J . MED. MICROBIOL.- VOL.
17 (1984). 217 235
C 1984 The Pathological Society of Great Britain and Ireland
OCCASIONAL REVIEW
COMPARATIVE STUDY OF T H E N A T U R E A N D BIOLOGICAL
ACTIVITIES OF BACTERIAL ENTEROTOXINS
C. G. GEMMELL
Department of Bacteriology, University of Glasgow, Royal Infirmary, Glasgow G4 OSF
Introduction
Within the last 20 years there has been a great increase in the number of aetiological
agents associated with gastrointestinal infection and intoxication. Concomitant with
this has been the realisation that many of the microorganisms responsible elaborate
enterotoxins that contribute to the pathogenesis of the infection. Most of these
enterotoxins give rise to a hypersecretion of fluid from the intestinal mucosa by a
mechanism that may or may not involve cytolytic damage to the epithelium lining the
intestinal tract (Keusch and Donta, 1975). Enterotoxins that are not cytotoxic are
now designated cytotonic, mainly because the efflux of water is passive and follows the
active secretion of chloride ions into the intestinal lumen (Field et al., 1969). In
general, these toxins are thought to exert their effects by perturbations of the cyclic
nucleotide system at the level of the cyclase. There are many similarities between the
different enterotoxins in terms of their activation of this system. In this review I shall
attempt to bring together the various features of the interactions of enterotoxins and
intestinal cells to achieve an integrated approach to the different modes of action.
Table I summarises the different types of enterotoxin which will be used to illustrate the
different ways in which they act.
Within the cytotonic group of enterotoxins, two types of toxin that differ in their
molecular size and mode of action within the intestinal cell can be distinguished (table
11). One group comprises enterotoxins with a mol. wt of 70 000-90 000, made up of
various active subunits and capable of activating adenylate cyclase. The enterotoxins
in the other group have a much lower mol. wt (5000-10000) and are capable of
activating guanylate cyclase instead of adenylate cyclase. Very much less is known
about the biological activities of the enterotoxins elaborated by gram-positive bacteria.
A more detailed picture of the structural features of the individual toxins has
emerged in recent years with the discovery that several subunits with different sizes and
biological functions make up the native enterotoxin molecules (table 111). As will be
shown later when the individual enterotoxins are discussed, there appear to be three
types of toxin present among the gram-negative intestinal pathogens. There are
heat-labile cytotonic enterotoxins with a subunit structure, heat-stable cytotonic
enterotoxins with no subunit structure and cytotoxic enterotoxins with a subunit
structure and active on subcellular functions. The enterotoxins of gram-positive
organisms appear to include substances with similar properties.
Much of our knowledge about the way in which bacterial enterotoxins act has been
Receitled 23 Nov. 1983; accepted 14 Dec. 1983
217
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
C . G . GEMMELL
218
TABLE
I
Classijkation of bacterial enterotoxins
- -
Source of enterotoxin
Type of
enterotoxin
Gram-positive
bacteria
-
Gram-negative
bacteria
~~
Cytotonic
Staph. aureus
C. dificile
toxin A
Cytotoxic
B. cereus
C . perfringens
A. hydrophila
E. coli
V. cholerae
Salmonella sp
Y . enterocolitica
Sh. dysenteriae
E. coli
TABLE
I1
Cytotonic toxins classijied according to molecular size and biological activity
Toxin
Molecular size
Biochemical
system
activated
V . cholerae
E. coli LT
Salmonella spp.
Sh. dysenteriae
84 000
85 000
90 000
70 000
Aden ylate
cyclase
b
%
1:
E. coli ST
Y. enterocolitica
Guanylate
cyclase
TABLE
I11
Sub-unit structure of certain bacterial enterotoxins
Producer
bacterium
Toxin
V. cholerae
Choleragen
E. coli
LT
Sh. dysenteriae
Dysentery (Shiga)
toxin
Number of
subunits
Molecular
size
1xA
5xB
1x A
5xB
1x A
4 or 5 x B
28 000
11 500
25 500
11 500
30 000
7000-11 000
TABLE
IV
Models used for enterotoxin assays in vitro and in vivo
In-vitro tests
In-vivo tests
~
Erythrocyte ghosts
Adrenal cells
Isolated fat cells
Lymphocytes
Fibroblasts
Suckling mice
Infant rabbit
Rabbit ileal loop
Skin permeability
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
BA CTERIA L EN TERO TOX I "
219
derived from the use of laboratory models-ither
whole animals or isolated tissue
cells. A list of the commonly used models is shown in table IV. For example, Y 1
adrenal cells derived from mouse adrenal cortex grow as a confluent monolayer which,
in the presence of enterotoxin, undergoes morphological and metabolic changes that
include rounding up and detachment from the plate surface and secretion of increased
amounts of 6-4,3 ketosteroids. Similar changes can be induced by cyclic adenosine
monophosphate (CAMP) and adrenocorticotrophin (ACTH), and increased intracellular concentrations of CAMP are detectable in the intoxicated cells (Donta, King and
Sloper, 1973; Donta, 1974).
The different cell types have been used (Richards and Douglas, 1978) to study the
binding and physiological effects of choleragen and the heat-labile enterotoxin (LT) of
Escherichiu coli. Close similarities in membrane receptors have been shown and the
sensitivities of the cells to these toxins have been compared on the basis of the numbers
of cell receptors and activation sites on each cell type; e.g., toad erythrocytes, cultured
melanocytes and fibroblasts have 20 000-50 000 binding sites/cell but only 1000-3000
activation sites.
With the increasing interest in enteropathogenic microorganisms, their virulence
factors and their role in the disease process, it will be impossible in this review to cover
all aspects of infection and intoxication by these organisms. I shall concentrate on the
comparative aspects of the potent enterotoxins produced, in terms of their biological
activity, their active sites, receptors and substrates, leading to a discussion of their
mode of action at a molecular level where known and so attempt to draw together the
various strands of information available for the different toxins to produce an
integrated picture of their similarities and differences.
Vibrio cholerae
Of the various enterotoxins to be described in this review, that produced by Vibrio
cholerae has been well defined both in terms of its structure and mode of action. It
consists (fig. 1) of five B subunits (mol. wt 11 500 each) that are responsible for the
binding of the toxin to specific receptors on the brush border epithelium, and the A
subunit (mol. wt 28 000). The receptors are composed of GM1 ganglioside, of which
the lipid moiety is embedded in the lipid matrix of the cell membrane with the
hydrophobic oligosaccharide moiety sticking out from the outer surface of the cell
membrane (fig. 2). The B subunits stick to this oligosaccharide, one per receptor
molecule. After the toxin has bound to the membrane, translocation of the single A
subunit through the membrane takes place. The A subunit comprises a single
polypeptide that can split through a disulphide bridge into two fragments-A1 (mol.
wt 20 500) and A2 (mol. wt 7500).
The active toxin moiety is accessible to antitoxin for only 30-50 s after binding.
Thereafter, its presence remains undetected until at least 10 min have elapsed during
which activation of adenylate cyclase occurs. Little is known of what happens during
this time except that some evidence has been presented showing that some protein
synthesis occurs (Moss et ul., 1976).
Two hypotheses have been proposed: (a) the active moiety may become enclosed
within a vesicle and later released within the cytosol; or (b) it may pass directly through
the plasma membrane. The B subunits may facilitate these processes by forming a
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
220
C. G . GEMMELL
CHOLERA
TOX 1N
CHOLERAGEN (PI
84, o m
6.6)
11,500
21 .hnn
I
EACH
B I N D I N G TO CELL
7 ,n6o
I
SURFACE R E C E P T O R S
FUNCTION
ACTIVATES
ADENYL C Y C L A S E
UNCERTAIN
FIG. 1 .-Subunit
structure of cholera toxin.
channel through which the A subunit can pass in an unfolded form (Gill, 1976) with
fraction A2 leading the way (fig. 3). Gill et al. (1 98 1) have proposed that the B subunits
remain on the surface of the cell membrane and play no further part in the entry
process. Saturation of the binding sites by B subunits renders the cell immune to
cholera toxin. However, if the cells are broken open they become sensitive to
intoxication by fragment A1.
Once in the cytosol, the A1 fragment catalyses a transfer of ADP-ribose from NAD
to membrane-bound proteins, principally a part of the regulatory GTP-binding
component of adenylate cyclase; this target has a subunit structure of mol. wt 42 000
(Cassel and Pfeuffer, 1978). Such an event “locks” adenylate cyclase in an active form
by inhibiting an inherent feed-back regulatory mechanism by which stimulatory GTP
is hydrolysed to inhibitory G D P and inorganic phosphate; CAMP then accumulates
intracellularly (fig. 4). Hormones increase the cyclase activity by increasing the rate of
GTP binding. Cholera toxin, on the other hand, inhibits the GTPase “turn-of’
reaction and thus increases the activity of adenylate cyclase (Cassel and Selinger,
1977).
Ceramide-Glucose-Galactose-N-Acetyl
I
galactosamine-Galactose
N- Acetyl
neuraminic acid
FIG.2.-Structure
of ganglioside G M I , the cell receptor for cholera toxin.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
22 1
BACTERIAL E N T E R 0 T O X I N S
I
FIG.3.-Binding to (I) and translocation through (11) the mammalian cell membrane of cholera toxin. AI,
A2 and B represent subunits of cholera toxin; = ganglioside, G M 1; 0 = PhOSPhOIiPid molecule;
0 =protein molecule.
P
FIG.4.-Cholera toxin and the adenylate cyclase system. A I =subunit of toxin, C =component responsible
for conversion of ATP to CAMP, R = component responsible for GTP-dependent regulation of enzyme
activity, H = hormone receptor.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
222
C. G. G E M M E L L
The accumulation of CAMP stimulates active secretion of chloride and bicarbonate
by crypt cells and inhibits the normal absorption of chloride coupled to sodium ions by
villus cells. Because cholera toxin increases cAMP in both cell types, the diarrhoea
probably results from both mechanisms acting simultaneously (Field, 1981). In
addition, it has been postulated that cholera toxin potentiates the release of serotonin
from the enterochromaffin cells of the small intestine resulting in further fluid secretion
by a nervous reflex (Cassuto et al., 1979).
Activation of membrane-bound adenylate cyclase is dose-dependent (Bennett and
Cuatrecasas, 1975) and results in the conversion of ATP to CAMP. This activation is
distinct from that caused by epinephrine, prostaglandins, glucagon (Berkenbile and
Delaney, 1976), catecholamines or ACTH. Cholera toxin exerts its maximum effect
after 1-3 h and cAMP levels remain high for 12 h whereas E. coli LT acts best in 15-90
min and activation is short-lived (Finkelstein et al., 1976). Several enterotoxins can
also potentiate hormonal stimulation of adenylate cyclase. The variety of physiological changes induced in different types of cell is summarised in table V.
TABLE
V
Interaction of cholera toxin with isolated cell functions
~
Cell system
~~
Effect of toxin
Mouse adrenal
cells
Chinese hamster
ovary cells
Enterochromaffin
cells
Morphological change;
steroidogenesis increased
Morphological change;
cAMP increased
Granule release;
cAMP increased
Leukocytes
Lysosomal enzyme
release reduced
Donta et al.
(1 973)
Guerrant et
al. (1974)
Fujita et
al. (1 974)
Bourne et
al. (1973)
Zurier et
al. (1974)
Shigella dysenteriae
The molecular structure and mechanism of action of Shigella dysenteriae toxin
(Shiga toxin) has been largely elucidated in the last 4 years. Like cholera toxin and E.
coli LT, it consists of two kinds of polypeptide chain (Olsnes and Eiklid, 1980; Olsnes,
Reisbig and Eiklid, 1981). One of the chains, the A chain or its active A, fragment,
enters into the enterocyte cytoplasm and inhibits protein biosynthesis by inactivating
catalytically, the 60s ribosomal sub-units (Brown, Rothman and Doctor, 1980;
Reisbig, Olsnes and Eiklid, 1981). Only one A chain is present in each molecule of
toxin. The remainder of the toxin consists of six or seven B chains that are probably
concerned in the binding of the toxin to sites at the cell surface (Eiklid and Olsnes,
1980). So far, only certain primate-derived epithelial cell lines bind the toxin. A
possible mechanism has been proposed (Keusch, Donahue-Rolfe and Jacewicz, 1981)
whereby the toxin penetrates the intestinal cell and releases its active subunit A so that
protein synthesis can be impaired (figs 5A and 5B).
There is conflicting evidence about whether the toxin directly stimulates adenylate
cyclase activity leading to increased levels of cAMP and fluid accumulation in the gut
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
BA CTERIA L E N T E R 0 TOXINS
223
w
Ribosome
i\ f
1.
2.
3.
FIG.5A.-Penetration of Shiga toxin into intestinal cells and its biological activity therein. Modified from
Keusch, Donohue-Rolfe and Jacewicz (1981). A =component of toxin associated with biological
activity; B = component of toxin associated with binding; A1A2 = subunits of component A.
lumen (Donowitz, Keusch and Binder, 1975; Charney et al., 1976). Injection of the
toxin cause haemorrhages in the intestinal mucosa, probably as a result of its cytotoxic
activity, and it seems more likely, therefore, that any changes in fluid balance,
adenylate cyclase activity and CAMPlevels might be a secondary phenomenon (table
VI). As well as inducing fluid accumulation in rabbit ileal loops, the toxin is also lethal
to monkeys, rabbits and mice (Keusch et al., 1972) and the intoxicated animals show
evidence of neurological disorder, viz., hind leg paralysis.
In an elegant study, Eiklid and Olsnes (1983) have shown that the cytotoxic,
enterotoxic and neurotoxic activities of the purified toxin are engendered by the same
FIG.5B.-Sequence of events during internalisation and activation of Shiga toxin. The diagram represents
the fusion of lysosomes (small circles) with a pinocytic vesicle containing the toxin (large circle).
Lysosomal enzymes are thought to cleave the A and B components within the pinocytic vesicle. The
active subunit A1 is finally released into the cytosol to inhibit ribosomal activity.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
C . G . GEMMELL
224
TABLEVI
Sequence of events in Sh. dysenteriae intoxication
1.
2.
3.
4.
5.
6.
7.
8.
Binds to cell receptors by B subunits
Penetration of A subunit into cytosol
enclosed in membrane-bound vesicle
Migration of lysosomes towards vesicle
Nicking of active moiety, Ai, of toxin by
proteolytic enzymes released by lysosomes
Biological activity of A1 by binding to 60s
ribosomal subunit
Impairment of protein biosynthesis
Cell damage that may lead to death
Secondary accumulation of CAMPand fluid
release into intestinal lumen
molecule. They showed that the relative concentrations of the three activities
remained the same during purification of the toxin and were equally destroyed by
heating to 80°C and precipitation by neutralising antibody. In summary, therefore,
this enterotoxin ought to be regarded as a cytotoxic enterotoxin that is potentially
lethal, and is different in its mechanism of action from either cholera toxin or E. coli
LT.
Salmonella species
Until the mid 1960’s, there was little evidence to suggest that the pathogenesis of
human salmonellosis involved anything more than endotoxin (Kohler and Bohl, 1966;
Thomlinson and Buxton, 1963). Koupal and Deibel (1975) identified an envelopeassociated enterotoxin that caused fluid secretion in newborn mice and this finding
started a different train of events that resulted in the discovery of two skin permeability
factors in rabbits (Sandefur and Peterson, 1976). The salmonella toxin(s) closely
resembled cholera toxin and E. coli LT in its biological activities in vitro. However, the
salmonella toxin failed to induce fluid and electrolyte secretion in rabbit intestinal loop
tests (Molina and Peterson, 1980). More recently, Baloda et al. (1983) examined
various strains of Salmonella enteritidis and S. typhimurium and described the presence
therein of a heat-labile enterotoxin that gave a positive intestinal loop assay, a positive
rabbit skin permeability test and morphological changes in Y 1 adrenal cells and CHO
tissue cells. Their strains had been isolated in India whereas those of Peterson and his
colleagues were derived from American sources. Thus there may be geographical
differences amongst Salmonella strains in their enterotoxinogenicity. It appears that
the two toxins differ in their sensitivity to proteolytic digestion (Peterson, 1980).
Because the salmonella enterotoxin can bind to G M I ganglioside, it is likely that it
may act at the cellular and sub-cellular level in the same way as cholera toxin.
However, no subunit structure has yet been described in this enterotoxin.
Escherichia coli
The association between travellers’ diarrhoea and toxigenic E. coli is well
recognised and is usually the result of the action of heat stable (ST) or heat-labile (LT)
enterotoxin or both on the small intestine (Sack, 1975). ST from E. coli strains of
human origin is a simple protein with a mol. wt of c. 2500 (Madsen and Knoop, 1980).
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
BACTERIAL E N T E R 0 TOXINS
225
Specific binding of ST to the rat intestinal cells and, in particular, brush border
membranes has been demonstrated (Giannella, Luttrell and Drake, 1980) and a
proteinaceous receptor has been implicated in the binding process. Thereafter, the
toxin stimulates guanylate cyclase, so increasing the intracellular concentration of
cGMP (Newsome, Burgess and Mullan, 1978).
Usually the biological activity of the toxin is measured in uiuo and there is evidence
to suggest that calcium and prostaglandins play a part in the secretion process
(Guerrant et al., 1980) and a mechanism of action of ST has been proposed by
Greenberg and Guerrant (198 1) which is outlined in table VII. Hypotheses of the roles
for free radicals and the activation of prostaglandin synthesis have developed from the
findings that (a) butylated hydroxy anisole (BHA), a free radical scavenger, can inhibit
the action of ST, and (b) indomethacin, by inhibiting cyclo-oxygenase, which is
involved in the arachidonic acid cascade, can impair ST action. In summary,
therefore, ST causes fluid secretion within the intestinal lumen by activation of
guanylate cyclase and impairment of chloride absorption. Experiments with a variety
of pharmacologically active inhibitors indicate that ST may act through the
prostaglandin synthesis pathway, may involve free-radical activation of guanylate
cyclase and may be regulated by intracellular calmodulin.
On the basis of significant differences in heat stability and amino-acid composition,
the production of two distinct STs by E. coli has been recognised (Kapitany et al.,
1979). Likewise, differences between Yersinia enterocolitica ST (see later) and E. coli
ST may reflect a family of ST molecules that are similar in their biological activity and
mechanism of action but somewhat different in structure.
In contrast to the relatively small amount of information available about E. coli ST,
LT has received considerable attention because of its close similarity with cholera
toxin. By cloning of the LT structural gene and translation of the chimeric plasmid
produced in E. coli minicells, the structure of LT has been shown (Dallas, Gill and
Falkow, 1979) to consist of a protein of mol. wt 25 500 that has adenylate-cyclase-stimulating activity as well as NADase activity and several copies of a protein of mol. wt
1 1 500 that is bound to Y 1 adrenal cells; both proteins react with anticholera toxin sera
(Dallas and Falkow, 1979). LT cross-reacts immunologically mainly with antisera
directed against the B subunit of cholera toxin; this indicates some homology between
the binding domains of each toxin (Klipstein and Engert, 1977). Like cholera toxin,
TABLE
VII
Mechanism of action of E. coli S T
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Ingestion of approximately lo7 cells with
plasmid-encoded ST and colonisation factor
Adherence and multiplication in upper small
bowel where ST is produced
Possible binding of ST to tissue specific receptor
Activation of prostaglandin pathway
Release of free radicals
Activation of guanylate cyclase
Cyclic GMP synthesis
Possible activation of protein kinase
Impaired chloride absorption
Isotonic fluid secretion (non-inflammatory
diarrhoea)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
226
C . G . GEMMELL
LT needs NAD+ to activate adenylate cyclase and can also potentiate the effect of
hormones on this enzyme.
As well as the immunological similarities between the B subunits of E. coli LT and
cholera toxin, LT appears to bind specifically to ganglioside GMI but it has been
suggested that the binding is less avid because the amounts of GM1 needed to neutralise
its biological activity are somewhat larger than for cholera toxin (Zenser and Metzger,
1974). Another difference between the binding of the two toxins is that the biological
activity of E. coli LT is not blocked to the same extent by cholera toxoid
(choleragenoid) as is the activity of cholera toxin (Holmgren, 1973). Otherwise it
appears that E. coli LT and cholera toxin bear a close resemblance to each other.
A significant immunological cross-reaction between cholera toxin and E. coli LT
has been recognised (Holmgren, 1980) to the extent that both the A and B sub-units of
the two toxins are antigenically similar. Using an enzyme-linked immunosorbent
assay (ELISA), Holmgren and Svennerholm (1979) observed that monospecific
anti-cholera A and anti-cholera B antibodies both reacted with E. coli LT bound to
GMI ganglioside adsorbed onto plastic. It is likely that E. coli LT, as well as sharing
determinants with cholera toxin, also contains unique antigenic determinants.
The production of enterotoxins by non-invasive enterotoxigenic strains of E. coli is
responsible for intestinal fluid secretion and, hence, diarrhoea. In contrast, other E.
coli strains that are serologically related to the shigellae (Formal et al., 1978) are able to
penetrate and multiply within colonic epithelial cells (Dupont et al., 1971). Strains of
a third group recognised by serotype as enteropathogenic E. coli (EPEC), have been
incriminated as agents of epidemic and endemic diarrhoea1 disease in infants (table
VIII) but they are not enteroinvasive nor do they produce LT or ST (Gurwith,
Wiseman and Chow, 1977). Many EPEC and some other E. coli intestinal isolates
that produce the classic enterotoxins also make a cell-associated cytotoxin that can be
neutralised by antitoxin prepared against the purified toxin of Sh. dysenteriae (O’Brien
et al., 1982). Moreover, like shigella extracts, cell lysates of certain EPEC strains are
enterotoxic for rabbit ileal loops and lethal and paralytic for mice. This cytotoxin is
probably a recognisable virulence determinant in EPEC strains and this suggestion is
supported by the destruction of intestinal epithelial cell microvilli seen in small bowel
biopsies from infants with EPEC diarrhoea (Clausen and Christie, 1982). The
Shiga-like toxin has been isolated and purified (O’Brien and La Veck, 1983) and shown
to be composed of one A subunit (mol. wt 31 500k 1000) and several copies of a B
subunit (mol. wt 4000), making a native toxin molecule of mol. wt 48 000.
The A subunit can be nicked by trypsin or by endogenous proteases to form an A1
component (mol. wt 27000) and an A2 component (mol. wt 4500). There are,
therefore, many similarities between this toxin and that produced by Sh. dysenteriae
TABLEVIII
E. coli serotypes and diurrhoeal diseuse
Pathogenic
group
Enteropathogenic
Enterotoxigenic
Enteroinvasive
I
0-serotypes that have the given
pathogenic mechanism
26, 55, 86, 1 1 1, I 14, 119, 125, 126, 127, 128, 142
6, 8, 15, 25, 27, 78, 148, 149
28ac, 112ac, 124, 136, 143, 144, 152, 164
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
BACTERIAL ENTEROTOXINS
227
and antibody to the latter toxin can neutralise the former. So far, no studies have been
published on the mechanism of action of the E. colicytotoxin but it is likely that there is
a family of Shiga-like toxins as there is a family of cholera-like toxins.
Aeromonas hydrophila
Another member of the Vibrionaceae, Aeromonas hydrophila produces both a
heat-stable and a heat-labile enterotoxin. Strains of this species that have caused
opportunistic infection in compromised patients and diarrhoea in normal people
produce only the heat-labile enterotoxin (Wadstrom, Ljungh and Wretlind, 1976;
Annapurna and Sanyal, 1977) whereas strains isolated from fish produce both
enterotoxins (Boulanger, Lallier and Cousineau, 1977). Little is known about the
heat-stable enterotoxin, but several interesting facts about the heat-labile enterotoxin
are known (Ljungh, Eneroth and Wadstrom, 1982a and b; Ljungh and Kronevi, 1982).
The biological features resemble those of cholera enterotoxin and E. coli LT but there
are several significant differences (table IX). Some of these differences may be
attributable to the nature of the toxin itself. It has a much lower mol. wt, only 15 000,
has no apparent subunit structure and is very sensitive to proteolytic digestion which
inactivates it. The native form of this toxin may have a higher mol. wt and be broken
TABLEIX
Biological properties of A . hydrophila enterotoxin
Test system
Result
Lethality
Fluid accumulation (rabbit ileal loop)
Adrenal Y 1 cell elongation
Rabbit skin permeability
Potentiation of CAMP synthesis
Steroidogenesis in adrenal cells
Neutralisation by gangliosides
Neutralisation of effects by antiserum to
cholera toxin or E. coli LT
negative
positive
posi t/ve
positive
positive
positive
negative
negative
down immediately after intracellular biosynthesis. Such a hypothesis would account
for the inability of gangliosides or antibody to cholera toxin or E. coli LT to neutralise
its biological effects. It should also be noted that the enterotoxin induces cytotoxic
changes in adrenal Y 1 cells and elongation of Chinese hamster ovary cells similar to
those produced by cholera toxin and E. coli LT (Ljungh, Wretlind and Mollby, 1981).
However, the sensitivity of the adrenal cells to aeromonas toxin was much lower than
their sensitivity to cholera toxin, but P-glucosidase or P-galactosidase treatment of the
cells increased their sensitivity to the aeromonas toxin.
Similar differences in the degree of induction of intracellular CAMP, but not of
cGMP, by the toxin in either intestinal or adrenal cells were noted by Ljungh, Eneroth
and Wadstrom (1982b). Aeromonas enterotoxin was further shown to induce
steroidogenesis in adrenal Y 1 cells but was less effective on the basis of equivalent
weights than cholera toxin, although the lag phase was significantly shorter with
aeromonas enterotoxin than with either cholera toxin or E. coli LT.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
228
C. G. G E M M E L L
Yersinia enterocolitica
This organism has been recognised recently as an important cause of gastroenteritis
in children (Kohl, Jacobson and Nahmias, 1976) and part of its pathogenicity has been
attributed to its ability to elaborate a low mol. wt (9700) heat-stable enterotoxin
(Okamoto et al., 1982). This toxin has strong similarities with E. coli ST in its activity
in the suckling mouse assay (EDmi, = 25 ng) through causing fluid accumulation within
the intestine by activation of cGMP (Robins-Browne et al., 1979), and in its failure to
induce morphological changes in Chinese hamster ovary cells or Y 1 adrenal cells.
Compared with the enterotoxins elaborated by gram-negative bacteria, the
enterotoxins of the gram-positive bacteria (table X) have only recently been subjected
to the same kind of analysis, with Clostridium perfringens enterotoxin foremost.
TABLEX
Enterotoxins of gram-positive bacteria
~
Producer
species
Mol. wt
Isoelectric point (PI)
Staph. aureus
B. cereus
C. perfringens
C. dificile
28 000-35 000
55000-60000
35 000
450 000
74-8.6
4.85
4.3
5-5
Number of
electrophoretic
forms
Several
One
Two
One
Clostridum perfringens
This enterotoxin, although cytotoxic, probably falls into a separate group from the
other cytotoxic enterotoxins because it inhibits glucose uptake, energy production and
macromolecular synthesis within the intestinal epithelium (McDonel and Duncan,
1976, 1977; McDonel, 1979).
The toxin has been estimated to have a mol. wt of 35 000 and binds to intestinal
cells with a suggested binding site density of 1 . 9 6 lo6
~ sites/cell (McDonel, 1979).
However, this binding is not reversible, which suggests that some form of trapping
mechanism occurs within the cell. This would account for the fact that the toxin
produces functional “holes” in Vero (African green monkey) cells according to the
methods described by Thelestam and Mollby (1976). The “holes” allowed the release
of nearly 100% of the 14C-aminobutyric acid (mol. wt 103) and 50-60% of 3H-uridine
(mol. wt 224) from the cells. Markers with higher molecular sizes were not released,
which suggests that the functional holes were small but, nevertheless, large enough or
frequent enough to cause changes in cell morphology, viability and macromolecular
synthesis. Within the intestinal mucosa, levels of CAMP remain unaltered throughout
enterotoxin treatment. Using everted ileal sacs from rats, McDonel and Duncan
(1976) showed that the toxin causes a significant reduction in oxygen consumption by
the sacs. Cellular glycolysis was unaffected but oxidative metabolism (e.g., oxygen
consumption by isolated mitochondria) was impaired. Such damage to the respiratory control within the intestinal cells may or may not be secondary to the structural
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
229
BA CTERIAL E N T E R 0 TOXINS
damage caused, especially a t the villus tips (McDonel et al., 1978) but nevertheless may
result in the fluid and electrolyte changes that culminate in diarrhoea (fig. 6).
An amphipathic structural model has been proposed for the enterotoxin, with one
tightly formed hydrophobic domain and another, less rigid, hydrophilic domain
(Granum and Whitaker, 1980). Limited proteolysis of the toxin has been reported to
activate the enterotoxin about threefold (Granum, Whitaker and Skjelkviile, 1981). In
particular, trypsination yields a polypeptide of 16 000 mol. wt that can inhibit cell-free
protein biosynthesis (Granum, 1982) and intracellular protein biosynthesis may be
inhibited after the entry of part of the complete enterotoxin molecule into the cell
cytosol. It has been suggested that the hydrophilic part of the molecule is responsible
for binding to cell receptors and the hydrophobic moiety is released through the
membrane. Amino acid sequencing of the N-terminal end of the trypsin-activated
toxin (Richardson and Granum, 1983) revealed certain similarities between C.
perfringens enterotoxin and the corresponding region of sub-unit B of cholera toxin
which is, itself, associated with membrane binding.
Staphylococcus aureus
Several antigenically distinct staphylococcal enterotoxins are produced by a
significant percentage of clinical isolates of Staph. aureus (Bergdoll, 1976). A
summary of the physicochemical features of the different serotypes of toxin is
presented in table XI. Symptoms of intoxication in man include vomiting, diarrhoea
and nausea and all serotypes produce similar clinical features.
However, although much is known about the structural features of the various
A
8
mse
serosal
side
mucosal
side
serosal
side
mucosal
side
FIG.6.-Fluid and electrolyte changes in intestinal mucosa induced by C. perfringens enterotoxin. Modified
from McDonel(l979). A represents the normal flow of electrolytes across the intestinal epithelium; B
represents the electrolyte flow induced by C. perfringens enterotoxin.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
C. G. G E M M E L L
230
TABLE
XI
Staphylococcal enterotoxins
Serological
type of
enterotoxin
A
B
C
c2
E
F
Mol. wt
Isoelectric point
(PI) of
main form
Number
of
electrophoretic
forms
27 800
28 366
34 100
34 100
29 600
22 000
7.3
8.6
8.6
7.0
7.0
7.2
4
4
3
3-8
not known
1
enterotoxin types the biochemical mechanism by which the toxins act is largely
unknown. One suggestion for their activity has been direct cytotoxicity causing
changes in intestinal transport of fluids and electrolytes (Sullivan, 1969; Huang, Chen
and Rout, 1974). Contrary evidence has recently been presented by Buxser and
Bonventre (198 1) who studied the putative leakage of several radioactively-labelled
markers from human embryonic intestinal epithelial cells (Henle 407) by enterotoxin.
Enterotoxin failed to produce any “functional holes” in the cell membranes and, even
in the presence of several staphylococcal cytotoxins, no synergistic cellular damage was
found. Nor did the enterotoxin affect the ability of the Henle cells to take up amino
acids or to synthesize protein, RNA or DNA.
Bacillus cereus
Culture filtrates of isolates of B. cereus from cases of gastroenteritis contain, among
other exo-products, an enterotoxin of mol. wt 55 000 (Spira and Goepfert, 1975) which
resembles several of the enterotoxins of gram-negative bacteria. It causes fluid
accumulation in the ligated rabbit ileum, activates adenylate cyclase and enhances
vascular permeability in rabbit skin. It is distinct from a low mol. wt (5000)
heat-stable product of other strains of B. cereus associated with food poisoning
(Turnbull et al., 1979) which is responsible for vomiting but not diarrhoea in the
affected individuals.
Clostridium dificile
It is only within the last 10 years that our understanding of the pathogenesis of
antibiotic-associated diarrhoea and pseudomembraneous colitis (PMC) has been
developed. In particular C. dzficile has been identified as the aetiological agent. It
produces two antigenically and biologically distinct toxins; toxin A (an enterotoxin)
causes the accumulation of haemorrhagic fluid in rabbit ileal loops and caecitis in
hamsters, and toxin B (a cytotoxin) can damage a wide spectrum of tissue culture cells.
Toxin A has been purified (Sullivan, Pellett and Wilkins, 1982) and shown to be a
hydrophobic protein of mol. wt 550000. Its primary action is to cause fluid
accumulation within the intestine but not by activation of adenylate cyclase. In
addition it has some cytotoxic activity but very much less than seen with toxin B. Toxin
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
BA CTERIA L E N T E R 0 TOXINS
23 1
B is also a hydrophobic protein with a mol. wt of 350 000. PMC is caused by the
production of large amounts of both toxins in the adult colon. In the somewhat
similar disease in hamsters, both toxins seem to be of equal importance because the
animals have to be immunised against both to survive (Libby, Jortner and Wilkins,
1982). In man, toxin A perpetuates the diarrhoea and increases the permeability of the
intestinal epithelium while toxin B causes areas of necrosis and haemorrhage. Both
toxins might thereafter enter the circulation where they could cause lethal effects far
removed from the intestinal tract.
Summary
It is apparent that there are considerable similarities between many of the
enterotoxins produced by enteric pathogens. Although the effect of most of these
toxins is restricted to the intestine in vivo, many cells are also sensitive to intoxication in
vitro. The resultant in-vitro biochemical changes may have no pathological
significance but serve to underline the central role of cyclic nucleotides in cellular fluid
regulation. The biological activity of these enterotoxins is the result of interaction
with membrane-bound adenylate cyclase, leading to persistent elevation of intracellular levels of CAMP. Stimulation of adenylate cyclase occurs consistently after a
characteristic lag phase which varies somewhat between toxins. The duration and
degree of stimulation of adenylate cyclase by the various toxins may point to possible
differences in affinity, dissociation and mechanism of activation of the cyclase
molecule. Subtle events at, or within, the cell membrane must occur during
intoxication and may include complex associations of toxin with membrane lipid and
protein components.
The heat-labile toxins of V. cholerae, E. coli, Salmonella spp., A . hydrophila and Y .
enterocolitica have much in common in their structures, membrane receptors and
biochemical modes of action. Similarly the heat-stable toxins of E. coli and Y .
enterocolitica, match each other in their biological activities. Classified along with the
enterotoxin of C . perfringens, the enterotoxin produced by Sh. dysenteriae (and
possibly some strains of E. coli) appears to differ from the other enterotoxins by acting
on protein biosynthesis primarily and not on the nucleotide cyclase activation systems.
In another category must be placed the various enterotoxins produced by Staph. aureus
until more is known. Surprisingly little research has been directed towards the
elucidation of their mode of action, although much is known of their serological and
structural differences. Evidence to date suggests that staphylococcal enterotoxins
differ from the other diarrhoeagenic agents discussed in this review.
The structural and immunological similarities between the various heat-labile
enterotoxins suggest a common genetic origin with gene transfer between the different
bacterial species being responsible for the spread of enterotoxigenicity. It is possible
that many of the “newer” enterotoxins owe their origin to genetic recombination with
the “older” enteropathogens like V . cholerae.
REFERENCES
Annapurna E, Sanyal S C 1977 Enterotoxicity of Aerornonas hydrophila. Journal o f Medical
Microbiology 10:317-323.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
232
C. G . G E M M E L L
Baloda S B, Faris A, Krovacek K, Wadstrom T 1983 Cytotonic enterotoxins and cytotoxic
factors produced by Salmonella enteritidis and Salmonella typhimurium. Toxicon 2 1:
785-796.
Bennett V, Cuatrecasas P 1975 Mechanism of activation of adenylate cyclase by Vibrio cholerae
enterotoxin. Journal of Membrane Biology 22:29-52.
Bergdoll M S 1976 Concerning the structure of the staphylococcal enterotoxins. In: Jeljaszewicz
J. (ed) Staphylococci and staphylococcal disease. Zentralblatt fur Bakteriologie, Parasitenkunde, Infektionskrankheiten und Hygiene, I, Abt. Orig. Suppl. 5:583-586.
Berkenbile F, Delaney R 1976 Stimulation of adenylate cyclase by Vibrio cholerae toxin and its
active subunit. Journal of Infectious Diseases 133 (Suppl.):82-88.
Brown J E, Rothman S W, Doctor B P 1980 Inhibition of protein synthesis in intact HeLa cells
by Shigella dysenteriae 1 toxin. Infection and Immunity 29:98- 107.
Boulanger Y, Lallier R, Cousineau G 1977 Isolation of enterotoxigenic Aeromonus from fish.
Canadian Journal of Microbiology 23: 1 161- 1164.
Bourne H R, Lehrer R I, Lichtenstein M, Weissmann G, Zurier R 1973 Effects of cholera
enterotoxin on adenosine 3’,5’-monophosphate and neutrophil function. Comparison with
other compounds which stimulate leukocyte adenyl cyclase. Journal of Clinical Investigation 52:698-708.
Buxser S, Bonventre P F 1981 Staphylococcal enterotoxins fail to disrupt membrane
integrity or synthetic functions of Henle 407 intestinal cells. Infection and Immunity
3 1:929-934.
Cassel D, Pfeuffer T 1978 Mechanism of cholera toxin action: covalent modification of the
guanyl nucleotide-binding protein of the adenylate cyclase system. Proceedings of the
National Academy of Sciences of the USA 75:2669-2673.
Cassel D, Selinger Z 1977 Mechanism of adenylate cyclase activation by cholera toxin:
inhibition of GTP hydrolysis at the regulatory site. Proceedings of the National Academy of
Sciences of the USA 74:3307-331 1.
Cassuto J, Jodal M, Tuttle R, Lundgren 0 1979 The effect of lidocaine on the secretion induced
by cholera toxin in the cat small intestine. Experientia 3511467-1468.
Charney A N, Gots R E, Formal S B, Giannella R A 1976 Activation of intestinal mucosal
adenylate cyclase by Shigella dysenteriae 1 enterotoxin. Gastroenterology 70: 1085-1090.
Clausen C R, Christie D L 1982 Chronic diarrhea in infants caused by adherent enteropathogenic Escherichiu coli. Journal of Pediatrics 100:358-361.
Dallas W S, Falkow S 1979 The molecular nature of heat-labile enterotoxin (LT) of Escherichia
coli. Nature 277:406-407.
Dallas W S, Gill D M, Falkow S 1979 Cistrons encoding Escherichiu coli heat-labile toxin.
Journal of Bacteriology 139:850-858.
Donowitz M, Keusch G T, Binder H J 1975 Effect of Shigella enterotoxin on electrolyte
transport in rabbit ileum. Gastroenterology 69: 1230-1237.
Donta S T 1974 Differentiation between the steroidogenic effects of cholera enterotoxin and
adrenocorticotropin through use of a mutant adrenal cell line. Journal of Infectious
Diseases 129:728-731.
Donta S T, King M, Sloper K 1973 Induction of steroidogenesis in tissue culture by cholera
enterotoxin. Nature New Biology 243:246-247.
Dupont H L, Formal S B, Hornick R B, Synder M J, Libonati J P, Sheahan D G, Labrec E H,
Kalas J P 1971 Pathogenesis of Escherichia colidiarrhea. New England Journal of Medicine
285: 1-9.
Eiklid K, Olsnes S 1980 Interaction of Shigella shigae cytotoxin with receptors on sensitive and
insensitive cells. Journal of Receptor Research 1: 199-213.
Eiklid K, Olsnes S 1983 Animal toxicity of Shigella dysenteriae cytotoxin: evidence that the
neurotoxic enterotoxic, and cytotoxic activities are due to one toxin. Journal of
Immunology 130:380-384.
Field M 1981 Secretion of electrolytes and water by mammalian small intestine. In: Johnson, L R
(ed) Physiology of the gastrointestinal tract. Raven Press, New York, pp 963-982.
Field M, Fromm D, Wallace C K, Greenough W B 1969 Stimulation of active chloride secretion
in small intestine by cholera exotoxin. Journal of Clinical Investigation 48:24A.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
BA CTERIAL E N T E R 0 TOXINS
233
Finkelstein R A, Larue M K, Johnston D W, Vasil M L, Cho G J, Jones J R 1976 Isolation and
properties of heat-labile enterotoxin(s) from enterotoxigenic Escherichia coli. Journal of
Infectious Diseases. 133 (Suppl.): 120-137.
Formal S B, O’Brien A D, Gemski P, Doctor B P 1978 Invasive Escherichia coli. Journal of the
American Veterinary Medical Association 173:596-598.
Fujita T, Osaka M, Yanatori Y 1974 Granule release of enterochromaffin (EC) cells by cholera
enterotoxin in the rabbit. Archives of Histology Japan 36:367-378.
Giannella R A, Luttrell M, Drake K 1980 Binding of pure Escherichia coli heat-stable
enterotoxin to isolated rat intestinal villus cells. Clinical Research 28:764A.
Gill D M 1976 The arrangement of subunits in cholera toxin. Biochemistry 15:1242-1248.
Gill D M, Hope J A, Meren R, Jacobs D S 1981 Cholera toxin’s interactions with cell
membranes. In: Middlebrook J L, Kahn L D (eds) Receptor-mediated binding and
internalization of toxins and hormones. Academic Press, New York, pp 113-121
Granum P E 1982 Inhibition of protein synthesis by a tryptic polypeptide of Clostridium
perfringens type A enterotoxin. Biochimica et Biophysica Acta 706:6- 1 1.
Granum P E, Whitaker J R 1980 Perturbation of the structure of Clostridium perfringens
enterotoxin by sodium dodecyl sulphate, guanidine hydrochloride, pH and temperature.
Journal of Food Biochemistry 4:219-234.
Granum P E, Whitaker J R, Skjelkvile R 1981 Trypsin activation of enterotoxin from
Clostridium perfringens type A. Fragmentation and some physicochemical properties.
Biochimica et Biophysica Acta 668:325-332.
Greenberg R W, Guerrant R L 1981 E. coli heat-stable enterotoxin. Pharmacology and
Therapeutics 13:507-531.
Guerrant R L, Brunton L L, Schnaitman T C, Rebhun L I, Gilman A G 1974 Cyclic adenosine
monophosphate and alteration of Chinese hamster ovary cell morphology: a rapid sensitive
in vitro assay for the enterotoxins of Vibrio cholerae and Escherichia coli. Infection and
Immunity 10:320-327.
Guerrant R L, Hughes J M, Chang B, Robertson D C, Murad F 1980 Activation of
intestinal guanylate cyclase by heat-stable enterotoxin of Escherichia coli: Studies of
tissue specificity, potential receptors and intermediates. Journal of Infectious Diseases
142:220-228.
Gurwith M J, Wiseman D A, Chow P 1977 Clinical and laboratory assessment of the
pathogenicity of serotyped enteropathogenic Escherichia coli. Journal of Infectious
Diseases 135:736-743.
Holmgren J 1973 Comparison of the tissue receptors for Vibrio cholerae and Escherichia coli
enterotoxins by means of gangliosides and natural cholera toxoid. Infection and Immunity
81851-859.
Holmgren J 1980 Cellular action and pathophysiological effects of cholera toxin. In: Smith H,
Skehel J J, Turner M J (eds) The molecular basis of microbial pathogenicity. Verlag Chemie
GmbH, Weinheim, pp 269-284.
Holmgren J, Svennerholm A M 1979 Immunological cross-reactivity between Escherichia coli
heat labile enterotoxin and cholera toxin A and toxin B subunits. Current Microbiology
2155-58.
Huang K C, Chen S T, Rout W R 1974 Effect of staphylococcal enterotoxins A, B and C on ion
transport and permeability across the flounder intestine. Proceedings of the Society for
Experimental Biology and Medicine 147:250-254.
Kapitany R A, Scoot A, Forsyth G W, Mckenzie S L, Worthington R W 1979 Evidence for two
heat-stable enterotoxins produced by enterotoxigenic Escherichia coli. Infection and
Immunity 24:965-966.
Keusch G T, Donohue-Rolfe A, Jacewicz M 1981 Shigella toxin(s): description and role in
diarrhea and dysentery. Pharmacology and Therapeutics 15:403-438.
Keusch G T, Donta S T 1975 Classification of enterotoxins on the basis of activity in cell culture.
Journal of Infectious Diseases 131:58-63.
Keusch G T, Grady G F, Mata L J, Mciver J 1972 The pathogenesis of Shigella diarrhea 1.
Enterotoxin production by Shigella dysenteriae 1. Journal of Clinical Investigation
5 1:1212-121 8.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
234
C. G . GEMMELL
Klipstein F A, Engert R F 1977Immunological interrelationships between cholera toxin and the
heat-labile and heat-stable enterotoxins of coliform bacteria. Infection and Immunity
181110-117.
Kohl S, Jacobson J A, Nahmias A 1976 Yersinia enterocolitica infections in children. Journal of
Pediatrics 89:77-79.
Kohler E M, Bohl E H 1966 Studies of Escherichia coli in gnotobiotic pigs. I Experimental
reproduction of colibacillosis. Canadian Journal of Comparative Medicine 30: 199-203.
Koupal L R, Deibel R H 1975 Assay, characterization and localization of an enterotoxin
produced by Salmonella. Infection and Immunity 11:14-22.
Libby J M, Jortner B S, Wilkins T D 1982 Effects of the two toxins of Clostridium dificile in
antibiotic-associated cecitis in hamsters. Infection and Immunity 36:822-829.
Ljungh A, Eneroth P, Wadstrom T 1982a Steroid secretion in adrenal Y l cells exposed to
Aeromonas hydrophila enterotoxin. FEMS Microbiology Letters 15:141-144.
Ljungh A, Eneroth P, Wadstrom T 1982b Cytotoxic enterotoxin from Aeromonas hydrophila.
Toxicon 20:787-794.
Ljungh A, Kronevi T 1982 Aeromonas hydrophila toxins-intestinal fluid accumulation and
mucosal injury in animal models. Toxicon 20:397-407.
Ljungh A, Wretlind B, Mollby R 1981 Separation and characterisation of enterotoxin and two
haemolysins from Aeromonas hydrophila. Acta Pathologica et Microbiologica Scandinavica 89:387-397.
McDonel J L 1979 Molecular mode of action of Clostridium perfringens enterotoxin. American
Journal of Clinical Nutrition 32:2 10-218.
McDonel J L, Duncan C L 1976 Mechanism of action of Clostridium perfringens enterotoxin in
the rat ileum: protein synthesis and cyclic 3'-5' adenosine monophosphate. Proceedings of
the 11th Joint Conference on Cholera, held November 1975, New Orleans; US-Japan
Cooperative Medical Science Program, DHEW, Bethesda, MD, pp 297-305.
McDonel J L, Duncan C L 1977 Effect of Clostridiumperfringens enterotoxin on mitochondria1
respiration. Infection and Immunity 15:999-1001.
McDonel J L, Chang L W, Pounds J G, Duncan C L 1978 The effects of Clostridium perfringens
enterotoxin on rat and rabbit ileum: an electron microscopic study. Laboratory Investigation 39:210-218.
Madsen G L, Knoop F C 1980 Physiochemical properties of a heat-stable enterotoxin produced
by Escherichia coli of human origin. Infection and Immunity 28: 1051-1053.
Molina N C, Peterson J W 1980 Cholera toxin-like toxin released by Salmonella species in the
presence of Mitomycin C. Infection and Immunity 30:224-230.
Moss J, Fishman P H, Manganiello V C, Vaughan M, Brady R 0 1976 Functional incorporation
of ganglioside into intact cells: induction of choleragen responsiveness. Proceedings of the
National Academy of Sciences of the USA 73:1034-1037.
Newsome P M, Burgess M N, Mullan N A 1978Effect of Escherichia coli heat-stable enterotoxin
on cyclic GMP levels in mouse intestine. Infection and Immunity 22:290-291.
O'Brien A D, Laveck G D 1983Purification and characterization of a Shigella dysenteriae 1-like
toxin produced by Escherichia coli. Infection and Immunity 40:675-683.
O'Brien, A D, Laveck G D, Thompson M R, Formal S B 1982 Production of Shigella dysenteriae
type 1-like cytotoxin by Escherichia coli. Journal of Infectious Diseases 146:763-769.
Okamoto K, Inoue T, Shimizu K, Hara S, Miyama A 1982 Further purification and
characterization of heat-stable enterotoxin produced by Yersinia enterocolitica. Infection
and Immunity 35:958-964.
Olsnes S, Eiklid K 1980 Isolation and characterization of Shigella shigae cytotoxin. Journal of
Biological Chemistry 255:284-289.
Olsnes S, Reisbig R, Eiklid K 1981 Subunit structure of Shigella cytotoxin. Journal of Biological
Chemistry 256:8732-8738.
Peterson J W 1980 Salmonella toxin. Pharmacology and Therapeutics 1 1:719-724.
Reisbig R, Olsnes S, Eiklid K 1981The cytotoxic activity of Shigella toxin. Evidence for catalytic
inactivation of the 60s ribosomal subunit. Journal of Biological Chemistry 25637398744.
Richards K L, Douglas S D 1978 Pathophysiological effects of Vibrio cholerae and
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09
B A CTERIAL E N T E R 0 TOXINS
235
enterotoxigenic Escherichia coli and their exotoxins on eucaryotic cells. Microbiological
Reviews 42:592-613.
Richardson M, Granum P E 1983 Sequence of the amino-terminal part of enterotoxin from
Clostridium perfringens type A: identification of points of trypsin activation. Infection and
Immunity 40:943-949.
Robins-Browne R M, Still C S, Miliotis M D, Koornhof H J 1979 Mechanism of action of
Yersinia enterocolitica enterotoxin. Infection and Immunity 25:680-684.
Sack R B 1975 Human diarrheal disease caused by enterotoxigenic Escherichia coli. Annual
Review of Microbiology 29:333-353.
Sandefur P D, Peterson J W 1976 Isolation of skin permeability factors from culture filtrates of
Salmonella typhimurium. Infection and Immunity 14:671-679.
Spira W M, Goepfert J M 1975 Biological characteristics of an enterotoxin produced by Bacillus
cereus. Canadian Journal of Microbiology 2 1:1236-1246.
Sullivan R 1969 Effects of enterotoxin B on intestinal transport in vitro. Proceedings of the
Society for Experimental Biology and Medicine 131:1 159-1 162.
Sullivan N M, Pellett S, Wilkins T D 1982 Purification and characterization of toxins A and B of
Clostridium dzjicile. Infection and Immunity 35: 1032-1040.
Thelestam M, Mollby R 1976 Cytotoxic effects on the plasma membrane of human diploid
fibroblasts, a comparative study of leakage tests. Medical Biology 54:39-49.
Thomlinson J R, Buxton A 1963 Anaphylaxis in pigs and its relationship to the pathogenesis of
oedema disease and gastroenteritis associated with Escherichia coli. Immunology
6: 126-139.
Turnbull P C B, Kramer J M, Jorgensen K, Gilbert R J, Melling J 1979 Properties and
production characteristics of vomiting, diarrhoea1 and necrotizing toxins of Bacillus cereus.
American Journal of Clinical Nutrition 32:2 19-228.
Wadstrom T, Ljungh A, Wretlind B 1976 Enterotoxin, haemolysin and cytotoxic protein in
Aeromonas hydrophila from human infections. Acta Pathologica et Microbiologica
Scandinavica, Section B 84: 1 12-1 14.
Zenser T V, Metzger J F 1974 Comparison of the action of Escherichia coli enterotoxin on the
thymocyte adenylate cyclase-cyclic adenosine monophosphate system to that of cholera
toxin and prostaglandin E 1. Infection and Immunity 10503-509.
Zurier R B, Weissmann G, Hoffstein S, Kammerman S, Tai H H 1974 Mechanisms oflysosomal
enzyme release from human leukocytes. I1 Effects ofcAMP and cGMP, autonomic agonists
and agents which affect micro-tubule function. Journal of Clinical Investigation
531297-309.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Thu, 15 Jun 2017 13:04:09