WO1994001562A1

WO1994001562A1 - Bivalent living vaccines against bacterial intestinal pathogenic agents, process for preparing the same, plasmids and strains useful as base material

Info

Publication number: WO1994001562A1
Application number: PCT/EP1993/001715
Authority: WO
Inventors: Kenneth Timmis; Himanshu Brahmbhatt
Original assignee: Kenneth Timmis; Himanshu Brahmbhatt
Priority date: 1992-07-03
Filing date: 1993-07-02
Publication date: 1994-01-20
Also published as: DE4221840A1; AU4563793A

Abstract

A bivalent S. typhi living vaccine is disclosed. The micro-organism contained in this vaccine is capable of expressing a complete LPS with a S.-typhi core and heterologous O-polysaccharides. The micro-organism can be produced by transferring the rfa-L gene from E. coli and the gene region of a cross-reacting species whose genetic products catalyze the biosynthesis of O-lipopolysaccharides, into S. typhi.

Description

Title.

Bivalent live vaccine against bacterial intestinal pathogens, manufacturing processes as well as plasmids and strains as starting material.

Description.

The bacillary dysentery (Shigellenruhr) is an invasive disease of the colon in humans and higher primates and is transmitted by the fecal-oral route. It is highly infectious and 10 bacteria can lead to the illness of a healthy person. The dysentery is one of those diseases that are traditionally associated with poor hygiene, overpopulation and stress. Consequently, the majority of cases occur in developing countries with a poorly trained health system, but there have also been outbreaks of this disease in connection with war zones and intellectual currents in developed countries. It is caused by various species of Sh ige l la and enteroinvasive strains of Escherich ia co li (EIEC), which are present in endemic areas, while the most severe form of this disease is caused by S. dysenteriae 1. Infections caused in particular by S. dysenteriae 1 can furthermore lead to serious complications such as hemolytic uraemic syndrome (HUS), hypoproteinemia, leukemoid reactions and sepsis. According to a recent estimate, there are more than 100 million cases of bacillary dysentery worldwide and approximately 600,000 deaths a year. Great problems in the clinical treatment of the disease are caused by the fact that it cannot be treated by rehydration therapy and that most Shige 17a isolates are resistant to a large number of antibiotics, in particular to those which are generally available in developing countries. The widespread use and severity of Sh ige 7 la loose and the difficulties associated with their treatment are the reasons why the development of effective Sh ige 7 la vaccines is urgently needed.

Previous studies of vaccinations against bacterial intestinal pathogens suggested that protection against the pathogen was associated with O-antigen specificity (Mel et al. 1965; Formal et al. 1966; Dupont et al. 1969; Formal and Levine 1984). Recent attempts to construct bivalent vaccine strains based on the attenuated Salmon l la typhi Ty21a strain (currently used as an oral live vaccine against typhoid; Germanier and Furer 1976) as carriers for heterologous O-antigens such as O-antigens from Sh ige l la sonne i (Formal et al. 1981), Shige l la flexneri 2a (Baron et al. 1987; Macpherson et al. 1991), Sh ige l la dysenteriae 1 (Mills et al. 1988) and Vibrio cho Ierae serotype Inhaba (Tacket et al. 1990; Attridge et al. 1990) failed due to the fact that the heterologous O-polysaccharide is not bound to the core lipid A. from S. typh i. The foreign O-antigen therefore remains excluded from the bacterial cell as non-imogenic hapten polysaccharide and encloses the bacterium loosely like capsule material.

In contrast, the heterologous O-polysaccharides bind to the core lipid A structures of E. coli K-12 (Sansonetti et al. 1983; Stur et al 1986; Yoshida et al. 1991). The Shige l la dys¬ enteriae 1 O antigen also binds to the core / lipid A of Salmon l la typh i mu rium and S. dub l in (Mills et al. 1988). Attempts have already been made to solve this problem in that the S. typh i Ty21a rfa gene location (which codes for the enzymes involved in core biosynthesis) by conjugative DNA transfer through the analogous region from E. col i K-12 was replaced (Tacket et al. 1990). The hybrid Ty21a strain was able to link the V. cholerae O polysaccharide to the E. Col i K-12 core, both of which were expressed in Ty21a.

It is an object of the present invention to provide a vaccine which offers protection against intestinal pathogens, namely by providing an attenuated bivalent vaccine strain which expresses complete LPS, which is composed of a core lipid A and bound to it, heterologous O-polysaccharides, which carry the O-antigenicity.

This object was achieved by the provision of a Salmon I la type I live vaccine which possesses the properties mentioned.

In order to define the precise modification of the S. tp / 77 core or core, which is necessary for the binding of heterologous O-polysaccharides to the core lipid A, a more detailed analysis of the E. coli K-12 rfa Locally carried out. Through complementation analysis in defined rfa mutants of Salmon l la typhimurium we were able to identify the enzymes involved in the biosynthesis of the outer core of K-12 and the 0-antigen / core ligase. The results of this study show that it is not necessary to modify the structure of the S. typh i nucleus and that the S. dysenteriae 1 O-pol.ysaccharide is used by means of a single K-12 rfa function, i.e. the O antigen / core ligase to which S. typhi core can be bound. The bivalent vaccine strain expresses the O-polysaccharides of both S. dysenteriae 1 and S. typhi as complete LPS molecules.

A gene has now been identified that is found in £. co7 / K-12 rfa gene location, which codes for the LPS core biosynthesis, is located, namely the rfaL gene, which codes for the enzyme O-antigen / core ligase. This enzyme is along with the rfaT gene product involved in the catalysis of the linkage of the O-polysaccharide with the LPS core. The E. coli K-12 rfa gene was cloned, its sequence was located and transferred to a suitable cassette that can be transferred to a variety of possible vaccine strains. If one transfers the cloned rfaL gene to Sa7 mone l la typhi Ty21a, which carries the biosynthesis genes (rfb rfp gene cassette) for the Sh ige l la dysenteriae 1 O antigen on its chromosomes, it causes the linkage of the S. dys enteriae 1 O antigen with the S. typh i Ty21a LPS core. Accordingly, this hybrid vaccine strain expresses the O antigens of both S. typhi and S. dysenteriae as complete LPS.

It is now no longer impossible to construct bivalent vaccines, which was not feasible because heterologous O-polysaccharides do not bind to the LPS core of the antigen carrier strain S. typh i Ty21a. The construction can be carried out in a simple manner with the aid of E. coli K-12 O antigen / core ligase, since our studies show that this enzyme shows relaxed weakened specificity and is therefore able to remove the S. dysenteriae 1 O -Antigen to bind to the Ty21a core. Because of its relaxed specificity, this enzyme should also be able to generate various heterologous O-antigens, such as those from Sh ige l la sonne i, Sh ige l la f lexneri, Sh ige l la boydi i (the main causers of shigellose ), Vibrio c / 7θ7erae strains (the cause of cholera), Salmon l! to link a paratyph i strains (the etiological cause of paratyphus) and many from the gut pathogenic Escherichia coli strains. The hybrid vaccine strains can be used for oral immunization and should be able to elicit protective local immune responses in the intestinal mucosa that are directed against the predominant cell surface antigen, the O-antigen. This present invention relates not only to S. typh i Ty21a as an antigen carrier, but to any other attenuated S. typh / strain. Many such attenuated S. typh i strains have been constructed recently, for example the Vi + strain (Cryz et al. 1989), the cya, crp-λutante (Curtiss et al. 1987), the aroA- Mutant (Edward and Stoc¬ ker 1988), the arok, aroC and aroA, aroD double mutants (Dougan et al. 1988, Miller et al. 1989), the arc, aroD double mutant (Levine et al. 1990 ), the pboP mutant (Fields et al. 1989) and the ompR mutant (Dor an et al. 1989). Other recently developed carrier strains can also be used, for example the Vibrio cho lerae mutant CVD 103-Hg ^R ; A-, B + (Levine et al. 1988a, 1988b). Some of these mutants are currently being tested on human subjects, and at least one of them has been selected by the FDA (Food and Drug Administration, USA) for use in extensive human field trials.

Materials and methods.

Strains, plasmids and media

The plasmid pMN6 carries the gnd-Ger ^~ of E. coli K-12 (which codes for the 6-phosphogluconate dehydrogenase) and was a gift from Richard Wolf Jr. (Nasoff and Wolf 1980; Nasoff et al. 1984). The plasmid pSS37 carries the rfb-rfp gene cassette (which codes for the 0-antigen biosynthetic functions of Sh ige l la dysenteriae 1) and was obtained in our laboratory (Sturm and Timmis 1986). The plasmid pLPF / Ars is a chromosomal integration vector based on transposons with an arsenic resistance marker, usable for the selection of recombinant strains which carry the inserted DNA in the chromosome (Herrero et al. 1990). The plasmid pGP704 (Miller and Mekalanos 1988) carries an Ampici11 resistance marker, the RP4 / nob site and a Polylinker with cloning sites. It also acts as a suicide vector since it depends on the pir function for replication and can therefore only be kept in strains which carry pir. The strain SM10 pir (.thi-, thr, leu, tonk, lacY, εupE, reck:: RP4-2-Tc:: Mu, km ^r , pir) (Miller and Mekalanos 1988) was used as the host strain for the plasmid pGP704 and whose derivatives are used. All other plasmids were kept in strain CC118 ((ara-leu) araD, lacXlA, galE, galK, phok20, thi- ~ \, rpεE, rpoB, argE (km) reck) (Manoil and Beckwith 1985). The plasmids pLOF / Ars, pG704 and the strains SM10 pir and CC118 were kindly left to us by Victor de Lorenzo. The E. coli K-12 cosmid clones (No. 68 and No. 195, produced in the cosmid vector pJB8) were kindly provided by RP Birkenbihl (Birkenbihl and Vielmetter 1989). The 10.5 kB long plasmid pRK415 (Keen et al. 1988) carries the 7a promoter, which transcribes from the HindIII to the EcoRI cloning site in the polylinker. It is a mobilizable plasmid with tetracycline resistance. Salmonel la typhi Ty21a is a galactose epimerase-free (galE) mutant of S. typhi (currently used as a live vaccine against typhoid fever (Germanier and Furer 1975) and was a gift from Stanley Cryz Jr. The S. typhimurium hεdR -, hεdM + (SL5285 ^, Salmonella typhimi rium rfa mutant SL1655 (metk22, trpCZ, H1-b, H2-e, n, x 'cured of Fels 2' fla-66, rpsL120, xy7-404, /7.etE551 , 7 ^' 7v-452, hsdT6, hsdSk2S, rfaG3037; Bullas and Colεon 1975; Kadam et al. 1975), SL3769 (p rE +, rfaG471), SL3748 (pyrE +, rfa {R-res-2) 432 = rfaI432), SL3750 (pyrE +, rfaJ417) and SL3749 (p rE +, rfa446) (Roantree et al. 1977) were kindly provided by KE Sanderson The Salmonella typhimurium rfa mutants TV148 (rfaI432; Subbaiah and Stocker 1964) and SL1032 {trp , met, rfaG, εt ^R ; Osborn 1968) were a gift from AA Lindberg. The plasmids pBR322 (Bolivar et al. 1977) and pUC19 (Yanish-Perron et al. 1985) were used as a cloning vector for general purposes.

All E. coli and S. typh i mur ium strains were grown in Luria medium or agar (Miller 1972). Brain-heart infusion medium (BHI; Difco), with or without 1.5% agar, was used for the cultivation of S. typhi Ty21a and its derivatives. If necessary for the detection of the complete LPS, the BHI medium was supplemented by 0.005% D-galactose. Wilson Blair bismuth sulfate agar (obtained from BioMerieux, France) was used in conjugation experiments between E. coli donor cells and S. typh i Ty21a recipients. This medium only supports the growth of Sa Imone77a strains (results in black colonies) and enables the selection of the E. coli donor with the aid of the counter. If necessary, the media were supplemented by antibiotics: ampicillin (50 μg / ml), chloramphenicol (30 μg / ml), tetracycline (15 μg / ml) and kanamycin (50 μg / ml).

Recombinant DNA technology.

The construction of plasmids, their isolation and purification as well as the hybridization in the Southern blot was carried out according to conventional molecular cloning techniques (Maniatis et al. 1982). The restriction endonucleases, the Klenow fragment of E. coli DNA polymerase I and the T ₄ DNA ligase were obtained from Boehringer GmbH (Mannheim, Germany). The phoshorylated Xbal linkers were obtained from New England Biolabs (Beverly, Mass., USA). All reagents were used in accordance with the manufacturer. The transformation of the plasmids was carried out according to the Hanahan (1983) method. The plasmid mobilization and the evaluation of transposition events were carried out as described by Herero et al. (1990). Isolation and cleaning of LPS.

LPS was isolated and purified from bacterial strains using the Westphal and Jann (1965) phenol-water method, while LPS was prepared for routine analysis using the method described by Hitchcock and Brown (1983).

SDS-polyacrylamide gel electrophoresis, silver staining and western blotting from LPS.

Polyacrylamide gels were produced and used with the buffer system from Laemmli (Laemmli 1970). The "mini-protean II" device from Bio-Rad was used for electrophoresis. Separating gels were prepared either with 12% acrylamide and 0.1% SDS (for the detection of the complete LPS) or with 15% acrylamide (for the detection of mutant LPS). The samples were mixed 1: 1 with sample buffer (containing 4% SDS) and boiled for 5 minutes, and amounts of 20 µg were applied to the gel. The electrophoresis was carried out with a constant current of 20 mA until the color trace penetrated the separating gel and 50 mA until the color trace reached the bottom of the gel. The LPS bands were visualized using the silver staining method of Tεai and Fraεch (1982) or using the "semi-dry blotting" device (obtained from Biometra) on membranes made of Immobilon PVDF (polyvinylidene difluoride) (obtained from Millipore ) transferred and detected with polyclonal O-Antiεerum against either S. dysenteriae 1, Salmon I la typh i (anti-09) or Salmon I la typh i-murium (anti-04.5), the method of Sturm et al. (1984) was used. The PVDF membranes were used instead of nitrocellulose because it has been found to have a much better LPS binding and retention capacity. sit. The polyclonal O-antisera were obtained from Behringwerke AG, Marburg, Germany.

In the following, the various steps according to the invention are explained in detail in addition to a number of scientifically relevant measures.

Integration of the rfb-rfp cassette into the chromosome of Salmonella typhi Ty21a and expression of S. dysenteriae 1 O antigen in S. typhi Ty21a.

Previous studies have shown that the rfb-rfp cassette (which codes for the Shige l la dysenteriae 1 O-antigen biosynthesis functions) in the plasmid pSS37 is then unstable if it is in S. typhi Ty21a or in S. typhimurium (Mills et al. 1988). The plasmid also carries a chloramphenicol resistance marker, which is disadvantageous because markers with antibiotic resistance are undesirable in vaccine strains for medical or veterinary use. The choice was therefore made to integrate the rfb-rfp cassette into da Chromos of S. typhi Ty21a, using two different techniques: (i) interface-specific homologous recombination and (ii) transposon-mediated random integration . The interface-specific, homologous recombination is preferred since the foreign DNA can be inserted into a non-essential part of the chromosome without destroying functions which are required for the invasive capacity of the vaccine strain. The gnd gene (which codes for the 6-phosphogluconate dehydrogenase, part of the pentose-phosphate cycle) was selected for the site-specific recombination (see FIG. 1, which shows the plasmids). This gene (1.404 kb) is in Sa7- one l la and in E. Col i between the his and rf b genes, and DNA sequence analysis has shown that this region is highly conserved in both species (Barcak and Wolf 1988; Reeves and Stevenεon 1989) . The plasmid pMN6 (Fig. 1) carries the

along with some flanking DNA. The 741 bp long gnd fragment, which was found after a PstI / pr? I partial digestion of pMN6, was purified by gel electrophoresis and subcloned into PstI / piI sites of the plasmid pUC19, whereby the plasmid pHBIOO was obtained. The gπd segment was further subcloned as an EcoRI / Spbl fragment into the corresponding interface of the suicide vector pGP704, the plasmid pHB101 being formed. The Xbal section of the vector was then removed by bal digestion of the plasmid pHB101, filling to blunt ends with the Klenow fragment of DNA polymerase I and religating. The plasmid pHB102 formed was cleaved with EcoRV, ligated to phosphorylated-bal linkers, digested and ligated with Xbal, resulting in the plasmid pHB103, which carries a single Xbal cleavage site on the gnd gene. Then the plasmid pSS37 was digested with EcoRV, ligated to the phospellated Xbal linker and digested with Xbal, and the rfb-rfp cassette was subcloned into the plasmid pHB103 digested with Xbal, whereby plasmid pHB107 was formed. In this plasmid, the rfb-rfp genes are flanked by the left and right half of the gnd gene. This plasmid was transformed into the SM10 pir strain for use as a donor strain for conjugation when mating with S. typhi, Ty21a.

The corresponding genes from the plasmid pHB107 were used for the construction of a plasmid on a transpoison basis (ε. 1, which shows the plasmid) for the chromosomal random integration of the rfb-rfp cassette in S. typhi Ty21a cut out as a bal fragment and subcloned into the only Xbal view interface of the plasmid pLOF / Arε. The plasmid pHB120 formed was transformed into the SM10 pir strain and used as a donor for conjugation when mating with the S. typhi Ty21a strain. The pro- Conjugation products were plated on arsenic-containing Wilεon Blair agar, and the colonies obtained were first checked for their sensitivity to ampicillin (to determine the loss of plasmid DNA). The positive isolates were tested for the expression of O-antigen by both S. typhi and S. dysenteriae, by means of SDS-PAGE / Western blotting and point blotting (data not specified) using the corresponding O-antisera . The results showed that the hybride S. typh i Ty21a strain (designated H4098) expressed both homologous and heterologous O-polysaccharides. The strain expressed the complete LPS of S. typh i, because the complete LPS "ladder" could be observed in SDS-PAGE / Western blotting. The O-polysaccharide from S. dysenteriae 1 was - although expressed - not bound to the core lipid A of the host, since the Western blotting with O-antiserum from S. dysenteriae 1 showed only a weak running track in the upper one Part of the gel.

Identification and genetic analysis of the E. coli K-12 rfa cos-id clones.

Structural studies on the LPS core of S. typhimurium and E. coli K-12 (Fig. 3) show that part of the backbone-forming chain (GluI-HepII-HepI-KDO) is the same in both species. However, the core structures closest to the O antigens are different, which makes cross-complementation studies difficult. It has been suggested that the sugar transferase specified by rfaG (UDP-glucose: (heptosyl) LPS-1, 3-glucosyltransferase), which adds Glul to HepII, should be present in both species, while that of rfal specified (UDP-galactose: (glucosyl) LPS-1, 3-galactosyl-transferase), which Galll binds to Glcl, and the (UDP-glucose: (galactosyl) LPS-1, 2-glucosyltransferase) specified by rfaJ who have favourited GlcII Galll adds that they should only be present in S. typh imurium (Austin et al. 1990).

Analysis of the core structures suggests that the possible reasons why the S. dysenteriae 1 O antigen does not bind to the S. tyhph i nucleus could either be because the S. typh i cannot bind to the heterologous O-antigen in any way, as long as it is not modified to the structure found in E. coli K-12, or that the O-antigen / core ligase, which is the gene product von rfaL from S. typh i has a stringent specificity for the O-antigen of S. typh i, while the enzymes encoded by rfaL from E. co7 / K-12 and S. ty¬ ph imurium are a relaxed, weakened species ¬ fity, which enables the linking of heterologous O-polysaccharides to the respective core structures. The former possibility seems unlikely since the O antigen from S. dysenteriae 1 to the nucleus of S. typh imur ium binds and it is known from the core structure analysis that the Salmon l lae have identical core structures.

Therefore, it seems unlikely that the core of S. typh i should not bind the O-antigen from S. dysenteriae due to certain εpecific structures at the core terminus. However, if the latter proposal is correct, then it should be possible to bind the O-pole sacchaπd of S. dysenteriae to the S. typh i nucleus simply by having the rfaL gene of E. co77 K-12 incorporated into the S. typh i strain. Therefore, the path was chosen first, the rfa genes from E. co7 / K-12, which code for the biosynthesis of the outer core (the enzymes coded by rfaG, rfaM and rfaN) and the O-antigen core coded by rfaL - identify ligases.

The 8.5 kB 5 'end of the rfa locus of E. co77 K-12 has already been cloned and characterized on the genetic level (Austin et al. 1990). These studies led to the construction a protein map (using transcription-translation in vitro) of the 5 'end (8.5 kB) of the rfa locus in E. co7 / K-12, and the identification of one of the rfa genes from rfaG (coding for a 39 kD protein) apparently seems to have been successful there.

In order to obtain the complete rfa locus, the cosmid gene bank of E. co7 / K-12, which was created by Birkenbihl and colleagues (Birkenbihl and Vielmetter 1989), was analyzed. The comparative analysis of the known EcoRI restriction map of the Coεmid bank with an EcoRI map of the rfa locus of E. co77 K-12 and flanking DNA allowed the identification of two cosmid clones, No. 68 and No. 195 (FIG. 2), which were believed to carry the full rfa locus. The restriction enzyme analysis of clones 68 and 195 showed that the restriction cleavage sites within the 5'-sided 8.5 kB region of the gene location were identical to those identified by Austin and colleagues (1990). Partial digestion with EcoRI and subsequent religion of the cosmid clone 195 led to the formation of the plasmid pHB115 (FIG. 2), in which part of the DNA flanking the rfa locus was eliminated.

In order to clarify which enzyme functions, which are involved in the synthesis and assembly of the outer nucleus, are expressed by the 5 'half of rfa, a number of subclones were generated with the aid of the known restriction interfaces in this region (Auεtin et al. 1990, namely (i) the digested fragment of plasmid pHB115 partially digested by Sa7I Bglll, sub-cloned into the Sa71 / ßa / τ-HI sections of the plasmid pBR322 (plasmid pHB127); (ii) the Sa II / Aval fragment, subcloned into the Sa7I /, 4vaI interfaces of the plasmid pBR322 (plasmid pHB130); and (iii) the Sa 11 / Bg III fragment, subcloned into the Sa7I / ßa / τ] HI sites of the plasmid pBR322 (plasmid pHB128). Identification of the rfaG function of E. coli K-12 by complement analysis.

The plasmids pHB127, pHB128 and pHB130 were transformed into defined, rfaG mutant strains of S. typh imurium: SL1655, SL3769, SL1032. Since it was not known whether these strains would have a functional rfb locus (which codes for O-antigen biosynthesis functions), they were co-transformed with plasmid pSS37 to produce the O-antigen of Sh ige l la dysenteriae 1 to win. Should the rfa subclones provide the enzyme function that was missing in the mutant and should the E. co77 enzyme be able to complement the analogue Salmon 17a function, then the rfa enzymes remaining from Salmon la should (provided that they should still be functional in the mutated rfa locus of the Sa Imone 17a host) should complete the residue of the complemented nucleus. The O antigen from S. dysenteriae (the enzymes being encoded by the plasmid pSS37) should be able to bind to the complemented nucleus with the formation of complete LPS. If the Salmon la rfb genes are actually also functional, then both homologous and heterologous LPS species should be synthesized by the hybrid strain.

The LPS profiles of the hybrid strains were analyzed by SDS-PAGE and subsequent silver staining, as well as by Western blotting with both homologous and heterologous O-specific antisera. The results show (FIG. 4.1; gels only shown by way of example) that the plasmids pHB127 and pHB130 complemented the rfaG defect in the mutants SL1655, SL3769 and SL1032 (FIG. 4.1; only the data for the plasmid pHB130 are shown in strain SL3769, since the others were identical). In each of the rfaG mutants, the presence of one of the two plasmids (pHB127 or pHB130) was sufficient to complete the LPS (FIG. 4.1, lane B), which suggests that, despite the defective rfa locus, this Strains that have intact rfb genes are un perform their function. However, the plasmid pHB128 did not complement the rfaG mutation, which suggests that the coding region of rfaG via the

extends in the direction of the 5 'end of rfa. The presence of only the plasmid pSS37 (which codes for the O-antigen biosynthetic functions of S. dysenteriae) in all these mutant strains did not allow the heterologous O-polysaccharides to bind to the mutated core structures, as expected (FIG. 4.1 , Track C) since the O-anti gene could not bind to the incomplete core.

This result suggested that either the rfaG function was located between the ßg <7II and / Aval interfaces (positions 4.9 and 7.7, Fig. 2), or that the coding region was located via the ß57II interface le reaches out.

Identification of the jrfaM and rfaN enzyme functions of E. coli K-12.

Previous studies have shown that a Clarke carbon plasmid pLC10-7 (which carries part of the rfa locus of E. co77 K-12; Clarke and Carbon 1976) complements the rfal and rfaJ defects in Salmon la (which could be determined by appropriate enzyme tests) and the E. co77 K-12 analogue from Salmon l la rfal alε rfaM (UDP-glucose: (glycosyl) LPS-1, 3-glucosyltransferase; Creeger and Rothfield 1979) terminated. Auεtin and al. (1990) further suggested that the rfaJ analog be terminated as rfaN (UDP-Glucoεe: (Glucoεyl) LPS-1, 2 Glucoεyltransferase). Sanderεon hypothesized that the pLC10-7 complementation of Salmon l la rfal and rfaJ defects could indicate that the complemented nucleus of E. co77 type is and contains three glucose residues to which Salmon l la O-anti gene is then bound (Austin et al. 1990). In order to determine whether the plasmids pHB127, pHB128 and pHB130 carry the rfaM and rfa-N enzyme functions, the corresponding plasmids were transformed into rfa mutant strains of Salamone l la typhimurium: SL3748 (rfal-), TV148 (rfal -), SL3750 (rfaJ-). The LPS profiles on SDS-PAGE gels stained with silver showed that the rfal defects in the strains SL3748 and TV148 could only be complemented by the plasmid pHB127 (FIG. 4.2, lanes G and J; the data are only given for the plasmids pHB130 and pHB127 in strain SL3748). Interestingly, the presence of the plasmid pHB127 in the rfal mutated strain led to the expression of very small amounts of complete LPS (Fig. 4.2, lane G), which may suggest that the rfb locus in this strain (a) "leak" mutation (en) carries.

The rfaJ mutation was also complemented by only the plasmid pHB127 (FIG. 4.3, lane P). In this strain the presence of the plasmid pHB127 alone did not lead to complete LPS, which shows that the rfb site is mutated and not functional.

These results suggest that the rfaM and rfaN functions are encoded by the .4va7I-ßg7II fragment (positions 7, 7 and 11, 1; FIG. 2); however, it is not known whether one of these two genes overlaps the / Aval interface (7,7). Transcription / translation studies carried out in vitro (Austin et al. 1990) show that only a single protein is found within the -4va7I / ßg7II fragment (44 kD). Although these authors were able to observe a further 30 kD polypeptide, they suggested that it could be a degradation product of the 44 kD protein. Our results allow the assumption that genes that code for at least two functional enzymes (functions encoded by rfaM and by rfaN) are arranged in this area. It is more likely that the 30 kD protein is a polypeptide encoded in this area than it is a degradation product. The combination of the results provides the conclusion that the 30 kD and 44 kD proteins are the gene products of rfaM and rfaN.

Identification of the rfaL enzyme function of E. coli K-12.

The plasmids pHB127, pHB128 and pHB130 were transformed into an rfaL-mutant strain of S. typh imurium, SL3749 (rfaL-), and the purified LPS was analyzed by SDS-PAGE. The results showed (data not given) that none of the plasmids complemented the rfaL mutants. Only the core / lipid A band could be observed after the silver staining.

Previous studies have shown that the arrangement of the rfa genes on the chromosome map of Salmon l la typhimurium is cysE- rfaF- (- rfaJ rfal rfaL) - rfaG-pyrE (Kuo and Stocker 1972; Sanderson and Saeed 1972a; 1972b). This suggested that in the event that the card arrangement in E. co 17-K-12 should be similar, the rfaL gene in the 5'-proximal region of the 3'-half of rfa (position 9.7 to 18 , 4) could be arranged.

A restriction enzyme analysis of the 3 '8.7 kB region of the rfa site had not yet been carried out, with the exception of the 2.037 kB long 3' end of the rfa region where the rfaD gene (which is for ADP -L-glycero-D-mannoheptose-6-epimerase) (Fig. 2; Pegues et al. 1990). Therefore, the W / nc / III / EcoRI fragment of the plasmid pHB115 was subcloned into the HindIII / EccRI interfaces of the plasmid pRK415, resulting in the plasmid pHB133, which carries the inserted DNA under the control of the external 7ac promoter . This was therefore necessary because it was not known whether the inserted DNA carries an internal promoter or not. However, the analysis of the plasmid pHB133 in rfaL mutant strains of S. typhimurium was not possible since the strain SL5283 (S. typhimurium, restriction negative) was not transformed with this plasmid could, which suggests that for some reason this region was lethal in S. typhimurium.

For the purpose of the further production of subclones in which the toxicity should no longer be present, a detailed restriction enzyme analysis of the plasmid pHB133 was carried out, and the resulting map (FIG. 2) was mixed with the known restriction enzyme. Interfaces in the rfaD region compared. It was determined that all the interfaces in this region were conserved in the plasmid pHB133, which further confirmed that clone no. 195 actually encompasses the rfa locus.

Three further subclones of the plasmid pHB133 were formed: (i) the WfncIII / EcoRV fragment, subcloned in HindIII / EcoRV interfaces of the plasmid pBR322 (plasmid pHB135), (ii) the plasmid pHB133, cleaved with CB13 and religated (pH713) and (iii) the plasmid pHB133, cut with Mlul and religated (plasmid pHB137). All three plasmids could be transformed into rfaL mutants of S. typhimurium, which suggests that the element causing the toxicity was removed. SDS-PAGE analysis of LPS profiles of the hybrid strains showed that the plasmid pHB137 complemented the rfaL defect in strain SL3749 (data not shown), while the plasmids pHB135 and pHB136 did not complement this defect. This suggests that the rfaL enzyme function of E. cσ7 / -K-12 is encoded by the C7al / M7ul fragment (positions 13.0 to 14.9; FIG. 2). However, it cannot be excluded that the coding area overlaps the Clal site. It also suggested that the M7 uI / Eco I fragment (positions 14.9 to 18.4; FIG. 2) is not necessary for the rfaL function. The plasmid pHB137 was therefore cleaved with EcoRI and Mlul, blunt-ended with the aid of the Klenow fragment of the DNA polymer and religated, the plasmid pHB139 being formed. The filled interface regenerates an EcoRI interface in this plasmid. To make it more precise to localize, a further set of subclones of the plasmid pHB139 was produced, the plasmid being cleaved with HindIII / BglII, blunt-ended with Klenow polymerase and religated. This gave pHB140.

JL's Immunological Electron Microscopy Analysis. typhi Ty21a strains that contain the O antigens of both ∑ _s . dysenteriae 1 as well as from J. typhi bound to the ∑. Express Typhi Core / Lipid A.

The plasmid pHB139, which expresses the rfaL gene product, was first transformed into the E. co77 strain S600, since this carries the helper plasmid for mobilizing plasmids based on the RK2 vector, for example pRK415. The plasmid pHB139 was transferred in a transfer by conjugation (mating) with the donor strain S600 / pHB139 into the strain H4098. The ex-conjugants were selected on Wilson Blair agar containing tetracycline.

Control and test strains (H4098 and H4098 / pHB139), grown in BHI medium with and without galactose, were first tested by agglutination on a slide with both O-antisera from S. typh i and S. dysenteriae 1, and the result showed that the strain H4098 / pHB139 agglutinated very strongly with both O-anti-sera, while the strain H4098 agglutinated only with the O-anti-serum from S. typhi.

Whole cells of the strains S. typhi Ty21a, E. cσ7 / -K-12 / pSS37, S. typhi Ty21a / pSS37, H-4098 and H4098 / pHB139 were identified by indirect immunogold electron microscopy with O-antiεeren of both S. dysenteriae 1 and also analyzed by S. typh i. The results showed that in the absence of galactose, all S. typh i strains reacted weakly with the S. typhi O antiserum (Fig. 5A.1, 5A.2, 5B1, 5B2), and in the presence of galactose, the cells were marked much more strongly (FIGS. 5A.3, 5A.7, 5B.3, 5B.7). This is to be expected since S. typh i Ty21a has a ga E mutation and consequently needs exogenous galactose to complete its LPS. In contrast, all S. typhi strains with and without galactose did not react with the S. dysenteriae O-antiserum, except for those carrying plasmid pHB139 (Fig. 5A.2, 5A.4, 5A.6, 5A .8, 5B.2, 5B.4, 5B.6 and 5B.8), which is also to be expected, since this expresses in the S. typhi Ty21a / pSS37 and H4098 strains (with rfb-rfp on the chromosome) Mated O-polysaccharide from S. dysenteriae is not bound to the nucleus and is consequently washed out during the preparation of the cells for immunoelectron microscopy. The positive control, the strain E. co7 / K-12, showed no reaction with the O antiserum from S. typh i (FIG. 5B.9), but reacted very strongly with the O antiserum from S. dysenteriae ( 5B.10), which is also to be expected, since in E. co7f-K-12 the S. dysenteriae 1 O-polysaccharide is bound to the nucleus. In the presence of galactose, test strain H4098 / pHB139 reacted strongly with both O-anti-sera (FIGS. 5B.7, 5B.8), and in a double-labeling experiment in which both O-anti-sera were combined with the test strain , the results showed that both O-antisera reacted strongly with the same bacterial cell (FIGS. 5C.2, 5C.3). These results make it highly probable that in the presence of the plasmid pHB139 strain H4098 was able to bind the O-antigen of S. dysenteriae 1 to the core / lipid A of the host, probably with the help of the rfaL gene product, namely the O-antigen core / Ligase.

Result

A variety of strategies are currently used to isolate Salmonella strains for use as carriers of foreign protection anti- to weaken genes, with the aim of eliciting a local immune response against the carrier and heterologous antigens by oral immunization with the hybrid strains (an overview can be found in Curtiss et al. 1989). Using this strategy, protection against human intestinal pathogens should be possible if attenuated S. typh i are used as carriers, since this strain colonizes the human intestinal tract and persists and persists in the intestinal lymphoid tissue (GALT). In fact, S. typh i Ty21a has been successfully used as a vaccine against typhoid fever (Wahden et al. 1982, Levine et al. 1987; Ferreccio et al. 1989; Levine et al. 1990a), and is the only carrier strain which has so far been identified by the FDA (Food and Drug Administration, USA) has been released for human use. Other weakened S. typt? / Strains have recently been designed and are currently being tested on human test subjects. The weakened mutations in these strains either inactivate the biosynthetic pathway of aromatic amino acids such as. Examples are arok, AroC or aroD (Edwarde and Stocker 1988; Dougan et al. 1988, Miller et al. 1989; Levine et al 1990b) or the cya, crp mutants (Curtiεε et al. 1987), which cyclic AMP and the CRP -Receptor protein for the expression of numerous genes, some of which code for determinants that play a role in pathogenicity. 2,3-Dihydroxybenzoate, para-aminobenzoic acid (auxotrophies resulting from aro mutations), cyclic AMP and the CRP protein are substrates which are not available to the bacteria in mammalian tissue, so that the bacteria cannot proliferate within mammalian cells. However, they survive and thrive long enough intracellularly to stimulate a protective immune response.

In an attempt to construct bivalent vaccines based on S. typhi carrier strains, the heterologous O-antigens, for example those of the intestinal pathogens S. dysenteriae 1, S. sonne i, S. flexneri, V. cho lerae or E CO7 _express.,. loading A major difficulty is that the heterologous O-polysaccharide is not bound to the core lipid A of S. typhi. Since it is present as a hapten, the O-antigen does not stimulate a protective immune response when the hybrid strains are used as oral vaccines. However, since the heterologous O-antigens bind to the core lipid A of E. co7 / -K-12, it was of interest to identify the nature of the modification of the S. typh i-Coreε that is required in order to to bind heterologous O-polyaccharides to the core lipid A of this host.

In order to define the precise modification of the Ty21a LPS core, which is necessary to bind the O-antigen of, for example, S. dysenteriae, the known structures of Salmon I la and E. co7 / -K-12 were compared. It appears that the only differences exist in the outer region of the nucleus, where the core terminus of K-12 carries GlcI-GlcII-GlcIII-GlcNAc, while the Sa one 17a core carries Glcl-GalII-GlcII-GlcNAc. In Salmon la, the Galll and GlcII substitutions are catalyzed by the rfal and rfaJ gene products, while the GlcII and GlcIII substitutions of the K-12 nucleus are catalyzed by the rfaM and rfaN gene products become. It can be assumed that the rfaL and rfPT gene products catalyze the linkage of the O-antigen to the nucleus.

The rfa locus of E. co 1 iK-12 was isolated from a gene bank of cosmid clones, which had been designed by Birkenbihl and colleagues (1989), and analyzed at the genetic level. The rfaG, rfaM, rfaN and rfaL genes were identified by cross-complementation, and the Salmonone l la typh imurium strains mutated in the corresponding genes of the rfa region were analyzed for the production of complete LPS by analyzing the K- 12 rfa analogs were transferred to the mutant strains. After the rfaL gene from K-12 had been transferred to S. typh i Ty21a, which carries the rfb-rfp gene cassette in the chromosome, it was found that the hybrid strain was the O-polyεaccharide from S. dysenteriae 1 linked to the core / lipid A of S. typhi. The carrier strain Ty21a also expressed homologous LPS. It therefore appears that it is not necessary to modify the S. typh i nucleus in any way, and that one possible reason why heterologous O-polysaccharides are not bound to the Ty21a nucleus is that the O-antigen / core ligase (the rfaL gene product) has a narrow, stringent specificity and therefore only the homologous O-antigen can bind to the core. On the other hand, the O-antigen / core ligases of E. cσ7 / -K-12 and S. typhimurium could show expanded, relaxed specificity and consequently be able to attach a large number of heterologous O-polysaccharides to the core structures of the host. These properties are of interest not only from the point of view of vaccine development, but also with regard to the evolution of the O-antigen / core ligase. Indeed, according to our results, there appears to be a relaxed specificity in these enzymes with regard to the cross-complementation of the outer core structures, since the substitution of the rfa function of the Salmon l la typh i murium mutatite by an analogue from E. co7 / -K-12 leads to the completion of the core structure, whereby essentially an enzyme from E. cσ7 / -K-12 is involved, while the rest comes from S. typhimurium. This is of interest from the evolutionary point of view, since it is well known that the carbohydrate chain of LPS is extremely heterogeneous. If the enzymes involved in the biosynthesis have relaxed specificities, then natural gene transfer from cross-reacting species would lead to the evolution of new serotypes.

Another problem that arises when constructing a bivalent S. dysenteriae 1 / S. typb / vaccine occurred was that the plasmid pSS37, which carries the rfb-rfp genes coding for the O-polysaccharides of S. dysenteriae 1, is unstable in S. typhi Ty21a. In the present case, this problem was solved by using a chromosomal integration vector system based on transposons to insert the rfb-rfp genes into the chromoson of Ty21a. animals. The _S elektions marker was a herbicide resistance gene encoding mercury resistance, the use of An ^¬ tibiotikumresistenz markers has been avoided, the yes zinstamm in vacuo ^¬ undesirable.

Literature:

Aronson BD, Ravinkar PD, Somerville RL (1988) Nucieotide sequence of the 2-amino-3-ketobutyrate coenzyme A ligase (ώl) gene of E. coli. Nucleic Acids Res. 16: 3586.

Attridge SR, Daniels D, Morona JK, Morona R (1990) Suπace co-expression of Vibrio cholerae and Salmonella typhi O-antigens on Ty21a clone EX210. Microbial Pat. 8: 177-188.

Austin EA, Graves JF, Hite LA, Parker CT, Sciinaitman CA (1990) Genetic analysis of lipopolysaccharide core biosynthesis by Escherichia coli K-12: Insertion mutagenesis of the rfa locus. J. Bacteriol. 17i 5312-5325.

Bachmann BJ (1990) Linkage map of Escherichia coli K-12, Edition 8. Microbiol Rev 54: 130-197.

Barcak GJ, Wolf RE Jr., (1988) Comparative nucleotide sequence analysis of growth-rate-regulated grtd alleles from natural isolates of Escherichia coli and from Salmonella typhimurium T-2. J. Bacteriol. 170: 372-379. Baron LS, Kopecko DJ, Formal SB, Seid R, Guerry P, Powell C (1987) Introduction of Shigella flexneri 2a type and group antigen genes into oral typhoid vaccine strain Salmonella typhi Ty21a. Infect. Immune. 55: 2797-2801.

Birkenbihl RP, Vielmetter W (1989) Cosmid-derived map of £. coli strain BHB2600 in co parison to the map of strain W3110. Nucleic Acids Res. 17: 5057-5068.

Boiteux S, O'Conner TR, Laval J (1987) For amidopyrimidine-DNA glycosyiase of Escherichia coli: cioning and sequencing of the fpg structural gene and overproduction of the protein. EMBO J 6: 3177-3183.

Bolivar F, Rodriguez RL, Betlach MC, Boyer HW (1977) Construction and characterization of new cioning vehicles. I. Ampicillin-resistant derivatives of plasmid pMB9. Genes ≥ 75-93.

Clarke L, Carbon J (1976) Cell 9: 91-94.

Creeger ES, Rothfield LI (1979) Cioning of genes for bacterial glycosyi transferases. I. Selection of hybrid plasmids carrying genes for two glycosyi transferases. J. Biol. Che. 254: 804-810.

Cryz SJ Jr., Furer E, Baron LS, Noon KF, Rubin FA, Kopecko DJ (1989) Construction and characterization of a Vi-positive variant of the Salmonella typhi live oral vaccine strain Ty21a. Infect. Immune. 57: 3863-3878.

Curtiss R III, Kelly SM (1987) Salmonella typhimurium deletion mutants lacking adenylate cyciase and cyclic AMP receptor protein are avirulent and immunogenic. Infect. Immune. 55: 3035-3043.

Curtiss R III. Kelly SM. Guiig PA. Nakayama K (1989) Selective deiivery of aniigens by recombinant bacteria. Curr. Top. Microbioi. Immunoi. 146: 35-49. Dorman CJ, Chatfield S, Higgins CF, Hayward C, Dougan G (1989) Characterization of porin and ompR mutants of a virulent strain of Salmonella typhimurium: ompR mutants are attenuated in vivo. Infect. Immune. 57: 2136-2140.

Dougan G, Chatfield S, Pickard D, Bester J, O'Callaghan D, Maskell D (1988) Construction and characterization of vaccine strains of Salmonella harbor mutations in two different aro genes. J. Infect. Dis. 158: 1329-1335.

Dupont HL, Hornick RB, Dawkins AT, Snyder MJ, Formal SB (1969) The response of man to virulent Shigella flexneή 2a. J. Infect. Dis. 119: 296-299.

Edwards MF, Stocker BAD (1988) Construction of del aroA his del pur strains of Salmonella typhi. J. Bact. 170: 3991-3995.

Ferreccio C, Levine MM, Rodriguez H, Contreras R, Cliilean Typhoid Committee (1989) Comparative efficacy of two, three or four doses of Ty21a live oral typhoid vaccine in enteric-coated capsules. A field triai in an endemic area. J. Infect. Dis. 159: 766-769.

Fields PI, Groisman EA, Heffron F (1989) A Salmonella locus that controls resistancs to microbicidal proteins from phagocytic cεils. Science 243: 1059-1062.

Formal SB, Kent TH, May HC, Paimer A, Falkow S, LaBrec EH (1966) Protection of monkeys against experimental shigellosis with a living attenuated oral polyvalent dysentery vaccine. J. Bacteriol. 92: 17-22.

Formal SB, Baron LS, Kopecko DJ, Washington O, Powell C, Life CA (1981) Construction of a potential bivalent vaccine strain: Introduction of Shigella sonnei Form I antigen genes into the galE Salmonella typhi Ty21a typhoid vaccine strain. Infect. Immune. 34: 746-750. Formal SB, Levine MM (1984) Shigeilosis. In Bacterial Vaccines. Germanier R. (ed.) London. Academic Press, pp. 167-186.

Germanier R, Furer E (1975) Isolation and characterization of GalE mutant Ty21a of Salmonella typhi: a candidate strain for a live, oral typhoid vaccine. J. Infect. Dis. 131: 553-558.

Hanahan D (1983) Studies on transformation of Escherichia coli with piasmids. J. Mol. Biol. 166: 557-580.

Herrero M, Lorenzo VD, Timmis KN (1990) Transposon vectors containing non-antibiotic resistance selection markers for cioning and stable chromosomai insertion of foreign genes in gram-negative bacteria. J. Bacteriol. 172: 6557-6567.

Hitchcock PJ, Brown TM (1983) Morphological heterogenicity among Salmonella lipopolysaccharide chemotypes in silver stained polyacryla ide gels. J. Bacteriol. 154: 269-277.

Kadam SK, Rehemtulla A, Sanderson KE (1985) Goning of rfcG, B, I, J genes for glycosyltransferase enzymes for synthesis of the lipopolysaccharide core of Salmonella typhimurium. J. Bacteriol. 161: 277-284.

Keen NT, Ta aki S, Kobayashi D, Trollinger D (1988) Improved broad-host-range piasmids for DNA cioning in gram-negative bacteria. Gene 70: 191-197.

Kuo TT, Stocker BAD (1972) Mapping of rfa genes in Salmonella typhimurium by ES 18 and P22 transduction and by conjugation. J. Bacteriol. 112: 48-57.

Laemmli UK (1970) Cleavage of structurai proteins during the assembly of the head of bacteriophage T4. Nature (London) 227: 680-685. Levine MM, Black RE, Ferreccio C, Germanier R, Chilean Typhoid Committee (1987) Large-scale field trial of Ty21a live oral typhoid vaccine in enteric-coated capsule formation. Lancet 1: 1049-1052.

Levine MM, Kaper JB, Herrington D, Losonsky G, Morris JG, Clements ML, Black RE, Tall B, Hall R (1988a) Volunteer studies of deletion mutants of Vibrio cholerae 01 prepared by recombinant techniques. Infect. Immune. 56: 161.

Levine MM, Kaper JB, Herrington DA, Ketley J, Losonsky G, Tacket CO, Tall B, Cryz SJ (1988b) Safety, immunogenicity, and efficacy of recombinant live oral cholera vaccines, CVD 103 and CVD 103-HgR. Lancet Ifc 467.

Levine MM, Ferreccio C, Cryz S, Ortiz S (1990a) Comparison of enteric-coated capsules and liquid formulation of Ty21a typhoid vaccine in a randomized controlled field trial. Lancet (in press).

Levine MM, Hone D, Tacket C, Ferreccio C, Cryz S (1990b) Clinicai and field trials with attenuated Salmonella typhi as live oral vaccines and as "carrier" vaccines. Res. Micro biol. 141: 807-816.

Macpherson DF, Morona R, Beger DW, Cheah KC, Manning PA (1991) Genetic analysis of the rfb region of Shigella flexnert encoding the Y serotype O- antigen specificity. Molec. Microbiol. 5: 1491-1499.

Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cioning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

Manoil C, Beckwith J (1985) TnphoA: a transposon probe for protein export signals. Proc. Natl. Acad. Be. UNITED STATES. 82: 8129-8133. Mel DM, Terzin AL, Vuksic L (1965) Studies on vaccination against bacillary dysentery. 3. Effective oral immunization against Shigella flexneri 2a in a field trial. Bull, of World Health Organization 32: 647-655.

Miller JH (1972) Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

Miller VL, Mekaianos JJ (1988) A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and viruience determinants in Vibrio cholerae requires toxR. J. Bacteriol. 170: 2575-2583.

Miller IA, Chatfield S, Dougan G, Desilva L, Joysey HS, Hormaeche C (1989) Bacteriophage P22 as a vehicle for tranducing cosmid gene banks between smooth strains of Salmonella typhimurium: use in identifying a role for aroD in attenuating virulent Salmonella strains. Mol. Gen Genetics 215: 312-316.

Mills SD, Sekizaki T, Gonzaiez-Carrero MI, Timmis KN (1988) Analysis and genetic manipulation of Shigella viruience determinants for vaccine development. Vaccine 6: 116-122.

Nasoff MS, Wolf RE Jr, (1980) Molecular cioning, correlation of genetic and restriction maps, and determination of the direction of transcription of gnd of Escherichia coli. J. Bacteriol. 143: 731-741.

Osborn MJ (1968) Nature (London) 217: 1-8.

Nasoff MS, Baker HV, Wolf RE Jr (1984) DNA sequence of the Escherichia coli gene, gnd, for 6-phosphogluconate dehydrogenase. Gene 21: 253-264.

Pegues JC, Chen L, Gordon AW, Ding L, Coleman Jr. WG (1990) Cioning, expression and characterization of the Escherichia coli K-12 rfaD gene. J. Bacteriol. 172: 4652-4660. Reeves P, Stevenson G (1989) Cioning and nucieotide sequence of the Salmonella typhimurium LT-2 gnd gene and its homoiogy with the corresponding sequence of Escherichia coli K-12. 217: 182-184.

Roantree RJ, Kuo TT, MacPhee DG (1977) The effect of defined lipopolysaccharide core defects upon antibiotic resistances of Salmonella typhimurium. J. Gen. Microbiol. 103: 223-234.

Sanderson KE, Saeed YA (1972a) P22-mediated transduction analysis of the rough A (rfa) region of the chromosome of Salmonella typhimurium. J. Bacteriol. 112: 58-63.

Sanderson KE, Saeed YA (1972b) Insertion of the F factor into the cluster of rfa (rough A) genes of Salmonella typhimurium. J. Bacteriol. 112: 64-73.

Sansonetti PJ, Haie TL, Dammin GJ, Kapfer C, CoUins HH Jr, Formal SB (1983) Alterations in the pathogenicity of Escherichia coli K-12 after transfer of plasmid and chromosomal genes from Shigella flexneri. Infect. Immune. 39: 1392-1402.

Sturm S, Fortnagel P, Timmis KN (1984) Immunoblotting procedure for the analysis of electrophoretically fractionated bacterial lipopolysaccharides. Arch. Microbiol. 140: 198-201.

Sturm S, Timmis KN (1986) Cioning of the rfb gene region of Shigella dysenteriae 1 and construction of an rfb-rfp gene cassette for the deveiopment of lipopolysaccharide-based live anti-dysentery vaccines. Microbial path. 1: 289-297.

Sturm S, Jann B, Jann K, Fortnagel P, Timmis KN (1986) Genetic and biochemical analysis of Shigella dysenteriae 1 O antigen poiysaccharide biosvnthesis in Escherichia coli K-12 9 kb plasmid of S. dysenteriae 1 determines addition of a galactose residue to the lipopolysaccharide core. Microb. Pathog. 1: 299-306. Subbaiah TV, Stocker BAD (1964) Rough mutants of Salmonella typhimurium. I. Genetics. xNature (London) 201: 1298-1299.

Tacket CO, Forrest B, Morona R, Attridge SR, Labrooy J, Tall BD, Reymann M, Rowley D, Levine MM (1990) Safety, immunogeniciiy, and efficacy against Cholera challenge in humans of a Typhoid-Cholera hybrid vaccine derived from Salmonella typhi Ty21a. Infect. Immune. 58: 1620-1627.

Tsai CM, Frasch CF (1982) A sensitive silver stain for detecting lipopoiysaccharides in polyaccrylaxnide gels. Anal. Biochem. 119: 115-119.

Wahden MH, Serie C, Cerisier S, Sallam S, Germanier R (1982) A conirolled field trial of live Salmonella typhi strain Ty21a oral vaccine against typhoid three-year results. J. Infect. Dis. 145: 292-295.

Westphal O, Jann K (1965) Bacterial lipopoiysaccharides. Extraction with phenol-water and further applications of the procedure. Methods Carbohydr. Chem. 5: 83-9 l. Y

Yanisch- Perron C, Vieira J, Messing J (1985) Improved M13 phage cioning vectors and host strains: nucieotide sequences of the M13mpl8 and pUC19 vectors. Gene 33: 103-119.

Yoshida Y, Okamura N, Kato J, Watanabe H (1991) Molecular cioning and characterization of form I antigen genes of Shigella sonnei. J. Gen. Microbiol. 137: 867-874. Figure description.

1 is a schematic representation of the plasmid construction for the integration of the rfb-rfp cassette into the chromosome of S. typh i Ty21a. For some plasmids there is a linear representation which shows the relevant segment of the inserted DNA.

2 shows a physical map of the E. co7 / -K-12 rfa gene location and plasmid constructions within this region. The top row shows the chromosome map of E. co 1 / -K-12 in the area of the 82nd minute (part of which has already been shown by Austin et al., 1990). The rfa locus is from

(which codes for the 2-amino-3-ketobutyrate coenzyme A ligase; Aronson et al. 1988) and the fpg gene (which codes for the formamidopyrimidine DNA glycosylase; Boiteux et al. 1987) flanked. The location of the fpg gene is not shown in the eighth edition of the E. co7 / -K-12 linkage map (Bachmann 1990); However, if one evaluates the mapping and DNA sequencing data from Boiteux and colleagues (1987), one can clearly see that the fpg gene is the ßa / nHI cleavage site in the δ'-proximal region of the ε rfa -Genorteε spanned, as shown above. The evaluation of the DNA sequence of the fpg gene and the assumption that the 5 'end of rfa starts directly in the 3' direction next to the fpg gene should suggest that the 5 'end of rfa should be at least 300 Bp to the left of the ßa / 7 * .I interface (position 0). The restriction interfaces are the following: (A) Apal, (Ac) Accl, (Av) Aval, (B) ßstEII, (Bg) ßg7II, (Bm) BamHI, (C) C7al, (E) EcoRI, (EV ) EcoRV, (H) Hindlll, (Hc) Hindi, (Hp) Hpal, (M) Mlul, (N) Ncol, (P) PstI, (Pv) PvuII, (S) Sa71, (Sc) Seal, (X ) Xbal, (Xh) Xhol. The positions (values in kB) of the essential restriction enzyme interfaces (referred to in the text) are given below the respective enzymes, the ßamHI interface in the fpg gene with the position "0" and the EcoRI interface between tdh and rfa with 18.4. The restriction enzyme cleavage sites shown between ßamHI (0) and BglII (1,1) were obtained from Austin and Kolleg (1990). No restriction enzyme cleavage sites for BarriΛI, Kpnl, Ncol, PstI, Sacl, Sa71 and Xbal were found in the region between Hindlll (9.7) and EcoRI (18.4). The enzymes AccI, Aval, ßstEII, HincII, Ncol and PvuII, which were used in the previously published map (Austin et al. 1990), were not used for mapping interfaces between Hindlll (9,7) and EcoRI (18.4) was used. The Apal, Clal, EcoRV, Seal and Xhol enzymes have not been mapped in the region between BarriM (0) and Bglll (11.1) and there are no Mlul sites in the same region. The presence or absence of the rfaG, I, J, and L enzymes (labeled + and -, respectively) that are encoded by the different clones is shown.

3 shows the chemical structures of the LPS nuclei of E. cσ7 / -K-12 and Salmon l la typhimurium together with the respective rfa genes which code for the enzymes which the nucleus on the respective step -Biosyntheεe are involved. O-antigen is ligated to the nucleus using two enzymes, one of which is encoded by rfaL and the other by an rfb gene, rfp-T. Dashed bonds represent incomplete substitutions.

Abbreviations: KDO: 3-Deεoxy-D-manno-2-octulosonic acid; P: phosphate; Etn: ethanolamine; Hep: heptose; Glcu: glucose; GlcNAc: N-acetylglucosamine; Gal: galactose.

FIG. 4 shows complementation studies of E. co7 / -K-12 rfa clones, expressed in defined S. typhimurium rfa mutants.

4.1, 4.2 and 4.3 are SDS-PAGE gels stained with silver and represent LPS profiles. Fig. 4.1. Lane (A) SL3769 (rfaG-); (B) SL3769 / pHB130; (C)

SL3769 / pSS37; (D) SL3769 / pHB130 + pSS37.

Fig. 4.2: Lane (E) SL 3748 (rfal-); (F) SL3748 / pHB130; (G)

SL3748 / pHB127; (H) SL3748 / pSS37; (I) SL3748 / pHB130 + pSS37; (J)

SL3748 / pHB127 + pSS37.

Fig. 4.3: Lane (K) SL3750 (rfaJ-); (L) SL3750 / pHB130; (M)

SL3750 / pHB127; (N) SL3750 / pSS37; (0) SL3750 / pHB130 + pSS37; (P)

SL3750 (pHB127 + pSS37.

5 shows an immunogold marking of whole bacterial cells, which shows the binding of the O-polysaccharide from S. dysenteriae to the strain Salmon I la typhi Ty21a.

5A: The upper and the lower horizontal row show bacteria from the strain S. typhi Ty21a and S. typhi Ty21a / pSS37, respectively. The numbers 1, 3, 5 and 7 are marked with S. typhi O antiserum, and the numbers 2, 4, 6 and 8 are marked with S. dysenteriae 1 O antiserum. The numbers 1, 2, 5 and 6 were grown without galactose, the numbers 3, 4, 7 and 8 in the presence of galactose.

Fig. 5B: The top two horizontal rows show the strains H4098 and H4098 / pHB139, and the bacteria are labeled with O-antisera from either S. typhi or S. dysenteriae, arranged in each case perpendicular to one another. The numbers 1, 2, 5 and 6 were grown without galactose, the numbers 3, 4, 7 and 8 in the presence of galactose. Numbers 9 and 10 show the positive control strain E. co7 / -K-12 / pSS37, labeled with O antisera from S. typhi and S. dysenteriae, respectively.

5C: The strain H4098 / pHB139 is shown here. Number 1 shows bacteria which had been grown without galactose and labeled twice with O-anti serum from S. typhi (10 nm gold particles) and O-anti serum from S. dysenteriae (5 nm gold particles). The bacteria of number 2 were in the presence of Ga lactose and duplicate marked as described above. The O-antiserum from S. dysenteriae was used first, followed by the O-antiserum from S. typhi.

Number 3 differs from number 2 only in that the order of the antisera has been reversed.

Living material

S. typhi H4098 S. typhi Ty21a ATCC 33 459

E. coli K12 for sale E. coli S600 I buy E. coli SMpir

pHB 139 as H4098 pSS 37 pLOF / Ars DSM 7071 pHB 120

rfaL gene see pHB 139

Claims

1. Bivalent live vaccine against bacterial intestinal pathogens, characterized in that the microorganism contained therein expresses complete, bivalent lipopolysaccharide (LPS) with a core lipid A of the microorganism and heterologous O-polysaccharides.

2. Bivalent live vaccines according to claim 1, characterized in that εie is a S. typh / live vaccine.

3. Bivalent live vaccine according to claim 1 or 2, characterized in that it is a S. typhi, in particular a S. ty¬ ph i Ty21a (galE) live vaccine or a Vibrio cho lerae- live vaccine.

4. Bivalent S. typh / live vaccine according to one of the preceding claims, characterized in that the heterologous O-polyaccharides are those of the species S. dysenteriae 1, Sh ige l la sonnei, Sh ige l la f lexneri, Shige l la boydi i, Vibrio cho lerae or Salmon l la paratyphi ε are or from a gut-pathogenic Escherichia co7 / strain.

5. Bivalent live vaccine according to claim 4, characterized gekenn¬ characterized in that the heterologous O-Polyεaccharide those of S. dysenteriae 1 εind.

6. Bivalent live vaccine according to one of claims 1 to 5, characterized in that εie is produced from the S. t pΛZ sta m H409S.

7. Bivalent S. typh / live vaccine according to one of claims 1 to 6, characterized in that the genetic material of the vaccine carries the rfaL gene from E. co7 / K-12 or a gene hybridizable therewith.

8. Divalent live vaccine according to any one of claims 1 to 7, da¬ by in that the gene hybridizable with the gene rfaL- at a concentration of 1 M NaCl at a temperature of 30, preferably 40 and particularly 50 ^* C is hybridizable.

9. Bivalent live vaccine according to claim 7 or 8, gekennzeich¬ net by the rfaL gene of the plasmid pHB139.

10. Bivalent live vaccine according to claim 8 or 9, characterized in that it is or contains H4098 / pHB139.

11. A method for producing a vaccine according to one of the preceding claims, characterized in that

(a) the rfaL gene from E. col i K-12 or the gene hybridizable with the rfaL gene is transferred into the microorganism of the live vaccine and

(b) the gene coding for the O-antigen biosynthesis functions of the species whose O-polysaccharides are to be expressed is transferred into the microorganism of the live vaccine.

12. The method according to claim 11, characterized in that for stage (a) a cosmid clone is selected from a gene bank of E. co7 / K-12, which bears the rfaL locus, generates a plasmid with this locus by partial digestion and religion and is cloned if necessary, this plasmid is cut once or several times into a further plasmid containing this locus, religated and subcloned in order to remove any toxicities, and the plasmid is introduced into the microorganism of the live vaccine.

13. The method according to claim 12, characterized in that the microorganism of the live vaccine is S-typhi.

14. The method according to claim 12 or 13, characterized in that the plasmid is transformed into E. co7 / S600 and then transferred by conjugative mating into the microorganism of the live vaccine.

15. The method according to any one of claims 12 to 14, characterized ge indicates that the plasmid pHB139 is introduced into the microorganism of the live vaccine.

16. The method according to claim 11, characterized in that for stage (b) the gene locus coding for the O-antigen biosynthetic functions is transferred from S. dysenteriae to S. typhi.

17. The method according to claim 16, characterized in that this gene locus is the (for example originating from pSS37) rfb-rfp cassette.

18. The method according to claim 17, characterized in that the rfb-rfp cassette is inserted in S. typh i Ty21a with the aid of a chro osomal integration vector system based on transposons.

19. The method according to claim 18, characterized in that with the aid of the vector pLOF / Ars a rfb-rfp cassette carrying Plasmid is formed as a chromosomal integration vector system.

20. The method according to claim 19, characterized in that the transposon-based integration vector system is pHB120.

21. The method according to claim 20, characterized in that pHB120 is transformed into SMIOpir and transferred to S. typhi Ty21a by conjugative mating.

22. pHB120

23. pHB139

24. Strain H4098

25. Strain H4098 / pHB139