Nealson and Hastings

MICROBIOLoGIcAL REVIEWS, Dec. 1979, p.
496-518
0146-0749/79/04-0496/23$02.00/0
Vol. 43, No.4
Bacterial Bioluminescence: Its Control and Ecological
Significance
K. H. NEALSON'* AND J. W. HASTINGS2
Marine Biology Department, Scripps Institution of Oceanography, La Jolla, California 92093' and The
Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138'
INTRODUCTION
... 496
BIOCH EMI STRY
497
TAXONOMIC RELATIONSHPS OF LUMINOUS BACTERIA
................. 499
Marine Forms ............................................................. 499
Terrestrial and Freshwater Forms
.......................................... 501
CONTROL OF THE SYNTHESIS AND ACTIVITY OF THE LUMINESCENT
SYSTEM
50 1
Autoinduction
...... 501
Catabolite Repression
...................................................... 503
......................................... 503
Composition of the Growth Medium
Oxygen
.......... 504
Pyruvate and Excretion Products
....................
................. 504
Dark (K) Variants of Luminous Bacteria ........ 504
HABITATS AND DISTRIBUTION OF LUMINOUS BACTERIA
............... 505
Free Living ................................................................. 505
................................... 507
Saprophytic Forms: Natural Enrichments
Commensal Forms
.......................................................... 508
508
Parasites
Marine
508
508
..
Nonmarine .
..................................................... 509
Light Organ Symbionts
Teleost fishes ............................................................. 509
Squid
5 10
............................ ..................... 510
Nonculturable symbionts
512
Endosymbionts
.............................. ......... 512
FUNCTIONS OF BIOLUMINESCENCE
...................................... 512
Vestigial or Nonfunctional Hypotheses
513
Biochemical Functions
........................................................ 513
Biological Functions
................................................ 513
light organ luminescence
.................................
...... 513
Dispersion and propagation
............ 514
PERSPECTIVES: LUMINOUS AND NONLUMINOUS BACTERIA
......................................................... 515
LITERATURE CMD
tional studies. It is now clear that light organ

symbioses represent specific advantageous asTen years ago there existed very little data sociations between luminous bacteria and fish
relevant to the ecology of the bioluminescent and that some species of luminous bacteria are
bacteria and the functional importance of light not found as light organ symbionts. This has
emission. Although this lack of data did not raised the question of whether the light from
dampen enthusiasm for speculation as to the bacteria that are not light organ symbionts
function(s) and selective advantage of light emis- might have adaptive functions, such as enhancsion in bacteria, it did make the formulation of ing the dispersion and propagation of these bachypotheses difficult. Over the past decade, how- teria.
Likewise, the results of investigations into the
ever, a considerable amount of new knowledge
has accumulated, the synthesis and analysis of mechanisms controlling bioluminescence that
which provide a foundation for new perspectives were directed at physiological questions have
ecological implications, the most important of
and postulates.
The results of taxonomic studies that were which is that these bacteria may be luminous
directed primarily at problems of systematics under some conditions and nonluminous under
have ecological implications and have provided others; i.e., the system is inducible and represthe tools for extensive ecological and distribu- sible.
INTRODUCTION
496
VOL. 43, 1979
These perspectives and other specific postulates presented here concerning the nature and
functions of bacterial bioluminescence provide
new approaches for data collection and experimental work.
BACTERIAL BIOLUMINESCENCE
497
lished data). All exhibit high specificity for

FMNH2 and a long-chain aliphatic aldehyde (8
to 16 carbons) (41, 97). For luciferases from
different strains of a given species, the kinetics
of decay with a given aldehyde are similar (J.
M. Fitzgerald, Ph.D. thesis, Monash University,
Victoria, Australia, 1979; Nealson, unpublished
data); with dodecanal as the aldehyde in the in
vitro assay, all luciferases from Photobacterium
isolates so far examined exhibit fast decay,
whereas all luciferases from Beneckea exhibit
slow kinetics (Fig. 1). The turnover rates of the
luciferases thus form the basis for rapid distinction among the major bacterial groups (64).
The subunit structure of luciferase, as exemplified by the luciferase of B. harveyi, is a heterodimer having a molecular weight of 79,000,
whose two nonidentical subunits (alpha, 42,000
daltons; beta, 37,000 daltons) can be isolated in
quantity by diethylaminoethyl-Sephadex chromatography in 5 M urea. Individual subunits are
inactive but can be recombined to yield a recon-
BIOCHEMISTRY
The biochemistry of bacterial bioluminescence has been reviewed recently (40); the short
summary below serves to introduce the subject
and to provide a perspective for the subsequent
sections.
The light-emitting reaction of luminous bacteria involves a luciferase-catalyzed oxidation of
reduced flavin mononucleotide (FMNH2) by
molecular oxygen, with the concomitant oxidation of a long-chain aliphatic aldehyde, probably
tetradecanal (93). Although the biochemical system is unique to bacteria, it is nevertheless found
in higher organisms by virtue of the occurrence
of luminous bacteria in mutualisms of several
types (see below). Recent work has also shown
that bacterial luciferase is in fact not confined to
100
the bioluminescent species of bacteria; many
FMNH2+ 02+ RCHO LUCIFERASE
closely related nonluminous species contain luFMN + H20 + RCOOH + 2hvu50
ciferase in low but readily detectable amounts
(K. H. Nealson and D. S. Walton, Abstr. Annu.
Meet. Am. Soc. Microbiol. 1978, 1131, p. 102).
Luciferase can be viewed as a mixed-function
20
oxidase, involving the oxidations of FMNH2 and
a long-chain aliphatic aldehyde in a reaction
that is analogous to the reactions of certain
B\\dhorveyi k=0.12 sec'
\
flavoprotein hydroxylases (37). An unusual feaS splendido k= 0.11 sec'1
ture of the reaction is its inherent slowness; at czI5
room temperature the time required for a single
luciferase cycle to occur is on the order of 10 s
(38), making it one of the slowest enzymes
- '0
P phosphoreum k= 0 78 secc
known. A practical consequence of this is that
there are intermediates in the reaction with long
lifetimes. Indeed, by using low temperatures, z 10
P /elognothi k= 0 64 sec'
these have been isolated, purified, and charac- J
terized (2, 9, 37a); the initial chemical species is
P. logel k 0.94 sec
5
postulated to be a luciferase-bound 4a-peroxy
P fischeri k= 0 sec1
adduct of FMNH2 (23).
Luciferases from several different luminous
t
bacteria have been isolated, purified, and stud3U
25
C
20
10
15
3
T ME (seconds)
ied; these bacteria include five marine species
(Beneckea [Vibrio] harveyi, Beneckea splenFIG. 1. Luciferase kinetics in the nonturnover asdida, Photobacterium fischeri, Photobacterium say. The kinetics of the decay of luminescence for the
phosphoreum, and Photobacterium leiognathi) luciferase reaction in vitro (the reaction sequence is
and two terrestrial-freshwater species (Vibrio shown in the upper right) when initiated with FMNH2
cholerae biotype albensis and Xenorhabdus lu- in the presence of dodecanal are shown for luciferases
from several species of marine bacteria. The
minescencens) (See below). Although there are isolated
apparent first-order rate constants (k) for the decay
significant differences, all bacterial luciferases of
luminescence are indicated for two species not
possess a heterodimeric structure and have mo- graphed. All Beneckea species display slow kinetics,
lecular weights in the range 76,000 + 4,000 (40; whereas all Photobacterium species show fast kinetE. G. Ruby and J. W. Hastings, unpublished ics. RCHO, Long-chain aliphatic aldehyde; FMN,
data; K. Kopecky and K. H. Nealson, unpub- flavin mononucleotide.
x
ll.x
498
HASTINGS AND NEALSON
stituted luciferase with activity and properties

like those of the native material. From mutant
analyses (15, 16) and chemical modification
studies (60) it is postulated that the catalytic
properties reside on the alpha subunit. Although
the beta subunit is required for luciferase activity, its specific function has not been elucidated.
Recent determinations of the amino acid sequences of the luciferase subunits of P. fischeri
and B. harveyi indicate clear relatedness between the luciferases of the two genera and also
between the two subunits (1). From these data
it is possible to postulate that the luciferases in
these two species evolved from the same monomer. Complementation studies with subunits
from different species, coupled with amino acid
sequence data, should lead to insight into the
evolution of this protein. For example, active
hybrid luciferases have been formed by renaturing the subunits from the three different Photobacterium species (Ruby and Hastings, unpublished data). All combinations except P. fischeri alpha subunit with P. phosphoreum beta
subunit give active hybrids. In the hybrids, the
properties related to catalytic function (turnover
time) are determined by the species contributing
the alpha subunit. Similarly, subunits ofB. harveyi luciferase form active hybrids with those
from both V. cholerae biotype albensis and X.
luminescens (Kopecky and Nealson, unpublished data). No active hybrids between Photobacterium and Beneckea luciferase subunits
have yet been reported.
The overall reaction of bacterial luciferase
results in the production of light. As Fig. 2 shows,
the emitter (VI*) is a luciferase-flavin complex,
but its specific structure is not yet known. Aldehyde is not required in the light-emitting reaction until the peroxy-FMNH2 (II) has been
formed; in the absence of aldehyde (in vitro) this
intermediate (II) breaks down without significant light emission. Mutants in which the synthesis of natural aldehyde is blocked are nonluminous, but normal or even higher levels of
active luciferase occur (40). Whether electrons
are shunted via this dark pathway in such mutants is not known. In the presence of aldehyde,
and presumably in vivo, the light-emitting reaction results in the concomitant formation of
the corresponding long-chain fatty acid. Evidence both in vivo (93) and in vitro (59) indicates
that this fatty acid is reduced back to the aldehyde (Fig. 2).
In cells the bacterial bioluminescent system
can be viewed as a branch (at the level of pyridine nucleotides) of the electron transport pathway, in which electrons from reduced substrates
are shunted via reduced flavin to oxygen (Fig.
MICROBIOL. REV.
2). The affinity of the system for oxygen is very

high; at very low oxygen concentrations, where
cytochrome electron flow may be completely
blocked, luminescence can occur without appreciable diminution (32). Under microaerophilic
conditions, such a pathway might be advantageous (see below). Since the emission of light
requires molecular oxygen, during strict anaerobiosis luminescence ceases, and upon readmission of oxygen an excess flash of luminescence
occurs. This can now be interpreted to be the
result of the accumulation of luciferase-reduced
flavin complex I (L-FH2 in Fig. 2).
In very bright cells growing in a rich medium,
it has been estimated that luciferase accounts
for as much as 5% of the cellular protein (40a)
and that luminescence accounts for as much as
20% of the total oxygen consumption (32). The
luminescence system may thus represent a significant energetic drain. However, in minimal
media, when luminescence is much less, the
energetic requirements are also much less. It has
been estimated, on the basis of energy charge
analysis and light measurements (47), that for
very bright strains the energy committed to
bioluminescence is as much as 10% of the total,
Substrcte
dJbstrcte
educed
A TP
NV
--
JADH
NAD
K-
cytochromees
I7
FMN-
----FMN
+
(I)
H20
H20
[FH
FMNH
aTP
/l02
FH2
NV
NA
PRODUCTS
-- RCOOH
(It)
2
2
rnf46 >^+
2
2)PCFH4O'
-
tF1:
(VI*)
,le
noX
Ideyly
L + FMN + H 202
-_
C 2 2
RCHC -
(IM)
22,
lq-)
FIG. 2. Pathway of bacterial bioluminescence. The
pathway of reactants and proposed intermediates is
shown in relation to cellular metabolism and electron
flow. For each cycle of light production, one reduced
flavin (FMNH?) and one reduced pyridine nucleotide (reduced nicotinamide adenine dinucleotide
[NADH2i, used for aldehyde regeneration) are presumably required. NAD, Nicotinamide adenine dinucleotide; ATP, adenosine triphosphate; FMN,
flavin mononucleotide; RCHO, long-chain aliphaic
aldehyde.
VOL. 43, 1979
499
fructose, glycerol, and N-acetylglucosamine, and

all produce an extracellular chitinase (90).
The current classification and division into
genera are based on (i) the mode of flagellation,
(ii) the moles percent guanine plus cytosine content of the DNA, (iii) in vitro deoxyribonucleic
acid (DNA)-DNA hybridization and DNA-ribosomal ribonucleic acid hybridization, and (iv)
nutritional and enzymatic properties. Based on
TAXONOMIC RELATIONSHIPS OF THE these criteria, four species are placed in the
genus Photobacterium and two are placed in
LUMINOUS BACTERIA
Beneckea. All luminous Photobacterium speMarine Forms
cies, except the newly described Photobacterium
Over the past decade, the taxonomic status of logei, are known to occur in specific association
the marine luminous bacteria has been clarified with a higher organism, whereas no Beneckea
(6-8, 78, 80), allowing for a better appreciation species have been found as light organ symof the key physiological and ecological proper- bionts.
P. fischeri cells are yellow pigmented and rod
ties of the different bacterial groups (40). This
work has resulted in a wider recognition of the shaped and have a tuft of sheathed flagella (Fig.
fact that a specific bacterial taxon may have 3). They are restricted to the marine environment and have a specific requirement for sodium
some members that exhibit bioluminescence and
ion for growth (79). They occur as free-living
others that do not (Table 1).
As shown in Table 1, there are six recognized forms and as specific symbionts of monocentrid
marine luminous species, all gram-negative mo- fishes (21, 86).
P. logei is a recently identified species closely
tile rods that share a number of morphological,
physiological, and biochemical properties with related to P. fischeri (3). This species is adapted
the Enterobacteriaceae and the Vibrionaceae. to colder waters, showing an ability to grow at
They are all facultative anaerobes capable of 4C and an inability to grow at 30C. In most
growth at 25C on glucose, mannose, galactose, other aspects it is similar to P. fischeri.
whereas for other bacteria (normal luminous

forms) the energetic drain is much less (0.001%
or less). Understanding the energetic requirements of the synthesis and activity of the luminous system under different conditions will be
required for understanding the relationship between the biochemistry of the system and its
ecological significance.
TABLE 1. Luminous and nonluminous species of luminous genera

Nonluminous species
Habitat
Genus
DNA guanine
plus cytosine
Luminous
content
species
Ability to
(mol %)
Name
induce lu- Production

minescence SYStem in B.
of bacterial
luciferaseb
harveyia
Marine
Nonmarine
Photobacterium
39-44
P. fischeri, P. logei, P.
phosphoreum, P.
P. angustum
Beneckea (Vibrio)Y
45-48
B. harveyi, B. splendida
B. alginolytica
B. parahaemolytica
B. anguillara
B. natriegens
B. campbelli
B. vulnifica
B. nigrapulchrituda
B. pelagia
B. nereida
V. cholerae
+
+
+
+
+
+
+
+
+
+
NTd
NT
NT
NT
leiognathi
Vibrio
47-49
Xenorhabdus'
43-44
(Achromobacter)
V. cholerae biotype albensis

X. luminescens strain
Hb (ATCC 29999),
strains ATCC 19061,
X-II, All, Dn, R
X. nematophilus NC19 (ATCC 129304)

Hb-B, Hm
See reference 25.

Nealson and Walton, Abstr. Annu. Meet. Am. Soc. Microbiol. 1978, 1131, p. 102.
The nomenclature at the generic level for these species is still unsettled. For a discussion of this, see reference 7.
d
NT, Not tested.
'Achromobacter is an earlier, now obsolete assignment. See reference 92 for full data on the genus Xenorhabdus.
b See
500
P. phosphoreum and P. leiognathi, which are

taxonomically similar, may also be distinguished
by differences in optimal growth temperatures.
P. phosphoreum is enriched for by incubation at
low temperatures (4C), and its growth is often
inhibited by temperatures above 25C. P. leiognathi strains grow well in laboratory culture at
temperatures up to 35C and are commonly
found in warm tropical waters. Both species are
obligate marine forms with specific requirements
for sodium ion. They are short rods with one to
three unsheathed polar flagella and, when grown
on glucose, possess bright refractile granules of
poly-beta-hydroxybutyric acid. Both species
also occur as specific light organ symbionts, P.
leiognathi with members of the family Leiognathidae (82) and P. phosphoreum with a variety of midwater and deepwater fishes (45, 84).
The luminous Beneckea species are quite similar and were separated into two taxa only after
extensive nutritional analyses (78). B. harveyi
cells are straight rods, whereas B. splendida
FIG. 3. Electron micrographs of negataively

stained preparations of luminous bacteria. Bars =
0.5 ,um (A) B. harveyi, showing a single sheathed,
polar flagellum and unsheathed peritrichous flagella. (B) P. fischeri, showing a tuft of sheathed
flagella. (C) P. phosphoreum, showing a tuft of unsheathed flagella. (D) P. leiognathi, showing a single
unsheathed polar flagellum. (These pictures were
supplied by P. Baumann. Used with permission; from
reference 78.)
MICROBIOL. REV.
cells are more curved; both have one to three

sheathed polar flagella. Many but not all strains
of B. harveyi have sheathed peritrichous flagella
that are induced by growth on solid media (Fig.
1). Both species have specific growth requirements for sodium ion. B. harveyi occurs in surface and coastal seawaters in relatively high
numbers in a seasonal fashion (88; E. G. Ruby,
Ph.D. thesis, Scripps Institution of Oceanography, La Jolla, Calif., 1977), whereas the distribution of B. splendida has not been documented. Many nonluminous strains that are taxonomically indistinguishable by DNA-DNA hybridization from B. harveyi have been isolated
from seawater and fish surfaces (8, 78). It is
possible that these strains contain very low levels of luciferase, as has been found in some
nonluminous species (Nealson and Walton,
Abstr. Annu. Meet. Am. Soc. Microbiol. 1978),
but this question has not been examined.
No Beneckea species have been isolated as
specific light organ symbionts, whereas with the
exception of P. logei, all luminous Photobacterium species have been. This raises the possibility
that luminous members of the genus Beneckea
may be classed as free living and that Photobacterium species may be classed as symbionts (39,
40). Among the taxonomic features that distinguish the two genera are large differences in
nutritional versatility. The Beneckea strains are
more nutritionally versatile, capable of utilizing
28 to 45 (depending on the strain) of 147 different
organic compounds as the sole carbon and energy source, whereas the Photobacterium
strains, including most of those isolated directly
from light organs, can utilize only between 7 and
22 of these compounds (78). This suggests that
nutritional factors are not involved in strain
selection and maintenance in the light organs
and that the light organ is not the only niche
available for the symbiotic species. The fact that
Beneckea isolates are common in surface and
coastal waters (and not deeper) (88, 100; Ruby,
Ph.D. thesis) may be related to nutritional factors, but much more information is needed. For
instance, the capacity of these species to grow
under low-nutrient (oligotrophic) conditions has
not yet been studied in detail.
Of 150 isolates from the light organ of Monocentris japonica, all were P. fischeri and grew
well on a defined minimal medium without supplements (86). Similarly, all 723 P. leiognathi
isolates from many leiognathid fishes (82) and
152 P. phosphoreum isolates from midwater macrourids and opisthoproctids (84) were prototrophs. On the other hand, naturally occurring
auxotrophs of P. phosphoreum do occur, as first
reported by Doudoroff (19). Reichelt and Baumann (78) found that about 45% of 78 isolates of
VOL. 43, 1979
P. phosphoreum from seawater samples echibited some growth requirement, most commonly
methionine. Of 500 P. phosphoreum strains isolated from the Atlantic Ocean at depths between
200 and 1,000 m Ruby et al. (E. G. Ruby, E. P.
Greenberg, and J. W. Hastings, Appl. Environ.
Microbiol., in press) found that 117 would not
grow on minimal medium. Of these, 52 were
chosen at random and tested; 51 grew on minimal medium supplemented with methionine.
However, Ruby and Nealson (88) found no
strains that required growth factors among approximately 2,300 fresh isolates from coastal waters at San Diego, Calif. Only 16 of these were
P. phosphoreum; 651 were B. harveyi, and 1,601
were P. fischeri.
As mentioned above, a useful distinction between the two genera relates to the kinetics of
the reaction of the luciferase when assayed in
vitro. Although the luciferases from all species
require the same components, the decay kinetics
(enzyme turnover time) for the reaction in vitro
using dodecanal are slow for luciferases from
Beneckea and fast for those from Photobacterium (Fig. 3). Thus, the luciferase assay can be
used for rapid screening at the generic level.
Since this enzymatic pathway is apparently
unique to bacteria, the existence of bacterial
luciferase can be used as evidence for the presence of luminescent bacteria, even when they
cannot be cultured (D. Cohn, G. Leisman, and
K. H. Nealson, Abstr., Annu. Meet. Am. Soc.
Photobiol. 1979). The reaction has also provided
evidence for the presence of a functional luciferase gene even when luminescence in vivo is
not detectable (Nealson and Walton, Abstr.
Annu. Meet. Am. Soc. Microbiol. 1978).
Terrestrial and Freshwater Forms
The existence of nonmarine luminous bacteria, in particular the occurrence of a freshwater
strain of Vibrio albensis (now called Vibrio
cholerae, sometimes called V. cholerae biotype
albensis), is indicated from the earlier literature
(32). This strain has a low (50 mM) requirement
for sodium ion and a DNA guanine plus cytosine
content of 47.8 mol%; DNA-DNA hybridization
studies indicate that it hybridizes 100% with the
type strain of V. cholerae, whereas Beneckea
strains hybridize poorly with V. cholerae (80).
However, luciferase from this strain exhibits Beneckea-type (slow) kinetics (Kopecky and Nealson, unpublished data). Neither the distribution
nor the physiology of luminous members of this
species has been studied.
Recently, luminescence has been reported to
occur in some members of a group of bacteria
which are symbiotic with soil nematodes and
collaborate with the nematodes in the parasiti-
501
zation of insects (52, 76; see below). The biochemistry of light emission is similar to that of
the marine forms, and the reaction exhibits slow
(Beneckea-like) kinetics (75). A new genus (Xenorhabdus) with luminous (X. luminescens) and
nonluminous (X. nematophilus) members has
been created to accomodate these bacteria (Table 1) (92). Members of this genus have a DNA
guanine plus cytosine content of 43 to 44 mol%
and share many properties with other members
of the Enterobacteriaceae, in which group the
genus is placed. The cells are unusually large for
luminous bacteria (length, 5 to 10 ,um), pigmented, and chitinase and oxidase negative, and
they prefer low salt concentrations; growth is
inhibited at 3% NaCl. Both luminous and nonluminous species produce antibiotic-like substances that presumably are responsible for the
fact that infected dead insects do not putrefy
(52). Several of these inhibitory compounds have
been isolated and purified from the growth medium (B. Fenical, personal communication).
CONTROL OF THE SYNTHESIS AND

ACTIVITY OF THE LUMINESCENT
SYSTEM
Knowledge of control mechanisms is important to the understanding of the selective advantage of luminescence and the ecology of the
bacteria. Most of the effects elucidated involve
control of synthesis, and up to now there is no
case known in which luciferase synthesis is affected without a concomitant effect on the synthesis of aldehyde (e.g., the biosynthesis of luciferase and the enzymes involved in aldehyde
production are controlled in parallel). Thus, if
luciferase synthesis is low, its ability to act in
vivo is also lowered by virtue of substrate (al-
dehyde) limitation.
Autoinduction
The most obvious and perhaps the most significant physiological control of light emission
involves control of the synthesis of the luminescent system by the bacteria themselves. This
has been referred to as autoinduction. It was
first described by Nealson et al. (69) and was
postulated to involve a substance (autoinducer)
that accumulates in the growth medium and
induces the synthesis of the components of the
luminescence system. This hypothesis is supported by the demonstration that both P. fischeri and B. harveyi cells produce a substance
which induces bioluminescence at low cell density and which may be isolated and concentrated
from the culture medium (20, 25, 63). Although
the autoinducer is apparently a different compound in Photobacterium and Beneckea species
502
MICROBIOL. REV.
(the substances do not cross-react), many nonluminous Beneckea species produce autoinducer
activity in large amounts (some more than B.
harveyi) which induces the luminescent system
in B. harveyi (25). Several strains of P. fischeri
have been found to differ in the amount of
autoinducer produced (63); the amount of luminescence in vivo is correlated with autoinducer
levels.
In a previous review (40) there was a report
based on unpublished data that autoinducer
from P. fischeri had a molecular weight of about
159 and properties similar to those of indole
acetaldehyde. This is now believed to have been
in error; the autoinducer was a minor contaminant in the fraction under study (A. Eberhard,
personal communication). Separation of the active fraction has now been achieved, and its
structure determination is under way (Eberhard,
personal communication).
Support for the autoinduction hypothesis has
been obtained recently from two types of studies.
In the first, P. fischeri cells were grown in a
chemostat in which the cell density was regulated by carbon (glycerol) limitation (R. A. Rossen and K. H. Nealson, Abstr. Annu. Meet. Am.
Soc. Photobiol. 1979, p. 150-151). At steadystate cell densities of 108 ml-' or greater, the
cells remained brightly luminous at an intensity
cells/ml
fischer,
per cell independent of cell density (Fig. 4A).

Below this density the luminescence per cell
dropped sharply (by a factor of 100 with a 10fold decrease in cell number). At cell densities
below 5 x 106 cells per ml no light emission was
detectable. When purified autoinducer was then
added, the luminescence increased by several
orders of magnitude within hours (Rosson and
Nealson, Abstr. Annu. Meet. Am. Soc. Photobiol. 1979). A different approach with B. harveyi
provided a similar conclusion (J. W. Hastings
and S. Ulitzur, Abstr. Annu. Meet. Am. Soc.
Photobiol. 1979, p. 66). By repeated dilution,
cells were maintained at densities below 10 ml-1;
luminescence per cell rapidly decreased to less
than 1% of that seen after induction (Fig. 4B).
Autoinduction thus occurs in members of two
different marine genera and may occur in all,
since similar density-dependent effects upon the
development of luminescence have been observed in all luminous species. However, it has
been reported that strains of both P. phosphoreum (96) and P. leiognathi (48) exhibit constitutive luciferase synthesis, and autoinducer has
not actually been isolated from either of these
species or from Xenorhabdus species.
It thus appears that at low cell densities or
under ecological conditions where the autoinducer cannot accumulate, luminous bacteria are
light/cell
o4
109
o
u-t
is
a':iV
ax
41~~~~~~~~~~~~~~~~i
~.i
Time (days)
i0s
Time
doays)
1i
12
FIG. 4. Loss of luminescence in cells growing at low cell densities. (A) Chemostat experiment with P.
fischeri. Each day the carbon (glycerol) concentration was lowered by a factor of two, and the cells of P.
fischeri were allowed to grow at that concentration. When the cell density dropped below about 108 cells per
ml, the luminescence per cell decreased; at a cell density of 3 x 106 cells per ml, light emission was
undetectable, i.e., less than 10 quanta/s per cell. Addition of purified autoinducer (arrow) restored bioluminescence in the culture at this density. (B) Cells of B. harveyi were maintained at low density by repetitive
subculturing rather than in a chemostat; no nutrient limitation was used. When the culture reached a density
of about 107 cells per ml, it was diluted to approximately 102 cells per ml in the same medium. After the first
dilution, luminescence per cell fell to and remained at a low level through all subsequent dilutions. Either the
addition of inducer (arrow A) or allowing the cells to continue growth (arrow B) resulted in the induction of
bioluminescence.
VOL. 43, 1979
repressed for luciferase synthesis and are thus

nonluminescent. This consideration is especially
applicable to the case of bacteria growing unconfined, as, for example, free in seawater. Kelly
and Tett (49) have proposed that luminous bacteria might provide a background glow in the
ocean. Direct observations are lacking, and this
would not be expected to occur unless the bacterial counts are very high, the water contains
autoinducer, or a bacterial strain that does not
require autoinducer is involved. One test of the
hypothesis was reported by Booth and Nealson
(C. K. Booth and K. H. Nealson, Biophys. J. 15:
53a, 1975). Light intensity measurements of seawater samples known to contain luminous bacteria (by subsequent plating) showed that the
bacteria were nonluminous or that the luminosity was at least 100-fold less than the light intensity of the same cells after growth in the laboratory.
On the other hand, autoinduction presumably
occurs for associated bacteria, whether they are
light organ symbionts, parasites, gut bacteria, or
saprophytes. In such cases, cell density becomes
high so that autoinducer can accumulate to levels sufficient for the induction of the luminescence system. It should be stressed here that
since the control by autoinduction is at the level
of transcription (63, 69), response to autoinducer
can be measured only after some minutes. Thus,
it is not a mechanism that a host fish might
exploit for the immediate control of luminescence from the light organ. On the other hand,
for long-term situations, such as conserving energy during day-night cycles, it is not inconceivable that the level of bacterial luminescence
could be modulated by a host fish interacting at
the level of autoinduction. Such ideas have never
been tested.
Catabolite Repression
Inducible enzymes are synthesized and have
a function under some conditions but not under
others (56). For many such enzymes induction is
mediated by some specific nutrient for which
catabolic enzymes are not produced constitutively. Moreover, the induced synthesis of the
relevant enzyme is often repressed by glucose,
even in the presence of inducer, and this repression can be overcome by exogenous cyclic adenosine monophosphate (cAMP); glucose repression and reversal by cAMP are referred to as
catabolite repression. The synthesis of the luminescence system in B. harveyi is subject to
catabolite repression (66). Furthermore, cAMP
binding protein occurs in large amounts in B.
harveyi, and it is immunologically homologous
with the cAMP binding proteins of Escherichia
coli and several other enteric bacteria (73). By
503
analogy, luciferase in B. harveyi has been hypothesized to be a nonessential enzyme, which

is functional under certain, presumably advantageous, conditions and repressed under others.
B. harveyi mutants that are resistant to catabolite repression have been isolated as "bright on
glucose" phenotypes, but are not yet well characterized (P. Lin, M. H. Saier, Jr., and K. H.
Nealson, Fed. Proc. 35:1361, 1976; K. H. Nealson and J. W. Hastings, Bacteriol. Proc., p. 219,
1970). Ulitzur and Yashphe (94) isolated a mutant that is nonluminous unless supplied with
exogenous cAMP. As expected, the mutation is
pleiotropic; the strain is deficient for a variety of
other functions for which cAMP is normally
required in enteric bacteria (56).
The Photobacterium species are quite different, but not yet fully characterized. Henry and
Michelson (42) reported that a strain of P. leiognathi displays a strong permanent glucose
repression that is not reversed by cAMP. Makemson (57) reported glucose repression that is
not reversed by cAMP in P. phosphoreum. P.
fischeri exhibits a transient repression of the
development of luminescence when switched
from glycerol- to glucose-containing media. This
effect is not reversible by cAMP, and once the
cells are glucose adapted, no further effect is
seen (86). Xenorhabdus species have not been
studied in this regard.
Composition of the Growth Medium

In addition to control by catabolite repression,
the synthesis of the luminescence system can be
dramatically affected by the composition of the
growth medium. In B. harveyi, growth in a minimal medium with glycerol, which does not repress the luminous system, is much slower than
growth in a complete medium, and the bioluminescence is severely repressed. Although a lag
in the development of luminescence that is characteristic of autoinduction occurs, only very
small amounts (1% or less) of luciferase are
actually produced (69). Under these conditions,
additional autoinducer does not increase luciferase synthesis or luminescence, but arginine
and the salt (NaCl) concentration do have pronounced effects.
Added arginine dramatically stimulates the
development of luminescence of B. harveyi
growing in a minimal medium (17, 69, 98). Citrulline and argininosuccinate (the immediate
biosynthetic precursors of arginine) also stimulate luminescence, as does proline, but to a lesser
degree; the mode of action is not known. In
addition, B. harveyi is considerably brighter
when grown in minimal medium with low salt
(1 instead of 3% NaCl) (K. H. Nealson, Ph.D.
thesis, University of Chicago, Chicago, Ill., 1969;
504
C. A. Waters, Ph.D. thesis, Harvard University,

Cambridge, Mass., 1974). In some nonluminous
Beneckea species that contain luciferase, its synthesis is also enhanced by growth in a medium
with low salt (Nealson and Walton, Abstr. Annu.
Meet. Am. Soc. Microbiol. 1978, I131, p. 102).
Although the salt effect has not been investigated in detail, it may be related to the arginine
effect, since cells grown in low salt exhibit less
stimulation by arginine. Furthermore, mutants
that have escaped the arginine requirement and
are thus bright when grown on a minimal medium without arginine exhibit little or no enhancement of luminescence when grown in low
salt. Neither the salt effect nor the arginine
effect occurs in a complete medium (Nealson,
Ph.D. thesis; Waters, Ph.D. thesis).
Whether these phenomena occur in the other
luminous species and, if so, to what extent is not
known. From recent studies it appears that light
emission per cell is also quite low for P. leiognathi grown in a minimal medium (Rossen and
Nealson, Abstr. Annu. Meet. Am. Soc. Photobiol. 1979) and that this is a species that grows
well under conditions of low nutrients (100).
MICROBIOL. REV.
may lead to significant advances in the understanding of the ecology of these bacteria.
Pyruvate and Excretion Products
Oxygen
When P. fischeri is grown in a medium containing glucose or one of several other sugars as
the carbon source, pyruvate accumulates in the
growth medium (87; Ruby, Ph.D. thesis). Unless
the medium is strongly buffered, the pH drops
and luminescence ceases. P. fischeri grows with
pyruvate as the sole source of carbon, so its
excretion during growth on other substrates is
unusual. It may simply be a case of unbalanced
growth, but the mechanism has yet to be elucidated. Pyruvate excretion has also been observed in Beneckea species (both luminous and
nonluminous) and P. phosphoreum, but not in
P. leiognathi. Whether pyruvate excretion is
ecologically important to the luminous bacteria
under any conditions or is an artifact of the
laboratory studies awaits the demonstration of
pyruvate excretion under in situ conditions.
However, nutrient exchange is a common feature of many symbioses, and pyruvate may represent such a nutrient, as proposed for one model
of symbiosis of luminous bacteria with luminous
fish (65, 86).
When stab cultures of different species of luminous bacteria are examined, two different patterns of growth and luminescence are seen. In
some (B. harveyi and P. leiognathi), luminescence is maximal at the surface of the agar,
where oxygen concentration is high. In others
(P. fischeri and P. phosphoreum), maximal luminescence occurs deep in the stab, where
growth is inhibited and oxygen concentration is
low (67). Analysis of this phenomenon has
shown that oxygen is important as a controlling
element in the synthesis of the luminescence
system. For P. fischeri and P. phosphoreum, at
low concentrations where growth is limited, luciferase synthesis is not halted, resulting in cells
with high luciferase content. Quite a different
response occurs in B. harveyi and P. leiognathi,
in which there is a tight coupling between
growth and luciferase systhesis. Both properties
are simultaneously limited, so that cells grown
under oxygen-limiting conditions develop the
same amount of luciferase as do well-aerated
cells at the same cell density.
These differences may have clear ecological
implications and suggest that the different species occupy niches that are different with regard
to oxygen concentration or the necessity for the
presence of the luminescence system or both.
For instance, the light organ may represent a
niche where the oxygen concentration under
which the bacteria are maintained could be controlled by the host fish. Experiments in this area
Dark (K) Variants of Luminous Bacteria

No review of the physiology of luminous bacteria is complete without consideration of a confusing and poorly understood property noted
and studied by Beijerinck at the turn of the
century (10, 11); namely, luminous bacteria have
the capacity, indeed the proclivity, to form spontaneous dark variants (51). Many years later, in
a class experiment regularly included in the summer course of Van Niel at Pacific Grove, such
dark strains of luminous bacteria were isolated.
Bright, freshly isolated luminous bacteria, when
subcultured at high temperatures, reliably
yielded dark variants. The interest of Van Niel
in this related, at least in part, to his experience
as the curator of the Delft culture collection.
There he had repeatedly encountered the appearance of dark strains in his cultures of luminous bacteria and was obliged to reisolate single
bright colonies upon each transfer (C. B. Van
Niel, personal communication).
We have conducted some studies with spontaneous dark variants of B. harveyi, which are
referred to as K variants. If a liquid culture is
maintained without shaking for several days
after it reaches stationary phase, it is common
to find K variants (51). Similar dark variants are
also readily isolated after treatment with acridine dyes (68; Nealson, Ph.D. thesis). The K
variants are not completely nonluminous; all
VOL. 43, 1979
exhibit a dim light emission, the exact level of

which may vary between 10-2 and 10-5 of the
wild-type level. Thus, they do not lack the luciferase gene; they simply express it at extremely
low levels. They are low in luciferase, in antigenically active luciferase cross-reacting material
(CRM), and in all or many of the other components of the luminescence system (60). Failure
to synthesize the luminescence system is not
attributable to the lack of autoinducer production, nor is luciferase synthesis stimulated by
arginine or cAMP.
The K variants are genetically stable #nd have
several other altered properties. In addition to
their deficiencies in bioluminetcence, they are
altered in colonial morphology, in flagellation,
and in phage sensitivity. The wild-type colonies
are convex, creamy, and translucent, whereas
the K-variant colonies are flat and more transparent. The K variants are also sensitive to
anaerobic conditions; they die rapidly, whereas
the wild-type cells survive well (Nealson, Ph.D.
thesis). Such conditions are the only ones that
have been found that result in selection of bright
bacteria over darks.
Beijerinck (10, 11) noted that dark variants
occasionally reverted back to the wild-type
bright state. When this occurs, all of the altered
characteristics revert simultaneously (51; Nealson, Ph.D. thesis). The appearance of dark forms
is more frequent in freshly isolated strains, even
at room temperature, compared with laboratory
strains that have been subcultured repeatedly
(62). In cultures of either bright or dark forms
maintained under conditions of rapid growth or
vigorous aeration or both, the appearance of the
alternate form is not common, whereas partial
anaerobiosis (as in dense cultures without aeration) and higher temperatures appear to favor
the appearance of variant forms. Neither genetic
transfer between bright forms and dark forms
nor a specific inhibitory or toxic factor that
might be produced by either form could be demonstrated (Nealson, Ph.D. thesis). These phenomena have not been studied in detail, and the
mechanism by which dark variants are formed
and by which reversion occurs is not understood.
In other species of luminous bacteria, the formation of K variants is known, but has not been
studied. Many dim and dark variants have been
isolated from P. leiognathi (39) and P. phosphoreum (B. M. Tebo and K. H. Nealson, unpublished data), but their properties are not well
defmed. Understanding the tendency of the various species to form dark variants and the factors
that control the reversible transformation between bright and dark forms is necessary if the
ecology of the luminous bacteria is to be understood.
505
HABITATS AND DISTRIBUTION OF

LUMINOUS BACTERIA
Luminous bacteria are widely distributed in
many different marine habitats, whereas they
are less abundant in nonmarine environments
(32). A tabulation of these habitats and associations as they are now known is presented in
Table 2.
Free Living
It is well documented that luminous bacteria
are ubiquitous in seawater as free-living forms,
but only recently has it been shown that most
are unattached and that the free-living populations are dynamic and change in a regular fashion (88) (Fig. 5). Coastal waters near San Diego,
Calif., regularly contained between 1 x 103 and
6 x 103 luminous bacteria per liter over a 2-year
sampling period. In the summer, the luminous
isolates were predominantly Beneckea spp.,
whereas in the winter P. fischeri was more abundant. Summer and winter populations at Woods
Hole, Mass., showed similar seasonal patterns
(Ruby and Hastings, unpublished data).
Like many marine heterotrophic bacteria
(102), the luminous forms occur at a much lower
abundance in the surface waters of the open
ocean than in coastal waters. In a recent quantitative study (Ruby et al., Appl. Environ. Microbiol., in press) it was found that in open ocean
surface waters at several stations in the eastern
Atlantic Ocean, B. harveyi showed marked seasonal changes, ranging from 2 to 300 colonyforming units per liter. In the same study, the
species distribution as a function of depth was
also examined. Midwater samples (100 to 1,000
m) in certain areas of the Atlantic Ocean contain
P. phosphoreum in relatively high numbers (20
to 100 cells per liter) irrespective of season and
very few of any of the other luminous species.
Since P. phosphoreum does not occur (less than
1 cell per liter) in the surface waters in those
areas, there is an actual increase in the numbers
of this species with depth, as both temperature
and (presumably) dissolved nutrient levels decrease (Fig. 6). The numbers of luminous bacteria decrease below 1,000 m. The reason for the
abundance of P. phosphoreum in the midwater
region is not known, but it may be related to the
association of luminous bacteria with higher organisms, especially fish, either as light organ or
gut symbionts or both. Such associations would
result in the constant release of the bacteria into
the seawater.
Other data are limited. Samples of seawater
from the Arctic Ocean (J. Baross, personal communication) and Antarctic Ocean (72) are reported to contain primarily P. phosphoreum and
nes1-uv*kaAtjo
506
MICROBIOL. REV.
TABLE 2. Habitats and associations of luminous bacteria
Saprophytic
Nonspecific symbionts
Commensal
Parasitic
Species
Habitat or host
Mode
Nonsymbionts
Free living
Seawater
Soil (?), freshwater (?)
Marine animals, fish, salt meat(?)
Wounds, meat
All Photobacterium and Beneckea spp.

X. luminescens(?)
All Photobacterium and Beneckea spp.
X. luminescens(?)
Digestive tracts, marine fish and

invertebrates, outer surfaces
of marine animals
Marine crustacea
Terrestrial and freshwater animals
All Photobacterium and Beneckea spp.'

P. fischeri, P. phosphoreum, B. harveyi
X. luminescens, others cultured, but not
identified
Specific symbionts
Symbiotic/parasitic
Light organ exosymbionts
Nematode/caterpillar
Teleost fishes (11 families)
Squid
(Euprymna scalopes)
(Heteroteuthis hawaiiensis)
endosym- Tunicates, squid(?)
X. luminescens
P. fischeri, P. phosphoreum, P. leiognathi; some isolates not identified, some
not cultured
P. fischerib
Not cultured
Not cultured or identified
Light organ
biontsc
a B. splendida and P. leiognathi have not been rigorously demonstrated in such habitats, but their presence
seems certain.
b The older literature (12, 32) contains reports that isolates cultured from squids exhibit characteristics that
are very different from "ordinary" marine luminous bacteria. Several different names were assigned, including
Vibrio pierantonii, which was later identified as P. fischeri (78). This has been confirmed with recent isolates
from Euprymna scalopes (G. Leisman, unpublished data).
'The occurrence of luminous endosymbionts has not been firmly established.
6
5
Cells
ml
4
3
2
rz/
/~
1975
s,
J
1976
1977
FIG. 5. Abundance of species of luminous bacteria

in surface seawater collected at the entrance of Mission Bay (San Diego, Calif.) on incoming tides at 3to 4-week intervals. Symbols: A, B. harveyi; A, P.
fischeri; El P. phosphoreum. Each sample was based
on taxonomic identification of between 80 and 100
isolates (64).
Tel Aviv, Israel, and for the Gulf of Elat. Some

of the patterns are complex, involving B. harveyi, P. leiognathi, and P. fischeri, although in a
saline lagoon only B. harveyi was present. These
authors postulate that several bacterial properties dictate the distribution of the species, including sensitivity to photoinactivation and
thermal inactivation, resistance to hypersalinity,
and the ability to grow under nutrient-poor conditions (89).
The occurrence of luminous bacteria in the
nematode-caterpillar relationships suggests that
these species might also occur as free-living soil
bacteria (Table 2), but they have never been
reported as such.
Saprophytic Forms: Natural Enrichments
If a swab from the surface of a marine animal,

especially a fish or a crustacean (but not a plant
or inanimate surface), is streaked onto seawater
nutrient agar plates, luminous colonies almost
always arise in numbers exceeding those that
might be obtained by plating seawater directly.
P. logei, but how these are distributed vertically The time-honored method for isolating lumiand seasonally is not yet documented. Yetinson nous bacteria involves allowing bacterial growth
and Shilo (100) have reported the patterns of on the surface of freshly killed fish or squid,
variation for the luminous bacteria in coastal followed by inspection and visual selection in
surface waters of the Mediterranean Sea near the dark for luminous colonies. Both Beneckea
VOL. 43, 1979
507
Total CFU/100ml Sea Water

50
100
150
U,
4-
E
r-C
T
0
40.
500
E
600
700
800
900
Temperature, OC
1,000
2,000
6,000
10,000
2 3 4 5 6
0
Luminous CFU/lOOml Sea Water
FIG. 6. Bacterial colony counts and water temperatures in the water column over the Puerto Rico Trench.
(A) Abundance of planktonic bacteria as a function of depth. Symbols: x, total colony-forming units (CFU);
A, P. phosphoreum; O1 P. leiognathi; 0, P. fischeri; 0, Beneckea spp. Data are from three hydrocasts made
over a period of 4 days at two stations. (B) Temperature profile of the water at the time of the hydrocasts.
and Photobacterium species may be obtained in

this manner; the temperature of incubation may
result in enrichment for one or another of the
different species (see above).
Meat, in general, is susceptible to overgrowth
by luminous bacteria. Salted hams and other
such stored meats have commonly been observed to emit light due to a growth of luminous
bacteria (32). Such observations have been less
frequent with the advent of refrigerators, espe-
cially those with automatic lights. The bacteria

are reputed to be harmless, in fact to be an
indication that no putrefaction has occurred.
There are also older accounts of the occurrence
of luminous bacteria in open human wounds,
reported especially from battlefield hospitals
during the 19th century (33); the presence of
luminous bacteria was taken as a good sign of
wound healing. It has always been tacitly (sometimes explicitly) assumed that the bacteria in-
508
volved in these associations were typical marine

forms. In fact, no isolates have been studied, so
there is no knowledge of their characteristics.
With the discovery of the terrestrial genus Xenorhabdus, it now seems possible that some of
the above cases may involve nonmarine luminous bacteria. The fact that Xenorhabdus isolates produce antibiotics is especially interesting
in connection with the report that there was no
putrefaction in the presence of the luminous
bacteria in question.
Commensal Forms
Quantitatively, the most important habitat of

luminous bacteria may be that of the gut tracts
of marine animals. It is not uncommon to find
between 5 x 106 and 5 x 107 luminous colonyforming units per m in the gut material of certain
marine fishes (85), and it may be that many such
fishes carry these bacteria as major enteric populations. All species of marine luminous bacteria
produce extracellular chitinase (78, 90) and are
thus conceivably important as gut symbionts in
the digestion of chitin (103; B. H. Robison and
J. G. Morin, Abstr. Annu. Meet. West. Soc.
Natural. 1977, 71a). Baross (unpublished data)
found that the gut contents of flatfish (Parophyrs vetulus and Citarichthys sordidus) contained about 108 bacteria per g, all of which were
luminous and chitinoclastic. Ruby and Morin
(85) studied Oxyjulis californica, Chromis
punctipinnis, and Argyropelecus hemigymnus
from the waters off southern California and
found B. harveyi, P. fischeri, and P. phosphoreum in different samples, with often just one of
the species predominating; the occurrence in the
fish was often correlated with the species composition of the planktonic luminous bacterial
population. Fecal pellets were luminescent and
also contained viable luminous bacteria and extractable luciferase (85). Luminous fecal pellets
have also been reported from Antarctic cod (77)
and from a midwater shrimp (95) although no
isolation of luminous bacteria was reported in
either case. Hastings and Mare'chal (unpublished data) plated and counted bacteria from
intestinal tracts of fish (A. hemigymnus and
Myctophum) coming from deep water in the
Straits of Messina, Sicily. In different samples
40 to 100% of the colony-forming units were
luminous and were a single species, P. phosphoreum. Invertebrates, such as squid, mussels, scallops, and crabs, also carry such bacteria (32; J.
Baross, P. A. Tester, and R. Y. Morita, Abstr.
Annu. Meet. Am. Soc. Microbiol. 1977, 11, p.
230; P. Dunlap, unpublished data).
MICROBIOL. REV.
Parasites
Marine. Although there are few recent studies on the subject, it is well known from the
older literature that luminous bacteria can infect
the tissues of a variety of marine crustaceans
(24, 32, 46). It has usually been possible to isolate
and grow the infecting organism in pure culture,
and the infections are apparently not species
specific. Different species of luminous bacteria
occur as parasites, and the same host may be
parasitized by different bacterial species (both
Photobacterium and Beneckea); the parasitism
is apparently correlated with season, corresponding to the abundance of free-living isolates
(Nealson, unpublished data). Moreover, parasitic bacteria apparently have a wide host range
and are capable of infecting many different crustaceans (32). Luminous bacteria grow in the
hemolymph of the infected animals, often making them easily visible. Whether an intermediate
host, such as is found in the nematode-caterpillar case (see below), is involved is not known.
There are current studies concerned with necrotic lesions in the Tanner crab exoskeleton
infected specifically with P. phosphoreum. The
extracellular bacterial chitinase may provide a
mechanism for the bacteria to inhabit this niche
(J. Baross, personal communication).
Nonmarine. There are many reports in the
older literature of parasitic infections of terrestrial organisms by luminous bacteria; the welldocumented hosts include mole crickets, mayflies, ants, wood lice, and millipedes (32). As
mentioned above, recent studies link the parasitism of caterpillars by luminous bacteria (X.
luminescens) with a mutualistic symbiosis between the bacteria and a specific genus of nematode (52, 76). The luminous bacteria are carried
symbiotically in the gut of a nematode and occur
there in small numbers, estimated to be less
than 100 cells per animal (G. Thomas, personal
communication); the nematode stage is not visibly luminous. In its juvenile stage the nematode
is ingested by the caterpillar and bores through
the gut wall, inoculating the hemolymph with
the bacteria, which then grow there and are
pathogenic for the caterpillar. As mentioned
above, the bacteria produce an antibiotic that
presumably prevents putrefaction. The life cycle
of the nematode is completed in the hemocoel,
and luminous bacteria are incorporated into the
progeny. Poinar (personal communication) has
proposed that the function of the luminescence
is to aid in the success of the nematode progeny
by attraction of additional potential hosts (other
caterpillars). It seems equally likely that the
light could aid in the dispersion of the nematode
VOL. 43, 1979
progeny via animals that are attracted by the

luminescence and feed on the caterpillars. Free
luminous bacteria ingested by the caterpillars
are not capable of killing the hosts, and without
the bacteria the nematode cannot complete its
life cycle (52, 76). In fact, bacteria-free nematodes are not pathogenic. However, the bacteria
may be cultured and are highly virulent when
injected directly into the hemocoel; as few as
one or two bacteria so injected are sufficient to
kill a caterpillar. The bacteria, classed in the
newly created genus Xenorhabdus (92), have
properties differing strikingly from those of Photobacterium and Beneckea species (52, 76), but
the luciferase and light-emitting reactions are
similar (75).
It is not possible to state whether the other
examples of terrestrial and freshwater animals
being infected by luminous bacteria (25, 32) are
due to similar nonmarine bacteria, but without
inquiry, they should certainly not be assumed to
be marine forms.
Light Organ Symbionts
Teleost fishes. Members of at least 30 genera
in 11 families of teleost fishes possess specialized
light organs in which luminous bacteria occur
(Fig. 7). It should be recalled that there are
many other luminous fishes (in approximately
25 families) that generate their own light and do
not utilize luminous bacteria (32, 45).
The bacteria that have been cultured from
P. phosphoreum
Macrourid
Merlucciid
P fischeri
Monocentrid
509
fish light organs are all members of the genus

Photobacterium and are species specific for a
particular host. The tropical and temperate
fishes, which occur in shallow and warmer waters, harbor P. leiognathi and P. fischeri, which
tolerate higher temperatures. The deep-sea and
midwater fishes, on the other hand, harbor psychrotrophic P. phosphoreum species as their
symbionts (73). P. logei is a newly identified
psychrotrophic species that has not yet been
reported to be specifically associated with a
higher organism (3).
Not all of the symbionts have been identified,
however; in some cases (Table 2) luminous bacteria that have been visualized in light organs
have proven impossible to culture (see below).
Species determination of these forms has not
been possible, but detection of bacterial luciferase has indicated that bacteria are indeed present. At least one group, the anomalopid fishes,
have luciferase with slow decay kinetics, suggesting that their symbiotic bacteria could be of
the Beneckea type (Cohn et al. Abstr. Annu.
Meet. Am. Soc. Photobiol., 1979).
In the symbiotic association mutual benefits
accrue; the bacteria are supplied with nutrients
and a protected environment, whereas the fish
is supplied with light. The function of the light
may be classed under one or more of three main
categories: attracting prey, assisting in escaping
or diverting predation, and communication (61).
For example, angler fish (Lophiiformes: Cera-
P. Ieiognothi
Leiognathid
Apogonid
Non -culturoble
Ceratioid
Not identified
Acropomatid
Anomolopid
Morid
Opisthoproctid
Trachichthyid
FIG. 7. The "ichthylicht". A diagrammatic fish is used to indicate the approximate locations, sizes, and
openings of the light organs of the several different families of luminous fishes that culture symbiotic luminous
bacteria as a source of light for the organ. The taxa of bacteria known to be specifically associated with each
fish group are listed at the bottom of the figure.
510
tioida) possess an illuminated fishing lure-like

structure that may attract would-be predators,
which are then converted into prey (71, 74).
Light emission is used to avoid predation in a
variety of ways, including frightening and diverting predators. Another technique to avoid
predation, uniquely possible with light emission,
involves counterillumination, namely ventral
emission of light during the day, the intensity
and color of which matches the down-welling
light that would otherwise silhouette the fish
(14, 36, 58, 95). The preponderant ventral location of luminescence in many families suggests
that this may be its function in these cases, and
the hypothesis has been supported experimentally (13, 95, 101). In the case of the pony fish
(Leiognathidae) the light organ is located deep
within the body as an outpocketing of the esophagus, and highly adapted optical mechanisms
allow this light to be presented as a uniform
ventral glow (36). Some of the light is initially
directed into the swim bladder, which is internally reflective, from whence it passes to the
entire posterior ventral surface by means of reflective and light-conducting fibers and tissues.
There are many variations on this theme in the
Leiognathidae.
Fish may also use the luminescence from light
organs for intraspecies communication and to
illuminate their surroundings, as in the flashlight
(anomalopid) and pinecone (monocentrid)
fishes.
Within the organs, the bacteria emit light
continuously, as they do in laboratory culture,
and in all cases the control of light emission
involves some specific host mechanism, which
may be a shutter, the rotation of the organ, or
chromatophores (45). Figure 8 shows progressively higher magnifications of the light organ of
Monocentris japonica, illustrating the several
different components of the organ. The bacterial
symbionts are extracellular and are maintained
in canals or tubules which communicate with
the exterior either directly (monocentrid, anomalopid, ceratioid, and macrourid fishes) or via
the intestinal tract (45). The bacteria are characteristically densely packed within the tubules
(109 to 1010 cells per ml of organ fluid); in the
leiognathids and monocentrids, microscopic and
viable counts give similar values, indicating that
nearly all of the cells are viable (39, 91). The
uniform microscopic appearance of the bacteria
in light organs and of the colonial isolates from
these organs suggested that the bacteria are a
single species (4, 5, 28); this has been confirmed
by taxonomic analysis, which has shown, however, that different colonial isolates from a given
organ may exhibit some different, but taxonomically inconsequential, specific characteristics
MICROBIOL. REV.
(21, 82, 84, 96).

Only a few of the numerous types of light
organs have been described in detail. In all
known cases, the organs are extensively vascularized, presumably for the supply of nutrients
and oxygen to the bacteria; control of the level
of the latter may be of physiological significance.
An interesting feature revealed by recent ultrastructural studies of some light organs is the
occurrence of large rather unusual mitochondrion-containing cells (Fig. 8). These have been
observed in both anomalopids (50) and monocentrids (91), but they apparently do not occur
in leiognathids (5). Their function may relate to
the maintenance of a low or proper oxygen tension (50), to the removal of bacterial excretion
products, or to energy generation, as proposed
by Nealson (65) in a model for symbiosis in the
Monocentridae.
Squid. Most luminous squid have their own
chemical mechanisms; only 2 (Loliginidae and
Sepiolidae) of the 19 luminous families have
genera that use luminous bacteria as the source
of light (44). The luminous bacteria have been
cultured from some but not all species. In the
older literature various names were assigned
that cannot be matched with current nomenclature (Table 3). The bacteria isolated from Euprymna scalopes have recently been identified
as P. fischeri (G. Leisman, unpublished data).
In the well-described cases, the anatomical
adaptations for the bacterial symbionts include
paired organs in the mantle cavity, which lie
against the ink sac near the anus (12, 53). Although the light is commonly emitted as a luminous cloud, in at least some species it can be
seen in the animal itself through the body wall
(43). The luminescence is presumably used to
frighten and confuse other organisms; in the
aphotic deep sea the usual black ink would of
course be ineffective.
Nonculturable symbionts. There are many
instances in which suspected bacterial symbionts have not been growable in host-free culture. For example, the symbionts of the anomalopid and ceratioid fishes have never been cultured, even though luminous bacteria have long
been believed to be involved in light emission,
based on both structural and physiological observations (29), but impossible to culture. Dilly
and Herring (18), on the basis of morphological
studies and the failure to culture luminous bacteria, concluded that the luminescence of the
squid Heteroteuthis dispar was probably not
bacterial in origin. For some such cases, it has
now been possible to demonstrate conclusively,
by biochemical analysis of light organ extracts,
whether a bacterial system is involved (Cohn et
al., Abstr. Annu. Meet. Am. Soc. Photobiol.
.m
FIG. 8. Structural features of the luminous organ of the luminous fish M. japonicus. (A) Line drawing of
fish and ventral view of lower jaw, showing the location of the light organs (solid black area). Bar = 1.0 cm.
(B) Scanning electron micrograph of the dorsal surface of the light organ. Numerous dermal papillae can be
seen. The emissary ducts from the light organ emerge at the tips of the four large papillae (arrows). Bar = 0.2
mm. (C) Light micrograph of a sagittal section of the lower jaw. m, Melanocytes; t, tubules with bacteria; b,
mandibular bone; d, dermal layer; p, dermal papillae. The arrow points to an emissary duct. Bar = 50 pm.
(D) Light micrograph showing the tubules of the light organ filled with bacteria. Tubules are lined with a
single layer of cuboidal epithelial cells that display loose nuclear chromatin and prominent nucleoli supported
by connective tissue cells. Blood capillaries are sparse and not readily visible. Bar = 15 /im. (E) Electron
micrograph showing the major features of the tubule epithelium. Epithelial cells that make up the lining of
the tubules have light-staining mitochondra (lm) with fine cristae. Epithelial cells that are further away from
the tubule lumen next to the blood capillaries have dark-staining mitochondria (dm) with thick cristae. t,
tubule with luminous bacteria; e, erythrocyte visible in capillary; n, nuclei of tubule epithelial cells. Bar = 1
,um.
511
512
TABLE 3. Bacterial luciferase asays in extracts of

light organs of eucaryotes
MICROBIOL. REV.
systems. Both Herring (43) and Galt (22) suggest

that a bacterial system is not involved in tunicate luminescence. Mackie and Bone (55), on
Luciferase
Organism
kineticsa the other hand, have published a detailed ultrastructural study of the light organ of a pyrosome,
Cephalopod (squid) (Heteroteuthis hawhich suggests that bacteria are indeed present;
waiiensis) ................. ......... Fast
they propose that bacteria are responsible for
Teleost (anomalopid) fishes
Photoblepharonpalpebratusb ...... ..... Slow
the light emission. In fact, bacterial luciferase
Anomalops katoptronb ................. Slow
activity has been demonstrated recently in exKryptophanaron alfredi ................ Slow
tracts of both salps (J. C. Makemson, M. HayTeleost (ceratiod) fishes
good, J. Dunlap, and J. W. Hastings, unpubCryptopsavus couesii ................ Fast
lished data) and pyrosomes (G. Leisman, D.
Ceratius holboelli ..... Fast
Cohn, and K. H. Nealson, unpublished data). In
Melanocetus johnsoni .................. Fast
the latter study, the luminescence was not due
........ Fast
Oneirodes sp. ........
to external luminous bacteria. Although such
Rate of decay of luminescence in the nonturnover adventitious luminous bacteria were present,
assay. See Fig. 2.
b
Previously reported to give light emission in the their luciferase activity was distinguishable, on
bacterial luciferase assay (29) but with no data on the basis of decay kinetics (slow), from that of
the pyrosome extracts (fast).
kinetics.
The possibility that bacteria provide the
1979) (Table 3). It is of interest that the lucifer- source of light for tunicates is of special interest
ases from the anomalopids exhibit slow kinetics, in connection with the manner by which the
since all other marine symbionts possess lucifer- luminescence is controlled. Pyrosomes respond
ases with fast kinetics. Perhaps the anomalopids to external stimuli, producing rapid (less than
are symbiotic with a different bacterial group, O.1-s) flashes of light. The nature of the control
more closely related to the Beneckea species. system is of great interest, as it relates to the
This would be of particular interest ichthyolog- degree of cellular integration of the symbionts.
ically, since the Anomalopidae is placed in the For example, does the host actually control the
Beryciformes (83) along with the Monocentridae biochemical flashing of the bacterial luciferase
and Trachichthyidae, both of which culture or merely mediate a constant bacterial glow by
Photobacterium species. Since bacterial lucifer- other methods? In other symbiotic systems, fish
ase is unique to procaryotes, its presence can and squid, for example, the light is controlled by
also be used to implicate the involvement of means of shutters or chromatophores. No such
symbiotic luminous bacteria in light emission structures are obvious in the pyrosomes, and the
(Cohn et al., Abstr. Annu. Meet. Am. Soc. Pho- possibility that the biochemistry of light emistobiol., 1979). The nonculturable symbiotic bac- sion of these endosymbionts is directly conteria are evidently highly host dependent, which trolled by the host is considered by Mackie and
is possibly related to nutritional or other factors Bone (55).
required for growth. This may represent an inFUNCTIONS OF BIOLUMINESCENCE
termediate stage in the evolution of endosymbiosis.
Endosymbionts. Harvey (32)' discusses but
Vestigial or Nonfunctional Hypotheses
largely discounts the possibility that luminous
Ten years ago the suggestion that the lumitunicates harbor endosymbiotic bacteria as their
source of light. Both pyrosomes and salps con- nescence system in bacteria is a vestigial one,
tain bacteria-like bodies; however, although having no present day function, was still being
many attempts have been made, bacteria have discussed (35). Knowledge of the elaborate
not been cultured from these organisms. Buch- mechanisms controlling the biosynthesis and acner (12) stated that symbiotic luminous bacteria tivity of luminescence now makes such a view
are definitely involved and that they are large unlikely, if not altogether untenable (40). Like(2 to 3 ,um by 10 to 30 ,im) sporeformers. The wise, the possibility that luminescence is the
spores are reported to be intracellular, and it has obligatory concomitant of some other functional
been suggested that they are responsible for the system has been discarded.
infection of eggs and the transmission of the
Biochemical Functions
symbionts. Both their size and their capacity to
Ten
would
years
ago the possibility that the emission
that
these
bacteria
indicate
form spores
be in a different group altogether and that if of light might relate to a biochemical (as distinct
such a symbiotic association occurs in tunicates, from a biological) function was put forward and
it would be fundamentally different from other discussed in detail (35). A specific suggestion
a
VOL. 43, 1979
made then was that a bioluminescence system

could allow a cell to carry out "photochemistry
in the dark," utilizing excited states produced
biochemically, possibly in the form of singlet
oxygen. According to this hypothesis, photochemistry would be an alternative to light emission, so that the functionally important condition need not be bioluminescence; in addition,
the light emitted would itself be nonfunctional.
No support for this idea has appeared, but to
our knowledge no good tests or experiments
have been carried out to examine it.
Another possible biochemical function is that
the light-emitting system can function as an
alternate pathway for electron flow, analogous
to the alternate carriers already known in E. coli
(27, 81), Klebsiella (30), and Beneckea natriegens (54). In luminous bacteria it is known that
luciferase functions in vivo at very low oxygen
concentrations (32, 34), even at concentrations
so low that electron flow via the cytochromes is
essentially stopped. Under these conditions, luminous bacteria may have an energetic advantage because the electron flow via luciferase
would allow for the reoxidation of reduced coenzymes and some adenosine triphosphate formation. Although the adenosine triphosphate
production associated with cytochrome electron
flow is lost, that associated with the flow from
pyridine nucleotide to flavin is not, so that the
phosphate-electron ratio might be changed from
3 to 1, a situation greatly favorable to fermentative metabolism. This mode of metabolism
might represent a major advantage under lowoxygen conditions with nonfermentable carbon
sources. Indeed, this might explain why bright
strains survive low-oxygen conditions better
than the dark K variants do (Nealson, Ph.D.
thesis).
From an ecological point of view, the ability
to function well under microaerophilic conditions may be of significance for both the symbiotic and the saprophytic luminous bacteria.
Within light organs, where symbiotic bacteria
are cultured, the oxygen concentration may be
poised at a low level. In the saprophytic state,
luminous bacteria live as heterotrophs associated with organic matter. This ecological niche
may also involve a low oxygen tension, so a
physiological adaptation to respire under such
conditions may be advantageous ecologically.
Biological Functions
Most authors favor the idea that it is the
emitted light that is of functional importance
and that its perception by some other organism(s) initiates phenomena which ultimately
serve to the advantage of the light-emitting bacteria. It seems unlikely that intraspecific effects
513
occur, since the behavior of luminous bacteria is

unaffected by visible light (32).
light organ luminescence. It was known as
early as 1921 that light organs of certain fish
harbor luminous bacteria as the source of their
luminescence (31); bacteria from Monocentris
were cultured a few years later (99). The function of the light is known in many of these cases
and relates to the behavior of the host, not the
bacteria (see above). The host possesses the
capability to control the light via shutters and
other mechanisms. In return for light production, the bacteria are provided with nutrients,
oxygen, and a protected niche. Descriptions of
such light organs and the functions of light emission in these cases are found elsewhere (4, 5, 28,
32, 40, 45, 50, 65, 91).
Dispersion and propagation. Luminous
bacteria apparently flourish and emit light in
numerous niches outside specific light organs,
some of which are in association with other
organisms (parasitic and commensal) and some
of which are not (saprophytic). One possible
view is that these bacteria are opportunistic
escapees or an inevitable overflow from the light
organ niche and that their light emission in these
other situations has no strong positive selective
advantage.
Another view, which we favor, is that there is
positive selection and that the selection mechanism may relate to the dispersion and propagation of the bacteria. If these bacteria are truly
marine enterobacteria (7), then it may be postulated that attraction of organisms that will
ingest them is advantageous. The facts that luminous bacteria are abundant in fish intestines
(85; Ruby, Ph.D. thesis) and that marked (mutant) strains of luminous bacteria survive passage through the gut tracts of fish (Ruby, Ph.D.
thesis) support this view. Luminous bacteria
growing on a substrate, whether it is a parasitized crustacean, the surface of a dead fish or
squid, or a fecal pellet, could produce sufficient
light to attract organisms to feed on the material,
thus enhancing the propagation of the bacteria
(40). In fact, based on analyses of fish gut contents, Robinson and Morin (Abstr. Annu. Meet.
West. Soc. Natural. 1977) have argued for such
an hypothesis, viewing the bacteria in the fish
guts as an aid to chitin digestion and the luminescence of fecal pellets as a selective advantage
to the bacteria for their own propagation.
For Beneckea species, which are not known to
occur as specific symbionts, this free-living mode
may represent a major and exhaustively exploited way of life. For Photobacterium species,
it may represent a light organ-independent mode
for which the property of luminescence may
nevertheless still have a positive selective value.
514
For the postulate to be valid, the light must

be seen. For the light to be seen there must be
more than a single bacterium, although the number is difficult to specify, given the complexities
of the optics (absorption, reflection, solid angle
subtended, etc.) and the distribution of the bacteria (size and shape of source) in any given
ecological situation. Bright, fully induced bacteria emit between 103 and 104 quanta/s per cell
(26). Photoreceptors can be sensitive to as few
as 5 to 10 quanta, perhaps fewer, but this means
quanta impinging on the photoreceptor after all
losses have been suffered. Moreover, photoreceptors adapt readily and are more sensitive to
changes in the light intensity than to absolute
levels, so the bacterial light emission, being continuous, must either be modulated by some
mechanism or be relatively bright in comparison
with the surrounding environment to be significant ecologically. With all this in mind, it seems
likely that a flux of 108 quanta/s per mm2 of
emitting surface is the minimum needed to be
seen. This would require 104 or 105 cells; if this
number corresponded to 1 mm3 of bacterial culture, it would represent a cell density of 107 or
108 cells per ml, which is just adequate for autoinduction. Both light organ cultures and saprophytic growths are two or more orders of
magnitude more dense and thus brighter than
the miniimum value discussed above (39, 86).
PERSPECTIVES: LUMINOUS AND
NONLUMINOUS BACTERIA
It now seems clear that the versatility of the
luminous species is not limited to alternate uses
of light. These bacteria are also capable of successfully exploiting habitats where there is apparently no positive selection for the ability to
emit light. One way to compete with nonluminous forms, assuming that the energetic drain
due to bioluminescence is a disadvantage, is to
have the capacity to block the synthesis of the
light-emitting system.
Thus, the phenomenon of autoinduction can
be viewed as a mechanism that allows for physiological adaptation to alternate environments.
The environment that is favorable for autoinduction is relatively nonspecific; it requires only
that the bacteria grow in a confined situation so
that the autoinducer accumulates. The fact that
autoinduction occurs in both Photobacterium
and Beneckea species suggests that light emission is functionally important and has a selective
advantage in situations other than light organs.
Under conditions of low numbers of bacteria,
when even bright bacteria would not be seen
anyway, the lack of autoinduction insures that
the cells are repressed for luminescence.
MICROBIOL. REV.
Why do nonluminous species of Beneckea

produce autoinducer that is active in inducing
the luminous strains? Are they inducing the
luminous strains under conditions where it will
not confer selective advantage, so that the nonluminous strains overgrow the luminous ones?
Or is it a bacterial mutualism where the two
species collaborate in the production of luminescence that profits both? There are surely numerous variations on these (hypothetical) scenarios, and it is not yet possible to arrive at any
favored hypothesis for any of the luminous species.
Catabolite repression represents still another
specific mechanism whereby control over the
synthesis of the bioluminescence system can be
exerted. The ecological significance of the distinctions between the Beneckea and Photobacterium species in this regard is not clear, although it may relate to the fact that the light
organ symbionts (Photobacterium species) receive nutrients from the host via the bloodstream, which may be high in glucose. Knowledge concerning the blood sugar(s) in the host
fish would be valuable. Repression of luminescence is obviously not consistent with light organ
function, but it may be important in the guts of
fish, where it could be mediated by the product
of chitin digestion (N-acetylglucosamine).
More permanent control of luminescence may
be achieved via the reversible transformation to
genetically stable dark forms, the K variants.
Under conditions where luminescence has no
selective advantage, the K variants are presumably selected for, and conversely, the bright revertants are selected for when luminescence is
advantageous. Thus, the propagation of either
form provides a potential inoculum for the other.
However, the factors that select for or against
luminescence are not known. If the ecology of
the luminous bacteria is to be understood, the
importance of the K variants must be established, as well as the factors that lead to their
formation and reversion. Are the naturally occurring dark bacteria that are taxonomically
identified as Beneckea and Photobacterium species merely naturally occurring K variants? Does
the differential sensitivity to anaerobic conditions in the two forms provide a mechanism
whereby the light organ can periodically purge
itself of any dim K variants? Is there a selective
advantage for the dark variants over the wildtype bright variants under any conditions? Does
the capacity to form stable dark variants allow
for life in some niche outside the symbiotic
mode, where luminescence is selected against?
Clearly, these and other such questions are central to the understanding of the ecology of the
VOL. 43, 1979
luminescent bacteria.
There are other relevant questions. What is
the cellular energetic commitment to bacterial
light emission? Are some of the luminous bacteria low-nutrient forms? Can they grow as freeliving forms in the ocean? What is the relationship between the various symbiotic partnerships
and the population densities of luminous bacteria? Have luminous bacteria established themselves in true endosymbiotic relationships in addition to the well-known extracellular ones?
What is the abundance, distribution, and significance of luminous bacteria in the gut tracts of
marine organisms? What are the physiological
mechanisms involved in the establishment,
maintenance, and control of the various symbioses?
Although the questions are many and the
firmly supported models are few, information
concerning the biochemistry, physiology, and
distribution of the luminous bacteria continues
to accumulate. As the data appear, many of the
questions posed above will be answered, and an
understanding of the ecology of these abundant,
widespread, and spectacular bacteria will be possible.
ACKNOWLEDGMENTS
Although we are responsible for the ideas and interpretations presented in this review, we were greatly
aided in their formulation and alteration by discussions with J. Morn, M. Shilo, P. Baumann, J. Reichelt,
R. Rosenblatt, and S. Ulitzur. In addition, unpublished
data and preprints were supplied by A. Eberhard, M.
Shilo, E. G. Ruby, E. P. Greenberg, G. Leisman, D.
Cohn, B. Tebo, G. Poinar, S. Ulitzur, D. Karl, K.
Kopecky, P. Dunlap, J. Morn, and J. Baross. Without
this assistance, it would have been impossible to do a
complete job on the review, and we are most grateful.
We acknowledge the support of National Science
Foundation grants PCM74-14788 (to K.H.N.) and
PCM77-19917 (to J.W.H.).
ADDENDUM IN PROOF
The compound responsible for the autoinducer activity from P. fischeri has now been purified and
identified as N-(3-oxohexanoyl)-3-aminodihydro2(3H)-furanone (N-fl-ketohexanoyl-homoserine lactone) (A. Eberhard, personal communication).
LITIERATURE CITED
1. Baldwin, T. O., M. M. Ziegler, and D. A.
Powers. 1979. Covalent structure of bacterial
luciferase: NH2 terminal sequence demonstrates subunit homology. Proc. Natl. Acad.
Sci. U.S.A. (in press).
2. Balny, C., and J. W. Hastings. 1975. Fluorescence and bioluminescence of bacterial luciferase intermediates. Biochemistry 14:47194723.
515
3. Bang, S., P. Baumann, and K. H. Nealson.

1978. Phenotypic characterization of Photobacterium logei (sp. nov.), a species related to
P. fischeri. Curr. Microbiol. 1:285-288.
4. Bassot, J. M. 1966. On the comparative morphology of some luminous organs, p. 557-610.
In F. H. Johnson and Y. Haneda (ed.), Bioluminescence in progress. Princeton University
Press, Princeton, N.J.
5. Bassot, J. M. 1975. Les organes lumineux a
bacteries symbiotiques de quelques teleosteens
Leiognathides. Arch. Zool. Exp. Gen. Notes
Rev. 116:359-373.
6. Baumann, L., and P. Baumann. 1976. Study of
the relationship among marine and terrestrial
enterobacteria by means of in vitro DNA/ribosomal RNA hybridization. Microbios Lett.
3:11-20.
7. Baumann, P., and L. Baumann. 1977. Biology
of the marine enterobacteria: genera Beneckea
and Photobacterium. Annu. Rev. Microbiol.
31:39-61.
8. Baumann, P., L. Baumann, and M. Mandel.
1971. Taxonomy of marine bacteria: the genus
Beneckea. J. Bacteriol. 107:269-294.
9. Beevar, J., S. Tu, and J. W. Hastings. 1978.
Activity and stability of the luciferase-flavin
intermediate. Biochemistry 17:1807-1812.
10. Beijerinck, M. W. 1889. Le Photobacterium luminosum, bacterie lumineuse de la Mer du
Nord. Arch. Neerl. Sci. Exactes Nat. 23:401415.
11. Beijerinek, M. W. 1916. Die Leuchtbakterien
der Nordsee im August und September. Folia
Microbiol. (Delft) 4:15-40.
12. Buchner, P. 1965. Endosymbiosis of animals
with plant microorganisms. John Wiley & Sons,
New York.
13. Case, J. F., J. Warner, A. T. Barnes, and M.
Lowenstine. 1977. Bioluminescence of lantern
fish (Mycotophidae) in response to changes in
light intensity. Nature (London) 265:179-181.
14. Clarke, W. D. 1963. Function of bioluminescence
in mesopelagic organisms. Nature (London)
198:1244-1246.
15. Cline, T. W., and J. W. Hastings. 1971. Temperature-sensitive mutants in the bacterial bioluminescence system. Proc. Natl. Acad. Sci.
U.S.A. 68:500-504.
16. Cline, T. W., and J. W. Hastings. 1972. Mutationally altered bacterial luciferase: implications for subunit functions. Biochemistry 11:
3359-3370.
17. Coffey, J. J. 1967. Inducible synthesis of bacterial luciferase: specificity and kinetics of induction. J. Bacteriol. 94:1638-1647.
18. Diry, P. N., and P. J. Herring. 1974. The ocular
light organ of Bathothauma lyromma (Mollusca: Cephalopoda). J. Zool. 172:81-100.
19. Doudoroff, M. 1942. Studies on the luminous
bacteria. I. Nutritional requirements of some
species, with special reference to methionine.
J. Bacteriol. 44:451-459.
20. Eberhard, A. 1972. Inhibition and activation of
516
bacterial luciferase synthesis. J. Bacteriol. 109:

1101-1105.
21. Fitzgerald, J. M. 1977. Classification of luminous bacteria from the light organ of the Australian pinecone fish, Cleiodopus gloriamaris.
Arch. Microbiol. 112:153-156.
22. Galt, C. P. 1978. Bioluminescence: dual mechanism in a planktonic tunicate produces brilliant
surface display. Science 200:70-72.
23. Ghisla, S., J. W. Hastings, V. Fauvaudon,
and J. Lhoste. 1978. Structure of the oxygen
adduct intermediate in the bacterial luciferase
reaction; "3C nuclear magnetic resonance determination. Proc. Natl. Acad. Sci. U.S.A. 75:
5860-5863.
24. Giard, A. 1889. Sur l'infection phosphorescente
des Talitres et autres Crustace's. C.R. Acad.
Sci. 109:503-506.
25. Greenberg, E. P., J. W. Hastings, and S.
Ulitzur. 1979. Induction of luciferase synthesis
in Beneckea harveyi by other marine bacteria.
Arch. Microbiol. 120:87-91.
26. Haas, E. 1979. Bioluminescence from single bacterial cells exhibit no oscillation. Biophys. J.
(in press).
27. Haddock, B. A., and C. W. Jones. 1977. Bacterial respiration. Bacteriol Rev. 41:47-99.
28. Haneda, Y. 1966. On a luminous organ of the
Australian pinecone fish, Cleidopus gloriamaris, p. 547-555. In F. H. Johnson and Y.
Haneda (ed.), Bioluminescence in progress.
Princeton University Press, Princeton, N.J.
29. Haneda, Y., and F. I. Tsuji. 1971. Light production in the luminous fishes Photoblepharon
and Anomalops from the Banda Islands. Science 173:143-145.
30. Harrison, D. E. F., and S. J. Pirt. 1967. The
influence of dissolved oxygen concentration on
the respiration and glucose metabolism of
Klebsiella aerogenes during growth. J. Gen.
Microbiol. 46:193-211.
31. Harvey, E. N. 1921. A fish, with a luminous
organ, designed for the growth of luminous
bacteria. Science 53:314-315.
32. Harvey, E. N. 1952. Bioluminescence. Academic
Press Inc., New York.
33. Harvey, E. N. 1957. A history of luminescence.
American Philosophical Society, Philadelphia.
34. Hastings, J. W. 1952. Oxygen concentration and
bioluminescence intensity. I. Bacteria and
fungi. J. Cell. Comp. Physiol. 39:1-30.
35. Hastings, J. W. 1968. Bioluminescence. Annu.
Rev. Biochem. 37:597-630.
36. Hastings, J. W. 1971. Light to hide by: ventral
luminescence to camouflage the silhouette. Science 173:1016-1017.
37. Hastings, J. W., and C. Balny. 1975. The oxygenated bacterial luciferase-flavin intermediate. J. Biol. Chem. 250:7288-7293.
37a.Hastings, J. W., C. Balny, D. LePeuch, and
P. Douzou. 1973. Spectral properties of an
oxygenated luciferase-flavin intermediated isolated by low-temperature chromatography.
Proc. Natl. Acad. Sci. U.S.A. 70:2468-3472.
38. Hastings, J. W., and Q. Gibson. 1963. Inter-
MICROBIOL. REV.
mediates in the bioluminescent oxidation of
flavin mononucleotide. J. Biol. Chem. 238:
2537-2554.
39. Hastings, J. W., and G. Mitchell. 1971. Endosymbiotic bioluminescent bacteria from the
light organ of pony fish. Biol. Bull. (Woods
Hole, Mass.) 141:261-268.
40. Hastings, J. W., and K. H. Nealson. 1977.
Bacterial bioluminescence. Annu. Rev. Microbiol. 31:549-595.
40a.Hastings, J. W., W. H. Riley, and J. Massa.
1965. The purification, properties, and chemiluminescent quantum yield of bacterial luciferase. J. Biol. Chem. 240:1473-1481.
41. Hastings, J. W., J. Spudich, and G. Malnic.
1963. The influence of aldehyde chain length
upon the relative quantum yield of the bioluminescent reaction of Achromobacter fischeri.
J. Biol. Chem. 238:3100-3105.
42. Henry, J. P., and A. M. Michelson. 1970. Physiologie bacterienne-etudes de bioluminescence. Regulation de la biolum. bacterienne. C.
R. Acad. Sci. 270:1947-1949.
43. Herring, P. J. 1977. Luminescence in cephalopods and fish. Symp. Zool. Soc. London 38:
127-159.
44. Herring, P. J. 1978. Bioluminescence of invertebrates other than insects, p. 199-240. In P. J.
Herring (ed.), Bioluminescence in action. Academic Press Inc., London.
45. Herring, P. J., and J. G. Morin. 1978. Bioluminescence in fishes, p. 272-329. In P. J. Herring (ed.), Bioluminescence in action. Academic Press Inc., London.
46. Inman, 0. L. 1927. A pathogenic luminous bacterium. Biol. Bull. (Woods Hole, Mass.) 53:
197-200.
47. Karl, D. H., and K. H. Nealson. 1979. Regulation of cellular metabolism during synthesis
and expression of the luminous system in Beneckea and Photobacterium. J. Gen. Microbiol. (in press).
48. Katznelson, R., and S. UWitzur. 1977. Control
of luciferase in a newly isolated strain of Photobacterium leiognathi. Arch. Microbiol. 115:
347-351.
49. Kelly, M. G., and P. Tett. 1978. Bioluminescence in the ocean, p. 399417. In P. J. Herring
50.
51.
52.
53.
(ed.), Bioluminescence in action. Academic

Press Inc., London.
Kessel, M. 1978. The ultrastructure of the relationship between the luminous organ of the
teleost fish Photoblepharon palpebratus and
its symbiotic bacteria. Cytobiologie Z. Exp.
Zellforsch. 15:145-158.
Keynan, A., and J. W. Hastings 1961. The
isolation and characterization of dark mutants
of luminescent bacteria. Biol. Bull. (Woods
Hole, Mass.) 121:375.
Khan, A., and W. M. Brooks. 1976. A chromogenic bioluminescent bacterium associated
with the entomophilic nematode Chromonema
heliothidis. J. Invertebr. Pathol. 29:253-261.
Kishitani, T. 1932. Studien uber die Leuchtsymbiose von japanischen Sepien. Folia Anat. Jpn.
VOL. 43, 1979

10:315-418.
54. Linton, J. D., D. E. F. Harrison, and A. T.
Bull. 1977. Molar growth yields, respiration
and cytochrome profiles of Beneckea natriegens when grown under carbon limitation in a
chemostat. Arch. Microbiol. 115:135-142.
55. Mackie, G. O., and Q. Bone. 1978. Luminescence and associated effector activity in Pyrosoma (Tunicata: Pyrosomida). Proc. R. Soc.
London Ser. B 202:483-495.
56. Magasanik, B. 1970. Glucose effects: inducer
exclusion and repression, p. 189-219. In J. R.
Beckwith and D. Zipser (ed.), The lac operon.
Cold Spring Harbor Laboratory, Cold Spring
Harbor, N.Y.
57. Makemson, J. C. 1973. Control of in vivo luminescence in a psychrophilic marine Photobacterium. Arch. Microbiol. 93:347-358.
58. McAllister, D. E. 1967. The significance of ventral bioluminescence in fishes. J. Fish. Res.
Board Can. 24:537-554.
59. Meighen, E. A. 1979. Biosynthesis of aliphatic
aldehydes for the bacterial bioluminescent reaction stimulation by ATP and NADPH.
Bioch. Biophys. Res. Commun. 87:1080-1086.
60. Meighen, E. A., M. Z. Nicoli, and J. W. Hastings. 1971. Functional differences of the nonidentical subunits of bacterial luciferase. Properties of native and chemically modified bacterial luciferase. Biochemistry 10:4069-4073.
61. Morin, J. G., A. Harrington, N. Krieger, K.
H. Nealson, T. 0. Baldwin, and J. W. Hastings. 1975. Light for all reasons. Science 190:
74-76.
62. Nealson, K. 1972. Factors controlling the appearance of spontaneous dark mutants of luminous bacteria. Biol. Bull. (Woods Hole,
Mass.) 143:471-472.
63. Nealson, K. H. 1977. Autoinduction of bacterial
luciferase: occurrence, mechanism and significance. Arch. Microbiol. 112:73-79.
64. Nealson, K. H. 1978. Isolation, identification and
manipulation of luminous bacteria. Methods
Enzymol. 57:153-166.
65. Nealson, K. H. 1979. Alternative strategies of
symbiosis of marine luminous fishes harboring
light emitting bacteria. Trends Biochem. Sci.
4:105-110.
66. Nealson, K. H., A. Eberhard, and J. W. Hastings. 1972. Catabolite repression of bacterial
bioluminescence: functional implications. Proc.
Natl. Acad. Sci. U.S.A. 69:1037-1076.
67. Nealson, K. H., and J. W. Hastings. 1977. Low
oxygen is optimal for luciferase synthesis in
some bacteria: ecological implications. Arch.
68. Nealson, K. H., and A. Markovitz. 1970. Mutant analysis and enzyme subunit complementation in bacterial bioluminescence in Photobacterium. J. Bacteriol. 104:300-312.
69. Nealson, K., T. Platt, and J. W. Hastings.
1970. The cellular control of the synthesis and
activity of the bacterial luminescent system. J.
Bacteriol. 104:313-322.
70. Ne'eman, Z., S. Ulitzur, D. Branton, and J.
517
W. Hastings. 1977. Membrane polypeptides

co-induced with the bacterial bioluminescent
system. J. Biol. Chem. 252:5150-5154.
71. O'Day, W. T. 1974. Bacterial luminescence in
the deep-sea anglerfish Oneirodes acanthias
(Gilbert, 1915). Scientific Contribution 255.
Natural History Museum, Los Angeles, Calif.
72. Olson, R. J., B. M. Tebo, E. Mousalli, A. F.
Carlucci, F. Azam, and 0. Holm-Hansen.
1978. Microbial investigations in the Ross Sea.
Antarct. J. U.S. 13:130-131.
73. Perlman, R. L, and L. Pastan. 1968. Regulation
offB galactosidase synthesis in E. coli by cyclic
adenosine 3',5'-monophosphate. J. Biol. Chem.
243:5420-5427.
74. Pietach, T. W. 1974. Osteology and relationships
of ceratioid anglerfishes of the family Oneirodidae, with a review of the genus Oneirodes
Lutken. Bull. Los Angeles County Mus. Nat.
Sci. 28:113.
75. Poinar, G. O., G. Thomas, M. Haygood, and
K. H. Nealson. 1979. Properties of the luminescent bacterium associated with the terrestrial nematode, Heterorhabditis bacteriophora. Soil Biol. Biochem., in press.
76. Poinar, G. O., G. M. Thomas, and R. Hess.
1977. Characteristics of the specific bacterium
associated with Heterorhabditis bacteriophora. Nematologica 23:97-102.
77. Raymond, J. A., and A. L. DeVries. 1976.
Bioluminescence in McMurdo Sound, Antarctica. Limnol. Oceanogr. 21:599-602.
78. Reichelt, J. L., and P. Baumann. 1973. Taxonomy of the marine, luminous bacteria. Arch.
Mikrobiol. 94:283-330.
79. Reichelt, J. L, and P. Baumann. 1974. Effect
of sodium chloride on growth of heterotrophic
marine bacteria. Arch. Microbiol. 97:329-345.
80. Reichelt, J. L., P. Baumann, and L. Baumann. 1976. Study of genetic relationships
among marine species of the genera Benecka
and Photobacterium by means of in vitro
DNA/DNA hybridization. Arch. Microbiol.
110:101-120.
81. Reichelt, J. L., and H. W. Doelle. 1971. The
influence of dissolved oxygen concentration on
phosphofructokinase and the glucose metabolism of Escherichia coli K-12. Antonie van
Leeuwenhoek J. Microbiol. Serol. 37:497-506.
82. Reichelt, J., K. H. Nealson, and J. W. Hastings. 1977. The specificity of symbiosis: pony
fish and luminescent bacteria. Arch. Microbiol.
112:157-161.
83. Rosenblatt, R. H., and W. L Montgomery.
1976. Kryptophaneron harveyi, a new anomalopid fish from the eastem tropical Pacific,
and the evolution of the Anomalopida. Copeia
3:510-515.
84. Ruby, E. G., and J. G. Morin. 1978. Specificity
of symbiosis between deep-sea fishes and psychrotrophic luminous bacteria. Deep Sea Res.
25:161-167.
85. Ruby, E. G., and J. G. Morin. 1979. Luminous
enteric bacteria of marine fishes: a study of
their distribution, densities and dispersion.
518
Appl. Environ. Microbiol. 38:406-411.

86. Ruby, E. G., and K. H. Nealson. 1976. Symbiotic association of Photobacterium fischeri
with the marine luminous fish Monocentris
japonica: a model of symbiosis based on bacterial studies. Biol. Bull. (Woods Hole, Mass.)
151:574-586.
87. Ruby, E. G., and K. H. Nealson. 1977. Pyruvate
production and excretion by the luminous marine bacteria. Appl. Environ. Microbiol. 34:
164-169.
88. Ruby, E. G., and K. H. Nealson. 1978. Seasonal
changes in the species composition of luminous
bacteria in nearshore sea water. Limnol.
Oceanogr. 23:530-533.
89. Shilo, M., and T. Yetinson. 1979. Physiological
characteristics underlying the distribution patterns of luminous bacteria in the Mediterranean Sea and the Gulf of Elat. Appl. Environ.
90. Spencer, R. 1961. Chitinoclastic activity in the
luminous bacteria. Nature (London) 190:938.
91. Tebo, B. M., D. S. Linthicum, and K. H. Nealson. 1979. Luminous bacteria and light emitting fish: ultrastructure of the symbiosis.
BioSystems, in press.
92. Thomas, G. M., and G. 0. Poinar, Jr. 1979. A
new entomopathogenic, nematophiclic, bioluminescent bacterium, Xenorhabdus luminescens sp. nov. (Enterobacteriaceae: Eubacteriales) associated with the terrestrial nematode
Heterorhabditis bacteriophora. Int. J. Syst.
Bacteriol. 29:352-360.
93. Ulitzur, S., and J. W. Hastings. 1978. Myristic
acid stimulation of bacterial bioluminescence
of aldehyde mutants. Proc. Natl. Acad. Sci.
U.S.A. 75:266-269.
94. UHtzur, S., and J. Yashphe. 1975. An adenosine 3',5'-monophosphate-requiring mutant of
the luminous bacteria Beneckea harveyi.
MICROBIOL. REV.
Biochim. Biophys. Acta 404:321-328.
95. Warner, J. A., M. I. Latz, and J. F. Case. 1979.
Cryptic bioluminescence in a midwater shrimp.
Science 203:1109-1110.
96. Watanabe, T., N. Mimara, A. Takimoto, and
T. Nakamura. 1975. Luminescence and respiratory activities of Photobacteriumphosphoreum. J. Biochem. 77:1147-1155.
97. Watanabe, T., and T. Nakamura. 1972. Studies on luciferase from Photobacterium phosphoreum. II. Substrate specificity and stoichiometry of the reaction in vitro. J. Biochem.
72:647-653.
98. Waters, C. A., and J. W. Hastings. 1977. Mutants of luminous bacteria with an altered control of luciferase synthesis. J. Bacteriol. 131:
519-525.
99. Yasaki, Y. 1928. On the nature of the luminescence of the knight fish, Monocentris japonicus. J. Exp. Zool. 50:495-505.
100. Yetinson, T., and M. Shilo. 1979. Seasonal and
geographic distribution of luminous bacteria in
the eastern Mediterranean Sea and the Gulf of
Elat. Appl. Environ. Microbiol. 37:1230-1238.
101. Young, R. E., and C. R. F. Roper. 1976. Bioluminescent countershading in midwater animals: evidence from living squid. Science 191:
1046-1048.
102. Zobell, C. E. 1946. Marine microbiology. Chronica Botanica, Waltham, Mass.
103. Zobeli, C. E., and S. C. Rittenberg. 1938. The
occurrence and characteristics of chitinoclastic
bacteria in the sea. J. Bacteriol. 35:275-287.
ADDENDUM IN PROOF
The compound responsible for the autoinducer activity from P. Fischeri has now been purified and
identified as N-(3-oxohexanoyl)-3-aminodihydro2(3H)-furanone (N-/3-ketohexanoyl-homoserine lac-
tone) (A. Eberhard, personal communication).

Nealson and Hastings

Uploaded by

Copyright:

Available Formats

Nealson and Hastings

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

Nealson and Hastings

Uploaded by

Copyright:

Available Formats

MICROBIOLoGIcAL REVIEWS, Dec. 1979, p.

Vol. 43, No.4

Bacterial Bioluminescence: Its Control and Ecological

tional studies. It is now clear that light organ

VOL. 43, 1979

lished data). All exhibit high specificity for

HASTINGS AND NEALSON

stituted luciferase with activity and properties

2). The affinity of the system for oxygen is very

VOL. 43, 1979

fructose, glycerol, and N-acetylglucosamine, and

whereas for other bacteria (normal luminous

TABLE 1. Luminous and nonluminous species of luminous genera

induce lu- Production

V. cholerae biotype albensis

X. nematophilus NC19 (ATCC 129304)

See reference 25.

HASTINGS AND NEALSON

P. phosphoreum and P. leiognathi, which are

FIG. 3. Electron micrographs of negataively

cells are more curved; both have one to three

VOL. 43, 1979

CONTROL OF THE SYNTHESIS AND

HASTINGS AND NEALSON

per cell independent of cell density (Fig. 4A).

VOL. 43, 1979

repressed for luciferase synthesis and are thus

analogy, luciferase in B. harveyi has been hypothesized to be a nonessential enzyme, which

Composition of the Growth Medium

HASTINGS AND NEALSON

C. A. Waters, Ph.D. thesis, Harvard University,

Pyruvate and Excretion Products

Dark (K) Variants of Luminous Bacteria

VOL. 43, 1979

exhibit a dim light emission, the exact level of

HABITATS AND DISTRIBUTION OF

HASTINGS AND NEALSON

TABLE 2. Habitats and associations of luminous bacteria

All Photobacterium and Beneckea spp.

Digestive tracts, marine fish and

All Photobacterium and Beneckea spp.'

FIG. 5. Abundance of species of luminous bacteria

Tel Aviv, Israel, and for the Gulf of Elat. Some

Saprophytic Forms: Natural Enrichments

If a swab from the surface of a marine animal,

VOL. 43, 1979

Total CFU/100ml Sea Water

and Photobacterium species may be obtained in

cially those with automatic lights. The bacteria

HASTINGS AND NEALSON

volved in these associations were typical marine

Quantitatively, the most important habitat of

VOL. 43, 1979

progeny via animals that are attracted by the

fish light organs are all members of the genus

HASTINGS AND NEALSON

tioida) possess an illuminated fishing lure-like

(21, 82, 84, 96).

HASTINGS AND NEALSON