Isolated DNA molecule encoding ShET2 of Shigella flexneri 2a

Abstract

Substantially pure enterotoxins of Shigella flexneri 2a are described, along with a method for obtaining the same, antibodies having binding specificity to the enterotoxins and a method for use of the enterotoxins to develop a non-reactogenic Shigella flexneri 2a vaccine candidate.

Description

FIELD OF THE INVENTION

The present invention relates to two substantially pure enterotoxins of Shigella flexneri 2a (hereinafter "ShET1" and "ShET2"), a method for obtaining the same, antibodies having binding specificity to the enterotoxins and a method for use of the enterotoxins to develop a non-reactogenic Shigella flexneri 2a vaccine candidate.

BACKGROUND OF THE INVENTION

Much has been written about the molecular pathogenesis of Shigella with respect to the genes and gene products involved in their ability to invade epithelial cells, and thereby to cause dysentery (Makino et al, Microb. Pathog., 5:267-274 (1988); Sansonetti et al, Infect. Immun., 35:852-860 (1982); Hale et al, Infect. Immun., 40:340-350 (1983); Pal et al,. J. Clin. Microbiol., 27:561-563 (1989); and Venkatesan et al, Proc. Nat'l. Acad. Sci. U.S.A., 85:9317-9321 (1988)). In contrast, surprisingly little is known of the precise mechanisms by which Shigella cause watery diarrhea.

Although the cardinal feature of the pathogenesis of Shigella flexneri 2a infection involves the invasion of epithelial cells, because Shigella flexneri 2a can cause watery diarrhea, it has been hypothesized that Shigella flexneri 2a also produces an enterotoxin (Rout et al, Gastroenterology, 68:270-278 (1975); and Kinsey et al, Infect. Immun., 14:368-371 (1976)). More specifically, the following observations have suggested the existence of enterotoxins in Shigella flexneri 2a:

1. Clinically in humans, Shigella flexneri 2a infections are usually characterized by a period of watery diarrhea that precedes the onset of scanty dysenteric stools of blood and mucus (DuPont et al, J. Infect. Dis., 119:296-299 (1969); and Stoll et al, J. Infect. Dis., 146:177-183 (1982)). In mild cases, only watery diarrhea may occur, leading to a clinical picture undistinguishable from that due to enterotoxingenic E. coli infection (Taylor et al, J. Infect. Dis., 153:1132-1138 (1986); and Taylor et al, J. Clin. Microbiol., 26:1362-1366 (1988)). 2. When Shigella are fed to monkeys, three clinical syndromes are seen (Route et al, Gastroenterology, 68:270-278 (1975)). Some monkeys develop only dysentery; some exhibit only watery diarrhea and some exhibit watery diarrhea and dysentery. In vivo perfusion studies by Rout et al, Gastroenterology, 68:270-278 (1975), showed that net transport of water into the lumen of the colon occurs in all ill animals. In contrast, only in the jejunum of monkeys with overt watery diarrhea (alone or followed by dysentery) does there occur net secretion of water, sodium and chloride ions; such net transport does not occur in the jejunum of monkeys manifesting dysentery without watery diarrhea. Net secretion in the jejunum was not accompanied by abnormal histological findings in this anatomic site of the small intestine. 3. The net secretion of water and electrolytes into the jejunum of monkeys with watery diarrhea requires the passage of Shigella through the jejunum (Kinsey et al, Infect. Immun., 14:368-371 (1976)). This was demonstrated by bypassing the small intestine and inoculating Shigella directly into the cecum of monkeys. Of 16 monkeys who developed clinical illness, 15 manifested dysentery, ". . . only rarely preceded by mild diarrhea". Net secretion of water and sodium into the colon was recorded in ill monkeys that developed dysentery following intracecal inoculation, while no abnormalities of water or electrolyte transport were observed in the jejunum of the ill animals.

Together, these observations suggest that Shigella elaborate an enterotoxin that elicits secretion early in the infection as the organisms pass through the jejunum.

However, except for the cytotoxin/neurotoxin/enterotoxin elaborated by Shigella dysenteriae (O'Brien et al, Microbiol. Rev., 5:206-220 (1987); Keusch et al, Pharmac. Ther., 15:403-438 (1982); and Fontaine et al, Infect. Immun., 56:3099-3109 (1988)), but not by other Shigella species, little convincing proof has been generated to substantiate the contention that Shigella, other than Shigella dysenteriae, in fact produce enterotoxins.

More specifically, previous attempts in the art to detect enterotoxic activity in supernatants of Shigella flexneri 2a have yielded positive findings in only one instance. O'Brien et al, Infect. Immun., 15:796-798 (1977), partially purified a toxin produced by Shigella flexneri 2a strain M4243 that was 10 detectable in cell-free supernatants. This toxin stimulated fluid production in rabbit ileal loops, but was also cytotoxic for HeLa cells in monolayers and was lethal when inoculated intraperitoneally into mice. Further, it was not necessary to grow the bacteria in Fe.sup.++ -depleted medium in order to detect the enterotoxic activity. In addition, the cytotoxicity of the toxin described by O'Brien et al, supra, was neutralized by anti-sera to Shiga (Shigella dysenteriae 1) toxin.

Enterotoxic activity in cell-free supernatants of Shigella flexneri 2a and 3a was reported by Ketyi et al, Acta Microbiol. Acad. Sci. Hung., 25:165-171 (1978); Ketyi et al, Acta Microbiol. Acad. Sci. Hung., 25:219-227 (1978); and Ketyi et al, Acta Microbiol. Acad. Sci. Hung., 25:319-325 (1978). Filtered ultrasonic lysates of two Shigella flexneri 2a and 3a strains were founds to give rapid fluid accumulation in rabbit ileal loops (4 hour assay). However, the loops showed no fluid accumulation when examined at 18-24 hours after inoculation. Only three loops were inoculated for each of the two test strains, and when examined at 4 hours, only 2/3 for one strain and 1/3 for the other strain were positive. In addition, the Shigella were not cultured in Fe.sup.++ -depleted medium.

In the present invention, it was discovered for the first time that enterotoxic activity, which is clearly dissociated from cytotoxic activity, is expressed by Shigella flexneri 2a in the bacteria-free culture supernatant, and could be detected only after growth of the bacteria in Fe.sup.++ -depleted medium.

It has been reported that when grown in Fe.sup.++ -depleted medium, enteroinvasive Escherichia coli (EIEC) elaborate an enterotoxin (MW circa 68-80 kDa) that causes fluid accumulation in isolated rabbit ileal loops and an electrical response in Ussing chambers (Fasano et al, Infect. Immun., 58:3717-3723 (1990)). Based on the similarities known to exist between enteroinvasive E. coli and Shigella (Levine et al, J. Infect. Dis., 155:377-389 (1987)), it was postulated in the present invention that Shigella flexneri 2a would express an enterotoxin when grown in Fe.sup.++ -depleted medium.

In the present invention, it was unexpectedly disclosed that Shigella flexneri 2a produces two distinct enterotoxins, one encoded by the chromosome, and the other encoded by an invasiveness virulent plasmid. The latter enterotoxin was found in the present invention to be essentially the same as the EIEC enterotoxin.

SUMMARY OF THE INVENTION

An object of the present invention is to purify the two enterotoxins produced by Shigella flexneri 2a.

Another object of the present invention is to provide a method for culturing Shigella flexneri 2a so as to produce said enterotoxins.

A further object of the present invention is to provide antibodies having binding specificity for said enterotoxins.

An additional object is to identify, clone and sequence the genes encoding such enterotoxins.

Still another object of the present invention is provide Shigella flexneri 2a mutants which fail to produce at least one functional enterotoxin as a result of a mutation in a Shigella enterotoxin gene.

These and other objects of the present invention have been achieved in the detailed description of the invention provided below.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows the results of assays for enterotoxic activity in Ussing chambers when using culture supernatants of Shigella flexneri 2a strains M4243, M4243avir and BS103, the 30-100 kDa fraction of EIEC strain CVD/EI-34 (0136:H-) (as a positive control) and culture media (as a negative control). In these assays, variations in short-circuit current (delta Isc) were measured.

FIG. 2 shows the results of assays for enterotoxic activity in Ussing chambers when using Shigella flexneri 2a strain M4243 culture supernatant which was first neutralized with anti-sera against the Shigella flexneri 2a enterotoxins (anti-ShETs), with anti-sera against the EIEC enterotoxin (anti-EIET), with pre-challenge sera or post-challenge sera of volunteers challenged with wild-type Shigella flexneri 2a. In these assays, variations in short-circuit current (delta Isc) (FIGS. 2A and 2C), and transepithelial electrical potential differences (delta PD) (FIGS. 2B and 2D) were measured.

FIGS. 3A-3B show the molecular mass determination of the Shigella flexneri 2a strain M4243 enterotoxic moieties when assayed in rabbit ileal loops (FIG. 3A), and in Ussing chambers (FIG. 3B). In the rabbit ileal loop assays, fluid accumulations were measured and in the Ussing chambers, variations in short-circuit current (delta Isc) were measured.

FIG. 4 shows the results of assays for enterotoxic activity in Ussing chambers when using protein bands from SDS-PAGE obtained from strain M4243avir (containing only ShET1 enterotoxin) that represent the 65-75 kDa column fraction, an extract of an unused strip of nitrocellulose (negative control), and a sample representing the 65-75 kDa column fraction (positive control). Values are given as Isc variation (.mu.Amp/cm.sup.2).

FIG. 5 shows a restriction map for the fragment in pJS26 which contains the tie gene, as well as restriction maps for the relevant portions of plasmids derived from pJS26.

FIGS. 6A-6D (SEQ ID NO:1) show the DNA sequence of EIET enterotoxin encoded by enteroinvasive E. coli, as well as the determined amino acid sequence.

FIGS. 7A-7D (SEQ ID NO:2) show the DNA sequence of ShET2 enterotoxin located on the Shigella flexneri 2a invasiveness plasmid, as well as the determined amino acid sequence.

DETAILED DESCRIPTION OF THE INVENTION

In the present invention, the enterotoxins are obtained by culturing Shigella flexneri 2a in Fe.sup.++ -depleted medium and collecting the supernatant.

"Fe.sup.++ -depleted media" is an expression well-known and used in the art. This expression refers to iron-depleted media, such as syncase broth, treated, e.g., in CHELEX.RTM. (BioRad), a styrene divinyl benzene resin matrix with iminodiacetic acid exchange groups, to leave just traces of iron in the medium.

The particular culture medium employed is not critical to the present invention. Examples of such culture media include Fe.sup.++ -depleted syncase broth or L-broth plus ethylenediamine-N-N'-diacetic acid (EDDA). Fe.sup.++ -depleted syncase broth is the preferred culture medium since maximal production of the enterotoxins was obtained with this medium.

While the culture temperature and incubation period are not critical to the present invention, generally the culturing temperature will range from 30.degree. to 37.degree. C. preferably 36.degree. to 37.degree. C., and the incubation period will range from 24 to 72 hours, preferably 48 to 72 hours.

The enterotoxins can be purified from the supernatant by size exclusion and HPLC chromatography.

Shigella flexneri 2a is a well-known virulent Shigella serotype available from a variety of sources, such as the Center for Vaccine Development, the Centers for Disease Control, the Walter Reed Army Institute of Research, the Uniformed Services University of the Health Sciences, and the Institut Pasteur. The particular strain of Shigella flexneri 2a employed in the present invention is not critical thereto. Examples of such Shigella flexneri 2a strains include M4243, M4243avir, Shigella flexneri 2a Chile 747, Shigella flexneri 2a Chile 3480 (Ferreccio et al, Am. J. Epi., 134:614-627 (1991)), strain 2457T (Kotloff et al, Infect. Immun. 60:2218-2224 (1992)) and BS103 (Andrews et al, Infect., Immun., 59:1997-2005 (1991)). The preferred Shigella flexneri 2a strains employed in the present invention is Shigella flexneri 2a strain M4243 and M4243avir.

Shigella flexneri 2a strain M4243 and its plasmid-cured derivative M4243avir can be obtained from, e.g., Dr. Samuel B. Formal of the Walter Reed Army Institute of Research, Washington, D.C. BS103 can be obtained from Dr. Anthony Maurelli of the Uniformed Services University of the Health Sciences, Bethesda, MD.

The antibodies having binding specificity to the two enterotoxins of the present invention may be polyclonal or monoclonal. Polyclonal antibodies to the purified enterotoxins can be prepared by conventional means as described in Antibodies: A Laboratory Manual, Harlow and Lane, Eds., Cold Spring Harbor Laboratory Press (1988). Monoclonal antibodies to the purified enterotoxins can be prepared by conventional means as described in Kohler et al, Nature, 256:495-497 (1975).

Monoclonal antibodies obtained using purified enterotoxins may be used to induce passive immunity against Shigella enteric infection. Such antibodies will bind the Shigella flexneri 2a enterotoxins, thus preventing their interaction with the cellular receptor, and preventing the stimulation of water and electrolyte secretion.

The substantially pure enterotoxins of the present invention are also useful for the development of a non-reactogenic Shigella flexneri 2a candidate live oral vaccine. As background, in the United States, Shigella flexneri 2a is one of the most common serotype of Shigella associated with disease. In developing countries of the world, Shigella flexneri is the most common serogroup of Shigella causing diarrheal disease and Shigella flexneri 2a is often the single most common serotype. Prospective epidemiologic studies in a low socioeconomic community in Santiago, Chile, where Shigella infections are endemic, have shown that an initial clinical episode of shigellosis confers significant protection against subsequent disease due to the same serotype (Ferroccio et al, Am. J. Epidemiol., 134:614-627 (1991)). The immunizing effect of diarrheal illness due to wild-type Shigella has also been demonstrated in a volunteer model of experimental shigellosis where an initial clinical infection due to Shigella flexneri 2a (DuPont et al, J. Infect. Dis., 125:12-16 (1972)) or Shigella sonnei (Herrington et al, Vaccine, 8:353-357 (1990)) conferred significant protection against re-challenge with the homologous wild-type organism. Together these observations suggest that it may be possible to protect against shigellosis with a vaccine that requires only a single dose.

There have been many attempts to develop attenuated strains of Shigella to serve as vaccinees. Some attempts have met with limited success. In the 1960s, streptomycin-dependent strains of Shigella flexneri 2a and other serotypes were developed and utilized as live oral vaccines (Mel et al, Bull. WHO, 32:647-655 (1965); Mel et al, Bull. WHO, 39:375-380 (1968); and Mel et al, Acta Microbiol. Acad. Scient. Hung., 21:109-114 (1974)). These streptomycin-dependent strains were safe and conferred significant serotype-specific protection against shigellosis in most of the controlled field trials of efficacy that were carried out (Mel et al, Bull. WHO, 32:647-655 (1965); Mel et al, Bull. WHO, 39:375-380 (1968); Mel et al, Acta Microbiol. Acad. Scient. Hung., 21:109-114 (1974); and Levine et al, Am. J. Epidemiol., 133:424-429 (1976)). However, the streptomycin-dependent Shigella vaccines suffer from certain drawbacks. One is the fact that multiple spaced doses have to be given to confer protection (four doses over a two-week period containing large numbers (2-4.times.10.sup.10) of viable vaccine organisms). Moreover, protection is relatively short-lived. A booster dose has to be given after one year in order to maintain protection (Mel et al, Acta Microbiol. Acad. Scient. Hung., 21:109-114 (1974)). Colonial mutant Shigella flexneri 2a vaccine strain T.sub.32 described by Istrari et al, Arch. Roumaines Pathol. Exp. Microbiol., 24:677-686 (1985), is also well-tolerated and protective (Wang Bing Rui, Arch. Roumaines pathol. Exp. Microbiol., 43:285-289 (1984)), but still requires multiple doses.

Because of the above-mentioned drawbacks of the streptomycin-dependent and T.sub.32 vaccines of the 1960s, various investigators have attempted to make more immunogenic Shigella vaccines that can protect following the administration of just a single dose. The approaches taken have included:

(1) introducing specific segments of the chromosome of E. coli K-12 into Shigella by conjugation (Formal et al, Dev. Bios. Stand., 15:73-78 (1971); and Levine et al, J. Infect. DIS., 127:261-270 (1973)); (2) introducing DNA encoding protective Shigella antigens into E. coli K-12 (Formal et al, Infect. Immun., 46:465-469 (1984)); and (3) inactivating genes of the aromatic amino acid biosynthesis pathway, thereby rendering the Shigella nutritionally dependent on substrates that are not available in human tissues (Lindberg et al, Vaccine, 6:146-150 (1988); and Karnell et al, Rev. Infect. Dis., 13(4):S357-361 (1991)).

Regrettably, each of the above approaches has met with limitations. That is, hybrids in which Shigella carrying attenuating E. coli DNA are unstable, and can revert to full virulence (Levine et al, J. Infect. Dis., 127:261-270 (1973)). Further, the most recent generation of E. coli expressing Shigella antigens has been associated with side reactions in vaccinees, including fever, mild diarrhea and every dysentery in some individuals (Kotloff et al, J. Infect. Immun., 60:2218-2224 (1992)). Finally, some recipients of .DELTA.aroD Shigella flexneri developed mild diarrhea (Karnell et al, Rev. Infect Dis., 13(4):S357-361 (1991)). It has been hypothesized in the present application that the residual diarrhea encountered in these various Shigella flexneri candidate vaccine strains is likely due to the two enterotoxins.

Accordingly, Shigella flexneri 2a vaccine candidates can be constructed which, e.g., in addition to containing other attenuating mutations, express one or two toxoids, rather than the enterotoxins. This can be accomplished by deleting the portion of the enterotoxin genes that encodes the biologically active "toxic" site, leaving intact immunogenic sequences of the proteins.

More specifically, it is preferable to first construct a Shigella flexneri 2a strain in which deletion mutations are introduced in at least one aro gene (aroA, aroC, or aroD) of the Shigella chromosome, rendering the strain auxotrophic for para-aminobenzoic acid, a substrate that cannot be sufficiently scavenged in vivo in human. Strain CVD1202 prepared in Example 7 below contains an aroA deletion, and has been deposited at the American Type Culture Collection under ATCC No. 55517 on Dec. 1, 1993.

In addition, it is also preferable to have an independently attenuating, deletion mutation in the virG gene, which is found on the 140 MD invasiveness plasmid of Shigella flexneri 2a. This plasmid gene, also known as icsa (Sansonetti et al, Vaccine, 7:443-450 (1989)), is involved with the intracellular and intercellular spread of Shigella. This mutation is also present in CVD1202.

Recognizing that the vaccine candidate, e.g., CVD1202, may still not be sufficiently attenuated with just the aroA and VirG mutations (since the ability to produce enterotoxins remains intact), the enterotoxin genes can be mutated as a result of, e.g., the determination of the ShET1 and ShET2 gene sequences. One type of mutation, e.g., a deletion of substantially all of one or both of the enterotoxin genes, will totally inactivate enterotoxin production, resulting in a non-enterotoxinogenic strain. A second mutation, e.g., a deletion of part of one or both of the enterotoxin genes, will result in expression of toxoids, i.e., modified proteins that lacks the toxicity of the toxins but retain immunogenic moieties. This alternative mutation will result in a vaccine candidate strain that expresses, e.g., two toxoids. These toxoids can be used to induce active immunity against Shigella enteric infection.

The particular size of the deletion is not critical to the present invention, and can be readily determined based upon whether one desires to totally inactivate the enterotoxins, or simply produce toxoids.

The isolated DNA molecules of the present invention encoding the enterotoxin genes can be cloned in any suitable plasmid or vector, and used, e.g., to produce large amounts of DNA for use as probes or to integrate mutated enterotoxin genes into vaccine strains.

The expression "isolated" is used herein to mean set apart from its natural environment, e.g., the DNA molecules are separated from the parent chromosome or parent plasmid from which they were originally obtained in the present invention. Thus, "isolated" as used herein includes the presence of the DNA molecules in a foreign host or foreign plasmid.

The following examples are provided for illustrative purposes only and are in no way intended to limit the scope of the present invention.

EXAMPLE 1

Production of Enterotoxins

A. Preparation of Culture Filtrate Fraction

Shigella flexneri 2a strain M4243 and its plasmid-cured derivatives M4243avir and BS103, were grown overnight at 37.degree. C. with shaking (200 rpm) in 5.0 ml of CHELEX.RTM. (BioRad, Richmond, Calif.) treated, Fe.sup.++ -depleted syncase broth (O'Brien et al, J. Infect. Dis., 136:763-759 (1982)). CHELEX.RTM. binds to the iron present in the broth. All culture vessels employed were either new plastic or borosilicate glass soaked overnight in 6.0N HCl, and rinsed in distilled deionized water to ensure the absence of iron. 50 .mu.l of the resulting culture broth were then subcultured in 5.0 ml of Fe.sup.++ -depleted syncase broth in baffled Fernbach flasks, and incubated for an additional 48 hours under the above conditions. After 72 hours of incubation, the cultures were harvested by centrifugation at 12,000 .times.g for 20 minutes at 4.degree. C. and the supernatants were passed through a 0.45 m filter membrane (Millipore Products, Bedford, Mass.) to obtain a "sterile supernatant".

B. Rabbit Ileal Loop Test

Whole cultures of Shigella flexneri 2a strain M4243 and its plasmid-cured derivative M4243avir, along with their respective sterile supernatants, obtained as described above, were tested in a standard rabbit ileal loop test. Supernatants of EIEC strain CVD/EI-34 (0136:H-) (which induces fluid accumulation in rabbit ileal loops), and non-pathogenic E. coli HS, were also included in each experiment as positive and negative controls, respectively (Fasano et al, Infect. Immun., 58:3717-3723 (1991)). EIEC strain CVD/EI-34 (0136:H-) was obtained from the Center for Vaccine Development strain collection. E. coli HS was obtained from Dr. Herman Schneider, Walter Reed Army Institute of Research.

More specifically, male adult New Zealand white rabbits weighing 2-3 kg were starved for 24 hours but allowed water ad libitum. These animals were then anesthetized by intramuscular administration of a cocktail of 50 mg/kg ketamine and 1.0 mg/kg acepromazine, followed by intramuscular administration of 7.0 mg/kg xylazine.

Bacterial cultures were grown to reach 108-109 CFU/ml. Whole cultures, or the respective sterile supernatants, in a standard volume of 1.0 ml, were injected into the lumen of the intestine of the anesthetized rabbits near a tie closest to the mesoappendix (Moon et al, Ann. N.Y. Acad. Sci., 176:197-211 (1971)); a second tie was made to isolate the site of inoculation. Proceeding proximally along the ileum, a series of five to six loops 7-8 cm long separated by double ties were isolated and inoculated (Moon et al, Ann. N.Y. Acad. Sci., 176:197-211 (1971)). After 18 hours of incubation, the animals were sacrificed, the fluid volume and length of the loops were measured, and sections of intestine from each loop were fixed in 10% (v/v) formalinized saline and examined by light microscopy. The results of the loop test are shown in Experiment 1 in Table 1 below.

TABLE 1______________________________________Fluid Accumulation (ml/cm) in Rabbit Ileal Loops______________________________________Experiment 1M4243 bacteria (5) 1.06 .+-. 0.34*M4243 supernatant (5) 0.52 .+-. 0.10**M4243avir bacteria (5) 0.21 .+-. 0.50M4243avir supernatant (5) 0.24 .+-. 0.09HS supernatant (5) 0.09 .+-. 0.06Experiment 2M4243 supernatants:L broth, 24 hours (4) 0.01 .+-. 0.01L broth, 72 hours (4) 0.04 .+-. 0.03Minimal Fe.sup.++ broth, 24 hours (4) 0.43 .+-. 0.11*Minimal Fe.sup.++ broth, 72 hours (4) 0.47 .+-. 0.14*HS supernatant:Minimal Fe.sup.++ broth, 24 hours (4) 0.01 .+-. 0.01______________________________________ In the Table above, the results are expressed as mean .+-. SE for (n) animals. The bacterial cultures were grown for 72 hours unless otherwise indicated. *p < 0.01 compared to HS; **p < 0.05 compared to HS.

As shown in Experiment 1 in Table 1 above, the intestinal loops injected with the positive control, i.e., whole viable cultures of M4243, and sterile culture supernatant therefrom, showed pronounced fluid accumulation at 18 hours post-inoculation, with the whole viable culture showing a two-fold greater fluid accumulation. Further, as shown in Experiment 1 in Table 1 above, fluid accumulation induced by M4243avir (both whole culture and sterile supernatant) was slightly higher than the negative control strain HS.

The fluid to gut length recorded in the rabbit ileal loops, 0.5 ml/cm, measured using graduated syringes (fluid) and a scale (length), was substantially less than seen with enterohemorrhagic E. coli (EHEC) strain 933J, serotype (0157:H7), where ratios of 1.5-2.0 ml/cm occur. However, the recorded fluid to gut length measured using graduates syringes (fluid) and a scale (length) still represents definite evidence of net secretion and fluid accumulation. On histologic examination of the sections of intestine from each loop, severe tissue damage was observed with whole cultures of M4243, characterized by prominent necrosis of the luminal epithelium and marked villus atrophy. In contrast, with M4243 sterile culture supernatant, no tissue damage was detected. Further, no tissue damage was observed with whole cultures of M4243avir or sterile supernatants therefrom. Moreover, no tissue damage was observed with tissue incubated with the negative control strain HS.

To determine whether the time of incubation and the iron content in the medium are crucial for the full expression of this enterotoxic moiety, Shigella flexneri 2a strain M4243 was cultured in Fe.sup.++ -containing medium (L-broth) and Fe.sup.++ -depleted medium (syncase broth). After 24 and 72 hours of incubation for each medium, sterile, supernatants were obtained and then injected into ileal loops, as described above. The results are shown in Experiment 2 in Table 1 above.

As shown in Experiment 2 in Table 1 above, Fe.sup.++ -depleted culture conditions are required in order to detect expression of the enterotoxins. Further, enterotoxin expression was not notably affected by the length of incubation.

The results obtained in the rabbit ileal loop assay were compatible with elaboration of an enterotoxin by M4243.

C. Ussing Chambers

These experiments were performed as previously described by Guandalini et al, J. Pediatr. Gastroenterol. Nutr., 6:953-960 (1987). Briefly, male adult New Zealand white rabbits weighing 2-3 kg were anesthetized by methoxyflurane inhalation and then sacrificed by air embolism. A 20 cm segment of distal ileum was removed, opened along the mesenteric border, rinsed free of intestinal contents, and stripped of muscular and serosal layers. Four pieces of intestine so prepared were then mounted in lucite Ussing chambers (1.12 cm.sup.2 opening), and bathed in Ringer's solution containing 53 mM NaCl, 5.0 mM KCl, 30.5 mM Na.sub.2 SO.sub.4, 30.5 mM mannitol, 1.69 mM Na.sub.2 HPO.sub.4, 0.3 mM NaH.sub.2 PO.sub.4, 1.25 mM CaCl.sub.2, 1.1 mM MgCl.sub.2 and 25 mM NaHCO.sub.3. During the experiment, the tissue was kept at 37.degree. C. and gassed with 95% O.sub.2 -5% CO.sub.2. Once the tissue reached a steady-state condition, 300 .mu.l of either M4243, M4243avir or BS103 sterile supernatants from Fe.sup.++ -depleted cultures were added to the mucosal surface, resulting in a 1:33 dilution of the original culture filtrate concentration (0.3 ml into 10 ml of Ringer's solution). 300 .mu.l of either M4243, M4243avir or BS103 sterile supernatants were also added to the serosal side to preserve osmotic balance. Variation in transepithelial electrical potential difference (delta PD), total tissue conductance (Gt) and short-circuit current (delta Isc) were recorded. The 30-100 kDa supernatant fraction from EIEC (0136:H-), and CHELEX.RTM.-treated syncase broth (culture media) were also tested in the same manner as positive and negative controls, respectively. Four animals were employed for each test. The results are shown in FIG. 1.

As shown in FIG. 1, the overall increase in Isc was significantly greater for the M4243 supernatant as compared to the negative control (culture medium) (**=p<0.02), and similar in magnitude to that induced by the positive control (EIEC 0136:H-). On the other hand, supernatant from the plasmid-cured derivatives M4243avir and BS103 expressed significantly less enterotoxin in comparison with the plasmid-containing parent strain (*=p<0.05). However, the enterotoxic activity of the M4243avir and BS103 supernatants was nevertheless significantly greater than the negative control (culture medium) (*=p<0.05). Possible interpretations of such results include: (1) a plasmid-encoded regulation factor that regulates a chromosomal toxin gene; (2) multiple copies of the same gene located both on the S. flexneri 2a chromosome and the plasmid; or (3) a gene on the invasiveness plasmid encoding for a distinct enterotoxic factor. As discussed in detail below, this last hypothesis turned out to be correct.

The plasmid-cured derivative of strain M4243 showed less enterotoxic activity compared to the wild-type in both ileal loops and in Ussing chambers. Only in Ussing chambers did M4243avir induce changes that were significantly different from the negative control; this could be due to the higher sensitivity of the Ussing chamber technique as compared to the ileal loop assay. These data suggest that, while not absolutely necessary for the effect, the virulence plasmid of Shigella flexneri 2a M4243 enhances enterotoxic activity.

D. Enterotoxin Neutralization

EIEC (0136:H-) and Shigella flexneri 2a share many similarities, e.g., surface antigens, identical plasmids (pInv), clinical manifestations, etc. Thus, neutralization experiments were carried out to determine if there is any immunological relatedness between the enterotoxin produced by EIET (0136:H-) and the enterotoxins produced by M4243.

More specifically, 600 .mu.l of the 30-100 kDa fraction of M4243 sterile supernatant (see Section E. below) were incubated for 60 min at 37.degree. C. with 60 .mu.l of anti-ShET polyclonal sera (anti-Shigella flexneri 2a enterotoxins) or with anti-EIET polyclonal sera (anti-enteroinvasive E. coli enterotoxin) or with pre- or post-challenged convalescent sera.

Anti-ShET polyclonal sera, anti-EIET polyclonal sera and convalescent sera were obtained as described in Example 2.

The resulting samples were tested in Ussing chambers as described in Section C. above with half of each mixture added to each side of a chamber. The results are shown in FIGS. 2A-2D.

As shown in FIGS. 2A-2D, the electrical response in Ussing chambers was drastically reduced when M4243 supernatant was pre-incubated with polyclonal rabbit antibodies raised against the Shigella flexneri 2a enterotoxins (anti-ShETs) or with convalescent sera from volunteers who had been challenged with Shigella flexneri 2a. This neutralization was not observed in either of the pre-immune sera control experiments in which responses were similar to those seen when testing the active fraction alone.

Only a partial cross-neutralization was observed when the M4243 supernatant was pre-incubated with polyclonal antibodies raised against the enteroinvasive E. coli enterotoxin (anti-EIET). In FIGS. 2A-2D, the number of animals tested was 4. Values are mean .+-.SE. *=p<0.05 and **=p<0.02 compared to PBS (the negative control).

Taken together, the above results suggest that S. flexneri 2a supernatant probably contains two enterotoxin moieties, ShET1 (whose gene is located on S. flexneri 2a chromosome) and ShET2 (whose gene is located on the invasiveness plasmid). Both enterotoxins were neutralized when anti-S. flexneri 2a antiserum was used. The ability of EIEC antiserum to partially neutralize the S. flexneri supernatant enterotoxicity was found to be due to the high similarity (at least 99% homology) of the EIET gene with the ShET2 gene (see below).

E. Estimate of Molecular Mass

To obtain an estimate of the M.sub.r of the Shigella flexneri 2a enterotoxins, sterile supernatant of M4243 was fractionated by ultracentrifugation through DIAFLO ultrafiltration membranes (Amicon Corp., Danvers, Mass.). YM100 (100,000-MW cutoff) and YM30 (30,000-MW cutoff) membranes were utilized to produce fractions defined by these size limits. Membrane retentates were washed free of lower molecular weight species with phosphate buffered saline (pH 7.3) (PBS), by two successive 10:1 volume dilutions with PBS, reconcentration, and final reconstitution to the original volume in PBS.

The individual fractions, representing coarse molecular weight pools of >100 kDa, 30-100 kDa and 0.5-30 kDa, were tested for enterotoxic activity in Ussing chambers and ileal loops. The results are shown in FIGS. 3A-3B.

As shown in FIGS. 3A-3B, both ileal loop (FIG. 3A) and Ussing chamber (FIG. 3B) assays localized the active enterotoxic fraction within the 30-100 kDa size range. In FIGS. 3A-3B, the number of animals tested was 4. Values are means +SE. *=p<0.05 and **=p<0.02 compared to the other fractions and the negative control.

F. Cytotoxicity Assay

To establish whether there is a correlation between enterotoxic activity and cytotoxic activity, the following experiments were carried out.

A cell lysate was obtained as follows: Cultures from strain M4243 were harvested by centrifugation at 12,000 .times.g for 20 minutes at 4.degree. C. Supernatants were passed through a 0.45 .mu.m filter, and retained for assay. The bacterial cells were then washed twice in PBS, resuspended in 1.5 ml of PBS and disrupted in a French pressure cell at 12,000 lb/in.sup.2 to obtain a cell lysate (Fasano et al, Infect. Immun., 58:3717-3723 (1991)). The cell lysate was then mixed with 3.5 ml of PBS (final volume 5.0 ml), clarified by centrifugation at 18,000 .times.g for 20 minutes at 4.degree. C., and filter-sterilized using a 0.45 .mu.m membrane.

Fractions of the culture supernatant of strain M4243 were obtained as described in Section E. above.

Cytotoxicity assays were performed on the cell lysate and 3 different culture supernatant fractions (less than 30 kDa, 30-100 kDa, and more than 100 kDa), using Vero cells by the method of Gentry et al, J. Clin. Microbiol., 12:361-366 (1980)). Serial two-fold dilutions (1:2 to 1:64) of the culture supernatant fractions and cell lysate were tested, and the cytotoxic dose required to kill 50% of the Vero cells (CD50) was estimated spectrophotometrically Gentry et al, J. Clin. Microbiol., 2:361-366 (1980)).

Whole culture supernatants and cell lysates of enterohemorrhagic E. coli (EHEC) strain 933J, serotype 0157:H7, which elaborates Shiga-like toxin 1 (SLT1), were used as the positive control in the Vero cell cytotoxicity assay (Fasano et al, Infect. Immun., 58:3717-3723 (1991)). The whole supernatant of non-pathogenic E. coli strains HS, which has been used extensively as a negative control in assays of pathogenicity and in clinical studies (Levine et al, Lancet, I:1119-1122 (1978); and Levine et al, J. Infect. Dis., 148:699-709 (1983)), was used as a negative control in the Vero cell cytotoxicity assay.

Since the positive control (EHEC) killed more than 50% of the Vero cells at a 1:64 dilution, a 10-fold dilution of both supernatants and lysates from EHEC was tested. Cytotoxic titers were expressed as the reciprocal of the CD50/mg protein of the 30-100 kDa culture supernatant fraction or cell lysate; the protein content was measured by the method of Bradford, Anal. Biochem., 72:248-254 (1976).

Both supernatant and lysate of the positive control strain EHEC strain 933J serotype (0157:H7) showed a high level of cytotoxicity (0.5.times.10.sup.3 and 3.4.times.10.sup.4 CD.sub.50 /mg protein, respectively). In contrast, the supernatant of HS, the negative control, showed no cytotoxic activity. Against these two extremes, M4243 exhibited a low-level of cytotoxic activity which was restricted to the less than 30 kDa supernatant fraction (4.2.times.10.sup.2 CD50/mg protein) and the cell lysate (5.1.times.10.sup.2 CD.sub.50 /mg protein).

The cytotoxic assay described above was repeated, except that HeLa cells were substituted for Vero cells. As a result of this experiment, it was determined that the 30-100 kDa fraction obtained from Shigella flexneri 2a supernatant and cell lysate also does not possess any cytotoxic activity against HeLa cells. On the other hand, as expected, and consistent with the results obtained using Vero cells, only the less than 30 kDa supernatant fraction obtained from Shigella flexneri 2a possesses cytotoxic activity against HeLa cells (3.2.times.10.sup.2 CD.sub.50 /mg protein). Also as expected, the cell lysate fraction from Shigella flexneri 2a, which contains the less than 30 kDa fraction possesses cytotoxic activity against HeLa cells (4.4.times.10.sup.2 CD.sub.50 /mg protein).

Thus, the enterotoxin (30-100 kDa fraction) activity and cytotoxin (less than 30 kDa fraction) activity found in Shigella flexneri 2a are the result of two distinct moieties.

Hence, the enterotoxins appear to be responsible for the diarrhea induced by Shigella flexneri 2a, since the 30-100 kDa fraction (where the enterotoxic activity was localized) was responsible for fluid accumulation in rabbit ileal loops and in electrical responses in Ussing chambers.

EXAMPLE 2

Preparation of Antitoxin

A. Preparation of Antibodies in Rabbits

1.0 ml of the 30-100 kDa fraction from the supernatant of Shigella flexneri 2a strain M4243 that showed enterotoxic activity was mixed with an equal volume of Freund's complete adjuvant and inoculated intramuscularly in four separate sites in male New Zealand white rabbits. A booster dose (1.0 ml) was administered four weeks later, and one month thereafter the animals were bled to obtain antisera. Antisera to EIEC enterotoxin (EIET) from strain CVD/EI-34 (0136:H-) was prepared in the identical manner. These antisera are herein referred to as anti-Shigella flexneri 2a enterotoxins (anti-ShETs), and anti-enteroinvasive E. coli enterotoxin (anti-EIET).

B. Preparation of Antibodies in Humans

Pre- and post-challenged (convalescent) serum pools from 10 adult volunteers who developed diarrhea after ingesting Shigella flexneri 2a M4243 (Kotloff et al, Infect. Immun. 60:2218-2224 (1992)) were prepared for use in neutralization experiments in Ussing chambers (FIGS. 2C and 2D), and for Western immunoblots (FIG. 4).

EXAMPLE 3

Purification and Partial Sequencing of Shigella Enterotoxin 1 (ShET1)

A. Purification

Large-scale preparation of Shigella flexneri 2a enterotoxin was undertaken in order to obtain sufficient material for further characterization and analyses. Plasmid-cured S. flexneri 2a M4243avir was used in order to avoid expression of both ShET2 and plasmid-encoded membrane associated proteins (Hale et al, Infect. Immun., 50:620-629 1985)), which are known to be similar in size to the fractions exhibiting enterotoxic activity and to be antigenic in volunteers (Van De Verg et al, J. Infect. Dis., 66:158-161 (1992)).

More specifically, plasmid-cured Shigella flexneri 2a was inoculated into 30 liters of L-broth containing 25 .mu.g/ml of the iron-chelator, ethylenediamine-N-N'-diacetic acid (EDDA) (Rogers, Infect. Immun., 7:445-456 (1973)), and incubated overnight at 37.degree. C. in the New Brunswick Scientific 30 liter fermentor. Bacterial cells were removed by centrifugation at 5,000 .times.g in a Sharples industrial centrifuge, and the supernatant was filtered through a 0.45 .mu.m filter. This filtrate (approximately 30 liters) was fractionated to isolate and concentrate 100-fold the moieties falling within the 30-100 kDa range as described above, except Pellicon tangential flow cassettes (Millipore) were used for ultrafiltration processing of these larger volumes. This filtrate exhibited enterotoxic activity similar to levels observed for smaller batches employing the plasmid-cured strain.

A 10 ml aliquot of the 30-100 kDa concentrate was then further fractionated by replicate separations with an HPLC size exclusion column (SEC-2000, 7.5.times.600 cm with guard column, Phenomenex, Torrance, Calif.). Fractions were eluted from the column with PBS at 0.5 ml/min. The fractions containing moieties in the 65-75 kDa range were collected, pooled and concentrated by vacuum dialysis to 1.0 ml employing a 10 kDa membrane (MicroProDiCon, Spectrum Medical Industries, Los Angeles, Calif.). An aliquot of this material was reserved for enterotoxin assay, and the remainder was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli, Nature, 227:680-685 (1970)) using an 11 cm preparative well with peripheral marker lanes. The resultant 18 bands were transferred to a nitrocellulose membrane by the method of Towbin et al, (Towbin et al, Proc. Natl. Acad. Sci. U.S.A., 76:4350-4354 (1979)). Multiple 2 mm wide vertical strips of the nitrocellulose membrane were prepared and stained with colloidal gold (Aurodye, Janssen Pharmaceutical, Piscataway, N.J.) to visualize protein bands, or reacted with the pooled convalescent sera from humans challenged with Shigella flexneri 2a M4243 by Western immunoblotting techniques (Vial et al, J. Infect. Dis., 158:70-79 (1988)).

Five protein bands were identified by the convalescent serum Western strips indicating their antigenic relatedness. The five protein bands were aligned with the remainder of the nitrocellulose blot which had been reversibly stained with Ponceau S (colloidal gold (Harlow et al, Antibodies: A Laboratory Manual, p. 494 (1988)). Using a scalpel, bands of about 10 cm in length corresponding to immunoreactive material from each of the five protein bands were carefully excised by identification and alignment with the Western and protein stained strips. Material from each of these bands were eluted (Montelero, Electrophoresis, 8:432-438 (1987)) by dissolution of the nitrocellulose in 200 .mu.l of dimethyl sulfoxide, addition of four volumes of water to precipitate the nitrocellulose, followed by centrifugation at 10,000 .times.g, and dialysis of the supernatant against PBS.

Each sample, in addition to the reserved 65-75 kDa sizing column fraction, and material from a mock-blotted and extracted nitrocellulose strip as positive and negative controls, respectively, was then tested for enterotoxic activity in Ussing chambers, as discussed in Example 1 above. The results are shown in FIG. 4.

As shown in FIG. 4, three of the bands, of approximate MW 63 kDa, 53 kDa and 41 kDa, exhibited enterotoxic activity. Replicates of a band corresponding to a MW of 41 kDa showed a consistent mean rise in Isc of 70.4 Amp/cm.sup.2, whereas the 63 kDa and 53 kDa bands exhibited rises in Isc of 24.3 and 19.5 Amp/cm.sup.2, respectively The remaining two immunoreactive bands showed no enterotoxic activity.

The observation that convalescent sera from volunteers who were fed wild-type S. flexneri 2a contain antibodies that neutralize the enterotoxic activity S. flexneri 2a supernatants in Ussing chambers, and that specifically bind to immobilized protein shown to produce such activity, demonstrates that ShET1 is expressed in vivo where it elicits an immune response. Thus, it is likely that this enterotoxin plays a role in the pathogenesis of Shigella diarrhea in humans.

B. N-terminal Sequencing of ShET1

To obtain greater protein mass for sequencing, scale-up of the chromatographic procedure was preformed using Sephacryl S-200 (Pharmacia, Piscataway, N.J.) packed in a calibrated, 4.degree. C. jacketed, 5.times.100 cm XK 50/100 column (Pharmacia). The 65-75 kDa size fraction was handled as above except that a polyvinylidine diflouridine membrane (Immobilon, Millipore), was substituted for nitrocellulose for electrophoretic transfer. The three protein bands, identified as described above, were excised, extensively rinsed with distilled water and dried. Individual strips bearing the protein bands were then subjected to N-terminal sequencing on an Applied Biosystems model 477A sequencer, as described by Hall et al, J. Bacteriol., 171:6372-6374 (1989). The determined N-terminal sequence data are shown in Table 2 below.

TABLE 2__________________________________________________________________________Preliminary N-terminal amino acidsequence of Shigella enterotoxin 1MW of Proposedentero- A:Btoxic subunit N-terminal amino acid sequencemoiety ratio.sup..cndot. 1.sup. 2 3 4 5 6 7 8 9 10 11 12 13 14__________________________________________________________________________63 kDa A1:B3 Ala Pro Pro Val (SEQ ID NO:3) Asp.sup..sctn. Thr Leu53 kDa A1:B2 Ala Pro Pro Val (SEQ ID NO:3) Asp Thr Leu41 kDa A1:B1 Ala Pro Pro Val Pro Ile Asn Pro Ala Xaa Pro Ile Xaa Arg* Asp Thr Glu Phe Arg Arg__________________________________________________________________________ .sup..cndot. assuming an A subunit size of about 30 kDa and a B subunit size of about 11 kDa .sup. sequencing cycle number .sup..sctn. Duplicate amino acid signals detected for samples at position indicated *(SEQ ID NO:4)

As shown in Table 2 above, a definitive extended sequence could not be determined from the material available for any of the three bands. However, the identical putative amino acid sequence was found for the first four residues of all three bands. Moreover, the data derived suggested that two distinct N-termini were being identified. Notably, this was consistent for all three bands examined.

The University of Wisconsin package (Genetics Computer Group, Madison, Wis.) (Devereux et al, Nucleic Acids Res., 12:387-395 (1984)), data bases containing known protein sequences and untranslated DNA sequences were perused to identify those with potential amino acid homology to the putative N-terminal sequences acquired from the above samples. GenBank release 75.0 and PIR Protein 35.0 were also examined using the TFASTA and WORDSEARCH programs. No apparent regions of extensive alignment were found to exist. In addition, no substantial homology to known bacterial toxins was detected.

The common A:B.sub.n active:binding unit motif frequently encountered in bacterial enterotoxins, including cholera toxin (CT) (LoSpalluto et al, Biochem. Biophys. Acta, 257:158-166 (1972)), heat-labile enterotoxin (LT) of enterotoxigenic E. coli (Clements et al, Infect. Immun., 38:806-809 (1982)) and Shiga toxin of S. dysenteriae 1 (Olsnes et al, J. Biol. Chem., 256:8732-8738 (1981); and Seidah et al, J. Biol. Chem., 261:13928-13931 (1986)), may be reflected in the above data. That is, as proposed in Table 2, the apparent molecular sizes of active material are consistent with such stoichiometries based upon the sizes of the A (28-32 kDa) and B (7.7-11 kDa) subunits of the previously identified enterotoxins. By extension, a holotoxin consistent with a size of 65-75 kDa and an A1:B4 structure would be predicted by these conventions. These tentative configurations also satisfy the usual requirements for both a binding and an active domain that allow the enterotoxin to attach and gain entrance to enterocytes and to initiate events that culminate in intestinal secretion.

EXAMPLE 4

Gene sequencing of Enteroinvasive E. coli Enterotoxin

A genetic approach was employed to identify and clone the enterotoxin from enteroinvasive E. coli. More specifically, TnphoA insertion mutants were generated in EIEC strain EI-37 (0136:NM) (Fasano et al, Infect. Immun., 58:3717-3723 (1991)) as described by Taylor et al, J. Bacteriol., 171:1870-1978 (1989). The resulting TnphoA insertion mutants were screened for increased expression of alkaline phosphatase in low iron L-agar (containing 30 .mu.g/ml of EDDA) compared with standard L-agar. As a result, nine insertion mutants with increased expression of alkaline phosphatase were identified.

The supernatants from the resulting nine TnphoA insertion mutants were then tested in Ussing chambers as described above, and two of the mutants were found to have significantly less enterotoxic activity, as defined by changes in I.sub.sc, than the wild-type parent, suggesting that the phoA gene was inserted into the open reading frame that encodes enterotoxic activity.

DNA was then purified from the two mutants, and the purified DNA was digested with BamHI. The resulting DNA fragments, which flank the TnphoA insertions, were cloned into the BamHI site of vector pBluescript Sk/.+-. (Stratagene, La Jolla, Calif.). Then, the cloned DNA was hybridized against a pHC79 cosmid library of EIEC strain EI-34 (Fasano et al, Infect. Immun., 58:3717-3723 (1991)). The flanking DNA sequences from one of the two TnphoA insertion mutants were found to be homologous to nine cosmid clones. Random subcloning of these cosmid clones into pBluescript Sk/.+-. led to the identification of a 2.8 kb HindIII fragment which was found to encode enterotoxin activity in Ussing chambers. This fragment, when cloned into the HindIII site of pBluescript Sk/.+-., gave rise to pJS26 (FIG. 5). DH5.alpha. (Gibco/BRL Life Technologies, Gaithersberg, Md.) was transformed with pJS26, and found to confer reproducible increases in Isc in Ussing chambers.

The 2.8 kb HindIII fragment was manually sequenced, and two potential open reading frames (orf's), encoding predicted peptides of 62.8 kDa and 16.1 kDa were found (FIG. 5).

The 2.8 kb HindIII fragment was digested with ClaI and subcloned into HindIII- and ClaI-digested pBluescript Sk/.+-., to give rise to pJS264, which contained only the 62.8 kDa orf (FIG. 5). DH5.alpha. transformed with pJS264 exhibited rises in Isc in Ussing chambers similar to that found with the entire 2.8 kb HindIII fragment. This orf, whose DNA sequence, along with the determined amino acid sequence are shown in FIGS. 6A-6C (SEQ ID NO:1), was therefore designated tie (for "toxin invasive E. coli").

The 2.8 kb HindIII fragment was also digested with ClaI and subcloned into HindIII- and ClaI-digested pBluescript Sk/.+-., to give rise to pJS263, which contained only the 16.1 kDa orf (FIG. 5). DH5.alpha. transformed with pJS264 did not elicit rises in Isc in Ussing chambers.

A GenBank search for amino acid homology of the translated orf's revealed no significant identity to any known prokaryotic sequences.

The 2.8 kb HindIII fragment containing the tie gene was then digested with AccI and cloned into DH5.alpha. so as to obtain pJS261 (FIG. 5), which was then used to transform DH5.alpha.. The resulting transformant was also found to express enterotoxic activity when tested in Ussing chambers as described above.

In order to gauge the effect of the tie gene on secretory activity, a deletion mutation was constructed by digesting the tie gene in pJS26 with 10NdeI and SphI. The resulting plasmid was designated pJS26a (FIG. 5). This plasmid lacked the first two-thirds of the N-terminus of the open reading frame. This plasmid was then used to transform DH5.alpha., and tested in Ussing chambers as described above. The supernatant obtained from the pJS26a transformants elicited less response in the Ussing chamber assay when compared to pJS26, confirming that tie gene is the EIET structural gene.

Thus, unlike ShET1, which as discussed above is believed to be composed of A and B subunits, EIET is a single molecule.

EXAMPLE 5

Gene sequencing of Shigella enterotoxin 2 (ShET2)

As discussed above, Shigella and EIEC share some similarities. Thus, the orf containing the gene encoding the EIEC enterotoxin shown in FIGS. 6A-6C (SEQ ID NO:1) was used as a probe to determine whether Shigella has similar DNA sequences.

More specifically, purified genomic DNA was obtained from each of S. flexneri 2a M4243 and S. flexneri 2a M4243avir, digested with SalI, and then screened for hybridization with the tie gene. The DNA--DNA hybridization showed the presence of a single 3.5 kb band in genomic DNA from the wild-type strain, but not from the plasmid-cured derivative. This result suggests that the homologous DNA is located on the invasiveness plasmid.

The 3.5 kb SalI fragment was identified on the S. flexneri 2a M4243 plasmid by PCR using the following oligonucleotide primers that hybridize to the tie gene (CAGTGTATCACCACGAG (SEQ ID NO:13); and AAATTATCTACAGTCAG (SEQ ID NO:14)), and sequenced using an automated sequencer. The resulting DNA sequence, along with the determined amino acid sequence are shown in FIGS. 7A-7C (SEQ ID NO:2). As shown in FIGS. 7A-7C (SEQ ID NO:2), this fragment was found to contain a 1595 bp open reading frame and has at least 99% homology to the EIET gene. This Shigella gene encodes for a protein of a predicted MW of 63 kDa, and a pI of 6.36. No leader peptide was identified. The analysis of the peptide structure revealed three possible membrane spanning domains (amino acid positions 120-140, 260-300 and 480-520) and five cysteine residues. A predicted ribosome binding site is found at nucleotide positions 290-293. When the translation of this open reading frame was compared to the N-terminal sequence of ShET1 shown in Table 2, no homologies were found, suggesting that this gene, located on the S. flexneri 2a M4243 plasmid, encodes for a toxin (hereinafter named "ShET2") which is distinct from ShET1, but substantially identical to EIET.

Due to the similarity between the EIET gene and the ShET2 gene, it is evident that the gene located on S. flexneri 2a M4243 plasmid, i.e., that hybridized with EIET gene probe, is the ShET2 structural gene.

EXAMPLE 6

Use of EIEC enterotoxin gene as a DNA probe

The tie gene was used as a DNA probe, and hybridized against a collection of EIEC and Shigella strains under high stringency by the standard colony blot method. The results are shown in Table 3 below.

TABLE 3______________________________________Prevalence of tie Gene in E. coli and ShigellaColony Blot Hybridization with tie ProbeCategory Positive Negative % Positive______________________________________Shigella 27 7 80%EIEC 60 20 75%Other E. coli 0 110 0%______________________________________

As shown in Table 3 above, tie-homologous sequences are present in 80% (27/34) of Shigella strains, including members of all four Shigella species (flexneri, boydii, sonnei and dysenteriae), and 75% of EIEC. None of 110 E. coli tested, other than EIEC, carried homologous sequences.

EXAMPLE 7

Construction of the attenuated S. flexneri 2a strain CVD1202

S. flexneri 2a strain 2457T (Kotloff et al, Infect. Immun. 60:2218-2224 (1992)), known to be virulent based on experimental challenge studies in adult volunteers, was selected as the wild-type parent to be attenuated by introduction of a deletion in both the aroA and VirG genes.

More specifically, the aroA gene (Duncan et al, FEBS, 170:59-63 (1984)) was subjected to polymerase chain reactions in a Programmable Thermal Controller unit, using Tag polymerase and buffer obtained from Promega to obtain a deletion of 201 nucleotides in the aroA gene, which corresponds to a deletion of amino acids 168-231 of the encoded enzyme. In particular, the 5' end of the aroA gene was amplified with the upstream primer (TAATCGAATTCATGGAATCCCTGACGTTA) (SEQ ID NO:5) so as to introduce an EcoRI site, and with the downstream primer (GGTACCCCCAATATTAGGGCCATCAACGTCAACGTTGCCGCC) (SEQ ID NO:6) so as to introduce KpnI and SspI sites. The 3' end of the aroA gene was amplified with the upstream primer (AATATTGGGGGTACCGGTACTTATTTGGTCGAAGGCGATGCA) (SEQ ID NO:7) so as to introduce SspI and KpnI sites, and with the downstream primer (TGATAAGTCGACTCAGGCTGCCTGGCTAAT) (SEQ ID NO:8) so as to introduce a SalI site. Both segments were amplified for 30 cycles of 1 min at 94.degree. C., 2 min at 50.degree. C. and 4 min at 72.degree. C.

In a second PCR reaction, the 5' and 3' segments were fused, and the resulting fusion product was amplified in the same reaction. To facilitate this fusion, the first 15 cycles had an annealing temperature slope (1.degree. C./8 sec from 40.degree. C. to 50.degree. C.+50.degree. C. for 2 min), followed by 15 cycles with an annealing temperature of 55.degree. C. in which the new .DELTA.aroA gene was amplified. The aroA gene of Shigella was cloned into the EcoRI and SalI sites of the suicide plasmid pKTN701 (Hone et al, Vaccine, 9:810-816 (1991)) giving rise to suicide vector pSh.DELTA.aroA2, which was propagated in strain SY327 (Miller et al, J. Bactiol., 170:2575-2583 (1988)). The plasmid was electroporated into strain Sm10.lambda.pir (Miller et al, J. Bactiol., 170:2575-2583 (1988)). Sm10.lambda.pir(pSh.DELTA.aroA2) was used to conjugate the deletion cassete into strain 2457T.

Suicide vector pSh.DELTA.roA was integrated into the chromosome of strain 2457T to introduce the .DELTA.aroA mutation by homologous recombination, followed by tetracycline-sensitive enrichment using the procedures described for Salmonella by Hone et al, Vaccine, 810-816 (1991).

To screen for .DELTA.aroA mutants in which the aromatic amino acid biosynthesis pathway was inactivated, individual colonies, were grown overnight in Brain heart infusion (BHI) (BBL, Becton Dickinson, Cockeysville, MD) agar, and after carrying out the antibiotic-sensitive enrichment procedure, were inoculated onto replica plates containing Shigella minimum medium (SMM) consisting of 0.4 g NaCl, 8.4 g K.sub.2 HPO.sub.4, 3.6 g KH.sub.2 PO.sub.4, 0.8 g (NH.sub.4).sub.2 SO.sub.4, 2.5 g glucose, 0.05 g nicotonic acid, 0.05 g aspartic acid, 0.05 g serine and 15 g noble L-agar. SMM allows one to screen for .DELTA.aroA mutants colonies that cannot synthesize aromatic compounds de novo, and thus require exogenous aromatic compounds in order to grow. In this manner, the aroA strain, CVD1201, was obtained.

A deletion of 900 nucleotides in the virG gene (Lett et al, J. Bacteriol., 172:352-359 (1989)), which corresponds to a deletion of amino acids 341-640 of the 120 kDA VirG protein, was obtained by following steps analogous to that used for preparing the aroA mutation. More specifically, the 5' end of the virG gene was amplified with the upstream primer (GGGGAATTCCAAATTCACAAATTTTTTTGT) (SEQ ID NO:9) so as to introduce an EcoRI site, and with the downstream primer (TCCATGCCATTCATGGAGTATTAATGAATT) (SEQ ID NO:10). The 3' end of the virG gene was amplified with the upstream primer (CTCCATGAATGGCATGGAAAGGCGGAATA) (SEQ ID NO: 11), and the downstream primer (CGGGTCGACTCAGAAGGTATATTTCACACCCAA) (SEQ ID NO:12) so as to introduce a SalI site. Amplification and fusion of the virG 5' and 3' segments were performed using the same PCR cycles described above. The resulting new .DELTA.virG gene was cloned into the EcoRI and SalI sites of pKTN701, giving rise to pSh.DELTA.virG, which was propagated in strain SY327. Next, the chloramphenicol resistance gene in pSh.DELTA.virG amplified from pBR322, was replaced with the tetracycline resistance gene in plasmid pBR322 to give rise to the suicide vector pSh.DELTA.virG.sup.T, which was propagated in strain SY327. The plasmid was then electroporated into strain Sm10pir. Sm10.lambda.pir(pSh.DELTA.virG.sup.T) was used to conjugate the deletion cassete into the .DELTA.aroA strain, CVD1201.

Suicide vector pSh.DELTA.virG.sup. was then integrated into the virulence plasmid (.DELTA.virG) loci of the .DELTA.aroA strain, CVD1201, to introduce the .DELTA.virG mutation by homologous recombination, followed by chloramphenicol-sensitive enrichment using the procedures described for Salmonella by Hone et al, Vaccine, 9:810-816 (1991).

The presence of the 140 MD virulence plasmid was assessed by uptake of Congo red dye by the resulting .DELTA.aroA mutants when grown on 0.01% (w/v) Congo red dye (Sigma, St Louis, Mo.) tryptic soy agar (CR) (BBL) plates in the presence of 20 .mu.g chloramphenicol/ml of BHI agar.

In addition, to screen for .DELTA.virG mutants, individual colonies were grown overnight on BHI medium, and then plated in replicate on CR and BHI in the presence of 16 .mu.g/ml of tetracycline (Sigma).

Five-six colonies of every clone that was found to be chloramphenicol sensitive and Congo red positive, and that showed aromatic compound auxotrophy, were boiled in 100 .mu.l of sterile water for 10 min. The resulting supernatant was used as the template for PCR screening of .DELTA.aroA using the aroA primers shown above. In addition, five-six colonies of every clone that was tetracycline sensitive and Congo red positive were boiled in 100 .mu.l of sterile water for 10 min. The resulting supernatant was used as the template for PCR screening of VirG using the VirG 5' and 3' upstream primers.

In this manner the aroA VirG Shigella flexneri 2a mutant, CVD1202 (ATCC No. 55517), was isolated.

While the invention has been described in detail, and with reference to specific embodiments thereof, it will be apparent to one of ordinary skill in the art that various changes and modifications can be made therein without departing from the spirit and scope thereof.

__________________________________________________________________________SEQUENCE LISTING(1) GENERAL INFORMATION:(iii) NUMBER OF SEQUENCES: 14(2) INFORMATION FOR SEQ ID NO:1:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 2008 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: genomic DNA(iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO(vi) ORIGINAL SOURCE:(A) ORGANISM: Enteroinvasive E. coli(B) STRAIN: EI-37 (0136:NM)(xi) SEQUENCE DESCRIPTION: SEQ ID NO:1:ATCGATATATTGTTTATTGTCAGTATGGCTCAATGTGATAATAGTTGGAAAGTTTGATGG60GTTTCGCCCCGTTGTAGCGGTAGTCGACCCCGTTGT AGCGGTAGTCGAGCTGGAAGGTCT120TCAGGCACTGCTTACAGCGATAGAGCAGCCCCCCAGAACTGGAATGGCCGTTCCGATACC180CCCCTGAGTTTCAGAGTAACGGGGACAAACCACATCAATCTTTGCCATCAATCATCCAAA240GGGCAAAGAG TACAACAACACTAAGTCTGCGTCACAACCCATCAATGAAAGGAATATATA300CATATGCCATCAGTAAATTTAATCCCATCAAGGAAAATATGTTTGCAA348MetProSerValAsnLeuIleProSerArgLysIleCysLeu Gln151015AATATGATAAATAAAGACAACGTCTCTGTTGAGACAATCCAGTCTCTA396AsnMetIleAsnLysAspAsnValSerValGluThrIle GlnSerLeu202530TTGCACTCAAAACAATTGCCATATTTTTCTGACAAGAGGAGTTTTTTA444LeuHisSerLysGlnLeuProTyrPheSerAspLys ArgSerPheLeu354045TTAAATCTAAATTGCCAAGTTACCGATCACTCTGGAAGACTTATTGTC492LeuAsnLeuAsnCysGlnValThrAspHisSerGly ArgLeuIleVal505560TGTCGACATTTAGCTTCCTACTGGATAGCACAGTTTAACAAAAGTAGT540CysArgHisLeuAlaSerTyrTrpIleAlaGlnPheAsn LysSerSer657075GGTCACGTGGATTATCATCACTTTGCTTTTCCGGATGAAATTAAAAAT588GlyHisValAspTyrHisHisPheAlaPheProAspGluIleLys Asn80859095TATGTTTCAGTGAGTGAAGAAGAAAAGGCTATTAATGTGCCTGCTATT636TyrValSerValSerGluGluGluLysAlaIleAsnVal ProAlaIle100105110ATTTATTTTGTTGAAAACGGTTCATGGGGAGATATTATTTTTTATATT684IleTyrPheValGluAsnGlySerTrpGlyAspIle IlePheTyrIle115120125TTCAATGAAATGATTTTTCATTCCGAAAAAAGCAGAGCACTAGAAATA732PheAsnGluMetIlePheHisSerGluLysSerArg AlaLeuGluIle130135140AGTACATCAAATCACAATATGGCATTAGGCTTGAAGATTAAAGAAACT780SerThrSerAsnHisAsnMetAlaLeuGlyLeuLysIle LysGluThr145150155AAAAATGGGGGGGATTTTGTCATTCAGCTTTATGATCCCAACCATACA828LysAsnGlyGlyAspPheValIleGlnLeuTyrAspProAsnHis Thr160165170175GCAACTCATTTACGAGCAGAGTTTAACAAATTTAACTTAGCTAAAATA876AlaThrHisLeuArgAlaGluPheAsnLysPheAsnLeu AlaLysIle180185190AAAAAACTGACTGTAGATAATTTTCTTGATGAAAAACATCAGAAATGT924LysLysLeuThrValAspAsnPheLeuAspGluLys HisGlnLysCys195200205TATGGTCTTATATCCGACGGTATGTCTATATTTGTGGACAGACATACT972TyrGlyLeuIleSerAspGlyMetSerIlePheVal AspArgHisThr210215220CCAACAAGCATGTCCTCCATAATCAGATGGCCTAATAATTTACTTCAC1020ProThrSerMetSerSerIleIleArgTrpProAsnAsn LeuLeuHis225230235CCCAAAGTTATTTATCACGCGATGCGTATGGGATTGACTGAGCTAATC1068ProLysValIleTyrHisAlaMetArgMetGlyLeuThrGluLeu Ile240245250255CAAAAAGTAACAAGAGTCGTACAACTATCTGACCTTTCAGACAATACG1116GlnLysValThrArgValValGlnLeuSerAspLeuSer AspAsnThr260265270TTAGAATTACTTTTGGCAGCCAAAAATGACGATGGTTTGTCAGGATTG1164LeuGluLeuLeuLeuAlaAlaLysAsnAspAspGly LeuSerGlyLeu275280285CTTTTAGCTTTACAAAATGGGCATTCAGATACAATCTTAGCATACGGA1212LeuLeuAlaLeuGlnAsnGlyHisSerAspThrIle LeuAlaTyrGly290295300GAACTCCTGGAAACTTCTGGACTTAACCTTGATAAAACGGTAGAACTA1260GluLeuLeuGluThrSerGlyLeuAsnLeuAspLysThr ValGluLeu305310315CTAACTGCGGAAGGAATGGGAGGACGAATATCGGGTTTATCCCAAGCA1308LeuThrAlaGluGlyMetGlyGlyArgIleSerGlyLeuSerGln Ala320325330335CTTCAAAATGGGCATGCAGAAACTATCAAAACATACGGAAGGCTTCTC1356LeuGlnAsnGlyHisAlaGluThrIleLysThrTyrGly ArgLeuLeu340345350AAGAAGAGAGCAATAAATATCGAATACAATAAGCTGAAAAATTTGCTG1404LysLysArgAlaIleAsnIleGluTyrAsnLysLeu LysAsnLeuLeu355360365ACCGCTTATTATTATGATGAAGTACACAGACAGATACCTGGACTAATG1452ThrAlaTyrTyrTyrAspGluValHisArgGlnIle ProGlyLeuMet370375380TTTGCTCTTCAAAATGGACATGCAGATGCTATACGCGCATACGGTGAG1500PheAlaLeuGlnAsnGlyHisAlaAspAlaIleArgAla TyrGlyGlu385390395CTCATTCTTAGCCCCCCTCTCCTCAACTCAGAGGATATTGTAAATTTG1548LeuIleLeuSerProProLeuLeuAsnSerGluAspIleValAsn Leu400405410415CTGGCCTCAAGGAGATATGACAATGTTCCCGGACTTCTGTTAGCATTG1596LeuAlaSerArgArgTyrAspAsnValProGlyLeuLeu LeuAlaLeu420425430AATAATGGACAGGCTGATGCAATCTTAGCTTATGGTGATATCTTGAAT1644AsnAsnGlyGlnAlaAspAlaIleLeuAlaTyrGly AspIleLeuAsn435440445GAGGCAAAACTTAACTTGGATAAAAAAGCAGAGCTGTTAGAAGCGAAA1692GluAlaLysLeuAsnLeuAspLysLysAlaGluLeu LeuGluAlaLys450455460GATTCTAATGGTTTATCTGGATTGTTTGTAGCCTTGCATAATGGATGT1740AspSerAsnGlyLeuSerGlyLeuPheValAlaLeuHis AsnGlyCys465470475GTAGAAACAATTATTGCTTATGGGAAAATACTTCACACTGCAGACCTT1788ValGluThrIleIleAlaTyrGlyLysIleLeuHisThrAlaAsp Leu480485490495ACTCCACATCAGGCATCAAAATTACTGGCAGCAGAAGGCCCAAATGGG1836ThrProHisGlnAlaSerLysLeuLeuAlaAlaGluGly ProAsnGly500505510GTATCTGGATTAATTATAGCTTTTCAAAATAGGAATTTTGAGGCAATA1884ValSerGlyLeuIleIleAlaPheGlnAsnArgAsn PheGluAlaIle515520525AAAACTTATATGGGAATAATAAAAAATGAAAATATTACACCTGAAGAA1932LysThrTyrMetGlyIleIleLysAsnGluAsnIle ThrProGluGlu530535540ATAGCAGAACACTTGGACAAAAAAAATGGAAGTGATTTTCTAGAAATT1980IleAlaGluHisLeuAspLysLysAsnGlySerAspPhe LeuGluIle545550555ATGAAGAATATAAAAAGCTGAATATTAT2008MetLysAsnIleLysSer560565(2) INFORMATION FOR SEQ ID NO:2:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 1722 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: genomic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(vi) ORIGINAL SOURCE:(A) ORGANISM: Shigella flexneri 2a(B) STRAIN: M4243(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2:ACCCATCAATGAAAGGAATATATACATATGCCATCAGTAAATTTAATCCCA51MetProSerValAsnLeuIlePro15TCAAGGAAAATATGTTTGCAAAATATGATAAATAAAGACAACGTCTCT99SerArgLysIleCysLeuGlnAsnMetIleAsnLysAspAsnValSer101520GTTG AGACAATCCAGTCTCTATTGCACTCAAAACAATTGCCATATTTT147ValGluThrIleGlnSerLeuLeuHisSerLysGlnLeuProTyrPhe25303540TCTGACAAGAGGAGTTTTTTATTAAATCTAAATTGCCAAGTTACCGAT195SerAspLysArgSerPheLeuLeuAsnLeuAsnCysGlnValThrAsp4550 55CACTCTGGAAGACTTATTGTCTGTCGACATTTAGCTTCCTACTGGATA243HisSerGlyArgLeuIleValCysArgHisLeuAlaSerTyrTrpIle6065 70GCACAGTTTAACAAAAGTAGTGGTCACGTGGATTATCATCACTTTGCT291AlaGlnPheAsnLysSerSerGlyHisValAspTyrHisHisPheAla758085TTTCCGGATGAAATTAAAAATTATGTTTCAGTGAGTGAAGAAGAAAAG339PheProAspGluIleLysAsnTyrValSerValSerGluGluGluLys9095100GCTAT TAATGTGCCTGCTATTATTTATTTTGTTGAAAACGGTTCATGG387AlaIleAsnValProAlaIleIleTyrPheValGluAsnGlySerTrp105110115120GGAGATATTATTTTTTATATTTTCAATGAAATGATTTTTCATTCCGAA435GlyAspIleIlePheTyrIlePheAsnGluMetIlePheHisSerGlu1251301 35AAAAGCAGAGCACTAGAAATAAGTACATCAAATCACAATATGGCATTA483LysSerArgAlaLeuGluIleSerThrSerAsnHisAsnMetAlaLeu1401451 50GGCTTGAAGATTAAAGAAACTAAAAATGGGGGGGATTTTGTCATTCAG531GlyLeuLysIleLysGluThrLysAsnGlyGlyAspPheValIleGln155160165CTTTATGATCCCAACCATACAGCAACTCATTTACGAGCAGAGTTTAAC579LeuTyrAspProAsnHisThrAlaThrHisLeuArgAlaGluPheAsn170175180AAATT TAACTTAGCTAAAATAAAAAAACTGACTGTAGATAATTTTCTT627LysPheAsnLeuAlaLysIleLysLysLeuThrValAspAsnPheLeu185190195200GATGAAAAACATCAGAAATGTTATGGTCTTATATCCGACGGTATGTCT675AspGluLysHisGlnLysCysTyrGlyLeuIleSerAspGlyMetSer2052102 15ATATTTGTGGACAGACATACTCCAACAAGCATGTCCTCCATAATCAGA723IlePheValAspArgHisThrProThrSerMetSerSerIleIleArg2202252 30TGGCCTGATAATTTACTTCACCCCAAAGTTATTTATCACGCGATGCGT771TrpProAspAsnLeuLeuHisProLysValIleTyrHisAlaMetArg235240245ATGGGATTGACTGAGCTAATCCAAAAAGTAACAAGAGTCGTACAACTA819MetGlyLeuThrGluLeuIleGlnLysValThrArgValValGlnLeu250255260TCTGA CCTTTCAGACAATACGTTAGAATTACTTTTGGCAGCCAAAAAT867SerAspLeuSerAspAsnThrLeuGluLeuLeuLeuAlaAlaLysAsn265270275280GACGATGGTTTGTCAGGATTGCTTTTAGCTTTACAAAATGGGCATTCA915AspAspGlyLeuSerGlyLeuLeuLeuAlaLeuGlnAsnGlyHisSer2852902 95GATACAATCTTAGCATACGGAGAACTCTTGGAAACTTCTGGACTTAAC963AspThrIleLeuAlaTyrGlyGluLeuLeuGluThrSerGlyLeuAsn3003053 10CTTGATAAAACGGTAGAACTACTAACTGCGGAAGGAATGGGAGGACGA1011LeuAspLysThrValGluLeuLeuThrAlaGluGlyMetGlyGlyArg315320325ATATCGGGTTTATCCCAAGCACTTCAAAATGGGCATGCAGAAACTATC1059IleSerGlyLeuSerGlnAlaLeuGlnAsnGlyHisAlaGluThrIle330335340AAAAC ATACGGAAGGCTTCTCAAGAAGAGAGCAATAAATATCGAATAC1107LysThrTyrGlyArgLeuLeuLysLysArgAlaIleAsnIleGluTyr345350355360AATAAGCTGAAAAATTTGCTGACCGCTTATTATTATGATGAAGTACAC1155AsnLysLeuLysAsnLeuLeuThrAlaTyrTyrTyrAspGluValHis3653703 75AGACAGATACCCGGACTAATGTTTGCTCTTCAAAATGGACATGCAGAT1203ArgGlnIleProGlyLeuMetPheAlaLeuGlnAsnGlyHisAlaAsp3803853 90GCTATACGCGCATACGGTGAGCTCATTCTTAGCCCCCCTCTCCTCAAC1251AlaIleArgAlaTyrGlyGluLeuIleLeuSerProProLeuLeuAsn395400405TCAGAGGATATTGTAAATTTGCTGGCCTCAAGGAGATATGACAATGTT1299SerGluAspIleValAsnLeuLeuAlaSerArgArgTyrAspAsnVal410415420CCCGG ACTTCTGTTAGCATTGAATAATGGACAGGCTGATGCAATCTTA1347ProGlyLeuLeuLeuAlaLeuAsnAsnGlyGlnAlaAspAlaIleLeu425430435440GCTTATGGTGATATCTTGAATGAGGCAAAACTTAACTTGGATAAAAAA1395AlaTyrGlyAspIleLeuAsnGluAlaLysLeuAsnLeuAspLysLys4454504 55GCAGAGCTGTTAGAAGCGAAAGATTCTAATGGTTTATCTGGATTGTTT1443AlaGluLeuLeuGluAlaLysAspSerAsnGlyLeuSerGlyLeuPhe4604654 70GTAGCCTTGCATAATGGATGTGTAGAAACAATTATTGCTTATGGGAAA1491ValAlaLeuHisAsnGlyCysValGluThrIleIleAlaTyrGlyLys475480485ATACTTCACACTGCAGACCTTACTCCACATCAGGCATCAAAATTACTG1539IleLeuHisThrAlaAspLeuThrProHisGlnAlaSerLysLeuLeu490495500GCAGC AGAAGGCCCAAATGGGGTATCTGGATTAATTATAGCTTTTCAA1587AlaAlaGluGlyProAsnGlyValSerGlyLeuIleIleAlaPheGln505510515520AATAGGAATTTTGAGGCAATAAAAACTTATATGAAAATAATAAAAAAT1635AsnArgAsnPheGluAlaIleLysThrTyrMetLysIleIleLysAsn5255305 35GAAAATATTACACCTGAAGAAATAGCAGAACACTTGGACAAAAAAAAT1683GluAsnIleThrProGluGluIleAlaGluHisLeuAspLysLysAsn5405455 50GGAAGTGATTTTCTAGAAATTATGAAGAATATAAAAAGC1722GlySerAspPheLeuGluIleMetLysAsnIleLysSer555560565(2) INFORMATION FOR SEQ ID NO:3:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 4 amino acids(B) TYPE: amino acid(C) STRANDEDNESS: unknown(D) TOPOLOGY: linear(ii) MOLECULE TYPE: peptide(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(v) FRAGMENT TYPE: N-terminal fragment(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3:AlaProProVal 4(2) INFORMATION FOR SEQ ID NO:4:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 14amino acids(B) TYPE: amino acid(C) STRANDEDNESS: unknown(D) TOPOLOGY: linear(ii) MOLECULE TYPE: peptide(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(v) FRAGMENT TYPE: N-terminal fragment(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4:AlaProProValProIleAsnProAlaXaaProIleXaaArg14(2) INFORMATION FOR SEQ ID NO:5:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 29 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear (ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5:TAATCGAATTCATGGAATCCCTGACGTTA29(2) INFORMATION FOR SEQ ID NO:6:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 42 base pairs(B) TYPE: nucleic acid (C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:6:GGTACCCCCAATATTAGGGCCATCAACGTCAACGTTGCCGCC42(2) INFORMATION FOR SEQ ID NO:7:(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 42 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:7:AATATTGGGGGTACCGGTACTTATTTGGTCGAAGGCGATGCA 42(2) INFORMATION FOR SEQ ID NO:8:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 30 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:8:TGATAAGTCGACTCAGGCTGC CTGGCTAAT30(2) INFORMATION FOR SEQ ID NO:9:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 30 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO (xi) SEQUENCE DESCRIPTION: SEQ ID NO:9:GGGGAATTCCAAATTCACAAATTTTTTTGT30(2) INFORMATION FOR SEQ ID NO:10:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 30 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(i i) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10:TCCATGCCATTCATGGAGTATTAATGAATT30(2) INFORMATION FOR SEQ ID NO:11:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 29 base pairs(B) TYPE: nucleic acid (C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11:CTCCATGAATGGCATGGAAAGGCGGAATA29(2) INFORMATION FOR SEQ ID NO:12:(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 33 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:12:CGGGTCGACTCAGAAGGTATATTTCACACCCAA 33(2) INFORMATION FOR SEQ ID NO:13:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 17 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(xi) SEQUENCE DESCRIPTION: SEQ ID NO:13:CAGTGTATCACCACGAG 17(2) INFORMATION FOR SEQ ID NO:14:(i) SEQUENCE CHARACTERISTICS:(A) LENGTH: 17 base pairs(B) TYPE: nucleic acid(C) STRANDEDNESS: single(D) TOPOLOGY: linear(ii) MOLECULE TYPE: synthetic DNA(iii) HYPOTHETICAL: NO(iv) ANTI-SENSE: NO(x i) SEQUENCE DESCRIPTION: SEQ ID NO:14:AAATTATCTACAGTCAG17

Claims

1. An isolated DNA molecule encoding ShET2 which consists of the amino acid sequence encoded by the DNA of SEQ ID NO:2.

2. The isolated DNA molecule of claim 1, wherein said DNA molecule consists of the nucleotide sequence shown in SEQ ID NO:2.

3. A plasmid comprising the DNA of claim 1.

4. A plasmid comprising the DNA of claim 2.