WO1992011354A1 - VIBRIO CHOLERAE STRAINS DEFECTIVE IN irgA EXPRESSION, AND CHOLERA VACCINES DERIVED THEREFROM - Google Patents

Abstract

A Vibrio cholerae cell harboring a mutation which inhibits or prevents expression in the cell of a functional irgA gene product; a purified preparation of such mutant cells; and a cholera vaccine incorporating such mutant cells.

Description

VIBRIO CHOLERAE STRAINS DEFECTIVE IN irgA

EXPRESSION, AND CHOLERA VACCINES DERIVED THEREFROM

Background of the Invention

The field of the invention is Vibrio cholerae mutations. This invention was made with governmental support (Public Health Service grant AI27329 awarded by the National Institute of Allergy and Infectious

Diseases), and the government has certain rights to the invention.

Despite more than a century of research on cholera, this disease remains a major health problem in developing countries of Asia and Africa. It causes over 30,000 deaths per year in Bangladesh alone, and perhaps over 200,000 worldwide (Holmgren et al., 1989). One approach to the development of an effective cholera vaccine is the generation of Vibrio cholerae strains which are attenuated yet immunogenic: that is, they lack the virulent characteristics of the wild-type bacterium, while maintaining the latter organism's ability to stimulate an immune response that can protect the host from subsequent infection with wild-type V. cholerae . By the use of modern genetic engineering methods, mutant strains of V. cholerae have been created in which

specific genes (or critical portions of the genes) contributing to the virulence of the wild type organism are deleted. For example, deletion mutations in each of the two alleles encoding the toxic A subunit of cholera toxin (ctxA), but not in the genes encoding the

immunogenic B subunit, has yielded a number of mutant strains that have been tested in human volunteers

[Mekalanos et al., 1983; Mekalanos, U.S. Patent

No.4,882,278 (herein incorporated by reference); Kaper et al., 1984; Levine et al., 1988a,b; Herrington et al., 1988]. In addition, deletion mutations in two other V. cholerae genes, tcpA and toxR, have been described

(Taylor et al., 1987).

Summary of the Invention

It has been found that a strain of V. cholerae with an insertional inactivation of the single irgA locus has about a ten-fold reduced ability to colonize the intestines of test animals, compared to the colonization ability of the wild-type parental strain. In addition, this irgA mutant is about 100-fold less virulent than the wild-type parent strain, as assessed by LD₅₀ (the dosage level which is lethal to 50% of the test animals which receive it). When this mutation is combined in a single strain of V. cholerae with mutations which inactivate both of the ctxA alleles [or any other gene(s) the inactivation of which substantially diminishes virulence while maintaining immunogenicity], the resulting strain would exhibit significantly less virulence in test subjects than would a strain harboring only the irgA mutation or only the ctxA mutations. Lessened virulence means that the inoculated animal would be less likely to develop the side effects, such as diarrhea and fever, which are frequently seen with other live attenuated cholera vaccines. Mutations in genes other than irgA, such as irgB, which also result in a decrease in IrgA production, would also be useful in producing a strain of V. cholerae with substantially lowered virulence.

The invention features a V. cholerae cell (or a purified preparation of such cells) harboring a mutation which inhibits or prevents expression in the cell of a functional irgA gene product. This mutation could be in the irgA gene itself, in the irgB gene, or in any other gene which affects production of functional IrgA protein. For maximal stability of the mutation, it is preferably a deletion of a substantial amount of the affected gene (e.g., at least 25% of the coding sequence of the

affected gene, and more preferably at least 50%). This mutation may be the only such virulence-inhibiting mutation in the cell, or may be combined in the cell with a second mutation which inhibits production in the cell of a second V. cholerae virulence factor (besides IrgA), and potentially with additional mutations which have the effect of reducing the levels of still other virulence factors. V. cholerae virulence factors which have been identified so far include cholera toxin (subunits A and B), neuraminidase, hemolysin, and certain specific adhesins including toxin coregulated pilus. One such second mutation that is potentially useful in the

invention is a ctxA deletion mutant, such as in V.

cholerae strain 0395-N1 (Mekalanos, U.S. Patent No.

4,882,278). In this strain, a substantial amount of the coding sequence of each of the two ctxA alleles has been deleted, so that no functional cholera toxin A subunit is made by these cells.

As used herein, the term "functional irgA gene product" is taken to mean an outer membrane protein of approximate molecular weight 77 kDa (as determined by polyacrylamide gel electrophoresis) which functions as a V. cholerae virulence factor: that is, strains in which the 77 kDa protein is produced at a decreased level compared to the parental (or a wild-type) strain are less virulent in vivo than the parental (or wild-type) strain; such virulence can be conveniently assayed using the in vivo mouse model described in the Examples. A mutation is herein said to inhibit expression of a functional irgA gene product if a substantially lower amount (e.g., 50% or less) of the 77 kDa irgA gene product is produced by the mutant strain under low-iron culture conditions than is produced by the parental strain under identical culture conditions. (In vitro assays employing appropriate low-iron conditions are described in the Examples below.) The mutation is said to prevent

expression of a functional irgA gene product if no 77 kDa irgA gene product is detectable in a polyacrylamide gel electrophoresis analysis of outer membrane proteins produced by the mutant strain under low-iron conditions, as described in Example 1 below. A mutation is said to inhibit the production of a second virulence factor if, as a result of the mutation, the level of the second virulence factor in the mutant cell is decreased in comparison with the level in a cell of the parental strain, as measured by (a) a significant (e.g., at least 50%) decrease in virulence in the mutant strain compared to the parental strain, and (b) a significant (e.g., at least 50%) decrease in the amount of the polypeptide identified as the second virulence factor in the mutant strain compared to the parental strain. A mutation which leaves the cell incapable of producing a detectable amount of the virulence factor of interest (assayed by standard methods such as gel electrophoresis and

autoradiography under conditions such as those utilized in the Examples set forth below) is said to prevent such production. A preparation of cells is termed

"substantially purified" if contaminating V. cholerae cells without the desired mutant genotype constitute less than 1% (and preferably less than 0.1%) of the total number of cells in the preparation.

Also within the invention is a cholera vaccine made up of the mutant cells of the invention, and a method for inducing immunity to cholera in a mammal by administering to the mammal an effective amount of such cholera vaccine. The cells used for the vaccine are preferably alive and thus capable of colonizing the intestines of the innoculated animal, and may be combined if desired with vaccines targeted at other illnesses to make a single, multi-valent vaccine.

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

Detailed Description

The drawings are first described.

Drawings

Fig. 1 is an autoradiogram of an electrophoresis gel of outer membrane proteins for V. cholera 0395 wild-type cells grown in high-iron medium (lane 1); 0395 wild-type cells grown in low-iron medium (lane 2); MBG40 cells grown in high-iron medium (lane 3); and MBG40 cells grown in low-iron medium (lane 4).

Fig. 2 is a Southern blot analysis of chromosomal DNA from 10 mutant strains, digested with Eco RV and probed with a ³²P-labeled internal fragment of TnphoA, where lane 1 is from mutant strain MBG18; lane 2 is from MBG19; lane 3 is from MBG20; lane 4 is from MBG21; lane 5 is from MBG24; lane 6 is from MBG25; lane 7 is from

MBG25; lane 8 is from MBG34; lane 9 is from MBG37; and lane 10 is from MBG40; the numbers to the left of the gel indicate the sizes (in kilobases) of standards used.

Fig. 3 is an illustration of in vitro growth curves of 0395 wild-type cells and MBG40 cells (A) in LB medium with and without added dipyridyl, and (B) in T medium with and without added iron; where closed circles represent 0395 wild-type in high-iron medium, open circles represent 0395 wild-type in low-iron medium, closed triangles represent MBG40 in high-iron medium, and open triangles represent MBG40 in low-iron medium.

Fig. 4 is a diagram illustrating the construction of plasmids utilized in Example 2, where: the lightly-stippled bar represents MBG40 chromosomal DNA contained in the insert; the open bar represents DNA from TnphoA; the darkly-stippled bar (Km) represents the kanamycin resistance gene carried on TnphoA; the cross-hatched bar (Ap) represents the ampicillin resistance gene carried on pBR322; the arrow indicates the direction of

transcription of irgA; and B=BamHI, Bg=BglII, N=NheI. and Nr=NruI.

Fig. 5 is a Western blot analysis of proteins from MBG40 and 0395 prepared after growth in low- and high- iron media and probed with anti-PhoA antibody, where lane 1 represents whole-cell proteins of 0395 grown in high iron; lane 2, whole-cell proteins of MBG40 grown in high iron; lane 3, periplasmic extract of MBG40 grown in low iron; and lane 4, whole-cell proteins of MBG40 grown in low iron; the sizes of molecular weight standards (in kDa) are indicated on the left.

Fig. 6 is a restriction map of irgA and upstream chromosomal DNA in pMBG59.

Fig. 7 is a Northern blot analysis of RNA from MBG40 and 0395, probed with a ³²P-labeled fragment of irgA, where lane la is 0395 grown in high iron; lane lb is 0395 grown in low iron; lane 2a is MBG40 grown in high iron; and lane 2b is MBG40 grown in low iron; the

positions of single-stranded RNA molecular weight markers (in kb) are indicated on the left.

Fig. 8 is an illustration of the nucleotide sequence of irgA upstream of its fusion with phoA, the promoter region of irgA, and the large region of upstream DNA required for its transcription, starting at the ClaI site, with the first five nucleotides of the phoA

sequence after the fusion joint with irgA indicated by lower case letters at the end of the sequence

(corresponds to SEQ ID NO.: 1).

Fig. 9 is an illustration of nucleotide homology between the proposed iron regulatory region of irgA and the Fur box consensus sequence of E. coli [5,6], where arrows refer to areas of interrupted dyad symmetry.

Fig. 10 is a hydropathicity plot of the 151-residue amino-terminus of IrgA, where hydrophobicity is indicated by positive (+) data points and hydrophilicity by negative (-) data points, with each data point

representing the average of three consecutive residues.

Fig. 11 is an illustration of one region of homology between the peptide sequences of IrgA and FepA, with the number of residues from the amino terminus of each precursor protein to the start of each sequence indicated on the left.

Fig. 12 is a Northern blot analysis of RNA from 0395 and MBG40, probed with an oligonucleotide

complementary to the 5' terminus of irgB, where lane 1 represents 0395 grown in high iron; lane 2, 0395 grown in low iron; lane 3, MBG40 grown in high iron; and lane 4, MBG40 grown in low iron; the positions of single-stranded RNA molecular weight markers (in kilobases) are indicated on the left.

Fig. 13 is an illustration of the nucleotide sequence of the chromosomal DNA in pMBG59 (reading 5' to 3¹ from right to left in Fig. 1), starting downstream of the start site of irgA transcription and extending up to the ClaI restriction site in pMBG59, with the deduced amino acid sequence of IrgB shown in three-letter code (corresponds to SEQ ID NO.: 2).

Fig. 14 is a depiction of the overlapping, divergent promoters of irgA and irgB , with the promoter of irgB (-35, -10), start site of transcription of irgB (*), Shine-Dalgarno sequence (SD) of irgB , and open reading frame of irgB indicated on the upper strand, while the corresponding features of irgA are noted on the bottom strand. Fig. 15 is an illustration of the homology between the amino terminus of IrgB and several members of the LysR family of positive transcriptional activators, where amino acids are identified in single letter code and identical residues are enclosed in boxes; the conserved helix-turn-helix domain of these proteins is indicated above the sequences.

The Mutation

Included within the invention is any V. cholerae cell harboring a mutation which inhibits or prevents the cell from expressing a functional irgA gene product.

This includes all types of inactivating mutations of irgA or of any other locus necessary for expression of irgA. Such mutations may constitute, for example, insertions, deletions, or replacements of one or a few nucleotides, mutations that could result in frame shifts or in

relatively minor changes in critical portions of the affected polypeptide, which in the case of the irgA gene is the 77-kDa major iron-regulated outer membrane

protein. It is expected that nearly all mutations that result in a change in the amino acid sequence of this protein would decrease the protein's ability to carry out its function, and thus would be within the invention.

More preferably, however, the mutation would constitute a more substantial alteration in the affected gene which would be less likely to revert spontaneously to the wild-type phenotype. One such preferred mutation involves an insertion of a relatively long (e.g., 500 or more base pairs) segment of DNA directly into the irgA coding sequence, such as is described below in Example 1. Even more stable would be a mutation in which a substantial amount (e.g., more than 25%, and preferably at least 50%) or all of the coding sequence were deleted. If such a mutated gene expressed any polypeptide at all, it would be nonfunctional and probably degraded by cellular proteases shortly after synthesis. Such a deletion mutation can be accomplished by any of several methods known to those of ordinary skill in the art of genetic manipulations. One such method would begin with an inactivating insertion mutant prepared as described in Example 1, and would then apply the in vivo marker exchange technique described in Mekalanos, U.S. Patent No. 4,882,278, to produce the desired deletion mutation.

Mutations of a second V. cholerae gene, irgB, have also been shown to inhibit expression of irgA, and thus would be within the invention. The wild-type IrgB protein serves as a positive regulator of irgA

transcription, so that an inactivating mutation of irgB (such as the insertional mutation of irgB described in Example 3 below, or a deletion of any portion of irgB as described in Example 1) results in a nearly complete elimination of detectable expression of irgA. The marker exchange technique discussed above could be used to produce a deletion mutation in an irgB gene into which had been introduced an inactivating insertion in

accordance with Example 3.

Use

The V. cholerae cells of the invention are useful as sources of immunological protection against cholera, in particular as the basis of a live-cell vaccine capable of colonizing the inoculated animal's intestine and provoking a strong immune reaction. Appropriate dosages and conditions of administration of such a live,

attenuated vaccine are as described in Holem et al.,

Acute Enteric Infections in Children. New Prospects for Treatment and Prevention (1981) Elsevier/North-Holland Biomedical Press, Ch. 26, pp. 443 et seq. (Levine et al.). Experimental data

Examples 1-3 below describe experimental results in support of the claims.

Example 1: Identification of i rgA gene in V. cholerae , and construction of insertion mutation of irgA with decreased virulence in vivo

MATERIALS AND METHODS

Bacterial strains. V. cholerae 0395 Sm' was a gift of John

J. Mekalanos, and E. coli CC118 has been described previously (6). Mutant strain MBG 40 was derived from 0395 by

the methods described below. Strains were maintained at

-70 °C in LB medium (19) containing 15% glycerol.

Media. Two types of media were used to assess the effect

of iron concentration on gene expression: (i) LB medium

with or without the addition of the iron chelator 2,2-dipyridyl

(Sigma Chemical Co., St. Louis. Mo.) to a final concentration of 0.2 mM and (ii) Tris-buffered medium (T medium) (30)

prepared with highly purified water (Barnstead Nanopure

water purification system [Sybron, Boston, Mass.]) and

supplemented with 4 g of sucrose per liter, with or without

the addition of 36 μM FeSO₄. The concentrations of iron in

growth medium were verified by using 1,10-phenaπthrotine

(Aldrich Chemical Co., Inc., Milwaukee, Wis.) (18). By this

assay, T medium contained less than 0.5 μM iron and LB

medium contained 10 μM iron.

LB agar was used for high-iron plates. Chelex-LB agar

supplemented with 0.3% glucose was used for low-iron

plates. Chelex-LB agar was prepared by stirring the iron- chelating resin Chelex-100 (Bio-Rad Laboratories, Richmond, Calif.) in fivefold concentrated LB medium overnight

at 4ºC, filtering through Whatman filter paper (no. 1) to

remove the resin, and then adding highly purified water and

Noble agar (Difco Laboratories, Detroit, Mich.) before

sterilization. Streptomycin (100 μg/ml), kanamycin (45 μg/

ml), gentamkin (30 μg/ml), and 5-bromo- 4-chloro-3-indoIyI

phosphate (XP) (Amresco; 40 μg/ml) were added where

appropriate. XP is a chromogenic substrate for alkaline

phosphatase.

Genetic methods. The transposon vector TnphoA was used to obtain random insertions into the chromosome of V. cholerac 0395 by methods previously described (32. 33). Fusion-containing colonies were screened for the PhoA⁺ phenotype under low-iron growth conditions by the presence or absence of blue color on Chelex-LB agar containing streptomycin, kanamycin, gentacnicin, glucose, and XP. They were subsequently screened for iron regulation of alkaline phosphatase activity by streaking on Chelex-LB agar with the same supplements and by adding a filter paper disk spotted with 10 μl of 10 mM FeSO₄ in the streak. Strains that showed a distinct zone of white color surrounding the disk, with blue color peripherally, contained fusions in genes whose expression was negatively regulated by iron.

Confirmation of single TaphoA insertions into the chromosome of 0395 was examined by Southern hybridization of digests of chromosomal DNA by using the restriction enzymes XbaI and EcoRV, which do not cut within the TnphoA insert, and probing with an internal fragment of TcphoA that had been radioactively labeled by random primer extension with a commercially available kit (Prime Time [International Biotechnologies, Inc.]).

Assays. The enzymatic activity of alkaline phosphatase encoded on TnphoA permitted the comparison of fusion gene expression when strains were grown in low- versus high-iron media. Strains were grown overnight in LB medium with or without added dipyridyl and in T medium with or without added FeSO₄. Alkaline phosphatase activity was calculated as described previously (21) from measurement of hydrolysis of p-nitrophenyl phosphate (Amresco) by permeabilized cells. Activity was expressed in units per A₄₀₀ or bacterial cells, with 1 U of activity defined as by O'Callaghan et al. (23).

The amount of cholera toxin produced was assayed by using the GM-1 -dependent enzyme-linked immunosorbem assay as previously described by Holmgren (15) with slight modifications. Assays were performed on serial dilutions of supernatant fluids from cells grown to saturation in low- and high-iron media al 30ºC, using 96-well GM-1 ganglioside- coated microdilution plates.

Preparation and analysis of outer membrane proteins. Enriched outer membrane proteins were prepared by procedures previously described (14) from cells grown to late logarithmic phase in LB medium with and without added dipyridyl. Outer membrane proteins were separated on sodium dodecyl sulfate-10% polyacrylamide gels and stained with Coomassie blue.

Assessment of virulence by competition and LO₅₄ assays. 0395 wild type and mutant strain MBG40 both demonstrated equal plating efficiencies on LB agar and Chelex-LB agar (data not shown). Competition assays between strains were performed essentially as described by Freter et al. (12) and modified by Taylor et al. (33). In vitro competition was determined by growth at 37ºC for 24 h in LB medium with dipyridyl, from a starting density of 4 × 10⁵ CFU/ml. In vivo competition was determined by intraintestinal growth in 6- to 7-day-old suckling CD-1 mice (Charles River Breeding Labs, Inc., Wilmington, Mass.) inoculated orally with 4 × 10² CFU. The mice were sacrificed 24 h later. The input ratio was approximately 1.0 in both types of competition experiments; competitive indices were corrected for the input ratio. Viable cell counts and the ratio of the two strains were determined by plating dilutions onto Chelex-LB agar containing streptomycin, glucose, and XP and scoring for the PhoA ⁺ blue colony phenotype of the mutant TnphoA fusion strain and the PhoA- while colony phenotype of the wild type. The in vivo competitive index was the average of values from five individual mouse experiments.

The 50% lethal dose (LD₅₀) assays were performed by oral inoculation of 3- to 5-day-old suckling CD-1 mice with various doses of viable bacteria grown in LB medium at 30ºC, pelleted, washed twice in LB medium, and suspended in 0.15 M NaHCO₃ (pH 8.15). Four or more mice were used per dose of bacteria. Survival was determined at 36 h, and results were analyzed as described previously (26).

RESULTS

TaphoA is a derivative of the transposon Tn5 that contains a portion of phoA. the E. coli gene for alkaline phosphatase

(20). After random insertion of TaphoA into chromosomal DNA, those insertions that yield in-frame fusions between a target gene and phoA encode hybrid proteins that have a carboxy-terminal fragment of PhoA fused to an amino-terminal portion of the target protein product. These hybrid proteins display alkaline phosphatase activity only if the target gene encodes a protein expressed at the cell surface (a secreted, transmembrane, or outer membrane protein) and so provides the requisite signals for transport of the carboxy-terminal PhoA fragment into the periplasmic space (20). Because most bacterial virulence determinants are expressed at the cell surface, this technique selects for TnphoA insertions into such genes (11, 17, 25, 32, 33).

Isolation and characterization of TnphoA fusions in iron-regulated genes of V. cholerae. (i) Construction and isotation of TnphoA insertion mutant strains. After random insertion of TnphoA into the chromosome of V. cholerae 0395, colonies were screened for the PhoA⁺ phenotype by blue color on a low-iron agar plate containing XP. Of 300 individual blue colonies examined, we obtained 25 TnphoA insertion mutant Strains whose blue colony phenotype was repressed around an iron-containing disk.

(ii) Alkaline phosphatase assays, iron regulation of these TnphoA gene fusions was confirmed by measuring alkaline phosphatase activity after growth in LB medium with or without added dipyridyl (Table 1). All strains showed a significant increase in alkaline phosphatase activity in low-iron conditions compared with high-iron conditions, with induction ratios ranging from 6-fold (MBG38) to more than 850-fold (MBG40). The differing alkaline phosphatase activities and induction ratios of the 25 mutant strains suggest that we had isolated TnphoA insertions into a number of different genes on the V. cholerae chromosome. To confirm further the iron regulation of the TnphoA gene fusion in MBG 40, alkaline phosphatase activity of both this strain and wild- type 0395 was also determined after growth in T medium with and without added iron: alkaline phosphatase activity of

MBG40 increased from 1 U after growth in iron-supplemented T medium to 126 U after growth in T medium without added iron. As expected, wild-type 0395 had low levels of alkaline phosphatase activity in both low- and high-iron growth conditions.

(iii) Outer membrane proteins. Since most bacterial virulence determinants are expressed at the cell surface, we analyzed outer membrane proteins for wild-type 0395 and each of the 25 mυtant strains after growth in low- and high-iron media. In wild-type 0395. at least four new proteins of apparent molecular mass from 75 to 81 kDa appeared after growth in low-iron conditions compared with high-iron conditions (Fig. 1). The most prominent of these iron-regulated proteins, with an apparent molecular mass of 77 kDa. was lost in two of the 25 mutant strains MBG 15 (data not shown) and MBG40 (Fig. 1). Several other mutant strains had lost other individual iron-regulated outer membrane protein bands (data not shown).

(iv) Mapping of TnphoA insertions to single chromosomal fragments. Because XbaI and EcoR V do not cut within the TnphoA vector. Southern hybridization of XbaI- or EcoRV- digested chromosomal DNA. probed with an internal fragment of TnphoA, was performed for each of the 25 mutant strains. This analysis verified that only single TnphoA chromosomal insertions are present in each of the mutant strains (data not shown). Moreover, single TnphoA insertions were present in at least eight distinct XbaI chromosomal fragments, ranging in size from 9.0 to 30.0 kilobases of DNA (including the inserted TnphoA [7.6 kilobases]); there was no hybridization signal for chromosomal DNA of parent 0395 (data not shown). Figure 2 shows a Southern blot of EcoR V-digested chromosomal DNA from 10 of the mutant strains, including MBG40.

We selected strain MBG40 for further study for the following two reasons, (i) The expression of the gene fusion in this strain, as measured by alkaline phosphatase activity, was highly regulated by iron, with an induction ratio of more than 850 after growth in low-iron conditions compared with high-iron medium (Table 1). (ii) The TnphoA insertion in strain MBG40 was associated with loss of the major 77-kDa iron-regulated outer membrane protein of V. cholerae 0395 (Fig. 1). The iron-regulated gene in strain MBG40 that contains the TnphoA insertion was designated irgA.

In vitro characterization of iron-regulated (irgA) fusion strain MBG40. (i) In vitro growth curves. To determine whether MBG40 was defective for growth in vitro, growth curves of 0395 wild type and MBG40 in LB medium with and without added dipyridyl and in T medium with and without added iron were determined. The two strains grew similarly in each in vitro growth condition (Fig. 3). The enhancement of growth for both strains after the addition of iron to T medium demonstrates that growth in unsupplcmented T medium was indeed limited by iron.

(ii) In vitro competition assays. To determine whether MBG 40 was able to compete effectively in vitro with 0395 wild type, in vitro competition of the two strains was performed in LB medium with added dipyridyl. Low-iron media was chosen because previous data has suggested that the intraintestinal environment has limited iron availability (27). The in vitro competitive index of 0.98 (Table 2) con- firmed the results of the in vitro growth curves, suggesting the absence of an in vitro growth defect in the mutant strain.

(iii) Cholera toxin assays. To control for any differences in cholera toxin production that might affect the virulence of parental and mutant strains under different growth conditions, cholera toxin activity was assayed from the supernatant fluids of 0395 and MBG40 grown to saturation in T medium with and without added iron and in LB medium with and without added dipyridyl. Cholera toxin activity of the two strains was similar under the various growth conditions

(data not shown).

In vivo characterization of iron-regulated (irgA) fusion strain MBG40. (i) LD₅₀ assays. To determine whether irgA has a role in virulence of 0395, LD₅₀ assays of 0395 wild type and MBG40 were performed. The LD₅₀ of mutant strain

MBG40 was 3 × 10⁵ bacteria. compared with 4 × 10³ bacteria for parental strain 0395 (Table 2). This increase in LD₅₀ of 2 orders of magnitude clearly suggests that irgA is important for virulence in this animal model. Previous data have shown that 0395 strains carrying either a randomly selected chromosomal TnphoA insertion or TnphoA insertion in the structural gene for the outer membrane protein OmpV maintain full virulence for suckling mice (33), demonstrating that chromosomal TnphoA insertion by itself has no effect on virulence.

(ii) In vivo competition assays. Mutant strain MBG40 was tested for a colonization defect by an in vivo competition assay consisting of coinfection of the mutant with parental strain 0395 in suckling mice. The wild-type strain outcompeted the mutant almost 10-fold (in vivo competitive index,

0.11 (Table 2)). suggesting a probable colonization defect in the mutant strain.

LITERATURE CITED IN EXAMPLE 1

1. Actis, L. A., S. A. Potter, and J. H. Crosa. 1985. Iron-regulated outer membrane protein OM2 of Vibrio anguillarum is encoded by virulence plasmid pJMI. J. Bacteriol. 161:736-742.

2. Bagg, A., and J. B. Neilands. 1985. Mapping of a mutation affecting regulation of iron uptake systems in Eschericlua coli K-12. J. Bacteriol. 161:450-453.

3. Betley, M. J., V. L . Miller, and J. J. Mekalanos. 1986. Genetics of bacterial cnterotoxins. Annu. Rev. Microbiol. 40:377-605.

4. Bjorn, M. J., B. H. Iglewski, S. K. Ives, J. C. Sadoff, and M. L. Vasil. 1978. Effect of iron on yields of exotoxin A in cultures of Pseudomonas acruginosa PA-103. Infect, lmmun. 19:785-791.

5. Bjoro, M. J., P. A. Sokol, and B. H. Iglewski. 1979. Influence of iron on yields of extracellular products in Pseudomonas aeruginosa cultures. J. Bacteriol. 138:193-200.

6. Calderwood, S. B., and J. J. Mekalanos. 1987. Iron regulation of Shiga -like toxin expression in Esehtrichiu coli is mediated by the fur locus. J. Bacteriol. 169:4759-4764.

7. Calderwood, S. B., and J. J. Mekalanos. 1988. Confirmation of the Fur operator site by insertion of a synthetic oligonucleotide into an operon fusion plasmid. J. Bacteriol. 170:1015-1017.

8. Carnid, E., D. Mazigh, and H. H. Mollaret. 1987. Expression of iron-regulated proteins in Yersinia species and their relation to virulence. Infect. Immun. 55:277-280.

9. DeGrandis, S., J. Ginsberg, M. Toone. S.Climie, J. Friesen, and J. Brunton. 1987. Nucleotide sequence and promoter mapping of the Escherichia coli Shiga-like toxin operon of bacteriophage H-19B. J. Bacteriol. 169:4313-4319.

10. Dubos, R. J., and J. W. Geiger. 1946. Preparation and properties of Shiga toxin and toxoid. J. Exp. Med. 84:143-156.

11. Finlay, B. B., M. N. Starnbach, C. L. Francis, B. A. D. Stocker, S. Chatfield, G. Dougan, and S. Falkow. 1988. Identification and characterization of TnphoA mutants of Salmonella that are unable to pass through a polarized MDCK epithelial cell mono- layer. Mol. Microbiol. 2:757-766.

12. Freter, R., P. C. M. O'Brien, and M. S. Macsai. 1981. Role of chemotaxis in the association of motile bacteria with intestinal mucosa: in vivo studies, infect. Immun. 34:234-240.

13. Hantke, K. 1981. Regulation of ferric iron transport in Eschcrichia coli K12: isolation of a constitutive mutant. Mol. Gen. Genet. 182:288-292.

14. Hantke, K. 1984. Cloning of the represser protein of iron- regulated systems in Escherichia coli K12. Mol. Gen. Genet. 197:337-341.

15. Holmgren, J. 1973. Comparison of the tissue receptors for Vibrio cholerae and Escherichia coli enteroioxins by means of gangliosides and natural cholera toxoid. Infect. Immun. 8: 851-859.

16. Jonson, G., A. Svennerholm, and J. Holmgren. 1989. Vibrio cholerae expresses cell surface antigens during intestinal infection which are not expressed during in vitro culture. Infect. Immun.57:1809-1815.

17. Knapp. S., and J. J. Mekalanos. 1988. Two trans -acting regulatory genes (vir and madi control antigenic modulation in Bordetella pertussis. J. Bactenol. 170:5059-5066.

18. Krishna Murti, G. S. R., A. V. Moharir, and V. A. K. Sarma.

1970 Spectrophotometric determination of iron with or- thophenanthroline. Microchem. J. 15:585-589.

19. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular claning: a laboratory manual. Cold Spring Harbor Laboratory. C d Spring Harbor. N.Y.

20. Manoil, C., and J. Beckwith. 1985. TnphoA: a transposon probe for protein export signals. Proc. Natl. Acad. Sci. USA 82: 8129-8133.

21. Michaelis, S., H. Inouye, D. Oliver, and J. Beckwith. 1983. Mutations that alter the signal sequence of alkaline phosphatase in Escherichia coli. J. Bacteriol. 154:366-374.

22. Neilands, J. B. 1982. Microbial envelope proteins related to iron. Annu. Rev. Microbiol. 36:285-309.

23. O'Callaghan, C. H., A. Morris, S. M. Kirby, and A. H. Shingler.

1972. Novel method for detection of β-lactamases by using a chromogenic ccphalosporin substrate. Antimicrob. Agents Che- mother. 1:283-288.

24. Pappenheimer, A. M., Jr. 1955. The pathogenesis of diphtheria. Symp. Soc. Cen. Microbiol. 5:40-56.

25. Peterson, K. M., and J. J. Mekalanos. 1988. Characterization of the Vibrio cholerae ToxR regulon: identification of novel genes involved in intestinal colonization. Infect. Immun.56:2822-2829.

26. Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty percent endpoints. Am. J. Hyg. 27:493-497.

27. Sciortino, C. V., and R. A. Finkelstein. 1983. Vibrio cholerae expresses iron-regulated outer membrane proteins in vivo.

Infect. Immun. 42:990-996

28. Sigel, S. P., and S. M. Payne. 1982. Effect of iron limitation on growth. siderophore production, and expression of outer membrane proteins of Vibrio cholerae. J. Bacteriol. 150:148-155. 29. Sigel, S. P., J. A. Stoebner, and S. M. Payne. 1985. Iron- vibriobactin transport system is not required for virulence of Vibrio cholerae. Infect. Immun. 47:360-362.

30. Simon, E. H., and I. Tessman. 1963. Thymidine -requiring mutants of phage T4. Proc. Natl. Acad. Sci. USA 50:526-532. 31. Stoebner, J. A., and S. M. Payne. 1988. Iron-regulated hemolysin production and utilization of heme and hemoglobin by Vibrio cholerae. Infect. Immun. 56:2891-2895.

32. Taylor, R. K., C. Manoil, and J. J. Mekalanos. 1989. Broad- host-range vectors for delivery of TnphoA: use in genetic analysis of secreted virulence determinants of Vibrio cholerae.

J. Bacteriol. 171:1870-1878.

33. Taylor, R. K., V. L. Miller, D. B. Furlong, and J. J. Mekalanos.

1987. Use of phoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin. Proc. Natl. Acad. Sci. USA 84:2833-2837. Example 2; Transcriptional regulation of ircrA, and construction of deletion mutants of gene (irσB) encoding the irgA transcriptional regulator protein.

Bacterial strains. The V. cholerae wild-type strain used in this study was classical Ogawa strain 0395 Sm^r [22]. V. cholerae strain MB640 is 0395 irgA: :TnphoA [10]. E. coli strain CC118 is a phoA deletion derivative of MC1000 [21], and E. coli strain SM10 has been

described previously [35]. E. coli strain DH5α was obtained from Bethesda Research Laboratories Life

Technologies, Inc. (Gaithersburg, Maryland).

Media. Two types of liquid media were used to assess the effect of iron concentration on gene

expression as previously described [10]: (i) LB medium with or without the addition of the iron chelator 2,2- dipyridyl (final concentration 0.2 mM) and (ii) trisbuffered medium (T medium) with or without the addition of 36 μM FeSO₄. For E. coli strains, T medium was supplemented with thiamine (10 μg/ml) and the L-amino acids arginine and leucine (40 μg/ml). For citrate utilization assays, sodium citrate was added to T media at a final concentration of 10mM.

Solid media containing a normal concentration of iron were made using LB agar. Low-iron LB plates

supplemented with 0.3% glucose were made after overnight treatment with Chelex-100 (Bio-Rad Laboratories,

Richmond, CA), as described previously [10]. Ampicillin (100 μg/ml), tetracycline (15 μg/ml), kanamycin (45 μg/mg), streptomycin (100 μg/ml), and 5-bromo-4-chloro-3-indolyl phosphate (XP) (Amresco; 40 μg/ml) were added to the media as appropriate. XP is a chromogenic

substrate that allows blue-white colony screening on agar plates for alkaline phosphatase activity. Construction of niasmids. pSBC34 is a derivative of the broad host range plasmid pRK290 [8] modified to contain a gene fusion between the iron-regulated Shiga-like toxin I A gene (slt-IA) from E. coli and

phoA, the E. coli gene for alkaline phosphatase. To construct pSBC34. the 5.5 kilobase-pair (kbp) BamHI fragment of pSC105 (Δ2) [5].

encompassing the slt-I promoter, slt-IA fused to phoA. and the

downstream kanamycin resistance marker, was purified by gel

electroeiution. This fragment was ligated into the unique BglII restriction site of pRK290, with selection for tetracycline resistance (encoded by pRK290) and kanamycin resistance. pRK290 and its

derivatives are not self-transmissible, but can be mobilized at high frequency if supplied with mobilization functions on RP4 in trans.

pSBC34 was therefore transformed into SM10. a derivative of E. coli strain C600 that contains a chromosomally-integrated RP4-2 (Tc::Mu). SM10(pSBC34) was conjugated with V. cholerae 0395, with double

selection for tetracycline resistance (encoded by pSBC3.) and

streptomycin resistance (encoded by 0395).

Strain MBC40 contains a chromosomal gene fusion between irgA and phoA. constructed by TnphoA mutagenesis [10]. This gene fusion is contained within a 10 kbp BamHI restriction fragment, extending from a BamHI site in the chromosome approximately 5 kbp upstream of the fusion joint to the unique BamHI site within TnphoA , approximately 5 kbp downstream of the fusion joint. The BamHI site within TnphoA is located downstream of the end of phoA and the kanamycin resistance marker of the transposon.

This gene fusion was cloned into pUC19 as follows. Chromosomal DNA from MBC40 was digested with BamHI and ligated into the BamHI site on pUC18. The Ligation mix was electroporated into DHSa according to methods described below. Colonies containing the correct clone

(pMBG47) were isolated by selecting for ampicillin resistance (carried on pUC19) and kanamycin resistance (carried on TnphoA). pMBC47 was then digested with BamHI and the 10 kbp fragment separated from the parent plasmid by gel electroelution and ligated into the BamHI site on pBR322. creating plasmid p.MBG53 (Fig.4). Presence of the correct insert in pMBC53 was confirmed by sequencing the junction betveen pBR322 and the TnphoA end of the fragment. Subsequent subcloning of pMBG53 to make pMBG55 and pMBC59 was performed as shown in Figure 4 pMBG59 contains an insert of approximately 1.8 kbp of chromosomal DNA fused to 2.9 kbp of TnphoA. All plasmids were propagated in E. coli strain CC118, which contains a chromosomal deletion of alkaline phosphatase.

Plasmid subclones pHBG57 and pMBG58 were derived from pHBG55 by deletion of portions of the chromosomal insert from each of two BglII restriction sites within the insert to the BamHI restriction site at the junction of the insert with pBR322. Plasmid subclones pMBG102. pHBG10S. pMBC109. and pMBC110 were similarly derived from pMBG59 by deletion of portions of the chromosomal insert from AceI, Nco,. ClaI, and HincII restriction sites respectively within the insert to the NruI site in pBR322. Plasmid subclone pHBG103 was derived from pHBG59 by deletion from the BalI site within the chromosomal insert to the BalI site in pBR322.

Genetic methods. Except as noted, plasmid constructs were transformed into E. coli strains and conjugated into V. cholerae 0395 by mobilization from the E. coli strain SM10 by standard techniques Electroporation used in the transformation of pMBGf7 into CC118was performed in a Gene Pulser (BioRad) according to the manufacturer's protocol. Electroporation conditions were 2.300 V at 25-uF

capacitance, producing a time constant of 4.8 ms.

Assays. The enzymatic activity of alkaline phosphatase encoded on

TnphoA permitted screening of plasmid subclones containing the

irzA'-'phoA gene fusion for the presence or absence of fusion gene expression under low-iron growth conditions by streaking E. coli strain CC118 containing subclones cnto Chelex-LB agar containing XP. Colonies having a PhoA+ phenotype are blue on XP. whereas colonies having a PhoA- phenotype are white.

Similarly, the enzymatic activity of alkaline phosphatase encoded on TnphoA permitted the comparison of fusion gene expression in liquid media when V. cholerae strain MBG40 or E. coli strain CC118 carrying plasmid subclones of the gene fusion were grown in low- versus high-iron conditions. Strains were grown overnight in T medium with or without added FeSO₄. Alkaline phosphatase activity was determined as described previously [23] from measurement of hydrolysis of

p-nitrophenyl phosphate (Amresco) by permeabilized cells. Activity was normalized to the A₆₀₀ of the bacterial cells and defined as

described by O'Callaghan et al. [27].

DNA and RNA analysis. DNA and RNA analysis, including preparation of DNA and RNA. restriction mapping, and Northern blot analysis, were performed according co standard molecular biological techniques [31]. For Northern bloc analysis, an equivalent amount of RNA. as calculated from OD₂₆₀, was loaded into each lane. DNA sequencing was performed using Sequenase DNA Sequencing Kit (United States Biothemical Corporation. Cleveland. Ohio). Primer extension and primer extension sequencing were performed essentially as described by Miller et al.

[24]. except that the oligonucleotide primers were hybridized to RNA in 0.4 a NaCl and 40 mH Pipes (pH 6.4). without formamide. at 60ºC for two hours. RNasin and AMV reverse transcriptase were obtained from

Bethesda Research Laboratories Life Technologies. Inc. (Gaithersburg, Maryland).

Synthetic oligonucleotides for use both as probes for Northern blot analysis and as primers for DNA sequencing and primer extension were the generous gift of Brian Seed.

Protein analysis, whole cell and periplasmic proteins were prepared following growth in low- and high-iron media as described previously [16]. Proteins were separated on a sodium dodecyl sulfate-12.5% polyacrylamide gel, transferred to a NitroScreen West membrane (Dupont, Boston, Mass.) using a semidry blotting apparatus (Hoefer Scientific Instruments, San Francisco. Ca.). Immunoreactive proteins were visualized by sequential incubation with polyclonal rabbit anti-PhoA antibody (a generous gift of John J. Mekalanos) and goat anti-rabbit immunoglobulin-conjugated alkaline phosphatase (Sigma Chemical Co., St. Louis. Missouri), followed by staining for alkaline phosphatase activity as described previously [25].

Protein database searches. Database searches and protein alignments were done using IBI-Pustell Sequence Analysis software (International Biotechnologies. Inc., New Haven. Conn.) to search the NBRF Protein Database (Release 19) using the FASTP algorithm for protein homology (19). The hydropathicity index profile of IrgA was calculated by the technique of Kyte-Doolittle [18] using IBI -Pustell Sequence Analysis software. RESULTS Reciprocity of iron regulation between V . cholerae and E. coli. In order to determine whether or not the mechanism of iron regulation in V. cholerae is similar to the mechanism of iron regulation in E. coli. we compared iron regulation of the V. cholerae gene irgA in an E. coli background to iron regulation of the E. coli gene slt-IA in a V.

cholerae background, using alkaline phosphatase activities of the respective gene fusions following growth in low- and high-iron media (Table 3). These assays demonstrated reciprocal iron regulation of each gene fusion in the heterologous background, suggesting chat V.

cholerae and E. coli share a common mechanism of iron regulation. We attempted to transform the irgA'-'phoA gene fusion on pMBG53 into an E. coli fur° background, but were unable to establish a stable

transformant, perhaps due to toxicity of the overexpressed fusion protein in the fur° background.

Mapping of irgA by analysis of protein products and RNA

transcripts.

(i) Western blot analysis of IrgA'-'PhoA fusion protein. In order to determine the size of the IrgA'-'FhoA fusion protein, we performed Western blot analysis of whole ceil and periplasmic proteins from MBC40 and 0395 following growth in low-and high-iron media, using an anti-PhoA antibody probe (Fig.5). As expected, strain 0395 produced no immunoreactive protein and production of IrgA'-'PhoA by MBG40 was iron-regulated. The IrgA'-'PhoA fusion protein had an apparent moiecular mass of 70 kDa in whole cell extracts (Fig. 5. Lane 4), with several smaller immunoreactive species probably representing fusion protein breakdown produces. The fusion protein was transported to the periplasmic space (Fig. 5. Lane 3). but we were unable to resolve with certainty whether the largest band in periplasmic extracts was the same size as in whole cell extracts or slightly smaller, as would occur with proteolytic processing by signal peptidase. Since the PhoA portion of the fusion protein has a predicted molecular mass of approximately 50 kDa. the upstream IrgA portion of the fusion procein would have a predicted molecular mass of approximately 20 kDa. Therefore, plasmid pMBG59. which contains approximately 1.8 kbp of DNA upstream of the fusion joint, should contain the 5' terminus of irgA, plus more than 1.2 kbp of additional upstream DNA.

(ii) Verification that pHBC59 contains those sequences necessary for the expression and iron regulation of ireA. In order to verify that pMBG59 contained the elements essential to expression and iron regulation of irgA, we determined alkaline phosphatase activities in low- and high-iron media of V. cholerae strain MBG40 and E. coli strain CC118 containing plasmid subclones pMBG47. pMBG53, pMBG55. and pMBC59. Alkaline phosphatase activity of MBG40 increased from 1 U after growth in iron-supplemented T media to 126 ϋ after growth in T media without added iron, while activity of CC118 containing each of the plasmid subclones increased from approximately 70 U after growth in

iron-supplemented T media to approximately 200 U after growth in T media without iron. Therefore, expression and iron regulation of the gene fusion on each of these plasmid subclones is similar, suggesting that they each contain the DNA sequences necessary for the expression and iron regulation of irgA. The difference in induction ratio of iron regulation between CC118 containing the plasmid subclones and MBG40 is probably due to a plasmid copy number effect and less efficient

regulation of a V. cholerae gene in an E. coli background.

(iii) Deletion subcloninc and mapping of fusion gene expression and iron regulation. In order to determine the extent of chromosomal DNA upstream of the fusion joint required for fusion gene expression and iron regulation, alkaline phosphatase assays were performed on CC118 containing plasmid deletion subclones pMBG109. pMBC110. pMBG105. pMBC102. pMBG103. pMBC58, and pMBG57 after growth in low- and high-iron T media (Fig. 6). Alkaline phosphatase activity of CC118 containing pMBG109 was comparable to that of pMBC59, but alkaline phosphatase activity of all subclones having less than 1.5 kbp of chromosomal DNA upstream of the fusion joint was minimal. This amount of upstream chromosomal DNA is substantially more than that predicted to encode IrgA based on Uestern blot analysis.

(lv) Northern blot analysis of the irgA transcript. Northern blot analysis of the irgA transcript was performed to determine the size of the irgA transcript and to distinguish whether or not iron regulation of irgA occurs at the transcriptional level. RNA was prepared from strains MBG40 and 0395 following growth in low- and high-iron media. The blot was probed with the restriction fragment located between the

HindIII and SmaI sites on the insert of pMBC59 (Fig.6). A single band is seen at 2.2 kilobases (kb) in RNA from 0395 grown under low-iron conditions (Fig. 7, Lane 1b), and two less intense bands are seen at 2.5 and 2.1 kb in RNA from MBC40 grown under low-iron conditions (Lane 2b). No bands are seen in either strain grown under high-iron conditions (Lanes 1a and 2a). demonstrating that iron regulation occurs ac the transcriptional level. After DNA sequencing, these results were confirmed by probing with a synthetic oligonucleotide shown to be entirely internal to irgA (data not shown).

To investigate the basis of the two bands seen in MBG40. we synthesized oligonucleotides complementary to phoA and to an open reading frame immediately downstream of phoA, as reported by Chang et al. [7]. Probing additional Northern blots with these two

oligonucleotides demonstrated that the two bands seen in MBG40 are not a result of two transcriptional start sites for irgA, but a result of two downstream transcription termination sites within the TnphoA portion of the insert, one after phoA and the second located 400 bp further downstream, after the next open reading frame beyond phoA (data not shown). The 2.1 kb band from MBG40 is the transcript formed when termination occurs at a site located approximately 1.4 kb downstream from the fusion joint, and the 2.5 kb band is the transcript formed when termination occurs ac a site approximately 1.8 kbp downstream from the fusion joint.

Comparison of the size of the IrgA'-'PhoA fusion protein by Western bloc analysis (70 kDa) with the size of the smaller transcript seen in MBC40 by Northern blot analysis (2.1 kb) suggests that the RNA transcript for irgA is monocistronic and has a transcription start site located near the 5' BglII restriction site in the DNA upstream of the fusion joint (Fig.6). Localization of the transcription start site to this region was confirmed by probing Northern blots of MBG40 RNA with synthetic oligonucleotides complementary to DNA sequence located betveen the HindIII and upstream BglII restriction sites (irgA

transcript not seen) and co DNA sequence located approximately 220 bp downstream of this BglII site (irgA transcript seen: data not shown). These data, coupled with the alkaline phosphatase assays of the plasmid deletion subclones (Fig.6). suggest that approximately 900 bp of DNA upstream of the predicted transcription start site is required for production of an active IrgA' -'PhoA fusion protein.

(v) Northern blot analysis of irgA transcription from plasmid deletion subclones. In order to determine whether the 900 bp region of DNA upstream of the predicted transcription start site is required for transcription of irgA or for some step in post-transcriptional

processing or activation of the fusion protein, we analysed RNA prepared from CC118 containing the plasmid deletion subclones in parallel with MBG40 RNA, following growth in low-iron media. The Northern blot was probed with the HindIII-SmaI restriction fragment used previously. Identical double bands, representing the two irgA transcripts, were seen in lanes containing RNA from MBG40 and RNA from CC118 containing pMBG53 , pMBG59. and pMBC109, but were absent in all other lanes (Fig.6). Thus, plasmid deletion subclones having full iron-regulated alkaline phosphatase activity also had the irgA

transcripts, and those lacking alkaline phosphatase activity lacked the irgA transcripts. The presence of approximately equal amounts of RNA In the lanes of this Northern blot was confirmed for the plasmid deletion subclones by reprobing the membrane with a restriction fragment internal to the ampicillin resistance gene from pBR322 (data not shown). These data suggest that the 900 bp of DNA upstream of the predicted transcription start site are necessary for transcription of irzA, rather than for a post-transcriptional event. The face that the irgA transcripts seen in E. coli CC118 containing the active plasmid deletion subclones are idencical co the cranscripes seen with the irgA: :phoA fusion on the chromosome of V. cholerae strain HBG40

suggests chat the results with CC118 (Fig.6 ) are not a plasmid-related

artifact nor unique to E. coli rather than V. cholerae.

DNA sequence of the irgA'-'phoA fusion and upstream DNA. Figure 8 (correspond to SEQ ID NO.: 1)

shows the DNA sequence of the chromosomal insert of pMBG59 from the

Clal site to the fusion with phoA. A 453 bp open reading frame, which

is in-frame with phoA. begins at position 1077. represencing the 5'

portion of irgA. A Shine-Dalgarno sequence is located just upstream of

the initiating methionine. The size of irgA upstream of the fusion

joint with phoA is slightly smaller chan we had predicted by Western

blot of IrgA'-'PhoA and by Northern blot of the irgA transcript, but

within reasonable experimental error.

The 900 bp region of DNA upstream of the BglII site that is

required for cranscripcion of irgA includes an 894 bp open reading

frame in inverse orientation to irgA (Fig.6). This open reading frame

begins at position 932, 144 bp upstream of the irgA open reading frame,

and terminates at position 39. just downstream of the ClaI restriction

site.

Primer extension analysis of the start site of irgA transcription.

Primer extension analysis of RNA from MBG40 grown in low iron was done

using two distinct synthetic oligonucleotides complementary to DNA

sequences located 20 bases and 64 bases downstream of the methionine

start codon (data not shown). Both oligonucleotides identified the

same approximate transcription start site, which is indicated by an

asterisk in Figure 8. A - 10 box. reasonably homologous to the E. coli

consensus sequence [14], was located approximately 7 bases upstream of

the transcription start site, but no consensus -35 box was identified. A 19 bp interrupted dvad symmetric sequence, homologous co the Fur binαing consensus sequence of E. coli (5.6]. was located immediately downstream of the transcription start site (Fig. 9: indicated by

converging horizontal arrows in Fig.8).

IrgA protein analysis. The predicted amino acid sequence encoded by the portion of irgA located upstream of the fusion joint with phoA is shown in Figure 8.(corresponding to SEQ ID NO.: 1).

(i) Hydropathicity index. The hydropathicity index profile of this portion of IrgA is shown in Figure 10 (a positive index indicates hydrophobic residues and a negative index indicates hydrophilic

residues). A stretch of hydrophobic residues, consistent with a signal sequence, is seen at the amino-terminus. This is followed by a second stretch of hydrophobic residues. 24 amino acids in length, suggestive of a possible transmembrane domain.

(ii) Homology of IrgA to E. coli ferrienterochelin receptor. The

151 amino acids at the amino-terminus of IrgA were analyzed using the FAST? algorithm for homologous proteins in the NBRF Protein Database. The best match in this search was the E . coli ferrienterothelin receptor. FepA. an 80 kDa iron- regulated oucer membrane procein [20]. The optimized score between the amino terminus of IrgA and FepA was

185. There were several regions of amino acid homology at roughly the same positions in each protein, including one stretch of ten

consecutive idencical amino acid residues (Fig.11).

Citrate utilization by MBG40. In addition to an iron assimilation system that utilizes the siderophore vibriobaccin. V. cholerae has been shown to have a ferric cicrace iron uptake system. enabling it to utilize citrace for grouch in low iron media [34]. We therefore evaluated whether or not MBG40 was able to utilize citrate to enhance growth in low-iron media and whether or not the expression of irgA. as determined by alkaline phosphatase assay, was altered by the addition of citrace. Cell density and alkaline phosphatase activity were measured for MBG40 after growth in low- and high- iron media with and without the addition of citrate (Table A). The addition of citrate to iron-deficient media partially restores growth of the mutant.

demonstrating that the mutant is able to utilize citrate. Further, alkaline phosphatase activity of the irgA'-'phoA fusion was not regulated in response to citrate. These data suggest that the ferric citrate iron uptake system is intact in the mutant MBC40 and that irgA is not involved in ferric citrate uptake.

LITERATURE CITED IN EXAMPLE 2

1. Bagg. A., and J.B. Neilands. 1985. Mapping of a mutation affecting regulation of iron uptake systems in Escherichia coli K-12. J.

Bacteriol. 161:450-453. 2. Bindereif. A., and J.B. Neilands. 1985. Promoter mapping and

transcriptional regulation of the iron assimilation system of plasmid ColV-K30 in Escherichia coli K-12. J. Bacteriol.

162:1039-1046. 3. Bjorn. M.J., B.H. Iglewski, S.K. Ives. J.C. Sadoff. and M.L. Vasil.

1978. Effect of iron on yields of exocoxin A in culcures of

Pseudomonas aeruginosa PA-103. Infect. Immun. 19:785-791.

4. Bjorn. M.J.. P.A. Sokol. and B.H. Iglewski. 1979. Influence of iron on yields of extracellular products in Pseudomonas aeruginosa cultures. J. Bacteriol. 138:193-200.

5. Calderwood, S.B.. and J.J. Mekalanos. 1987. Iron regulation of

Shiga-like coxin expression in Escherichia coli is mediaced by the fur locus. J. Bacteriol. 169:4759-4764. 6. Calderwood. S.B.. and J.J. Mekalanos. 1988. Confirmation of the Fur operacor site by insertion of a synthetic oligonucleotide into an operon fusion plasmid. J. Bacceriol. 170:1015-1017. 7. Chang. C.N., W. J. Kuang. and E.Y. Chen. 1986. Nucleocide sequence of the alkaline phosphatase gene of Escherichia coli. Gene.

44:121-125. 8. Ditta. G., S. Stanfield. D. Corbin. D.R. Helinski. 1980. Broad host range DNA cloning system for Gram-negative bacteria: construction of a gene bank of Rhizobium meliloti. 77:7347-7351.

9. Dubos. R.J., and J.W. Geiger. 1946. Preparation and propercies of Shiga toxin and toxoid. J. Exp. Med. 84:143-156.

10. Goldberg, K.B., V.J. DiRica. and S.B. Calderwood. Idendficacion of an iron-regulaced virulence decerminanc in Vibrio cholerae. using TnphoA mutagenesis. 1990. Infect. Immun. 58:55-60.

11. Griffiths. G.L.. S.P. Sigel. S.M. Payne, and J.B. Neilands. 1984.

Vibriobactin. a siderophore from Vibrio cholerae. J. Biol. Chem. 259:383-385. 12. Hantke, K. 1981. Regulation of ferric iron transport in Escherichia coli K12: isolation of a constitutive mutant. Mol. Gen. Genet. 182:288-292.

13. Hantke. K. 1984. Cloning of the repressor protein gene of

iron-regulated systems in Escherichia coli K12. Mol. Gen. Genet.

197:337-341. 14. Hawley, D.K.. and W.R. McClure 1983. Compilation and analysis of Escherichia coli promoter DNA sequences. Nucleic Acids Res.

11:2237-2255. 15. Henikoff, S., G.W. Haughn, J.M. Calvo. and J.C. Wallace. 1988. A large family of bacterial activator proteins. Proc. Natl. Acad. Sci. USA. 85:6602-6606.

16. Hovde, C.J., S.B. Calderwood. J.J. Mekalanos. and R.J. Collier.

1988. Evidence chat glutamic acid 167 is an active-site residue of Shiga-like toxin I. Proc. Natl. Acad. Sci. USA. 85:2568-2572.

17. Jonson, G., A. M. Svennerholm, and J. Holmgren. 1989. Vibrio

cholerae expresses cell surface antigens during incescinal infeccion which are noc expressed during in vitro culture. Infect. Immun. 57:1809-1815.

18. Kyte. J., and R.F. Dooliccle. 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157:109-132.

19. Lipman, D.J., and W.R. Pearson. 1985. Rapid and sensitive protein similarity searthes. Science. 227:1435-1441.

20. Lundrigan, M.D.. and R.J. Kadner. 1986. Nucleotide sequence of the gene for the ferrienterothelin receptor FepA in Escherichia coli.

J. Biol. Chera. 261:10797-10801. 21. Manoil. C. , and J. Beckwich. 1985. TnphoA: a cransposon probe for procein export signals. Proc. Natl. Acad. Sci. USA. 82:8129-8133.

22. Mekalanos. J.J., D.J. Swartz, G.D.N. Pearson. N. Harford. F.

Groyne, and M. de Wilde. 1983. Cholera coxin genes: nucleocide sequence, delecion analysis and vaccine developmenc. Nature.

306:551-557.

23. Michaelis, S., H. Inouye, D. Oliver, and J. Beckwich. 1983.

Mucacions chac alcer the signal sequence of alkaline phosphacase in Escherichia coli. J. Bacteriol. 154:366-374.

24. Miller, S.I., S.M. Landfear, and D.F. Wirth. 1986. Cloning and

characterization of a Leishmania gene encoding a RNA spliced leader sequence. Nucleic Acids Res. 14:7341-7360.

25. Miller, V.L. , R.K. Taylor, and J.J. Mekalanos. 1987. Cholera toxin transcripcional accivacor ToxR is a cransmerabrane DNA binding protein. Cell. 48:271-279.

26. Neilands, J.B. 1982. Microbial envelope proceins relaced co iron.

Ann. Rev. Microbiol. 36:285-309.

27. O'Callaghan, C.H.. A. Morris. S.M. Kirby. and A.H. Shingler. 1972.

Novel raechod for detection of /3-lactamases by using a

chromogenic cephalosporin substrate. Antiraicrob. Agents and Chemother. 1:283-288. Soc. Gen. Microbiol. 5:40-56

29. Peterson, K.M., and J.J. Mekalanos. 1988. Characterization of the Vibrio cholerae ToxR regulon: identification of novel genes involved in intestinal colonization. Infect. Immun. 56:2822-2829.

30. Pettis. G.S., T.J. Brickman, and M.A. Mclntosh. 1988.

Transcriptional mapping and nucleotide sequence of the Escherichia coli fepA-fes enterobaccin region. J. Biol. Chem. 263:18857-18863.

31. Sambrook, J., E.F. Fritsch, and T. Maniacis. 1989. Molecular

cloning: a laboracory manual. 2nd ed. Cold Spring Harbor Laboracory Press. Cold Spring Harbor. N.Y.

32. Sciortino. C.V., and R.A. Finkelstein. 1983. Vibrio cholerae

expresses iron-regulated outer membrane proteins in vivo. Infect. Immun. 42:990-996. 33. Sigel. S.P., and S.M. Payne. 1982. Effect of iron limitation on growth, siderophore production, and expression of outer membrane proteins of Vibrio cholerae. J. Bacteriol. 150:148-155.

34. Sigel. S.P., J.A. Stoebner, and S.M. Payne. 1985. Iron-vibriobaccin transport system is not required for virulence of Vibrio cholerae.

Infect. Immun. 47:360-362. 35. Simon. R., U. Priefer. and A. Pύhler. 1983. A broad host range mobilization system for in vivo genetic engineering: cransposon mutagenesis in Cram-negative bacteria. Bio/Technology. 1:784-791. 36. Stoebner. J.A., and S.M. Payne. 1988. Iron-regulated hemolysin production and utilization of heme and hemoglobin by Vibrio cholerae. Infect. Immun. 56:2891-2895.

37. Taylor, R.K.. V.L. Miller, D.B. Furlong, and J.J. Mekalanos. 1987.

Use of phoA gene fusions to identify a pilus colonization factor coordinately regulated with cholera toxin. Proc. Natl . Acad. Sci. USA. 84 :2833-2837.

Example 3: Cloning and characterization of irgB , and insertional inactivation of irgB in V. cholerae

MATERIALS AND METHODS

Bacterial strains. The V. cholerae wild-type strain used in this study was classical Ogawa strain 0395 Sm^r. V. cholerae strain MBG40 is 0395 irgA::TnphoA. E. coli strain CC118 is a phoA deletion derivative of MC1000 (13), and E. coli strains SY327 λ pir and SM10 λ pir have been described previously (14). V. cholerae strains MBG259 and MBG260 were constructed as described in

Genetic Methods below.

Media. Two types of liquid media were used to evaluate the effect of iron concentration on gene

expression: (i) LB medium with or without the addition of the iron chelator 2,2-dipyridyl (final concentration 0.2 mM), and (ii) tris-buffered medium (T medium) with or without the addition of FeSO₄ (final concentration 36 μM). For E. coli strain CC118, T medium was supplemented with thiamine (10 μg/ml) and the L-amino acids arginine and leucine (40 μg/ml).

Construction of plasmids. Strain MBG40 contains a chromosomal gene fusion between irgA and phoA,

constructed by TnphoA mutagenesis. Plasmids pMBG59, pMBG110, pMBG103, and pMBG58 contain subclones of the intact irgA::TnphoA gene fusion from MBG40 into plasmid pBR322, with decreasing amounts of chromosomal DNA upstream of irgA as described in Example 2. Plasmid pMBG59 contains all of the open reading frame upstream of irgA (subsequently named irgB, see Results below), while plasmids pMBG110, pMBG103, and pMBG58 contain

progressively less upstream DNA and larger deletions of irgB (Fig. 6).

Plasmids pSBC45 and pSBC46 are derivatives of pACYC184, a plasmid encoding chloramphenicol resistance, which is compatible with pBR322 and its derivatives.

Plasmids pSBC45 and pSBC46 were constructed by isolating the NruI-SmaI fragment of plasmid pMBG59 by

electroelution from a gel and ligating it into Nrul

digested pACYC184; the two plasmids differ only in the orientation of the inserted fragment. The Nrul-Smal fragment of pMBG59 contains the intact gene irgB (Fig. 6, see Results below).

Plasmid pMBG111 was derived from pGP704, a broad host range plasmid containing the ampicillin resistance gene from pBR322, the mobilization domain of plasmid RP4 (15), the origin of replication from plasmid R6K (16), and a polylinker from M13 tgl31 (Amersham Corp.).

Plasmid pGP704 was a gift of Gregory D. N. Pearson and is itself derived from plasmid pJM703.1 (14). Plasmid pGP704 and its derivatives are able to replicate only in strains containing the pir gene, which encodes the π protein necessary for the function of the R6K origin (16). To construct pMBGlll, a 676 bp HincII-BglII fragment of pMBG59 internal to irgB (Fig. 6) was ligated into the EcoRV and Bglll sites of the pGP704 polylinker.

Genetic methods. V. cholerae strains MBG259 and MBG260, which contain insertion mutations in irgB, were constructed from strains 0395 and MBG40, respectively, in the following manner. Plasmid pMBGlll was transferred from strain SY327 λ pir into SM10 λ pir by

transformation. SM10 λ pir contains a chromosomally-integrated RP4-2 (Tc::Mu), which encodes trans-acting factors necessary to mobilize pGP704 derivatives into a broad range of recipients without RP4 itself being transferred (14). SM10 λ pir containing pMBG111 was conjugated separately with 0395 and MBG40, with double selection for ampicillin resistance (encoded by pMBG111) and streptomycin resistance (encoded by 0395 or MBG40). Because 0395 and MBG40 do not contain the pir gene, pMBGlll is unable to replicate in either of these

strains, so that doubly-resistant colonies arise by homologous recombination between the internal fragment of irgB on pMBG111 and the corresponding chromosomal gene on the recipient, causing insertional inactivation of irgB. To confirm that chromosomal integration occurred within irgB on the recipient, we performed Southern

hybridization of chromosomal DNA digests using EcoRV. an enzyme that does not cut within either irgB or pMBG111. The blot was probed with the HincII-BglII fragment of irgB, radioactively labeled by random primer extension with a commercial kit (Prime Time, International

Biotechnologies, Inc.).

Plasmids were transformed into E. coli strains by standard techniques (17). Plasmids pACYC184 and pSBC45 were introduced into V. cholerae strains 0395, MBG40, MBG259, and MBG260 by electroporation, using the protocol of the manufacturer (Gene Pulser, Bio-Rad, Rockville Centre, NY), with the exception of substitution of 2mM CaCl₂ as the buffer for resuspending cells during

preparation, rather than water or HEPES buffer.

Assays. The enzymatic activity of alkaline phosphatase encoded on TnphoA permitted the comparison of fusion gene expression when strains were grown in low- versus high-iron media. Strains were grown overnight in either LB medium with or without added 2,2-dipyridyl or T medium with or without added FeSO₄. Alkaline phosphatase activity was determined as described in Example 1.

DNA and RNA analysis. Analysis of DNA and RNA, including DNA and RNA preparation, restriction mapping, DNA sequencing, Northern blot analysis and primer

extension were performed as described in Example 2. For Northern blot analysis, an equivalent quantity of RNA, as calculated from OD₂₆₀, was loaded into each lane. Synthetic oligonucleotides used as probes for Northern blot analysis and as primers for DNA sequencing and primer extension were the generous gift of Brian Seed.

Protein analysis and protein database searches. Protein analysis and protein database searches were performed using IBI-Pustell Sequence Analysis software (International Biotechnologies, Inc., New Haven, CT.). The hydropathicity index profile of IrgB was calculated by the formula of Kyte-Doolittle (18). The protein secondary structure prediction of IrgB was calculated by the algorithm of Chou-Fasman (19). Database searches and protein alignments were performed by searching the NBRF Protein Database (Release 19) using the FASTP algorithm for protein homology (20).

RESULTS

Northern blot analysis of the open reading frame upstream of irgA. In order to determine whether an RNA transcript was associated with the upstream open reading frame, and if present, to determine the size of the transcript and whether or not transcription was regulated by iron, we performed Northern blot analysis of RNA prepared from strains 0395 and MBG40 following growth in low- and high-iron media (Fig. 12). The blot was probed with an oligonucleotide complementary to the DNA sequence near the 5' terminus of the open reading frame. A single band of approximately 1.1 kilobases (kb) in size was seen in RNA prepared from 0395 and MBG40 grown under low-iron conditions (Fig. 12, Lanes 2 and 4), demonstrating that a transcript was associated with the open reading frame. No bands were seen in either strain grown under high-iron conditions (Fig. 12, Lanes 1 and 3), demonstrating that transcription of the open reading frame was

negatively regulated by iron. The size of the transcript, in relation to the size of the open reading frame (894 bp, see below), suggests that the transcript is monocistronic. The iron-regulated gene encoded by this open reading frame was designated irgB.

Trans-complementation of irgB and irgA. We have previously demonstrated, by deletion subcloning of pMBG59 in an E. coli background, that deletion of any portion of irgB eliminates transcription of irgA (Example 2) . We wished to examine whether irgB could restore expression of irgA in trans.

(i) Trans-complementation in an E. coli

background. Alkaline phosphatase assays were performed for E. coli strain CC118 containing plasmid pMBG59 and plasmid deletion subclones pMBG110, pMBG103, and pMBG58, alone and in combination with pSBC45 (which carries an intact irgB gene) following growth in low- and high-iron media (Table 5). CC118 containing pMBG59, which carries all of irgB and the irgA'-'phoA fusion, had significant iron-regulated alkaline phosphatase activity, as

described in Example 2. CC118 containing pMBGHO, pMBG103, or pMBG58, each with a progressive deletion of irgB (Fig. 6), had lost alkaline phosphatase activity that could be partially or completely restored by pSBC45 in trans. Restoration of iron regulation by pSBC45 in trans was variable or incomplete in the E. coli

background with high copy number plasmids.

Complementation of pMBG110, pMBG103, or pMBG58 with the vector plasmid pACYC184 in CC118 did not restore alkaline phosphatase activity (data not shown). In addition, alkaline phosphatase activities of strains complemented with pSBC46 (the intact irgB gene cloned in the opposite orientation in pACYC184) were similar to those obtained for strains complemented with pSBC45 (data also not shown). (ii) Trans-complementation in V. cholerae. In order to examine trans-complementation of irgB and irgA in the V. cholerae background, we constructed the irgB mutant strains MBG260 and MBG259 from MBG40 and 0395 respectively. Highly expressed, iron-regulated alkaline phosphatase activity seen in strain MBG40 was almost completely eliminated with the introduction of the irgB mutation to make strain MBG260 (Table 5). Iron-regulated alkaline phosphatase activity was completely restored by the introduction of irgB in trans on plasmid pSBC45

(Table 5), while the introduction of the vector pACYC184 had no effect (data not shown). Strains 0395 and MBG259, which do not contain an irgA'-'phoA fusion, had

negligible alkaline phosphatase activity, with or without the introduced plasmids (data not shown). The higher alkaline phosphatase activities seen in MBG260(pSBC45) and MBG40(pSBC45), as compared to MBG40 (Table 5), may result from the high copy number of irgB carried on pSBC45. These data suggest that irgB is a trans-acting factor that positively regulates irgA.

DNA sequence and deduced protein sequence of irgB. Fig. 13 (SEQ ID NO.: 2) shows the DNA sequence of the chromosomal insert of pMBG59 (reading 5' to 3' from right to left in Fig. 6), starting approximately 60 bp beyond the transcription start site of irgA and extending up to the Clal site of pMBG59, including the 894 bp open reading frame of irgB. A Shine Dalgarno sequence is indicated just upstream of the initiating methionine (21). A perfect inverted repeat, suggestive of a

transcription terminator, is indicated just beyond the termination codon of the irgB open reading frame.

Primer extension analysis of the start site of irgB transcription. Primer extension analysis of RNA prepared from MBG40 and 0395 following growth in low-iron media was done using a synthetic oligonucleotide complementary to the DNA sequence located between 3 bases upstream and 17 bases downstream of the methionine start codon (data not shown). The same transcription start site was identified in both MBG40 and 0395 and is

indicated by an asterisk in Figure 13. A promoter homologous to the E. coli consensus sequence (22) was located upstream of the transcription start site

(Fig. 13; SEQ ID NO.: 2). The 19 bp interrupted dyad symmetric sequence that is homologous to the Fur binding consensus sequence of E. coli (5,23), and is located immediately downstream of the irgA transcription start site, also overlaps the irgB transcription start site and -10 box (Fig. 13; SEQ ID NO.: 2). Fig. 14 shows the overlapping but divergent irgA and irgB promoters, and the location of the Fur-like box in relation to each.

IrgB protein analysis. The predicted amino acid sequence of IrgB is shown below the nucleotide sequence in Fig. 13 (SEQ ID NO.: 2).

(i) Hydropathicity profile. The hydropathicity profile of IrgB showed no stretches of hydrophobic residues that would be suggestive of either a signal sequence or a transmembrane domain (data not shown).

(ii) Homology of IrgB to the LysR family of positive transcriptional activators. Comparison of IrgB to the NBRF Protein Database demonstrated significant homology between IrgB and the LysR family of positive transcriptional activator proteins in bacteria (24). The best match in this family was to the E. coli positive activator protein IlvY; the optimized score between IrgB and IlvY was 217. The regions of highest homology between IrgB and the LysR family were near the amino terminus and were at roughly the same positions in each protein. The homology near the amino terminus of IrgB with several members of the LysR family is shown in Fig. 15. All of the proteins shown are of approximately the same size (IrgB is 298 amino acids; IlvY, LysR, AmpR, NodD, and CysB are 297, 300, 291, 300, and 301 amino acids, respectively).

(iii) Prediction of the secondary structure of IrgB. The secondary structure of IrgB was predicted using the Chou-Fasman algorithm (19). A helix-turn-helix motif was seen in the same region of IrgB as in the other members of the LysR family (Fig. 15) (24).

LITERATURE CITED IN EXAMPLE 3

1. Taylor. R. K.. Miller, V. L., Furlong, D. B. & Mekalanos, J. J.

(1987) Proc. Natl. Acad. Sci. USA 84, 2833-2837.

2. Pecerson, K. M. & Mekalanos J. J. (1988) Infect. Immun. 56.

2822-2829.

3. Pappenheimer. Jr. A. M. (1955) Svmp. Soc. Gen. Microbiol. S,

40-56.

4. Dubos, R. J. & Geiger, J. W. (1946) J. Exp. Med. 84, 143-156.

5. Calderwood. S. B. & Mekalanos, J. J. (1987) J. Bacceriol. 169.

4759-4764.

6. Bjorn. M. J., Iglewski, B. H., Ives. S. K.. Sadoff. J. C. 6c Vasil, M. L. (1978) Infect. Immun. 19, 785-791.

7. Bjorn, M. J., Sokol, P. A. & Iglewski, B. H. (1979) J. Bacteriol.

138. 193-200.

8. Actis, L. A., Potter. S. A. & Crosa, J. H. (1985) J. Bacteriol.

161, 736-742.

9. Carniel. E., Mazigh. D. & Mollaret. H. H. (1987) Infect. Immun.

55. 277-280.

10. Goldberg. M. B., DiRiCa, V. J. & Calderwood, S. B. (1990) Infect.

Immun. 58. 55-60.

11. Goldberg. M.B., Boyko, S.A. & Calderwood, S.B. J. Bacceriol. (In press).

12. Mekalanos, J. J., Swartz, D. J., Pearson, G. D. N., Harford. N., Groyne, F. & de Wilde. M. (1983) Nature 306, 551-557.

13. Manoil. C. & Beckwich, J. (1985) Proc. Natl. Acad. Sci. USA 82, 8129-8133. 14. Miller. V. I. & Mekalanos. J. J. (1988) J. Bacteriol. 170, 2575-2583.

15. Simon, R., Priefer. U. & Puhler, A. (1983) Bio/Technology 1.

784-791.

16. Kolter. R. , Inuzuka, M. & Helsinki. D. R. (1978) Cell 15.

1199-1208.

17. Sambrook, J., Fricsch, E. F. & Maniatis, T. (1989) Molecular

Cloning: A Laboratory Manual. 2nd ed. (Cold Spring Harbor Lab

Press, Cold Spring Harbor, NY).

18. Kyte. J. & Dooliccle, R. F. (1982) J. Mol- Biol. 157, 109-132.

19. Chou. P. & Fasman, G. D. (1974) Biochemistry 13, 222-245.

20. Lipman, D. J. & Pearson. W. R. (1985) Science 227. 1435-1441.

21. Shine. J. & Dalgarno. L. (1975) Nature 254, 34-38.

22. Hawley, D. K. & McClure. W. R. (1983) Nucleic Acids Res. 11.

2237-2255.

23. Calderwood, S. B. & Mekalanos, J. J. (1988) J. Bacteriol. 170.

1015-1017.

24. Henikoff, S., Haughn, G. W., Calvo, J. M. & Wallace, J. C. (1988) Proc. Natl. Acad. Sci. USA 85. 6602-6606.

25. Pettis. G. S., Brickman. T. J. & McIntosh, M. A. (1988) J. Biol.

Chera. 263, 18857-18863.

TABLE 2

TABLE 2 . Vinilence assays of wild-type and mutant strains

Steam Competive Competitive index LD ₅₀ strain In vitro In vivo (no. of bacteria)

0395 4 × 10³ MBG40 0395 0.98 0.11 3 × 10⁵

TABLE 3 Alkaline phosphatase assays in T media with and without added iron.

Strain Alkaline phosphacase activity (U/A₆₀₀)

With iron Uithouc iron

0395 (pSBC34) 82 496

CC118 (pMBG53) 63 277

TABLE 4 Growth and alkaline phosphatase activity of MBG40 grown overnight in low and high iron T media, with and without added citrate.

Growth conditions A₆₀₀ Alkaline phosphatase activity (U/A₆₀₀)

Low iron, without cicrace 0.397 320

Low iron, with cicrace 0.801 167

High iron, wichouc cicrace 1.431 1

High iron, with cicrace 1.663 1

TABLE 5. Complementation of irgB and irgA' -'phoA

in E. coli and V. cholerae.

Alkaline phosphatase assays in low and high iron media.

Strain Alkaline phosphatase activity (U/A₆₀₀)

Low iron High iron

Complementation in E. coli

CC118(pMBG59) 222 60

CC118(pKBG110) 4 1

CC118(pMBG103) 3 1

CC118(pMBG58) 1 1

CC118(pSBC45) 0 1

CC118(pMBG110.pSBC45) 209 11°

CC118(pHBG103,pSBC45) 124 68

CC1188(pMBG58.pSBC45) 63 108

Complementation in V. cholerae

MBG-0 683 3

MBG260 4 1

MBG40(pSBC45) 1420 11

MBG260(pSBC45) 1573 8

Assays of E. coli strains were performed following growth in

T medium with and without added FeSO^; assays of V. cholerae strains were performed following growth in LB medium with and without added dipyridyl. SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: Goldberg, Marcia

Calderwood, Stephen B.

Mekalanos, John J.

(ii) TITLE OF INVENTION: VIBRIO CHOLERAE STRAINS

DEFECTIVE IN irgA

EXPRESSION, AND CHOLERA VACCINES DERIVED THEREFROM

(iii) NUMBER OF SEQUENCES: 2

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSE E Fish & Richardson

(B) STREET: 225 Franklin Street

(C) CITY: Boston

(D) STATE: Massachusetts

(B) COUNTRY: U.S.A.

(F) ZIP: 02110-2804

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: 3.5" Diskette, 1.44 Mb storage

(B) COMPUTER: IBM PS/2 Model 50Z or 55SX

(C) OPERATING SYSTEM: IBM P.C. DOS (Version 3.30)

(D) SOFTWARE: WordPerfect (Version 5.0)

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER: 07/629,102

(B) FILING DATE: December 18, 1990

(C) CLASSIFICATION:

(vii) PRIOR APPLICATION DATA:

(A) APPLICATION NUMBER:

(B) FILING DATE:

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: Clark, Paul T.

(B) REGISTRATION NUMBER: 30,162

(C) REFERENCE/DOCKET NUMBER: 00786/058001

(ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (617) 542-5070

(B) TELEFAX: (617) 542-8906 (C) TELEX : 200154

(2) INFORMATION FOR SEQ ID NO: 1 :

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 1530

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1:

ATCGATGATA AAAAATCCCG CTGCGGCGGG ATTTTTTATT GCCACTCATC GGGCCTTGCT 60

TGGCGGAGCG CATCAATAAA TAGGCGCAGC CGAAGTGGGT GACGACCGAG CGGATAGAAG 120

CAGTTGATTT CTGTTGGCTG TGATTGCCAT CCGTTGACGC AAGGAATGAG GCTGCCCGGA 180

TGCGCCGTTT CAAAACCATT GGCAAACCAA GTGGGAAGCA AACCAATACC ACGACCTTTA 240

GCAATCGCAT CGGCTTGCAT GGCAAGATTA TCGCTTTGTA AACGACTCTC TAGTGCTGGC 300

AGTGAATAAC TGCCGAACTC TGGATGGTGC AGTTCAAGCT CCGCGCGCCG ACAAGCAATA 360

AAATCAATCC ATGGGTGATG AATCAGCTCA CGAGGATGGG TCGGTTTATC TCGATGGGCC 420

AAATATTTGG GAGAGGCGTA AGTGGCATAG CGCCAATAGC CTAAGCGTTC TTTGCGATAA 480

CCCATGGGGG CGGCGTGTTC AATCCAAATG ATCAAATCGG GCTCAAACAC CTCATCACTG 540

TGTTGAAACT GGCTGAGTAG ACGGATCTTC AATGTCGAAT GCTGCTGCAT AAACTCATCC 600

AATACTTGGC TGAGCCAGCC GCGGATCAAA TTGGGGTGTA CCACCAGCGT GAGTTCGCCA 660

GTCACTTGAT TGTTCAATTC TTGCAACGCT TCCTGACTTT TATTGGCCAG TTCAAGTAGT 720

TGCTCCGAGT AAACCGCAAA CACTTCTCCT GCTTTGGTGA GCGTTAAGCG GTTGCCTTGA 780

CGCATCAACA AGCTTTGTCC CAAGTCCTCT TCAAGTTGCG CCAAACGGCG ACTCAGGGTG 840

GATTTAGGCT GTTCAAGCGC TTTGGCAGCG GCAGTCAGGC TCTTATGTTG GCAAAGCGCA 900

TGGAAAGCTT TTACGGCGCT GAGATCTTGC ATAGGTATTT GACCCTTAAA GAATAATTAC 960

CACAGACGTT CCATATTTGG ACCGAACTAT TCCATGTGTC GATCTATCTC CAGTACAGAA 1020

TATATGAATA ATCCGCTTCT GAAATTAAGA ATAATTATCA TTTAAAGGAG TGGTAA ATG 1079

Met

1

TCC AGA TTC AAT CCA TCC CCC GTC AGT TTA TCT GTG ACA CTA GGC TTA 1127 Ser Arg Phe Asn Pro Ser Pro Val Ser Leu Ser Val Thr Leu Gly Leu

5 10 15

ATG TTT TCG GCT AGC GCT TTT GCT CAA GAC GCG ACG AAA ACG GAT GAA 1175 Met Phe Ser Ala Ser Ala Phe Ala Gin Asp Ala Thr Lys Thr Asp Glu

20 25 30

ACC ATG GTG GTC ACT GCG GCG GGA TAC GCG CAA GTG ATT CAA AAT GCA 1223 Thr Met Val Val Thr Ala Ala Gly Tyr Ala Gin Val Ile Gln Asn Ala

35 40 45

CCA GCC AGT ATC AGT GTG ATT TCA AGA GAA GAT CTG GAA TCT CGC TAT 1271 Pro Ala Ser Ile Ser Val Ile Ser Arg Glu Asp Leu Glu Ser Arg Tyr

50 55 60 65

TAC CGT GAT GTG ACC GAT GCG CTA AAA AGC GTA CCG GGT GTG ACA GTC 1319 Tyr Arg Asp Val Thr Asp Ala Leu Lys Ser Val Pro Gly Val Thr Val

70 75 80

ACC GGA GGG GGC GAT ACT ACC GAT ATC AGC ATT CGT GGT ATG GGA TCA 1367 Thr Gly Gly Gly Asp Thr Thr Asp Ile Ser Ile Arg Gly Met Gly Ser

85 90 95

AAC TAT ACT CTT ATC TTG GTG GAT GGT AAG CGC CAA ACC TCA CGC CAG 1415 Asn Tyr Thr Leu Ile Leu Val Asp Gly Lys Arg Gin Thr Ser Arg Gin

100 105 110

ACC CGT CCA AAC AGC GAT GGC CCG GGC ATT GAG CAA GGT TGG TTA CCG 1463 Thr Arg Pro Asn Ser Asp Gly Pro Gly Ile Glu Gin Gly Trp Leu Pro

115 120 125

CCA CTG CAA GCG ATT GAA CGT ATC GAG GTG ATC CGT GGC CCG ATG TCT 1511 Pro Leu Gin Ala I le Glu Arg I le Glu Val I le Arg Gly Pro Met Ser

130 135 140 145

ACG CTG TAC GGC TCG GAT G 1530

Thr Leu Tyr Gly Ser Asp

150

(2) INFORMATION FOR SEQ ID NO: 2 :

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 1100

(B) TYPE: nucleic acid

(C) STRANDEDNESS: Single

(D) TOPOLOGY:linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2:

GGGGGATGGA TTGAATCTGG ACATTTACCA CTCCTTTAAA TGATAATTAT TCTTAATTTC 60

AGAAGCGGAT TATTCATATA TTCTGTACTG GAGATAGATC GACACATGGA ATAGTTCGGT 120

CCAAATATGG AACGTCTGTG GTAATTATTC TTTAAGGGTC AAATACCT ATG CAA GAT 177

Met Gln Asp

1

CTC AGC GCC GTA AAA GCT TTC CAT GCG CTT TGC CAA CAT AAG AGC CTG 225 Leu Ser Ala Val Lys Ala Phe His Ala Leu Cys Gln His Lys Ser Leu

5 10 15

ACT GCC GCT GCC AAA GCG CTT GAA CAG CCT AAA TCC ACC CTG AGT CGC 273 Thr Ala Ala Ala Lys Ala Leu Glu Gln Pro Lys Ser Thr Leu Ser Arg

20 25 30 35

CGT TTG GCG CAA CTT GAA GAG GAC TTG GGA CAA AGC TTG TTG ATG CGT 321 Arg Leu Ala Gln Leu Glu Glu Asp Leu Gly Gln Ser Leu Leu Met Arg

40 45 50

CAA GGC AAC CGC TTA ACG CTC ACC AAA GCA GGA GAA GTG TTT GCG GTT 369 Gln Gly Asn Arg Leu Thr Leu Thr Lys Ala Gly Glu Val Phe Ala Val

55 60 65

TAC TCG GAG CAA CTA CTT GAA CTG GCC AAT AAA AGT CAG GAA GCG TTG 417 Tyr Ser Glu Gin Leu Leu Glu Leu Ala Asn Lys Ser Gln Glu Ala Leu

70 75 80

CAA GAA TTG AAC AAT CAA GTG ACT GGC GAA CTC ACG CTG GTG GTA CAC 465 Gin Glu Leu Asn Asn Gln Val Thr Gly Glu Leu Thr Leu Val Val His

85 90 95

CCC AAT TTG ATC CGC GGC TGG CTC AGC CAA GTA TTG GAT GAG TTT ATG 513 Pro Asn Leu I le Arg Gly Trp Leu Ser Gln Val Leu Asp Glu Phe Met

100 105 110 115

CAG CAG CAT TCG ACA TTG AAG ATC CGT CTA CTC AGC CAG TTT CAA CAC 561 Gln Gln His Ser Thr Leu Lys Ile Arg Leu Leu Ser Gln Phe Gin His

120 125 130

AGT GAT GAG GTG TTT GAG CCC GAT TTG ATC ATT TGG ATT GAA CAC GCC 609 Ser Asp Glu Val Phe Glu Pro Asp Leu I le I le Trp I le Glu His Ala 135 140 145

GCC CCC ATG GGT TAT CGC AAA GAA CGC TTA GGC TAT TGG CGC TAT GCC 657 Ala Pro Met Gly Tyr Arg Lys Glu Arg Leu Gly Tyr Trp Arg Tyr Ala

150 155 160

ACT TAC GCC TCT CCC AAA TAT TTG GCC CAT CGA GAT AAA CCG ACC CAT 705 Thr Tyr Ala Ser Pro Lys Tyr Leu Ala His Arg Asp Lys Pro Thr His

165 170 175

CCT CGT GAG CTG ATT CAT CAC CCA TGG ATT GAT TTT ATT GCT TGT CGG 753 Pro Arg Glu Leu I le His His Pro Trp I le Asp Phe Ile Ala Cys Arg

180 185 190 195

CGC GCG GAG CTT GAA CTG CAC CAT CCA GAG TTC GGC AGT TAT TCA CTG 801 Arg Ala Glu Leu Glu Leu His His Pro Glu Phe Gly Ser Tyr Ser Leu

200 205 210

CCA GCA CTA GAG AGT CGT TTA CAA AGC GAT AAT CTT GCC ATG CAA GCC 849 Pro Ala Leu Glu Ser Arg Leu Gln Ser Asp Asn Leu Ala Met Gln Ala

215 220 225

GAT GCG ATT GCT AAA GGT CGT GGT ATT GGT TTG CTT CCC ACT TGG TTT 897 Asp Ala I le Ala Lys Gly Arg Gly I le Gly Leu Leu Pro Thr Trp Phe

230 235 240

GCC AAT GGT TTT GAA ACG GCG CAT CCG GGC AGC CTC ATT CCT TGC GTC 945 Ala Asn Gly Phe Glu Thr Ala His Pro Gly Ser Leu I le Pro Cys Val

245 250 255

AAC GGA TGG CAA TCA CAG CCA ACA GAA ATC AAC TGC TTC TAT CCG CTC 993 Asn Gly Trp Gln Ser Gln Pro Thr Glu I le Asn Cys Phe Tyr Pro Leu

260 265 270 275

GGT CGT CAC CCA CTT CGG CTG CGC CTA TTT ATT GAT GCG CTC CGC CAA 1041 Gly Arg His Pro Leu Arg Leu Arg Leu Phe I le Asp Ala Leu Arg Gln

280 285 290

GCA AGG CCC GAT GAG TGG CAA TAA AAAATCCCGC CGCAGCGGGA TTTTTTATCA 1095 Ala Arg Pro Asp Glu Trp Gln

295

TCGAT 1100

Claims

CLAIMS 1. A V. cholerae cell harboring a mutation which inhibits expression in said cell of a functional irgA gene product.

2. The cell of claim 1, wherein said mutation prevents said expression.

3. The cell of claim 1, wherein said mutation is in said cell's irgA gene.

4. The cell of claim 3, wherein said mutation is a deletion.

5. The cell of claim 4, wherein at least 25% of the coding sequence of said irgA gene has been deleted.

6. The cell of claim 1, wherein said mutation is in said cell's irgB gene.

7. The cell of claim 6, wherein said mutation is a deletion.

8. The cell of claim 7, wherein at least 25% of the coding sequence of said irgB gene has been deleted.

9. A substantially purified preparation of V. cholerae cells, wherein each such cell harbors a mutation which inhibits expression in said cell of a functional irgA gene product.

10. The preparation of claim 9 , wherein said mutation prevents said expression.

11. The preparation of claim 9, wherein said mutation is in said cell's irgrϋ gene.

12. The preparation of claim 11, wherein said mutation is a deletion.

13. The preparation of claim 12, wherein at least 25% of the coding sequence of said irgA gene has been deleted.

14. The preparation of claim 9, wherein said mutation is in said cell's irgB gene.

15. The preparation of claim 14, wherein said mutation is a deletion.

16. The preparation of claim 15, wherein at least 25% of the coding sequence of said irgB gene has been deleted.

17. A cholera vaccine comprising live attenuated V. cholerae cells, each of which cells harbors a mutation which inhibits expression in said cell of a functional irgA gene product.

18. The cholera vaccine of claim 17, wherein said mutation prevents said expression.

19. The cholera vaccine of claim 17, wherein said mutation is a deletion in said cell's irgA gene.

20. The cholera vaccine of claim 17, wherein said mutation is a deletion in said cell's irgB gene.

21. The cell of claim 1, wherein said cell harbors a second mutation which inhibits production of a second V. cholerae virulence factor in said cell .

22. The cell of claim 21, wherein said second mutation prevents said production.

23. The cell of claim 22, wherein said second virulence factor is the A subunit of cholera toxin and said first and second mutations are deletions.

24. The cell of claim 23, wherein said first mutation is in said cell's irgA gene.

25. The cell of claim 23, wherein said first mutation is in said cell's irgB gene.

26. The preparation of claim 9, wherein each of said cells harbors a second mutation which inhibits production of a second V. cholerae virulence factor in said cell.

27. The preparation of claim 26, wherein said second mutation prevents said production.

28. The preparation of claim 26, wherein said second virulence factor is the A subunit of cholera toxin and said first and second mutations are deletions.

29. The preparation of claim 28, wherein said first mutation is in said cell's irgA gene.

30. The preparation of claim 28, wherein said first mutation is in said cell's irgB gene.

31. The cholera vaccine of claim 17, wherein each of said cells harbors a second mutation which inhibits production of a second V. cholerae virulence factor in said cell.

32. The cholera vaccine of claim 31, wherein said second mutation prevents said production.

33. The cholera vaccine of claim 32, wherein said second virulence factor is the A subunit of cholera toxin and said first and second mutations are deletions.

34. The cholera vaccine of claim 31, wherein said first mutation is in said cell's irgA gene.

35. The cholera vaccine of claim 31, wherein said first mutation is in said cell's irgB gene.

36. A cell derived from V. cholerae strain 0395-N1, wherein said cell harbors a deletion mutation which inhibits the expression in said cell of an active irgA gene product.

37. A cholera vaccine comprising the cell of claim 36.

38. A method for inducing immunity to cholera in a mammal, which method comprises administering to said mammal an effective amount of the cholera vaccine of claim 31.