. Extracellular phospholipase cyclic AMP- dependent protein kinase signaling
activity has been identified in Cryptococcus strains pathway is a major pathway that controls both
collected from clinical sources (Chen et al. 1997a, mating and the production of virulence factors in
b). The amount of phospholipase activity pro- var. grubii (D’Souza et al. 2001; Pukkila-Worley
duced by individual isolates has been correlated and Alspaugh 2004; Hicks et al. 2004). A G-protein
with their ability to cause disseminated disease. In coupled receptor detects the signal (Alspaugh
addition, a phospholipase gene from Cryptococcus et al. 1997). An adenylyl cyclase is not essential
(PLB1) has been cloned and its activity shown to for vegetative growth in C. neoformans, but has
be necessary for the full pathogenic potential of the been shown to regulate mating, production of
organism (Cox et al. 2001). Animals injected with melanin and capsule, and virulence in a mouse
plb1 deletion strains survived longer than those model (Alspaugh et al. 2002). A catalytic subunit
injected with wild-type strains. In the rabbit model of the cAMP-dependent protein kinase (PKA1) was
of CNS disease, animals injected inter-cisternally shown to be essential for production of virulence
with the mutant strains had significantly smaller factors, virulence and mating, and the deletion of
fungal burden in the CSF than did those of control the regulatory subunit of the protein kinase (PKR1)
animals. C. neoformans produces eicosanoids that was shown to have constitutive hyper-expression
may act as signaling molecules between fungal of capsule and melanin production, and was
cells, or between the host and the fungus (Noverr hyper-virulent in mouse models of cryptococcal
et al. 2001). Phospholipase has been implicated in infection (D’Souza et al. 2001). Interestingly, there
prostaglandin production, plb1 mutants exhibit is a striking difference in this pathway between
a defect in eicosanoid production (Noverr et al. the var. grubii strain, H99 and the var. neoformans
2003), and it has been hypothesized that secreted strain JEC21 (Hicks et al. 2004). Deletion of
phospholipases may help cleave the substrates the PKA1 gene in JEC21 did not result in the
needed for prostaglandin production from the reduction of melanin and capsule. Searches of
host membranes. the genome databases revealed a second PKA
paralog (PKA2). Deletion of PKA2 in H99 did not
affect virulence factor production, but deletion
D. Urease of PKA2 in JEC21 reduced expression of both
capsule biosynthesis and melanin. Interestingly,
Urease production has been shown to be involved in
deletion of PKA2 in JEC21 did not affect virulence
the pathogenesis of Cryptococcus. A Cryptococcus
in a mouse model, nor did deletion of PKR1 result
urease gene (URE1) has been cloned and sequenced
in hyper-virulence.
(Cox et al. 2000). The deletion of this gene produced
The calcium/calmodulin signal transduction
mutants that were significantly less virulent than
pathway also impacts virulence. Strains deleted for
were the wild-type or replacement strains when
calcineurin are avirulent and cannot grow at higher
infected by the lateral tail vein or the inhalation
temperatures (reviewed in Lengeler et al. 2000a;
route in mice. In contrast, rabbits infected inter-
Kraus and Heitman 2003). The protein kinase C
cisternally with mutant or wild-type strains did
(PKC) signal transduction pathway has been ex-
not have significantly different fungal burdens in
tensively studied in other fungi for its role in cell
their CSF. Thus, the ure1 phenotype may be subject
wall biogenesis. Deletion of the MAP kinase MPK1
to species- or organ-type variation. Recent studies
reduces cell integrity and virulence in C. neofor-
in mice have shown that ure1 strains grow well in
mans (Kraus et al. 2003). The deletion of homologs
the brain when they are inoculated directly, but do
of other genes in the PKC pathway including MKK2,
not disseminate to the brain from other inoculation
ROM2, SIT4, and LRG1 results in sensitivity to cell
sites (Olszewski et al. 2004), suggesting a role for
wall inhibitors and reduced virulence in a mouse
urease in dissemination.
model (Gerik and Lodge, unpublished data).
E. Signal Transduction Pathways

F. Mating Type
Signal transduction pathways control expression
of virulence factors including capsule biosynthesis Virulence and mating type have been linked in var.
and melanin production (for review, see Lengeler neoformans, but not in var. grubii. The vast ma-
et al. 2000a). Extensive work has shown that the jority of clinical isolates are Mat a, suggesting that
there was a significant difference in virulence between the strains, or to differences in gene expres-
tween the two mating types. The congenic mating sion (Loftus et al. 2005).
pair from var. neoformans, JEC21 and JEC20, were
tested for virulence in a mouse model, and it was
shown that JEC21 was more virulent in a mouse H. Survival in a Phagocytic Environment
model (Kwon-Chung et al. 1992). This result has and Resistance to Oxidative
been repeated in other laboratories. However, con- and Nitrosative Stress
genic pairs of var. grubii strain K99a and K99α
had identical virulence in mouse models and rab- An additional virulence-related trait of C. neofor-
bit models of pathogenesis (Neilson et al. 2003). mans is the ability of cells to survive in phagocytic
Genomic analyses may help to elucidate the dif- cells and invade endothelial and epithelial cells. C.
ference in virulence between the serotype D, var. neoformans has been demonstrated to survive in
neoformans mating type and the serotype A, var. the phagolysosome of human and rat macrophages
grubii strains. (Levitz et al 1999; Goldman et al. 2000). Persis-
tence in this compartment appears to rely on the
fusion of lysosomes with the phagosome contain-
ing the engulfed fungal cell. This fusion leads to
G. Taxonomy an acidification of the phagolysosome, which is
necessary for the continued survival of the or-
The epidemiology and environmental niche of the ganism (Levitz et al. 1997). C. neoformans grows
C. neoformans and C. gattii are significantly differ- well in medium with ambient pHs as low as 3.0
ent. C. gattii is largely found in subtropical parts (Lodge, unpublished data). C. neoformans cells per-
of the world, is found in conjunction with euca- sist inside phagolysosomes where they grow and
lyptus trees, and infects immunocompetent hosts. produce extracellular polysaccharide that is con-
Var. grubii and var. neoformans are found world- tained in additional vacuoles (Feldmesser et al.
wide, are often associated with soil contaminated 2000, 2001). Eventually, the host macrophage is
with bird droppings, and infect immunocompro- weakened and dies, liberating the resident fun-
mised hosts. A recent study examined var. grubii gal cells. C. neoformans is also capable of adher-
clinical isolates that were obtained from apparently ing to and invading endothelial and epithelial cells
healthy individuals (D’Souza et al. 2004). The au- (Merkel and Cunningham 1992; Goldman et al.
thors showed that the isolates were not hybrids of 1994, 2000; Ibrahim et al. 1995). Once bound, the
C. gattii and var. grubii, nor did they observe PKR1 cells are internalized by phagocytosis. The phago-
mutations that would be consistent with hyper- some then is transported through the cell and pre-
virulence. However, perturbations in cAMP sig- sumably exits from the cell. The incorporation of
naling, which is known to regulate expression of the fungus into the epithelial cell is associated with
virulence traits, were present in these strains. cell injury and death. This injury appears to be as-
Var. grubii and var. neoformans are also dis- sociated in part with the active production of some
tinct from each other. In general, the var. grubii factor, and is increased by the presence of capsular
strains are more virulent in animal models and constituents. Thus, C. neoformans has the capacity
are more prevalent as clinical isolates. Var. neofor- to evade the immune system by surviving in phago-
mans strains require higher inoculums in order to cytic cells. Once liberated, these cells can then es-
cause the same level of disease as that of a var. gru- cape the confining granulomas and invade adja-
bii strain, and var. neoformans is virulent in the cent epithelial cells. Because C. neoformans cells
rabbit model of meningitis. There is an ongoing are also capable of traversing endothelial cells, C.
effort to compare the genomes from the three sub- neoformans can invade small blood vessels as well
species in order to identify candidate genes that as escape from them. These phenotypes undoubt-
might play a major role in these phenotypic dif- edly play pivotal roles in Cryptococcal pathogene-
ferences. Genome comparison of two var. neofor- sis.
mans strains with very different virulence proper- One aspect of survival inside a phagocytic
ties, JEC21 and B3501, is complete. There are only cell that is amenable to genomic approaches,
a few genes that are unique to each strain, suggest- and has been characterized to a certain extent,
ing that the differences in virulence are probably is the ability to resist oxidative and nitrosative
due either to single nucleotide polymorphisms be- stresses. Macrophages and neutrophils produce
numerous reactive oxygen, reactive nitrogen mouse strain was less susceptible to the fhb1∆
and reactive chlorinating species in order to kill strain than to wild-type C. neoformans, but an
invading microorganisms. Since C. neoformans iNOS mouse was equally susceptible to the fhb1∆
can survive inside this environment, it must and wild-type fungus. In contrast, the GNO1 gene
have mechanisms to resist the reactive molecular was not required for in vitro NO resistance or for
species (reviewed in Missall et al. 2004a). Based on virulence (de Jesus-Berrios et al. 2003).
work in other systems, one can predict the kinds Taken together, many virulence-related genes
of proteins that may be important for resistance have been identified in C. neoformans by more tra-
to various reactive species. Analysis of the C. ditional methods, and genomic information is ac-
neoformans genomic sequences shows that there celerating and enhancing these studies. The abil-
are two superoxide dismutases to protect from ity to proliferate in the lung, parasitize phagocytic
superoxide. To reduce peroxides, there are four cells, escape from the lung, cross the blood–brain
catalases, several classes of peroxidases includ- barrier, and proliferate in the brain most probably
ing thiol peroxidases, glutathione peroxidases, requires more cryptococcal genes that just those
glutaredoxins and cytochrome c peroxidases. that have already been described. Genome-wide
To protect against reactive nitrogen species, C. analysis such as transcriptome or proteomic stud-
neoformans has a flavohemoglobin denitrosylase ies of expression during macrophage engulfment
and an S-nitrosoglutathione (GSNO) reductase. or during pathogenesis will probably reveal genes
C. neoformans has a Cu, Zn superoxide dismu- that could be important for C. neoformans survival
tase and a Mn superoxide dismutase. The Cu,Zn in vivo.
superoxide dismutase gene (SOD1) has been exten-
sively characterized. It was cloned, and sod1 dele-
tions were made in both var. grubii (Cox et al. 2003) VII. Post-Genomic Approaches
and C. gattii (Narasipura et al. 2003). sod1∆ mutants for Identification of Virulence
are less resistant to superoxide in vitro, less viru-
Genes and Their Use
lent in a mouse model, and less able to survive in
macrophage-like cell lines or human neutrophils. in C. neoformans
As described below, a thiol peroxidase was
shown by proteomic analysis to be up-regulated at A. Functional Genomic Techniques
high temperature, and has been shown to be im-
portant for resistance to oxidative and nitrosative Post-genomic approaches to analyzing biological
stress, survival in macrophages, and virulence in function, regulatory networks and processes often
a mouse model (Missall et al. 2004b). This thiol include techniques that permit global analysis of
peroxidase is predicted to require thioredoxin gene expression either at the RNA level or at the
for recycling. The entire thioredoxin pathway, protein level. The analysis of protein or mRNA ex-
including Tsa1, two thioredoxins and thioredoxin pression during the modulation of external con-
reductase, and the glutathione pathway, including ditions or during particular developmental states
three glutaredoxins and two glutathione peroxi- can provide useful clues about what genes might be
dases, are being analyzed for their contributions important for particular functions. Another use-
to resistance to oxidative and nitrosative stresses. ful approach is the systematic generation of gene
Both pathways play a role in resistance to oxidative knockouts to analyze the contribution of many
and nitrosative stress, and they both influence the genes to a particular phenotype.
survival of C. neoformans in a macrophage-like
cell line (Missall and Lodge 2005b; Missall et al. 1. Transcriptome Analysis
2005b).
Gene deletion of the C. neoformans flavo- a) Microarrays
hemoglobin denitrosylase gene (FHB1) and the Microarray analysis is a powerful technique that
GSNO reductase (GNO1) demonstrated that the allows one to simultaneously analyze comparative
FHB1 gene was important for resistance to NO (de RNA expression levels from thousands of genes. Es-
Jesus-Berrios et al. 2003). Deletion of the FHB1 sentially, DNA corresponding to the different genes
gene also reduced survival in macrophages and is spotted on a solid support, often a glass slide. The
virulence in a mouse model, and this reduction was DNA can be PCR products from genomic clones,
dependent on host NO production. A wild-type cDNA clones or oligonucleotides. RNA is isolated
from organisms or tissues grown under particu- ber of times that a tag is found is directly propor-
lar conditions. The RNA is labeled with fluorescent tional to the expression level of that gene. SAGE
dyes, and mixed with a differentially labeled control has the advantage that the libraries can be made
RNA and hybridized to the DNA on the slide. The before all genes in the genome are annotated, but
ratio of the experimental RNA to the control RNA the construction of the libraries is labor-intensive,
is measured by comparing the different fluores- and therefore the analysis of multiple replicates,
cent signals. Microarrays have been used in many either biological or technical, is rare.
other pathogenic organisms to delineate their de- Two studies have focused on the SAGE analy-
velopmental program and their response to envi- sis of C. neoformans. In the first, the transcriptome
ronmental stimuli such as macrophage engulfment, of C. neoformans grown at 25 ◦ C was compared to
iron limitation, antibiotics, or stress (reviewed in that at 37 ◦ C in two strains: the var. grubii clinical
Lorenz 2002; see, for example, Whiteway and Nan- isolate H99 and the var. neoformans strain B3501
tel, Chap. 8, this volume).
The first report of microarray analysis was re-
cently published. Kraus et al. (2004) used a >6000
element microarray consisting mostly of random
genomic clones to probe the expression profiles of
the strain H99 at 25 and 37 ◦ C. They reported that
the genes that were up-regulated at 37 ◦ C included
those important for cell wall biogenesis (chitin syn-
thases and a WSC domain protein), genes involved
in resistance to oxidative stress (catalases, super-
oxide dismutase, and other oxidases), genes in tre-
halose biosynthesis, a voltage-gated chloride chan-
nel important for melanin biosynthesis (CLC1),
and the transcription factor MGA2. The deletion of
MGA2 resulted in slower growth and a slight tem-
perature sensitivity at 37 ◦ C. Microarray analysis
of the mga2∆ mutant suggested that this transcrip-
tion factor is important for the regulation of lipid
biosynthesis.
Several laboratories have initiated microar-
ray studies using PCR products based on EST
sequences (J. Murphy, personal communication).
A spotted oligonucleotide microarray is being Fig. 13.2.a–f Serial analysis of gene expression (SAGE) li-
developed, and it should be available to the entire brary construction. a mRNA from cells or tissue is converted
academic community by 2005. The first arrays will to cDNA using biotinylated primers (dots). b DNA copy of
mRNA is cut with the anchoring enzyme (AE, in this ex-
carry 70-mer oligonucleotides representing genes ample, NlaIII) to make cohesive ends, and are purified by
in both serotype A and serotype D strains (C. Hull, binding to streptavidin coated magnetic beads. c The sam-
T. Doering and J. Lodge, personal communication). ple is split in two, and PCR primers A and B are ligated to the
cDNA. They contain a recognition site for a type IIs restric-
tion enzyme (tagging enzyme, TE, in this example, BsmF1)
b) SAGE and AE cohesive ends. The tagging enzyme cleaves 10 and
14 bp from the recognition site. d These primers are ligated
Serial analysis of gene expression (SAGE) is an- to the cDNA and then cut with the tagging enzyme that
other technique that is used to quantify changes in cleaves some distance away (9–14 bp) from the recognition
mRNA expression. This technique is diagrammed site, releasing the tagged end from the magnetic bead. The
tagged ends are purified using magnetic beads, and blunt-
in Fig. 13.2. RNA is isolated from cells grown un- ended. e The blunt tags are ligated together to form ditags.
der different conditions and cDNA is made from the Ditags are amplified by PCR using primers A and B. The tags
RNA. SAGE tags from the cDNAs are generated and are then released by digestion with anchoring enzyme and
ligated together. The tags from many genes are lig- primers removed with avidenated magnetic beads. f Ditags
are ligated into concatamers and cloned into a sequencing
ated into one large fragment, which is then cloned vector. These clones are sequenced, the number of times
into a plasmid vector and sequenced. The tags can each tag is sequenced recorded indicating the abundance of
be uniquely linked to a specific gene, and the num- the mRNA in the original RNA pool
(Steen et al. 2002). For H99, libraries containing colysis, mitochondrial function, lipid metabolism
over 30,000 tags were used in the analysis, and the and calmodulin-calcineurin signaling.
50 most abundant transcripts at 25 and 37 ◦ C were
identified, but the analysis of H99 was difficult due c) Other Methods of Global Gene Expression
to the lack of available genome data at that time. Li- Profiling
braries from B3501 were also generated. B3501 does
not grow as well at 37 ◦ C, and so the library from Methods to identify genes that were alternatively
the higher temperature was smaller. There were regulated under specific conditions, such as dif-
15,363 tags from the 37 ◦ C library and 65,399 tags ferential display or subtractive hybridization, have
from the 25 ◦ C library. The analysis of genes that been used extensively to discover genes that could
changed significantly showed that at 25 ◦ C, genes be important for virulence. These methods do not
for histones, sterol and lipid metabolism, and trans- require prior knowledge of the genome sequence,
porters were upregulated. Changes in the expres- but genome sequence helps to identify the entire
sion of lipid metabolism genes might be responsi- gene once a fragment is isolated. Once microarrays
ble for altered membrane fluidity at different tem- are readily available, these methods are likely to
peratures. At 37 ◦ C, genes for heat shock proteins, become less popular. Differential display uses non-
translational machinery, mitochondrial proteins, specific PCR primers to amplify sequences from
and stress response proteins were up-regulated. RNA isolated from cells grown under different con-
The second SAGE analysis focused on C. ne- ditions. The appearance of unique bands in one of
oformans gene expression in the rabbit model of the samples suggests that the RNA template corre-
meningitis (Steen et al. 2003). C. neoformans cells sponding to that band is up-regulated in that con-
were isolated from the cerebral spinal fluid of rab- dition.
bits. A SAGE library was generated and it was com- Using differential display, Rude et al. (2002)
pared to the in vitro libraries from the previous were able to show that the gene encoding isoci-
study. Surprisingly, the in vivo library shared more trate lyase (ICL1), a component of the glyoxylate
in common with the 25 ◦ C in vitro library with pathway, was up-regulated in strain H99 during
respect to the classes of genes that were repre- experimental cryptococcal meningitis in the rabbit
sented, sharing 1885 tags with the 25 ◦ C library model. Further characterization of this gene in C.
and only 632 with the 37 ◦ C library. Classes of neoformans showed that it was not required for vir-
genes that were abundant in vivo included those ulence, in contrast to results from other pathogenic
that encode proteins involved in protein biosyn- organisms.
thesis and catabolism, stress responses, respiration, Del Poeta and colleagues (Luberto et al. 2001)
signal transduction and transport. A few genes of had tested the role of inositol-phosphoryl ceramide
particular interest were abundant in vivo, includ- synthase (IPC1) in pathogenesis, and found that
ing mannitol-phosphate dehydrogenase, myoinos- reduced IPC1 expression resulted in a strain that
itol 1-phosphate synthase, and MP98, a secreted grew more slowly in macrophages and was less
mannoprotein that is a chitin deacetylase. virulent. Differential display of the IPC1 down-
A third study examined expression differences regulated strain was performed to identify targets
between C. neoformans in the presence and absence that might be responsible for this phenotype. This
of iron (Lian et al. 2005). In this study, genes that en- resulted in the identification of APP1, a novel reg-
coded putative components for iron transport and ulator of phagocytosis (Luberto et al. 2003).
homeostasis, including the FTR1 (iron permease) Lastly, a comparison of C. neoformans grown
gene, had higher transcript levels in the low-iron at 37 versus 25 ◦ C using a subtractive cDNA library
condition. In addition, lack of iron increased ex- showed that alternative oxidase was up-regulated
pression for putative extracellular mannoproteins at 37 ◦ C. The deletion of the AOX1 gene resulted in
including CIG1, and a glycosylphosphatidylinosi- a strain that was sensitive to oxidative stress and
tol transamidase, GPI8. Functional analyses of mu- was less virulent in the mouse model than wild type
tants in these genes showed that the ftr1∆ mutant (Akhter et al. 2003).
and cig1∆ mutant had impaired growth in low-
iron medium, and altered capsule regulation. The 2. Gene Expression Modulation
GPI8 gene appeared to be essential. Lastly, iron-
replete conditions led to elevated transcripts for With the advent of genome data, expression of spe-
genes for iron storage, nitrogen metabolism, gly- cific genes can be modulated systematically. Genes
can be “knocked out” in several ways. Genes can be viability of the organism as well as to characterize
targeted for disruption by homologous recombina- the effects of individual genes on the development
tion, their expression can be modulated by RNAi of this organism. Isolation of a tightly regulated
or antisense RNA, or an inducible promoter can be promoter from the copper-regulated CTR4 gene
used to control their transcription. (Ory et al. 2004) allows one to control RNAi
expression (Missall and Lodge, unpublished data).
An effort to generate libraries expressing RNAi
a) Manipulation of mRNA Levels for C. neoformans genes is underway (T. Doering,
Technology for gene expression modulation has personal communication).
also been developed for Cryptococcus. Cryptococ- A third method of analyzing the contribution of
cal gene expression can be post-transcriptionally specific genes to virulence or vegetative growth is to
modulated by two methods: antisense repression modulate their expression by placing an inducible
and RNA interference. RNA antisense technology promoter in front of the protein-coding sequence.
relies on the ability of RNA that is complementary Use of an inducible promoter requires that the start
to a target gene’s mRNA to bind to that message site for translation of the protein is known, and
and render it incompetent for translation and tar- that the locations of genes adjacent to the gene of
get it for degradation. The expression of the cal- interest are known so that the normal regulation of
cineurin A gene (CNA1) from var. neoformans and neighboring genes is unaffected. The gene encod-
the laccase (LAC1) gene from var. grubii have been ing topoisomerase was placed under the control of
repressed using antisense RNA to each message the GAL7 promoter, and although the expression
(Gorlach et al. 2002). In this system, cDNA for each was reduced, the cells were still viable (Del Poeta
gene was cloned in an antisense orientation in a vec- et al. 1999b). However, with a more tightly regulated
tor containing the inducible promoter GAL7. This promoter, the thioredoxin reductase gene has been
construct was used to transform the corresponding shown to be essential in C. neoformans (Missall and
variety of C. neoformans. When the transformed Lodge 2005a).
cells were grown on medium containing galactose,
mRNA levels of both genes were significantly re-
duced, producing the expected phenotypes of re- b) Large-Scale Gene Deletion
duced melanin production (LAC1) and a growth In comparison with another common fungal
defect at 37 ◦ C (CNA1). pathogen, Candida albicans, which has two alleles
Gene expression can also be modified by RNA of each gene, C. neoformans is haploid and this
interference (RNAi). RNAi relies on an organ- facilitates targeted gene disruptions because
ism’s ability to degrade double-stranded RNA there is only one allele of each gene. In order to
(dsRNA). In RNAi competent organisms, dsRNA perform reliable gene deletion experiments, the
is specifically degraded into short 21–25 bp frag- disrupting DNA must be specifically targeted to
ments. Any message that is homologous to these the appropriate gene. Targeted gene disruption
fragments is targeted for degradation (reviewed can be performed at high frequencies in the
in Hutvagner and Zamore 2002). RNAi has been yeast Saccharomyces cerevisiae by homologous
used to modulate the expression of the CAP59 and recombination (HR). However, the efficiency of
ADE2 genes in variety neoformans (Liu et al. 2002). homologous recombination in C. neoformans is
Inverted repeats of portions of the CAP59 or ADE2 significantly less than that seen in S. cerevisiae.
genes were cloned into a vector separated from The reported frequency of homologous recom-
each other by an unrelated sequence. Expression bination in C. neoformans varies considerably due
of this sequence was driven by the promoter of to some of the same factors as those listed above for
the constitutively expressed cryptococcal actin variation in transformation efficiencies, i.e., the re-
(ACT) gene. Cells with the appropriate mutant cipient strain, the origin of the transforming DNA,
phenotypes were recovered after transformation the transforming vector, and the transformation
with the RNAi constructs. Analysis of these cells procedure. Published HR frequencies vary from
indicated that mRNA expression for each targeted 0.008 to 50% (Salas et al. 1996; Fox et al. 2001), but
gene was significantly reduced compared to wild because of the various methodologies used in pub-
type. The ability to specifically and temporally lished gene replacement experiments, direct com-
modulate gene expression in Cryptococcus will parison of the HR frequencies may lead to incorrect
allow us to identify genes that are required for the inferences. It is clear that higher rates of homol-
ogous recombination are obtained using biolistic As demonstrated in S. cerevisiae, systematic

transformation. HR rates in var. neoformans us- gene deletions have been a valuable tool for analysis
ing electroporation vary from 0.008 to 0.1%, but of gene function (Winzeler et al. 1999; Giaever et al.
rates using biolistic transformation in var. neofor- 2002). Gene deletion experiments are most infor-
mans vary from 1 to 87% (Davidson et al. 2000, mative when the boundaries of the protein-coding
2002). sequence of the gene in question are known, so that
The effects of varying the parameters of the entire protein can be eliminated without delet-
a homologous recombination experiment are ing genetic information of the surrounding genes.
largely unknown. The minimum flanking se- Deletions also require the development of efficient
quence requirement for efficient homologous protocols for transformation, homologous recom-
recombination in serotype A has been investigated bination, and screening. In a haploid organism,
(Nelson et al. 2003). A systematic homologous gene deletions cannot be used to analyze the func-
recombination experiment using the selectable tion of genes essential for vegetative growth. How-
marker for hygromycin B resistance was per- ever, a stable diploid strain has been made for var.
formed. In these experiments two independent neoformans (Sia et al. 2000), and this has proved
loci, LAC1 and CAP59, were chosen for disruption. useful for the analysis of essential genes.
The CAP59 gene is involved with the production A large set of gene deletions is being generated
of the polysaccharide capsule and cap59 mutants by the Madhani, Janbon, Heitman, Lodge and other
have a rough and dry colonial phenotype, whereas laboratories, with the long-term goal of deleting ev-
lac1 mutants are unable to produce melanin when ery nonessential gene in the serotype A strain H99.
grown on L-dopa medium. Five constructs of Each of the deletions is being tagged with a unique
each disruption vector were created with 400, DNA sequence so that pools of mutants can be ana-
300, 200, 100, and 50 bp of flanking sequence on lyzed in vivo, similar to signature tagged mutagene-
each side of the selectable marker. In addition, sis strategies (Hensel et al. 1995; Nelson et al. 2001).
two asymmetric disruption vectors were created. In vitro and in vivo analysis of these gene deletions
Constructs with 50 bp or less of flanking sequence have already revealed that the PKC1 pathway is im-
had an HR frequency of less than 1% of the total portant for cell integrity and for virulence, and that
transformants whereas constructs with flanking the deletion of a homolog of SSD1 makes C. neofor-
sequences longer than 200 bp had HR frequencies mans more virulent (Gerik and Lodge, unpublished
greater than 8%. If unstable transformants were data).
removed from the total transformants, the HR
frequency of these constructs varied between 10
and 75%. Taken together, these data indicate that c) Genome-Wide Insertional Mutagenesis
C. neoformans has a relatively efficient HR system In addition to targeted gene disruption, random
that can be relied upon for the efficient creation of insertional mutagenesis strategies have been used
targeted gene disruptions. to disrupt genes in C. neoformans. A major advan-
A system has been devised for efficiently gen- tage of insertional mutagenesis is that insertions
erating constructs for use in targeted gene disrup- in essential genes can be obtained if they alter
tion using homologous recombination (Davidson expression or only partially abrogate function.
et al. 2002). In this system, homologous sequences Also, the construction of libraries of insertion
from C. neoformans var. neoformans and var. gru- mutants is relatively fast. One disadvantage can
bii were PCR amplified from genomic DNA using be the difficulty in cloning the flanking sequence
primers with short (35–42 bp) overlaps to a se- around the inserted DNA. C. neoformans can
lectable marker cassette at their 5 ends. These PCR insert multiple copies of the transforming DNA in
products were then “linked” together with the se- tandem, making cloning the flanking sequences
lectable marker cassette separating the two flank- difficult and time-consuming, although use of
ing sequences in a third PCR reaction. This prod- linear vectors has been more successful.
uct was then used to disrupt the genomic copy of A form of random mutagenesis has been used
the target gene. Because this method is not depen- to identify var. grubii mutants with altered viru-
dent on the cloning of the target locus to create lence in a mouse model of disseminated disease
the knockout vector, it is providing a starting point (Nelson et al. 2001). In this study, biolistic transfor-
for the high-throughput generation of gene dele- mation circular plasmids containing signature tag
tions. sequences (Hensel et al. 1995) were used to produce
random insertion mutants. Signature tagged mu- produced tandem arrays of plasmid, making
tagenesis (STM; Fig. 13.3) uses a series of uniquely identification of the flanking sequences difficult.
tagged plasmid vectors to mutagenize a target Random insertion mutagenesis was also used
genome, producing libraries of mutants. The to isolate genes associated with the production of
strength of STM is the ability to pool mutants, each laccase in var. grubii (Erickson et al. 2001; Zhu and
with a distinct tag, and analyze them simultane- Williamson 2003). In this study, a component of
ously for any observable phenotype. Comparisons the vesicular protein pump, Vph1p, and a CLC-type
of a pool grown in vitro to the same pool following chloride channel were identified. The inactivation
growth in the mouse will identify mutants that of the VPH1 gene by insertional mutagenesis re-
were eliminated in vivo. The power of STM is sulted in a reduced virulence phenotype character-
the ability to screen many individual mutants in ized by reduced production of the major virulence
a single mouse, thus increasing the efficiency of factors capsule and laccase as well as urease.
identifying rare mutants. In this study, several avir- Newer methods for transformation and the
ulent mutants and one hyper-virulent mutant were availability of genome sequence are improving the
identified. However, the method of transformation recovery of inserted sequences and the rapidity
with which the insertion sites are identified.
Mutagenesis using A. tumafaciens-mediated
transformation has been highly successful for
generating and recovering insertional mutants
(Idnurm et al. 2004). There are major advantages
of A. tumafaciens-mediated transformation.
1. All transformants are stable and have DNA in-
tegrated into the genome, rather than being
present as unstable, extrachromosomal copies
of the transformation vector.
2. The ends of the insertion are defined by T-DNA
borders.
3. Single copies of the transforming DNA are in-
tegrated into the genome.
These advantages reduce screening and allow for
easy cloning of the insertion. The limited data
available at this time suggest that insertion of the
T-DNA may not be completely random, as there
may be a slight sequence bias toward insertion
into promoter regions. Also, the phenotype was
only linked to the inserted marker in two of three
Fig. 13.3. Signature tagged mutagenesis and analysis of mutants tested.
pooled mutants. A unique signature tag (ST) oligonu- A major improvement in the analysis of
cleotide, (NK)20, is synthesized and incorporated into random insertions is the recent availability of
a plasmid vector. Each “tagged” vector is used to transform mating pairs for serotype A. This has permitted
cells producing pools of uniquely tagged mutants. Single
mutants from each pool (96 total) are assembled into rapid analysis to determine whether the marker is
an input pool. Tags present in the input pool are PCR linked to the phenotype, enabling one to confirm
amplified, labeled and used to probe a filter containing that the phenotype is caused by the insertion.
dot blots of each tag, forming the preinoculum filter. A comparison of a limited set of mutants generated
The pooled organisms are injected into a mouse and the by biolistic transformation of plasmid DNA and A.
infection allowed to proceed. Organisms are recovered
from tissue and the tags are amplified using common tumafaciens-mediated transformation suggested
primers, labeled and used to probe a duplicate filter that the rate of generating unlinked phenotypes
(postinoculum filter). Tags missing from the pool are was similar (Idnurm et al. 2004).
identified by a loss of signal on the blot. Mutants that over Screening 590 biolistically transformed
proliferate can also be identified by having a much stronger
signal compared to the input blot. Mutants that did not mutants for defects in melanin biosynthesis,
have a change in their virulence have a similar signal to sensitivity to NO and temperature sensitivity
that of the input blot produced four mutants with defects in melanin
formation, two that were unable to grow at 37 ◦ C, ing is based on abundance, and so typically, only
and two mutants that were hypersensitive to NO the more abundant cellular proteins are analyzed.
(Idnurm et al. 2004). Of these eight, five had Newer methods hold promise for analyzing differ-
phenotypes that were linked to the selectable ential expression of proteins. The isotope coded
marker. The mutants with melanin defects had affinity tag (ICAT) method covalently attaches tags
an insertion into a predicted gene of unknown to specific amino acids (e.g., cysteines). Two dif-
function or into a mitogen activated kinase gene ferent tags are used, so that two different protein
(MPK2). The NO-hypersensitive mutants had lysates can be labeled. The proteins are cleaved into
insertions into the FHB1 gene, which encodes peptides by a protease, and the labeled peptides
a flavohemoglobin shown to be important for purified. The two samples are mixed together and
resistance to NO stress. The temperature-sensitive separated by liquid chromatography. Peptides that
mutant had an insertion between a predicted are present in a different ratio than most of the
protein and a helicase. Screening 576 mutants gen- other peptides are chosen for sequencing.
erated by Agrobacterium-mediated transformation Peptide mass fingerprinting or internal se-
resulted in the isolation of three melanin-defective quencing can be used to identify the protein. For
strains. Two of these mutants had phenotypes that identification of C. neoformans proteins, both of
co-segregated with the selectable marker. One these methods require a high-quality database of
mutant had an insertion in the promoter region of predicted proteins. Peptide mass fingerprinting
the laccase gene, and the other in a voltage-gated uses the molecular mass of tryptic peptides from
chloride channel. the protein in question and matches them to
a database of proteins. With highly conserved
proteins, a heterologous database could provide
B. Proteomics a good match, but with divergent or unique
proteins, other databases would not be useful.
Direct analysis of protein expression levels is com- A database generated from translating all of the
plementary to the analysis of mRNA expression lev- open reading frames in the C. neoformans genome
els. For studies of C. neoformans virulence, the dis- was also reasonably useful, but due to the large
covery of proteins that are regulated during patho- number of introns, it failed to match many proteins
genesis will be important. Many genes are regulated that were present in the new annotated database
by transcriptional control, but others are regulated (Pusateri and Lodge, unpublished data).
by translational control, protein degradation or by Proteomics has been used extensively in other
regulation of activities by post-translational mod- fungal systems to identify cell wall proteins, to iden-
ifications. These would not be detected by tran- tify response to drugs, and to identify response to
scriptome analysis. It is clear that there is not al- stresses. To date, only one study has been published
ways a correlation between mRNA abundance and using proteomics in C. neoformans, but with the
protein abundance (Gygi et al. 1999). However, the creation of a well-annotated set of predicted pro-
use of proteomics to determine changes in protein teins, proteomic studies are likely to become more
expression in vivo can be problematic due to con- common.
tamination by host proteins. A proteomic comparison of C. neoformans
Techniques such as two-dimensional (2D) gel strain H99 grown at 25 vs. 37 ◦ C using 2D gel
electrophoresis or liquid chromatography that can analysis revealed that two proteins were highly up-
separate and quantitate proteins can be used to regulated at 37 ◦ C (Missall et al. 2004b; Fig. 13.4).
determine which proteins are regulated under spe- Both of these proteins were identified as thiol
cific conditions. There are advantages and disad- peroxidases, proteins that have been shown to
vantages to each system. 2D gels are relatively sim- be important for resistance to oxidative stress
ple to run and have been the mainstay of pro- in other systems. The deletion of both genes, as
teomic analysis, but the types of proteins that are well as a third thiol peroxidase that was identified
analyzed can be limited on 2D gels. Proteins with through a genomic search, demonstrated that only
high or low molecular weights, or proteins with one of these, Tsa1, had a role in protection against
pIs of less than 3 or greater than 10 can be diffi- exogenous oxidative stresses and in virulence.
cult to resolve. Different solubilization procedures The tsa1∆ strain also was sensitive to nitrosative
have helped with separation of membrane or cell stress. In addition, 2D gels of the lysates from the
wall proteins. The detection of proteins by stain- deletion strains demonstrated conclusively that
tributions to the process under study. Because of

the tools and genomic information that are avail-
able, C. neoformans is and will continue to be a good
model for fungal pathogenesis. Although some fea-
tures of C. neoformans are unique, many will be
shared with other fungi, and so many of the lessons
learned while examining the mechanisms of C. ne-
oformans pathogenesis will be applicable to other,
less tractable systems.
Fig. 13.4. Comparison of lysates from C. neoformans grown Acknowledgements. The genome project of Cryptococcus
at 25 vs. 37 ◦ C. The two thiol peroxidases that are up- neoformans has been a community effort, with many
regulated at higher temperature are indicated by arrows genome centers contributing to the work, including the
Institute for Genome Research, the Stanford Genome and
Technology Center, Duke University Center for Genome
Technology, the Broad Institute, the University of British
Columbia, and University of Oklahoma. Many thanks to
the correct identification of the protein spots had Joe Heitman, John Perfect, Juneann Murphy, Jim Kronstad,
been made – the spots that had been identified as June Kwon-Chung, Fred Dietrich, Richard Hyman, Brendan
Tsa1 and Tsa3 were absent in the lysates from the Loftus, and James Galagan for their major contributions to
appropriate deletion strain. Proteomic analyses the genome project. Research in J.K. Lodge’s laboratory is
of the response to oxidative and nitrosative stress supported by NIH grants R01 AI50184 and R01 AI051209.
is ongoing (T. Missall and J. Lodge, unpublished
data).
References
Abadi J, Nachman S, Kressel AB, Pirofski L (1999) Crypto-
VIII. Conclusions and Future Impact coccosis in children with AIDS. Clin Infect Dis 28:309–
313
The Cryptococcus neoformans genome project has Aberg JA, Price RW, Heeren DM, Bredt B (2002) A pi-
lot study of the discontinuation of antifungal therapy
already had a major impact on research on the for disseminated cryptococcal disease in patients with
virulence of this important fungal pathogen, but acquired immunodeficiency syndrome, following im-
we have only seen the tip of the iceberg. We have munologic response to antiretroviral therapy. J Infect
fairly complete sequence datasets on five different Dis 185:1179–1182
Akhter S, McDade HC, Gorlach JM, Heinrich G, Cox GM,
genomes, and an outstanding set of annotations Perfect JR (2003) Role of alternative oxidase gene in
based on multiple gene finders and a high level pathogenesis of Cryptococcus neoformans. Infect Im-
of EST coverage. Investigators have begun taking mun 71:5794–5802
advantage of this information by comparing the Alspaugh JA, Perfect JR, Heitman J (1997) Cryptococcus
JEC21 genome with the B3501 genome. Tools for neoformans mating and virulence are regulated by the
G-protein alpha subunit GPA1 and cAMP. Genes Dev
molecular analysis are in place including transfor- 11:3206–3217
mation, multiple positive selectable markers, ho- Alspaugh JA, Pukkila-Worley R, Harashima T, Cavallo LM,
mologous recombination, regulatable promoters, Funnell D, Cox GM, Perfect JR, Kronstad JW, Heit-
and isogenic mating pairs. Studies of global gene man J (2002) Adenylyl cyclase functions downstream
of the Galpha protein Gpa1 and controls mating and
expression analysis have begun, a systematic set of pathogenicity of Cryptococcus neoformans. Eukaryot
gene deletions has been initiated, comprehensive Cell 1:75–84
sets of random insertion mutants are being con- Aoki FH, Imai T, Tanaka R, Mikami Y, Taguchi H,
structed, and databases are ready for analysis of Nishimura NF, Nishimura K, Miyaji M, Schreiber AZ,
proteomic data. Within a year or two, there will be Branchini ML (1999) New PCR primer pairs specific
for Cryptococcus neoformans serotype A or B pre-
comparisons of the different serotypes that have pared on the basis of random amplified polymorphic
different epidemiology and environmental niches. DNA fingerprint pattern analyses. J Clin Microbiol
Microarrays will be available for the analysis of 37:315–320
growth in an animal or of responses to specific Apidianakis Y, Rahme LG, Heitman J, Ausubel FM, Calder-
wood SB, Mylonakis E (2004) Challenge of Drosophila
stimuli or genotypes. Once specific genes are iden- melanogaster with Cryptococcus neoformans and role
tified by microarray or proteomic analysis, well- of the innate immune response. Eukaryot Cell 3:413–
defined mutants will be available to test their con- 419
Bach MC, Sahyoun A, Adler JL, Schlesinger RM, Breman J, Chen LC, Goldman DL, Doering TL, Pirofski A, Casadevall
Madras P, P’eng F, Monaco AP (1973) High incidence A (1999) Antibody response to Cryptococcus neofor-
of fungus infections in renal transplantation patients mans proteins in rodents and humans. Infect Immun
treated with antilymphocyte and conventional 67:2218–2224
immunosuppression. Transplant Proc 5:549–553 Cherniak R, Sundstrom JB (1994) Polysaccharide antigens
Bar-Peled M, Griffith CL, Doering TL (2001) Functional of the capsule of Cryptococcus neoformans. Infect Im-
cloning and characterization of a UDP-glucuronic acid mun 62:1507–1512
decarboxylase: the pathogenic fungus Cryptococcus Cherniak R, Reiss E, Slodki ME, Plattner RD, Blumer SO
neoformans elucidates UDP-xylose synthesis. Proc (1980) Structure and antigenic activity of the capsular
Natl Acad Sci USA 98:12003–12008 polysaccharide of Cryptococcus neoformans serotype
Bhattacharjee AK, Bennett JE, Glaudemans CP (1984) Cap- A. Mol Immunol 17:1025–1032
sular polysaccharides of Cryptococcus neoformans. Cherniak R, Jones RG, Reiss E (1988) Structure determina-
Rev Infect Dis 6:619–624 tion of Cryptococcus neoformans serotype A-variant
Biondo C, Beninati C, Delfino D, Oggioni M, Mancuso G, glucuronoxylomannan by 13C-n.m.r. spectroscopy.
Midiri A, Bombaci M, Tomaselli G, Teti G (2002) Identi- Carbohydr Res 172:113–138
fication and cloning of a cryptococcal deacetylase that Cherniak R, Morris LC, Turner SH (1992) Glucuronoxy-
produces protective immune responses. Infect Immun lomannan of Cryptococcus neoformans serotype D:
70:2383–2391 structural analysis by gas-liquid chromatography-
Blackstock R, Murphy JW (1997) Secretion of the C3 com- mass spectrometry and by 13C-nuclear magnetic
ponent of complement by peritoneal cells cultured resonance spectroscopy. Carbohydr Res 223:263–
with encapsulated Cryptococcus neoformans. Infect 269
Immun 65:4114–4121 Chuck SL, Sande MA (1989) Infections with Cryptococcus
Bose I, Reese AJ, Ory JJ, Janbon G, Doering TL (2003) A yeast neoformans in the acquired immunodeficiency syn-
under cover: the capsule of Cryptococcus neoformans. drome. N Engl J Med 321:794–799
Eukaryot Cell 2:655–663 Clarke DL, Woodlee GL, McClelland CM, Seymour TS,
Bozzette S, Larsen RA, Chiu J, Leal MA, Jacobsen J, Roth- Wickes BL (2001) The Cryptococcus neoformans
man P, Robinson P, Gilbert G, McCutchan JA, Tilles JG STE11alpha gene is similar to other fungal mitogen-
et al. (1991) A placebo-controlled trial of maintenance activated protein kinase kinase kinase (MAPKKK)
therapy with fluconazole after treatment of crypto- genes but is mating type specific. Mol Microbiol
coccal meningitis in the acquired immunodeficiency 40:200–213
syndrome. N Engl J Med 324:580–584 Coenjaerts FE, Walenkamp AM, Mwinzi PN, Schar-
Cameron ML, Bartlett JA, Gallis HA, Waskin HA (1991) ringa J, Dekker HA, Van Strijp AG, Cherniak R,
Manifestations of pulmonary cryptococcosis in pa- HoepelmanAI (2001) Potent inhibition of neutrophil
tients with acquired immunodeficiency syndrome. Rev migration by cryptococcal mannoprotein-4-induced
Infect Dis 13:64–67 desensitization. J Immunol 167:3988–3995
Casadevall A, Perfect JR (1998) Cryptococcus neoformans. Cox GM, Rude TH, Dykstra CC, Perfect JR (1995) The actin
ASM Press, Washington, DC gene from Cryptococcus neoformans: structure and
Chang YC, Kwon-Chung KJ (1994) Complementation of phylogenetic analysis. J Med Vet Mycol 33:261–266
a capsule-deficient mutation of Cryptococcus neofor- Cox GM, Toffaletti DL, Perfect JR (1996) Dominant selection
mans restores its virulence. Mol Cell Biol 14:4912– system for use in Cryptococcus neoformans. J Med Vet
4919 Mycol 34:385–391
Chang YC, Kwon-Chung KJ (1998) Isolation of the third Cox GM, Mukherjee J, Cole GT, Casadevall A, Perfect JR
capsule-associated gene, CAP60, required for viru- (2000) Urease as a virulence factor in experimental
lence in Cryptococcus neoformans. Infect Immun cryptococcosis. Infect Immun 68:443–448
66:2230–2236 Cox GM, McDade HC, Chen SC, Tucker SC, Gottfredsson M,
Chang YC, Kwon-Chung KJ (1999) Isolation, characteri- Wright LC, Sorrell TC, Leidich SD, Casadevall A, Ghan-
zation, and localization of a capsule-associated gene, noum MA et al. (2001) Extracellular phospholipase
CAP10, of Cryptococcus neoformans. J Bacteriol activity is a virulence factor for Cryptococcus neofor-
181:5636–5643 mans. Mol Microbiol 39:166–175
Chang YC, Penoyer LA, Kwon-Chung KJ (1996) The second Cox GM, Harrison TS, McDade HC, Taborda CP, Heinrich G,
capsule gene of Cryptococcus neoformans, CAP64, is Casadevall A, Perfect JR (2003) Superoxide dismutase
essential for virulence. Infect Immun 64:1977–1983 influences the virulence of Cryptococcus neoformans
Chang YC, Wickes BL, Miller GF, Penoyer LA, Kwon-Chung by affecting growth within macrophages. Infect Im-
KF (2000) Cryptococcus neoformans STE12alpha reg- mun 71:173–180
ulates virulence but is not essential for mating. J Exp Cunha BA (2001a) Central nervous system infections in the
Med 191:871–882 compromised host: a diagnostic approach. Infect Dis
Chen SC, Muller M, Zhou JZ, Wright LC, Sorrell TC (1997a) Clin N Am 15:567–590
Phospholipase activity in Cryptococcus neoformans: Cunha BA (2001b) Pneumonias in the compromised host.
a new virulence factor? J Infect Dis 175:414–420 Infect Dis Clin N Am 15:591–612
Chen SC, Wright LC, Santangelo RT, Muller M, Moran VR, Currie BP, Freundlich LF, Casadevall A (1994) Restriction
Kuchel PW, Sorrell TC (1997b) Identification of extra- fragment length polymorphism analysis of Cryptococ-
cellular phospholipase B, lysophospholipase, and acyl- cus neoformans isolates from environmental (pigeon
transferase produced by Cryptococcus neoformans. excreta) and clinical sources in New York City. J Clin
Infect Immun 65:405–411 Microbiol 32:1188–1192
Davidson RC, Moore TD, Odom AR, Heitman J (2000) Char- Garcia-Rivera J, Chang YC, Kwon-Chung KJ, Casadevall A
acterization of the MFalpha pheromone of the human (2004) Cryptococcus neoformans CAP59 (or Cap59p)
fungal pathogen Cryptococcus neoformans. Mol Mi- is involved in the extracellular trafficking of capsular
crobiol 38:1017–1026 glucuronoxylomannan. Eukaryot Cell 3:385–392
Davidson RC, Blankenship JR, Kraus PR, de Jesus Giaever G, Chu AM, Ni L, Connelly C, Riles L et al. (2002)
Berrios M, Hull CM, D’Souza C, Wang P, Heitman J Functional profiling of the Saccharomyces cerevisiae
(2002) A PCR-based strategy to generate integrative genome. Nature 418:387–391
targeting alleles with large regions of homology. Goldman D, Lee SC, Casadevall A (1994) Pathogenesis of
Microbiology 148:2607–2615 pulmonary Cryptococcus neoformans infection in the
de Jesus-Berrios M, Liu L, Nussbaum JC, Cox GM, Stam- rat. Infect Immun 62:4755–4761
ler JS, Heitman J (2003) Enzymes that counteract ni- Goldman DL, Lee SC, Mednick AJ, Montella L, Casade-
trosative stress promote fungal virulence. Curr Biol vall A (2000) Persistent Cryptococcus neoformans pul-
13:1963–1968 monary infection in the rat is associated with intracel-
Del Poeta M, Toffaletti DL, Rude TH, Sparks SD, Heitman J, lular parasitism, decreased inducible nitric oxide syn-
Perfect JR (1999a) Cryptococcus neoformans differ- thase expression, and altered antibody responsiveness
ential gene expression detected in vitro and in vivo to cryptococcal polysaccharide. Infect Immun 68:832–
with green fluorescent protein. Infect Immun 67:1812– 838
1820 Goldman DL, Khine H, Abadi J, Lindenberg DJ, Pirofski L,
Del Poeta M, Toffaletti DL, Rude TH, Dykstra CC, Heit- Niang R, Casadevall A (2001) Serologic evidence
man J, Perfect JR (1999b) Topoisomerase I is essential for Cryptococcus neoformans infection in early
in Cryptococcus neoformans: role in pathobiology and childhood. Pediatrics 107:E66
as an antifungal target. Genetics 152:167–178 Gonzalez CE, Shetty D, Lewis LL, Mueller BU, Pizzo PA,
Doering TL, Nosanchuk JD, Roberts WK, Casadevall A Walsh TJ (1996) Cryptococcosis in human immunod-
(1999) Melanin as a potential cryptococcal defense eficiency virus-infected children. Pediatr Infect Dis J
against microbicidal proteins. Med Mycol 37:175–181 15:796–800
D’Souza CA, Alspaugh JA, Yue C, Harashima T, Cox GM, Per- Gorlach JM, McDade HC, Perfect JR, Cox GM (2002) Anti-
fect JR, Heitman J (2001) Cyclic AMP-dependent pro- sense repression in Cryptococcus neoformans as a lab-
tein kinase controls virulence of the fungal pathogen oratory tool and potential antifungal strategy. Micro-
Cryptococcus neoformans. Mol Cell Biol 21:3179–3191 biology 148:213–219
D’Souza CA, Hagen F, Boekhout T, Cox GM, Heitman J Granger DL, Perfect JR, Durack DT (1985) Virulence of
(2004) Investigation of the basis of virulence in Cryptococcus neoformans. Regulation of capsule syn-
serotype A strains of Cryptococcus neoformans thesis by carbon dioxide. J Clin Invest 76:508–516
from apparently immunocompetent individuals. Curr Graybill JR, Sobel J, Saag MS, van der Horst C, Powderly WG,
Genet 46:92–102 Cloud G, Riaser L, Hamill R, Dismukes WE (2000)
Edman JC, Kwon-Chung KJ (1990) Isolation of the URA5 Diagnosis and management of increased intracranial
gene from Cryptococcus neoformans var. neoformans pressure in patients with AIDS and cryptococcal
and its use as a selective marker for transformation. meningitis. Clin Infect Dis 30:47–54
Mol Cell Biol 10:4538–4544 Grossi P, Farina C, Fiocchi R, Dalla Gasperina D (2000)
Eng RH, Bishburg E, Smith SM, Kapila R (1986) Crypto- Prevalence and outcome of invasive fungal infections
coccal infections in patients with acquired immune in 1963 thoracic organ transplant recipients: a mul-
deficiency syndrome. Am J Med 81:19–23 ticenter retrospective study. Italian Study Group of
Erickson T, Liu L, Gueyikian A, Zhu X, Gibbons J, Fungal Infections in Thoracic Organ Transplant Re-
Williamson PR (2001) Multiple virulence factors of cipients. Transplantation 70:112–116
Cryptococcus neoformans are dependent on VPH1. Gygi SP, Rochon Y, Franza BR, Aebersold R (1999) Correla-
Mol Microbiol 42:1121–1131 tion between protein and mRNA abundance in yeast.
Feldmesser M, Kress Y, Novikoff P, Casadevall A (2000) Mol Cell Biol 19:1720–1730
Cryptococcus neoformans is a facultative intracellu- Hajjeh RA, Conn LA, Stephens DS, Baughman W, Hamill R,
lar pathogen in murine pulmonary infection. Infect Graviss E, Pappas PG, Thomas C, Reingold A, Rothrock
Immun 68:4225–4237 G et al. (1999) Cryptococcosis: population-based mul-
Feldmesser M, Tucker S, Casadevall A (2001) Intracellular tistate active surveillance and risk factors in human im-
parasitism of macrophages by Cryptococcus neofor- munodeficiency virus-infected persons. Cryptococcal
mans. Trends Microbiol 9:273–278 Active Surveillance Group. J Infect Dis 179:449–454
Fox DS, Cruz MC, Sia RA, Ke H, Cox GM, Cardenas ME, Heitman J, Casadevall A, Lodge JK, Perfect JR (1999)
Heitman J (2001) Calcineurin regulatory subunit is es- The Cryptococcus neoformans genome sequencing
sential for virulence and mediates interactions with project. Mycopathologia 148:1–7
FKBP12-FK506 in Cryptococcus neoformans. Mol Mi- Hensel M, Shea JE, Gleeson C, Jones MD, Dalton E,
crobiol 39:835–849 Holden DW (1995) Simultaneous identification of
Franzot SP, Salkin IF, Casadevall A (1999) Cryptococcus ne- bacterial virulence genes by negative selection.
oformans var. grubii: separate varietal status for Cryp- Science 269:400–403
tococcus neoformans serotype A isolates. J Clin Mi- Hicks JK, D’Souza CA, Cox GM, Heitman J (2004) Cyclic
crobiol 37:838–840 AMP-dependent protein kinase catalytic subunits have
Fujihara H, Kagaya K, Fukazawa Y (1997) Anti-chemotactic divergent roles in virulence factor production in two
activity of capsular polysaccharide of Cryptococcus varieties of the fungal pathogen Cryptococcus neofor-
neoformans in vitro. Microbiol Immunol 41:657–664 mans. Eukaryot Cell 3:14–26
Hospenthal DR, Bennett JE (2000) Persistence of crypto- Korth H, Pulverer G (1971) Pigment formation for differen-
coccomas on neuroimaging. Clin Infect Dis 31:1303– tiating Cryptococcus neoformans from Candida albi-
1306 cans. Appl Microbiol 21:541–542
Hua J, Meyer JD, Lodge JK (2000) Development of positive Kovacs JA, Kovacs AA, Polis M, Wright WC, Gill VJ, Tua-
selectable markers for the fungal pathogen Cryptococ- zon CU, Gelmann EP, Lane HC, Longfield R, Overturf
cus neoformans. Clin Diagn Lab Immunol 7:125–128 G et al. (1985) Cryptococcosis in the acquired immun-
Huffnagle GB, Chen GH, Curtis JL, McDonald RA, odeficiency syndrome. Ann Intern Med 103:533–538
Strieter RM, Toews GB (1995) Down-regulation of the Kozel TR, Gotschlich EC (1982) The capsule of Cryptococ-
afferent phase of T cell-mediated pulmonary inflam- cus neoformans passively inhibits phagocytosis of the
mation and immunity by a high melanin-producing yeast by macrophages. J Immunol 129:1675–1680
strain of Cryptococcus neoformans. J Immunol Kozel TR, Gulley WF, Cazin J Jr (1977) Immune response to
155:3507–3516 Cryptococcus neoformans soluble polysaccharide: im-
Hull CM, Davidson RC, Heitman J (2002) Cell identity and munological unresponsiveness. Infect Immun 18:701–
sexual development in Cryptococcus neoformans are 707
controlled by the mating-type-specific homeodomain Kozel TR, Levitz SM, Dromer F, Gates MA, Thorkildson P,
protein Sxi1alpha. Genes Dev 16:3046–3060 Janbon G (2003) Antigenic and biological character-
Hull CM, Cox GM, Heitman J (2004) The alpha-specific istics of mutant strains of Cryptococcus neoformans
cell identity factor Sxi1alpha is not required for lacking capsular O acetylation or xylosyl side chains.
virulence of Cryptococcus neoformans. Infect Immun Infect Immun 71:2868–2875
72:3643–3645 Kraus PR, Heitman J (2003) Coping with stress: calmod-
Husain S, Wagener MM, Singh N (2001) Cryptococcus neo- ulin and calcineurin in model and pathogenic fungi.
formans infection in organ transplant recipients: vari- Biochem Biophys Res Commun 311:1151–1157
ables influencing clinical characteristics and outcome. Kraus PR, Fox DS, Cox GM, Heitman J (2003) The Cryp-
Emerg Infect Dis 7:375–381 tococcus neoformans MAP kinase Mpk1 regulates cell
Hutvagner G, Zamore PD (2002) RNAi: nature abhors integrity in response to antifungal drugs and loss of
a double-strand. Curr Opin Genet Dev 12:225–232 calcineurin function. Mol Microbiol 48:1377–1387
Ibrahim AS, Filler SG, Alcouloumre MS, Kozel TR, Kraus PR, Boily MJ, Giles SS, Stajich JE, Allen A, Cox GM,
Edwards JE, Ghannoum MA (1995) Adherence to Dietrich FS, Perfect JR, Heitman J (2004) Identification
and damage of endothelial cells by Cryptococcus of Cryptococcus neoformans temperature-regulated
neoformans in vitro: role of the capsule. Infect Immun genes with a genomic-DNA microarray. Eukaryot Cell
63:4368–4374 3:1249–1260
Idnurm A, Reedy JL, Nussbaum JC, Heitman J (2004) Kupfer DM, Drabenstot SD, Buchanan KL, Lai H, Zhu H,
Cryptococcus neoformans virulence gene discovery Dyer DW, Roe BA, Murphy JW (2004) Introns and splic-
through insertional mutagenesis. Eukaryot Cell ing elements of five diverse fungal organisms. Eukaryot
3:420–429 Cell 3:1088–1100
Ikeda R, Sugita T, Jacobson ES, Shinoda T (2002) Laccase Kwon-Chung KJ (1975) A new genus, filobasidiella, the
and melanization in clinically important Cryptococcus perfect state of Cryptococcus neoformans. Mycologia
species other than Cryptococcus neoformans. J Clin 67:1197–1200
Microbiol 40:1214–1218 Kwon-Chung KJ, Bennett JE (1978) Distribution of alpha
Jacobson ES, Emory HS (1991) Catecholamine uptake, and alpha mating types of Cryptococcus neoformans
melanization, and oxygen toxicity in Cryptococcus among natural and clinical isolates. Am J Epidemiol
neoformans. J Bacteriol 173:401–403 108:337–340
Jacobson ES, Tinnell SB (1993) Antioxidant function of fun- Kwon-Chung KJ, Rhodes JC (1986) Encapsulation and
gal melanin. J Bacteriol 175:7102–7104 melanin formation as indicators of virulence in
Janbon G, Himmelreich U, Moyrand F, Improvisi L, Dromer Cryptococcus neoformans. Infect Immun 51:218–
F (2001) Cas1p is a membrane protein necessary for the 223
O-acetylation of the Cryptococcus neoformans capsu- Kwon-Chung KJ, Polacheck I, Popkin TJ (1982) Melanin-
lar polysaccharide. Mol Microbiol 42:453–467 lacking mutants of Cryptococcus neoformans and
Kaplan MH, Rosen PP, Armstrong D (1977) Cryptococcosis their virulence for mice. J Bacteriol 150:1414–1421
in a cancer hospital: clinical and pathological correlates Kwon-Chung KJ, Edman JC, Wickes BL (1992) Genetic asso-
in forty-six patients. Cancer 39:2265–2274 ciation of mating types and virulence in Cryptococcus
Kappe R, Muller J (1991) Rapid clearance of Candida albi- neoformans. Infect Immun 60:602–605
cans mannan antigens by liver and spleen in contrast Lee SC, Casadevall A (1996) Polysaccharide antigen in brain
to prolonged circulation of Cryptococcus neoformans tissue of AIDS patients with cryptococcal meningitis.
antigens. J Clin Microbiol 29:1665–1669 Clin Infect Dis 23:194–195
Karos M, Chang YC, McClelland CM, Clarke DL, Fu J, Lendvai N, Casadevall A, Liang Z, Goldman DL, Mukherjee J,
Wickes BL, Kwon-Chung KJ (2000) Mapping of the Zuckier L (1998) Effect of immune mechanisms on the
Cryptococcus neoformans MATalpha locus: presence pharmacokinetics and organ distribution of crypto-
of mating type-specific mitogen-activated protein ki- coccal polysaccharide. J Infect Dis 177:1647–1659
nase cascade homologs. J Bacteriol 182:6222–6227 Lengeler KB, Davidson RC, D’Souza C, Harashima T,
Kontoyiannis DP, Peitsch WK, Reddy BT, Whimbey EE, Shen WC, Wang P, Pan X, Waugh M, Heitman J
Han XY, Bodey GP, Rolston KV (2001) Cryptococco- (2000a) Signal transduction cascades regulating
sis in patients with cancer. Clin Infect Dis 32:E145– fungal development and virulence. Microbiol Mol Biol
150 Rev 64:746–785
Lengeler KB, Wang P, Cox GM, Perfect JR, Heitman J (2000b) nun YA, Balish E et al. (2003) Identification of App1 as
Identification of the MATa mating-type locus of Cryp- a regulator of phagocytosis and virulence of Crypto-
tococcus neoformans reveals a serotype A MATa strain coccus neoformans. J Clin Invest 112:1080–1094
thought to have been extinct. Proc Natl Acad Sci USA Manfredi R, Moroni M, Mazzoni A, Nanetti A, Donati M,
97:14455–14460 Mastroianni A, Coronado OV, Chiodo F (1996) Isolated
Lengeler KB, Fox DS, Fraser JA, Allen A, Forrester K, Diet- detection of crytpcococcal polysaccharide antigen in
rich FS, Heitman J (2002) Mating-type locus of Cryp- cerebrospinal fluid samples from patients with AIDS.
tococcus neoformans: a step in the evolution of sex Clin Infect Dis 23:849–850
chromosomes. Eukaryot Cell 1:704–718 Marra RE, Huang JC, Fung E, Nielsen K, Heitman J,
Levitz SM, Harrison TS, Tabuni A, Liu X (1997) Chloro- Vilgalys R, Mitchell TG (2004) A genetic linkage map
quine induces human mononuclear phagocytes to in- of Cryptococcus neoformans variety neoformans
hibit and kill Cryptococcus neoformans by a mech- serotype D (Filobasidiella neoformans). Genetics
anism independent of iron deprivation. J Clin Invest 167:619–631
100:1640–1646 McDade HC, Cox GM (2001) A new dominant selectable
Levitz SM, Nong SH, Seetoo KF, Harrison TS, Speizer RA, marker for use in Cryptococcus neoformans. Med My-
Simons ER (1999) Cryptococcus neoformans resides col 39:151–154
in an acidic phagolysosome of human macrophages. Merkel GJ, Cunningham RK (1992) The interaction of Cryp-
Infect Immun 67:885–890 tococcus neoformans with primary rat lung cell cul-
Levitz SM, Nong S, Mansour MK, Huang C, Specht CA tures. J Med Vet Mycol 30:115–121
(2001) Molecular characterization of a mannoprotein Missall TA, Lodge JK (2005a) Thioredoxin reductase is es-
with homology to chitin deacetylases that stimulates T sential for viability in the fungal pathogen, Cryptococ-
cell responses to Cryptococcus neoformans. Proc Natl cus neoformans. Eukaryot Cell 4:487–489
Acad Sci USA 98:10422–10427 Missall TA, Lodge JK (2005b) Function of the thioredoxin
Lian T, Simmer MI, D’Souza CA, Steen BR, Zuyder- proteins in Cryptococcus neoformans during stress or
duyn SD, Jones SJ, Marra MA, Kronstad JW (2005) virulence and regulation by putative transcriptional
Iron-regulated transcription and capsule formation in modulators. Mol Microbiol 57:847–858
the fungal pathogen Cryptococcus neoformans. Mol Missall TA, Lodge JK, McEwen JE (2004a) Mechanisms of
Microbiol 55:1452–1472 resistance to oxidative and nitrosative stress: implica-
Liliang PC, Liang CL, Chang WN, Lu K, Lu CH (2002) Use tions for fungal survival in mammalian hosts. Eukaryot
of ventriculoperitoneal shunts to treat uncontrollable Cell 3:835–846
intracranial hypertension in patients who have crypto- Missall TA, Pusateri ME, Lodge JK (2004b) Thiol peroxidase
coccal meningitis without hydrocephalus. Clin Infect is critical for virulence and resistance to nitric oxide
Dis 34:e64–e68 and peroxide in the fungal pathogen, Cryptococcus
Lindsley MD, Hurst SF, Iqbal NJ, Morrison CJ (2001) Rapid neoformans. Mol Microbiol 51:1447–1458
identification of dimorphic and yeast-like fungal Missall TA, Moran JM, Corbett JA, Lodge JK (2005a) Distinct
pathogens using specific DNA probes. J Clin Microbiol stress responses of two functional laccases in Crypto-
39:3505–3511 coccus neoformans is revealed in the absence of the
Litvintseva AP, Marra RA, Nielsen K, Heitman J, Vilgalys R, thiol-specific antioxidant, Tsa1. Eukaryot Cell 4:202–
Mitchell TG (2003) Evidence of sexual recombination 208
among Cryptococcus neoformans serotype A isolates Missall TA, Cherry-Harris JF, Lodge JK (2005b) Two glu-
in sub-Saharan Africa. Eukaryot Cell 2:1162–1168 tathione peroxidases in the fungal pathogen, Crypto-
Liu L, Tewari RP, Williamson PR (1999a) Laccase protects coccus neoformans, are expressed in the presence of
Cryptococcus neoformans from antifungal activity of specific substrates. Microbiology 151:2573–2591
alveolar macrophages. Infect Immun 67:6034–6039 Mitchell DH, Sorrell TC, Allworth AM, Heath CH, McGre-
Liu L, Wakamatsu K, Ito S, Williamson PR (1999b) Cat- gor AR, Papanaoum K, Richards MJ, Gottlieb T (1995)
echolamine oxidative products, but not melanin, are Cryptococcal disease of the CNS in immunocompe-
produced by Cryptococcus neoformans during neu- tent hosts: influence of cryptococcal variety on clinical
ropathogenesis in mice. Infect Immun 67:108–112 manifestations and outcome. Clin Infect Dis 20:611–
Liu H, Cottrell TR, Pierini LM, Goldman WE, Doering TL 616
(2002) RNA interference in the pathogenic fungus Mody CH, Syme RM (1993) Effect of polysaccharide capsule
Cryptococcus neoformans. Genetics 160:463–470 and methods of preparation on human lymphocyte
Loftus B, Eula Fung E, Roncaglia P, Rowley D, Amedeo P, proliferation in response to Cryptococcus neoformans.
Bruno D, Vamathevan J, Miranda M, Anderson I, Fraser Infect Immun 61:464–469
JA et al. (2005) The genome and transcriptome of Monari C, Kozel TR, Bistoni F, Vecchiarelli A (2002) Modula-
Cryptococcus neoformans, a basidiomycetous fungal tion of C5aR expression on human neutrophils by en-
pathogen of humans. Science 307:1321–1324 capsulated and acapsular Cryptococcus neoformans.
Lorenz MC (2002) Genomic approaches to fungal Infect Immun 70:3363–3370
pathogenicity. Curr Opin Microbiol 5:372–378 Moosa MY, Coovadia YM (1997) Cryptococcal meningi-
Luberto C, Toffaletti DL, Wills EA, Tucker SC, Casadevall A, tis in Durban, South Africa: a comparison of clini-
Perfect JR, Hannun YA, Del Poeta M (2001) Roles cal features, laboratory findings and outcome for hu-
for inositol-phosphoryl ceramide synthase 1 (IPC1) in man immunodeficiency virus (HIV)-positive and HIV-
pathogenesis of C. neoformans. Genes Dev 15:201–202 negative patients. Clin Infect Dis 24:131–134
Luberto C, Martinez-Marino B, Taraskiewicz D, Bolanos B, Moyrand F, Klaproth B, Himmelreich U, Dromer F, Janbon G
Chitano P, Toffaletti DL, Cox GM, Perfect JR, Han- (2002) Isolation and characterization of capsule struc-
ture mutant strains of Cryptococcus neoformans. Mol induction and regulation of immunity to Cryptococcus
Microbiol. 2045:837–849 neoformans. Infect Immun 69:2808–2814
Mylonakis E, Ausubel FM, Perfect JR, Heitman J, Calder- Polacheck I, Hearing VJ, Kwon-Chung KJ (1982) Biochem-
wood SB (2002) Killing of Caenorhabditis elegans by ical studies of phenoloxidase and utilization of cate-
Cryptococcus neoformans as a model of yeast patho- cholamines in Cryptococcus neoformans. J Bacteriol
genesis. Proc Natl Acad Sci USA 99:15675–15680 150:1212–1220
Narasipura SD, Ault JG, Behr MJ, Chaturvedi V, Chaturvedi S Posteraro B, Sanguinetti M, Masucci L, Romano L,
(2003) Characterization of Cu,Zn superoxide dismu- Morace G, Fadda G (2000) Reverse cross blot hy-
tase (SOD1) gene knock-out mutant of Cryptococcus bridization assay for rapid detection of PCR-amplified
neoformans var. gattii: role in biology and virulence. DNA from Candida species, Cryptococcus neo-
Mol Microbiol 47:1681–1694 formans, and Saccharomyces cerevisiae in clinical
Neilson K, Cox GM, Wang P, Toffaletti DL, Perfect JR, Heit- samples. J Clin Microbiol 38:1609–1614
man J (2003) Sexual cycle of Cryptococcus neoformans Powderly WG (1996) Recent advances in the management
var. grubii and virulence of congenic a and alpha iso- of cryptococcal meningitis in patients with AIDS. Clin
lates. Infect Immun 71:4831–4841 Infect Dis 22:S119–123
Nelson RT, Hua J, Pryor B, Lodge JK (2001) Identification of Powderly WG, Saag MS, Cloud GA, Robinson P, Meyer RD,
virulence mutants of the fungal pathogen Cryptococ- Jacobson JM, Graybill JR, Sugar AM, McAuliffe VJ, Fol-
cus neoformans using signature-tagged mutagenesis. lansbee SE et al. (1992) A controlled trial of flucona-
Genetics 157:935–947 zole or amphotericin B to prevent relapse of crypto-
Nelson RT, Pryor BA, Lodge JK (2003) Sequence length re- coccal meningitis in patients with the acquired im-
quired for homologous recombination in Cryptococ- munodeficiency syndrome. The NIAID AIDS Clinical
cus neoformans. Fungal Genet Biol 38:1–9 Trials Group and Mycoses Study Group. N Engl J Med
Nosanchuk JD, Rudolph J, Rosas AL, Casadevall A (1999a) 326:793–798
Evidence that Cryptococcus neoformans is melanized Powderly WG, Cloud G, Dismukes WE, Saag MS (1994) Mea-
in pigeon excreta: implications for pathogenesis. Infect surement of cryptococcal antigen in serum and cere-
Immun 67:5477–5479 bralspinal fluid: value in the management of AIDS-
Nosanchuk JD, Valadon P, Feldmesser M, Casadevall A associated cryptococcal meningitis. Clin Infect Dis
(1999b) Melanization of Cryptococcus neoformans in 18:789–792
murine infection. Mol Cell Biol 19:745–750 Pukkila-Worley R, Alspaugh JA (2004) Cyclic AMP signaling
Nosanchuk JD, Rosas AL, Lee SC, Casadevall A (2000) in Cryptococcus neoformans. FEMS Yeast Res 4:361–
Melanisation of Cryptococcus neoformans in human 367
brain tissue. Lancet 355:2049–2050 Pukkila-Worley R, Gerrald QD, Kraus PR, Boily MJ,
Noverr MC, Phare SM, Toews GB, Coffey MJ, Huffnagle GB Davis MJ, Giles SS, Cox GM, Heitman J, Alspaugh JA
(2001) Pathogenic yeasts Cryptococcus neoformans (2005) Transcriptional network of multiple capsule
and Candida albicans produce immunomodulatory and melanin genes governed by the Cryptococcus
prostaglandins. Infect Immun 69:2957–2963 neoformans cyclic AMP cascade. Eukaryot Cell
Noverr MC, Cox GM, Perfect JR, Huffnagle GB (2003) Role 4:190–201
of PLB1 in pulmonary inflammation and cryptococcal Rabkin JM, Oroloff SL, Corless CL, Benner KG, Flora KD,
eicosanoid production. Infect Immun 71:1538–1547 Rosen HR, Olyaei AJ (2000) Association of fungal in-
Olszewski MA, Noverr MC, Chen GH, Toews GB, Cox GM, fection and increased mortality in liver transplant re-
Perfect JR, Huffnagle GB (2004) Urease expression by cipients. Am J Surg 179:426–430
Cryptococcus neoformans promotes microvascular se- Rappelli P, Are R, Casu G, Fiori PL, Cappuccinelli P, Aceti A
questration, thereby enhancing central nervous system (1998) Development of a nested PCR for detection of
invasion. Am J Pathol 164:1761–1771 Cryptococcus neoformans in cerebrospinal fluid. J Clin
Ory JJ, Griffith CL, Doering TL (2004) An efficiently regu- Microbiol 36:3438–3440
lated promoter system for Cryptococcus neoformans Reese AJ, Doering TL (2003) Cell wall alpha-1,3-glucan is
utilizing the CTR4 promoter. Yeast 21:919–926 required to anchor the Cryptococcus neoformans cap-
Pappas PG, Perfect JR, Cloud GA, Larsen RA, Pankey GA, sule. Mol Microbiol 50:1401–1409
Lancaster DJ, Henderson H, Kauffman CA, Haas DW, Retini C, Vecchiarelli A, Monari C, Bistoni F, Kozel TR
Saccente M et al. (2001) Cryptococcosis in human im- (1998) Encapsulation of Cryptococcus neoformans
munodeficiency virus-negative patients in the era of with glucuronoxylomannan inhibits the antigen-
effective azole therapy. Clin Infect Dis 33:690–699 presenting capacity of monocytes. Infect Immun
Perfect JR, Lang SD, Durack DT (1980) Chronic cryptococ- 66:664–669
cal meningitis: a new experimental model in rabbits. Rhodes JC, Polacheck I, Kwon-Chung KJ (1982) Phenoloxi-
Am J Pathol 101:177–194 dase activity and virulence in isogenic strains of Cryp-
Perfect JR, Magee BB, Magee PT (1989) Separation of chro- tococcus neoformans. Infect Immun 36:1175–1184
mosomes of Cryptococcus neoformans by pulsed field Richardson MD, White LJ, McKay IC, Shankland GS (1993)
gel electrophoresis. Infect Immun 57:2624–2627 Differential binding of acapsulate and encapsulated
Pfrommer GS, Dickens SM, Wilson MA, Young BJ, Kozel TR strains of Cryptococcus neoformans to human neu-
(1993) Accelerated decay of C3b to iC3b when C3b is trophils. J Med Vet Mycol 31:189–199
bound to the Cryptococcus neoformans capsule. Infect Rosas AL, Nosanchuk JD, Feldmesser M, Cox GM, Mc-
Immun 61:4360–4366 Dade HC, Casadevall A (2000) Synthesis of polymer-
Pietrella D, Cherniak R, Strappini C, Perito S, Mosci P, Bis- ized melanin by Cryptococcus neoformans in infected
toni F, Vecchiarelli A (2001) Role of mannoprotein in rodents. Infect Immun 68:2845–2853
Rude TH, Toffaletti DL, Cox GM, Perfect JR (2002) Rela- Sundstrom JB, Cherniak R (1993) T-cell-dependent and
tionship of the glyoxylate pathway to the pathogenesis T-cell-independent mechanisms of tolerance to
of Cryptococcus neoformans. Infect Immun 70:5684– glucuronoxylomannan of Cryptococcus neoformans
5694 serotype A. Infect Immun 61:1340–1345
Saag MS, Graybill RJ, Larsen RA, Pappas PG, Perfect JR, Syme RM, Bruno TF, Kozel TR, Mody CH (1999) The
Powderly WG, Sobel JD, Dismukes WE (2000) Prac- capsule of Cryptococcus neoformans reduces T-
tice guidelines for the management of cryptococcal lymphocyte proliferation by reducing phagocytosis,
disease. Infectious Diseases Society of America. Clin which can be restored with anticapsular antibody.
Infect Dis 30:710–718 Infect Immun 67:4620–4627
Saito Y, Osabe S, Kuno H, Kaji M, Oizumi K (1999) Rapid Tanner DC, Weinstein MP, Fedorciw B, Joho KL, Thorpe JJ,
diagnosis of cryptococcal meningitis by microscopic Reller L (1994) Comparison of commercial kits for
examination of centrifuged cerebrospinal fluid sedi- detection of cryptococcal antigen. J Clin Microbiol
ment. J Neurol Sci 164:72–75 32:1680–1684
Salas SD, Bennett JE, Kwon-Chung KJ, Perfect JR, Tenney AR, Brown RH, Vaske C, Lodge JK, Doering TL,
Williamson PR (1996) Effect of the laccase gene Brent MR (2004) Gene prediction and verification
CNLAC1, on virulence of Cryptococcus neoformans. in a compact genome with numerous small introns.
J Exp Med 184:377–386 Genome Res 14:2330–2335
Sax PE (2001) Opportunistic infections in HIV dis- Toffaletti DL, Rude TH, Johnston SA, Durack DT, Perfect JR
ease: down but not out. Infect Dis Clin N Am (1993) Gene transfer in Cryptococcus neoformans by
15:433–455 use of biolistic delivery of DNA. J Bacteriol 175:1405–
Schein JE, Tangen KL, Chiu R, Shin H, Lengeler KB, 1411
MacDonald WK, Bosdet I, Heitman J, Jones SJ, Tscharke RL, Lazera M, Chang YC, Wickes BL, Kwon-Chung
Marra MA et al. (2002) Physical maps for genome KJ (2003) Haploid fruiting in Cryptococcus neofor-
analysis of serotype A and D strains of the fungal mans is not mating type alpha-specific. Fungal Genet
pathogen Cryptococcus neoformans. Genome Res Biol 39:230–237
12:1445–1453 Tucker SC, Casadevall A (2002) Replication of Cryptococcus
Shaw CE, Kapica L (1972) Production of diagnostic pig- neoformans in macrophages is accompanied by phago-
ment by phenoloxidase activity of Cryptococcus neo- somal permeabilization and accumulation of vesicles
formans. Appl Microbiol 24:824–830 containing polysaccharide in the cytoplasm. Proc Natl
Sia RA, Lengeler KB, Heitman J (2000) Diploid strains Acad Sci USA 99:3165–3170
of the pathogenic basidiomycete Cryptococcus van Duin D, Casadevall A, Nosanchuk JD (2002) Melaniza-
neoformans are thermally dimorphic. Fungal Genet tion of Cryptococcus neoformans and Histoplasma
Biol 29:153–163 capsulatum reduces their susceptibilities to am-
Snydman DR (2001) Epidemiology of infections after solid- photericin B and caspofungin. Antimicrob Agents
organ transplantation. Clin Infect Dis 33 suppl 1:S5–S8 Chemother 46:3394–3400
Speed B, Dunt D (1995) Clinical and host differences be- Varma A, Kwon-Chung KJ (1999) Characterization of the
tween infections with the two varieties of Cryptococ- glyceraldehyde-3-phosphate dehydrogenase gene [cor-
cus neoformans. Clin Infect Dis 21:28–34 rection of glyceraldehyde-3-phosphate gene] and the
Steen BR, Lian T, Zuyderduyn S, MacDonald WK, Marra M, use of its promoter for heterologous expression in
Jones SJ, Kronstad JW (2002) Temperature-regulated Cryptococcus neoformans, a human pathogen. Gene
transcription in the pathogenic fungus Cryptococcus 232:155–163
neoformans. Genome Res 12:1386–1400 Vartivarian SE, Anaissie E, Cowart RE, Sprigg HA,
Steen BR, Zuyderduyn S, Toffaletti DL, Marra M, Jones SJ, Tingler MJ, Jacobson ES (1993) Regulation of cryp-
Perfect JR, Kronstad JW (2003) Cryptococcus neofor- tococcal capsular polysaccharide by iron. J Infect Dis
mans gene expression during experimental cryptococ- 167:186–190
cal meningitis. Eukaryot Cell 2:1336–1349 Vecchiarelli, A, Retini C, Casadevall A, Monari C,
Steenbergen JN, Shuman HA, Casadevall A (2001) Cryp- Pietrella D, Kozel TR (1998) Involvement of C3a
tococcus neoformans interactions with amoebae sug- and C5a in interleukin-8 secretion by human
gest an explanation for its virulence and intracellular polymorphonuclear cells in response to capsular
pathogenic strategy in macrophages. Proc Natl Acad material of Cryptococcus neoformans. Infect Immun
Sci USA 98:15245–15250 66:4324–4330
Steenbergen JN, Nosanchuk JD, Malliaris SD, Casadevall Viviani MA, Esposto MC, Cogliati M, Montagna MT, Wickes
A (2003) Cryptococcus neoformans virulence is BL (2001) Isolation of a Cryptococcus neoformans
enhanced after growth in the genetically malleable serotype A MATa strain from the Italian environment.
host Dictyostelium discoideum. Infect Immun Med Mycol 39:383–386
71:4862–4872 Viviani MA, Nikolova R, Esposto MC, Prinz G, Cogliati M
Stephen C, Lester S, Black W, Fyfe M, Raverty S (2002) (2003) First European case of serotype A MATa
Multispecies outbreak of cryptococcosis on southern Cryptococcus neoformans infection. Emerg Infect Dis
Vancouver Island, British Columbia. Can Vet J 43:792– 9:1179–1180
794 Wang Y, Casadevall A (1994) Decreased susceptibility of
Sudarshan SR, Davidson RC, Heitman J, Alspaugh JA (1999) melanized Cryptococcus neoformans to UV light. Appl
Molecular analysis of the Cryptococcus neoformans Environ Microbiol 60:3864–3866
ADE2 gene, a selectable marker for transformation and Wasser L, Talavera W (1987) Pulmonary cryptococcosis in
gene disruption. Fungal Genet Biol 27:36–48 AIDS. Chest 92:692–695
Wheat LJ, Smith EJ, Sathapatayavongs B, Batteiger B, Filo RS, Wilson MA, Kozel TR (1992) Contribution of antibody in
Leapman SB, French MV (1983) Histoplasmosis in re- normal human serum to early deposition of C3 onto
nal allograft recipients: two large urban outbreaks. encapsulated and nonencapsulated Cryptococcus ne-
Arch Intern Med 143:703–707 oformans. Infect Immun 60:754–761
White MH, Armstrong D (1994) Cryptococcosis. Infect Dis Wilson DE, Bennett JE, Bailey JW (1968) Serologic grouping
Clin N Am 8:383–398 of Cryptococcus neoformans. Proc Soc Exp Biol Med
White A, Cirrincione C, Blevins A, Armstrong D (1992) 127:820–823
Cryptococcal meningitis: outcome in patients with Winzeler EA, Shoemaker DD, Astromoff A, Liang H, An-
AIDS and patients with neoplastic disease. J Infect Dis derson K, Andre B, Bangham R, Benito R, Boeke JD,
165:960–963 Bussey H et al. (1999) Functional characterization of
Wickes BL, Edman JC (1995) The Cryptococcus neoformans the S. cerevisiae genome by gene deletion and parallel
GAL7 gene and its use as an inducible promoter. Mol analysis. Science 285:901–906
Microbiol 16:1099–1109 Yan Z, Li X, Xu J (2002) Geographic distribution of mating
Wickes BL, Moore TD, Kwon-Chung KJ (1994) Comparison type alleles of Cryptococcus neoformans in four areas
of the electrophoretic karyotypes and chromosomal of the United States. J Clin Microbiol 40:965–972
location of ten genes in the two varieties of Cryptococ- Yue C, Cavallo LM, Alspaugh JA, Wang P, Cox GM, Perfect JR,
cus neoformans. Microbiology 140:543–550 Heitman J (1999) The STE12alpha homolog is required
Wickes BL, Mayorga ME, Edman U, Edman JC (1996) Di- for haploid filamentation but largely dispensable for
morphism and haploid fruiting in Cryptococcus neo- mating and virulence in Cryptococcus neoformans.
formans: association with the alpha-mating type. Proc Genetics 153:1601–1615
Natl Acad Sci USA 93:7327–7331 Zhu X, Williamson PR (2003) A CLC-type chloride chan-
Williamson PR (1994) Biochemical and molecular char- nel gene is required for laccase activity and virulence
acterization of the diphenol oxidase of Cryptococ- in Cryptococcus neoformans. Mol Microbiol 50:1271–
cus neoformans: identification as a laccase. J Bacteriol 1281
176:656–664 Zhu X, Williamson PR (2004) Role of laccase in the biol-
Wills EA, Roberts IS, Del Poeta M, Rivera J, Casadevall A, ogy and virulence of Cryptococcus neoformans. FEMS
Cox GM, Perfect JR (2001) Identification and char- Yeast Res 5:1–10
acterization of the Cryptococcus neoformans phos- Zhu X, Gibbons J, Garcia-Rivera J, Casadevall A,
phomannose isomerase-encoding gene, MAN1, and Williamson PR (2001) Laccase of Cryptococcus
its impact on pathogenicity. Mol Microbiol 40:610– neoformans is a cell wall-associated virulence factor.
620 Infect Immun 69:5589–5596
Biosystematic Index
Candida glabrata, 215 Colletrotrichum gloeosporioides, 207

Colletotrichum lagenarium, 46
Aequoria victoriae, 36 Coniochaete tetraspora, 117
Agaricus bisporus, 118 Coprinus cinereus, 35, 36, 46, 117, 118, 123
Agrobacterium, 79 Coprinus domesticus, 118
Agrobacterium tumefaciens, 37, 232 Cryphonectria parasitica, 133
Anopheles gambiae, 38 Cryptococcus neoformans, 36, 41, 163, 237, 239, 240,
Arabidopsis, 68 242–256, 258, 259
Arabidopsis thaliana, 38, 128, 207 var. gattii, 242, 243, 245, 246, 250, 251
Ascomycetes, 54, 58 var. grubii, 242, 243, 245, 246, 250, 251, 253,
Ascomycota, 98 255, 256
Ashbya gossypii, 19–21, 23, 24, 29–31, 36 var. neoformans, 242, 243, 245, 246, 250, 251,
Aspergillus, 224, 230 253, 255, 256
Aspergillus awamori, 79
Aspergillus flavus, 80 Debaryomyces hansenii, 28, 36, 164–166
Aspergillus fumigatus, 41, 75, 116, 120, 121, 124, Dictyostelium discoideum, 115, 242
131, 163 Drosophila, 53, 55, 57, 58, 61, 62, 68, 71, 119, 137
Aspergillus nidulans, 36, 41, 46, 62, 75, 121–124, Drosophila melanogaster, 38, 206, 242
130–132, 189, 207, 208, 229
Aspergillus niger, 75 Encephalitozoon cuniculi, 36, 40
Aspergillus oryzae, 75 Escherichia coli, 13, 39, 40, 179, 186
Aspergillus sojae, 87
Fusarium, 58, 62
Basidiomycetes, 54 Fusarium graminearum, 36, 38, 41, 42, 46
Blastomyces dermatiditis, 222, 232 Fusarium oxysporum, 37, 46, 79, 207
Blastomyces dermatitidis, 225
Gonyaulax, 53
Blumeria graminis, 37, 42
Botrytis cinerea, 36, 41, 42, 46 Haemophilus influenzae, 39, 40
Heterobasidion annosum, 117
Caenorhabditis, 58 Histoplasma capsulatum, 163, 171, 221–233
Caenorhabditis elegans, 39, 57, 125, 242 var. dubosii, 227
Candida albicans, 19, 20, 36, 41, 119–121, 123, 124, var. farciminosum, 226
127, 128, 131, 133, 134, 148–151, 155–157,
163–179, 189, 191–199, 205–210, 214, 215 Kluyveromyces lactis, 19, 20, 23, 24, 28–31, 36, 126,
Candida dubliniensis, 164, 168, 172 164, 165
Candida famata, 164 Kluyveromyces waltii, 19, 20, 23, 24, 28–30, 36
Candida glabrata, 19–21, 23, 24, 28–30, 36, 164, 165,
205–210, 214, 215 Leptosphaeria, 62
Chromocrea, 62 Loculoascomyetes, 62
Coccidioides immitis, 36, 222
Cochliobolus, 58 Magnaporthe, 58, 62
Cochliobolus heterostrophus, 37, 46 Magnaporthe grisea, 35–38, 41, 42, 46, 117
268 Biosystematic Index
Mammalia, 98, 103, 104 Saccharomyces cerevisiae, 3–8, 11–13, 19–21, 23, 24,
Mucor racemosus, 123, 124, 126, 128, 132, 133 26, 28–31, 35, 36, 38, 41, 42, 77, 81, 98, 101,
Mus musculus, 38 102, 104–106, 115–128, 130–133, 135–137,
Mycobacterium tuberculosis, 169 147–149, 151, 155, 156, 164–170, 172–174,
Mycosphaerella graminicola, 37, 41, 42 179, 186–189, 191–193, 195–197, 205–211,
214, 215
Neurospora crassa, 36, 38, 41, 42, 46, 53, 55, 58, 60, Saccharomyces kluyveri, 36
70, 75, 84, 117, 128, 131, 132 Saccharomyces kudriavzevii, 19, 29, 36
Saccharomyces mikatae, 19, 20, 36, 164
Otidia onotica, 118 Saccharomyces paradoxus, 19, 20, 36, 164
Paracoccidioides brasiliensis, 171, 222, 225 Salmonella, 229
Penicillium chrysogenum, 116 Schizosaccharomyces, 57
Penicillium marneffei, 84, 171, 209 Schizosaccharomyces pombe, 36, 38, 41, 98, 100–104,
Peziza ostracoderma, 118 106, 109, 117, 122, 123, 125, 128, 131–133,
Phanerochaete chrysosporium, 35, 36, 38, 75 136, 137, 165, 197
Phytophthora infestans, 35 Sclerotinia sclerotiorum, 207
Pichia angusta, 30 Sordaria macrospora, 70
Pichia pastoris, 128, 136, 137 Sordariaceae, 62
Pichia stipitis, 36 Stereum hirsutum, 117
Podospora anserina, 115, 128 Streptococcus thermophilus, 186
Psathyrella candolleana, 118 Stropharia rugoso-annulata, 118
Pyrenomycetes, 55, 62 Tremella mesenterica, 118
Pyrenophora teres, 46 Trichoderma, 58
Trichoderma reesei, 89
Rattus norvegicus, 38
Renilla reniformis, 186 Ustilago maydis, 36, 41, 46
Rhizopus oryzae, 36
Yarrowia lipolytica, 19, 20, 36, 164, 165
Saccharomyces, 28, 57, 58
Saccharomyces bayanus, 19, 20, 26, 36, 164 Zygosaccharomyces bailii, 119
Saccharomyces castellii, 19, 21, 23, 28–30 Zygosaccharomyces rouxii, 26, 29
Subject Index
2D gel electrophoresis, 189 antifungal drugs, 114, 120, 121, 124, 133, 137, 185,
2D protein gel, 258 187, 199
antifungal therapy, 187
A. tumefaciens-mediated transformation (ATMT), antioxidant, 120, 126, 127, 129, 130, 133
37 antisense, 60, 65
AAF1, 175 antisense RNA, 255
abstinence, 26 apoptosis, 113–138, 179
Ace2, 206, 207, 210, 214, 215 apoptosis-inducing factor (AIF), 114, 115, 122, 133
actin polarisation, 84 appressoria, 117
adenine proteins and genes, 190 arginine biosynthesis, 192
ADH2, 28 assembly, 149
adhesins, 168, 171, 172, 185, 193, 196 aureobasidin A, 124
Affymetrix, 148 autolysis, 115, 116, 132
gene chips, 79, 80 autophagy, 114, 115, 136
agglutinin-like sequences (Als), 166–169, 171–173,
175 Bak, 118, 132, 136
aging, 116 Bax, 118, 129–131, 133, 136
ALB1, 38 Bayesian probability, 42
alcA, 78, 84, 91 Bcl-2, 114, 125, 129, 133, 136
alcohol dehydrogenase, 78 BemA, 84
alleles β-defensins, 177, 178
null, 28 β-oxidation, 170
α-(1,3)-glucan, 225 biofilms, 150, 155, 185, 192, 194
alternative animal model bioinformatics, 7, 11, 12
amoeba, 242 biological networks, 3, 8, 11
C. elegans, 242 BiP, 82, 83
Dictyostelium, 242 BLASTER, 78
Drosophila, 242 blood infections, 192
alternative oxidase, 254 BNI1, 84, 85
AMA1, 79 branching frequency, 90
amino acid biosynthesis, 42, 188, 189, 191, 192 BUD genes, 90
3-aminotroazole (3AT), 188–191 BUF1, 38
amphotericin B, 120, 124, 240, 241 bZIP, 88
Amt1, 208
animal model calcineurin, 119, 120, 122, 127
mouse, 241 calnexin, 82, 83, 136
rabbit, 241, 254 cAMP, 115, 126, 127, 129, 133, 154, 155
annexin-labelling, 116, 119, 121, 135 candida genome database (CGD), 186
annotation, 224, 227 CandidaDB, 186
cDNA, 247 candidiasis, 185
software, 247 capsule, 247, 248
annotation databases, 165 carbon catabolite repression, 87
270 Subject Index
carbon metabolism, 193–197 core and specific stress response genes, 100–102,
cardiolipin, 136 105
caspase, 114–116, 118, 120–122, 127–132, 136, 137 cortical markers, 84
cassette vectors, 77 cotA, 91
CBP1, 231 Csa1, 167
Cdc24, 84 Csy1, 192
CDC4, 26 CTA1, 178
Cdc42, 84 CTS1, 30
CDC48, 122, 124 CUG decoding, 186
cDNA, 149 cyclophilin, 83
cell devision cycle, 117, 125, 127, 134, 137 cytochrome c, 114, 118–120, 122, 124, 130, 136
cell factories, 87
cell growth, 99, 100, 102, 104, 108 data repositories, 7, 8
cell wall, 156, 167, 171–173, 175, 253 data standards, 12
cellular networks, 3, 10, 12, 13 development, 55, 56, 58, 61, 65, 66, 68, 69
cellulose, 87 asexual, 55, 56, 61, 65–67
cellulose-binding domains/modules, 87 sexual, conidia, 55, 56, 66, 67, 69
central nervous system (CNS), 240 diagnosis, 238, 239
centromeres, 23 CRAG (cryptococcal antigen), 239, 240
ceramides, 121, 128 DNA based detection, 240
cerebral spinal fluid, 239 india ink, 239
ceremide biosynthesis, 254 differential display, 133, 254
chaperones, 82, 88, 89 differential targeting, 84
checkpoints, 88 dimorphic fungus, 221, 222, 225, 232
CHEF gels, 245 dimorphism, 166, 171
ChIP-on-chip, 108 dithiothreitol, 80, 88
chitin synthase, 172 DNA damage, 100–102
chlamydospore, 157, 194 DNA repair, 125
chromatin modification in stress, 107 Downs strain, 225, 226
chromatin remodelling, 107 doxycycline-conditional mutants, 198
chromophore drug resistance, 194
FAD, 53, 65 duplication cycle, 89
chronological aging, 116 duplication, whole genome, 21
chymosin, 87 dynein mutations, 84
clinical isolates, 230
clock E-selectin, 177
biological, 55, 57, 60 EAP1, 175
circadian, 54, 62, 63, 66, 67, 70 Efg1, 173, 192–198
molecular, 20, 53 Efh1, 194
clock-controlled gene (ccg), 54, 65–68 eIF2, 188
clusters of orthologous groups (COGs), 39 entrainment, 54, 61, 65
codon bias, 58 environmental signals, 63
colinearity, 20 blue light, 54, 56, 64, 65, 69
colony, 138 light, 54–56, 60, 63–65, 69
colony morphology, 150 temperature, 63–65, 67, 68
commensal, 185, 194 environmental stress response, 133
comparative genomics, 38, 164, 166, 179 ER, 81–83, 88
complementation, 79 retention signal, 82, 83
conditional promoter, 78, 84 stress, 80, 88, 89
conidia, 221, 224, 229, 230 ERAD, 82, 88
contig, 149 essential genes, 77, 134
COPI coated vesicles, 83 evolution, 35–41, 197
COPII vesicles, 83 exocyst, 85
Subject Index 271
exoproteome, 80 genome(s), 57–59, 65, 66, 68, 147, 148, 163–167, 170,
expressed sequence tags (ESTs), 38, 41–46, 59, 60, 172–175, 179
67, 68, 246 genome-scale modell, 11–13
expression cloning, 79 genome-sequencing projects, 225–227
genomic databases, 6–8, 10, 12
feedback loop, 53, 54, 61–64, 69 genomics, 3–13, 54, 58–60
fitness attributes, 197, 198 germination, 229
Fkh2, 206, 207 GFP fusion, 84
flavohemoglobin, 252, 258 glaA, 80
fluconazole, 124, 240, 241 glass slide arrays, 79
flucytocine, 240, 241 glucan synthase, 172
foldases, 82, 88 glucoamylase, 87
FRQ, 54, 61–63, 65, 69 gluconeogenesis, 193, 197
frq, 54, 61–63, 65, 68 glucose repression, 187, 188
functional genomics, 3–6, 8, 11–13 glycolysis, 193
functional genomics levels, 3–6, 8, 11, 13 glycosyl hydrolases, 86
fungal infections, 185 glycosylation, 82, 87
fungicidal, 121, 124, 127, 133, 138 glyoxylate cycle, 170, 178, 187, 193, 196, 197
fungistatic, 127, 133, 138 Golgi, 77, 83, 88
fusion PCR, 77 Gpr1p, 170
GRACE (gene replacement and conditional expres-
G186A Histoplasma capsulatum strain, 225, 226 sion) method, 149, 173
G217B Histoplasma capsulatum strain, 225, 226 green fluorescent protein, 244
GAL loci, 29 GTPase, 83, 84, 153
GAS1, 37
GAS2, 37 HacA, 88, 89
gating, 65 haploid fruiting, 243
GCN response, 188, 189, 191, 192 haploinsufficiency, 149, 174
Gcn2, 188 Hbr, 90
Gcn4, 188, 189, 191, 192, 198 hbrA, 90
Gcn4 proteome, 189, 191 hbrB, 90
Gcn4 transcriptome, 189, 191 heat shock, 100, 101, 105
gene chips, 88 heavy metal, 100, 101
gene content, 28 het-c, 117
gene copy number, 87 heterokaryons, 77
gene disruption, 55, 78, 149, 150, 231, 255, 256 heterologous enzymes, 87
gene expression, 6–12, 151–156 heterologous proteins, 76, 87, 88
gene fusion, 87 heterothallism, 29
gene knockout, 150, 255, 256 histatins, 124
gene knockouts, 55, 232 histone deacetylase, 89
gene loss, 29 histone(s), 39, 107, 125
gene networks, 8 Histoplasma capsulatum phylogeny, 226
gene ontology terms, 154 HO gene, 26
gene order, 20 homologous recombination, 77, 255, 256
gene silencing, 58, 59 homothallism, 29
co-suppression, 55 horizontal gene transfer, 41
quelling, 55, 59 Hsp70, 83
genetic mapping, 245 HtrA2, 132
genetics, 243 Hwp1, 167, 171–173, 176
genome annotation, 59, 60, 69 hydrogen peroxide stress, 120, 126, 129, 131–133
genome sequencing, 186 hydrolytic enzymes, 84
genome websites, 36 hyperbranching, 85, 90
genome wide location analysis (GLA), 209–213 mutations, 90
272 Subject Index
hypha specific genes, 151–154, 193 messenger RNA (mRNA), 102–104, 106
hyphae, 115, 150–154 metabolic control, 4, 9
hyphal branching, 90 metabolic control analysis (MCA), 4, 9
hyphal development, 81 metabolic engineering, 4, 8, 13
hyphal growth, 75 metabolic fingerprinting, 7
hyphal morphology, 77 metabolic flux analysis (MFA), 8
Hyr1, 167, 171–173 metabolic footprinting, 7
metabolic networks, 3, 8, 9
IAP (inhibitor of apoptosis) protein, 132 metabolic pathways, 4, 7–9
ICAM-1, 177 metabolic regulation, 185–188, 191–194, 196–199
ICAT, 258 metabolism, 58, 63, 67, 69, 186–189, 192–199
ICL1, 170, 178, 196 metabolite profiling, 81
IgA, 169, 177 metabolites, 3–5, 7–11
immunocompetence, 185 metabolites role in regulation, 9, 13
infection structures, 36 metabolomics, 3, 8, 9, 12, 13
insertional mutagenesis, 36, 79, 232, 256, 257 metacaspase, 116, 120–123, 125, 130, 131, 134, 137
integrative biology, 3, 8, 10–13 methionine biosynthesis, 192
intein, 29 microarray(s), 6, 7, 11, 44, 81, 126, 129, 133, 148,
introns, 57, 70 151–155, 164–166, 173, 174, 179, 223, 253
ionising radiation, 102, 103 data standards, 98
iron acquisition, 228 experimental design, 100
iron homeostasis, 254 genomic approach, 97, 103, 104
iron metabolism, 168 microbial eukaryotes (COGEME), 41, 42
iron transport, 254 Mig1, 187, 196
itraconazole, 124, 240 model organisms, 97, 98, 103
IVET (in vivo expression technology), 174 modelling, models, 4, 8–13
molecular toolbox, 186
LAC4, 29 morphogenesis, 127, 150–157, 192–194, 198
LAC12, 29 MPG1, 37
laccase, 117, 129, 249 MSN2/4, 135
lactoferrin, 168, 177, 178 MSUD, 59
large-scale gene deletion, 187 Munich Information Centre for Protein Sequences
leader sequence, 81 (MIPS), 41
lethal mutations, 124, 134, 135 mutant strains libraries, 149
light-regulated genes, 54, 65 myoA, 84
light-responsive element (LRE), 64, 65 myosin, 84
lipase, 165, 167–169, 177
lovastatin, 81, 126 N-terminal secretion signals, 81
necrosis, 119–121, 133, 134
macrophage(s), 151, 192, 227–229 network, 8
MALDI-TOF, 80 network biology, 12
mannoprotein(s), 171 networks, 3, 4, 7–13
MAP kinase, 86 neutropenia, 185
MAT locus, 26, 29 neutrophils, 192, 193
mathematical models, 4, 11 Nrg1, 173, 174, 195, 196, 198
mating, 148, 156, 157, 169, 170 nucleases, 132
mating type, 231, 243, 250
Mds3, 206–208 ohnologs, 21
MEC1, 123, 125 oligonucleotide array, 148
meiosis, 55, 56, 59, 89, 221 opaque cells, 150
meiotic cell death, 117 oral candidiasis, 185, 192
melanin, 225, 247, 249, 253, 256, 257 orphan genes, 41
meningitis, 239 orthologous, 39
Subject Index 273
oscillator, 54, 63, 70 PLD1, 166

osmotic stress, 98, 100, 101, 103, 105, 107 Pld1p, 166
osmotin, 121, 126–129 PMK1, 38, 44, 46
over-expression induced lethality, 134 pneumonia, 239
oxidative stress, 99–103, 129 polar growth, 77, 79, 84, 85, 90
polarity defects, 83
p53, 137 polarity establishment, 84, 90
paralogous, 39 polyploidy
parasexual cycle, 77, 186 allo, 24
PAS domain, 61, 65 auto, 24
LOV domain, 65 post-ER secretion, 83
pathogenic, 147 post-transcriptional regulation, 103, 106, 107
patient population post-translational translocation, 81
HIV, 238 PPI, 82, 83
HIV negative, 238 pradimicin A, 121
therapy, antineoplastic, 238 predictive algorithms, 164, 167
tissue transplantation, 238 programmed cell death (PCD), 113–129, 131–138
PCR, 77 promoter exchange, 78
PCR-based gene disruption, 186 promoter(s), 58, 60, 64, 65, 68, 69, 87, 244, 255
PDE1, 37 proteases, 87, 88
pectinases, 87 proteasome, 82, 88, 125, 134, 135
peptide mass fingerprinting, 258 protein disulfide isomerase, 82
period length, 61, 62, 66, 69 protein folding, 82
PERK, 89 protein kinase, 63
permeability transition, 114, 118–120, 131 calcium/calmodulin-dependent kinase, 64
peroxidase, 252 casein kinase Ia, 64
thiol, 252, 258 casein kinase II, 63, 64
pH regulation, 228 protein kinase A (PKA), 115, 122, 126, 133, 135
phagocytosis, 163, 167, 170, 177–179, 192, 227, 228 protein kinase B (PKB), 123, 127
phalloidin, 84 protein kinase C (PKC), 66, 128
phase protein networks, 8
phase-shift, 54, 56, 62, 69 protein secretion, 75, 81, 89
phenanthroline, 121 proteinases, 86
phenotypic switching, 170, 171, 194 proteome microarrays, 6
pheromone, 118, 120, 127, 128, 156 proteomics, 3–12, 80, 134, 185–199, 258, 259
pheromone signalling, 231 AACT, 106
phosphatase, 155 ICAT, 106
phosphoinositide 3-OH kinase, 127 mass spectrometry, 106
phospholipase(s), 86, 249 pseudogene, 20, 26, 101
phosphorylation, 63–66 pseudohyphae, 150, 151, 154, 191, 192
photoadaption, 57, 63 pseudohyphal growth, 89
photobiology, 57, 58, 60, 62, 68, 70 PTH11, 37
photoreceptor, 54, 57, 65, 66, 70 purine biosynthesis, 189
phototropism, 57 pyrG, 78
Phr1, 172, 173
Phr2, 172, 173 quality control system, 82
phylogenetics, 39, 40
phylogenomics, 39 Rac, 84
physical mapping, 245 race tubes, 62
phytosphingosine, 121 Rap1, 208, 209
piecemeal microautophagy, 115 Ras superfamily, 83
plant pathogens, 35–46 Ras-signalling, 115, 123, 126, 127, 129, 133, 135, 195
plasma membrane, 84 Rbr1p, 173
274 Subject Index
Rbt1, 167, 173 secretory pathway, 76, 81, 88

Rbt5, 167, 172 secretory vesicles, 85
reactive nitrogen species, 229 selectable markers, 243, 244
reactive oxygen species (ROS), 115–121, 124–130, sepA, 84, 85
132, 135, 136 septation, 77
recombinant PCR, 77, 78 septum formation, 85
reconstituted human epithelium (RHE), 175, 177 sequencing, shotgun, 245
regulatory networks, 5, 11–13 serotype, 242
relics, 26 sexual cycle, 148
repeat-induced point mutation (RIP), 41, 55, 58, 59 signal anchors, 81, 87
replicating vectors, 79 signal peptides, 81, 87
replicative aging, 116 signal recognition particle, 81
repressor, 151, 154 signal sequences, 82
respiratory burst, 228 signal transduction
restriction enzyme-mediated DNA integration calmodulin, 250
(REMI), 37, 78 PKA, 250
Rfg1, 188, 198 PKC, 256
rhythms signalling, 98–105, 107, 108
circadian, 53, 61, 70 signature tagged mutagenesis, 78, 257
non-circadian, 61, 62, 70 SIR1, 29
ribosomal protein genes, 191 Snf1, 188
ribosome, 106 Snf3p, 170
Rim101, 173, 206–208 SOD1, 124
RNA interference, 231, 232 somatic incompatibility, 116
RNA-binding proteins, 103 sphingolipids, 128
RNAi, 107, 255 Spitzenkörper, 85
RSY1, 38 splice junctions, 247
RT-PCR, 174, 177 splicing, 58, 60, 64, 66
SrgA, 85
Saccharomyces cerevisiae starch, 87
characteristics, 4, 6, 8, 11, 13 Ste12, 209–215
collections of yeast mutants, 6, 7 STRE-elements, 126, 129
databases and resources, 6–8 stress, 88
functional genomics. State of the art, 4, 5 nitrosative, 251, 258
GRAS status, 4 oxidative, 251, 258
model eukaryote, 3, 13 stress activated MAP kinase, 99, 100, 107
reference biological system, 3, 4 stress activated protein kinase (SAPK), 98, 104
SAGE, 253 stress response, 127, 129, 133, 193, 194, 197, 198
saprophytes, 35, 40–42 stress response genes, 100–103, 105
SarA, 83 stress signalling pathway
Sec complex, 81, 83 activation threshold, 107
Sec3, 86 evolutionarily conservation, 98, 104
Sec4, 83, 85 feedback, 104
Sec61p complex, 82 STRING, 215
Sec63, 83 subtelomeres, 21
secondary metabolic pathway genes, 81 sugar-induced death, 118
secreted aspartic proteinase(s) (Sap), 166, 168, 169, suicide, 114, 120, 126, 132, 138
171, 177, 197 superoxide dismutase(s) (Sod), 167, 168, 171, 172,
secretion, 85, 89 178, 252
secreted proteins, 86 symptoms, 238
stress, 79 synteny, 21
secretion genes, 81 systemic infections, 185, 191, 194
secretome, 167 systems biology, 3, 5, 7–13
Subject Index 275
T-DNA, 232, 233 two-color array, 148

tandem repeats, 28 type 2A protein, 155
temperature, 222, 225
compensation, 61, 62, 66, 70 UAU cassette, 150
thioredoxin, 252 UBC3, 38
domain, 82 ubiquitin(s), 39, 88
TOR signalling, 115 ubiquitination, 115, 125, 134, 135
Tox locus, 37 unfolded protein response, 80, 83, 88, 89
trans-Golgi, 84 URA1, 28
transcript profiling, 3–7, 9, 10, 12, 102, 114, 133, urease, 250
134, 148, 149, 151–156, 185–198 UV irradiation, 118
transcription, 54, 61, 64–66, 68, 205–215 vacuolar protein sorting, 84, 91
transcription factor(s), 88, 97–103, 107, 108, 151, vaginal candidiasis, 185, 192
154, 155, 227 vesicle docking, 85
transcription network/program, 97, 101, 102, 104, vesicle supply centre, 85
108 vesicle transport, 84
transcriptional regulation, 97, 100–103, 106–108 virulence, 35–46, 150, 164, 167–171, 173–175, 179,
transcriptional repression, 101, 102 185, 191, 205–215
transcriptional repressor, 195 virulence attributes, 185, 197
transformation, 55, 59, 77 virulence genes, 197
Agrobacterium tumafaciens mediated, 244, virulence regulators, 197
258 VMA1 gene, 29
biolistic, 244, 256, 257 VPS33, 91
electroporation, 244, 256 VVD, 65, 70
translation inhibitors, 121
translational attenuation, 89 WC-1, 54, 61, 63–66, 68, 70
translational regulation, 106 WC-2, 54, 61, 64–66, 70
translocations, 20 WCC, 64–66, 71
translocon, 82 weak acids response, 119
transposable elements, 79 white cells, 150
transposition, 79 white–opaque transition, 156
transposon mutagenesis, 36, 78, 155, 174, 198 WU24 Histoplasma capsulatum strain, 226
transposon-arrayed gene knockout (TAG-KO), 37
TRAPP I, 83 XYZ3, 20, 26
TRAPP II, 83
tricarboxylic acid (TCA) cycle, 193 YBR184W, 30
tropomyosin, 84 YCA1 (yeast metacaspase), 116, 119, 123, 131, 132
TUNEL (terminal deoxynucleotidyl transferase- yeast form, 150–154
mediated dUTP nick-end labelling), yeast homologues, 90
116–121, 135 yeast-to-hyphal transition, 151–154
tunicamycin, 88 YKL023W, 30
TUP1, 174, 195–198 Ypt1, 83

(Themy 4

Uploaded by

Copyright:

Available Formats

(Themy 4

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Themy 4

Uploaded by

Copyright:

Available Formats

The Mycota

I Growth, Differentiation and Sexuality

II Genetics and Biotechnology

III Biochemistry and Molecular Biology

IV Environmental and Microbial Relationships

VI Human and Animal Relationships

VII Systematics and Evolution

VIII Biology of the Fungal Cell

XII Human Fungal Pathogens

XIII Fungal Genomics

With 37 Figures, 6 in Color, and 16 Tables

Library of Congress Control Number: 2005928810

ISBN-10 3-540-25594-X Springer Berlin Heidelberg New York

Springer is a part of Springer Science+Business Media

Editor: Dr. Dieter Czeschlik, Heidelberg, Germany

Printed on acid-free paper 31/3152 5 4 3 2 1 0

Alistair J.P. Brown

In early 1989, encouraged by Dieter Czeschlik, Springer-Verlag, Paul A. Lemke and

Bochum, Germany Karl Esser

Aberdeen, UK Alistair J.P. Brown

Biocemistry and Molecular Genetics

1 Metabolomics and Systems Biology in Saccharomyces cerevisiae

2 Genome Evolution in Hemiascomycete Yeasts

3 Investigating the Evolution of Fungal Virulence by Functional Genomics

Fungal Rythms and Responses

4 Circadian Rhythms, Photobiology and Functional Genomics in Neurospora

5 Genomics of Protein Secretion and Hyphal Growth in Aspergillus

6 The Genomics of Stress Response in Fission Yeast

7 Programmed Cell Death and Apoptosis in Fungi

8 Genomic Analysis of Cellular Morphology in Candida albicans

9 Postgenomic Approaches to Analyse Candida albicans Pathogenicity

10 Integration of Metabolism with Virulence in Candida albicans

11 Regulators of Candida glabrata Pathogenicity

12 Using Genomics to Study the Life Cycle of Histoplasma capsulatum

13 Cryptococcus neoformans Pathogenicity

David B. Archer (e-mail: david.archer@nottingham.ac.uk, Tel.: +44-115-9513313)

Jürg Bähler (e-mail: jurg@sanger.ac.uk, Tel.: +44-1223-494861) The Wellcome Trust

Madhumita C. Barooah School of Biological and Chemical Sciences, University of

Alistair J.P. Brown (e-mail: al.brown@abdn.ac.uk, Tel.: +44-1224-555883,

Juan I. Castrillo (e-mail: juan.i.castrillo@man.ac.uk) Faculty of Life Sciences, Michael

Jay C. Dunlap Departments of Genetics and Biochemistry, 704 Remsen, Dartmouth

Ken Haynes (e-mail: k.haynes@imperial.ac.uk, Tel.: +44-20-83831245,

Bernhard Hube (e-mail: HubeB@rki.de, Tel.: +49-1888-7542116,

Jennifer K. Lodge (e-mail: lodgejk@slu.edu, Tel.: +1-314-5778143,

Jennifer J. Loros (e-mail: jennifer.loros@dartmouth.edu, Tel.: +1-603-6501154,

Louis J. Montcalm Department of Genetics, Smurﬁt Institute, University of Dublin,

Carol A. Munro (e-mail: c.a.munro@abdn.ac.uk, Tel.: +44-1224-555882,

André Nantel (e-mail: andre@bri.nrc.ca) Biotechnology Research Institute, National

Rex T. Nelson Edward A. Doisy Department of Biochemistry and Molecular Biology,

Stephen G. Oliver (e-mail: steve.oliver@manchester.ac.uk, Tel.: +44-161-2751578;

Mark Ramsdale (e-mail: m.ramsdale@abdn.ac.uk, Tel.: +44-1224-555882,

Anita Sil (e-mail: sil@cgl.ucsf.edu, Tel.: +1-415-5021805, Fax: +1-415-476820) Depart-

Darren M. Soanes School of Biological and Chemical Sciences, University of Exeter,

Nicholas J. Talbot (e-mail: n.j.talbot@exeter.ac.uk, Tel.: +44-1392-264673) School of

Geoffrey Turner (e-mail: g.turner@shef.ac.uk, Tel.: +44-114-2226211) Department

Malcolm Whiteway (e-mail: Malcolm.Whiteway@cnrc-nrc.gc.ca,

Ken Wolfe (e-mail: khwolfe@tcd.ie, Tel.: +353-1-6081253, Fax: +353-1-6798588) De-

J.I. Castrillo1 , S.G. Oliver1

CONTENTS more integrative way (Oliver 1997, 2002; Oliver