Infection, Genetics and Evolution 11 (2011) 654–662

Contents lists available at ScienceDirect

Infection, Genetics and Evolution

journal homepage: www.elsevier.com/locate/meegid

Animal and human pathogenic Escherichia coli strains share common genetic
backgrounds
Olivier Clermont a, Maiwenn Olier b, Claire Hoede a, Laure Diancourt c, Sylvain Brisse c,
Monique Keroudean b, Jérémy Glodt a, Bertrand Picard d, Eric Oswald b,e, Erick Denamur a,*
a
UMR722, INSERM and Université Paris Diderot, Site Xavier Bichat, 16 rue Henri Huchard, 75018 Paris, France
b
INRA, UMR 1225, 31076 Toulouse, France
c
Genotyping of Pathogens and Public Health, Institut Pasteur, 75724 Paris, France
d
UMR722, INSERM and Université Paris Nord, Site Xavier Bichat, 75018 Paris, France
e
Laboratoire de Bactériologie-Hygiène, CHU de Toulouse, Institut Fédératif de Biologie, 31059 Toulouse, France

A R T I C L E I N F O A B S T R A C T

Article history: Escherichia coli is a versatile species encompassing both commensals of the digestive tracts of many
Received 12 October 2010 vertebrates, including humans, and pathogenic strains causing various intra- and extraintestinal
Received in revised form 4 February 2011 infections. Despite extensive gene flow between strains, the E. coli species has a globally clonal
Accepted 7 February 2011
population structure, consisting of distinct phylogenetic groups. Little is known about the relationships
Available online 13 February 2011
between phylogenetic groups and host specificity. We therefore used multilocus sequence typing (MLST)
to investigate phylogenetic relationships and evaluated the virulence gene content of 35 E. coli strains
Keywords:
representative of the diverse diseases encountered in domestic animals. We compared these strains with
Escherichia coli
Phylogeny
a panel of 101 human pathogenic and 98 non-human and human commensal strains representative of
Host specificity the phylogenetic and pathovar diversity of this species. A global factorial analysis of correspondence
Pathogenic indicated that extraintestinal infections were caused mostly by phylogenetic group B2 strains, whereas
intraintestinal infections were caused mostly by phylogenetic group A/B1/E strains, with strains
responsible from extraintestinal or intraintestinal infections having specific virulence factors. It was not
possible to distinguish between strains of human and animal origin. A detailed phylogenetic analysis of
the MLST data showed that numerous pathogenic animal and human strains are very closely related, and
had a number of virulence genes in common. However, a set of specific adhesins was identified in animal
non-B2 group strains of all pathotypes. In conclusion, human and animal pathogenic strains share
common genetic backgrounds, but non-B2 strains of different origins seem to have different sets of
adhesins that could be involved in host specificity.
ß 2011 Elsevier B.V. All rights reserved.

1. Introduction Despite the high degree of gene flow, the population structure of this
species remains mostly clonal (Touchon et al., 2009), with the clear
Escherichia coli is one of the most versatile bacterial species. It delineation of at least six principal phylogenetic groups (A, B1, B2, D,
alternates between its primary habitat, the gut of vertebrates, where E and F) (Jaureguy et al., 2008; Tenaillon et al., 2010) and the Shigella
it lives as a commensal (Tenaillon et al., 2010), and its secondary strains, which belong to the E. coli species but cluster outside the
habitat, water and sediment (Savageau, 1983). It may also function principal phylogenetic groups (Escobar-Paramo et al., 2003; Pupo
as an intra- and extraintestinal pathogen in humans and many other et al., 2000). It has been shown that genetic background plays a role
animal species (Kaper et al., 2004). This diversity of lifestyles is in the acquisition, retention and expression of foreign DNA (Escobar-
achieved through a high degree of genome plasticity, with gene Paramo et al., 2004).
losses and gains, through horizontal transfer (Rasko et al., 2008; Besides the Shigella strains that are clearly restricted to human
Touchon et al., 2009). This species has a core genome of less than host and have inactivated numerous genes during their evolutionary
2000 genes, but more than 10,000 genes in total (Rasko et al., 2008; history (Denamur et al., 2010), it has been suggested that some group
Touchon et al., 2009). Thus, the diverse phenotypes observed result B2 strains of the O81 serogroup may be specific for humans
principally from a large number of different gene combinations. (Clermont et al., 2008) and that some group B1 strains with the hly
gene may be specific for animals (Escobar-Paramo et al., 2006).
However, little is known about the relationships between phyloge-
* Corresponding author. Tel.: +33 1 57 27 75 34. netic groups and host specificity. In this context, the extent to
E-mail address: erick.denamur@inserm.fr (E. Denamur). which bacterial strains from infected humans and animals are

1567-1348/$ – see front matter ß 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.meegid.2011.02.005
 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662 655

phylogenetically related is unclear. Several studies based on intraintestinal [afaD, ipaH, stx1, stx2, eltB (LT), estA (ST), bfpA, eae,
serotyping, multilocus enzyme electrophoresis, outer membrane aaiC and aatA] (Table S2) infections by PCR, as previously described
protein profiles, pulsed-field gel electrophoresis, ribotyping, ran- (Escobar-Paramo et al., 2004). The extraintestinal genes tested
domly amplified polymorphic DNA, phylogenetic group affiliation correspond to the main classes of extraintestinal VFs, i.e. adhesin,
and virulence gene content have shown close relationships between toxin, iron capture system, protectin. The intraintestinal genes
human and animal isolates (Achtman et al., 1986; Cherifi et al., 1991, tested allow the classification in the main intestinal E. coli
1994; Ewers et al., 2007; Girardeau et al., 2003, 2005; Johnson et al., pathovars, i.e. ETEC, EPEC, EHEC, EIEC, EAEC and DAEC (Kaper
2001, 2008; Mariani-Kurkdjian et al., 1993; Maynard et al., 2004; et al., 2004). We also used PCR, as previously described (Bertin
Moulin-Schouleur et al., 2006; Pradel et al., 2001; Wu et al., 2008). et al., 1996; Boerlin et al., 2005; Dow et al., 2005; Franck et al.,
More recently, multilocus sequence typing (MLST) has been used to 1998; Imberechts et al., 1994), to check for the presence of
study the phylogenetic relationships between strains in more detail. adhesins classically associated with animal-specific pathogenic
These studies have focused principally on extraintestinal pathogenic strains of E. coli: K99 (fanA), K88 (faeG), F17 (f17A), F18 (fedA) and
E. coli (ExPEC) [including avian pathogenic E. coli (APEC) in particular] Afr2 (afr2G), referred latter on as ‘‘animal adhesins’’.
(Mora et al., 2009; Moulin-Schouleur et al., 2006, 2007) and
enterohemorrhagic E. coli (EHEC) (Feng et al., 2007; Newton et al., 2.3. PCR O-typing
2009) strains. Large amounts of data have been amassed, but these
findings are fragmented and difficult to compare, as different typing O-type was determined by an allele-specific PCR (Clermont
approaches and non-redundant sets of strains were used in the et al., 2007) using the primers given in Table S3. We assessed 28 O-
various studies. Consequently, there is currently no overview of the types with this assay. These 28 O-types were selected based on the
global relationships between animal and human pathogenic strains O-types already reported using the classical serological method for
in the framework of the phylogeny of the E. coli species as a whole. the other strains of the collection.
The aim of this work was to use MLST to study the phylogenetic
relationships and to assess the virulence gene content of 35 E. coli 2.4. MLST
strains representative of the diverse diseases encountered in
domestic animals, comparing these strains with a panel of 101 MLST was performed with partial dinB, icdA, pabB, polB, putP, trpA,
human pathogenic and 98 non human and human commensal trpB, and uidA sequences (Jaureguy et al., 2008). Allele sequences and
strains representative of the phylogenetic diversity of the species sequence types (STs) are available from Institut Pasteur’s MLST
and including well characterized archetypal strains. website, at www.pasteur.fr/mlst. Phylogenetic analysis was per-
formed with the concatenated sequences of the eight genes, by the
2. Materials and methods maximum likelihood (ML) method, as implemented in the PHYML
program (Guindon et al., 2005), as well as by the neighbor joining
2.1. Bacterial strains (NJ) and maximum parsimony (MP) methods using MEGA4 (Tamura
et al., 2007), with E. fergusonii as the outgroup.
We studied 234 E. coli strains and one strain of Escherichia
fergusonii, the closest relative of E. coli (Lawrence et al., 1991) (Table 2.5. Factorial analysis of correspondence (FAC)
S1). Five groups of E. coli strains were represented: (i) a panel of 35
strains pathogenic in animals and representative of the various FAC was used to describe associations between the different
diseases encountered in domestic animal species (from 8 birds and data sets. FAC uses a covariance matrix based on Chi squared
27 mammals) comprising 15 ExPEC/APEC (the APEC strains distances (Greenacre, 1992). This computation method determines
originating from the 8 birds) and 20 intraintestinal pathogenic E. a plane defined by two principal axes of the analysis. The first axis
coli (InPEC) [8 enterotoxigenic E. coli (ETEC), 4 enteropathogenic E. (F1) accounts for most of the variance, and the second axis (F2),
coli (EPEC), 5 Shiga toxin-producing E. coli (STEC)/EHEC and 3 orthogonal to F1, accounts for the largest part of the variance not
unclassified InPEC] strains (Table 1), (ii) a panel of 93 pathogenic accounted for by F1. FAC was conducted with SPAD.N 4.5 software
human strains comprising 43 ExPEC [29, 8 and 6 involved in urinary (Cisia, Saint Mandé, France), based on a two-way table. This table
tract infection (UTI), newborn meningitis (NBM), septicemia and had 234 rows, one for each E. coli strain, and 35 columns,
miscellaneous infections, respectively] and 50 InPEC [7 ETEC, 6 EPEC, corresponding to 35 variables: human/animal origin, commensal,
8 EHEC, 10 enteroaggregative E. coli (EAEC), 1 enteroinvasive E. coli ExPEC and InPEC characters, seven phylogenetic groups corre-
(EIEC), 16 diffusely adherent E. coli (DAEC) and 2 unclassified InPEC], sponding to A, B1, B2, C, D, E, F and ungrouped (UG) strains
(iii) 45 non human mammalian commensal strains, (iv) 53 human according to the MLST data (Escobar-Paramo et al., 2004; Jaureguy
commensal strains and (v) 8 human InPEC strains for which the et al., 2008), and 23 VFs (neuC, sfa/foc, iroN, papC, papGI, papGII,
complete genome was available (Ogura et al., 2009; Rasko et al., papGIII, hlyC, cnf1, hra, fyuA, ‘‘animal adhesins’’ (AnAd), afaD, ipaH,
2008); these strains were typed in silico in this study. The strains stx1, stx2, estA, eltB, bfpA, eae, aaiC and aatA). The data in this table
from groups (ii), (iii) and (iv) in this list originated mostly from three were attributed a binary code: ‘‘1’’ for present and ‘‘0’’ for absent.
published collections (Escobar-Paramo et al., 2004; Le Gall et al., The loading score for each variable on the plane of the variables
2007; Ochman and Selander, 1984) and, with the strains from group (factors F1 and F2, respectively) can be inferred from the
(v), may be considered representative of the phylogenetic and coordinates of the X and Y variables on the F1/F2 plane. Moreover,
pathovar diversity of E. coli. They encompass archetypal strains for the data corresponding to active variables were calculated directly
various diseases, and complete genome sequences are available for in the FAC, whereas the illustrative variables were only projected
some, in addition to group (v) strains. onto the plane and not included in the computation.

2.2. Virulence factor (VF) screening 3. Results and discussion

We tested for the presence of virulence factors involved in 3.1. Multidimensional analysis
extraintestinal (neuC, kpsE, sfa/foc, iroN, aer, iha, papC, papGI, papGII,
papGIII, hly, cnf1, hra, sat, ire, usp, chromosomal ompT, ibeA, malX, We assessed the global relationships between the phylogenetic
irp2, fyuA and traT) (Diard et al., 2007; Johnson et al., 2006) and groups, the VF content and origin (human versus animal and
656 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662

Table 1
Main characteristics of the 35 pathogenic animal E. coli strains studied.

1 Table 1. Main characteristics of the 35 pathogenic animal E. coli strains studied

Phylo-
genetic

estA (ST)
eltB (LT)
papG III
papG II
papG I
neuC b

sfa/foc
group STa

papC

afaD
ipaH

aaiC
kpsE

hlyC

bfpA

aatA
Country Patho- "Animal"

iroN

fyuA
cnf1

stx1
stx2
hra

eae
aer
Strain ID Host O-type
origin genicity and adhesin
sub-
group
25KH9 Bos taurus Belgium ETEC AI 2 101 - - - + - + - - - - - + - F17a - - - - - - - - - -
S1191 Sus scrofa USA STEC AI 2 139 - - - - - - - - - - - - - F18 - - - + - + - - - -
510 Bos taurus Belgium ETEC AI 160 101 - - - - - - - - - - - + - K99 - - - - - + - - - -
431 Sus scrofa USA ETEC AI 160 101 - - - - - - - - - - - + - K99 - - - - - + - - - -
262KH89 Bos taurus Belgium Diarrhea AI 230 26 - - - - - - - - - + - - - K88 - - - - - - - - + -
255/1-1 Bos taurus France EHEC AI 140 Unknown - - - - - - - - - - - - - - + - + - - - - - - -
126A Bos taurus Belgium ETEC UG 228 8 - - - - - - - - - - - + - K99 - - - - - + - - - -
86-1390 Sus scrofa Canada EPEC B1 245 45b - - - - - - - - - - - - + - - - - - - - - + - -

193 Bos taurus USA EHEC B1 246 26 - - - - + - - - - - - + + - - - - - - - - + - -

Northern
C/15333 Bos taurus EPEC B1 232 26 - - - - - - - - - + - - + - - - - - - - - + - -
Ireland
Oryctolagus
RDEC-1 USA EPEC B1 493 15 - - - - + - - - - - - - + - - - - - - - - + - -
cuniculus
5131 Sus scrofa Canada ExPEC B1 248 115 - - + + + + - - + - - + - - - - - - - - - - - -
987 Sus scrofa USA ETEC B1 249 9 - - - - - - - - - - - - - K99 - - - - - + - - - -
31A Bos taurus France Diarrhea B1 NC 153 - - - - + + - - - - - + + F17c/GAF - - - - - - - - - -
211 Bos taurus Belgium ETEC B1 247 Rough + + - - - - - - - - - + - - - - - - - - - - - -
S5 Ovis aries UK ExPEC B1 266 15 + + - - - - - - - + - - + F17b - - - - - - - - - -
Oryctolagus
E22 France EPEC B1 135 103 - - - - - - - - - - - - - AFR2 - - - - - - - + - -
cuniculus
111KH86 Bos taurus Belgium Diarrhea B1 227 Unknown - - - - - - - - - - - + - F111 - - - - - - - - - -
789 Gallus gallus Israel APEC C 148 78 - - + + + - - - - - - - + - - - - - - - - - - -
BEN 0265 Gallus gallus France APEC C 147 78 - - - + + - - - - - - - + - - - - - - - - - - -
1404 Bos taurus France ExPEC C 159 78 - - - - - - - - - + - - + F17b - - - - - - - - - -
239KH89 Bos taurus Belgium ExPEC C 229 Unknown - - - - + + - - + + + + + - - - - - - - - - - -
DEC7a Sus scrofa USA ETEC C 157 157 - - - - - - - - - + - - + K88 - - - - + + - - - -
248/1-2 Bos taurus France EHEC C 270 Unknown - - - + - + - - - - - + + - + - + - - - - - - -

G7 Sus scrofa UK ETEC C NC 8 - - - - - - - - - + - - + K88 - - - - + - - - - -

DEC4a Bos taurus Argentina EHEC E 284 157 - - - - - - - - - - - - - - - - + - - - - + - -
BEN 1189 Gallus gallus Belgium APEC D 145 78 - - - + + - - - - - - - - - - - - - - - - - - -
BM2-1 Bos taurus France ExPEC B2 II 150 2 - - + + - - - - - - - + + - - - - - - - - - - -
M623 Sus scrofa Spain ExPEC B2 IV 223 2 + + - + - + - - + + + + + - - - - - - - - - - -
28C Sus scrofa Spain ExPEC B2 VII 141 75 - - + + - + - - + + + + + - - - - - - - - - - -
APECO1 Gallus gallus USA APEC B2 IX 418 1 + + - + + + - + - - - - + - - - - - - - - - - -
Meleagridis
BEN 0139 France APEC B2 IX 146 2 + + - + + - - - - - - - + - - - - - - - - - - -
gallopavo
BEN 2908 Gallus gallus France APEC B2 IX 146 2 + + + + + - - - - - - + + F111 - - - - - - - - - -
BEN 0374 Gallus gallus Spain APEC B2 IX 146 18 + + + + + - - - - - - - + - - - - - - - - - - -
BEN 0079 Gallus gallus France APEC B2 IX 146 18 + + + + + - - - - + + + + - - - - - - - - - - -

a
ST numbers correspond to the IP scheme (www.pasteur.fr/mlst). NC: non-coded because more than one gene over the 8 studied are not amplifiable.
b
Colors correspond to the different PAIs and plasmid determined as in (Bingen-Bidois et al., 2002): orange, PAI IIJ96; green, PAI III536; violet, HPI; grey, plasmid origin.

commensal versus pathogenic) of the strains, by carrying out a FAC aer, afaD and aaiC were projected onto the negative values of the F2
with phylogenetic groups and VFs as active variables and the axis. The variables human, animal and commmensal were grouped
animal/human, commensal, ExPEC and InPEC variables as illustra- around the origin of the axes and it was therefore not possible to
tive variables. Strains were assigned to phylogenetic groups on the differentiate between them in this FAC (Fig. 1). When projecting
basis of MLST data. Seven groups were considered: the six the strains on the plane, most of the ExPEC strains had positive F1
previously recognized groups (A, B1, B2, D, E and F) and the C coordinates (Fig. 2A), whereas the InPEC strains had negative
group plus ungrouped strains (see below for the definition of the C coordinates on this axis (Fig. 2B). Moreover, it was not possible to
group and ungrouped strains). On the F1/F2 plane, which distinguish between ExPEC and InPEC strains of human and animal
accounted for 29.37% of the total variance, the variables papGI, origin in this FAC (Fig. 2).
papGIII, cnf1, sfa/sfoc, hlyC, iroN, hra, papC, neuC, kpsE, fyuA, B2 group This global analysis suggests that extraintestinal infections are
and ExPEC were projected onto the positive values of the F1 axis, caused principally by B2 strains, which have many extraintestinal
whereas the variables stx2, E group, eae, stx1, bfpA, eltB, estA, B1 VFs, whereas intraintestinal infections are caused mostly by A/B1/E
group, InPEC, AnAd, ipaH, aatA and A group were projected onto the strains exhibiting intestinal VFs. It was not possible to distinguish
negative values of this axis. The variables F group, papGII, D group, between strains of human and animal origin.
 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662 657

Fig. 1. Factorial analysis of correspondence (FAC) of the 234 E. coli strains. Projections of the 35 variables – human/animal origin, commensal, ExPEC and InPEC characters, 7
phylogenetic groups (A, B1, B2, C, D, E and F) and ungrouped strains (UG) and 23 VFs (see Section 2) – onto the F1/F2 plane calculated in the FAC (Greenacre, 1992). AnAd:
Animal adhesin.

3.2. Fine-scale phylogenetic analysis Two subgroups (I and II) were identified within the A phylogenetic
group, the subgroup II corresponding to the strains with an A0
We investigated the relationships between human and animal genotype with the PCR triplex typing method, i.e. the absence of
pathogenic strains in more details, at the clonal level and within amplification of any of the genetic markers (chuA, yjaA and
the species as a whole, by carrying out MLST analysis on the 234 TSPE4.C2) (Clermont et al., 2000; Gordon et al., 2008). No
strains. Three main MLST schemes are currently available for E. coli, phylogenetically robust subgroup was observed within the B1
i.e. the scheme used in this work (Institut Pasteur, IP, scheme) and D phylogenetic groups, only some strains were grouped with
(Jaureguy et al., 2008), the Whittam scheme (http://www.shiga- high bootstrap values. The phylogenetic groups and subgroups
tox.net/ecmlst/cgi-bin/dbquery) (Reid et al., 2000) and the Acht- delineated above were retrieved when the phylogenetic tree from
man scheme (http://mlst.ucc.i.e./mlst/dbs/Ecoli) (Wirth et al., the concatenated sequences was reconstructed using NJ (Fig. S1)
2006), each using a different combination of genes. The MLST and MP (Fig. S2) methods. Furthermore, a minimum spanning tree
data are usually studied in two ways, with the alleles at different analysis based on allelic profile data gave similar results for the
loci providing an allelic profile, which defines the sequence type shallow phylogenetic grouping (data not shown). The ST numbers
(ST). No weighting is given to take into account the number of of the strains using the IP scheme is given in Table 1 and Table S1,
nucleotide differences between the alleles. Alternatively, nucleo- as well as, when available, the correspondence with the ST using
tide sequences may be used for phylogenetic reconstructions the Achtman scheme (Table S1).
(Tenaillon et al., 2010). The results obtained with these three With this phylogenetic approach, we clearly identified animal
schemes are highly correlated (Gordon et al., 2008), suggesting and human strains belonging to the same phylogenetic subgroups
that the clonal structure of the species is robust. or clonal complexes (closely related STs), as presented below.
We chose to use the set of genes described in a previous study Both animal and human ExPEC strains were found in all but three
(Jaureguy et al., 2008), as we have previously used this MLST of the subgroups of the B2 phylogenetic group (I, V and VIII) (Table 1
scheme to characterize a unique set of strains representative of the and Table S1, Fig. 3). The most highly represented subgroup was
phylogenetic and lifestyle diversity of the species. Large amounts subgroup IX, which corresponds to ST95 of the Achtman scheme
of data are available for these strains. The phylogenetic tree (Wirth et al., 2006), ST29 of the Whittam scheme (Newton et al.,
reconstructed from the concatenated sequences using the ML 2009), and the B2-1 group of a previous study (Moulin-Schouleur
method (Fig. 3) showed the major phylogenetic groups previously et al., 2007). This subgroup encompasses the APEC strains of
described (Tenaillon et al., 2010): A, B1, B2, D, E and F. An serogroup O1, O2 and O18 (Johnson et al., 2007; Mora et al., 2009;
additional group, closely related to the B1 group but identified as A Moulin-Schouleur et al., 2007) and the archetypal human strains
by PCR triplex phylogrouping (Clermont et al., 2000), was also UTI89 (Chen et al., 2006), a strain causing urinary tract infection, and
identified and called C (Escobar-Paramo et al., 2004; Moissenet RS218, an isolate from a neonate with meningitis (Xie et al., 2006),
et al., 2010). Only six strains (ECOR42, ECOR31, 126A, DEC9a, 101-1 both of serotype O18:K1:H7. Non-human mammalian ExPEC strains
and DAEC5) were not included in these groups and were classified were also found in subgroup II (O2-type strain, archetypal UTI
as ungrouped, indicating the robustness of the phylogenetic human strain CFT073 of O6-type), subgroup IV (O2-type strain,
classification. The F and B2 groups are the most basal, having archetypal UTI human strain IAI74 of O2-type) and subgroup VII
emerged first, whereas the A and B1/C groups diverged more (O75-type strain, archetypal human UTI strain IH11128 of O75-
recently. Furthermore, a clear genetic structure was identified type). Furthermore, strains of serotype O6:H31, which have been
within the B2 phylogenetic group, with at least nine subgroups (I– reported in dog urinary tract infections (Cherifi et al., 1991; Johnson
IX) in addition to the EPEC 1 cluster commonly represented by the et al., 2001), belong to subgroup III (archetypal human UTI strain 536
O127:H6 strain E2348/69 (Le Gall et al., 2007; Reid et al., 2000). of O6-type) and O4:H5 strains, also isolated in urinary tract
658 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662

Fig. 2. Factorial analysis of correspondence (FAC) of the 234 E. coli strains. Projections of (A) the ExPEC strains (black symbols) of animal (triangle) and human (square) origin,
the non-ExPEC strains (white symbols) and (B) the InPEC strains of animal (triangle) and human (square) origin and the non-InPEC strains (white symbols) onto the F1/F2
plane calculated in the FAC (Greenacre, 1992).

infections in dog and cat (Johnson et al., 2001), belong to subgroup VI relationship was identified between the animal ExPEC strain
(archetypal human UTI strain J96 of O4-type) (Le Gall et al., 2007) (BEN1189) and human strains, although numerous human ExPEC
(data not shown). As expected, no pathogenic animal strain strains belong to this group (Bingen et al., 1998; Picard et al., 1999)
belonged to subgroup VIII, as this clone has been described as (Fig. 3). Conversely, the group C is clonal with short branch lengths
commensal in humans (Clermont et al., 2008). Similarly only human (Fig. 3). Thus, the numerous APEC strains of the O78-type assigned
EPEC strains testing positive for bfpA by PCR were present in the EPEC to phylogroup A by the triplex PCR method (Ewers et al., 2007;
1 cluster. Johnson et al., 2008) probably belong to this clone. One of the group
Non B2 ExPEC animal strains of the O78-type belong to C strains tested, the bovine ExPEC strain 1404, which carries the Vir
phylogenetic groups D (one APEC strain) and C (2 APEC and 2 plasmid, has been shown to be closely related to avian and human
bovine ExPEC strains) (Table 1). Group D is diverse, and no close ExPEC strains, as it belongs to esterase electrophoretic type 2
 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662 659

Fig. 3. Phylogenetic tree of the 234 E. coli strains studied, reconstructed from the partial sequences of 8 housekeeping genes (www.pasteur.fr/mlst) by PHYML (Guindon et al.,
2005) and rooted on E. fergusonii. Bootstrap values are indicated at the corresponding nodes only when they exceed 70%. Pathogenic strains are indicated in bold and boxed in
grey. The name of the strain is followed by H (for human origin) or A (for animal origin). The phylogenetic groups and subgroups are indicated on the right part of the figure.

(Cherifi et al., 1994). The group C strains in our collection also shown to be virulent in a mouse model of septicemia (Picard et al.,
include one human ExPEC strain, ECOR72, and animal and human 1999). One ExPEC strain from pig (5131) and another from sheep
InPEC strains. A case of neonatal meningitis due to a group C strain (S5), belong to the B1 group (Table 1) and appear closely related to
has recently been reported (Moissenet et al., 2010). This strain was human InPEC strains (Fig. 3).
660 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662

Animal InPEC strains were found to belong to phylogenetic associated to the animal pathogenic strains of non-B2 phylogenetic
groups A, B1, C and E. Group E, which includes human and animal groups.
EHEC O157:H7 strains, is clonal, whereas groups A and B1 are more
diverse, with longer branch lengths (Fig. 3). All animal InPEC (ETEC 3.4. Concluding remarks
and EHEC) strains of phylogenetic group A belong to subgroup I,
which also contains human InPEC (DAEC and EAEC with the Our data for a representative set of animal pathogenic strains
archetypal EAEC strain JM221) and ExPEC strains, as well as the analyzed within the context of the overall phylogeny of the E. coli
human commensal laboratory derived strain K-12 (Fig. 3). Within species clearly show that human and animal pathogenic strains
the B1 group, bovine O26-type EHEC (193) and EPEC (C/15333), share common genetic backgrounds, i.e. are very closely related by
together with O45b-type (86-1390) and O15-type EPEC (RDEC-1) MLST analysis. Human and animal strains causing the same disease
strains, are closely related to human O26-type and O111-type in different hosts share a common pool of virulence genes, but a set
EHEC strains. The swine ETEC strain 987 appears closely related to of adhesins specific to animal non-B2 strains was identified.
the human EPEC strain E110019, as are the bovine InPEC 111KH86 It has been suggested that, on many occasions, strains derived
and the human ETEC DEC13a strains (Fig. 3). The rabbit EPEC strain from a common recent ancestor have become specialized for a
E22 (O103-type) is very closely related, with a high bootstrap particular host through subtle genetic changes (Ron, 2006). Our data
value, to the human O103-type EHEC strain 12009, as previously indicate that the gain (or loss) of few genes, as genes coding for
reported (Mariani-Kurkdjian et al., 1993) and to the human O111/ adhesins, could participate to the host specificity. But more subtle
O128-EPEC 2 strains B171 (Rasko et al., 2008), DEC12a and DEC11a changes as single nucleotide polymorphisms (SNPs) in coding or
(Czeczulin et al., 1999; Reid et al., 2000) (Fig. 3). Animal ETEC and regulatory regions, in multiple combinations, could also be involved
EHEC strains were found in group C, together with human InPEC in the host specificity. The recent identification of numerous SNPs in
and animal and human ExPEC strains, as stated above (Table 1). O157:H7 EHEC strains specific for humans or cattle, including a SNP
In sum, MLST analysis corroborates the FAC analysis by showing in the translocated intimin receptor protein (Clawson et al., 2009),
that animal and human pathogenic strains are in most of the cases and of specific gene expression patterns in B2 phylogenetic group
closely phylogenetically related and that the B2 phylogenetic APEC and human urinary tract infection strains in the chicken and
group encompasses mainly ExPEC strains. UTI mouse models (Zhao et al., 2009) is consistent with this
hypothesis. Complete genome sequences for many animal and
3.3. Fine-scale VF pattern analysis human pathogenic strains, generated by ‘‘next-generation’’ sequenc-
ing technologies (MacLean et al., 2009), will facilitate identification
Animal ExPEC strains have a variable pattern of extraintestinal of the subtle genetic elements involved in host specialization.
VFs, whether considered individually or in the pathogenicity island
(PAI) context, even within a clonal lineage (Table 1). Different
Ethical statement
patterns were also observed between human and animal strains of
the same lineage (for example, within B2 subgroups II, VII and IX;
There is no need for ethical statement. The work is based on
Table 1 and Table S1). Furthermore, no particular pattern of VFs
pathogenic animal strains and previously published collection of
was found to be specific to a particular type of extraintestinal
strains. There is no animal model experiment.
disease (e.g. UTI, NBM, septicemia or avian colibacillosis), as
previously reported (Bauchard et al., 2010; Ewers et al., 2004;
Johnson et al., 2007; Kariyawasam et al., 2007; Mokady et al., 2005; Acknowledgments
Moulin-Schouleur et al., 2006). This is in agreement with the recent
findings based on comparative genomics showing that multiple We thank J. Fairbrother, J. Mainil and M. Moulin-Schouleur for
genetic paths of gain and loss of genes can lead to convergent providing us with E. coli strains. ED was supported in part by the
phenotypes (Mokady et al., 2005; Touchon et al., 2009). The ‘‘Fondation pour la Recherche Médicale’’.
intraintestinal VFs studied were pathotype-specific and therefore
did not distinguish between human and animal strains. In a very Appendix A. Supplementary data
small number of strains in our collection, we did not find the
expected pathotype-defining VFs (ST, LT, Bfp). However, in these Supplementary data associated with this article can be found, in
cases, the VFs were plasmid-encoded and the plasmid was the online version, at doi:10.1016/j.meegid.2011.02.005.
probably lost during the many subcultures of the strains (Table 1).
By contrast, adhesins usually described in pathogenic animal References
strains (see Section 2), were found in 17 animal strains, of which 15
are pathogenic, but only in four human strains: the F111 adhesin in Achtman, M., Heuzenroeder, M., Kusecek, B., et al., 1986. Clonal analysis of
Escherichia coli O2:K1 isolated from diseased humans and animals. Infect.
one DAEC strain (DAEC221), one EAEC strain (11074) and one Immun. 51, 268–276.
ExPEC strain (ECOR71), and the F17c/GAF adhesin in one ExPEC Bauchard, P., Germon, P., Brée, A., Oswald, E., Hacker, J., Dobrindt, U., 2010.
strain (IAI64) (Table 1 and Table S1). The expression of the F17c/ Pathogenomic comparison of human extraintestinal and avian pathogenic
Escherichia coli – search for factors involved in host specificity or zoonotic
GAF adhesin, which is specific for terminal N-acetyl-D-glucosamine potential. Microb. Pathog. 49, 105–115.
and blood group M antigen, by human strains (particularly those Bertin, Y., Martin, C., Oswald, E., Girardeau, J.P., 1996. Rapid and specific detection of
isolated from cases of urinary tract infection) is well documented, F17-related pilin and adhesin genes in diarrheic and septicemic Escherichia coli
strains by multiplex PCR. J. Clin. Microbiol. 34, 2921–2928.
although the reported incidence is low (Rhen et al., 1986). By
Bingen, E., Picard, B., Brahimi, N., Mathy, S., Desjardins, P., Elion, J., Denamur, E.,
contrast, to our knowledge, the presence of genes encoding F111 1998. Phylogenetic analysis of Escherichia coli strains causing neonatal menin-
adhesins has not previously been reported in any E. coli strains gitis suggests horizontal gene transfer from a predominant pool of highly
virulent B2 group strains. J. Infect. Dis. 177, 642–650.
pathogenic in humans. As suggested by the FAC (Fig. 1), these
Bingen-Bidois, M., Clermont, O., Bonacorsi, S., Terki, M., Brahimi, N., Loukil, C.,
‘‘animal’’ adhesins were found almost exclusively in non-B2 Barraud, D., Bingen, E., 2002. Phylogenetic analysis and prevalence of urosepsis
strains, as only three of the 21 strains concerned belonged to strains of Escherichia coli bearing pathogenicity island-like domains. Infect.
phylogenetic group B2 (animal ExPEC strain BEN2908, animal Immun. 70, 3216–3226.
Boerlin, P., Travis, R., Gyles, C.L., et al., 2005. Antimicrobial resistance and virulence
commensal strain ASP47b and human ExPEC strain IAI64) (Table 1 genes of Escherichia coli isolates from swine in Ontario. Appl. Environ. Microbiol.
and Table S1). In sum, a set of specific adhesins was preferentially 71, 6753–6761.
 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662 661

Chen, S.L., Hung, C.S., Xu, J., et al., 2006. Identification of genes subject to positive Johnson, J.R., Clermont, O., Menard, M., Kuskowski, M.A., Picard, B., Denamur, E.,
selection in uropathogenic strains of Escherichia coli: a comparative genomics 2006. Experimental mouse lethality of Escherichia coli isolates, in relation to
approach. Proc. Natl. Acad. Sci. U.S.A. 103, 5977–5982. accessory traits, phylogenetic group, and ecological source. J. Infect. Dis. 194,
Cherifi, A., Contrepois, M., Picard, B., Goullet, P., Orskov, I., Orskov, F., De Rycke, J., 1141–1150.
1991. Clonal relationships among Escherichia coli serogroup 06 isolates from Johnson, T.J., Kariyawasam, S., Wannemuehler, Y., et al., 2007. The genome
human and animal infections. FEMS Microbiol. Lett. 64, 225–230. sequence of avian pathogenic Escherichia coli strain O1:K1:H7 shares strong
Cherifi, A., Contrepois, M., Picard, B., Goullet, P., Orskov, I., Orskov, F., 1994. Clonal similarities with human extraintestinal pathogenic E. coli genomes. J. Bacteriol.
relationships among Escherichia coli serogroup O78 isolates from human and 189, 3228–3236.
animal infections. J. Clin. Microbiol. 32, 1197–1202. Johnson, T.J., Wannemuehler, Y., Johnson, S.J., Stell, A.L., Doetkott, C., Johnson, J.R.,
Clawson, M.L., Keen, J.E., Smith, T.P., Durso, L.M., McDaneld, T.G., Mandrell, R.E., Kim, K.S., Spanjaard, L., Nolan, L.K., 2008. Comparison of extraintestinal patho-
Davis, M.A., Bono, J.L., 2009. Phylogenetic classification of Escherichia coli genic Escherichia coli strains from human and avian sources reveals a mixed
O157:H7 strains of human and bovine origin using a novel set of nucleotide subset representing potential zoonotic pathogens. Appl. Environ. Microbiol. 74,
polymorphisms. Genome Biol. 10, R56. 7043–7050.
Clermont, O., Bonacorsi, S., Bingen, E., 2000. Rapid and simple determination of the Kaper, J.B., Nataro, J.P., Mobley, H.L., 2004. Pathogenic Escherichia coli. Nat. Rev.
Escherichia coli phylogenetic group. Appl. Environ. Microbiol. 66, 4555–4558. Microbiol. 2, 123–140.
Clermont, O., Johnson, J.R., Menard, M., Denamur, E., 2007. Determination of Kariyawasam, S., Scaccianoce, J.A., Nolan, L.K., 2007. Common and specific genomic
Escherichia coli O types by allele-specific polymerase chain reaction: application sequences of avian and human extraintestinal pathogenic Escherichia coli as
to the O types involved in human septicemia. Diagn. Microbiol. Infect. Dis. 57, determined by genomic subtractive hybridization. BMC Microbiol. 7, 81.
129–136. Lawrence, J.G., Ochman, H., Hartl, D.L., 1991. Molecular and evolutionary relation-
Clermont, O., Lescat, M., O’Brien, C.L., Gordon, D.M., Tenaillon, O., Denamur, E., 2008. ships among enteric bacteria. J. Gen. Microbiol. 137, 1911–1921.
Evidence for a human-specific Escherichia coli clone. Environ. Microbiol. 10, Le Gall, T., Clermont, O., Gouriou, S., Picard, B., Nassif, X., Denamur, E., Tenaillon, O.,
1000–1006. 2007. Extraintestinal virulence is a coincidental by-product of commensalism
Czeczulin, J.R., Whittam, T.S., Henderson, I.R., Navarro-Garcia, F., Nataro, J.P., 1999. in B2 phylogenetic group Escherichia coli strains. Mol. Biol. Evol. 24, 2373–2384.
Phylogenetic analysis of enteroaggregative and diffusely adherent Escherichia MacLean, D., Jones, J.D., Studholme, D.J., 2009. Application of ‘next-generation’
coli. Infect. Immun. 67, 2692–2699. sequencing technologies to microbial genetics. Nat. Rev. Microbiol. 7, 287–296.
Denamur, E., Picard, B., Tenaillon, O., 2010. Population genetics of pathogenic Mariani-Kurkdjian, P., Denamur, E., Milon, A., et al., 1993. Identification of a clone of
Escherichia coli. In: Robinson DA, F.D., Feil, E.J. (Eds.), Bacterial Population Escherichia coli O103:H2 as a potential agent of hemolytic-uremic syndrome in
Genetics in Infectious Disease. Wiley-Blackwell, Hoboken, pp. 269–286. France. J. Clin. Microbiol. 31, 296–301.
Diard, M., Baeriswyl, S., Clermont, O., Gouriou, S., Picard, B., Taddei, F., Denamur, E., Maynard, C., Bekal, S., Sanschagrin, F., Levesque, R.C., Brousseau, R., Masson, L.,
Matic, I., 2007. Caenorhabditis elegans as a simple model to study phenotypic Lariviere, S., Harel, J., 2004. Heterogeneity among virulence and antimicrobial
and genetic virulence determinants of extraintestinal pathogenic Escherichia resistance gene profiles of extraintestinal Escherichia coli isolates of animal and
coli. Microbes Infect. 9, 214–223. human origin. J. Clin. Microbiol. 42, 5444–5452.
Dow, M.A., Toth, I., Alexa, P., Davies, M., Malik, A., Oswald, E., Nagy, B., 2005. Moissenet, D., Salauze, B., Clermont, O., Bingen, E., Arlet, G., Denamur, E., Merens, A.,
Predominance of afr2 and ral fimbrial genes related to those encoding the K88 Mitanchez, D., Vu-Thien, H., 2010. Meningitis caused by Escherichia coli pro-
and CS31A fimbrial adhesins in enteropathogenic Escherichia coli isolates from ducing TEM-52 extended-spectrum beta-lactamase within an extensive out-
rabbits with postweaning diarrhea in Central Europe. J. Clin. Microbiol. 43, break in a neonatal ward: epidemiological investigation and characterization of
1366–1371. the strain. J. Clin. Microbiol. 48, 2459–2463.
Escobar-Paramo, P., Giudicelli, C., Parsot, C., Denamur, E., 2003. The evolutionary Mokady, D., Gophna, U., Ron, E.Z., 2005. Extensive gene diversity in septicemic
history of Shigella and enteroinvasive Escherichia coli revised. J. Mol. Evol. 57, Escherichia coli strains. J. Clin. Microbiol. 43, 66–73.
140–148. Mora, A., Lopez, C., Dabhi, G., et al., 2009. Extraintestinal pathogenic Escherichia coli
Escobar-Paramo, P., Clermont, O., Blanc-Potard, A.B., Bui, H., Le Bouguenec, C., O1:K1:H7/NM from human and avian origin: detection of clonal groups B2 ST95
Denamur, E., 2004. A specific genetic background is required for acquisition and D ST59 with different host distribution. BMC Microbiol. 9, 132.
and expression of virulence factors in Escherichia coli. Mol. Biol. Evol. 21, 1085– Moulin-Schouleur, M., Schouler, C., Tailliez, P., et al., 2006. Common virulence
1094. factors and genetic relationships between O18:K1:H7 Escherichia coli isolates of
Escobar-Paramo, P., Le Menac’h, A., Le Gall, T., Amorin, C., Gouriou, S., Picard, B., human and avian origin. J. Clin. Microbiol. 44, 3484–3492.
Skurnik, D., Denamur, E., 2006. Identification of forces shaping the commensal Moulin-Schouleur, M., Reperant, M., Laurent, S., Bree, A., Mignon-Grasteau, S.,
Escherichia coli genetic structure by comparing animal and human isolates. Germon, P., Rasschaert, D., Schouler, C., 2007. Extraintestinal pathogenic
Environ. Microbiol. 8, 1975–1984. Escherichia coli strains of avian and human origin: link between phylogenetic
Ewers, C., Janssen, T., Kiessling, S., Philipp, H.C., Wieler, L.H., 2004. Molecular relationships and common virulence patterns. J. Clin. Microbiol. 45, 3366–
epidemiology of avian pathogenic Escherichia coli (APEC) isolated from coli- 3376.
septicemia in poultry. Vet. Microbiol. 104, 91–101. Newton, H.J., Sloan, J., Bulach, D.M., et al., 2009. Shiga toxin-producing Escherichia
Ewers, C., Li, G., Wilking, H., et al., 2007. Avian pathogenic, uropathogenic, newborn coli strains negative for locus of enterocyte effacement. Emerg. Infect. Dis. 15,
meningitis-causing Escherichia coli: how closely related are they? Int. J. Med. 372–380.
Microbiol. 297, 163–176. Ochman, H., Selander, R.K., 1984. Standard reference strains of Escherichia coli from
Feng, P.C., Monday, S.R., Lacher, D.W., et al., 2007. Genetic diversity among clonal natural populations. J. Bacteriol. 157, 690–693.
lineages within Escherichia coli O157:H7 stepwise evolutionary model. Emerg. Ogura, Y., Ooka, T., Iguchi, A., et al., 2009. Comparative genomics reveal the
Infect. Dis. 13, 1701–1706. mechanism of the parallel evolution of O157 and non-O157 enterohemorrhagic
Franck, S.M., Bosworth, B.T., Moon, H.W., 1998. Multiplex PCR for enterotoxigenic, Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 106, 17939–17944.
attaching and effacing, and Shiga toxin-producing Escherichia coli strains from Picard, B., Garcia, J.S., Gouriou, S., Duriez, P., Brahimi, N., Bingen, E., Elion, J.,
calves. J. Clin. Microbiol. 36, 1795–1797. Denamur, E., 1999. The link between phylogeny and virulence in Escherichia
Girardeau, J.P., Lalioui, L., Said, A.M., De Champs, C., Le Bouguenec, C., 2003. coli extraintestinal infection. Infect. Immun. 67, 546–553.
Extended virulence genotype of pathogenic Escherichia coli isolates carrying Pradel, N., Boukhors, K., Bertin, Y., Forestier, C., Martin, C., Livrelli, V., 2001.
the afa-8 operon: evidence of similarities between isolates from humans and Heterogeneity of Shiga toxin-producing Escherichia coli strains isolated from
animals with extraintestinal infections. J. Clin. Microbiol. 41, 218–226. hemolytic-uremic syndrome patients, cattle, and food samples in central
Girardeau, J.P., Dalmasso, A., Bertin, Y., Ducrot, C., Bord, S., Livrelli, V., Vernozy- France. Appl. Environ. Microbiol. 67, 2460–2468.
Rozand, C., Martin, C., 2005. Association of virulence genotype with phyloge- Pupo, G.M., Lan, R., Reeves, P.R., 2000. Multiple independent origins of Shigella
netic background in comparison to different seropathotypes of Shiga toxin- clones of Escherichia coli and convergent evolution of many of their character-
producing Escherichia coli isolates. J. Clin. Microbiol. 43, 6098–6107. istics. Proc. Natl. Acad. Sci. U.S.A. 97, 10567–10572.
Gordon, D.M., Clermont, O., Tolley, H., Denamur, E., 2008. Assigning Escherichia coli Rasko, D.A., Rosovitz, M.J., Myers, G.S., et al., 2008. The pangenome structure of
strains to phylogenetic groups: multi-locus sequence typing versus the PCR Escherichia coli: comparative genomic analysis of E. coli commensal and patho-
triplex method. Environ. Microbiol. 10, 2484–2496. genic isolates. J. Bacteriol. 190, 6881–6893.
Greenacre, M., 1992. Correspondence analysis in medical research. Stat. Methods Reid, S.D., Herbelin, C.J., Bumbaugh, A.C., Selander, R.K., Whittam, T.S., 2000. Parallel
Med. Res. 1, 97–117. evolution of virulence in pathogenic Escherichia coli. Nature 406, 64–67.
Guindon, S., Lethiec, F., Duroux, P., Gascuel, O., 2005. PHYML online – a web server Rhen, M., Klemm, P., Korhonen, T.K., 1986. Identification of two new hemagglutinins
for fast maximum likelihood-based phylogenetic inference. Nucleic Acids Res. of Escherichia coli, N-acetyl-D-glucosamine-specific fimbriae and a blood group
33, W557–559. M-specific agglutinin, by cloning the corresponding genes in Escherichia coli K-
Imberechts, H., Van Pelt, N., De Greve, H., Lintermans, P., 1994. Sequences related to 12. J. Bacteriol. 168, 1234–1242.
the major subunit gene fedA of F107 fimbriae in porcine Escherichia coli strains Ron, E.Z., 2006. Host specificity of septicemic Escherichia coli: human and avian
that express adhesive fimbriae. FEMS Microbiol. Lett. 119, 309–314. pathogens. Curr. Opin. Microbiol. 9, 28–32.
Jaureguy, F., Landreau, L., Passet, V., et al., 2008. Phylogenetic and genomic diversity Savageau, M.A., 1983. Escherichia coli habitats, cell types, and moleculat mechan-
of human bacteremic Escherichia coli strains. BMC Genomics 9, 560. isms of gene control. Am. Nat. 122, 732–744.
Johnson, J.R., Delavari, P., Stell, A.L., Whittam, T.S., Carlino, U., Russo, T.A., 2001. Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: Molecular Evolutionary
Molecular comparison of extraintestinal Escherichia coli isolates of the same Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599.
electrophoretic lineages from humans and domestic animals. J. Infect. Dis. 183, Tenaillon, O., Skurnik, D., Picard, B., Denamur, E., 2010. The population genetics of
154–159. commensal Escherichia coli. Nat. Rev. Microbiol. 8, 207–217.
662 O. Clermont et al. / Infection, Genetics and Evolution 11 (2011) 654–662

Touchon, M., Hoede, C., Tenaillon, O., et al., 2009. Organised genome dynamics in Xie, Y., Kolisnychenko, V., Paul-Satyaseela, M., Elliott, S., Parthasarathy, G., Yao, Y.,
the Escherichia coli species results in highly diverse adaptive paths. PLoS Genet. Plunkett 3rd, G., Blattner, F.R., Kim, K.S., 2006. Identification and characteriza-
5, e1000344. tion of Escherichia coli RS218-derived islands in the pathogenesis of E. coli
Wirth, T., Falush, D., Lan, R., et al., 2006. Sex and virulence in Escherichia coli: an meningitis. J. Infect. Dis. 194, 358–364.
evolutionary perspective. Mol. Microbiol. 60, 1136–1151. Zhao, L., Gao, S., Huan, H., Xu, X., Zhu, X., Yang, W., Gao, Q., Liu, X., 2009. Comparison
Wu, G., Carter, B., Mafura, M., Liebana, E., Woodward, M.J., Anjum, M.F., 2008. of virulence factors and expression of specific genes between uropathogenic
Genetic diversity among Escherichia coli O157:H7 isolates and identification of Escherichia coli and avian pathogenic E. coli in a murine urinary tract infection
genes linked to human infections. Infect. Immun. 76, 845–856. model and a chicken challenge model. Microbiology 155, 1634–1644.