WO1998059053A1 - Immonogenic fragments of toxin a of clostridium difficile - Google Patents

Abstract

The present invention relates to the 14 repeat C-terminal region of Clostridium difficile Toxin A that is effective in generating anti-toxin A antibodies.

Description

IMMUNOGENIC FRAGMENTS OF TOXIN A OF CLOSTRIDIUM DIFFICILE

The present invention relates to immunogenic fragments of Toxin A of

Clost^ήdium difficile, methods for their preparation and their use as vaccines.

Clostήdium difficile is an important gastrointestinal pathogen causing infection in both hospitals and tertiary care centres. Disease generally occurs when the gastrointestinal flora of individuals is disrupted through the use of broad- spectrum antibiotics such as cephalosporins and ampicillin, which subsequently allows C. difficile to flourish within the intestine. The organism can form spores and contaminate the hands and clothing of health-care personnel quite readily, thereby posing a significant health care problem C. difficile causes a range of antibiotic associated disease, with symptoms ranging from mild diarrhoea to pseudomembraneous colitis which can be life threatening. Common pathological features of the disease include fluid accumulation, inflammation and necrosis of gastrointestinal mucosa. These pathological symptoms have been shown to be caused by two large proteinaceous toxins produced by C. difficile, termed toxin A and toxin B.

Toxin A is slightly larger than toxin B having molecular weights of 308 kDa and 270 kDa respectively. Studies carried out in various animal models of C. difficile disease have shown toxin A to be the primary mediator of tissue damage within the intestine. Since toxin B has been shown to be an extremely potent cytotoxin against several cell lines in vitro, it is believed that toxin B acts after this initial toxin A-mediated damage, thus exacerbating the mucosal tissue damage.

A striking feature of both toxin A and B is the repetitive nature of the amino acid sequences located at the carboxyl terminus of the protein. In the case of toxin A, there are 38 tandem repeat sequences which are classified into class I or class II repeats based on their size. The class II repeats are further subdivided into 4 groups on sequence homology. Work carried out with a recombinant peptide coding for 33 of the repeats showed the receptor-binding domain of the toxin to be located within this repeat region. Antiserum raised against this region was found to be able to neutralise the cytotoxic activity of toxin A (Lyerly et al Curr. Microbiol. 21 29-33 (1990)). In addition, studies utilising a synthetic decapeptide corresponding to a hydrophilic sequence conserved within the class IIB repeats showed that even this shoπ sequence possessed a receptor-binding capability, and that antiserum raised against the peptide could partially inhibit the binding and cytotoxic activity of whole toxin A (Wren et al Infect. Immun. 59 3151-3155 (1991)). These studies show the potential the carboxy-terminal repeat region has as a vaccine against C. difficile disease.

Although toxin A has been shown to be immunogenic in both animals and humans, the vast majority of animal vaccine studies performed to date have used parenteral routes of immunisation which generate a systemic anti-toxin response (Lyerly et al Curr. Microbiol. 21 29-33 (1990)). Subsequently, when these animals are challenged with intact C. difficile, only partial levels of protection are seen since toxin-mediated damage is allowed to occur at the gastrointestinal mucosa due to the absence of a relevant anti-toxin response initiated by the mucosal immune system. The majority of C. difficile vaccine studies have utilised chemically detoxified whole toxin as the immunogen. However, this is undesirable due to the random structural and chemical modifications which occur to the toxin during this treatment, plus the possibility of residual toxicity within the protein. The repeat region of toxin A has been shown to be an ideal vaccine candidate due to its immunogenicity and non-toxicity. However, the protein encoding the entire repeat region is relatively large and therefore would be expected to be problematic with regard to retaining its structural integrity within recombinant expression systems. There is therefore a need to achieve a compromise between the requirement for immunogemcity and structural integrity of a protein to be used as pan of a vaccine. International PCT Application WO96/12802 discloses fusion proteins comprising the C-teπninus repeats of C. difficile. These constructs are shown to be capable of generating anti-toxin A antibodies upon administration to Syrian golden hamsters. The whole C-terminal region i.e. comprising at least 36 repeats is described as necessary in order to efficiently prepare an agent against C. difficile. Sub-fragments of this region were shown not to be suitable as they were insoluble, unstable or failed to generate a suitable response. Of the whole C-terminal region, the N-terminal part was shown to be the least essential and the C-terminal part as critical. There was no disclosure of a C-terminal sub- fragment that could efficiently generate antibodies against C. difficile toxin A.

It has now been surprisingly found that a particular fragment of the C-terminal region can be expressed within recombinant expression systems to produce non- toxic immunogenic proteins which have the ability to act as a vaccine against C. difficile infection.

According to a first aspect of me present invention, there is provided a molecule which:

a) comprises an amino acid sequence as shown in Figure 6 ;

b) has one or more amino acid substitutions, deletions or insertions relative to a sequence as defined in a) above;

c) a fragment of a sequence as defined in a) or b) above, which is at least ten amino acids long;

d) comprises a multiple of a sequence as defined in a), b) or c); or

a sequence substantially homologous thereto and wherein said molecule is capable of eliciting an immune response in an animal.

The skilled person is able to determine whether or not a paπicular fragment is immunogenic by standard immunogenicity assessment techniques known in the art. Molecules of the present invention may be in any appropriate form. They may be proteins, polypeptides, or peptides and may be fused to other moieties. As will be described below, the amino acid sequence shown in Figure 6 (SEQ ID No. 1 ) comprises 14 repeats from the C-terminal region of C. difficile toxin A. This sequence has been demonstrated to be superior to the whole C-terminal repeat region and other sub-fragments thereof in generating immunity to C. difficile.

The molecules of the present invention may be provided in substantially pure form. Thus a molecule of the present invention may be provided in a composition in which it is the predominant component present (i.e. it is present

at a level of at least 50% ; preferably at least 75 % , at least 90% , or at least 95 % ; when determined on a weight/weight basis excluding solvents or carriers).

In order to more fully appreciate the present invention, molecules within the scope of a), b) or c) above will now be discussed in greater detail.

A molecule within the scope of a) may consist of the paπicular amino acid sequence given in Figure 6 , or may have an additional N- terminal and/or an additional C-terminal amino acid sequence.

Additional N-terminal or C-terminal sequences may be provided for various reasons. Techniques for providing such additional sequences are well known in the art. Additional sequences may be provided in order to alter the characteristics of a paπicular polypeptide. This can be useful in improving expression or regulation of expression in paπicular expression systems. For example, an additional sequence may provide some protection against proteolytic cleavage. This has been done for the hormone Somatostatin by fusing it at its N-terminus to pan of the β-galactosidase enzyme (Itakwa et al., Science 198: 105-63 (1977)).

Additional sequences can also be useful in altering the propeπies of a polypeptide to aid in identification or purification. For example a fusion protein may be provided in which a polypeptide is linked to a moiety capable of being isolated by affinity chromatography. The moiety may be an antigen or an epitope and the affinity column may comprise immobilised antibodies or immobilised antibody fragments which bind to said antigen or epitope (desirably with a high degree of specificity). The fusion protein can usually be eluted from the column by addition of an appropriate buffer.

The molecules of the present invention may be formulated as vaccines. Fusion proteins comprising these molecules can be prepared with immunogenic peptides from other sources. An example of such a fusion protein comprises a molecule of the present invention and tetanus toxin, suitably the immunogenic fragment C of tetanus toxin (Khan et al PNAS USA 91 11261-11265 (1994)). Other fusion proteins may comprise immunogenic peptides commonly used in vaccines against Haemophilus influenzae, Helicobacter pylori, diphtheria, cholera, whooping-cough or typhoid.

Additional N-terminal or C-terminal sequences may, however, be present simply as a result of a paπicular technique used to obtain a molecule of the present invention and need not provide any paπicular advantageous characteristic to the molecule of the present invention. Such molecules are within the scope of the present invention.

Whatever additional N- terminal or C-terminal sequence is present, it is prefeπed that the resultant molecule has at least a substantial propoπion of the immunogenic activity of the molecules having the amino acid sequence shown in Figure 6. The term "at least a substantial propoπion of activity" when used herein means at least 50% of the activity of a given molecule (preferably at least 75% of said activity, more preferably at least 90% of said activity, and most preferably the same level of activity or a greater level of activity).

Turning now to the molecules defined in b) above, it will be appreciated by the person skilled in the aπ that these molecules are variants of the molecules given in a) above, provided that such variants have the activity of being immunogenic. The skilled person will appreciate that various changes can often be made to the amino acid sequence of a molecule which has a paπicular activity to produce variants (sometimes known as "muteins") having at least a propoπion of said activity, and preferably having a substantial propoπion of said activity. Such variants of the molecules described in a) above are within the scope of the present invention and are discussed in greater detail below. They include allelic and non-allelic variants.

An example of a variant of the present invention is a molecule as defined in a) above, apaπ from the substitution of one or more amino acids with one or more other amino acids. The skilled person is aware that various amino acids have similar propeπies. One or more such amino acids of a molecule can often be substituted by one or more other such amino acids without eliminating a desired activirv of that molecule.

Thus the amino acids glycine, alanine, vaiine, leucine and isoleucine can often be substituted for one another (amino acids having aliphatic side chains). Of these possible substitutions it is preferred that glycine and alanine are used to substimte for one another (since they have relatively shoπ side chains) and that vaiine, leucine and isoleucine are used to substitute for one anodier (since they have larger aliphatic side chains which are hydrophobic).

Other amino acids which can often be substituted for one another include: phenylalanine, tyrosine and tryptophan (amino acids having aromatic side chains); lysine, arginine and histidine (amino acids having basic side chains); aspartate and glutamate (amino acids having acidic side chains); asparagine and glutamine (amino acids having amide side chains); and cysteine and methionine (amino acids having sulphur containing side chains). Substitutions of this nature are often refeπed to as "conservative" or "semi-conservative" amino acid substitutions.

Amino acid deletions or inseπions may also be made relative to the amino acid sequence given in a) above. Thus, for example, amino acids which do not have a substantial effect on the activity of the polypeptide. or at least which do not eliminate such activity, may be deleted. Such deletions can be advantageous since the overall length and the molecular weight of a polypeptide can be reduced whilst still retaining activity. This can enable the amount of polypeptide required for a particular purpose to be reduced, e.g. dosage levels can be reduced.

Amino acid inseπions relative to the sequence given in a) above can also be made. This may be done to alter the propeπies of a molecule of the present invention (e.g. to assist in identification, purification or expression, as explained above in relation to fusion proteins). Amino acid changes relative to die sequence given in a) above can be made using any suitable technique e.g. by using site-directed mutagenesis.

It should be appreciated that amino acid substitutions or inseπions within die scope of the present invention can be made using naturally occurring or non- naturally occurring amino acids, bom D- and L-amino acids.

As discussed supra, it is often advantageous to reduce the length of a polypeptide, provided that the resultant reduced length polypeptide still has a desired activity. Feature c) of the present invention therefore covers fragments of polypeptides a) or b) above, provided that such fragments have immunogenic activity. It is preferable that the fragments retain the N-terminal portion of SEQ ID No . 1.

Where the molecule comprises a multiple of a sequence as defined in a), b) or c) according to feature d), the number of sequences is suitably enough to generate an immune reaction. The number of multiples may be in the range of 1 to 10, generally 1 to 5 and preferably 2 to 4.

The skilled person can determine whether or not a paπicular fragment has this activity using the techniques disclosed above. Preferred fragments are at least 10 amino acids long. They may be at least 20, at least 50 or at least 100 amino acids long. The present invention includes molecules substantially homologous to those defined above.. Whatever amino acid changes are made (whether by means of substitution, inseπion or deletion), preferred polypeptides of the present invention have at least 50% sequence identity with a polypeptide as defined in a) above more preferably the degree of sequence identity is at least 70% , 75 % or

80% . Sequence identities of at least 90% or at least 95% are most preferred.

The degree of amino acid sequence identity can be calculated using a program such as "bestfit" (Smith and Waterman, Advances in Applied Mathematics, 482- 489 (1981)) to find the best segment of similarity between any two sequences. The alignment is based on maximising the score achieved using a matrix of amino acid similarities, such as that described by Schwarz and Dayhof (1979) Atlas of Protein Sequence and Structure, Dayhof, M.O.. Ed pp 353-358.

Where high degrees of sequence identity are present there will be relatively few differences in amino acid sequence. Thus for example they may be less than 20, less than 10, or even less than 5 differences.

Molecules in accordance with the present invention are also capable of eliciting an immune response in a mammal. The immune response may be humoral (including cell-mediated) or innate, and found at bod systemic and mucosal sites (Immunology ed. Roitt et al, Gower Medical Publishing, London, Fifth edition, (1997)). The animal may be a mammal, suitably a human or a non- human mammal, including dogs, cats, cows or bulls, sheep, horses, rabbits, llamas, rats or mice. Alternatively, the animal may be a bird species, such as poultry, including chickens or turkeys.

According to a second aspect of the present invention, there is provided the use of a molecule as previously defined in the preparation of an agent for the prophylaxis or treatment of a C. difficile infection. Therapeutic molecules of the present invention may be used in the treatment of a human or non-human animal suffering from infection with C. difficile. The treatment may be prophylactic or may be in respect of an existing condition. The molecules of the present invention may also be used in the manufacture of a medicament for the treatment of C. difficile infection. Where the medicament is to be used as a prophylactic, it may be conveniently formulated as a vaccine using vaccine preparations known in the aπ.

Formulations for use as vaccines may be prepared in conjunction with several different adjuvants or live delivery vectors, e.g. attenuated live vectors, preferably rationally attenuated live vectors. Several bacterial toxins such as cholera toxin (CT) from Vibrio cholerae and heat-labile toxin from Escherichia coli have been shown to be efficient adjuvants when administered in small amounts at mucosal surfaces (Nedrud et al J. Immunol. 139 3484 (1987)),

Clements et al Vaccme 6 269 (1988)). Preferably, the adjuvant is a heat-labile E-coh toxin. Various antigens have also been delivered to die immune system encapsulated within biodegradable microspheres (Eldridge et al Curr. Top. Microbiol. Immunol. 146 59 (1989)) and phospholipid vesicles or liposomes (Ward et al Micro. Path. 21 499-512 (1996)). Rationally attenuated live vectors have been shown to be efficient at inducing protective immune responses against heterologous antigens expressed within the attenuated cell. An example of a suitable rationally attenuated live vector is Salmonella typhimurium BRD509 ( roA αroD) (Strugnell et al Infect Immun. 60 3994-4002 (1992)), BRD915 (Johnson et al Mol. Microb. 5 401-407 (1991)) or S. typhimurium BRD916 (btrA) (Tackett et al Infect Immun. 65 452- 456 ( 1 97)). It is preferable that the vector used is a Salmonella. As the attenuated vectors are often efficacious vaccines in their own right, the use of — — — such vectors also extends to the preparation of multivalent vaccines, for example, the single-dose tetanus vaccine (Chatfield et al Bio/Tech 10 888-892 (1992)). The inclusion of more than one heterologous antigen within die same strain allows this principle to be extended even further (Khan et al Proc. Natl. Acad. Sci. USA 91 11261-11265 (1994)) and the recent development of safer attenuated S. τyphi strains means that this approach can be used in human trials (Tackett et al Infect. Immun. 65 452-456 (1997)). This aspect of the present invention therefore extends to a vaccine comprising a fusion polypeptide based

on a molecule of the present invention which may also be effective against more than one cause of infection, e.g. a vaccine may be effective against one or more of C. difficile. Clostridium tetani (tetanus), Salmonella typhi (typhoid).

The medicament will usually be supplied as pan of a sterile, pharmaceutical composition which will normally include a pharmaceutically acceptable carrier. This pharmaceutical composition may be in any suitable form, (depending upon the desired method of administering it to a patient).

It may be provided in unit dosage form, will generally be provided in a sealed container and may be provided as paπ of a kit. Such a kit would normally (although not necessarily) include instructions for use. It may include a plurality of said unit dosage forms.

The pharmaceutical composition may be adapted for administration by any appropriate route, for example by the oral (including buccal or sublingual), rectal, nasal, topical (including buccal, sublingual or transdermal), intragastric, vaginal or parenteral (including subcutaneous, intramuscular, intravenous or intradermal) route. Such compositions may be prepared by any method known in die aπ of phannacy, for example by admixing the active ingredient with the caπier(s) or excipient(s) under sterile conditions. The molecules of the present invention need not be exclusively expressed by an attenuated Salmonella vector system and may be administered directly with a suitable adjuvant. The molecules may be admixed with CT, LT or detoxified derivatives of these adjuvants (Douce et al Proc. Natl. Acad. Sci. USA 92 1644-1648 (1995)) and inoculated by any desired route described above. Inoculation via intranasal, vaginal or rectal routes may be conveniently employed to promote a mucosal immune response. To raise a systemic immune response, for example to generate polyclonal antiserum, the molecules of the present invention can be admixed with Freund's adjuvant and injected subcutaneously, intraperitoneally, intramuscularly, intravenously or intradermally. Pharmaceutical compositions adapted for oral administration may be presented as discrete units such as capsules or tablets; as powders or granules: as solutions, syrups or suspensions (in aqueous or non-aqueous liquids; or as edible foams or whips; or as emulsions). Suitable excipients for tablets or hard gelatine capsules include lactose, maize starch or derivatives thereof, stearic acid or salts thereof. Suitable excipients for use with soft gelatine capsules include for example vegetable oils, waxes, fats, semi-solid, or liquid polyols etc.

For the preparation of solutions and syrups, excipients which may be used include for example water, polyols and sugars. For the preparation of suspensions oils (e.g. vegetable oils) may be used to provide oil-in-water or water in oil suspensions.

Pharmaceutical compositions adapted for transdermal administration may be presented as discrete patches intended to remain in intimate contact with the epidermis of the recipient for a prolonged period of time. For example, the active ingredient may be delivered from the patch by iontophoresis as generally described in Pharmaceutical Research, 3 (6) 318 (1986).

Pharmaceutical compositions adapted for topical administration may be formulated as ointments, creams, suspensions, lotions, powders, solutions, pastes, gels, sprays, aerosols or oils. For infections of the eye or other external tissues, for example mouth and skin, die compositions are preferably applied as a topical ointment or cream. When formulated in an ointment, the active ingredient may be employed with either a paraffinic or a water-miscible ointment base. Alternatively, the active ingredient may be formulated in a cream with an oil-in-water cream base or a water-in-oil base. Pharmaceutical compositions adapted for topical administration to the eye include eye drops wherein the active ingredient is dissolved or suspended in a suitable carrier, especially an aqueous solvent. Pharmaceutical compositions adapted for topical administration in the mouth include lozenges, pastilles and mouth washes.

Pharmaceutical compositions adapted for rectal administration may be presented as suppositories or enemas.

Pharmaceutical compositions adapted for nasal administration wherein the carrier is a solid include a coarse powder having a paπicle size for example in the range 20 to 500 microns which is administered in the manner in which snuff is taken, i.e. by rapid inhalation dirough the nasal passage from a container of the powder held close up to the nose. Suitable compositions wherein die carrier is a liquid, for administration as a nasal spray or as nasal drops, include aqueous or oil solutions of the active ingredient.

Pharmaceutical compositions adapted for administration by inhalation include fine paπicle dusts or mists which may be generated by means of various types of metered dose pressurised aerosols, nebulizers or insufflators.

Pharmaceutical compositions adapted for vaginal administration may be presented as pessaries, tampons, creams, gels, pastes, foams or spray formulations.

Pharmaceutical compositions adapted for parenteral administration include aqueous and non-aqueous sterile injection solution which may contain anti- oxidants, buffers, bacteriostats and solutes which render the formulation substantially isotonic with the blood of the intended recipient; and aqueous and non-aqueous sterile suspensions which may include suspending agents and thickening agents. Excipients which may be used for injectable solutions include water, alcohols, polyols, glycerine and vegetable oils, for example. The compositions may be presented in unit-dose or multi-dose containers, for example sealed ampoules and vials, and may be stored in a freeze-dried (lyophilized) condition requiring only the addition of the sterile liquid carried, for example water for injections, immediately prior to use. Extemporaneous injection solutions and suspensions may be prepared from sterile powders, granules and tablets.

The pharmaceutical compositions may contain preserving agents, solubilising agents, stabilising agents, wetting agents, emulsifiers. sweeteners, colourants, odourants, salts (molecules of the present invention may themselves be provided in the form of a pharmaceutically acceptable salt), buffers, coating agents or antioxidants. They may also contain therapeutically active agents in addition to d e molecule of die present invention.

In addition to die uses discussed above in relation to treatments, molecules of the present invention can be used in diagnosis. For example, in the diagnosis of C. difficile infection in a biological sample from an affected subject. This aspect of the invention therefore also extends to a kit for the diagnosis of a C. difficile infection, the kit comprising a molecule as defined above. The method of diagnosis may comprise the use of a molecule of the present invention in an assay for the detection of circulating antibodies in an infected subject. The means for detection may comprise the use of ELISA, fluorescence based or radioimmuno assay (RIA) techniques and die assay can be performed in a suitable biological sample, including blood, saliva, tears, urine, faeces, sweat, semen or milk.

One further use of ie molecules of me present invention is in raising or selecting antibodies. According to a third aspect of the present invention there is provided an antibody to a molecule as previously defined. The present invention therefore includes antibodies which bind to a molecule of die present invention. Preferred antibodies bind specifically to molecules of the present invention so iat they can be used to purify such molecules. The antibodies may be monoclonal or polyclonal. Polyclonal antibodies can be raised by stimulating their production in a suitable animal host (e.g. a mouse, rat, guinea pig. rabbit, sheep, goat or monkey) when a molecule of the present invention is injected into the animal. If necessary an adjuvant may be administered together with a molecule of the present invention. The antibodies can then be purified by virtue of their binding to a molecule of the present invention.

Monoclonal antibodies can be produced from hybridomas. These can be formed by fusing myeloma cells and spleen cells which produce die desired antibody in order to form an immortal cell line. This is the well known Kohler & Milstein technique (Nature 256 52-55 (1975)).

Techniques for producing monoclonal and polyclonal antibodies which bind to a paπicular protein are now well developed in die aπ. They are discussed in standard immunology textbooks, for example in Roitt et al Immunology fifth edition (1997), Gower Medical Publishing, London.

In addition to whole antibodies, the present invention includes derivatives thereof which are capable of binding to molecules of me present invention.

Thus d e present invention includes antibody fragments and synthetic constructs.

Examples of antibody fragments and syndietic constructs are given by Dougall et al in Tibtech 12 372-379 (September 1994). Antibody fragments include, for example, Fab, F(ab')₂ and Fv fragments (see Roitt et al [supra]). Fv fragments can be modified to produce a syndietic construct known as a single chain Fv

(scFv) molecule. This includes a peptide linker covalently joining V_h and V, regions which contribute to the stability of the molecule.

Other syndietic constructs include CDR peptides. These are synthetic peptides comprising antigen binding determinants. Peptide mimetics may also be used. These molecules are usually conformationally restricted organic rings which mimic the structure of a CDR loop and which include antigen-interactive side chains.

Syndietic constructs include chimaeric molecules. Thus, for example, humanised (or primatised) antibodies or derivatives thereof are within the scope of the present invention. An example of a humanised antibody is an antibody having human framework regions, but rodent hypervariable regions.

Synthetic constructs also include molecules comprising a covalently linked moiety which provides the molecule widi some desirable property in addition to antigen binding. For example the moiety may be a label (e.g. a fluorescent or radioactive label) or a pharmaceutically active agent.

The antibodies or derivatives thereof of the present invention have a wide variety of uses. They can be used in purification and/or identification of the molecules of the present invention. Thus diey may be used in diagnosis of a C. difficile infection in a biological sample from an affected subject. They can be provided in d e form of a kit for screening for the molecules of ie present invention.

According to a fourth aspect of the present invention, iere is provided the use of an antibody as previously defined in the preparation of an agent for the prophylaxis or treatment of a C. difficile infection.

According to a fifth aspect of the present invention diere is provided a mediod of diagnosing a Clostridium difficile infection, the mediod comprising me step of contacting an optionally labelled antibody as previously defined with a biological sample. The biological sample may be blood, saliva, tears, urine, faeces, sweat, semen or milk. The method can also be used on samples of foodsmffs or other samples taken from the environment such as water to determine d e presence or otherwise of C. difficile. According to a sixth aspect of the present invention there is provided a recombinant DNA construct comprising nucleotides

7159-8118 of C. difficile as shown in Figure 1. Suitable recombinant DNA constructs comprising such a sequence may be introduced into host cells to enable the expression of molecules of the present inventions using techniques known to the person skilled in die aπ. Examples of suitable vectors are pRSET-A or pTECH-1. Recombinant DNA constructs prepared in accordance with the present invention include p5/6. The fragment expressed by p5/6 comprises 14 C-terminal toxin A repeats.

Molecules comprising amino acid sequences in accordance with the present application can be syndiesised by any convenient technique. Peptides and polypeptides may be syndiesised by chemical routes using routine procedures. Alternatively, the molecules may be synthesised by expressing a nucleotide sequence encoding d e peptides in a suitable host cell. According to a sixth aspect of e present invention diere is provided a method for me preparation of a fragment of C. difficile Toxin A, comprising the step of expressing a nucleotide sequence encoding a molecule as previously defined.

Typically d e nucleotide sequence will be incorporated into a vector appropriate for expression in the desired host cell. Suitable expression systems include but are not limited to prokaryotes such as E. coli, S. typhi, or S. typhimurium, preferably an attenuated host cell. Alternatively, expression may be in suitable eukaryotic systems including yeast such as S. cerevisiae, S. pombe, insect cells, plant cells such as Arabidopsis thaliana or tobacco, or mammalian cells such as COS, CHO, Vero or HeLa. The choice of vector and appropriate promoter and regulatory sequences will generally depend on the expression system being used at d e time. Techniques for cloning, expressing and purifying proteins and polypeptides are well known to die skilled person. Various such techniques are disclosed, for example, in Sambrook et al [Molecular Cloning 2nd Edition, Cold Spring Harbor Laboratory Press (1989)]; in Old & Primrose [Principles of Gene Manipulation 5th Edition, Blackwell Scientific Publications (1994); and in Stryer [Biochemistry 4th Edition, W H Freeman and Company (1995)].

By using appropriate expression systems molecules of the present invention may be expressed in glycosylated or non-glycosvlated form. Non-glycosvlated forms can be produced by expression in pro kary otic hosts, such as E. coli. Polypeptides comprising N-terminal methionine may be produced using certain expression systems, whilst in odiers the mature polypeptide will lack this residue. Preferred techniques for cloning, expressing and purifying a molecule of the present invention are summarised below.

In addition to nucleic acid molecules coding for molecules according to the present invention, refeπed to herein as "coding" nucleic acid molecules die present invention also includes nucleic acid molecules complementary thereto. Thus, for example, both strands of a double stranded nucleic acid molecule are included widiin the scope of ie present invention (whedier or not they are associated with one anodier). Also included are mRNA molecules and complementary DNA molecules (e.g. cDNA molecules).

Nucleic acid molecules which can hybridise to any of the nucleic acid molecules discussed above are also covered by d e present invention. Such nucleic acid molecules are refeπed to herein as "hybridising" nucleic acid molecules. Hybridising nucleic acid molecules can be useful as probes or primers, for example. Desirably such hybridising molecules are at least 10 nucleotides in length and preferably are at least 25 or at least 50 nucleotides in length. The hybridising nucleic acid molecules preferably hybridise to nucleic acids within the scope of a) or b) above specifically.

Desirably the hybridising molecules will hybridise to such molecules under stringent hybridisation conditions. One example of stringent hybridisation conditions is where attempted hybridisation is carried out at a temperature of from about 35°C to about 65°C using a salt solution which is about 0.9 molar. However, the skilled person will be able to vary such conditions as appropriate in order to take into account variables such as probe length, base composition, type of ions present, etc.

A hybridising nucleic acid molecule of die present invention may have a high degree of sequence identity along its lengdi with a nucleic acid molecule widiin the scope of a) or b) above (e.g. at least 50%, at least 75% or at least 90% sequence identity). As will be appreciated by the skilled person, the higher the sequence identity a given single stranded nucleic acid molecule has widi anodier nucleic acid molecule, die greater the likelihood diat it wdl hybridise to a nucleic acid molecule which is complementary to that other nucleic acid molecule under appropriate conditions.

In view of die foregoing description the skilled person will appreciate that a large number of nucleic acids are within die scope of ie present invention. Unless the context indicates odierwise, nucleic acid molecules of the present invention may have one or more of the following characteristics:

1) they may be DNA or RNA;

2) they may be single or double stranded;

3) they may be provided in recombinant form i.e. covalently linked to a 5' and/or a 3 ' flanking sequence to provide a molecule which does not occur in nature; 4) they may be provided without 5' and/or 3' flanking sequences which normally occur in nature; 5) iey may be provided in substantially pure form. Thus they may be provided in a form which is substantial!) free from contaminating proteins and/or from other nucleic acids; 6) diey may be provided with introns or without mtrons (e.g. as cDNA).

According to a sevendi aspect of the invention there is provided a vaccine formulation comprising a recombinant DNA construct as previously defined, optionally togedier with one or more carriers or adjuvants. Such vaccine formulations may also be employed in accordance with d e uses or methods described above. The recombinant DNA construct may be suitably prepared using pRc/CMV (Invitrogen) for vaccines in accordance with this aspect of the invention. Such DNA vaccines may be administered by any convenient route as described above but administration by subcutaneous or intramuscular injection may be prefeπed (Wolff et al Science 247 1465 (1990)).

It has also been observed that the molecules described above, are capable, when co-expressed as a fusion protein or administered in a mixture together with a second peptide or protein, of increasing the antibody response to the co- administered peptide or protein. Thus, a further embodiment of the preferred invention relates to the use of the molecules described above as an adjuvant in the administrations of peptides or proteins. Prefeπed aspects of the second and subsequent aspects of die present invention are as for the first aspect mutatis mutandis.

The present invention will now be described, by way of example only, with reference to the accompanying drawings and Examples, wherein:

FIGURE 1 shows a map of Toxin A specific fragments amplified using PCR.

FIGURE 2 shows an SDS-PAGE analysis of attenuated S. typhimurium expressing Toxin A fragments from whole cell lysates expressing toxin A constructs under aerobic and anaerobic conditions. Lane (1) BRD915 only; lane (2) BRD915 + pTECH-1 ; lane (3) BRD915 ÷ pTA2: lane (4) BRD915 + p5/6; lane (5) BRD915 + p5/7; lane (6) BRD915 + p9/10

(comparative control). Apparent molecular weights of toxin A fusion proteins are shown in kDa.

FIGURE 3 shows ELISA titres of toxin specific antibody in the serum of BALB/c mice immunised intragastricallv with 2 doses of S. typhimurium

BRD915 expressing C. difficile toxin A fragments. Both anti-toxin A

(filled bar) and anti-tetanus toxin (clear bar) are shown. Error bars represent the standard deviation of the mean titre from five mice. Detection limit 1:50.

FIGURE 4 shows binding activity of Toxin fragments as measured by agglutination of rabbit erythrocytes.

FIGURE 5 shows the predicted amino acid sequence of PCR product fragment pTA2 resulting from expression of nucleotides 5983-6594 of the sequence of the Toxin A gene from C. difficile strain VPI 10463 (Dove et al Infection and Immunity 58 (2) 480-488 (1990)).

FIGURE 6 shows die predicted amino acid sequence of PCR product fragment p5/6 resulting from expression of nucleotides 7159-8118 of the sequence of the Toxin A gene from C. difficile strain VPI 10463 (Dove et al (1990)).

FIGURE 7 shows die predicted amino acid sequence of PCR product fragment p5/7 resulting from expression of nucleotides 6748-8118 of the sequence of the Toxin A gene from C. difficile strain VPI 10463 (Dove et al (1990)).

FIGURE 8 shows the predicted amino acid sequence of PCR product fragment p9/10 resulting from expression of nucleotides 5530-8115 of the sequence of the Toxin A gene from C. difficile strain VPI 10463 (Dove et α/ (1990)).

FIGURE 9 shows the predicted N-terminal sequence of PCR product fragment p9/10 which was not encoded widiin vector pTECH-1.

FIGURE 10 shows the mean ELISA titres of toxin A specific total antibody found in serum of BALB/c mice immunised intragastrically with S. typhimurium BRD509 expressing 14 C. difficile toxin A repeats. Samples were harvested before immunisation (day 0), 28 days after first dose (day 28), 35 days after the second dose (day 63), and 22 days after a subcutaneous boost witii 0.5μg of purified p5/6 fragment C fusion protein (day 85). Eπor bars represent the standard eπor of the mean. Detection limit 1:50.

FIGURE 11 shows the mean ELISA titres of tetanus toxin-specific total antibody found in serum of BALB/c mice immunised intragastrically with S. typhimurium BRD509 expressing 14 C. difficile toxin A repeats. Samples were harvested before immunisation (day 0), 28 days after first dose (day 28), 35 days after the second dose (day 63), and 22 days after a subcutaneous boost with 0.5μg of purified p5/6 fragment C fusion protein (day 85). Eπor bars represent the standard eπor of the mean. Detection limit 1:50. FIGURE 12 shows the mean ELISA titres of toxin A-specific IgA antibody found in gastric lavage samples of BALB/c mice immunised intragastrically with S. typhimurium BRD509 expressing 14 C. difficile toxin A repeats. Samples were harvested before immunisation (day 0), 28 days after first dose (day 28), and 22 days after a subcutaneous boost with 0.5 μg of purified p5/6-fragment C fusion protein (day 85). Eπor bars represent the standard error of the mean. Detection limit 1:2.

FIGURE 13 shows die mean ELISA titres of tetanus toxin-specific IgA antibody found in gastric lavage samples of BALB/c mice immunised intragastrically with S. typhimurium BRD509 expressing 14 C. difficile toxin A repeats. Samples were harvested before immunisation (day 0), 28 days after first dose (day 28), and 22 days after a subcutaneous boost witii 0.5μg of purified p5/6-fragment C fusion protein (day 85). Eπor bars represent the standard error of the mean. Detection limit 1:2.

FIGURE 14 shows a plasmid map of vector pTECH-1.

FIGURE 15 shows a plasmid map of vector pRSET-A (Invitrogen) .

FIG. 16. SDS-PAGE analysis (10% polyacrylamide) of thyroglobulin affinity purified p56HIS (A) and p56TETC (B). Lane 1, cell lysates before affinity column; Lane 2, cell lysates after affinity column; Lane 3, purified protein collected from column (approximately 3 μg). Apparent molecular masses are shown in kDa. Pre-stained molecular weight markers are also shown (M). FIG. 17. Immunoblot of affinity purified p56HIS (Lane 1) and p56TETC (Lane 2) against the toxin A specific monoclonal antibody PCG-4 (A), or anti-TT polyclonal antiserum (B). Apparent molecular weights are shown in kDa. Molecular weight markers are shown (M).

FIG. 18. Mean anti-toxin A total immunoglobulin responses in the serum of intransal immunised BALB/c mice. Antibody titres were measured by ELISA in serum taken after 1 dose (day 19), 2 doses (day 34) and 3 doses (day 47) of antigen. Mean titres are shown ± SD from five mice. Individual pre-immune titres have been subtracted from each coπesponding mouse.

FIG. 19. Mucosal IgA responses against toxin A after 3 intranasal doses of antigen in nasal and pulmonary lavage. Responses show the variation in die immune responses between individual mice in each group. Bars represent mean antibody titres.

FIG. 20. Mean anti-TT total immunoglobulin responses in the serum of intransal immunised BALB/c mice. Antibody titres were measured by ELISA in serum taken after 1 dose (day 19), 2 doses (day 34) and 3 doses (day 47) of antigen. Mean titres are shown ± SD from five mice. Individual pre-immune titres have been subtracted from each coπesponding mouse.

FIG. 21. Mucosal IgA responses against tetanus toxin after 3 intranasal doses of antigen in nasal and pulmonary lavage. Responses show the variation in the immune responses between individual mice in each group. Bars represent mean antibody titres. Example 1 : Amino acid composition of toxin A fragments The primer sequences used to generate die toxin A fragments were based on d e entire sequence of toxin A from C. difficile strain VPI 10463 published by Dove et al, 1990 Infection and Immunity, vol. 58(2), page 480-488. The predicted amino acid sequences shown in Figures 5, 6, 7, 8 and 9 are also from Dove et al, 1990. The vector pTECH-1 is described in Khan et al Proc. Natl. Acad. Sci. USA 91 11261-11265 (1994), Chatfield et al Bio/technology 10 888-892 (1992) and Oxer et al Nucleic Acids Research 19 (11) 2889-2892 (1991). (In

tiiese studies, die vector pTECH-1 was obtained from Medeva, Leatherhead.

Surrey, GB)

The sequences shown for fragments pTA2 (8 repeats), p5/6 (14 repeats) and p5/7 (20 repeats) are the predicted amino acid sequences contained on die PCR- generated fragments mat have been cloned into botii pTECH-1 and pRSET-A vectors (Invitrogen). In die case of ie comparative control fragment p9/10 (36 repeats), the fragment cloned into die pTECH-1 vector had 29 of die amino- terminal amino acids missing as a result of restriction enzyme treatment. This is in contrast to the fragment cloned into pRSET-A which contained the full complement as shown.

Fragment pTA2

This PCR product contained the nucleotides 5983-6594 inclusively. The predicted amino acid sequence encoded by this fragment is as shown in Figure 5.

Fragment p5/6

This PCR product contained the nucleotides 7159-8118 inclusively. The predicted amino acid sequence encoded by this fragment is as shown in Figure 6.

Fragment p5/7

This PCR product contained die nucleotides 6748-8118 inclusive. The predicted amino acid sequence encoded by tiiis fragment is as shown in Figure 7.

Comparative control - Fragment p9/10

This PCR product contained me nucleotides 5530-8115 inclusive. The predicted amino acid sequence encoded by tiiis fragment is as shown in Figure 8. The fragment encoded widiin the pTECH-1 vector lacked the following sequence from the N-terminus as shown in Figure 9.

Example 2: Expression of toxin A fragments in E. coli Utilising the polymerase chain reaction (PCR), four overlapping DNA fragments which spanned the entire C-terminal repeat region of toxin A were amplified from C. difficile strain VPI 10463 (Figure 1). The toxin fragments encoded for 8, 14, 20 and 36 whole toxin A repeats and were labelled pTA2, p5/6, p5/7, and p9/10 respectively. These fragments were subsequently cloned into two expression vectors, pTECH-1 and pRSET-A and the integrity of the constructs was verified by di-deoxy terminal sequencing (ABI Prism). When expressed within E. coli (XL2 Ultra competent - Stratagene), the pTECH-1 expression systems generated toxin A fragments with die non-toxic fragment C of Tetanus toxin attached to their N-terminus. When me pRSET-A based constructs were expressed in E. coli (BL21(DE3) - Novagen), the toxin A fragments had six Histidine residues attached to their N-terminus. SDS-PAGE analysis showed tiiat all four toxin A fragments were successfully expressed in botii vectors, generating recombinant protein with the expected molecular weights. All of the fragments reacted witii a monoclonal antibody specific for toxin A (PCG-4) in a Western blot showing die repeats to be expressed in an immunologically reactive form. Cellular fractionation studies also showed all constructs to expressed widiin die cellular cytoplasm in a soluble form. This solubility is important as it allows the antigen to be in a desirable configuration for immunisation.

Example 3 : Expression of toxin A fragments in S. typhimurium The generation of a toxin-neutralising response at die gastrointestinal mucosa appears to be important in protecting against C. difficile disease. In order to facilitate the induction of an anti-mucosal response, the various pTECH-1 constructs containing the toxin fragments were introduced into an attenuated strain of Salmonella typhimurium The strain chosen contained an attenuating lesion within the htrA stress protem gene. Attenuated Salmonella were chosen as they have been shown to be an efficient mucosal delivery vehicle for other heterologous antigens. Expression of the toxm A fusion protems from the pTECH-1 constructs is driven by the nirB promoter which is only switched on under anaerobic conditions. (In diese studies Salmonella typhimurium (htrA) strain BRD915 was obtained from Medeva, Leadierhead, Suπev , GB)

The plasmid constructs pTA2, p5/6, p5/7 and p9/10 were introduced into the S. typhimurium htrA mutant stram (BRD915), a vaccine strain known to be efficacious against murme typhoid. SDS-PAGE analysis showed that all 4 toxm constructs were only expressed when grown under these conditions in vitro (Figure 2), and tiiat they reacted with monoclonal antibody PCG-4 in Western blots.

Example 4 Characterisation of immunogenicity of toxin A fragments To evaluate the immunogenic potential of the toxm fragments, six groups of female BALB/c mice (5 per group) were immunised intragastrically with the following inoculums:

Group 1 - S typhimurium only

Group 2 - S. typhimurium expressmg pTECH-1

Group 3 - S. typhimurium expressmg pTA2 (8 repeats) Group 4 - S. typhimurium expressmg p5/6 (14 repeats)

Group 5 - S typhimurium expressing p5/7 (20 repeats)

Group 6 - S. typhimurium expressmg p9/10 (36 repeats)

Each mouse received 3-4 x 10¹⁰ cfu per moculum. Two doses were given, the second after 28 days. Animals were terminated 56 days after the initial dose, and serum plus intestinal washes were collected from each animal. These samples were prepared to investigate the efficiency of the constructs in stimulating die immune system.

The anti-toxin A response elicited by die various fragments were evaluated by ELISA. Wells of an EIA/RIA 96 well plate (Costar) were coated overnight with purified whole toxin A (0.05μg/well) and blocked for 1 hour witii 1 % (w/v) BSA in PBS (Douce et al Proc. Natl. Acad. Sci. USA 92 1644-1648 (1995)). After washing 3 times with 0.1 % (v/v) Tween-20 in PBS, the wells were incubated for 2 hours with serum taken from each animal serially diluted five-fold in PBS. The wells were washed as before, and an anti-mouse horseradish peroxidase conjugate was added and allowed to react for 2 hours. Bound antibody was visualised using o-phenylenediamine substrate and quantified by reading the absorbance at 490 nm. ELISA titres were determined as die reciprocal of the serum dilution which gave an absorbance of 0.3 units above die background.

Figure 3 depicts the mean anti-toxin A antibody titres elicited by all six groups of animals. It is clear mat die antibody levels induced by die toxin A constructs containing 8, 20 and 36 repeats were relatively low, and were not significantly higher than the background level obtained widi the control groups. However, the construct expressing the 14 carboxy-terminal repeats did give a positive antitoxin A response and elicited a mean anti-toxin A titre which was over 4-fold higher than that of background. Thus, it appears that a fragment expressing 14 toxin A repeats of which only 4 are of the IIB class may be optimum within this system with regard to generating an anti-toxin A response within serum. Studies are cuπently in progress to evaluate the potency of this construct is stimulating die mucosal immune system. Example 5: In vitro characterisation of Salmonella-expressed toxin A-fragment

C fusion proteins

Immunogenicity studies have identified a shoπ protein encoding 4 class IIB repeats as being immunogenic when expressed widiin rationally attenuated S. typhimurium and given orally to mice. This is in contrast to fragments expressing 3, 7 and 12 IIB repeats which were not overtly immunogenic. Since the IIB repeats are known to bind to a characterised toxin A receptor, the toxin fragments were analysed for receptor-binding function in an established assay of toxin A-mediated binding. The toxin fragments were expressed from the pTECH-1 plasmid within S. typhimurium in vitro and die cells harvested. Sphaeroplasts were made from these cells and die soluble material released from the cell with 4 pulses of ultrasonic power. Fractions containing 50 micrograms of soluble material were serially diluted 2-fold in PBS widiin a 96 well plate, and reacted with 50 microlitres of a 2% (w/v) suspension of rabbit erythrocytes at 4°C. The level of binding was quantified as d e reciprocal of the highest dilution of sample which gave 100% agglutination of erythrocytes/50 micrograms of total soluble material. Although not strictly quantitative due to the different levels of expression of the toxin fragments, me assay did show quite clearly that die construct containing 14 toxin A repeats gave a significantly higher level of binding, approximately 14-fold greater than seen with the other constructs (Figure 4). A particular point of interest is mat the fragment containing 8 repeats gave no binding activity. Since the fragments comprising of 8 and 14 toxin repeats contain 3 and 4 class IIB repeats respectively, it appears that of die fragments expressed from the pTECH-I system, mat 4 class IIB repeats is the minimum requirement to promote binding to rabbit erythrocytes. Alternatively, it may be that die 14 toxin A repeats found in mis fragment generate a protein which is in an optimum configuration to promote receptor binding. In order to be an effective delivery system in vivo, the toxin A-expressing plasmids must be stable within the Salmonella in die absence of any antibiotic pressure. Continuous passage of the recombinant Salmonella in vitro within antibiotic-free medium showed all 4 constructs to be efficiently retained by strain BRD915 even after 7 days. The plasmids containing 14 and 20 repeats were present in 87 % and 82 % of cells respectively which was comparable to the 88% observed with pTECH-1 control, while the constructs containing 8 and 36 repeats were found in 30% and 46% of cells respectively.

These studies have successfully isolated 4 overlapping fragments from the C- terminus of toxin A of C. difficile and expressed them as fusions to the immunogenic, non-toxic component of tetanus toxin widiin a htrA mutant strain of S. typhimurium. Expression of the fusion proteins was successfully controlled by die anaerobically-inducible nirB promoter resulting in high level retention of the constructs by Salmonella in die absence of an antibiotic pressure which is essential for efficient antigen delivery in vivo. The toxin fragments were analysed for retention or receptor-binding function in an haemagglutination assay, which is an established assay of toxin A-mediated binding. Aldiough not strictly quantitative due to die different levels of fusion protein expressed within die Salmonella, the p5/6 construct containing 14 toxin A repeats promoted a significantly higher level of haemagglutination than the otiiers. This suggests that the conformation of die C-terminal repeats may be more important than the number of repetitive sequences when expressed as recombinant proteins. Intragastric vaccination of mice with an S. typhimurium htrA mutant expressing diese fusion proteins resulted in poor seroconversion with regard to tetanus toxin. This was unexpected as previous studies using die pTECH-1 vector within a different attenuated strain (S. typhimurium aroA) generated anti-tetanus toxin titres significantly higher than the titres seen in die present study (Khan et al PNAS USA 91 11261-11265 (1994)). However, using the btrA-mutant

Salmonella strain in the present study for delivering toxin A-fragment C fusions to the immune system, a positive anti-toxin A serum response was seen with the fusion protein containing 14 toxin A repeats. This fusion also gave elevated levels of haemagglutination, implying mat immunogenicity may be correlated to efficient receptor binding. In an attempt to further increase die anti-toxin antibody response, these fusion proteins have also been delivered to the gastrointestinal mucosa within die αro-mutated S. typhimurium strains which appear to be more proficient at mucosal delivery of antigen.

Example 6: Intragastric immunisation with S. typhimurium BRD509 (aroA aroD) expressing the p5/6 construct

The immunisation experiments mentioned above identified the p5/6 construct encoding me 14 toxin A repeats fused to fragment C of tetanus toxin as being immunogenic when delivered to die gastrointestinal mucosa within S. typhimurium BRD915 (htrA). In an attempt to increase further the immune- stimulating property of this protein, die p5/6 plasmid was introduced into the vaccine strain S. typhimurium BRD509 (aroA aroD) (In these studies, Salmonella typhimurium (aroA aroD) strain BRD509 was obtained from Medeva, Leadierhead, Suπey, GB). SDS-PAGE analysis showed expression of die toxin A-fragment C fusion protein only when cells were grown under anaerobic conditions in vitro, with d e levels of expressed protein being similar to that seen witii S. typhimurium BRD915. The fusion protein also reacted witii monoclonal antibody PCG-4 in Western blot, and promoted agglutination of rabbit erythrocytes (2% v/v) at a similar level to the 14 repeat protein expressed within BRD915.

To evaluate die immunogenic potential of die 14 repeat fragment expressed widiin S. typhimurium BRD509, tiiree groups of female BALB/c mice were immunised intragastrically with the following inoculums:

Group 1 - S. typhimurium only (n=5) Group 2 - S. typhimurium expressing pTECH-1 (n= 10) Group 3 - S. typhimurium expressing p5/6 (14 repeats) (n= 10)

On day 0, each mouse received 1-2 x 10¹⁰ cfu per inoculum. After 28 days, five animals from groups 2 and 3 were terminated, while the remaining animals in all 3 groups received a second dose. After 63 days, all animals were immunised subcutaneously witii 0.5μg of the 14 repeat fragment which had been expressed from die pRSET-A vector in E. coli and affinity' purified. All animals were terminated 85 days after the initial dose. Serum plus intestinal washes were collected from each terminated animal.

Botii the anti-toxin A and die anti-tetanus toxin response elicited by die 14 repeats were evaluated by ELISA. Wells of an EIA/RIA 96 well plate (Costar) were coated overnight with eitiier purified whole toxin A (0.15μg/well) or tetanus toxoid (0.5μg/well) and d e procedure performed as previously using serum taken from all animals at days 1, 28, 63 and 85.

Figure 10 depicts die mean anti-toxin A titres found in die serum taken from all tiiree groups of animals. It is clear that even after one dose (day 28), animals which were immunised with S. typhimurium BRD509 expressing the 14 repeats elicited an anti-toxin A response, generating a mean serum titre approximately 8-fold higher than die control groups. This antibody response was boosted witii a second dose (day 63), increasing the mean fold-difference to over 200. A low dose of the 14 toxin A repeat protein was also administered witiiout the S. typhmurium BRD509 delivery system. The amount of purified material administered (0.5μg) appeared to be sub-immunogenic, as the level of anti-toxin A antibody produced by both control groups after immunisation did not increase significantly (1- and 4-fold increase for Group 1 and Group 2. respectively). In addition, the antibodies that were produced were present at a lower level than diat seen after one oral dose of die p5/6 construct. In contrast, mice which had received two doses of the S. ryphimurium j)5/6 vaccine augmented tiieir antitoxin A response, producing levels of anti-toxin A antibody 7-fold higher than prior to the subcutaneous boost.

The levels of serum anti-tetanus toxin antibody produced by mice immunised with botii die pTECH-1 and p5/6 constructs were similar, and did not increase after die second oral dose (Figure 11). In contrast to the anti-toxin A response, the levels of anti-tetanus toxin antibodies appeared to be unaffected by the subcutaneous injection of purified protein.

To determine how efficient die immunisation schedule had been in generating a local antibody response at die gastrointestinal mucosa, gastric lavage samples were taken from all terminated animals and die levels of toxin-specific IgA determined by ELISA. In order to allow for sampling eπor, antibody levels were expressed as ELISA titre/μg IgA using the method of Douce et al (Proc. Natl. Acad. Sci. USA 92 1644-1648 (1995)). Figure 12 shows that mice immunised witii one dose of the p5/6 construct generated a mean titre of antitoxin A IgA approximately 8-fold higher than the control groups. This difference increased even further to 60-fold after 2 oral doses and a boost of purified 14 repeat protein. Similarly, levels of anti-tetanus toxin IgA were over 40-fold higher than the control group after one dose of both pTECH-1 and p5/6 (Figure 13), and were further boosted as a result of the second oral dose and subcutaneous boost.

In conclusion, diese studies show diat die p5/6 construct expressing 14 toxin A repeats is immunogenic when administered orally to mice within the BRD509 strain of S. typhimurium. The immunisation resulted in positive seroconversion for both anti-toxin A and anti-tetanus toxin antibodies after one dose. The levels of anti-toxin A antibodies were boosted after a second oral dose, and increased further after injection with a sub-immunogenic dose of purified 14 repeat protein. Importantly, a local IgA immune response was also elicited, generating botii anti-toxin A and tetanus toxin specific IgA at the gastrointestinal mucosa.

Example 7: Intra-naasal immunisation with S. typhimurium BRD509 expressing the p5/6 construct.

Bacterial strains and plasmids. Salmonella typhimurium LB5010 (galE) and BRD915 (htrA) and plasmid pTECH-1 were kind gifts from Steve Chatfield. Medeva Vaccine Research Unit, Imperial College. London, U.K. E.coli BL21 (DE3) was obtained from Novagen. and plasmid pRSET-A was supplied by Invitrogen (De Schelp, The Netherlands). Bacteria were routinely cultivated in either Luria Broth (LB) or on LB agar with or without ampicillin (1 OOμg/ml).

DNA manipulation.

Restriction enzymes and DNA ligase were purchased from Promega (Southampton, U.K.) and used according to the manufacturers instructions. DNA which had been subjected to restriction enzyme treatment was purified using either S-300 HR Microspin™ columns (Pharmacia), or Prep-A-Gene™ purification resin (Bio Rad. Hemel Hempstead. U.K.). PCR was carried out using a Perkin Elmer 9600 cycle sequencer and Taq DNA Polymerase as described by the manufacturer (Appligene Oncor. U.K.). DNA cycle- sequencing was performed with an ABI PRISM® reaction kit (Perkin Elmer Applied Biosystems, Warrington, U.K.), and analysed with an Applied Biosystems 373 A DNA sequencer.

Amplification and cloning of toxin A repeat sequence into pTECH-1 and pRSET-A expression vectors. Chromosomal DNA was isolated from C. difficile strain VPI 10463 using the method of Wren et al. (1987.J. Clin. Microbiol. 25:2402-2404) . A 960 bp DNA fragment encoding for the 14 carboxyl terminus repeats of toxin A (base pairs 7159-81 18 inclusive) was amplified using PCR and the following primers which were based on the published toxin A sequence from strain VPI

10463 ( Infect. Immun. 58:480-488.) : sense strand 5'- ACTlCTAG_AGCCTCAACTGGTTATACAAGT-3\ anti-sense strand 5'- ATAACTAGIAGGGGCTTTTACTCCATCAAC-3'. Underlining denotes Xbal restriction site within sense strand primer and Spel restriction site within anti-sense strand primer.

The amplified toxin A sequence was subcloned into the pTAg cloning vector (R&D Systems Europe Ltd, Abingdon, U.K.), excised with Xbal-Spel and inserted into Spel digested pTECH-1 downstream of the TETC encoding sequence (1994 Proc. Natl. Acad. Sci. USA. 91: 1 1261-11265) . This construct was named p56TETC. In order to redone the toxin A fragment into the pRSET-A vector, the entire toxin A sequence was first excised from the pTECH-1 vector with S el and Xbal and subcloned into ^bαl-digested pUC18 (Promega). Clones containing the toxin A fragment in the coπect orientation were isolated, the toxin A sequence excised with BamUl and Hindlll and then inserted into similarly digested pRSET-A to create plasmid p56HIS. All recombinant plasmids were initially transformed into Epicurian coli® XL2- blue MRF^" ultracompetent cells (Stratagene, Cambridge, U.K.), and e integrity of the cloning junctions confirmed by Dye Terminator cycle sequencing. Expression and affinity purification of recombinant toxin A C-terminal repeat proteins.

Plasmid p56TETC was first electroporated into S. typhimurium LB5010 using a Gene Pulser apparatus (Bio Rad) (1992 Vaccine. 10:53-60) and then introduced into S. typhimurium BRD915 (htrA) via P22 bacteriophage transduction (1993 Infect. Immun. 61 :5374-5380) . To induce protein expression, recombinant S. typhimurium BRD915 were incubated at 37°C witii aeration in LB-ampicillin until early/mid log phase (OD600 0.3-0.5). The anaerobically-inducible nirB promoter located on the pTECH-1 plasmid was then activated by incubating the culture for a further 24 h without shaking in an anaerobic atmosphere (Mark III work station, Don Whitley Scientific Limited, U.K.). E.coli BL21(DE3) transformed witii p56HIS were grown in LB-ampicillin until mid log phase and expression of the plasmid encoded protein induced by the addition of isopropyl-b-D-thiogalactopyranoside (IPTG) to a final concentration of 1 mM and incubating for a further 5 h. After recombinant protein induction, both the S. typhimurium and E.coli cells were harvested and sphaeroplasts generated as described by Ward et al. (1996 Microb. Pathog. 21 :499-512) . Fractions (10 ml) of the sphaeroplasts were disrupted with three 50 second bursts of ultrasonic power (Ultrasonic Processor, Jencons Scientific Ltd, U.K.) and the cellular debris removed at 3,000 x g for 15 min. The soluble fraction was recovered after centrifugation for 30 min at 12.000 x g, and filtered through a 0.22 μm membrane (Millipore).

Both of the recombinant toxin A fragments were purified using Bovine thyroglobulin affinity chromatography using the method essentially described by Kirvan and Wilkins (1987 Infect. Immun. 55: 1873-1877) . Soluble protein harvested as described above was diluted with an equal volume of TBS (0.05 M Tris. 0.15 M NaCl. pH 7.0) and cooled to 4°C. The chilled material was then passed through 4 ml of pre-chilled Affi-Gel 15 resin (Bio Rad) which had been coupled to 140 mg of bovine thyroglobulin (Sigma). After 4 repeat applications of the cellular lysate at 4°C, the column was washed with at least 40 bed volumes of cold TBS. and the bound material eluted with TBS pre- warmed to 37°C. Purified material collected was concentrated witii Centriplus™ 30 centrifugal concentrators (Amicon, Stonehouse, U.K.), and stored in aliquots at -70°C.

Polyacrylamide gel electrophoresis (SDS-PAGE) and Immunoblot.

Proteins were separated using 10% (w/v) polyacrylamide gels and the discontinuous buffer system of Laemmeli et al. (1970 Namre. 227:680-685) . Prior to SDS-PAGE analysis, the protein content of samples was determined using the BCA assay (Pierce &Warriner Ltd, U.K.). Proteins separated by SDS-PAGE were transferred to Hybond C nitrocellulose membrane (Amersham Life Science Ltd, U.K.) and processed as described ( Christodoulides, M.et al.. 1993 J. Gen. Microiol. 139:1729-1738) . Membranes were probed with either the monoclonal antibody PCG-4 specific for the C- terminal repeat region of toxin A, or a polyclonal rabbit anti-TETC antiserum.

Haemmagglutination activity of toxin A fragments.

Samples of protein were serially diluted 2-fold in PBS within a 96 well U- bottomed plate (Sterilin) and then reacted with 25 μl of a 2% (w/v) suspension of washed rabbit erythrocytes at 4^C (TCS Microbiology, Botolph Claydon, U.K.) for 18 h. Wells containing the highest dilution of protein able to promote 100%> agglutination of erythrocytes were taken as the endpoint, and each assay was performed in duplicate.

Animals and immunisation.

Groups of five female BALB/c mice aged 6-8 weeks old (Harlan Olac,

Bicester, U.K.) were immunised intranasally with combinations of either 10 μg of p56HIS or p56TETC, 5 μg of TETC, and l μg of E.coli LT. Each inoculum was diluted to a final volume of 30 μl in PBS (pH 7.2), and 15 μl administered to each nostril of lightly anaesthetised animals (Halothane. Rhone Merieux). Mice were immunised with 3 identical doses administered on days 0, 20 and 35. Serum samples were collected from the tail vein of each animal 1 day prior to immunisation. All mice were exsanguinated on day 47 by cardiac puncture and nasal, lung and intestinal lavage carried out with 0.1% (w/v) BSA in PBS as described (Douce G et al., 1995 Proc. Natl. Acad. Sci. USA. 92:1644-1648) . For intestinal lavage, all samples were stored witii 1 mM PMSF to inhibit intestinal proteases.

ELISA.

Anti-toxin A and anti-tetanus toxoid (TT) antibody responses widiin individual serum samples were determined by ELISA. These assays were performed as described by Douce et al. (1997 Infect. Immun. 65:2821-2828) , using ELISA wells coated overnight at 4°C with either purified whole toxin A (0.15 μg/well in 0.1 M NaHCOs, pH 9.5 ) or formalin inactivated TT (0.5 μg/well in PBS, pH 7.2). ELISA titres were determined as the reciprocal of the highest serum dilution which gave an absorbance value of 0.5 units above the background for toxin A. and 0.3 units above the background for TT. All titres were standardised against either monoclonal antibody PCG-4, or polyclonal anti-TETC positive control antiserum.

Measurement of anti-toxin antibody levels in mucosal secretions.

The levels of both anti-toxin A and anti-TT specific IgA within the mucosal lavage samples were determined by ELISA using the same antigen coating concentration as above and the method of Douce et al. ( Douce G et al.,1995 Proc. Natl. Acad. Sci. USA. 92:1644-1648) . ELISA titres were calculated as the reciprocal of the highest dilution which gave an absorbance 0.2 units above the background.

Antibody-mediated toxin A neutralisation.

The toxin A neutralising properties of both antiserum and mucosal lavage samples was determined using CHO-K1 cells (Tucker KD et al., 1990 J. Clin. Microbiol. 28:869-871) and thyroglobulin affinity-purified toxin A (Kirvan HC et al.. 1987 Infect. Immun. 55:1873-1877) . In brief. CHO-K1 cells (passage number 28-38) were grown in Ham's F12 medium (Sigma) containing 10% (v/v) foetal bovine serum and 1 mM L-glutamine and supplemented with streptomycin/neomycin/penicillin (Sigma). Freshly trypsinised CHO-K1 cells were seeded into 96 well trays (Corning Costar) at 3 x 10^ cells per well and allowed to recover for 24 h at 37°C in a 5%> CO2 atmosphere. The test sample was serially diluted two-fold in growth medium, and allowed to react with toxin A (0.6 μg/ml final concentration) for 90 min at 37°C. The toxin A-antiserum mixtures were then added to the CHO-K1 cells to give a final total volume of 100 μl for antiserum or 125 μl for lung lavage, and the cellular morphology noted after a 24 h incubation at 37°C and 5% CO2- The neutralising titre was taken as the highest dilution of sample to prevent 100% cellular rounding. All samples were tested in duplicate, and individual assays standardised against a positive control rabbit antiserum which had been raised against a conserved decapeptide located within the C- terminal region of toxin A (Wren BW et al., Infect. Immun. 59:3151-3155) .

Statistical analysis. Unpaired Student's t-test was used to compare unrelated groups of data. If the SD were found to be significantly different, the Mann-Whitney nonparametric test was used to determine the statistical relatedness of the data groups. P values < 0.05 were considered to be statistically significant. Statistical analysis was caπied out with the InStat statistical package (Sigma).

RESULTS

Construction, expression and affinity purification of recombinant 14 C- terminal toxin A repeat protein.

The mucosal immunogenicity of the 14 C-terminal toxin A repeats was determined using purified recombinant proteins containing the toxin repeats. Proteins were generated by cloning a PCR fragment encoding the entire 14 C- terminal repeats of toxin A (aa 2387-2706 inclusive) into both pRSET-A and pTECH-1 vectors to produce two constructs p56HIS and p56TETC. Clone p56HIS in E.coli generated a recombinant 14 toxin A repeat protein with six adjacent histidine residues attached to the N-terminus. Construct p56TETC when expressed in S. typhimurium produced a fusion between the 14 toxin A repeats and TETC of tetanus toxin. SDS-PAGE analysis of E.coli cellular lysates expressing p56HIS showed the presence of an additional protein with a mean apparent molecular mass of 42 kDa (calculated from three separate gels) which coπesponded to the toxin A-polyhistidine fusion (Fig. 16A, lane 1). Similar analysis of recombinant S. typhimurium lysates showed an additional band with a mean apparent molecular mass of 83 kDa which coπesponded to the toxin A-TETC fusion protein (Fig. 16B, lane 1) .

Both the p56HIS and p56TETC fusion proteins were successfully purified by utilising the inherent affinity the toxin A repeat region has for the trisaccharide Ga l-3, Gal_βl-4, GlcNac present within bovine thyroglobulin. SDS-PAGE analysis showed both the affinity purified p56HIS (Fig. 16A, lane 3) and p56TETC (Fig. 16B, lane 3) proteins to be relatively free from other E.coli or S. typhimurium proteins respectively. Immunoblot analysis showed both proteins to be immunoreactive with monoclonal antibody PCG-4 which is specific for the toxin A repeat region (Fig. 17A). The p56HIS purified material also appeared to contain a weaker immunoreactive 41 kDa protein. This protein also reacted with a monoclonal antibody specific for the polyhistidine tag attached to the N-terminus of the toxin A repeats, indicating that limited proteolytic degradation had occuπed at the C-terminus of the 42 kDa protein (data not shown). The p56TETC protein also reacted with a polyclonal anti- TETC antiserum in immunoblot to generate a 83 kDa immunoreactive band plus several weaker bands of lower molecular weight (Fig. 17B). As these weaker bands did not react with the PCG-4 monoclonal antibody, limited proteolysis of TETC appeared to have occuπed. Haemagglutination and cytotoxic properties of the 14 C-terminal toxin A repeats.

As the attachment of either the polyhistidine tag or TETC to the N-terminus of the 14 toxin A C-terminal repeats could compromise the ability of the repeats to bind to known toxin A receptors, the binding of these recombinant proteins to the Gal« 1-3, Galpl-4, GlcNac trisaccharide toxin A receptor present on the surface of rabbit erythrocytes was determined (Kirvan HC et al., 1986 Infect. Immun. 53:573-581) . This data shows that although there was no significant difference in the minimum amount of p56HIS and p56TETC required to bind to the erythrocyte surface and promote 100% haemagglutination as measured by Student^'s t-test (P > 0.05) (Table 1), both recombinant proteins failed to agglutinated the erythrocytes as effectively as whole toxin A. In both cases, the recombinants were over a 100 fold less efficient than the native whole toxin.

TABLE 1. Comparison of the minimum amount of purified 14 C-terminal toxin A repeat protein (pMole) required to promote 100% agglutination of rabbit erythrocytes (2% (v/v)). Values shown represent mean ± SEM from three separate experiments, each one performed in duplicate using protein taken from 3 different batches of purified protein.

Protein preparation Amount of protein (pMole) required to promote 100% haemagglutination of rabbit RBC p56HIS 3.61 ± 2.77 p56TETC 5.28 ± 1.74 Toxin A 0.05 ± 0.002 Tetanus toxoid No haemmagglutination ^a

^a 10 pMole tested

The toxin A receptor present on rabbit erythrocytes is also found on the surface of several other cell types such as CHO-K1 cells, thus allowing toxin A-mediated cytotoxicity to be quantified in vitro ( Katoh T et al., 1986 FEMS. Microbiol. Lett. 34:241-244) . Cytotoxicity was tested by incubation of both p56HIS and p56TETC with monolayers of CHO-K1 cells. Even when 10 pM of purified protein was tested no cytotoxic activity was observed, thus confirming the 14 C-terminal repeats to be non-toxigenic (data not shown). In contrast. 0.2 pM of whole toxin A was sufficient to promote cytotoxicity.

Anti-toxin A responses in mice immunised i.n. with toxin A repeat proteins. Mice were immunised i.n. with 3 doses of either the p56HIS or p56TETC affinity purified C-terminal toxin A fragments as described. To circumvent the potential problem of poor immunogenicity of the toxin A fragments when given i.n., LT was also co-administered with the proteins as a mucosal adjuvant. ELISA analysis of serum samples taken after 1, 2 and 3 doses showed the p56HIS protein containing the 14 C-terminal toxin A repeats to be immunogenic, inducing low levels of serum anti-toxin A antibody after a single dose which were boosted with subsequent doses (Fig. 18). The responses after three doses were significantly higher (P < 0.05) than the background titres seen in control mice immunised with TETC alone (data not shown). In contrast, the responses following co-administration of the p56HIS with LT were very impressive, with titres of anti-toxin A over 4000 fold greater than generated by the recombinant protein alone (P < 0.05). Interestingly, the genetic coupling of TETC to the toxin A repeats (p56TETC) was also shown to increase anti-toxin A responses, raising the mean anti-toxin A titre approximately 8 fold higher than the recombinant alone after 3 doses (P < 0.05). Anti-toxin A titres induced by p56TETC genetic fusion were also further increased by the addition of LT.

Serum samples containing high titres of LT-specific antibody were unable to cross-react with toxin A in ELISA (data not shown).

Anti-toxin A mucosal IgA responses after immunisation with toxin A repeat proteins.

To evaluate whether p56HIS was also capable of stimulating the production of toxin A specific IgA at the mucosal surface, lavage samples were collected from immunised mice and analysed by ELISA. The results showed p56HIS to be incapable of generating significant levels of toxin A-specific IgA at these mucosal surfaces (Fig. 19). However, the addition of TETC to the 14 toxin A repeats as a fusion (p56TETC) resulted in significant (P < 0.05) levels of toxin A-specific IgA in both nose and lung lavage samples. The addition of LT to both recombinant proteins significantiy (P < 0.05) increased the levels of toxin A-specific IgA in both the nose and lung samples.

Lavage samples were also taken from the small intestine of all immunised mice and the level of toxin A-specific IgA determined as above. Even in the presence of LT, both p56HIS and p56TETC did not generate significant levels of toxin-specific IgA (data not shown).

Toxin A neutralisation with serum and mucosal antibodies harvested from i.n. immunised mice.

Toxin A exhibits a cytotoxic effect on CHO-K1 cells in vitro which can be neutralised by specific antibodies (f atoh T et al., 1986 FEMS. Microbiol. Lett. 34:241-244 and Wren SB et al., 1991 Infect. Immun. 59:3151-3155) . Therefore, serum and mucosal lavage samples were tested for the ability to neutralise the cytotoxic activity of whole toxin A. Serum harvested from all five mice immunised with p56HIS were unable to neutralise toxin A (Table 2). However, serum from one out of the five mice which received p56TETC was able to neutralise toxin A. The addition of LT to p56HIS significantly increased the mean neutralising titre, with all five immunised mice generating toxin neutralising antibody, while four out of the five mice immunised witii p56TETC + LT produced toxin neutralising antibodies.

TABLE 2. Toxin A neutralising properties of both serum and mucosal antibodies harvested after 3 intranasal doses of antigen. Toxin-neutralising titres were scored as the highest dilution of antiserum or mucosal lavage to promote 100%) neutralisation of toxin A (60 ng/well) as measured against CHO-Kl cells in vitro. Mean neutralising titres ± SD for five mice are shown, with each assay being performed in duplicate.

Lung lavage samples were also tested for toxin A neutralising activity since these samples had been shown to contain high levels of toxin A-specific IgA by ELISA. Although the level of neutralisation was lower than that seen with the coπesponding antiserum, single mice from the p56TETC + LT immunised group did generate sufficient levels of IgA at the lung mucosa to neutralise toxin A (Table 2). However, the highest neutralisation titre was seen with lavage sample taken from p56HIS + LT immunised mice, with two out of five samples possessing toxin A neutralising activity.

Anti-tetanus serum responses in mice immunised i.n. with toxin A repeat proteins.

In addition to generating anti-toxin A antibodies, the level of anti-TT serum antibodies was also determined (Fig. 20). The control mice immunised witii TETC alone generated levels of anti-TT in the blood which were higher than expected. However, although the mean antibody titres induced by p56TETC after 3 i.n. doses were similar to those obtained with TETC alone (P > 0.05), when p56HIS was co-administered with TETC. the mean titres of anti-TT antibody were found to be approximately 4 fold higher after both 2 and 3 i.n. doses than the titres seen with TETC only. Again, the addition of LT to both p56TETC and p56HIS + TETC significantly (P < 0.05) increased the anti-TT antibody titres, even after 2 doses. LT-specific antibodies were unable to cross-react with TT in ELISA (data not shown).

Anti-tetanus IgA responses in mice immunised i.n. with toxin A repeat proteins. In contrast to the serum responses, mice immunised with TETC alone induced very poor immune responses at both the nose and lung mucosal surfaces (Fig. 21). However, mice immunised with either p56HIS + TETC + LT or p56TETC +LT induced strong anti-TT responses in the nose and lungs. These titres were very consistent within the group of 5 mice. Interestingly both nose and lung lavage samples taken from mice immunised with p56HIS + TETC were also shown to contain very high levels of anti-TT specific IgA. These mean antibody titres were higher than those obtained with TETC alone in both nasal washes (P < 0.05) and lung washes (P = 0.05). Although the spread of the response was much greater in mice immunised with D56HIS + TETC, the mean response of the group was not significantly (P > 0.05) different from the mice co-immunised with LT as an adjuvant.

SEQUENCE LISTING

SEQ ID No. 1

ASTGYTSING KHFYFNTDGI MQIGVFKGPN GFEYFAPANT DANNIEGQAI

LYQNKFLTLN GKKYYFGSDS KAVTGLRTID GKKYYFNTNT AVAVTGWQTI

NGKKYYFNTN TSIASTGYTI ISGKHFYFNT DGIMQIGVFK GPDGFEYFAP

ANTDANNIEG QAIRYQNRFL YLHDNIYYFG NNSKAATGWV TIDGNRYYEE

PNTAMGANGY KTIDNKNFYF RNGLPQIGVF KGSNGFEYFA PANTD ANNIE

GQAIRYQNRF LHLLGKIYYF GNNSKAVTGW QTINGKVYYF

MPDTAMAAAG GLFEIDGVIY FFGVDGVKAP

Claims

1. A molecule which: a) comprises an amino acid sequence as shown in

Figure 6; b) has one or more amino acid substitutions, deletions or insertions relative to the sequence defined in a) above; c) is a fragment of a sequence as defined in a) or b) above, which is at least ten amino acids long; d) comprises a multiple of a sequence as defined in a), b) or c); or a sequence substantially homologous thereto wherein said molecule is capable of eliciting an immune response in an animal.

2. A molecule as claimed in claim 1, in which the molecule is a fusion protein.

3. A molecule as claimed in claim 2, in which the fusion protein comprises tetanus toxin fragment C.

4. The use of a molecule as defined in any one of claims 1 to 3 in the preparation of an agent for the prophylaxis or treatment of a Clostridium difficile infection.

5. An antibody to a molecule as defined in any one of claims 1 to 3.

6. The use of an antibody as defined in claim 5 in the preparation of an agent for the prophylaxis or treatment of a Clostridium difficile infection.

7. The use of a molecule as defined in any one of claims 1 to 3 in the preparation of a vaccine against C. difficile infection.

8. A vaccine formulation comprising a molecule as defined in any one of claims 1 to 3, optionally together with one or more caπiers and/or adjuvants.

9. A vaccine formulation as claimed in claim 8, in which the adjuvant comprises cholera toxin (CT), heat-labile toxin (LT), or a detoxified derivative thereof.

10. A vaccine formulation comprising a molecule as defined in any one of claims 1 to 3, in which the molecule is expressed in a live delivery vector.

11. A vaccine formulation as claimed in claim 10, in which die live delivery vector is Salmonella typhi or Salmonella typhimurium.

12. A vaccine formulation comprising a molecule as defined in any one of claims 10 and 11, optionally together with one or more carriers and/or adjuvants.

13. A vaccine formulation as claimed in claim 12 , in which die adjuvant comprises cholera toxin (CT), heat-labile toxin (LT), or a detoxified derivative thereof.

14. A recombinant DNA construct comprising nucleotides 7159-8118 of C. difficile as shown in Figure 1.

15. A method for the preparation of a fragment of toxin A of C. difficile, comprising the step of expressing a nucleotide sequence encoding a molecule as defined in any one of claims 1 to 3 in a host cell.

16. A method as claimed in claim 14 , in which the host cell is E. coli,

Salmonella typhi or Salmonella typhimurium.

17. A method of immunising a subject against infection by C. difficile, which comprises administering to the subject a vaccine formulation as defmed in any one of claims 8 to 13.

18. A method for the prophylaxis or treatment of C. difficile infection which comprises the step of administering to a subject a vaccine formulation as defined in any one of claims 8 to 13.

19. The use of a molecule as defined in any one of claims 1 to 3 in the diagnosis of C. difficile infection in a biological sample.

20. The use of an antibody as claimed in claim 5 in die diagnosis of C. difficile infection in a biological sample,

21. A use as claimed in claim 19 or claim 20 in which the biological sample is blood, saliva, tears, urine, faeces, semen or milk.

22. A vaccine formulation comprising a recombinant DNA construct as claimed in claim 14 , optionally together with one or more caπiers or adjuvants.

23. A kit for use in the diagnosis of C. difficile infection comprising a molecule as defined in any one of claims 1 to 3.

24. A method of diagnosing a Clostridium difficile infection, the method comprising the step of contacting an antibody as defined in claim 5 with a biological sample from an affected individual.

25. Use of a molecule as defined in claim 1. as adjuvant in the administration of a peptide or protein.

26. Use according to claim 25, wherein the molecule as defined in claim 1 and the peptide or protein are co-administered in a mixture.

27. Use according to claim 25, wherein the molecule as defined in claim 1 and the peptide or protein are administered as a fusion protein.

28. Use according to any of claims 25 to 27, wherein the molecule comprises an amino acid sequence as shown in Figure 6.