Amino-terminal domain of the El Tor haemolysin of Vibrio cholerae O1 is expressed in classical strains and is cytotoxic

Amino-terminal domain of the El Tor haemolysin of Vibrio cholerae O1 is expressed in classical strains and is cytotoxic R i c h a r d A. A i m * , G r a h a m M a y r h o f e r , I e v a K o t l a r s k i a n d P a u l A. M a n n i n g *

Previous studies have shown that the classical isolates ofVibrio cholerae possess an 11 bp deletion in the structural gene f o r the El Tor haemolysin leading to the production o f a 27 k D a non-haemolytic truncated product HlyA* compared to the 82 k D a haemolysin, HlyA. These studies were designed to assess whether this truncated product had an), biological activity. A Km R cartridge was introduced into the hlyA gene effectively eliminating the haemolysin. This was recombined into the chromosome o f a varieO, o / strains and isogenic pairs were examined in a number o f systems. These studies suggest that the haemolytic (cytolytic) domain o f HlyA resides at the C-terminus and that the N-terminus, which is conserved as HlyA* in classical strains, possesses enterotoxic (cytotoxic) activity. Experiments with the cholera-toxinless vaccine candidate JBK70 and its hlyA:.'Km R mutant suggest that HlyA* may be responsible for the residual diarrhoea observed in cholera-toxinless vaccine strains. Keywords: Vibrio cholerae; haemolysin; El Tor; classical; toxin; virulence; vaccine

INTRODUCTION Cholera toxin (CT)-producing Vibrio cholerae O1 are responsible for cholera in human beings. Some non-O1 V. cholerae isolates produce an enterotoxin that is similar to CT 1 a; however, other isolates from clinical cases of' diarrhoea do not produce this enterotoxin indicating that there are other factors which can generate a diarrhoeal response. These non-O1 V. cholerae strains have been shown to produce a haemolysin, and, although its role in pathogenesis remains unclear, it has been implicated as a potential diarrhoeagenic factor 4. The haemolysin produced by non-O1 isolates has been shown by Yamamoto et al. 5 to be biologically, physicochemically and immunologically indistinguishable from the haemolysin that is produced by E1 Tor strains of V. cholerae O 1. This is supported by Southern hybridization analyses which imply that the genes are essentially identical 6'~. The haemolytic factor from V. cholerae O1 characterized by Honda and Finkelstein8 was cytotoxic, cardiotoxic and rapidly lethal for mice. McCardell et al. 9 and Spira et al. ~° have described a cytolysin that is produced by non-O1 and O1 strains of V. cholerae. This cytolysin causes fluid accumulation in the rabbit gut as well as being cytotoxic for Chinese hamster ovary (CHO) and Y-1 adrenal cells, and is proposed to be identical to the haemolysin produced by E1 Tor strains of V. cholerae Department of Microbiology and Immunology, University of Adelaide, Adelaide, SA 5001, Australia. *Present address: Department of Biochemistry and Microbiology, University of Victoria, Victoria BC, Canada. tTo whom correspondence should be addressed. (Received 27 July 1990; revised 20 December 1990; accepted 8 January 1991) 0264-410x/91/080588-07 ~ 1991 Butterworth-HeinemannLtd 588 Vaccine, Vol. 9, August 1991

O1. Haemolysins possessing cytotoxic activity have been described in a broad range of Vibrio species which include V. vulnificus 11, V. damsela 12, V. fluvialis 13, V. parahaemolyticus 14 and V. furnissii 13 Several V. cholerae O1 vaccine strains have been constructed using recombinant techniques to remove either the A or B subunit or both, of the cholera enterotoxin 15. When these strains were tested in volunteers, greater than 50% developed mild clinical diarrhoea 16. Further derivatives of these strains that had been made Hly- by deleting the small HpaI-HpaI fragment in the structural gene for the El Tor haemolysin were also tested and no relief from the mild diarrhoea was seen ~. The studies undertaken here show that the haemolysin of V. cholerae O1, biotype E1 Tor is both cytotoxic and cytolytic, and that the loss of the whole haemolytic determinant in a non-toxigenic strain results in a lack of fluid secretion in the rabbit ileal loop. These data also demonstrate that this enterotoxic activity in the haemolysin is associated with the domain which is conserved and expressed in classical strains. MATERIALS AND METHODS

Bacterial strains, media and plasmids V. cholerae strains O17 (El Tor, Ogawa) and 569B (classical, Inaba) were from our laboratory stocks. V. cholerae strains 1074-78, JBK70 (El Tor, Inaba), N16961 (El Tor, Inaba) and H-1 (El Tor, Ogawa) were provided by J. Kaper (Center for Vaccine Development, University of Maryland, Baltimore, MD) and D. Sharma (this department), respectively. Nutrient media were Luria

Vibrio cholerae haernolysin: R.A. Aim et al.

broth as described by Maniatis et al. ~8 and nutrient agar (Blood agar base, Difco) made as recommended without the addition of blood. CFA agar was made according to Evans et al. 19. Plasmid pRK290 was as described by Ditta et al. 2°. Plasmid pPH1J121 was obtained from J. Mekalanos (Department of Molecular Biology, Harvard Medical School, Boston, MA). Plasmid pME30522 was provided by R. Schmitt (Department of Genetics, University of Regensburg, Germany), while pGB213 was obtained from G. Churchward (Department of Molecular Biology, University of Geneva, Switzerland). J225 contains the Km R cartridge (Pharmacia) cloned into pUC19 and was obtained from J. Hackett (this department). Plasmid pPM431 which contains the 6.3 kb PstI fragment harbouring the hly locus of V. cholerae strain 017 cloned into pBR322 has been previously described 24.

SDS-polyacryalmide gel electrophoresis Electrophoresis of proteins was performed in SDS on 14% polyacrylamide gels using a modification of the procedure described by Lugtenberg et al. zs as described previously by Achtman et al. 26. Western blotting was performed according to Towbin et al. 2~.

Bacterial conjugation Exponential phase cultures of the appropriate donor and recipient strains were mixed in a ratio of 1:10, concentrated by centrifugation (3000if, 5 rain) and spread on cellulose acetate filters (Millipore Corp.) placed on a nutrient agar plate. After incubation for 4-5 h at 37°C, the cells were resuspended in 10 ml saline (0.9% w/v NaC1) and plated onto selective media.

Southern hybridization Transfer of DNA from agarose gels to nitrocellulose sheets as described by Southern 28 and radioactive labelling of the probe were done by the procedure of Maniatis et al. ~8.

Liquid haemolysis assay Bacterial cultures at various stages of growth were tested as follows: cells were collected by centrifugation (45 s, Eppendorf 5414) and the supernatant transferred to a clean tube and kept for analysis. The cells were resuspended in fresh Luria broth and subjected to sonication (60 x 1 s pulses, Branson Cell Disruptor) on ice. The cell contents, extracellular growth medium and whole bacterial cultures at the various stages of growth were subjected to analysis. These samples were mixed with an equal volume of 5% sheep red blood cells in PBS and incubated at 37°C. At each designated time point 300 #1 was taken and spun in a microfuge for 45 s. The supernatant (200 #1) was removed and placed in a 96-well tissue culture tray (Flow Laboratories) and twofold across the tray. Samples were read at 414 nm on a ELISA reader (Titertek Multiscan, Flow Laboratories) to determine the amount of haemoglobin released.

Infant mouse cholera model The infant mouse assay was as described by Ujiiye et al. 29 with modifications previously described by

Attridge and Rowley 3°. Exponential phase cultures of the strains to be tested were fed intragastrically to infant

mice (3 days old) that had been removed from food or water for 6 h. The mice were held at 25°C for 48 h. Viable counts were determined by plating dilutions of 10 -5 onto nutrient agar plates and incubating overnight at 37°C. The intestines of the mice were removed post mortem and placed in phosphate-buffered formalin for 48 h. The tissue was set in resin, thin sections cut, stained with haematoxylin and eosin and inspected using an Olympus BH-2 light microscope.

Rabbit ileal loop assay Enteropathogenicity in ileal loops of adult rabbits was carried out essentially as described by Sinha and Srivastava 31. The ileum was ligated into 8 cm segments. Bacterial strains were grown to mid-exponential phase in Luria broth and 1.0 ml samples were injected into closed ileal loops. Accumulation of fluid was scored as positive and loops devoid of fluid as negative. Luria broth was used as a negative control and to assess leakage between ties. The amount of fluid accumulated in the loops was removed by syringes and the volume determined. Viable counts were performed to determine the c.f.u, in the inoculum injected into each loop. Rabbits were derived of food for 24 h and of water for 12 h prior to the experiments.

Cytolytic assays The s 1Cr release assay used was performed as described previously by Brunner et al. 32 using a cultured tumour line of mouse cells, P815. Cells were grown in RPMI 1640 (Gibco Laboratories, NY) and incubated at 37°C in 5% CO2. HEp-2 human epithelial cells were cultured as above and approximately 18 h prior to the experiment, eightwell microscope slide chambers were inoculated with 1 2 × 105 HEp-2 cells and incubated at 37°C in 5% CO z. Infection of cultured HEp-2 cells was done as follows. Bacteria were grown overnight at 25°C on CFA agar and a loopful resuspended in 200 pl phosphate-buffered saline (PBS). HEp-2 cell monolayers were washed with PBS and then incubated with 50 #1 of the bacterial suspension for 15 min at 37°C. The monolayers were washed with PBS and the ceils fixed with methanol at - 2 0 ° C for 2 rain. The monolayers were visualized by light microscopy after staining with Giemsa stain for 5 min.

Construction of a hlyA mutant in V. cholerae The structural gene for the E1 Tor haemolysin, hlyA, has been cloned and sequenced 24'33'34. These and other analyses 35 demonstrated that classical strains all possess an 11 bp deletion leading to the production of a truncated protein HlyA* of 27 kDa, compared to the 82 kDa wild type HIyA protein. In order to analyse the function of both the HlyA protein and the truncated product HlyA*, a Km R cartridge and cloned into the unique M l u I site within the gene and which interrupted the coding sequence. Figure 1 shows the construction of plasmid pPM2018 which contains the P s t I - E c o R I hlyA fragment (with synthetic EcoRI linkers) that was inactivated by the insertion of a Km R cartridge. This was subsequently cloned into the broad host-range vector pRK290 to give plasmid pPM2019 (Figure 2). This construct was transferred to various V. cholerae strains by conjugation. After allowing recombination, this construct was expelled

Vaccine, Vol. 9, August 1991 589

Vibrio cholerae haemolysin: R.A. Aim et al.

EcoRI

Pst I Eco RI

• BomHI

MluI

P s t I ~

pPM431 10,2 kb

EcoF

~

Digest with Pstl and

Psi I est with Pstl and

EcoRI

EcoRI

pPM2OII

7.2kb

~EcoRl

Eco R I Pst I

~

hlyA

PstI

Pst I /

/

Digest with PsH~, blunt end and add Mlu I l i n k e r s ~

~gest

with Mlul

Eco R I

Eco Rr

Digest with Pst I, blunt end and add EcoRI linkers

pPM2OI7 8.7 kb hlyA

KmR Mlu

MluI

oRl MluT

Plasmids pBR322 and pPM431 were digested with Pstl and EcoRI and the hlyA encoding fragment cloned into pBR322 to give plasmid Km R cartridge was removed from plasmid J225 as a Pstl fragment, blunt ended with Klenow fragment and synthetic Mlul linkers added. Plasmid pPM2011 was cleaved with Mlyl and the Km R cartridge inserted to give plasmid pPM2017. Plasmid pPM2017 was digested with Psi1, blunt ended and synthetic EcoRI linkers added to produce plasmid pPM2018. This allowed the inactivated hlyA gene (hlyA::Km R) to be removed as an EcoRI fragment, and inserted into the wide-host range plasmid pRK290 to generate pPM2019 Figure 1

pPM2011. The

from the V. cholerae strains by the introduction of the incompatible plasmid pPH I JI by conjugation. Selection for Km R and the Gm" encoded by pPH1JI ensured that a double crossover recombination event had occurred replacing the intact hly + gene with the inactivated copy. To confirm this construction, a number of tests were performed. Whole genomic D N A digested with PstI was examined by Southern hybridization using the intact hlyA + gene cloned into plasmid pBG2 (pPM2008) as a probe (Figure 3). Plasmid pPH1JI was expelled from the strains using a further IncP plasmid, pME305 which is temperature sensitive for replication, permitting it to be cured by growth at 42°C. The parental strains and their respective mutants were also subjected to haemolysis assays (data not shown), and Western blots of trichloroacetic acid-precipitated culture supernatants were performed using an antiserum to the purified haemolysin (data not shown). These experiments confirm the absence

590

Vaccine, Vol. 9, August 1991

of both haemolytic activity and the HlyA protein as a result of introducing the chromosomal mutation in hlyA.

RESULTS Effect of the hlyA mutation on virulence in the infant mouse cholera model V. cholerae strains O17 and 569B and their respective hlyA::Km R mutants (V760 and V759) were tested for virulence in the infant mouse cholera model. After intragastric inoculation, the mice were checked regularly and their state of health noted (Figure 4). At the doses tested (5 × l0 6) the hlyA::Km R mutant (V759) of hypertoxigenic classical strain 569B killed at a reproducibly slower rate than the parent strain. However, this strain is non-motile, and hence a large inoculum was required to determine the LD~0. In contrast, the hlyA::Km R

Vibrio cholerae haemolysin: R.A. Aim et al.

haemolysis assays on culture supernatants, where it was noted that the haemolytic activity of JBK70 and Inaba N16961 was significantly lower than that of O17. If the haemolysin were responsible for lysis of P815 cells, then the lytic activity of these strains would be expected to be much lower than that of O17. The classical strain 569B and a non-haemolytic E1 Tor strain H-1 did not cause any significant lysis.

Mob pPM2019

HIyA is cytotoxic and cytolytic for cultured HEp-2 cells

/

\

/

\

/

\

Km R ~///////

I

I

I

MluI

MluI

HpoI

I EcoRI

ATG

171 ATG

I

I Eco RI

Hpol

TA A

Monolayers of cultured HEp-2 cells were prepared and incubated with V. cholerae strains and their hlyA::Km R mutants. Incubation of the HEp-2 cell monolayers with O17 led to disruption and significant cell lysis, whereas the hlyA::Km R mutant (V760) produced very little damage (Figure 6). Both the non-haemolytic classical strain 569B and its mutant (V759) did not produce any HEp-2 cell lysis after the incubation; however, there appeared to be changes in the morphology of the HEp-2 cells in the presence of 569B but not V759, suggesting that the HlyA* was having a cytotoxic effect.

I---72_I--1 TGA

Figure 2 Construction of the hlyA::Km R mutation. Plasmid pPM2019 is a derivative of the broad host-range plasmid pRK290 which contains the hlyA gene with the Km Rcartridge inserted into the Mlul site. This plasmid was used for allelic exchange with both El Tor and classical strains and leads to the inactivation of HlyA and HlyA*, respectively

mutant (V760) of the E1 Tor strain O17 was clearly affected in its virulence for mice. When both 569B and O17 or their mutants were administered at a higher dose (5 × 10~), the increased amounts of cholera toxin present appeared to effectively mask the loss of the haemolysin. At the time of death, the intestines of the infant mice were removed and prepared for histological examination (Figure 5). Marked histological differences were detected in the intestines of mice infected with 569B, O17, their hlyA::Km R mutants and an uninfected control. Both 569B and O17 showed significant sloughing off of the epithelium compared to their respective mutants. In all cases the villi were swollen to fill the gut lumen.

=

~-

rw o

uJ

± A

@0 I"-

GO t'--

CO O

O O hi

-

8.57 7.20 6.05 4.90 :3.55 2.65

Effect of HlyA on cultured cells The marked intestinal epithelial damage corresponding to the presence of both HlyA and HlyA* led to further investigation of the cytolytic activity of the haemolysin. A mouse antigen-presenting turnout cell line (P815) was internally labelled with 5~Cr. Culture supernates from exponentially growing V. cholerae cells were incubated with the 5~Cr-labelled P815 for 4 h and the amount of 5~Cr released was measured. The supernatants of these strains were diluted and checked for lyric activity (Table 1). This provides an indication of the percentage of lysed 51Cr-P815 cells. Culture supernatants of the haemolytic strain O17 and the non-toxigenic environmental strain 1074 78 readily lysed the cultured cells, whereas their respective hlyA::Km R mutants (V760 and V738) had lost this ability. The cholera toxin-deleted candidate vaccine strain JBK70, used by Levine et al. ~6, and its parent strain Inaba N16961 displayed reduced lyric activity which was confirmed when plated on blood agar by the production of a smaller zone of haemolysis than that of' O17. A similar result was obtained by performing

1.73 1.61 1.29 1.19 0.99

Figure 3 Whole genomic DNA of the V. cholerae strains and their hlyA::Km R mutants were digested with Pstl and electrophoresed on a 0.8% (w/v) agarose gel. After transfer to nitrocellulose, the filter was probed with [3=P]-dCTP nick translated pPM2008, washed and subjected to autoradiography. The 6.4 kb Pstl fragment harbouring the hly locus (open arrowhead) present in the wild-type strains is absent in all of the hlyA::Km" mutants, but a new 7.9 kb fragment representing the 6.4 kb Pstl fragment containing the Km R cartridge was detected (solid arrowhead). A cross-reacting band corresponding to pPH1JI can be detected in all of the mutants. Bacillus subtilis phase SPP1 DNA digested with EcoRI was used as a standard and the size of the fragments generated (kb) are shown on the right

Vaccine, Vol. 9, August 1991 591

Vibrio cholerae haemolysin: R.A. Aim et al. I00

Effect of X

V. cholerae strains O17, 569B and JBK70 and their respective hlyA::Km R mutants (V760, V759 and V736) were introduced into ligated ileal segments of adult rabbits. The amounts of fluid secreted into the rabbit gut were examined after 5 and 12 h (Table 2). After 5 h fluid accumulation was detected with strain 569B but not with V759. Only after 12 h did both 569B and V759 cause equivalent amounts of fluid accumulation. The El Tor strain O 17 produces a very viscous mucoid fluid after 12 h

x'X /

f

80

hlyA mutants in rabbit ileal loops

PX

60-

So

,....,

40Table

1

Percentage 51Cr release from culture supernatant Percentage released a (s.d.)

20 Strain

0

I

l

I0

20

30

40

50

Time ( h ) Figure 4 The V. cholerae strains O17 ( × ), 569B ( 0 ) and their respective hlyA::Km R mutants, V760 ([~) and V759 (A), were administered intragastrically to infant mice. The number of surviving mice at each time point was recorded. The mice were fed with 5 × 10" organisms

1/2

1/4

1/32

O17 V760 (017hlyA::Km R)

77.1 (5.3) 0.8 (3.2)

81.4 (0.5) 0.2 (2.2)

75.0 (1.7)

1074-78 V738 (1074 78hlyA::Km R)

51.5 (6.6) 0.6 (1.7)

26.9 (1.3) 1.2 (2.2)

4.5 (1.9)

N16961 V735 (N16961hlyA::Km R)

36.1 (3.3) 1.8 (2.4)

11.6 (5.0) - 6 . 5 (1.9)

JBK70 V736 (JBK7OhlyA::Km R)

18.3 (4.6) 4.5 (1.0)

1.4 (3.9) - 2 . 5 (2.5)

569B V759 (569hlyA::Km R)

- 1.0 (3.03) - 4 . 7 (1.2)

-- 0.76 (3.2) nd

H-1 V737 (H-lhlyA::Km R)

1.0 (3.0) 1.2 (1.1)

nd nd

s.d., standard deviation; nd, not determined "The culture supernatants were diluted 1/2, 1/4 or 1/32 and the amount of spontaneous release has been deducted

Figure 5 The intestines of infant mice were removed post mortem, preserved in formalin, set in resin and thin sections were cut. The sections were stained with haematoxylin and eosin and examined by light microscopy. The marked histological differences between saline (a), uninfected (b), 569B (c), 569B hlyA::Km R (d), O17 (e) and O17 hlyA::Km R (f) are apparent. (e) and (c) display marked swelling of the intestinal villi which was absent from the villi of mice infected with the mutants. The epithelial border in (e) has also degenerated when compared with the mutant strain in (f). (Magnification × 250)

592

V a c c i n e , Vol. 9, A u g u s t

1991

Figure 6 Monolayers of cultured HEp-2 cells, a human epithelial cell line, were incubated with V. cholerae strains O17 and 569B and their hlyA::Km R mutants, V760 and V759. (a), uninfected monolayer; (b) and (c), monolayers infected with strains O17 and V760, respectively. The cells in (b) show extensive lysis and only cell debris can be seen after microscopic examination. (d) and (e), monolayers infected with strains 569B and V759, respectively. No significant lysis is seen with strain 569B or its hlyA::Km R mutant, although some cell elongation is seen in (d) but not (e). (Magnification × 250)

Vibrio cholerae haemolysin: R.A. Aim et al. Table 2

Fluid accumulation in rabbit ideal loops Fluid accumulation (ml cm 1)

Strain

5h

12 h

Luria broth O17 V760 (017hlyA::Km R) 569B V759 (569BhlyA:: Km R) JBK70 V736 (JBK7OhlyA::Km R)

0 0.26 0 0.89 0 nd nd

0 1.53 1.65 1.75 1.81 0.26 0

nd, not determined

and the fluid clearly contained free haemoglobin when compared to the non-haemolytic classical strain. However, the O17 hlyA::Km R mutant (V760) caused no fluid accumulation after 5 h and the fluid seen after 12 h was less viscous and lacking any apparent free haemoglobin. At 12 h the non-toxigenic strain JBJ70 caused mild fluid accumulation, whereas the hlyA ::Km Rmutant (V736) did not cause fluid accumulation. Sections of the rabbit ileum were taken and prepared for histological examination and the cell morphology paralleled that seen in the infant mice intestines.

quently constructed a mutation at the MluI site to remove most of HlyA*. A general problem with all immunogenic, live, attenuated V. cholerae vaccine candidates to date is that they produce mild diarrhoea in a significant proportion •of volunteers, making them unacceptable as a public health measure against cholera 16. Our observations suggest that HlyA* is expressed in classical strains and, of course, this activity is also contained within HIyA in E1 Tor strains. This activity leads to early fluid accumulation in rabbit ileal loops but is masked later by the effects of cholera toxin, especially in hypertoxigenic strains such as 569B or when large inocula are used. It can also readily be envisaged that elimination of such an early effect by the introduction of a hlyA mutation could account for the time shift in the death curves of infected infant mice. Both the ability of JBK70 to produce HlyA* and the mild fluid accumulation seen in rabbit ileal loops, are eliminated with the introduction of the hlyA mutation. We conclude that HlyA* is, at least in animal models, responsible for the residual toxicity and diarrhoea observed in the candidate vaccine strains constructed to date.

DISCUSSION

ACKNOWLEDGEMENTS

It has previously been shown that a cytotoxin seen in V. cholerae of both the O1 and non-O1 serovars stimulates fluid secretion in the rabbit ileal loop assay, and that this cytotoxin appears identical to the E1 Tor haemolysin 36. To ascertain whether the haemolysin was responsible for these activities, as well as being an alternative diarrhoeagenic factor in non-toxigenic candidate vaccine strains, a specific mutant in the structural gene, hlyA, was constructed. An insertion mutation was introduced into the MluI site since this would not only prevent the HlyA protein being made but also the HlyA* product seen in the classical V. cholerae strain 569B by Alm et al. 3a. The experiments reported here clearly indicate that the E1 T o t haemolysin, HlyA, of V. cholerae has both cytotoxic (enterotoxic) and cytolytic (haemolytic) activity. The cytolytic activity is not associated with the truncated protein, HlyA*, produced by classical strains, typified here by strain 569B. This is consistent with our previous observations 24'33 which suggest that the ability to lyse sheep erythrocytes rests in the C-terminal portion of HlyA, corresponding to the region not expressed in HlyA* of classical strains. Together with other data our observations demonstrate that the N-terminal region of HlyA which is conserved in classical strains is in the form of HlyA .35, and the data here suggest that it has enterotoxic activity. The candidate vaccine strain JBK70 was shown to induce a good immune response and provide significant protection against challenge with wild type Inaba N 16961 but it caused mild diarrhoea in > 5 0 % of recipients 16. Kaper et al. 17 investigated the role of the E1 Tor haemolysin in JBK70 by creating a small 398 bp deletion between the two HpaI sites within the hlyA gene (Figure 2). When this new recombinant strain, CVD104, was fed to the volunteers, the residual diarrhoea still occurred 16. Since the removal of this HpaI fragment would only decrease HlyA* by 2.5 kDa, we felt that this may not satisfactorily inactivate the HlyA* product, and conse-

The authors are grateful to the National Health and Medical Research Council of Australia and the Clive and Vera Ramaciotti Foundations for their support. RAA was supported by a Commonwealth Postgraduate Research Award. The technical assistance of Chris Cursaro, Marj Quin and Garry Penney was greatly appreciated. REFERENCES 1

Craig, J.P., Yamamoto, K., Takeda, Y. and Miwatani, T. Production of cholera-like enterotoxin by a strain of Vibrio cholerae non-O1 isolated from the environment. Infect. Immun. 1981, 34, 90-97 2 Yamamoto, K., Takeda, Y., Miwatani, T. and Craig, J.P. Evidence that a non-O1 Vibrio cholerae produces enterotoxin that is identical to cholera enterotoxin. Infect. Immun. 1983, 39, 1128-1135 3 Zinnaka, Y. and Carpenter, C.C.J. An enterotoxin produced by non-cholerae vibrios. Johns Hopkins Med. J. 1972, 131,403-411 4 Mclntyre, O.R. and Feeley, J.C. Characteristics of non-cholera vibrios isolated from cases of human diarrhoea. Bull. WHO 1965, 32, 627-632 5 Yamamoto, K., Ichinose, Y., Nakasone, N., Tanabe, M., Nagahama, M., Sakurai, J. and Iwanaga, M. Identity of hemolysins produced by Vibrio cholerae non-O1 and V. cholerae O1, biotype El Tor. Infect. Immun. 1986, 51,927-931 6 Goldberg, S.L. and Murphy, J.R. Molecular cloning of the hemolysin determinant from Vibrio cholerae El Tor. J. Bacteriol. 1984, 160, 23~244 7 Brown, M.H. and Manning, P.A. Haemolysin genes of Vibrio cholerae: presence of homologous DNA in non-haemolytic O1 and haemolytic non-O1 strains. FEMS Microbiol. Lett. 1985, 30, 197-201 8 Honda, T. and Finkelstein, R.A. Purification and characterization of a hemolysin produced by Vibrio cholerae biotype El Tor: Another toxic substance produced by cholera vibrios. Infect. Immun. 1979, 26, 1020-1027 9 McCardell, B.A., Madden, J.M. and Shah, D.B. Isolation and characterization of a cytolysin produced by Vibrio cholerae serogroup non-O1. Can. J. Microbiol. 1985, 31,711-720 10 Spira, W.M., Sack, D.A., Fedorka-Cray, P.J., Sanyal, S., Madden, J. and McCardell, B. Description of a possible new extracellular virulence factor in nontoxigenic Vibrio cholerae O1. In: Advances in Research on Cholera and Related Diarrheas, Vol. 3 (Eds Kuwahara, S. and Pierce, N.F.) KTK Scientific Publishers, Tokyo, 1986, pp. 267-270

Vacci ne, Vol. 9, A u g u s t 1991

593

V i b r i o c h o l e r a e h a e m o l y s i n : R.A. A i m et al. 11

12 13

14 15

16

17

18

19

20

21

22

23

594

Wright, A.C., Morris, J.G., Maneval, D.R., Richardson, K. and Kaper, J.B. Cloning of the cytotoxin-hemolysin gene of Vibrio vulnificus. Infect. Immun. 1985, 50, 922 924 Kreger, A.S., Cytolytic activity and virulence of Vibrio damsela. Infect. Immun. 1984, 44, 326-331 Wall, V.W., Kreger, A.S. and Richardson, S.H. Production and partial characterization of a Vibrio fluvialis cytotoxin. Infect. Immun. 1984, 46, 773777 Joseph, S.W., Colwell, R.R. and Kaper, J.B. Vibrio parahaemolyticus and related halophilic vibrios. Crit. Rev. Microbiol. 1983,18, 77-124 Kaper, J.B., Lockman, H., Baldini, M. and Levine, M.M. Recombinant nontoxinogenic Vibrio cholerae strains as attenuated cholera vaccine candidates. Nature (London) 1984, 398, 65,5-658 Levine, M.M., Kaper, J.B., Herrington, D., Losonsky, G., Morris, J.G., Clements, M.L. et al. Volunteer studies of deletion mutants of Vibrio cholerae O1 prepared by recombinant techniques. Infect. Immun. 1988, 56, 161 167 Kaper, J.B., Mobley, H.L.T., Michaeiski, J.M., Herrington, D.A. and Levine, M.M. Recent advances in developing a safe and effective live oral attenuated Vibrio cholerae vaccine. In: Advances in Research on Cholera and Related Diarrheas Vol. 6, KTK Scientific Publications, Tokyo, 1986 Maniatis, T., Fritsch, E.F. and Sambrook, J. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982 Evans, D.G., Evans, D.J. Jr, and Tjoa, W. Hemagglutination of human group A erythrocytes by enterotogenic and enteropathogenic Escherichia coil isolated from adults with diarrhea: correlation with colonization factor. Infect. Immun. 1977, 18, 33(F337 Ditta, G., Stanfield, S., Corbin, D. and Helinski, D.R. Broadhost range DNA cloning system for Gram negative bacteria: Construction of a gene bank of Rhizobium meliloti. Proc. Natl Acad. Sci. USA 1980, 77, 7347 7351 Beringer, J.E., Beynon, J.L., Buchanan-Wollaston, A.V. and Johnston, A.W.B. Transfer of the drug resistance transposon Tn5 to Rhizobium. Nature (London) 1978, 276, 633-634 Rella, M., Mercenier, A. and Haas, D. Transposon insertion mutagenesis of Pseudomonas aeruginosa with a Tn5 derivative: application to physical mapping of the arc gene cluster. Gene 1985, 33, 29,~303 Churchward, G., Belin, D. and Nagamine, Y. A pSC101-derived plasmid which shows no sequence homology to other commonly used cloning vectors. Gene 1984, 31, 165-171

V a c c i n e , Vol. 9, A u g u s t 1991

24

25

26

27

28

29

30 31

32

33

34

35 36

Manning, P.A., Brown, M.H. and Heuzenroeder, M.W. Cloning of the structural gene (hly) for the haemolysin of Vibrio cholerae El Tor strain O17. Gene 1984, 31, 225-231 Lugtenberg, B., Meijers, J., Peters, R., van der Hoek, P. and van Alphen, L. Electrophoretic resolution of the major outer membrane proteins of Escherichia coil K-12 into four bands. FEBS Lett. 1975, 58, 254-258 Achtman, M., Schwuchow, S., Helmuth, R., Morelli, G. and Manning, P.A. Cell-celt interactions in conjugating Escherichia coil: Con mutants and stabilization of mating aggregates. Mol. Gen. Genet. 1978, 164, 171-183 Towbin, H., Staehlin, T. and Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Prec. Natl Acad. Sci. USA 1979, 76, 43~354 Southern, E.M. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 1975, 98, 503517 Ujiiye, A., Nakatomi, M., Sogame, S., Iwanaga, M. and Kobari, K. Experimental cholera in mice. I. First report on the oral infection. Trop. Med. 1968, 18, 6,5-71 Attridge, S.R. and Rowley, D.R. The role of the flagellum in adherence of Vibrio cholerae. J. Infect. Dis. 1983, 147, 864qB72 Sinha, V.B. and Srivastava, B. Suppression of pathogenicity by P and V plasmids in Vibrio cholerae. J. Gen. Microbiol. 1978, 104, 251255 Brunner, K., Mauel, J., Cerottini, J.C. and Chapuis, B. Quantitative assay of the lytic action of immune lymphoid cells on 51Cr-labelled allogenic target cells in vitro; inhibition by isoantibody and by drugs. Immunology 1968, 14, 181 196 Aim, R.A., Stroeher, U.H. and Manning, P.A. Extraceliular proteins of Vibrio cholerae: nucleotide sequence of the structural gene (hlyA) for the haemolysin of the haemolytic El Tor strain O17 and characterization of the hlyA mutation in the non-haemolytic classical strain 569B. Molec. Microbiol. 1988, 2, 481~88 Rader, A.E. and Murphy, J.R. Nucleotide sequences and comparison of the hemolysin determinants of Vibrio cholerae El Tor RV79(Hly +) and RV79(Hly-) and Classical 569B(Hly-). Infect. Immun. 1988, 58, 1414~1419 Aim, R.A. and Manning, P.A. Biotype specific probe for Vibrio cholerae serogroup O1. J. Clin. Microbiol. 1990, 28, 823~824 Madden, JiM, McCardell, B.A. and Shah, D.B. Cytotoxin production by members of genus Vibrio. Lancet 1984, i, 1217 1218