WO2001094599A1 - Gene expression cassette and its use - Google Patents

Abstract

A gene expression cassette comprising a secretory leader sequence encoding a signal peptide from Clostridium difficile having an amino acid sequence selected from SEQ ID NO: 1-12 and signal peptides of analogous exported clostridial N-acetylmuramoyl-L-alanine amidase-like proteins, linked to a DNA sequence encoding a heterologous polypeptide, and optionally a part of SEQ ID NO: 22-233, is described. The gene expression cassette is inserted into a vector, and the vector is used to transformed a host organism. Compositions, formulations, vaccines and medicaments based on spores of such engineered host organisms are used e.g. for colonization of a mammal.

Description

Gene expression cassette and its use

The present invention relates to a gene expression cassette and in particular to the use of the cassette in methods for presenting polypeptides on the surface of bacterial cells and/or secreting them into the surroundings of the latter. The invention further relates to gene expression constructs that are used to transform bacterial host cells. Uses of the invention include immunisation, in particular mucosal immunisation, induction of immunological tolerance and anti-tumour therapy in humans and animals. The intended vaccines and anti- cancer agents will also make use of bacterial spores produced by clostridia, e.g. Clostridium difficile, for both industrial production of the vaccine and for local production of the desired polypeptides at the body sites desired.

Vaccines against infection represent the greatest advance in medicine with unparalleled impact on morbidity and mortality at relatively low cost. Despite their cost- effectiveness, the cost associated with modern vaccines is still of concern and limits their use, particularly in developing countries. A number of factors contribute to the cost of injectable vaccines including the requirements for vaccines with defined sub-cellular components, for purity and sterility of the vaccine preparations, for testing of administration routes and combinations with different adjuvants, for maintaining the cold chain in distribution of the vaccines, and for using sterile syringes and needles. The need for repeated vaccinations also contributes to increased costs. Furthermore, for many infectious diseases a vaccine has not yet been made available

Also, in recent years great interest has been shown in mucosal immunisation, i.e. the exposure of mucosal surfaces to an antigen to elicit a general humoral and mucosal immune response, i.e. also at distant sites (Mucosal Immunology, Ed. P.L.Ogra et al., Acad. Press,

1999). Although this field is in its infancy, two such products are so far on the market, namely the oral polio vaccine and an oral (drinkable) vaccine against cholera and diarrhea due to Escherichia coli. The latter is an inactivated vaccine containing killed Vibrio cholerae organisms plus the cholera toxin subunit B which is non-toxic, immunogenic and shared between the cholera toxin and the toxin of "enterotoxigenic E. coli" (ETEC), the main cause of "travellers diarrhea". It is hoped that mucosal immunisation via for example intranasal or peroral administration of the antigen will provide a real alternative to injectable vaccines. As most microbes invade humans and animals via mucous membranes we anticipate that the mucosal route will turn out to be the superior alternative for vaccination in many instances. The scientific consensus at this point appears to be that live vaccines are potentially superior candidates for single-dose long lasting vaccination, because the carrier organism will continue to produce the antigen and boost immunity in vivo. However, experience with the bacterial carriers currently studied, for example the intestinal bacteria Salmonella and E. coli is that they are not generally classified as safe and are further difficult to distribute. Therefore, probiotic organisms such as Lactobacilli have been proposed as suitable carriers of foreign antigens. Problems have been identified, these include limited shelf-life unless freeze-dried which reduces viability, and difficulties in colonising the gut of the human recipient and evoking an immune response.

A classical way of enhancing the immunogenicity of vaccines is co-administration of so-called adjuvants. These may be molecules such as aluminium hydroxide or lipid vesicles that increase the exposure time for the vaccine by slowing its removal from the site of injection or "danger molecules" of microbial origin that increase the immune response in a non-specific way. Thus, in recent years it has been found that adjuvants also act by evoking production of immunomodulatory peptides called cytokines and chemokines (Brewer JM, Alexander J, Cytokines Cell Mol Ther 4:233-246, 1997. Ulanova M, Classical and non- classical antigen-presenting molecules in immune responsiveness, Thesis, Gδteborg University, Sweden, 2000, ISBN 91-628-4228-5). Due to possible adverse effects the use of cytokines themselves as adjuvants must await a better understanding of optimal selection and dosing of these molecules together with vaccines. It is likely that applying cytokine adjuvants by the mucosal route is less critical than parenteral administration from a side effect point of view. Interest has also been shown in the extracellular presentation of foreign protein antigens by bacteria, included as part of bacterial surface layer proteins. A surface layer (S- layer) protein is herein defined as any molecule of proteinaceous nature, including e.g. protein, glyco- or lipoprotein occurring in the outer layer of a bacterium and capable of being exposed on the surface of the bacterium. S-layer proteins are a main constituent of the cell wall of some gram-positive bacterial genera. They may be continuously and spontaneously produced in larger amounts than any other class of protein in the cell. WO-95/19371 describes a fusion protein of at least a part of a S- layer protein and a heterologous peptide, the intention being that the polypeptide is expressed and presented on the surface of the cell. A range of bacterial hosts is mentioned including Staphylococcus, Streptococcus, Bacillus, Clostridium and Listeriα. A preference for Bacillus is stated and the examples use B. sphaericus. Elsewhere, WO-97/28263 describes processes for the recombinant preparation of S-layer proteins in gram-negative host cells. It is suggested that these proteins could include antigenic species. FR-A-2778922 describes the use of genes which regulate the synthesis of toxin products in Clostridium bacteria, to produce polypeptides.

We first set out to investigate toxin release and to identify other extracellular proteins produced by Clostridium difficile (C. difficile). C. difficile is an anaerobic spore forming pathogen causing C. difficile associated diarrhea (CD AD) and pseudomembraneous colitis (PMC) by producing two toxins, A and B .

In contrast to studies by Kamiya et al (J.Med.Microbiol.,1992, 37, 206-210) and Ketley et al (J.Med.Microbiol.,1984, 18, 385-391) we found that the accumulation of extracellular toxins is not accompanied by cell lysis suggesting a toxin export mechanism. Our first detailed analysis (described in Example 9 below) of proteins occurring extracellularly as well as in the cell wall and membrane fraction of strains VPI 10643 and 630 and analysis of the DNA/genes encoding these proteins revealed a genomic segment containing eighteen genes (Figure 1A). Seven proteins (ORFl,3,5-7,9 and 11), when compared with publicly available sequence showed some homology to N-acetyl muramoyl L- alanine amidase (CwlB/LytC) and modifier protein of major autolysin (LytB) from B. subtilis, as well as S- layer proteins from Lactobacillus spp (Tables 1 and 4). The amidase motif was located either at the C-terminal or the N-terminal end of the S-layer protein ORFs (see examples in Figure IB). Other ORFs showed similarity to genes involved in polypeptide secretion (ORF2/secA), polysaccharide and capsule synthesis, and possibly glucosylation of the S-layer proteins. Further database searches indicated that the amidase motif confers anchorage of the S-layer proteins to the clostridial cell wall peptidoglycan-teichoic acid.

A search in the revised C. difficile database revealed five additional genes upstream of ORF1 which had similarities to the previously found ones, i.e. they had a two-domain architecture one showing homology to the CwlB/LytC and LytB proteins. These ORFs thus had the putative cell wall binding amidase motif typical of the other S-layer ORFs and were designated D, E, G, H and I by us (Fig. 2 and Table 1).

Significantly, we found that the N-termini of all S-layer ORFs contained a typical signal peptide for export via .fee-dependent secretion and that for e.g. ORF1 the predicted signal peptide cleavage site (Figure 3) was identical to that found in the protein sequence . Furthermore, the secreted ORFl product was further cleaved into two peptides in strain 630, the C-terminal one containing the N-acetyl muramoyl L-alanine amidase like sequence (Figure IB and Example 9). Although we identified the cleavage point, the kinetics and precise mechanism of this proteolytic event remains unknown. It is likely that the S-layer protein cleavage product containing the amidase motif provides cell- wall binding, whereas the other peptide showing more sequence variability is more surface exposed and providing antigen variation between strains (serotypes). We found no significant match between sequences within ORFl and the S-layer homology motif (SLH domain) found in most presently known S-layer proteins although a weak similarity to an S-layer protein from Lactobacillus helveticus was found for the N-terminal part of ORFl (Figure IB). Also, apart from the shared amidase motif, we found no significant homology beetween the different S- layer ORFs that could suggest how the variable cleavage product(s) get anchored to the cell- wall binding peptide (inner S-layer) or to each other to form the outermost layer recently distinguished by electron microscopy (Cerquetti M et al. Characterization of surface layer proteins from different C. difficile clinical isolates. Microb Pathogenesis 28: 363-372, 2000). Our subsequent analysis of 21 C. difficile serogroup type strains, probably representing all major genetic lineages of the species, indicated that a gene segment corresponding to ORFl and its upstream region plus ORF 2 is generally present in C. difficile (Example 1C). The N-terminal part of ORF6 showed homology to eukaryotic cysteine proteases (Fig.

IB). ORF5 has been suggested to be involved in adhesion to epithelial cells (Abstract; The Third International Meeting on the Molecular Genetics and Pathogenesis of the Clostridia, June 8-11, 2000, Chiba, Japan).

The present invention is based, at least in part, on the above discoveries. We have identified and developed a polypeptide expression and secretion system that may be used to produce a desired polypeptide on the surface of and/or into the surroundings of bacteria, for introduction into an appropriate mammal. The system may be used for example to initiate mucosal vaccination. A particular advantage of the system is that it may be used with any convenient Clostridium species, independently of any normal S-layer protein production. Furthermore, in case of C. difficile it is possible to use strains lacking the 5-gene toxicity cassette encoding the two major virulence factors toxins A and B and thus avoid the risk of CD AD when administering the engineered peptide producing strains to humans aged 2-4 years or more (neonates and small children are insensitive to the toxins) or animals. Therefore, in a first aspect of the invention we provide a gene expression cassette comprising a secretory leader sequence selected from any one of ORFl, ORF3, ORF5-7, ORF9 or ORFl 1 (SEQ JX> NO: 1 -7) (cf. Figure 1 and Table 1) of C. difficile strain 630 linked to a DNA sequence encoding a heterologous polypeptide. Alternatively, the secretory leader sequence is from any one of ORF D, E, G, H and I (SEQ ID NO: 8 - 12) (cf. Figure 2 and Table 1) or from any analogous S-layer ORF taken from any C. difficile strain.

By "heterologous" we mean a nucleic acid sequence or protein not native to the clostridial strain being used.

Use of each of the secretory leader sequences mentioned above represents a separate and independent aspect of the invention. The secretory leader sequence is preferably from ORFl.

A recent publication by Karjalainen et al, Infection and Immunity, May 2001, p3442- 3446 provides confirmation and analysis of most of the ORFD - ORFl - ORFl - ORF11 S- layer gene cluster in C. difficile strain 630. The nucleotide and polypeptide sequences disclosed by Karj alainen et al are incorporated herein by reference.

In a further aspect of the invention the gene expression cassette further includes a promoter of prokaryotic origin. The promoter is preferably a strong promoter and in general is placed 5' of the secretory leader sequence in the gene cassette.

In a further aspect of the invention the gene expression cassette further includes a DNA sequence encoding at least a functional portion of an S-layer protein of C. difficile fused to a nucleic acid coding sequence coding for a heterologous polypeptide such that the resulting fusion polypeptide will be expressed and presented on the outer surface of the host cell harbouring the cassette. If desired the polypeptide can also be released from the bacteria, e g. by excluding the S-layer amidase motif from the construct (cf. Figure. 3). In a further aspect of the invention the engineered gene expression cassette optionally further comprises at least a functional part of the secretory (secA) gene represented by ORF2. This may be used to complement or replace the function of the normal C. difficile sec gene in order to ensure efficient translocation of the peptide(s) produced by the cassette across the cytoplasmic membrane. An example of a preferred gene expression cassette is conveniently illustrated in

Figure 3.

The promoter in the gene expression cassette is conveniently a strong promoter, this may be the native promoter for ORFs 1 - 12 of C. difficile of strain 630 (Table 1). Alternatively, the promoter sequence is from any one of ORF D - 1 (cf. Table 1), alternatively from any other analogous S-layer ORF from a C. difficile strain or from another gene, preferably from this species (see Specific description Bl). The promoter may thus be another prokaryotic promoter that is strong, inducible or constitutive, and functional in the polypeptide producing bacterium. In all potential applications a distinct advantage of this cassette is the very large amounts of protein produced and exported.

The gene expression cassette is conveniently placed in a vector or specifically a plasmid carrying a transposon belonging to for example the Tn916, Tn5387 or the Tn5398 families. After transfection of a C. difficile host organism these transposons are able to insert themselves into its chromosome thereby making the engineered cassette a stable trait of the bacterium (cf. Figure 4). For other Clostridia other vectors may be preferable, e.g. the engineered shuttle plasmid pJTR750. Unlike the C. difficile plasmids currently available, this vector can replicate within both an E. coli and a C. perfringens host and is not dependent on integration of the plasmid into the host chromosome. Any convenient Clostridium species may be used, to date over 70 species have been defined by rRNA sequence analysis. These include C. difficile and classical pathogens as C. perfringens, C. tetani and C. botulinum, also C. acetobutylicum that is being genetically manipulated and used for industrial production of acetic acid and C. beijerinckii that has been transformed with E. coli genes.

C. perfringens is currently the species most amenable to genetic engineering. It is a normal, moderate level, fecal coloniser of most, if not all, humans. C. difficile is found in the fecal flora of most newborns, less often in adults but commonly in hospitalized individuals. As C. difficile is an early, normally colonising intestinal organism and even toxigenic strains are unable to cause CD AD in newborns and infants up to 2-4 years of age, we believe that recombinant C. difficile producing desired antigens and adjuvants is suitable for oral vaccination at any convenient time after birth.

Whereas C. perfringens normally produces many toxins about half of wild C. difficile strains are genetically non-toxigenic, which may be an advantage from a safety point of view. C. difficile toxin negative strains are preferred as host cells for the gene expression cassettes of this invention, at least for individuals aged 2-4 years or more (see above). The nucleic acid sequence coding for a heterologous polypeptide is placed in the gene expression cassette before or after insertion into a convenient vector or plasmid. The insertion points for the nucleic acid sequence are at the discretion of the skilled scientist, there may be in the variable or in the constant region of the relevant ORF nucleotide sequence. Routine experimentation may be used to determine convenient and particular insertion points. In Figure 3 we disclose polypeptide cleavage sites that need to be taken into consideration (See Specific description B3).

Examples of convenient plasmids include those mentioned in Figure 4, for example pCI195 and pSMB47. Convenient transposons include those belonging to the Tn916, Tn 5397 and Tn5398 families for transfection into C. difficile and for example pJTR750 for C. perfingens or other Clostridia. Any convenient heterologous nucleic acid sequence may be placed into the gene expression cassette. In a further aspect of the invention we provide a vector or plasmid comprising a gene cassette of the invention. The vector or plasmid may then be transfected into a convenient host using techniques known in the art (see for example: Gene 82: 327-333, 1989). For C. difficile it is at present preferred to introduce the plasmid into a Bacillus species such as B. subtilis and then transfer the target DNA by filter mating (conjugation) into a convenient C. difficile strain (outlined in Fig. 4). This will generally require the use of a conjugative transposon-bearing plasmid such as pCI195 or pSMB47 (J. Antimicrob. Chemother. 35: 305-315, 1995; FEMS Microbiol. Lett., 168: 259-268, 1998; D. Lyras, J. I. Rood, Clostridial genetics, in Gram-positive pathogens, ed. V. A. Fischetti, Am. Soc. Microbiol, 2000). However, we anticipate that further materials and procedures will become available for the direct introduction of plasmids or other foreign DNAs into Clostridia and particularly C. difficile. For e.g. C. perfringens such vectors and techniques are to some extent already available.

In a further aspect of the invention we provide a Clostridial bacterium transformed with a gene expression cassette of the invention encoding the desired fusion peptide or entirely heterologous polypeptide(s).

The transformed Clostridial bacterium, when administered orally to any convenient mammal such as a human or animal will lead to the intestinal colonization, production and presentation of the desired polypeptide particularly in the large bowel that is the natural site of colonization of C. difficile. The bowel wall is surrounded by an immense immune apparatus, the so-called Peyer's patches and thus, specialized in mounting immune responses of various types. Large bowel colonization by a clostridial vaccine or peptide producer strain thus enables a much longer immune stimulus than a traditional injection. In contrast to clostridia, the alternative and much studied S-layer producers for vaccine purposes, Bacillus spp, are free-living, obligate aerobic bacteria and unable to replicate in the anaerobic bowel lumen and thus, unable to colonize a recipient mammal. For clostridial colonization and peptide delivery in hypoxic tissues iv administration is used.

It will be appreciated that Clostridia carrying the gene expression cassette of this invention including DNA encoding different heterologous peptides allows the highly efficient production and export of these polypeptides in hypoxic tissues after iv administration, or into the gut, particularly the colon, of the orally colonized individual for a variety of prophylactic or therapeutic uses. Another advantage of this gene cassette for expression of heterologous peptides is its versatility, i.e. that it is normally used to produce and export peptides of varying size and having completely different amino acid sequences, in their N-terminal or C-terminal end.

In further independent aspects of the invention the recombinant gene expression cassette is used to produce in the gut, for example (i) peptides and enzymes for therapy and prophylaxis of various diseases, e.g. peptides having specific antimicrobial activity, cytokines against inflammatory bowel disease, and β-lactamases to prevent diarrhea due to antibiotic therapy

(ii) single, fusion or multiple polypeptide antigens of microbial, animal or mammalian origin for neonatal immune balancing, vaccination against infections, allergy, metabolic or auto-immune disease, cancer, (infertility, and drug addiction, (iii) carrier molecules (so-called adjuvants) separate or fused to the antigen in order to . amplify or modulate the immune response to the antigen in a desired way according to

(ii), e.g. a strong IgA response against a mucosal invader.

In a still further aspect of the invention the gene expression cassette of the invention may be used to provide recombinant clostridia for local production of peptides in tissues after iv administration of their spores (see below), for example for the prophylaxis and/or treatment of fibrinolysis in arterial or venous occlusion and/or for revitalising gangrenous and/or necrotic tissue in various diseases. Furthermore, for anti-tumour therapy by local production of

(i) immune stimulating human peptides for improving tumour host defence, (ii) enzymes that convert a pro-drug to a cytostatic agent inside a tumour (thus avoiding systemic side effects)

(iii) cytotoxins of e.g. bacterial origin to destroy tumour cells (iv) angiogenesis inhibitors at local concentrations enough to prevent local blood vessel formation and thus, tumour growth (v) signal transduction inhibitors.

In a further aspect of the invention we provide a pharmaceutical or veterinary composition which comprises a transformed viable Clostridial cell with the ability to present and/or to secrete the desired polypeptide together with a pharmaceutically or veterinary acceptable carrier or diluent.

The composition may be formulated as a vaccine. The composition may be administered orally, or intranasally or alternatively, the polypeptide can be isolated, purified and administered parenterally, e.g subcutaneously or intramuscularly.

The amount of the desired peptide(s) presented and/or secreted by the transformed strain may be modulated in the body by using

(i) promoters with different strength (power),

(ii) a promoter or regulator responding to external stimuli (inducible, e.g. by a specific carbohydrate) normally present in the gut or administered together with the engineered bacterium,

(iii) different dosage regimens (number of bacteria per dose and doses per time period) or

(iv) methods that influence the ability of the strain to colonise and propagate in the gut for convenient periods of time. Relevant factors include ability to compete with other bacteria, adhere to mucosal cells, and to avoid expulsion by local immune response mechanisms. The latter can be achieved by exploiting induction of tolerance to the natural S- layer antigen of the strain in the neonate (see above and below) or by using its putative normal antigenic variation, suggested to us by the presence of the large number of S-layer ORFs present in C. difficile, either by allowing the vaccine strain to change serotype during natural long-term colonization or by repeated applications (colonizations) over time of different strains producing the same antigen but having a different serotype antigen.

The transformed Clostridia as anerobic organisms are conveniently produced by fermentation under for example low oxygen tension and purified and recovered as known in the art for native Clostridia, for example by washing and freeze-drying. They may be formulated together with excipients as needed, for example magnesium stearate, lactose, or carboxymethyl cellulose, into solid dosage forms, e.g. in capsules, predominantly for oral administration. The dosage forms may be protected against the acidity of the stomach by a suitable enteric coating, comprising for example Eudragite "S", Eudragite "L", cellulose acetate, cellulose phthalate or hydroxypropyl cellulose. A preferred dosage form comprises freeze-dried transformed Clostridia contained in vials or ampoules, optionally under inert gas. Preferably, the transformed Clostridia cells are administered orally or intranasally, as an aqueous, reconstituted suspension of the lyophihzed cells e.g in water or physiological saline, optionally with addition of pharmaceutically acceptable buffers, e.g. sodium bicarbonate, phosphate or citrate to keep the pH of the suspension between 6 and 8, preferably between 6.5 and 7.5.

The dosage forms produced as described above may comprise a mixture of viable and non- viable bacteria depending on the process and/or the storage conditions. The viable, transformed Clostridia will, after oral administration, become attached to those parts of the gut, for example the lower intestinal tract, which provide appropriate growing conditions and proliferate, producing the desired polypeptide in increasing amounts. This will provide for an enhanced and sustained physiological effect, for example immunisation, of the polypeptide. If exposure to defined amounts of the polypeptide is desired, non- viable transformed

Clostridia presenting the polypeptide can be administered. The non-viable cells can be obtained as known in the art, e.g. by exposing the live cells to agents, e.g. heat, formaldehyde, antibiotics or solvents, which kill them. It is also possible to use cell walls (sacculi) or to use S-layer fragments obtained by mechanical or other disruption of the bacterial cells. These agents can be formulated into pharmaceutical and veterinary compositions as described above for live transformed Clostridia.

By "secretion" or "release" we mean that the heterologous polypeptide is exported out from the host cell into the surrounding environment as a soluble antigen. This is conveniently achieved by fusing the DNA coding for the polypeptide to a DNA sequence coding for a signal peptide sequence of any one of ORFl,ORF3, ORF5-7, ORF 9 or ORFl 1 (SEQ ID NO: 1-7, cf. Figure 1 and Table 1) preferably to that of ORFl and expressing it as described above under control of a strong promoter and exporting it with the aid of the sec gene (ORF2) product. Alternatively, the DNA codes for a signal peptide sequence of any one of ORF D, E, G, H or I (SEQ ID NO: 8-12, cf. Figure 2 and Table 1), or that of any other suitable secreted bacterial protein.

By "presentation" we mean that the polypeptide is translocated across the cell membrane and presented on the surface of the bacterium in a sufficient manner for it to act as, for example, a particulate antigen. The DNA coding for the heterologous polypeptide may then for example be fused to a S-layer coding sequence, which codes at least for a functional cell wall binding portion of a S-layer protein of C. difficile (Figure 3) and expressed as described above to get exposure of the heterologous polypeptide on the outside of the host cell and thus, hooked to the amidase motif of the S-layer protein. Alternatively, omitting this motif from the construct in order to get increased release of the heterologous peptide (see Specific description B3).

The heterologous polypeptide may be a foreign epitope or immunogen giving rise to antibodies that protect against disease, we note that many antibodies elicited are not protective. It typically comprises an antigenic determinant of a pathogen. The pathogen may be a virus, bacterium, fungus, yeast or parasite. The antigen may also be a "self molecule for prevention or cure of disease (see below). The heterologous polypeptide may further be an antimicrobial peptide, e.g. for elimination of undesired microorganisms, and an anti-tumour peptide (see below) or a molecule that changes the immune response of the gut from a negative one, such as allergy or auto-immune tissue destruction, to a positive one, such as infection protection (see above). For example, Lactobacillus components are believed to prevent allergy development and live lactobacilli are currently given to infants in successful trials for prevention of allergy (Bjδrksten B, pers comm, and Kalliomaki et al. Lancet 357: 1076-1097, 2001). Cystein proteases such as cathepsin are thought to change the intestinal mucosal response to infection from a Th2 type (disease promoting) to a Thl response (infection protection). Alternatively, the polypeptide(s) may be enzyme(s) that improve digestion of food, or that together synthesize a polysaccharide antigen of a microorganism, an antibiotic, or a specific vitamin or other nutrient or hormone useful to the host mammal. An enzyme produced by the the engineered Clostridial bacterium may also be an antibiotic inactivating enzyme, e.g. a beta-lactamase, to be given together with or after the antibiotic for prevention of CD AD or non-specific antibiotic induced diarrhoea, common problems in hospitals today.

The heterologous polypeptide may also be a part of an antibody molecule. This may comprise the constant part in order for example to obtain an enhanced non-specific immune response or the response to a co-administered antigen (adjuvant effect). Alternatively, it may be the variable part directed against any surface or secreted component of a microorganism (toxin, antigen, adhesin) in order to prevent its ability to colonize and cause intestinal disease. The expression product of the cassette of the invention may also represent the immune stimulating part of allergy causing antigens lacking their IgE interacting part, thus evoking an antibody response but avoiding an allergic reaction (anti-allergy vaccination).

Whether the heterologous polypeptide is to be provided alone or fused with a carrier peptide, or presented cell-bound, released or both depends on its desired function. For example, for a polypeptide acting as an enzyme, free "secreted" molecules may be most effective, whereas in case of vaccination an antigen fused to a carrier peptide or being a cell- bound ("presented") polypeptide on a bacterium, strongly adhering to or being phagocytosed by the gut mucosa, may give the best mucosal immune response.

The immune response to a heterologous peptide may be increased by fusion to the repeating C-terminal sequences encoding the non-toxic motifs of the C. difficile toxins A and B that enable these to enter the colonic mucosal cells by receptor-mediated endocytosis, and/or to a portion of toxin B responsible for intracellular and intercellular spread of the antigen (see Barth et al below). Thus, by using adhering clostridial bacteria producing a desired heterologous peptide antigen fused to non- toxic parts of the C. difficile toxins, the mucosal immune reponse may be boosted (adjuvant effect).

A further improved immune response may be obtained by exploting the natural S- layer proteins of C. difficile that seems to anchor the organsim to the mucosa. It is likely that the amidase like fragments are directed inwards to provide cell wall anchorage, whereas the sequence unique fragments represent the the outermost portion of the S-layer protein serving as surface antigen (see above) and probably also as adhesin by which C. difficile attaches to the mucosal cell surface as recently suggested by Waligora et al. (Infect Imm 69:2144-2153, 2001). Thus, by switching between expression of its different S-layer ORFs over time each C. difficile strain may achieve surface antigen variation and thus, immune evasion and prolonged colonization in the gut. However, after deliberate colonization of a newborn with a certain C. difficile strain, particularly if expressing or being administered together with a "danger molecule", tolerance (see above and below) to its S-layer serotype protein may be obtained enabling the use of the same C. difficile serotype for efficient deliberate long-term colonization of this individual, e.g. for vaccination purposes, also during later periods of life. In a further aspect a carrier peptide or adjuvant, e.g. a "danger molecule" is used in addition to the desired heterologous polypeptide, administered or produced in vivo either as a separate molecule or fused to the principal (antigenic) polypeptide. This is in order to amplify desired specific immune responses for prevention or therapy of infection, or in the neonate also for shifting its general response towards anti-infection and tolerance of "self ' and therefore away from allergy and auto-immunity (see above and below). The "danger molecule" or adjuvant is a species that may stay in a human or animal body for a long time, such as up to one, three, six months or up to one year. Alternatively, or in addition, this species is capable of eliciting a stronger immune response that the desired heterologous polypeptide acting alone. "Danger molecules" are often of microbial origin, rapidly recognized and strongly reacted upon both by the innate/primitive and the trained/specialized immune system (see above). A convenient reference for this aspect of the invention is the thesis by Carola Rask of the Department of Medical Microbiology and Immunology at Gδteborg University, Sweden, ISBN 91-628-4497-0 "Cholera toxin B subunit as a carrier for inducing mucosal immunity and/or peripheral tolerance."

By using a mixture of different genetically engineered Clostridia strains, presenting and/or secreting different heterologous polypeptides, which act additively or synergistically, it is possible to achieve an enhanced physiological response. Thus oral immunization with a mixture of genetically engineered Clostridia strains, each presenting and/or secreting a different polypeptide derived from the same or another microbial pathogen, will provide both a broader immune response and a so called adjuvant effect, i.e. a more complete immune response and a better vaccination against the pathogen than obtained by using just one strain expressing a single immunogenic epitope of the pathogen.

In a further aspect of the invention we provide a medicament or therapeutic agent which comprises a Clostridial bacterium transformed with a gene cassette of the invention and capable of presenting on the surface of the bacterium and/or secreting a polypeptide in a human or animal body.

The medicament or therapeutic agent is conveniently a lyophilised powder for reconstitution as a suspension or for production of a solid pharmaceutical form such as a capsule or a tablet. The therapeutic agent can be administered orally or intranasally. For oral administration capsule or tablet formulations may be used. To protect the compositions against the acidity of the stomach buffer substances, e.g. sodium bicarbonate, may be used and/or the formulations may be covered with enteric coatings, e.g. Eudragite "S" or "L", cellulose acetate, cellulose phthalate or hydroxypropyl cellulose. A convenient way for oral administration of the therapeutic agents is to provide them as lyophilised powders, and shortly before administration to suspend these in for example water, fruit juice or physiological saline, optionally with addition of sodium bicarbonate or neutral citrate, or phosphate buffer to protect against the acidity of the stomach. Any convenient dose may be used, this may be in the range from 1 to 10^π bacteria, more conveniently we anticipate this to be in the range from about 10³ to about 10⁹ bacteria.

A principal use of the invention is in vaccination. Therefore in a further aspect we provide a vaccine which comprises a Clostridial bacterium transformed by a gene cassette of the invention and capable of secreting and/or presenting an antigen on the surface of the bacterium in a human or animal body.

In addition to improving the existing anti-infection vaccines and creating new ones, there is also current interest in several novel uses for vaccines.

Allergy One strategy is engineered anti-allergy vaccines containing the immunostimulatory part of each antigen but lacking the part which interacts with IgE and thus, normally elicits the allergic reaction. Another new approach is to induce an immune response towards human IgE, that normally governs the allergic response, by turning these molecules into "non-self ones e.g. by coupling to IgE of animal origin. The use of these hybrid IgE molecules as vaccine is expected to elicit production of anti IgE antibodies that thus, inactivate human IgE thereby preventing allergy.

Alternatively, allergy may be prevented by stimulating the immune apparatus of the newborn in such a way that cellular, IgG and IgA antibody responses to microbial antigens, i.e. anti-infection, will be preferred to IgE production against allergens (immune balancing). Auto-immune diseases. Another new proposed area for vaccines is to boost tolerance to "self antigens in utero and/or in the newborn in order to prevent later development of auto-immune disorders such as type 1 diabetes, rheumatoid arthritis, inflammatory bowel disease and multiple sclerosis. This may be achieved either by non-specific tilting of the newborn immune system towards anti-infection and away from auto-immunity and allergy as mentioned above, or by applying the "self molecule (e.g. human insulin or other beta cell antigens, connective or CNS tissue antigens) coupled to or together with a "danger molecule" of microbial origin (e.g. part of the tetanus or cholera toxin, see above) here in order to amplify the normal immunotolerance response to e.g. insulin and thus, the natural avoidance of juvenile diabetes.

Presnancv and metabolic diseases. In contrast to the newborn, exposure to a "self antigen especially when coupled to a "danger molecule" may in the adult individual lead to an immune response to the antigen rather than reinforced tolerance. Such vaccines boosting specific auto-immunity may be used for prophylaxis and therapy by eliciting antibodies directed against specific "self target molecules, such as sperm or egg components or human gonadotropin (hCG) to prevent fertility, enzymes in cholesterol biosynthesis to prevent arteriosclerosis, beta amyloid for prevention and cure of Alzheimer's disease, other brain proteins to counteract prion and Creutzfeld Jacobs disease.

Drus addiction. A further novel application of vaccines includes the use of drugs molecules such as nicotine or heroin as part of the antigen for induction of anti drug antibodies that block its activity and remove the drug and thereby abolishes its CNS effect, in order cure addiction.

Cancer. Novel multicomponent vaccines containing "danger molecules" may be of use also against cancer both by boosting the innate immune defense, by eliciting anti-tumour antibodies and cellular immune responses or by stimulating apoptosis of cancer cells.

Before the optimal anti-infection and other types of mucosal vaccines can be achieved a lot more has to be learned, e.g. about the interaction of C. difficile with M-cells and mucosal cells in the gut, the uptake, processing and presentation of each antigen, and optimizing the choice and presentation of adjuvant so that the immune response can be maximized and modulated in the desired way and thus, perfected for each application in order to obtain a desired cellular, IgA, IgG and subclass or combined response. Thus, the choice, size and form (soluble, particulate) of antigen, the vector (live?), the choice and form of carrier molecule (adjuvant, separate, fused to the antigen), recipient (mother, child, both, adult), timing of administration etc need to be tailored in each case. In the above approaches to vaccination against infection in mammals and for immune modulation in newborns and adults recombinant clostridia producing desired popypeptides may be of particular interest as live vectors for mucosal immunization since certain species, e.g. C. perfringens and C. difficile belongs to the normal gut flora. C. difficile is particularly common in the neonatal period, and toxin negative strains thus can be given orally at all ages without ethical concerns. C. difficile appears to be a particularly good candidate also for delivery of antigens for gut mucosal immunization as exposure to microscopic numbers of the organism during hospital stay resulting in asymptomatic carriage is enough to yield an immune response to its toxins (NEJM 2000). Furthermore, we have observed in animals that asymptomatic gut colonization by C. difficile results in an immune response also to its S - layer protein (see below). These responses are probably enhanced by the the non-toxic part of the C. difficile toxins that are used for their receptor mediated pinocytosis into the mucosal cells. Toxin B then can form membrane pores in the pinocytic vacuoles containing toxin and presumably also in phagocytic vacuoles containing whole bacteria (Barth H et al, Low pH induced formation of ion channels by C. difficile toxin B in target cells. J Biol Chem 276(14): 10670-10676,2001). Thereby the the toxin and other bacterial components may be released into the cytosol of the mucosal cells and may spread also to neighbouring cells including to antigen presenting cells and thus, enhancing an immune response. Such an unusual adjuvant effect of C. difficile toxin B obtained by breakage of phagocytic vacuoles and intercellular spread of internalized antigens and bacteria can alternatively also be obtained e.g. by including the membrane attacking peptide listeriolysin O from Listeria monocytogenes in a recombinant C. difficile strain in order to boost immunity as has been shown in experiments using other gut mucosal delivery systems (Dietrich G et al, From evil to good: a cytolysin in vaccine development, Trends in Microbiology 9:23-28, 2001).

Clostridia furthermore represent a unique torpedo able to deliver a desired heterologous polypeptide to hypoxic tissues such as tumours. This is because spores of these obligate anaerobic organisms given intravenously are known to settle and be able to germinate into growing clostridial cells in the hypoxic parts of tumours but not in healthy tissues. This phenomenon was described already in 1955 (reference 7 in Theys J et al, FEMS Immunol

Med Microbiol 30:37-41, 2001) and is currently being studied for therapeutic applications by several groups (see also Abstracts, The 3^rd Int Meeting on Molecular Genetics and Pathogenesis of the Clostridia, Chiba, Japan, June 8-11, 2000). Thus, anti-tumour peptides including apoptosis inducing peptides, cytokines, toxins and other proteins, such as enzymes locally converting pro-drugs to active anti-cancer chemotherapeutic agents, thus minimizing systemic side effects, all produced by recombinant clostridia inside tumours may become novel approaches to cure cancer. We propose that another approach to anti-tumour therapy is iv injection of spores of recombinant clostridia producing various angiogenesis inhibitors, mostly of peptide nature and currently used in many clinical trials (phase I: angiostatin, SU6688, combrestatin A-4 prodrug, PTK787/ZK2284; phase II: endostatin, anti-VEGF Ab, TNP-470, 11-12, 2-methoxyestradiol, squalamine, vitaxin, EMD 121974, COL-3, CGS- 27023 A, CAI; phase III: thalidomide, marimastat, INF-alfa, neovastat, BMS-275291, SU5416, AG3340, IM862 as summarized in Larsson H, Regulation of angiogenesis, Thesis, 2001, Uppsala University, Sweden, ISBN-91-554-4954-9). A problem in these studies and possible to overcome by local production of enzymes, toxins and other peptides is limited effect in vivo due to short half-life in serum and thus, insufficient local concentrations of e.g. the cytostatic agent or the anti-angiogenesis peptide. In the latter application recombinant clostridia may become particularly powerful (self- accelerating), because the effect of the peptide will further lower the oxygen tension and thus, enhance bacterial growth and further production of the peptide(s)

We further believe, that that iv administration of spores from recombinant clostridia may be used also against other diseases involving local tissue hypoxia such as fibrinolytic and other agents for venous or arterial occlusion, and oxygen releasing or other tissue vitalizing molecules for tissue necrosis and gangrene.

We now found and disclose that Clostridial spores may be used to deliver heterologous polypeptide(s) to a human or animal body. This is an important step forward. A spore is a dormant or resting state of a bacterial cell. Unlike bacterial spores from species belonging to the obligate aerobic genus Bacillus (see above), ingested Clostridial spores naturally germinate into vegetative bacteria that can grow anaerobically and naturally colonise a human or animal gut. Intake of the spores of the genetically engineered Clostridia is preferably through the oral route. Spores are able to resist stomach HC1 and digestive enzymes. Upon contact with bile they will germinate and establish themselves in the colonic flora as vegetative bacteria presenting and/or secreting for example the desired heterologous peptide in vivo.

Therefore, in a further aspect of the invention we provide a therapeutic agent which comprises spores of Clostridia transformed with a construct capable of expressing, secreting or presenting a heterologous polypeptide in the mammalian body after conversion (germination) to live (vegetative) bacteria.

The construct is preferably a recombinant gene cassette of the invention as outlined before. The mammalian body is preferably a human or animal.

The use of Clostridial spores has a number of advantages including low production cost, relative ease of production, very long shelf life independently of the mode of storage, ease of administration, production of antigen at the site of action, and an oral route of immunisation which may be superior to a parenteral one. As mentioned above for bacteria, spores are suitable for administration of mixtures (coctails) of recombinant Clostridia having different desired properties.

The use of live vaccines administered via the oral route may lead to further fecal-oral transmission and enhanced immunization of a population. On the other hand, this may also be considered as to be unwanted spread of genetically modified organisms in the environment. Spores of Clostridia survive readily in the environment whereas the vegetative forms have a very limited capability to survive in an oxygen-containing milieu. The invention may be further developed to create Clostridia that are unable to reconvert to spores, once they have germinated in the colon. One way is to modify a genetic element present in C. difficile that is similar to the so-called skin (Sigma K intervening) element of B. subtilis. This element truncates the sigma K factor necessary for sporulation, and becomes removed by a specific excision system during sporulation (Krogh, S. et al.(l 996) and Takemaru, K. et al.(1995)). By genetic modification of the excision system, and insertion of wild-type copies with inducible promoters, there is a possibility to create a host strain that is able to sporulate only during special conditions, e.g. in the presence of a special chemical (IPTG) or at low temperature (20 °C). Such construction would allow the production of spores in vitro, whereas no new spores are created in the vaccinated host. The spread of genetically modified Clostridic microorganisms to the environment would still occur, but the probability of survival of these organisms would in practice be very low or nil.

Spores of the transformed Clostridia are produced, purified and isolated in the same way as for native Clostridial strains. They may thus be readily obtained from a stationary phase culture for example by treatment with ethanol, acid or heat or by combinations of such measures followed by purification and isolation in a conventional way. As outgrown spores will have the same properties as the parental bacteria, purification of the spores may not even be necessary.

Pharmaceutical and veterinary compositions for oral administration, comprising spores of transformed Clostridia and pharmaceutically acceptable carriers, diluents and excepients are further provided by the invention. They have the ability to colonise the intestinal tract of humans and animals with live Clostridial bacteria producing and presenting or secreting the heterologous polypeptide coded for by the modified gene cassette provided by the invention or by any other construct. The pharmaceutical and veterinary compositions may comprise tablets, capsules, powder for reconstitution or any other form suitable for oral administration to humans or animals. Examples of pharmaceutically acceptable carriers and diluents are lactose and carboxymethyl cellulose. A convenient way of oral administration of these therapeutic agents is to provide them as lyophihzed, or just dried, powders; shortly before administration they are suspended in for example water, physiological saline or fruit juice. The dose is as indicated above for Clostridial bacteria.

In a further aspect of the invention we provide a method of treatment for the human or animal body, which comprises of administering a therapeutic agent comprising Clostridial spores capable of expressing a heterologous polypeptide in a human or animal body. As in the treatment with Clostridial bacteria described above it will be possible to treat with mixtures of transformed Clostridial spores to obtain in the body a mixture of different heterologous polypeptides, which may act synergistically or additively.

It will be appreciated that the therapy may be either prophylactic or therapeutic. The method may be applied to any convenient mammal such as a human or animal.

Convenient animals include domestic animals such as dogs and cats, also cattle, pigs, chicken and horses.

In a further aspect of the present invention we provide a method for immunisation which method comprises administering to a mammalian body Clostridial spores capable of expressing a heterologous antigen after germination.

Examples of convenient Clostridium spores include the spores of C. difficile and C. perfringens, which normally colonise the large intestine of man. For animals other Clostridia such as C. tetani may also be useful. Thus, the intensity and duration of antigen exposure in the gut (clostridial colonization) in a particular host can be varied by not only exploiting and manipulating e.g. adherence of C. difficile (see above), but also by selecting the appropriate Clostridium species with regard to the intended host mammal.

In a further aspect of the invention we provide Clostridial spores transformed with a gene expression cassette of the invention.

It will be appreciated that the methods and materials of the invention may also be used for other applications such as the display of antibodies and peptide libraries. They may also be used for screening proteins and antigens and also to provide a support for immobilising an enzyme, peptide and/or antigen.

Sequence numbers of proteins encoded by the ORFs of the invention Protein encoded by ORF SEQ ID NO:

ORF 1 SEQ ID NO: 22

ORF 3 SEQ ID NO: 23

ORF 5 SEQ ID NO: 24

ORF 6 SEQ ID NO: 25

ORF 7 SEQ ID NO: 26

ORF 9 SEQ ID NO: 27

ORF 11 SEQ ID NO: 28 Protein encoded bv ORF SEQ ID NO:

ORF D SEQ ID NO: 29

ORF E SEQ ID NO: 30

ORF G SEQ ID NO: 31

ORF H SEQ ID NO: 32

ORF I SEQ ID NO: 33

The present invention is particularly directed to a gene expression cassette comprising a secretory leader sequence encoding a signal peptide from Clostridium difficile having an amino acid sequence selected from the group consisting of SEQ ID NO: 1, SEQ JD NO: 2, SEQ ID NO: 3, SEQ ID NO: 4, SEQ ID NO: 5, SEQ ID NO: 6 , SEQ ID NO: 7 and signal peptides of analogous exported clostridial N-acetylmuramoyl-L-alanine amidase-like proteins, linked to a DNA sequence encoding a heterologous polypeptide. The signal peptides of the analogous clostridial N-acetylmuramoyl-L-alanine amidase-like proteins may also be selected from Clostridium difficile signal peptides having an amino acid sequence of any one of SEQ ID NO: 8, SEQ ID NO: 9, SEQ ID NO: 10, SEQ ID NO: 11, and SEQ ID NO: 12.

The gene expression cassette may further include a promoter of prokaryotic origin, e.g. selected from clostridial promoters comprising a nucleotide sequence selected from the group consisting of SEQ ID NO: 13 - 21, or from the promoters of ORFs 1-11 or D-I mentioned above. The gene expression cassette according to the invention may further include a DNA sequence encoding at least a cell wall binding portion of a protein of prokaryotic origin functioning in clostridia such that a fusion polypeptide may be presented on the outer surface of a host cell harbouring the cassette.

The gene expression cassette according to the invention may in particular include a DNA sequence encoding at least a functional cell wall binding portion of an S-layer protein of C. difficile selected from any one of the polypeptides having an amino acid sequence selected from SEQ ID NO: 22 - 33 such that a fusion polypeptide may also be presented on the outer surface of a host cell harbouring the cassette. The DNA encoding the cell wall binding portions of SEQ ID NO: 22-33 may be omitted such that the fusion peptide is secreted into the surrounding milieu by the host cell harbouring the cassette. Further, the gene expression cassette according to the invention may be such that the DNA sequence encoding the heterologous peptide is inserted at a point downstream the first (signal) proteolytic cleavage sites in the gene encoding a polypeptide having an amino acid sequence selected from SEQ ID NO: 22 - 33, optionally including or excluding its second cleavage site.

In addition, the gene expression cassette according to the invention may further comprise at least a functional part of a secretory (secA) gene recognizing the signal peptide, to allow translocation of a heterologous polypeptide and/or fusion polypeptide across the cytoplasmic membrane of a host cell harbouring the expression cassette. For example, the secretory gene may be from C. difficile and encode a polypeptide having the amino acid sequence SEQ ID NO: 34.

In a preferred embodiment of the invention the gene expression cassette is the one that is shown in Figure 3.

The invention is also directed to a vector comprising a gene expression cassette according to the invention, such as a plasmid.

The invention is further directed to host organism transformed with a vector according to the invention for expression of the heterologous polypeptide and/or fusion polypeptide.

In an embodiment of the invention the host organism is a Clostridium host organism transformed with a vector according to the invention for expression of the heterologous polypeptide and/or fusion polypeptide.

In a preferred embodiment, the host organism is C. difficile or C. perfringens.

Further, the invention is directed to a pharmaceutical or veterinary composition or formulation which comprises Clostridial cells transformed with a vector according to the invention, with the ability to present on the cell surface and/or to secrete an expressed heterologous polypeptide or fusion polypeptide, together with a pharmaceutically or veterinary acceptable carrier or diluent. Preferably, the composition or formulation is suitable for oral or intranasal administration. The composition or formulation according to the invention may further comprise, as adjuvants, non-toxic motifs of the C. difficile toxins A and/or B that enable the heterologous polypeptide and/or fusion polypeptide to enter the colonic mucosal cells of a mammal by receptor-mediated endocytosis, and/or a portion of toxin B responsible for its intracellular and intercellular spread. The composition or formulation according to the invention may alternatively additionally comprise a further fused or separate carrier peptide or adjuvant, in addition to the expressed heterologous polypeptide and/or fusion polypeptide, to elicit a stronger or differently directed immune response than that against the expressed heterologous polypeptide acting alone.

The invention is, in another aspect, directed to a vaccine which comprises a Clostridial bacterium transformed with a vector according to the invention and capable of presenting on the surface of the bacterium and or secreting an antigen in a human or animal body, and optionally an adjuvant described in conjunction with a composition or formulation of the invention. The vaccine may comprise a mixture of at least two differently engineered Clostridia strains, each capable of presenting on the surface of the bacteria and/or secreting a different heterologous polypeptide and/or fusion polypeptide. Further, the vaccine may comprise spores of Clostridia cells or bacteria transformed with a vector according to the invention and capable of germinating into cells which are able to grow, express, and present or secrete a heterologous polypeptide and/or fusion polypeptide, and optionally also an adjuvant described in conjunction with a composition or formulation of the invention, in a mammalian body. The vaccine may comprise a mixture of spores from at least two differently engineered Clostridia strains. Each of these strains is capable of presenting on the surface of the bacterium and/or secreting a different heterologous polypeptide and/or fusion polypeptide. The spores are preferably from C. difficile or C. perfringens.

The invention is in yet another aspect directed to a medicament which comprises a Clostridial bacterium transformed with a vector according to the invention and capable of presenting on the surface of the bacterium and/or secreting a heterologous polypeptide and/or fusion polypeptide in a human or animal body, and optionally an adjuvant described in conjunction with a composition or formulation of the invention. The medicament may comprise a mixture of at least two differently engineered Clostridia strains, each capable of presenting on the surface of the bacteria and/or secreting a different heterologous polypeptide and/or fusion polypeptide. Further, the medicament may comprise spores of Clostridia cells or bacteria transformed with a vector according to the invention and capable of germinating into cells which are able to grow, express, and present or secrete a heterologous polypeptide and/or fusion polypeptide , and optionally an adjuvant described in conjunction with a composition or formulation of the invention, in a mammalian body. The medicament may comprise a mixture of spores from at least two differently engineered Clostridia strains. Each of these strains is capable of presenting on the surface of the bacterium and/or secreting a different heterologous polypeptide and/or fusion polypeptide. The spores are preferably from C. difficile or C. perfringens. The invention is in still another aspect directed to a method for vaccination of a mammal, which comprises administering a therapeutically or prophylactically effective dose of a vaccine according to the invention to the mammal. Spores used in the vaccine are preferably from C. difficile or C. perfringens. The invention is also directed to a method for prophylactic or therapeutic treatment of a mammal, which comprises administering a therapeutically or prophylactically effective dose of a medicament according to the invention to the mammal. Spores used in the medicament are preferably from C. difficile or C. perfringens.

The invention is additionally directed to a C. difficile-associated diarrhea (CD AD) vaccine comprising spores according to the invention and capable of expressing, after germination,

(i) relevant parts of the C. difficile toxins, alone or together with a (ii) suitable adjuvant to provide an IgA response to the toxin antigenic epitopes and (iii) S-layer protein antigenic variants (serotype antigens) or fimbrial antigens to obtain, after administration to a mammal, a polyvalent anti-S-layer (or anti-fimbrial) IgA response preventing C. difficile colonization of the mammal. The invention will now be illustrated but not limited by reference to the following Figures, Specific Descriptions, Tables, and Examples wherein: Figure 1A shows our first simple layout of the C. difficile strain 630 genomic segment encoding the S-layer genes. ORF2 represents secA and ORFs 1,3, 7-9 and 11 S-layer protein genes. For explanations of the these and other ORFs, see Table 1 and Example 9. Figure IB represents the result of comparisons between three of the S-layer ORFs with published sequences of other genes. The "amidase enhanced precursor" sequence is equivalent to the N-acetyl muramoyl L-alanine amidase motif mentioned in the text. Figure 2. Defining the upstream region of ORF 1-12. The figure illustrates additional information and genetic organisation of the C. difficile S-layer genes (cf. Figure 1), found after searches in the revised C. difficile database at the Sanger Centre. The genes upstream of ORF 1 to 12 are denoted A to I (see also Table 1). The numbers +1, +2 and +3 indicate the reading frame of the ORFs relative to the start point of the contig. ORFs D, E, G, H and I had the amidase motif typical of genes encoding the C. difficile S-layer proteins.

Figure 3 shows an example of a preferred gene expression cassette here taken from C. difficile strain 630 and containing a strong promoter, the secretory leader peptide from ORFl, the signal peptide cleavage site the area of insertion of foreign DNA encoding the heterologous peptide, the second (optional) peptide cleavage site in the N-acetyl muramoyl L- alanine amidase motif, and the secA gene (ORF2).

Figure 4 shows a preferred strategy for introducing a recombinant gene cassette of the invention back into C. difficile via B. subtilis . Figure 5. Further details of a particular C.difficile S - layer gene cassette. This is a 4960 bp cassette taken from strain 630 encoding an S-layer protein of 2160 bp in its original form (ORFl). The 210 bp region (pr, promoter) upstream of ORFl includes gene control elements for the S-layer protein included in the cassette. Also shown are an intervening 244 bp region and the 2346 bp sec A sequence. Figure 6. Strategy for the engineering of ORFl (Figure 5) to express a recombinant protein (for example as outlined in Example 2 and 2A). The 613 bp variable region (vr) is replaced by a foreign DNA. For example, three fragments encoding the Hepatitis B virus surface antigen (HBsAg) were selected: (i) the full length HBsAg that includes the pre SI, pre S2 and the S gene (1207 bp); (ii) the S gene (740 bp); and (iii) the subtype from the S gene (minimum antigenic epitope, 421 bp).

Figure 7. Cloning strategy for the construction of ORFl - secA (with and without the native promoter) together with the different lengths of the HBsAg antigenic loop (full length, S gene and Sub type - see legend to Fig. 6 above) using a PCR based method and cloning into the TA vector in E. coli. The primers indicated were used also for PCRs to help checking the correctness of the constructs. The expected and obtained constructs were 5564, 5097 and 4778 bp respectively.

Specific descriptions

A. Characterisation of the genomic segment responsible for S-layer protein expression in C. difficile (outlined in Fig. 1 and Fig. 2, see also Table 1 and Example 9):

1. The main surface layer proteins expressed by C. difficile strain 630 has been found to encoded by a single open reading frame (ORFl) encoding a 72 kDa protein. The gene product of ORFl is postranslationally cleaved at two sites yielding three different peptides; the leader peptide and the final S-layer proteins of apparent molecular weights of

36 kDa and 45 kDa. 2. The C-terminus of ORFl shows similarity to N-acetyl muramoyl L-alanine amidase, and the N-terminus shows weak similarity to surface layer proteins from L. helveticus (Fig. IB).

3. The gene immediately downstream of ORFl (ORF2) encodes the SecA protein responsible for secretion of proteins with signal peptides.

4. The genes downstream of secA encodes proteins with similarities to ORFl (ORF3, 5-7, 9 and 11, see Fig. 1).

5. ORFl is efficiently expressed and its product is efficiently exported in strain 630, whereas e.g. ORF3 is expressed more than 100-fold less in strain VPI 10643, indicating a strong termination between ORFl and 3 (as judged by identification of exported proteins by two- dimensional gel electrophoresis).

6. The upstream region including the putative promoter has not been characterized functionally, but the very high expression of ORFl in various growth conditions indicates the action of a strong, constitutive promoter. It is also active in E.coli (see Example 2). 7. In the revised C. difficile sequence database from Sanger Centre, the upstream region of ORF 1 was included and revealed 9 new ORFs (A-I) of which 5 (D, E, G, H and I) had the N-acetylmuramoyl L-alanine amidase motif typical of the C. difficile S-layer protein ORFs (Fig. 2 and Table 1). The putative promoter region for ORFl is thus situated between ORFl and ORF (See Example 1 A and Table 2). The S-layer proteins from strain VPI 10463 have similar molecular weights but different pi as compared to those of strain 630, and the N-terminal sequences of the two S-layer proteins from VPI 10463 showed no similaritiy with those of strain 630. Studies of strains from different serogroups showed that the S- layer proteins vary in pi and molecular weight. The downstream region of the gene segment may in part be more conserved, since the N-terminal sequence from another extracellular protein from strain VPI 10643 was identical to ORF3 of strain 630. Our results indicate that ORFl is located at part of the chromosome that is capable of expressing and exporting various S-layer proteins depending on the strain.

B. Designing a preferred engineered gene cassette based on the C. difficile S-layer gene segment data (outlined in Fig. 3).

1. Any strong prokaryotic promoter functional in Clostridia can be used to express the heterologous peptide, e.g. the promoter of ORFl or any of the promoters of genes encoding other highly expressed proteins in C. difficile such as certain electron transfer proteins (our unpublished data and Fix A and Fix B in Example 9) or ribosomal proteins. 2. A secretory leader peptide, preferably the leader peptide from ORFl, is fused with the heterologous peptide, to ensure its translocation across the cell membrane 3. Depending on the desired fate of e.g. the antigen (secreted, surface presented or both), the heterologous peptide is optionally fused to the amidase part of the S-layer protein optionally including the part involved in the proteolytic cleavage event (Figure 3). Thus, for maximum release of the peptide the secretory leader of e.g. ORFl may be sufficient. On the other hand, maximum cell- wall binding may require fusion to the amidase portion but omitting the proteolytic cleavage sequence in the middle of the gene (Figure 3). If both free and bound heterologous peptide is desired one recombinant cassette of each type present in the same Clostridium strain or a mix of two different strains, each harbouring one the recombinant cassettes, can be used. The peptide cleavage site may be exploited if for instance the antigen and an adjuvant are produced in a fused form, to obtain equal amounts of the two, but are desired as separate peptides on the outside of the producer bacterium. To what extent parts of the N-terminal (variable) portion of e.g. ORFl can be used to optimize the localization of the heterologous peptide requires further experimentation.

4. The secA gene is usually included in the construct to ensure efficient translocation of the polypeptide across the cell membrane. 5. The gene construct is made in plasmids suitable for transformation of both E.coli and C. perfringens (e.g. pJTR750 or 751) or in plasmids suitable for conjugation into C. difficile via B. subtilis (e.g. ρCI195 or ρSMB47, Figure 4).

Examples

Example 1

Defining a minimal gene segment responsible for production and export of the heterologous protein (an engineered gene cassette, see also Figure 3). A. Defining the promoter start (the 5' end). The promoter region may be further characterised in different C.difficile strains, for example by the following steps:

1. Defining the DNA upstream of ORF 1 by circular PCR. One or several restriction enzymes are used to cut the DNA outside ORFl. Useful enzymes are Ndel, TaqI, PstI, and BsmAI. A Southern blot is performed using a probe against ORFl to confirm size of the fragment (optimal size is around 1-3 kb). The cleaved genomic DNA is then ligated into circles. PCR is then performed using primers (directed outwards of each other) directed against ORFl. The PCR product is then cloned and sequenced. Optionally, PCR is performed with one primer directed against ORFl (lower primer) and different arbitrarily designed primers. The products are then cloned and sequenced.

2. Identification of the transcription start. Primer extension with primer located at the 5 prime end of the genomic segment is used.

3. By computerised search for homologies to known promoters. The new sequence data for strain 630 enabled a search for putative promoters upstream of ORF 1 and the result is shown in Table 2. Due to the AT rich genome, several putative promoters were found. The actual promoter start has to be experimentally determined as described in point 2 above.

B. Defining the termination point (the 3' end). Primers directed at different parts of ORFl, 2 and 3 to determine transcript abundance using Northern blots. Termination loops in the RNA may be identified by computer analysis.

C. Conservation of the S-layer locus in other C. difficile strains: characterization of part of the S-layer genomic segment (ORFl and proximal parts of its upstream region and of part of ORF2) in different C. difficile serogroups in order to check for conservation of this region of the genome.

Primer pairs that are directed against the identified upstream-transcription start region and the proximal part of ORF2 were designed. PCR was performed on different strains belonging to all serogroups to confirm the generality of the expression center of the S-layer locus.

(i) PCR performed with chromosomal DNA as template from the different serogroups of C. difficile.

The primers used:

CONS 1 : 5 prime- TAT AAT GTT GGG AGG AAT TTA AGA - 3 prime, total length 24 nt (5 prime end starts at 8th nt upstream of ORFl, ends at 32nd nt)

CONS2: 5 prime- CAA ATC CAA ATT CAC TAT TTG TAC - 3 prime, total length 24 nt (5 prime end starts at 2983rd nt downstream of ORFl, ends at 2959th nt) Total size of expected PCR pdt (from strain 630 sequence): 2975 bp (includes ORFl and the proximal part of ORF2).

The enzyme/system used: Expand™ Long Template PCR System from Boehringer Mannheim.

Reaction conditions (as specified by the manufacturer): In a total reaction volume of 50ml, 350mM dNTPs, 300nM primers, 50ng chromosomal DNA template, lx supplied PCR buffer with 1.75mM MgCl₂, and 2.5U of a mix of Taq and Pwo DNA polymerase. 10ml of the reaction mix was run on a 0.8% Agarose-TBE (Tris-Borate- EDTA buffer) to check for product.

PCR cycle conditions: Initial denaturation - 92°C for 2 mins For 30 cycles - denaturation - 92°C for 10 sees annealing - 40°C for 30 sees elongation - 68°C for 2 mins

Final elongation - 68°C for 2 mins

PCR results:

Sero group tested Product size No. of Expts

A none 3

B around 2900bp 2

C around 2900bp 2

D around 2900bp(lesser amt) 2

F around 2900bp 2

G around 2900bp 2

H pdt > 2900bp (lesser amt) 3

I around 2900bp 2

K around 2900bp 2

X around 2900bp 2

A2 around 2900bp (lesser amt) 3 Serogroup tested Product size No. of Expts

A3 around 2900bp (lesser amt) 2

A4 around 2900bp (lesser amt) 2 A5 none 2

A6 around 2900bp (lesser amt) 2

A8 around 2900bρ 2

A9 around 2900bp (lesser amt) 2

A10 around 2900bρ 2 SI around 2900bp 2

53 around 2900bp 2

54 none 2

VPI 10643 (G) around 2900bp 4, last PCR was weak

630 (X) around 2900bp (expected size) 5, last PCR was -ve Serogroups A, A5 and S4 did not give any PCR product with several attempts.

The PCR reactions were very sensitive to template condition, which had to be prepared fresh.

(ii) Restriction Enzyme Analysis of the PCR products.

12ul PCR reaction mix (reactions done above) from most of the serogroups (except A, A5-6, A9, and S4) were further subjected to Rsal (Boehringer Mannheim) and Sau3Al (Amersham Pharmacia Biotech) digestions. Total digestion reaction mixes were run on a 0.8% Agarose- TBE (Tris-Borate- EDTA buffer) to obtain respective digestion patterns.

Each serogroup tested appears to give rise to unique restriction pattern though the apparent size of the PCR product appears to be similar except for gp H. Conclusion:

It appears that the promoter-ORFl-ORF2 organization and the size of the genetic segment between the putative promoter and ORF2 (secA gene), i.e. the ORF 1 equivalent, is generally conserved in C. difficile. There may be sequence variation between ORF 1 eqivalents from strains of different serotype and also between two strains of the same serotype, presumably reflecting the variable (non-amidase) portion of these genes. Example 2

Cloning of the C. difficile promoter-ORFl-ORF2 "cassette" from strain 630 and construction of a recombinant cassette in E. coli encoding heterologous proteins to be transferred to C. perfringens and C. difficile and used for immunisation .

1. Primer pairs are designed that include the promoter region and part of ORFl including the leader peptide sequence (Figure 5). PCR is performed followed by cloning of the product into E. coli-C. perfringens shuttle vectors ρJIR750 or pJTR751 (Plasmid 229: 233-235, 1993) in frame of a reporter gene such as β-lactamase or at least a part of the hepatitis B virus (HBV) antigen. A convenient source of HBV antigen is the SMI strain no. 8423/87 having the genotype A and subtype adw2 (cf. Magnius et al, J.Gen. Virology, 1993, 74, 1341-1348). The plasmid is isolated from E. coli, purified and used to transform C. perfringens, and the engineered strain is isolated for further use. The secA gene is optionally included in the construction to optimise secretion. 2. Check for expression and secretion of the reporter gene using an HBV antigen based assay or a β-lactamase assay of the transformed E.coli and C. perfringens strain. Gnotobiotic mice and rats are fed with spores of the engineered Clostridium strain to obtain colonisation. Expression of antigen is checked by the HBV antigen based assay or β- lactamase assay in feces and immune response by antibody response in feces and in serum.

A. Cloning of ORFl - ORF2 (secA) in order to construct a fusion with a foreign antigen

PCR was performed with chromosomal DNA from Strain 630 as template. The primers used:

AMP1: 5' - GGAATT CCATGAATAAGAAAAATATAG CA- 3', total length 29 nt (5 'end starts at the first codon of ORFl, ends at 7th codon ) AMP2: 5' - CGG GAT CCC GTT TTT AGT TAA ATT TAT ATA AG - 3', total length 32 nt (5 'end starts at starts at the stop codon for secA) Analogous PCRs but with the first primer in the upstream region in order to include a putative native promoter were also performed.

Total size of expected PCR product (from strain 630 sequence) : 4770 bp (4960 bp including the promoter) (Figure 5). The enzyme/system used: Expand™ Long Template PCR System from Boehringer Mannheim. Reaction conditions (as specified by the manufacturer): In a total reaction volume of 50ul, 350mM dNTPs, 300nM primers, 150ng chromosomal DNA template from strain 630, I x supplied PCR buffer with 1.75mM MgCl₂, and 2.5U of a mix of Taq and Pwo DNA polymerase. lOul of the reaction mix was run on a 0.8% Agarose-TBE (Tris-Borate- EDTA buffer) to check for product.

PCR cycle conditions: Initial denaturation - 92°C for 2 mins For 30 cycles - denaturation - 92°C for 10 sees annealing - 40°C for 30 sees elongation - 68°C for 4 mins final elongation - 68°C for 5 mins Results:

The expected PCR products were obtained and cloned into pGEMT vector (Promega). The plasmid containing the insert will be subjected to partial digestion with PvuII enzyme (sites at position 282 and 895 of the insert) to eliminate the 613 bp internal fragment from ORFl, where the foreign antigen is planned to be inserted (Figure 6). The digestion time had to be standardised. The foreign antigen used was the hepatitis B virus (HBV)surface antigen (HbsAg).

Three HBsAg regions were used (Figure 6): (a) Total (413 aa, 3-1243 bp; 1240 bp) (b) Central (259 aa, 465-1243 bp; 778 bp)

(c) C-terminal (140 aa, 822-1243 bp; 421 bp) References for HBV antigen sequences include Prange et al, 1995, 14 (2), 247-256 and Chen et al, 1996, 93, 1997-2001.

Alternative antigens that may be used include relevant epitopes of the rota virus and hepatitis A virus.

B. Cloning of the C. difficile OKFl-secA cassette containing the three HBsAg DNAs in E. coli.

Result: The cloning strategy (Figure 7) using PCR and the TA vector was successful according to DNA analyses including agarose gel electrophoresis and PCR. C. ELISA to check for expression in E. coli of the three ORFl-HBsAg-sec<4 DNA constructs in the TA vector.

A commercially available ELISA (Abbott) was performed on sonicated samples of three overnight cultures of E. coli each containing one of the three different HBsAg DNAs inserted into ORFl-secA (with the native C.difficile ORFl promoter) and cloned into the TA vector. Results: Cut off value = 1.0, higher values are regarded as positive.

1.3 : Full length HBsAg(pre SI, pre S2 and S gene), 2.8 : S gene,

3.7 : Partial S gene(the minimum HBsAg epitope) a : Duplicates

# : Samples stored overnight in cold before analysis. NT : Not tested

* : Culture OD at the time of harvest. Cells from 1ml culture were pelleted and resuspended in 0.5ml of PBS before sonication.

The above experiment was repeated with cells from 5ml cultures resuspended intolml PBS to see if increased protein concentrations would give higher titer values. Here a control E. coli culture with the TA vector carrying the ORFl-secA cassette but without any HBsAg insert was also included.

Concusion:

All three constructs containing HBV DNA expressed both soluble and and particulate HBsAg. Thus, the cloning proved to be successful and the native C. difficile ORFl promoter was to some extent functional also in E. coli.

Example 3

Cloning of the ORF1-ORF2 (secA) "cassette" into C. difficile and construction of recombinant protein useful for immunisation studies.

1. Cloning of a gene expression cassette according to examples 1 and 2 in plasmids pCI195 or pSMB47 followed by transfer to a non-toxigenic strain of C. difficile by mating via B. subtilis is performed according to standard methods reported (J. Gen. Microbiol 136:

1343-1349, 1990; Plasmid 31: 320-323, 1994, see also Figure 4). The engineered strain is isolated for further use.

2. Check for expression, secretion and antibody response in vitro and in vivo (see Example 2).

A. Cloning of the ORFl-HBsAg-s<?c i constructs into the shuttle plasmid vector pJIR750

Results:

Ligation mixtures containing the desired recombinant plasmids were obtained an judged by agarose gel electrophoresis and PCR. However, upon transformation into E. coli the plasmid constructs were fragmented. This indicated that a plasmid replication machinery better at handling large plasmids in E. coli than that of pJIR750 (colE based) needs to be used. Also, attempts to transform C. perfringens with our recombinant pJIR750 plasmids are being performed.

Example 4 Production of transformed C. perfrinsens expressing and presenting foreign antigen for vaccination.

A. Live bacteria.

C. perfringens, transformed with a gene cassette coding for a foreign antigen fused to ORFl and obtained as in Example 2 is cultivated under anaerobic conditions in a fermenter until a cell density of at least 10⁷ bacteria per ml is obtained. The broth is cooled to 11°C, the bacteria recovered by centrifugation and the supernatant discarded. The pellet is twice washed with cold 0.1 M phosphate buffer, pH 7 and centrifuged. The final pellet is resuspended in the phosphate buffer to a concentration of about 10⁹ organisms per ml. One ml portions of the suspension are dispensed into glass ampoules and freeze-dried to remove the water. The final product is obtained by sealing of the ampoules in vacuo.

B. Bacterial envelopes.

Transformed C. perfringens bacteria are produced as in A above. The final pellet is suspended in 50 mM Tris-HCl, pH 7.2, and sonicated for 1-10 min (40 watt, Bransic Sonic Power co. Sonicator). Triton X-100 is added to a final concentration of 2% and the mixture incubated under stirring at 11°C for 30 min.. The cells are collected by centrifugation and washed three times with cold distilled water. The pellet is resuspended in 5 mM MgCl, containing DNase (1 lmg/ml) and RNase (1 lmg/ml) and incubated for 15 min at 11°C. The resulting envelopes are recovered by centrifugation, washed three times with cold distilled water, resuspended in cold distilled water and freeze-dried to give the envelopes as a powder suitable for formulation in capsules or tablets, for suspension in e.g. physiological saline for oral administration.

Example 5

Production of Clostridial spores, germination of spores in vitro and in vivo, and colonization and immune response to C. difficile in animals given spores orally Clostridial strain producing the heterologous peptide is allowed to grow anaerobically in Peptone- Yeast extract-glucose or another medium optimal for sporulation for 48-72h to ensure maximum conversion of the vegetative bacteria into spores during the stationary phase. The remaining vegetative bacteria are killed by heat or ethanol treatment, eliminated by the bacteriolytic enzymes lysozyme or lysostaphin and the remaining spores are purified by centrifugation. Results:

Spores from C. difficile strain 630 were readily observed on 3-day old blood agar plates as well as in PYG medium, whereas spores from C. perfringens NCTC 8798 were found only during growth in Duncan Strong (DS) medium, after premocubation in Fluid Thioglycolate (FTG) medium.

Prepared spores (washed in ethanol and resuspended in sterile PBS) were checked for germination ability on plates (TCCFA) as well as in gnotobiotic rats. Feces from 1 week old rats were positive for bacterial growth in feces 1-2 days after receiving C. difficile or C. perfringens spores orally, confirming the ability of the spores to germinate also in vivo and lead to colonization of the animal gut.

Antisera from 5 rats colonised for one week by C. difficile were pooled and used for Western blotting of C. difficile protein extracts. Western blotting revealed immunological reactions to bands corresponding to the C. difficile S-layer proteins confirming that antibodies were produced against these C. difficile antigens upon feeding with spores, spore germination and colonization of the animals.

Example 6

Production of transformed C. difficile spores for use in oral immunisation. A. Capsules.

The C. difficile spores obtained according to Example 5 are mixed together with Mg stearate (1%) and lactose (30%), granulated in ethanol and compressed to tablets, containing

10⁶ spores, which are filled into gelatine capsules.

B. Powder for suspension. The C. difficile spores obtained according to Example 5 are mixed together with Mg stearate (1%) and carboxymethyl cellulose (25%), granulated in ethanol. The granulate is dried and dispensed into vials to give an amount of about 10⁶ spores in each vial. For oral administration the content of the vial is suspended in water or for example orange juice immediately before intake.

Example 7 Use of S-layer genes for epidemiological typing.

Present methods to detect and follow the spread of certain C. difficile strains in the environment and between infected patients (i.e. "finge rinting") include e.g. serotyping and PCR ribotyping. PCR ribotyping is a PCR based approach to amplify the region between the 16S and 23 S genes of C. difficile, and which has been shown to resolve and detect over 100 different patterns or strains. Different serotypes are likely to represent differences of the surface-exposed proteins, i.e. variations of S-layer proteins among strains. Our results with PCR amplification and following cleavage with restriction enzymes indicate that this region is present in almost all serogroups and that the cleavage pattern also varies among these groups (see Example 1C, (ii)). Thus, a molecular method including PCR combined with restriction enzyme cleavage or direct sequencing of the variable part of the ORFl or another part of this segment may be a method which is faster and more reliable than serotyping and in particular also more reliable than PCR ribotyping for fingeφrinting.

Example 8 Vaccination against CD AD

Immunity to CD AD after an episode of the infection is regarded to be short (months). This may be due to that anti-toxin antibodies are mainly of the serum IgG classes and not the secreted IgA class made to protect the gut mucosal surface, because the toxins are internalized by the gut mucosal cells (see above) and not by the M-cells specialized in furthering an IgA response. A further problem may be that immunity to the 20 C. difficile S-layer serotypes is required for prevention of colonization and thus the best protection against infection. For these reasons, it is likely that injectable vaccines against CD AD based on the toxins and under development may turn out to offer poor protection.

As an alternative, we provide a polyvalent live oral vaccine containing (i) the most prevalent toxin-producing serotypes (S-layer variants), here attenuated by knock-out of their toxin genes, and (ii) carrying a recombinant ORFl -sec - cassette encoding relevant parts of the toxin genes and (iii) an adjuvant peptide ensuring uptake of the immunogenic toxin epitopes by e.g. M-cells in order to obtain an IgA anti-toxin response. Example 9

Analysis of extracellular and membrane fraction proteins in Clostridium difficile bv two- dimensional PAGE, N-terminal sequencing and data base searches - identification of genes encoding the S-layer proteins

Identification of extracellular proteins

Analysis of its extracellular protein pattern by 2-D PAGE showed that two proteins of 50 kDa and 36 kDa were very abundant. Subsequent analysis of membrane preparations from C. difficile VPI 10643 corroborated the almost exclusive fractionation of the 50 kDa and 36 kDa proteins into the membrane fraction, when compared with the soluble fraction. Also in strain 630 two proteins with similar molecular weights but with different pi were abundant in the extracellular as well as in the membrane fraction. Thus, the 50 kDa and 36 kDa proteins were likely to constitute the C. difficile surface(S)-layer proteins, which are known to be partially shed from the bacterial surface into the culture supernatant (Luckevich and .

Beveridge, 1989; Tsukagoshi et al, 1984). The N-terminal sequence of spot no. 1 from VPI 10463 did not show any homology to other proteins in the C. difficile strain 630 genome database (Table 3). The N-terminal part of spot no. 2 showed similarity to an open reading frame encoding a 72 kDa protein in the C. difficile genome database (Table 3; ORFl, see also. Figure 1). However, only nine out of 15 amino acids matched close to the N-terminus of ORFl. Strikingly, the N-terminal sequences of the corresponding proteins from strain 630 were different from those of VPI 10463 and both matched to ORFl but at two different positions (spot No.10 and 11 in Table 1).

Several proteins were specifically found in PY cultures, i.e. during high toxin production (Table 3, spot no. 3, 4, 5, and 6). The N-terminal sequence of spot no. 3 matched with an ORF of 47.5 kDa in the C. difficile genome database. This ORF showed weak homology to a hypothetical protein in the Plasmodium falciparum genome database. The N- terminal sequence of spot no. 4 matched with an ORF of 39 kDa that showed homology to a phage-like element PBSX protein (XkdK) from Bacillus subtilis. The N-terminal sequences of spot no. 5 and 6 matched with two ORFs of 38 and 22 kDa, respectively, and these ORFs had the highest similarity to the FixB and FixA proteins from Escherichia coli. Spot no. 7, 8 and 9 were absent in PY but abundant in culture medium from PYG cultures. Spot no. 7 and 8 matched to an ORF located on the same contig as ORFl (Table 3; ORF3). The N-terminus of spot no. 9 matched to a central part of ORFl, and is likely to be a proteolytic fragment of a protein encoded by ORFl.

Analysis of the surface layer genes The identification of the S-layer genes revealed a genomic segment including seven genes (ORFl, 3, 5-7, 9 11) with significant homology to N-acetyl muramoyl L-alanine amidase (CwlB/LytC) and modifier protein of major autolysin (LytB) from Bacillus subtilis (Fig.l, Table 4). In addition to the LytB/LytC similarity, the N-terminal part of ORF6 showed similarity to eukaryotic cysteine proteases, and the highly expressed ORFl (above) showed weak similarity to S-layer proteins from Lactobacillus and Streptococcus spp. (Fig. 1). The N- terminus of ORFl contained a typical signal peptide for export via the Sec-dependent secretion and the predicted cleavage site was identical to that found in the protein sequence (not shown). However, no typical protein cleavage site was identified within ORFl that would allow processing of the 72 kDa protein further to give the finally sized S-layer proteins found (50 and 36 kDa). Strikingly, no significant match between the C. difficile S-layer ORFs and the S-layer homology motif (SLH domain) found in all presently known S-layer proteins was obtained (not shown). Most of the remaining genes in this genomic segment showed similarity to genes involved in secretion, polysaccharide and capsule synthesis (Fig. 1; Table 4). At least 2 other genomic sequence segments were found that contained genes similar to CwlB/LytC, indicating a complex variability (not shown).

Summary of results

The most dominant surface-exposed protein in many bacterial species is the S-protein. This protein crystallizes into a regular monolayer on the outside surface of the bacteria: the S- layer. The S-layers satisfy multiple roles for the cell and function as protective coats, as structures involved in cell adhesion and surface recognition, as molecular seives, as molecular and ion traps, as scaffolding for enzymes and as virulence factors (Sleytr and Beveridge, 1999; Sara and Sleytr, 2000). Though all S-layers share general features (all are made of relatively large proteins, self-assemble and are paracrystalline), comparative studies indicate that S-layers are non-conserved structures and are of limited taxonomical value (Kuen and Lubitz, 1996; Sleytr et al. 1999). Chemical analysis and genetic studies of a variety of S- layers have shown that they are composed of a single, homogenous protein or glycoprotein species with molecular weights ranging from 40 to 170 kDa. The S-layers of Clostridium difficile (Takeoka et al., 1991) and Bacillus anthracis (Mesnage et al., 1998; Etienne- Toumelin et al., 1995) consist of two types of S-layer subunits which together form a defined lattice type but do not cross-react with polyclonal antibodies. Typically, S-layer proteins are often weakly acidic proteins (pis between 4 to 6), containing 40-60% hydrophobic amino acids, and possess few or no sulfur-containing amino acids (Messner, 1996). S-protein production is directed by single or multiple promoters in front of the S-protein gene, yielding stable rnRNAs. Most bacteria secrete S-proteins via the general secretory pathway (sec- pathway). Silent S-protein genes have been found in Campylobacter fetus and Lactobacillus acidophilus. These silent genes are placed in the expression site in a fraction of the bacterial population via inversion of a chromosomal segment (Boot and Pouwels, 1996). The S-layer has been detected in some C.difficile strains and preliminary characterization has been done from C.difficile C253 (Mauri et al., 1999). Another independent study purified and identified the S-layer subunits from C.difficile GAI 0714 (Takeoka et al., 1991). In both cases, the S-layer has been shown to be composed of two different protein subunits with apparent molecular weights of 36 kDa and 47 kDa (C.difficile C253) and 32 kDa and 45 kDa (C.difficile GAI 0714). The S-layer proteins from VPI 10463 and strain 630 was here found to be similar in size but with significant pi differences. The N- terminal sequences varied significantly especially for the larger protein. The N-terminal sequences as determined for these proteins also indicate that they are not identical. Those from strain 630 appear to be processed products from the same gene (ORFl, Table 3). The N- terminal sequences of spot 1 from VPI 10643 did not find any homologue in the strain 630 database corroborating earlier results that the higher molecular weight S-layer protein was sero-group specific (VPI 10643, group G and strain 630, group X) (Takeoka et al., 1991). Our current work indicated that this arrangement of two S-layer proteins was also true for all the different serotypes tested (data not shown). The spot 2 found a partial match with another ORF (ORF3) in the same contig as ORF 1.

Other ORFs located in the same contig also had similarities with ORFl and ORF3, whose C-terminal parts showed similarities to N-acetyl muramoyl L-alanine amidase (CwlB/LytC) and modifier protein of major autolysin (LytB) from Bacillus subtilis (Lazarevic et al., 1992), whereas the N-terminal part showed weak similarities to surface-layer proteins from Lactobacillus helveticus (Callegari et al., 1998) and Streptococcus spp. It is interesting to note that N-acetyl muramoyl L-alanine peptidoglycan amidase is the major autolysin of B. subtilis and has high affinity for cell walls, which is enhaced by the modifier protein, but small amounts of cell free autolysin can be detected in cultures of B. subtilis. Thus, the amidase-like motif that appears to be typical of C.difficile S-layer proteins probably confers their anchorage to the cell wall peptidoglycan-teichoic acid.

Considering the highly competitive situation of closely related organisms in their natural habitats, it is obvious that the S-layer surface has to contribute to diversification rather than to conservation. With respect to this, the importance of S-layer variation leading to the expression of alternative S-layer genes under different stress factors such as those imposed by the immune system of a host in response to an S-layered pathogen or drastic changes in the growth and environmental conditions for nonpathogens is conceivable (Dworkin and Blaser, 1997; Sara et al., 1996). This could probably explain the variation in S-layer proteins even amongst the same species as in the case of C. difficile.

Identification of spot 4 as having similarity with XkdK, a protein encoded by the phage- like element PBSX from Bacillus subtilis (Krogh et al., 1996) and being located in a contig with other ORFs having similarity with other PBSX encoded proteins is very interesting. PBSX is a bacteriophage-like bacteriocin, or phibacin, of B. subtilis 168 (Okamoto et al., 1968). When B. subtilis 168 cells are exposed to treatments that induce the SOS response (such as UV light, mitomycin C), the cells lyse after incubation of lh and release particles of PBSX (Seaman et al., 1962). The spot 4 is completely absent in PYG supernatants. Taken together this could indicate that toxin production (high in PY) and expression of this phage- like protein in C. difficile is a response to certain stress, environmental or otherwise, that decides whether it will resort to toxin expression, sporulation or both.

The N-terminal sequences of spots 7 (41 kDa) and 8 (38 kDa) are identical (Table 3) and correspond to the same ORF (ORF3, Fig 1), whose N-terminal part is similar to N-acetyl muramoyl L-alanine amidase (CwlB/LytC) from B. subtilis. However, the size of the proteins in the gel do not match the size expected from the ORF3 (encodes a 67.5 kDa protein). Both ORFl and ORF3 have clear signal sequences at the beginning which is missing in the protein spots sequenced, thus indicating that these are indeed secreted and processed following translation. This could also possibly explain the difference in size and pi for the two different spots, and the discrepancy between expected and observed molecular weights on SDS-gels. The spot 9 (24 kDa) has a N-terminal sequence (Table 1) which corresponds to an internal fragment of ORFl. The expected size of this fragment is around 21 kDa which corresponds closely with what is observed experimentally. Clearly there are post-translational processing events which could be enacted in the cell envelope or in the supernatant. It is important however to note that spots 7-9 are present in PYG supernatants only, when the cells start sporulating.

The spots 9 and 10 are also processed products from ORFl and are present in both PY supernatant and membrane fractions. However, these samples are obtained from strain 630. The results obtained thus far indicate that this operon (contig) (Fig 1) is present in both VPI 10643 and strain 630, but different ORFs are expressed by the two strains.

Experimental procedures

Strains and growth medium The toxin-producing C. difficile strain VPI 10463 (CCUG 19126, Culture Collection,

University of Gδteborg, Sweden) was grown in either PY, PYG (purchased from the Karolinska Hospital, Stockholm, Sweden) or SDM medium. SDM is identical to MADM (Karasawa et al., 1995; Yamakawa et al., 1994; Yamakawa et al., 1996), except that the concentrations of glycine and threonine were 100 mg/L and 200 mg/L, respectively, and that Ca-D-panthotenate, pyridoxine and biotin were used as the sole vitamin sources. PY(G) was prepared by adding cysteine (500 mg/L), boiling for 20 min while purging with an anaerobic gas mixture (10% CO₂, 10% H₂, 80%) N₂) for 20 min, sterilised by filtration (Acrodisc, Gelman sciences) and aliquoted into tubes with serum vial-style necks (Bellco Glass) while flushing with anaerobic gas. The tubes were closed with butyl stoppers secured with aluminium crimp seals. SDM was prepared accordingly.

Growth conditions, sampling and optical density measurements

For each experiment, a tube containing 20 ml SDM was inoculated with 0.2 ml thawed bacterial suspension (stored at -70°C) using a syringe and a needle that was passed through the rubber septum of the tube. To avoid entry of O₂, the syringe was equilibrated with anaerobic gas before inoculation. The tube was put horizontally on a rotary shaker (50 rpm, 37 °C), and on the next day, the culture was serially diluted into PY or PYG. On day three, samples were collected from the diluted cultures and OD was measured at 600 nm using a Hitachi U-1100 spectrophotometer.

Sample preparation and membrane fractionation

Culture samples were centrifuged at 16000 x g for 3 min, whereafter the supernatants were removed, filtered, and stored at -20°C for later analysis. The pellet was frozen at -20°C for 30 min or longer, thawed, dissolved in 1 ml sterile water and sonicated on ice for 3 x 30 s at 100 W (Labsonic 1510, B. Braun). Larger cell pellets, obtained from >1 ml culture, was sonicated for longer times. The cell extracts were centrifuged at 5000 x g for 5 min. The pellet was separated as the low speed pellet (LSP), and the supernatant was further centrifuged at 50000 x g for 20 min. The pellet was separated as the high speed pellet (HSP), and the supernatant (soluble fraction) was stored at -20°C. The LSP and the HSP were resuspended in lx PBS (Phosphate buffered saline). Protein amount was measured using a kit (Biorad) and a BSA standard curve according to the manufactureris instructions. The culture supernatants were precipitated using trichloroacetic acid (TCA) to a final concentration of 10%. The pellets were washed with ice-cold Acetone, air dried and finally resuspended in lx PBS to obtain the extracellular protein fraction. Protein estimation and analysis was carried out as described earlier.

Immunoprecipitation Immunoprecipitation was performed in microtiter wells coated with antibodies against toxin A (PCG-4, r-Biopharm) or toxin B (xxx, r-Biopharm), Ten μg/ml antibody in 0.04 M Na₂CO₃, 0.06 M NaHCO₃, pH 9.6 was added to microtiter wells and incubated for 1 h at 37°C and washed four times with PBS containing 0.05% (v/v) Tween-20, pH 7.4. The wells were loaded with cell extract, culture supernatant medium or PBS (negative control), incubated 90 min at 25°C, and washed four times with PBS. After addition of 50 μl SDS sample buffer solution (below) and heating for 5 min at 95°C, the precipitated proteins were analysed by SDS-PAGE.

SDS-polyacrylamide gel electrophoresis (SDS-PAGE SDS-PAGE was performed using pre-cast polyacrylamide gels (ExcelGel 8-18% gradient gels, Pharmacia Biotech) and a Multiphor II horizontal slab gel apparatus (Pharmacia Biotech) according to the manuals provided by the manufacturer. The samples were mixed 1 : 1 with SDS sample buffer solution (0.05 M Tris, 1% (w/v) SDS, 10 mM DTT, 0.01% (w/v) bromophenol blue, pH 7), incubated 5 min at 95°C, loaded onto the gels and run at 15°C. Chemicals were obtained from Sigma, and molecular weight markers from Pharmacia Biotech. The gels were stained with silver (PlusOne, Pharmacia Biotech) using a Hoefer automatic gel stainer (Pharmacia Biotech), digitised by scanning (Scanjet 3c/T, Hewlett- Packard), and transferred to ClarisDraw (Claris Software) on a Macintosh computer. Immunoblotting

Proteins were separated by SDS-PAGE transferred to polyvinylidene fluoride membranes (Immobilon P^SEQ, Millipore) using the Pharmacia Novablot transfer equipment and a continuous buffer system (39 mM glycine, 48 mM Tris, 0.0375% (w/v) SDS, 20% (v/v) methanol) according to the Multiphor II manual. The membranes were dried at 25°C for 1.5 h, blocked with 0.5% Tween-20 for 20 min, and then incubated with toxin A or toxin B antibodies (r-Biopharm, 0.2 μg/ml in TST buffer containing 0.05 M Tris, 0.5 M NaCl, 0.1 % Tween-20, pH 9) for 1 h. After three washes in TST, the membranes were incubated with horse-radish peroxidase conjugated anti-mouse antibodies (DAKOPATTS, diluted 1 : 10000 in TST) for 1 h and washed three times in TST. A chemiluminiscent signal (ECL Plus, Amersham) was used to detect the bands. The relative amount of toxin B was measured on scanned x-ray films using the Molecular Analyst software (Biorad).

Two-dimensional polyacrylamide gel electrophoresis (2-D PAGE)

Protein samples were obtained as described under sample preparation. For 2-D PAGE, 40 ml aliquots of each sample was mixed with 160 ml of buffer III [9.9M Urea, 4% (v/v) Igepal CA630, 2.2% (v/v) ampholytes 3-10, lOOmM DTT, 2% (w/v) CHAPS]. Protein mixtures were focused at 20°C on 180mm IPG Drystrip pH 4-7 (Pharmacia Biotech) using the = Multiphor II 2-D gel Kit according to the manufactureris instructions. The second dimension was run on 12% (linear) SDS acrylamide gels. The gels were stained with silver (PlusOne, Pharmacia Biotech) using a Hoeffer automatic gel stainer (Pharmacia Biotech). The chemicals were obtained from Sigma except for Pharmalytes (Pharmacia Biotech).

N-terminal sequencing

For N-terminal Sequence determination, gels were transferred to Immobilon-P polyvinyline difluoride membrane (Millipore) as described under immunoblotting. The membrane was stained with Coomassie blue; protein spots were excised for sequence determination. The protein spots cut from the transfer membrane were washed four times in 10% methanol and then dried and frozen. N-terminal sequence analysis was performed at the Protein Analysis Center, Karolinska Institute. Peptide sequences were matched against the C.difficile genome database (Sanger Center, UK). References

Bardwell, J.C.A., Lee, J.-O., Jander, G., Martin, N., Belin, D., and Beckwith, J. (1993) A pathway for disulfide bond formation in vivo. Proc Natl Acad Sci USA 90: 1038-1042. Bardwell, J.C.A., McGovern, K., and Beckwith, J. (1991) Identification of a protein required for disulfide bond formation in vivo. Cell 65: 581-589.

Boot, H.J., and Pouwels, P.H. (1996) Expression, secretion and antigenic variation of bacterial S-layer proteins. Mol Microbiol 21: 1117- 1123. Callegari, L., Riboli, B., Sanders, P., Coconelli, S., Kok, J., Venema, G., and Morelli L.(1998)

The S-layer gene of Lactobacillus helveticus CNRZ 892 : cloning, sequencing and heterologous expression. Microbiology 144: 719-726.

Cerquetti, M., Pantosti, A., Stefanelli, P., and Mastrantonio, P. (1992) Purification and characterization of an immunodominant 36 kD antigen present on the cell surface of

Clostridium difficile. Microb Pathogen 13: 271-279. Dauter, Z., Wilson, K.S., Sieker, L.C., Moulis, J.M., and Meyer, J. (1996) Zinc- and iron- rubredoxins from Clostridium pasteurianum at atomic resolution: a high-precision model for ZnS4 coordination unit in a protein. Proc Natl Acad Sci USA 93: 8836-8840. Dworkin, J., and Blaser, M.J. (1997) Molecular mechanisms of Campylobacter fetus surface layer protein expression. Mol Microbiol 26: 433-440. Economou, A. (1999) Following the leader : bacterial protein export through the Sec pathway. Trends Microbiol 7: 315-319.

Etienne-Toumelin, I., Sirard, J.-C, Duflot, E., Mock, M., and Fouet, A. (1995)

Characterization of the Bacillus anthracis S-layer : cloning and sequencing of the structural gene. J. Bacteriol. Ill : 614-620. Gamier, T., and Cole, S.T. (1988) Studies of UV-inducible promoters from Clostridium perfringens in vivo and in vitro. Mol Microbiol 2: 607-614.

Izard, J.W., and Kendall, D.A. (1994) Signal peptides: exquisitely designed transport promoters. Mol Microbiol 13: 765-73. Kamiya, S., Ogura, H., Meng, X.Q., and Nakamura, S. (1992) Correlation between cytotoxin production and sporulation in Clostrdidium difficile. J Med Microbiol 37: 206-210. Karasawa, T., Ikoma S., Yamakawa, K., and Nakamura, S. (1995) A defined medium for

Clostridium difficile. Microbiology 141: 371-375. Kawata, T., Takeoka, A., Takumi, K., and Masuda, K. (1984) Demonstration and preliminary characterization of a regular array in the cell wall of Clostridium difficile. FEMS Microbiol

Lett 24: 232-328. Klauser, T., Pohlner, J., and Meyer, T.F. (1993) The secretion pathway of IgA protease-type proteins in gram-negative bacteria. Bioessays 15: 799-805.

Krogh, S., M. O'Reilly, N. Nolan, and K.M. Devine. (1996) The phage-like element PBSX and part of the skin element, which are resident at different locations on the Bacillus subtilis chromosome, are highly homologous. Microbiology 142 : 2031-2040. Lazarevic, V., Margot, P., Suldo, B., and Karamata, D. (1992) Sequencing and analysis of the Bacillus subtilis lytRABC divergon : a regulatory unit encompassing the structural genes of the N-acetylmuromoyl-L-alanine amidase and its modifier. J Gen Microbiol 138: 1949-

1961. Luckevich, M.D., and Beveridge, TJ. (1989) Characterization of a dynamic S-layer on

Bacillus thuringiensis. J. Bacteriol. 171 : 6656-6667. Masuda, K., Itoh, M., and Kawata, T. (1989) Characterization and reassembly of a regular array in the cell wall of Clostridium difficile GAI 4131. Microbiol Immunol 33: 287-298. Mauri, P.L., Pietta, P.G., Maggioni, A., Cerquetti, M., Sebastianelli, A., Mastrantonio, P.

(1999) Characterization of surface layer proteins from Clostridium difficile by liquid chromatography/ electrospray ionization mass spectrometry. Rapid Commun. Mass Spectrom. 13: 695-703.

McFarland, L.V., Elmer, G.W., Stamm, W.E., and Mulligan, M.E. 1991. Correlation of immunoblot type, enterotoxin production, and cytotoxin production with clinical manifestations of Clostridium difficile infection in a cohort of hospitalized patients. Infect

Immun 59: 2456-2462. Mesnage, S., Tosi-Couture, E., Gounon, P., Mock, M., and Fouet, A. (1998) The capsule and the S-layer: two independent and yet compatible macromolecular structures in Bacillus anthracis. J. Bacteriol. 180 : 52-58. Messner, P. (1996) Chemical composition and biosynthesis of S-layers, p.35-76. In

U.B.Sleytr, P.Messner. D.Pum, and M.Sara (ed.), Crystalline bacterial cell surface layer proteins (S-layers). Academic Press, Landes Company, Austin, Tex.

Nicholls, L., Grant, T.H., and Robins-Browne, R.M. (2000) Identification of a novel genetic locus that is required for in vitro adhesion of a clinical isolate of enterohaemorrhagic

Escherichia coli to epithelial cells. Mol Microbiol 35: 275-288. Orme- Johnson, W.H., and Orme- Johnson, N.R. (1978) Overview of iron-sulfur proteins. Methods Enzymol 53: 259-268.

Pugsley, A.P. (1993) The complete general secretory pathway in Gram-negative bacteria. Microbiol Rev 57: 50-108. Salra, M., and Sleytr, U.B. (2000) S-layer proteins. J. Bacteriol. 182 : 859-868.

Salra, M., Kuen, B., Mayer, H.F., Mandl, F., Schuster, K.C., and Sleytr, U.B. (1996) Dynamics in oxygen induced changes in S-layer protein synthesis from Bacillus stearothermophillus PV72 and the S-layer deficient variant T5 in continuous culture and studies on the cell wall composition. J. Bacteriol. 178 : 2108-2117. Sellman, B.T., and Tweten, R.K. (1997) The propeptide of Clostridium septicum alpha toxin functions as an intramolecular chaperone and is a potent inhibitor of alpha toxin-dependent cytolysis. Mol Microbiol 25: 429-440. Simonen, M., and Palva, I. (1993) Protein secretion in Bacillus species. Microbiol Rev 57: 109-137. Takemaru, K., M. Mizuno, T. Sato, M. Takeuchi, and Y. Kobayashi. (1995) Complete nucleotide sequence of a skin element excised by DNA rearrangement during sporulation in Bacillus subtilis. Microbiology 141 : 323-327.

Takeoka, A., Takumi, K., Koga, T., and Kawata, T. (1991) Purification and characterization of S layer proteins from Clostridium difficile GAI 0114. J Gen Microbiol 137: 261-267. Takumi, K., Takeoka, A., and Kawata, T. (1987) Purification and immunochemical properties of a wall protein antigen from Clostridium difficile ATCC 110011. Microbiol Immunol 31:

837-849. Tsukagoshi, N., Tabata ., Takemura, T., Yamagata, H., and Udaka, S. (1984) Molecular cloning of a major cell wall protein gene from protein producing Bacillus brevis 47 and its expression in Escherichia coli and Bacillus subtilis. J. Bacteriol. 158 : 1054-1060.

Welch, R.A. (1991) Pore-forming cytolysins of gram-negative bacteria. Mol Microbiol 5: 21-

528. Table 1. Identification of ORF 1-12 and sequences upstream of ORFl in SEQ ID NO: 35 and predicted signal peptide sequences of the S-layer ORFs. See Figures 1 and 2.

Gene Frame from to Total bp (aa) Homoiogy/homologous motif (aa domain) Predicted signal peptide

A +2 227 1366 1140 tRNA- Guanine transglycolase

B +3 1500 1793 294 conserved protein (membrane bound)

C +3 1851 2213 363 Hypothetical protein YrzE

D +1 3046 4941 1896 (631) N-acetylmuramoyl-L-alanine amidase (aa 320-610) MLSNKK.11SMAIVTVIAGATVMSAAAPIFA4.DNTVTEN SEQ ID NO: 8

E +3 5373 6806 1434 (477) N-acetylmuramoyl-L-alanine amidase (aa 20-290) MKSTLGVENNMKNSKKILAIGLTLFLVMVNTPMV SA -LTSVE SEQ ID NO: 9

F +2 7187 131505964 Unknown

G +1 13360 15240 1881 (626) N-acetylmuramoyl-L-alanine amidase (aa 320-620) MNKRKSFIRTIAVSTMAVAVTGSATCAYAiAPVLQ SEQ ID NO: 10

H +1 15556 17208 1653 (550) N-acetylmuramoyl-L-alanine amidase (aa 5-370) MENrøπ^INIKYKNHQGDMKMN KILSLGLAVSLIL

VNFKSVNAiSSW SEQ ID NO: 11

I +3 17427 19013 1587 (528) N-acetylmuramoyl-L-alanine amidase (aa 20-250) MKVNKRVLSIGLAISLIMAGAPNINA-lLSSIE SEQ ID NO: 12

ORFl +3 19224 21383 2160 (719) N-acetylmuramoyl-L-alanine amidase (aa 380-719) MNKKNIAIAMSGLTVLASAAPWA ATTGT SEQ ID NO: 1

ORF2 +1 21628 23973 2346 SecA SEQ ID NO: 34

ORF3 +1 24388 26259 1872 (622) N-acetylmuramoyl-L-alanine amidase (aa 310-540) MNKKNLSVIMAAAMISTSVAPVFA4AETTQ SEQ ID NO: 2

ORF4 +2 26420 27073 654 Unknown

ORF5 +1 27106 28938 1833 (610) N-acetylmuramoyl-L-alanine amidase (aa 5-300) MKISKKIVSLLTMTFLTVTLYGNTSNAlSTKDT SEQ ID NO: 3

ORF6 +3 29613 32024 2412 (803) N-acetylmuramoyl-L-alanine amidase (aa 500-800) MRKYKSKKLSKLLALSTVCFLIVSTIPVSA4-ENHK SEQ ID NO: 4

ORF7 +2 32321 33898 578 (525) N-acetylmuramoyl-L-alanine amidase (aa 30-340) MKAPKTILTILTIALTLSSISIIPSYAiLTEEK SEQ ID NO: 5

ORF8 +3 34032 35351 1320 GDT1 protein

ORF9 +1 35590 37629 2040 (528) N-acetylmuramoyl-L-alanine amidase (aa 5-285) MRGDMMEKTTKXLATGMLSVAMVAP1WALA ENTT SEQ ID NO: 6

ORF10 +1 37873 38535 699 Hypothetical protein

ORF11 +2 38630 39685 1056 (351) N-acetylmuramoyl-L-alanine amidase (aa 40-340) MIKKISΗLSLVLLISISSTIGWA .DANPKR SEQ ID NO: 7

ORF12 +2 39800 41356 1557 Virulence Factor MviN aa = no. of amino acids or their position in translated ORF I = predicted cleavage point

Table 2. Possible promoter starts between ORF I and ORF 1. (1.00 is highest score). ORF 1 start codon is at position 19224.

Position Score Sequence

19026-19071 0.91 TAGTTTATTACATTTTAAAATTTAGGGTATAAAAACTTGTAAACTTGGAG SEQ ID NO : 13

19036-19081 1.00 CATTTTAAAATTTAGGGTATAAAAACTTGTAAACTTGGAGaAAATAATAA SEQ ID NO : 14

19059-19104 0.80 AACTTGTAAACTTGGAGaAAATAATAATTTAAAAAAATAGCTTGCAAaAA SEQ ID NO : 15

19067-19112 0.99 AACTTGGAGaAAATAATAATTTAAAAAAATAGCTTGCAAaAAGAATAAAA SEQ ID NO : 16

19083-19128 1.00 TAATTTAAAAAAATAGCTTGCAAaAAGAATAAAAATGGATTATTATAGAG SEQ ID NO : 17

19094-19139 0.99 AATAGCTTGCAAaAAGAATAAAAATGGATTATTATAGAGATGTGAGAaAT SEQ ID NO : 18

19123-19168 0.96 TATTATAGAGATGTGAGAaATATTagGaATATATGGATGATTATTCTATG SEQ ID NO : 19

19132-19177 0.85 GATGTGAGAaATATTagGaATATATGGATGATTATTCTATGtAcATAATA SEQ ID NO : 20

19161-19206 0.84 GATTATTCTATGtAcATAATAAAGAGATGTAATTTTAATATAATGTTGGG SEQ ID NO : 21

Table 3. Extracellular C. difficile proteins identified by N-terminal sequence analysis and genome database searches.

Spot MW Source N-terminal Sequence³ Highest match" Comment

1 50 PY AAKASIADENSPVKLT No match LKSDXKKDL

2 36 PY DDTKVETGDQGYTVV N-acetyl— (ORFl) Partly matched

3 47 PY SEKEILTARLAV Hypothetical protein Fragment

4 40 PY AIGLPSΓNISSK PBSX phage protein, XkdK

5 39 PY MXDIKLDXFXKX Fix B

6 30 PY MKILVXVKQVXX Fix A

7 41 PYG AETTQVKKETIT N-acetyl muramoyl-L-alanine amidase Fragment (ORF3)

8 38 PYG AETTQVKKETIT N-acetyl— (ORF3) Fragment

9 24 PYG TSLKIADEVGLD N-acetyl— (ORFl) Fragment

10 48 PY ANDTIASQDTPAK N-acetyl— (ORFl)/ S-layer protein Fragment (central part)

11 35 PY ATTGTQGYTWKNDD N-acetyl— (ORFl )/S-layer protein Fragment (N-terminal) underlined sequence indicates that only this part was found to be identical with the ORF in the C. difficile strain 630 genome database ^bSearch was made in the C. difficile strain 630 genome database by tBLASTn algorithm at www.. Contigs were exported to ORF finder at NCBI, and the entire ORF was subjected to further BLAST searches against the redundant database.

Table 4. Blast search summary of the ORFs in the C. difficile strain 630 S- layer contig. ORF numbers refer to those in Fig 1. ORF Length (amino acids) Match (highest E-value)¹ Source E value

1 719 N-acetylmuramoyl-L-alanine amidase B. subtilis 1x10 zr

2 781 Preprotein translocase SecA subunit B. subtilis 0.0

3 623 N-acetylmuramoyl-L-alanine amidase B. subtilis 1x10^' -24

4 217 No match

5 610 N-acetylmuramoyl-L-alanine amidase B. subtilis 2x10^" -42

6 803 N-acetylmuramoyl-L-alanine amidase B. subtilis 2x10- -36

7 525 N-acetylmuramoyl-L-alanine amidase B. subtilis 2xl0- ■39

8 439 GDT1 protein D. discoideum 2x10^'

9 679 N-acetylmuramoyl-L-alanine amidase B. subtilis 4xl0- ■56

10 232 Hypothetical protein B. subtilis 7x10 ,-- 5

11 351 N-acetylmuramoyl-L-alanine amidase B. subtilis 1x10 ,^"-π

12 518 Virulence factor MviN S. typhimurium 2x10^" •53

13 568 Phosphomannose murase B. subtilis ₁₀-159

14 352 Mannose- 1 -phosphate guanyltransferase E. coli 5xiσ -68

15 157 Cap8J S. pnemoniae 5x10^"

16 385 No match

17 279 Glucosyl transferase S. pnemoniae 5_Xr^r

Database search was made using the BLAST algorithm at http://www.ncbi.nlm.nih.gOv/.com. Matches with a higher E value than 0.001 was regarded as no match.

Claims

1. A gene expression cassette comprising a secretory leader sequence encoding a signal peptide from Clostridium difficile having an amino acid sequence selected from the group consisting of SEQ ID NO: 1 , SEQ ID NO: 2, SEQ ID NO: 3, SEQ ED NO: 4, SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7 and signal peptides of analogous exported clostridial N-acetylmuramoyl-L-alanine amidase-like proteins, linked to a DNA sequence encoding a heterologous polypeptide.

2. A gene expression cassette according to claim 1, wherein the signal peptides of the analogous clostridial N-acetylmuramoyl-L-alanine amidase-like proteins are selected from Clostridium difficile signal peptides having an amino acid sequence of any one of SEQ ID NO: 8, SEQ ID NO: 9, SEQ ID NO: 10, SEQ LD NO: 11, and SEQ ID NO: 12.

3. A gene expression cassette according to claim 1 or 2, which further includes a promoter of prokaryotic origin.

4. A gene expression cassette according to claim 3, wherein the promoter is selected from clostridial promoters comprising a nucleotide sequence selected from the group consisting of SEQ ID NO: 13 - 21.

5 . A gene expression cassette according to any one of claims 1-4 which further includes a DNA sequence encoding at least a cell wall binding portion of a protein of prokaryotic origin functioning in clostridia such that a fusion polypeptide may be presented on the outer surface of a host cell harbouring the cassette.

6. A gene expression cassette according to any one of claims 1-4, which further includes a DNA sequence encoding at least a functional cell wall binding portion of an S-layer protein of C. difficile selected from any one of the polypeptides having an amino acid sequence selected from SEQ ED NO: 22 - 33 such that a fusion polypeptide may be presented on the outer surface of a host cell harbouring the cassette.

7. A gene expression cassette according to claim 5 or 6, wherein DNA encoding the cell wall binding portions of SEQ ED NO: 22-33 has been omitted such that the fusion peptide is secreted into the surrounding milieu by the host cell harbouring the cassette.

8. A gene expression cassette according to claim 6 or 7, wherein the DNA sequence encoding the heterologous peptide is inserted at a point downstream the first (signal) proteolytic cleavage site in the gene encoding a polypeptide having an amino acid sequence selected from SEQ ID NO: 22 - 33, optionally including or excluding its second cleavage site.

9. A gene expression cassette according to any one of claims 1-8, which further comprises at least a functional part of a secretory (secA) gene recognizing the signal peptide, to allow translocation of a heterologous polypeptide and/or fusion polypeptide across the cytoplasmic membrane of a host cell harbouring the expression cassette.

10. A gene expression cassette according to claim 9, wherein the secretory gene is from

C. difficile and encodes a polypeptide having the amino acid sequence SEQ ED NO: 34.

11. A gene expression cassette as shown in Figure 3.

12. A vector comprising a gene expression cassette as claimed in any one of claims 1-11.

13. A vector according to claim 12, wherein the vector is a plasmid.

14. A host organism transformed with a vector according to claim 12 or 13 for expression of the heterologous polypeptide and/or fusion polypeptide.

15. A Clostridium host organism transformed with a vector according to claim 12 or 13 for expression of the heterologous polypeptide and/or fusion polypeptide.

16. A host organism as claimed in claim 15 which is C. difficile.

17. A host organism as claimed in claim 15 which is C. perfringens.

18. A pharmaceutical or veterinary composition or formulation which comprises Clostridial cells transformed with a vector according to claim 12 or 13, with the ability to present on the cell surface and/or to secrete an expressed heterologous polypeptide or fusion polypeptide, together with a pharmaceutically or veterinary acceptable carrier or diluent.

19. A composition or formulation according to claim 18, which is suitable for oral or intranasal administration.

20. A composition or formulation according to claim 18 or 19, which further comprises, as an adjuvant, non-toxic motifs of the C. difficile toxins A and/or B that enable the expressed heterologous polypeptide and/or fusion polypeptide to enter the colonic mucosal cells of a mammal by receptor-mediated endocytosis, and/or a portion of toxin B responsible for its intracellular and intercellular spread.

21. A composition or formulation according to claim 18 or 19, which further comprises a further fused or separate carrier peptide or adjuvant, in addition to the expressed heterologous polypeptide and/or fusion polypeptide, to elicit a stronger or differently directed immune response than that against the expressed heterologous polypeptide acting alone.

22. A vaccine which comprises a Clostridial bacterium transformed with a vector according to claim 12 or 13 and capable of presenting on the surface of the bacterium and/or secreting an antigen in a human or animal body, and optionally also an adjuvant described in claim 20 or 21.

23. A medicament which comprises a Clostridial bacterium transformed with a vector according to claim 12 or 13 and capable of presenting on the surface of the bacterium and/or secreting a heterologous polypeptide and/or fusion polypeptide in a human or animal body, and optionally also an adjuvant described in claim 20 or 21.

24. A vaccine according to claim 22, which comprises a mixture of at least two differently engineered Clostridia strains, each capable of presenting on the surface of the bacterium and/or secreting a different heterologous polypeptide and/or fusion polypeptide.

25. A medicament according to claim 23, which comprises a mixture of at least two differently engineered Clostridia strains, each capable of presenting on the surface of the bacterium and/or secreting a different heterologous polypeptide and/or fusion polypeptide.

26. A vaccine which comprises spores of Clostridia cells or bacteria transformed with a vector according to claim 12 or 13, and capable of germinating into cells which are able to grow, express, and present or secrete a heterologous polypeptide and/or fusion polypeptide, and optionally an adjuvant described in claim 20 or 21, in a mammalian body.

27. A medicament which comprises spores of Clostridia cells or bacteria transformed with a vector according to claim 12 or 13, and capable of germinating into cells which are able to grow, express, and present or secrete a heterologous polypeptide and/or fusion polypeptide and optionally an adjuvant described in claim 20 or 21, in a mammalian body.

28. A vaccine according to claim 26, which comprises a mixture of spores from at least two differently engineered Clostridia strains.

29. A medicament according to claim 27, which comprises a mixture of spores from at least two differently engineered Clostridia strains.

30. A method for vaccination of a mammal, which comprises administering a therapeutically or prophylactically effective dose of a vaccine according to any one of claims 22, 24, 26 and 28 to the mammal.

31. A method for prophylactic or therapeutic treatment of a mammal, which comprises administering a therapeutically or prophylactically effective dose of a medicament according to any one of claims 23, 25, 27 and 29 to the mammal.

32. A vaccine according to claim 26 or 28, a medicament according to claim 27 or 29, or a method according to claim 30 or 31, wherein the spores are from C. difficile or C. perfringens.

33. A C. ztϊczVe-associated diarrhea (CD AD) vaccine comprising spores according to claim 26 or 32 and capable of expressing, after germination,

(i) relevant parts of the C. difficile toxins, alone or together with a

(ii) suitable adjuvant to provide primarily an IgA response to the toxin antigenic epitopes and/or (iii) S-layer protein antigenic variants (serotype antigens) or fimbrial antigens firom C. difficile to obtain, after administration to a mammal, a polyvalent anti-S-layer (or anti- fimbrial) IgA response preventing C. difficile colonization of the mammal.