WO2006130925A1

WO2006130925A1 - Genetic manipulation of clostridium difficile

Info

Publication number: WO2006130925A1
Application number: PCT/AU2006/000800
Authority: WO
Inventors: Julian Ian Rood; Dena Lyras; Jennifer Ruth O'connor
Original assignee: Monash University
Priority date: 2005-06-10
Filing date: 2006-06-09
Publication date: 2006-12-14

Abstract

This invention relates to methods of genetic engineering of bacteria. In particular, the invention provides a method by which mutations in genes of the bacterium Clostridium difficile can be achieved in a targeted and reproducible fashion by insertional inactivation, optionally accompanied by introduction of a desired nucleic acid molecule of interest into the bacterial cell.

Description

GENETIC MANIPULATION OF CLOSTRIDIUM DIFFICILE

PRIORITY

This application claims priority from Australian provisional patent application No. 2005903063 dated 10 June 2005, the entire contents of which are incorporated herein by this reference.

FIELD

This invention relates to methods of genetic engineering of bacteria. In particular, the invention provides a method by which mutations in genes of the bacterium Clostridium difficile can be achieved in a targeted and reproducible fashion.

BACKGROUND

All references, including any patents or patent application, cited in this specification are hereby incorporated by reference to enable full understanding of the invention. Nevertheless, such references are not to be read as constituting an admission that any of these documents form part of the common general knowledge in the art, in Australia or in any other country. The discussion of the references states what their authors assert, and the applicants reserve the right to challenge the accuracy and pertinency of the cited documents. Clostridium difficile is a spore-forming bacterium which is widely distributed in the environment, and is known to survive for months in hospitals and long- term care facilities. It is the causative organism of C. difficile-associated diarrhoea (CDAD) , a form of infectious diarrhoea, and of pseudomembranous colitis and pseudomembranous enterocolitis. The major toxins of C. difficile can be detected in approximately 30% of cases of antibiotic-associated, colitis. Patients are at risk of developing CDAD when they are treated with antibiotics which alter the population of normal, protective bacteria which reside in the colon. Virtually all antibiotics have been implicated in causing CDAD, and the number of reported cases of CDAD infections has increased in recent years . C. difficile represents a significant risk to many hospital and long-term care patients, especially among the elderly. C. difficile is the most common cause of infectious, hospital-based diarrhoea and, although incidence is not officially tracked, may result in more than 400,000 cases of diarrhoea and colitis annually in the United States, leading to death in about 5,000 cases.

In the United States as many as 10 percent of all patients hospitalized for more than two days are afflicted with CDAD, which often adds up to two weeks to the length of patient hospitalization. This can add as much as $10,000 to $15,000 to the cost of a hospital stay. CDAD is also implicated in the majority of cases of antibiotic- associated colitis, which is one of the leading causes of morbidity and mortality among elderly debilitated patients . The rising incidence of CDAD has been attributed to the increasingly frequent prescription of broad-spectrum antibiotics for hospitalized patients. Over 80% of cases occur in the over-65s, and

C. difficile is now by far the commonest enteric pathogen isolated from such individuals. Elderly people in hospitals, nursing homes^', and other chrόnic-care facilities are at particular risk, and CDAD outbreaks can be devastating, both in terms of mortality and the cost of disease management in terms of disruption to services, patient isolation to a separate ward, revised supportive therapy, specific therapy to eliminate C. difficile, scrupulous hygiene in nursing, environmental decontamination, and ward closure. Whilst mild CDAD can often be treated by removing the provoking antibiotic so that the normal gut flora re-establishes and excludes C. difficile, this is not practicable if the patient is undergoing treatment for another infection.

C. difficile proliferates in the setting of altered normal colonic bacterial flora, which is most commonly due to the administration of broad-spectrum antibiotics. Once established in the colon, C. difficile produces toxins which disrupt the intestinal lining, causing cell death and inflammation which result in diarrhoea and colitis. Vancomycin is currently the only therapy approved by the U.S. Food and Drug Administration for treatment of CDAD . Metronidazole is also widely used in other countries . Even after successful treatment with the current standard of care, approximately 20 percent of patients experience a recurrence of CDAD which may require repeat hospitalization. In addition; a subset of patients with CDAD develop multiple recurrences of the disease, with symptoms that may persist for years. Treatment of severe or fulminant C. difficile colitis may require removal of part or all of the colon.

Consequently there have been intensive efforts to develop new therapies for treatment of CDAD, and a variety of anti-bacterial agents is in clinical trial. Antibodies against C. difficile toxins are also being developed. However, these efforts have been hampered by poor understanding of the underlying pathogenic processes involved in C. difficile infections, and the fact that genetic manipulation of C. difficile has been very difficult. This in turn has meant that, although a number of genes from this organism have been cloned, its complete gene sequence has been determined, and microarrays with most of the genes are available, there is very little information available about the role of many of these genes in the physiology or virulence of the organism. In particular, one standard method, insertional inactivation, which is widely used in such investigations with other organisms has not been available for use with C. difficile . Although there have been reports of insertional inactivation of individual genes in C. difficile, these have proved to be either non-reproducible or not generally- applicable. For example, a C. difficile suicide vector, ie . one which could not autonomously replicate in C. difficile, has been used to target the gldA gene; the vector was integrated into the chromosome via homologous recombination, resulting in insertional inactivation of the gldA gene (Liyanage, Kashket et al . 2001). Only the inactivation of a single gene in an avirulent strain was demonstrated. This paper is the only report of a suicide vector being used for insertional inactivation of a target gene. In fact, another group has recently reported that they attempted to use this method to inactivate another gene, and could not achieve the desired outcome, raising doubts as to the reproducibility and viability of this approach (Carter, Purdy et al . 2005) . We also have been unable to effect inactivation of C. difficile by this method. An alternative approach used introduction of a

Tn916-hased transposon into C. difficile to introduce an intact copy of the sigK gene into the chromosome of a C. difficile strain which had a naturally inactivated copy of sigK. It was expected that the transposon would insert into the chromosome at a defined location, specific to the nature of the transposon. However, homologous recombination occurred instead, resulting in an integration of the transposon into the site where there was homology between the transposon and the chromosome, ie. at the location of the sigK gene (Haraldsen and Sonenshein 2003) . Because this method has been employed in the study of only one gene, its reproducibility has not been established.

Another system is an extremely inefficient and cumbersome procedure based on the Gram-positive conjugative transposon, Tn916, which provides a means of introducing cloned fragments back into C. difficile (Mullany, Wilks et al. 1994). The frequency of transfer is, however, extremely low, being 10^"8 per donor, and the delivery system involved in introducing the transposon into C. difficile is cumbersome . Moreover the action of the transposon in this system is unpredictable, because the transposon can be incorporated into the chromosome with or without homologous DNA being present; thus the insertion does not necessarily occur by homologous recombination, and may occur at multiple sites.

In US Patent NO.5955368, the entire contents of which are incorporated herein by this reference, two of the present inventors and their coworkers disclosed a family of mobilizable conjugative transfer plasmids which could be used as shuttle vectors to introduce genes from Escherichia coli into bacteria of the genus Clostridium. A variant of this vector has been used in an avirulent C. difficile strain as a shuttle vector (Mani, Lyras et al . 2002). However, it has not been suggested that such a shuttle vector could be used as a suicide vector to inactivate target genes in C. difficile. The use of a shuttle vector to construct mutants of Clostridium acetobutylicum by insertional inactivation has been reported (Harris, Welker et al. 2002) . However, the organism used was a species which has different properties to those of C. difficile, in that it is not pathogenic and does not produce a toxin, the shuttle vector used was entirely different to the one used in the present invention, the vector was designed to make double crossover mutants rather than single cross-over mutants, and the vector was introduced by transformation rather than by conjugation.

International patent application No.

PCT/GBOl/01612 discloses a plasmid for transformation of C. difficile, comprising a C. difficile replicon, a restriction endonuclease site, and preferably also a selectable marker. It is proposed in PCT/GBOl/01612 that this vector could be used to introduce anti-sense DNA into C. difficile in order to inactivate target genes, or for deactivation by homologous recombination and insertional activation. However, transformation other than with anti- sense DNA was not demonstrated, and there was no evidence that homologous recombination was actually achieved. In fact, subsequent reports from this laboratory show that the method described in this specification is not useful for transformation, and did not refer to any homologous recombination or gene inactivation experiments (Purdy, O'Keeffe et al. 2002) . Thus hitherto it has not been possible to use the technique of reverse genetics reproducibly in order to study the specific contribution of individual genes to overall virulence of C. difficile . There is a need in the art for methods for reproducible methods of genetic manipulation of C. difficile in order to explore these questions, and to develop improved therapeutic methods and vaccines for this organism.

SUMMARY

We have now found that C. difficile genes can be mutated in a simple, targeted and reproducible fashion via insertional inactivation. When the E. coli-C. perfringens shuttle vector pJIR1456 is introduced into C. difficile, and selective pressure is applied, the vector is autonomously maintained. Surprisingly, however, we have found that when a pJIR1456-based shuttle vector carrying C. difficile DNA homologous to a region of the host genome is introduced into this host and selective pressure is applied then subsequently removed, the vector is unstable unless homologous recombination occurs between the vector and the chromosome. This results in incorporation of the vector into the chromosome at the site where the plasmid DNA is homologous to the chromosome. Consequently when the selective pressure is removed, plasmid DNA is passed on to subsequent generations only when the plasmid has integrated with the host chromosome. Thus in this system the shuttle vector is acting as a type of conditional lethal suicide vector which we have termed a recombination vector.

We have taken advantage of this phenomenon by specifically designing pJIR1456-based vectors to inactivate C. difficile genes of interest. To our knowledge, prior to the development of this technique no simple or reproducible method existed for the insertional inactivation of C. difficile genes in a virulent strain; additionally, our use of the shuttle vector, which can either autonomously replicate, or integrate with the chromosome, is unique in the C. difficile system.

In a first aspect, the invention provides a method of inactivating a target gene in a Clostridium difficile host cell, comprising the steps of (i) introducing a plasmid which is unstable in C. difficile into the host cell,

(ii) exposing the host cell to conditions which select for cells expressing the selectable marker, thereby providing ^' sufficient time to permit homologous recombination between the C. difficile DNA in the plasmid and C. difficile DNA in the host cell to occur, and (iii) removing the selective conditions, in which the plasmid comprises

(a) an origin of conjugative transfer capable of modulating conjugative transfer of the plasmid from a non- clostridial donor species into a clostridial species;

(b) a gene encoding a selectable marker which functions in both a Clostridium species and in the non- clostridial donor species; (c) an origin of replication which is functional in the non-clostridial donor species;

(d) an origin of replication which is functional in a Clostridium species; and

(e) a fragment of C. difficile DNA which is homologous to a region of DNA in the target gene.

In a second aspect the invention provides a method of inactivation of a target gene in a C. difficile host cell and introducing a nucleic acid molecule of interest into the cell, comprising the steps of (i) introducing a plasmid which is unstable in C. difficile into the host cell, (ii) exposing the host cell to selective conditions which select for cells expressing the selectable marker, thereby- providing sufficient time to permit homologous recombination between the C. difficile DNA in the plasmid and C. difficile DNA in the host cell to occur, and (iii) removing the selective conditions, in which the plasmid comprises

(a) an origin of conjugative transfer capable of modulating conjugative transfer of the plasmid from a non- clostridial donor species into a clostridial species; (b) a gene encoding a selectable marker which functions in both a Clostridium species and in the non- clostridial donor species;

(c) an origin of replication which is functional in the non-clostridial donor species; (d) an origin of replication which is functional in a Clostridium species,- e) a fragment of C. difficile DNA which is homologous to a region of DNA in the host cell; and f) a nucleic acid molecule of interest. In order to introduce a desired nucleic acid molecule of interest it is not essential for a host gene to be inactivated. Thus in a third, alternative aspect, the invention provides a method of introducing a nucleic acid molecule of interest into a C. difficile host cell, comprising the steps of

(i) introducing a plasmid into the host cell, (ii) exposing the host cell to selective conditions which select for cells expressing the selectable marker, thereby providing sufficient time to permit homologous recombination between the C. difficile DNA in the plasmid and C. difficile DNA in the host cell to occur, and (iii) removing the selective conditions, - S -

in which the plasmid is unstable in C. difficile, and comprises

(a) an origin of coήjugative transfer capable of modulating conjugative transfer of the plasmid from a non- clostridial donor species into a clostridial species;

(b) a gene encoding a selectable marker which functions in both a Clostridium species and in the non- clostridial donor species;

(c) an origin of replication which is functional in the non-clostridial donor species;

(d) an origin of replication which is functional in a Clostridium species; e) a fragment of C. difficile DNA which is homologous to a region of non-coding DNA or a repeated region of DNA in the host cell; and f) the nucleic acid molecule of interest.

In one embodiment the non-coding DNA encodes a ribosomal

RNA gene.

In all these three aspects the host cell may optionally subsequently be recovered, and/or subjected to further manipulation.

Any plasmid comprising the features identified above may be used. The plasmid may be of clostridial origin. Suitable plasmids include those described in US Patent No. 5995368, such as pJIR1457 and pJIR1456, and their derivatives . Optionally the plasmid also comprises a plasmid replication sequence and/or a multiple cloning site. However, the invention has enabled novel vectors to be constructed. Thus in a fourth aspect, the invention provides a mobilisable conjugative recombination vector for gene inactivation in C. difficile, comprising

(a) an origin of conjugative transfer capable of modulating conjugative transfer of the plasmid from a non- clostridial donor species into a clostridial species,-

(c) an origin of replication which is functional in the non-clostridial donor. species;

(d) an origin of replication which is functional in a Clostridium species;

(e) at least a fragment of C. difficile RgaR and RgbR response regulator DNA

(f) a nucleic acid segment which prevents read- through expression from the vector, and (g) a selectable marker.

In particular embodiments, the nucleic acid segment which prevents read-through expression from the vector may be an Ω transcriptional terminator cassette; the selectable marker may be an antibiotic resistance marker, such as a macrolide-lincosamide-streptogramin antimicrobial resistance marker; and the macrolide-lincosamide- streptogramin resistance marker may be ermQ. In one embodiment the origin of conjugative transfer may be RP4 oriT, as described in US5955368. In specific embodiments the conjugative recombination vector is a plasmid selected from the group consisting of pJIR2512, pJIR2515, pJIR2363, pJIR2364, pJIR2633, pJIR2634, pJIR3012, pJIR3013, pJIR3014 and pJIR3015. In a sixth aspect, the invention provides a complementation vector for C. difficile, comprising

(a) an origin of conjugative transfer capable of modulating conjugative transfer of the plasmid from a non- clostridial donor species into a clostridial species; (b) a gene encoding a selectable marker which functions in both a Clostridium species and in the non- clostridial donor species,-

(c) an origin of replication which is functional in the non-clostridial donor species,- (d) an origin of replication which is functional in a Clostridium species;

(e) a C. difficile RgaR or RgbR response regulator DNA, or a fragment thereof;

(f) a nucleic acid segment which prevents read- through expression from the vector, and

(g) a selectable marker. In one embodiment the complementation vector is plasmid pJIR34041 or plasmid pJIR34042.

In particular embodiments of all the foregoing aspects of the invention, the origin of conjugative transfer may be capable of modulating conjugative transfer of the plasmid from Escherichia coli into a clostridial species; the gene encoding a selectable marker may be a gene which functions in both clostridial hosts and in Escherichia coli; the origin of replication which is functional in the non-clostridial donor species may be functional in Escherichia coli; and the origin of replication which is functional in a Clostridium species may be functional in Clostridium perfringens, respectively.

In a seventh aspect, the invention provides a double crossover suicide vector for C. difficile, comprising

(a) a region of DNA upstream of a target gene;

(b) a region of DNA downstream of a target gene;

(c) a first antibiotic resistance marker which selects for integration into the chromosome, and (d) a second antibiotic resistance marker which selects against single crossovers and independently replicating plasmids .

In particular embodiments the target gene is rgaR; the region upstream of the rgaR gene may comprise the first 257 bp of the coding region; the first antibiotic resistance marker may be an ermQ erythromycin resistance cassette, and the second antibiotic resistance marker may be catP. In a specific embodiment the vector is pJIR3011, In an eighth aspect the invention provides a C. difficile cell in which a target gene has been subjected to insertional inactivation by homologous recombination, and into which a nucleic acid molecule of interest has optionally been introduced.

In a ninth aspect the invention provides a recombinant C. difficile strain transformed with or comprising a vector according to the invention. In particular embodiments the recombinant C. difficile strain is selected from the group consisting of JIR8149, JIR8150, JIR8218, JIR88223, JIR8226, JIR8145: JIR8094 (pJIR1456) , JIR8149 : JIR8094 (pJIR2634 ) , JIR8150.-JIR8094 (pJIR2633) , JIR8149 : JIR8094 (pJIR2634) , JIR8150: JIR8094 (pJIR2633 ) , JIR8218 : JIR8094ΩpJIR3014 , JIR8223: JIR8094ΩpJIR3015 , JIR8226: JIR8094ΩpJIR3015 , JIR8234: JIR8218, JIR8236 : JIR8218 , JIR8243 : JIR8223 , JIR8233: JIR8223, JIR8245 : JIR8226 , JIR8247 : JIR8226 , and JIR8292. In a tenth aspect the invention provides a composition comprising a C. difficile cell according to the invention, together with a pharmaceutically-acceptable carrier. The composition may be formulated as a vaccine, in which case it may also comprise an adjuvant. The composition may be administered orally, intranasally or parenterally, e.g. subcutaneousIy or intramuscularly.

In particular embodiments, the method of the invention results in production of a C. difficile host cell of reduced virulence, or a host cell in which virulence is abolished.

In one embodiment the gene which is inactivated is one which encodes a protein involved in virulence of C. difficile, for example toxin A, toxin B or a factor which regulates the expression of one or both of these toxins. Preferably virulence of the host cell is reduced or abolished.

It will be appreciated that the method of the invention provides a means of constructing a modified C. difficile host cell in which a desired target gene is inactivated, so that the function of this gene can be investigated, for example using microarrays, by virulence testing, by complementation studies and the like. Thus in an eleventh aspect the invention provides s a C. difficile cell in which a target gene has been subjected to insertional inactivation by homologous recombination, or a progeny cell, mutant or derivative of the cell.

It will be clearly understood that all progeny of a cell of the invention, and all mutants or derivatives of the cell of the invention which retain the biological characteristics of that cell, are within the scope of the invention.

It will also be appreciated that a C. difficile host cell rendered avirulent by this method is useful in a variety of applications, for example vaccines, modification of host immune responses, modification of bacterial responses and the like.

In one embodiment of all these aspects of the invention the gene which is inactivated is one which encodes a protein involved in virulence of C. difficile, for example toxin A, toxin B or a factor which regulates the expression of one or both of these toxins. Such factors include but are not limited to promoters of the genes for toxin A and toxin B, and sigma factors such as TcdR or the putative anti-sigma factor, TcdC. Sigma factors are known in the art (See for example Dupuy and Sonenshein, 1998; Mani and Dupuy, 2001; Mani et al. , 2002; Haraldsen and Sonenshein, 2003) . Preferably virulence of the host cell is reduced or abolished.

Preferably in the second and third aspects of the invention, the nucleic acid molecule of interest is under the control of a promoter of prokaryotic origin, more preferably one which is functional in a Clostridium species, such as the tnpX promoter from Clostridium perfringens or C. difficile.

The nucleic acid molecule of interest may be homologous or heterologous. A homologous nucleic acid molecule may be used in order to investigate the effect of increasing the expression of the corresponding protein or may encode a C. difficile antigen. A heterologous nucleic acid molecule may encode any desired peptide, polypeptide or protein, including but not limited to: antibodies or fragments thereof such as ScFv fragments, including antibodies or fragments thereof coupled to toxins of non-C. difficile origin; peptide or polypeptide ligands; immunomodulatory agents such as cytokines ; peptide or polypeptide antibiotics or bacteriocins ; peptide or polypeptide anti-cancer agents; antigens of bacterial, viral or cell surface origin; peptide antigens or epitopes; enzymes, such as β-lactamases to prevent diarrhoea due to antibiotic therapy; and regulatory factors

It will also be appreciated that the nucleic acid molecule of interest may encode a regulatory factor which modulates expression of a polypeptide or non-polypeptide component of the C. difficile cell, such as a structural component or cell surface antigen.

International Patent Application No. PCT/SEOl/01280, the entire contents of which are incorporated herein by this reference, lists a variety of proteins which may be desirable to express in C. difficile host cells, and proposes a number of uses for such modified C. difficile cells. This specification also discloses the use of secretory leader sequences encoding a signal peptide from an exported clostridial N-acetyl-muramoyl-L- alanineamidase-like protein which facilitates expression of a desired protein at the cell surface. Such secretory leader sequences may optionally also be incorporated into the plasmid of the invention.

In a twelfth aspect the invention provides an isolated C. difficile response regulator nucleic acid sequence comprising

(a) the nucleic acid sequence set out in SEQ ID NO: 57 or 58;

(b) a nucleic acid molecule which is able to hybridise to (a) under stringent conditions; or

(c) a nucleic acid molecule which has at least 75% sequence identity to (a) .

In a thirteenth aspect the invention provides an isolated nucleic acid molecule comprising a VirR box region CD0587, CD2098 agrB (CD2750) , CD1667/CD1668 intergenic, CD1511/CD1512 intergenic and nCcAGTTnTnCatttttAannACcAGTTntgcnn, which is capable of being specifically bound by an RgaR-like or VirR-like protein. The nucleic acid molecule may be a

DNA or an RNA. More preferably in (b) the nucleic acid molecule is able to hybridise under stringent conditions to the molecule of (a) . More preferably in (c) the nucleic acid molecule has at least 80%, even more preferably at least 90%, yet more preferably at least 95% sequence identity sequence identity to the molecule of (a) .

In a fourteenth aspect the invention provides a protein which is encoded by the nucleic acid molecule of the invention, or by a fragment or derivative thereof which retains the biological activity of the original nucleic acid molecule.

Amino acid sequence variants of these proteins are also within the scope of the invention, and these will generally share at least about 75%, preferably greater than 80%, and more preferably greater than 90% sequence identity with one or more of the amino acid sequences deduced from the nucleic acid sequences disclosed herein, after aligning the sequences to provide for maximum homology, for example as determined by the version described by Fitch, et al., Proc. Nat. Acad. Sci. USA 80:1382-1386 (1983), of the algorithm described by Needleman, et al . , J. MoI. Biol. 48:443-453 (1970) . In a fifteenth aspect the invention provides a method of assessing a parameter selected from the group consisting of host immune response, modification of a bacterial response, virulence, gene activation and gene complementation, comprising the use of a C. difficile cell according to the invention.

In a sixteenth aspect the invention provides a method of treatment or prophylaxis of a condition associated with C. difficile, comprising the step of administering an effective dose of a composition according to the invention to a subject suffering from or at risk of the condition.

In a seventeenth aspect the invention provides the use of a composition according to the invention in the manufacture of a medicament for the treatment or prophylaxis of a condition associated with C. difficile .

In an eighteenth aspect the invention provides an isolated C. difficile protein which is down- regulated by disruption of a virR-like gene such as rgaR, and which is encoded by a gene selected from the group consisting of CD0588, CD0590, CD2098 and CD2750, or a mutant or derivative of the gene .

In a nineteenth aspect the invention provides an isolated C. difficile operon comprising open reading frames CD0587-CD0590.

In a twentieth aspect the invention provides an isolated C. difficile AgrB quorum sensing protein encoded by CD2750, or by a mutant or derivative thereof.

In a twenty-first aspect the invention provides an isolated antibody specific for a C. difficile cell according to the invention or a protein according to the invention.

The method of the invention has wide implications for the field of research into C. difficile. Prior to our discovery, no-one was able to use the technique of reverse genetics reproducibly in order to study the specific contribution of individual genes to overall virulence of the organism, using a virulent C. difficile strain. This method will allow the insertional inactivation of selected virulence genes and the study of the resultant strains in an animal model, a step which was previously impossible.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows the results Southern blotting analysis of C. difficile mutant strains. DNA was digested with Seal and AsplOO restriction endonucleases . Lanes in Southern blots: (1) pJIR2634, Seal, rgaR vector; (2) pJIR2634, Asjp700; (3) pJIR2633 Seal, rgbR vector; (4) pJIR2633 Asp700; (5) JIR8094, Seal, C. difficile wild type strain; (6) JIR8094, Asp700; (7) JIR8149, Seal, C. difficile rgaR mutant; (8) JIR8149, AsplOO, (9) JIR8150, Seal, C. difficile rgbR mutant; (10) JIR8150, Asp700. Blot (a) probed with the rgaR gene, (b) probed with the rgbR gene, (c) probed with catP; (d) Schematic representation indicating the arrangement of the JIR8149 and JIR8150 chromosomal rgaR region and rgbR region, respectively.

Figure 2 is a schematic representation showing the single cross-over suicide vectors used for insertional inactivation of rgaR and rgbR. The RP4 origin of transfer was PCR amplified from pJIR1456 and cloned into the Xmnl site of the E. coli-C. perfringens shuttle vector, pJIR750, which encodes E. coli and C. perfringens origins of replication (oriEC and oriCP, respectively) , a C. perfringens pIP404 replication gene (rep) , and the thiamphenicol resistance-encoding gene, catP. This resulted in pJIR2816. Then PCR products corresponding to internal fragments from rgaR and rgbR were cloned into pJIR2816 Sphl/Xbal. Finally, the omega cassette was cloned next to the rgaR/rgbR fragment Xbal/Asp718, resulting in pJIR3015 and pJIR3014, respectively. Figure 3 shows the results of PCR analysis of

C. difficile rgaR mutant and complemented mutant strains. (Ml) NEB PCR molecular weight markers (M2) NEB 1 Kb ladder molecular weight markers (1) JIR8094, C. difficile wild type,- (2) JIR8223 C. difficile rgaR mutant 1; (3) JIR 8226, C. difficile rgaR mutant 2; (4) JIR8243, C. difficile rgaR mutant 1 (pMTL9301) ; (5) JIR8245 C. difficile rgaR mutant 2 (pMTL9301) ; (6) JIR8233, C. difficile rgaK mutant 1

(pJIR3041) complemented; (7) JIR8247 C. difficile rgaR mutant 2 (pJIR3041) complemented; (8) pJIR3041 complementation vector. Panel (a) : PCRs on the rgaR open reading frame; Panel (b) : PCRs across the left hand side of the single cross-over region; Panel (c) : PCRs across the right hand side of the single cross-over region; Panel (d) : schematic representation of the genomic regions amplified by the PCRs, the expected product sizes are indicated. Figure 4 shows the results of PCR analysis of C. difficile rgbR mutant and complemented mutant strains . (Ml) NEB 1 Kb ladder molecular weight markers; (M2) NEB PCR molecular weight markers; (1) JIR8094, C. difficile wild type; (2) JIR8218 C. difficile rgbR mutant; (3) JIR 8234, C. difficile rgbR mutant (pMTL9301) ; (4) JIR8236, C. difficile rgbR mutant (pJIR3042) complemented; (5) pJIR3042 complementation vector. Panel (a) : PCRs on the rgbR open reading frame; Panel (b) : PCRs across the left hand side of the single cross-over region; Panel (c) : PCRs across the right hand side of the single cross-over region; Panel (d) : Schematic representation of the genomic regions amplified by the PCRs, the expected product sizes are indicated.

Figure 5 shows the results of Southern blotting analysis of C. difficile rgaR mutant strains. DNA was digested with Sad and Ncol restriction endonucleases . Lanes in Southern blots: (1) pJIR3015, rgaR suicide vector; (2) pMTL9301 shuttle vector; (3) pJIR3041 complementation vector; (M) lambda-HindiII molecular weight markers, sizes as indicated; (4) JIR8094, C. difficile wild type strain; (5) JIR8223, rgaR mutant 1; (6) JIR8226,

C. difficile rgaR mutant 2; (7) JIR8243, C. difficile rgaR mutant 1 (pMTL9301) ; (8) JIR8233 C. difficile rgaR mutant 1 (pJIR3041) ; (9) JIR8245, C. difficile rgaR mutant 2 (pMTL9301) ; (10) JIR8247 C. difficile rgaR mutant 2 (pJIR3041) . Blot (a) was probed with the rgaR gene; (b) probed with the 3' end of the rgaR coding region; (c) probed with catP; (d) probed with repA from pMTL9301; (e) Schematic representation indicating the arrangement of the chromosomal rgaR region into which the single crossover occurred, expected sizes of restriction fragments are indicated, and the regions to which the probes are expected to hybridise are marked, expected size of the intact rgaR band in the wild type strain; JIR8094 is 19.5 kb.

Figure 6 shows the results of Southern blotting analysis of C. difficile rgbR mutant strains. DNA was digested with Hpal and Ncol restriction endonucleases . Lanes in Southern blots: (1) pJIR3014, rgbR suicide vector; (2) pMTL9301 shuttle vector; (3) pJIR3042 complementation vector; (M) lambda-Hindlll molecular weight markers, sizes as indicated; (4) JIR8094, C. difficile wild type strain; (5) JIR8218, rgbR mutant; (6) JIR8234, C. difficile rgbR mutant (pMTL9301) ; (7) JIR8236,

C. difficile rgbR mutant (pJIR3042) . Blot (a) probed with the rgbR gene; (b) probed with the 3' end of the rgbR coding region; (c) probed with catP; (d) probed with repA from pMTL9301; (e) Schematic representation indicating the arrangement of the chromosomal rgbR region into which the single crossover occurred, expected sizes of restriction fragments are indicated, and the regions to which the probes are expected to hybridise are marked, expected size of the intact rgbR band in the wild type strain, JIR8094 is 13.6 kb.

Figure 7A shows the results of Gel mobility shift analysis of VirR or RgaR proteins with the C. perfringens pfoA promoter region. The purified C. perfringens VirR (CPVirR) .and C. difficile RgaR proteins were incubated with a DIG-labelled 183 bp PCR fragment that contained the

C. perfringens VirR boxes located upstream of P_PfOA (Cheung & Rood, 2000) . CI - Complex I, CII - Complex II. NS - non- specific unlabelled DNA competitor (contains mutated VirR boxes, CCA -→TAG, in the conserved region of both VirR boxes) , S - specific unlabelled DNA competitor (wild-type VirR boxes) . 5 pmol of each unlabelled competitor was used. Figure 7B demonstrates binding of C. difficile

RgaR to its C. difficile DNA targets. The purified C. difficile RgaR protein was incubated with DIG- labelled DNA fragments from the upstream regions of the following putative genes: rpoA (203 bp) (negative control), CD0587 (210 bp) , CD2098 (187 bp) and agrB (179 bp) . NS - unlabelled non-specific competitor {rpoA fragment) , S - unlabelled specific competitor (same as the labelled DNA target in that lane) ; 10 pmol of each unlabelled competitor was used. Figure 8 is a schematic representation showing the rgaR double crossover plasmid, pJIR3011. C. difficile genomic DNA from the chromosomal regions up- and downstream of the rgaR gene, and the first 257 bp of the rgaR gene were cloned on either side of the ermQ erythromycin resistance cassette. This region was subcloned into pJIR1456, which is unstable in C. difficile. This resulted in pJIR3011, a C. difficile rgaR recombination vector that carries two antibiotic resistance markers, ermQ to select for integration into the chromosome, and catP (on the plasmid backbone) , to select against single crossovers and independently replicating plasmids .

Figure 9 shows the results of Southern blotting analysis of C. difficile rgaR double crossover mutant strain. Genomic DNA was digested with AsplOO. (1) pJIR3011, rgaR double crossover vector; (λ) λ-Hindlll molecular weight markers; (2) JIR8094 wild type (WT) strain; (3) JIR8292 rgaR mutant. (A) probed with the rgaR gene; (B) probed with ermQ; (C) probed with catP. (D) Schematic representation indicating the expected AsplOO- digested fragment sizes in the C. difficile wild type (WT) and rgaR mutant (mutant) strains. This schematic indicates that a double crossover event between the suicide plasmid, pJIR3011 and rgaR would lead to the deletion of the last 454 bp of the rgaR gene, corresponding to the C-terminal DNA-binding domain of the encoded protein. The catP gene is present on the backbone of the pJIR3011 suicide vector (panel (C) and see Figure 8) , but is not present in the double crossover strain.

Figure 1OA shows the results of gel mobility shift analysis of VirR and RgaR proteins with the C. perfringens pfoA promoter region. The purified C. perfringens VirR and C. difficile RgaR (CD3255) proteins were incubated with a DIG-labelled 183 bp PCR fragment which contained the C. perfringens VirR boxes located upstream of P_Pf_OA- CI - Complex I, CII - Complex II. NS - non-specific unlabelled DNA competitor (contains mutated VirR boxes, CCA →TAG, in the conserved region of both VirR boxes) , S - specific unlabelled DNA competitor (wild-type VirR boxes) . 5 pmol of each unlabelled competitor was used.

Figure 1OB illustrates binding of C. difficile RgaR to its C. difficile DNA targets. The purified

C. difficile RgaR protein was incubated with DIG-labelled DNA fragments from the upstream regions of the following putative genes: rpoA (203 bp) (negative control), CD0587 (210 bp) , CD2098 (187 bp) and agrB (179 bp) . NS - unlabelled non-specific competitor {rpoA fragment) , S - unlabelled specific competitor (same as the labelled DNA target in that lane) ; 10 pmol of each unlabelled competitor was used.

Figure 11 is a schematic representation of the C. difficile rgaR double cross-over plasmid, pJIR3011.

C. difficile genomic DNA from the chromosomal regions up- and downstream of the C. difficile rgaR gene, and the first 257 bp of the rgaR gene were cloned either side of the ermQ erythromycin resistance cassette. This region was subcloned into pJIR1456, which is unstable in C. difficile. This resulted in pJIR3011, a C. difficile rgaR suicide vector which carries two antibiotic resistance markers, ermQ to select for integration into the chromosome, and catP (on the plasmid backbone) , to select against single crossovers and independently replicating plasmids.

Figure 12 shows the results of Southern blotting analysis of a C. difficile rgaR double crossover mutant strain. Genomic DNA was digested with Asp700.

(1) pJIR3011, rgaR double crossover vector; (λ) λ-Hindlll molecular weight markers;

(2) JIR8094 wild type (WT) strain; (3) JIR8292 rgaR mutant.

(A) probed with the rgaR gene;

(B) probed with ermQ;

(C) probed with catP.

(D) schematic indicating the expected Λsp700-digested fragment sizes in the C. difficile wild type (WT) and rgaR mutant (mutant) strains. This schematic indicates that a double crossover event between the suicide plasmid, pJIR3011 and rgaR would lead to the deletion of the last 454 bp of the rgaR gene, corresponding to the C-terminal DNA-binding domain of the encoded protein. The catP gene is present on the backbone of the pJIR3011 suicide vector (panel (C) and see Figure 11) , but should not be present in the double crossover strain.

Figure 13 shows the results of Southern hybridisation analysis of C. difficile tcdA and tcdB mutants. Genomic DNA from all strains was digested with Xbal . Blot A was probed with an internal fragment from tcdA, blot B was probed with an internal fragment from tcdB, and blot C was probed with the catP gene. D is a schematic representation of the pathogenicity locus

(PaLoc) , digested with Xbal, showing expected fragment sizes for the wild type (WT) . ΔtcdA and ΔtcdB strains for both the tcdA and tcdB genes of each strain are indicated. The sites where the suicide plasmids are expected to be integrated into the PaLoc are indicated; the blocks with the letters A-C respectively indicate where the tcdA, tcdB and catP probes are expected to hybridise. The wild-type band in panel A is expected to be 1.74 kb in size, the same as in the tcdB mutant strains, but in this experiment the wild-type band did not appear; however, previous experiments have shown that it was present at the expected location.

Figure 14 shows the results of Western blot analysis of culture supernatants from the toxigenic wild type C. difficile parent strain JIR8094 (WT) , the non- toxigenic strain, CD37, two tcdA mutants (ΔtcdA) and two tcdB mutants {ΔtcdB) , demonstrating that toxin A is produced by the wild-type strain and the tcdB mutants, but not by the non-toxigenic strain or the tcdA mutants.

DETAILED DESCRIPTION

Risks for colonisation by C. difficile include the length of the hospital stay, sharing a room with a patient who is suffering CDAD, chemotherapy, an age of 65 years or over, intestinal surgery, the severity of the underlying disease and antibiotic therapy. Virtually all antibiotics have been implicated in C. difficile disease, although penicillin, clindamycin, cephalosporins and now fluoroquinolones are the most common culprits (Cartmill et al., 1994, Loo et al., 2005, McFarland et al . , 1990, Spencer, 1998) . In addition, resistance to the two antimicrobials of choice for treatment of CDAD has been reported: metronidazole resistance has been documented on numerous occasions, and one report has identified vancomycin intermediate-resistant C. difficile isolates. However, currently the detection of resistance to these two antimicrobials is rare, and they remain effective. CDAD can also be induced by antibiotics to which the invading C. difficile strain is sensitive, indicating that the onset and progression of C. difficile disease involves complex interactions and includes a number of factors, many of which have not yet been satisfactorily explained. In the absence of an epidemic, the carriage rate in hospitals can vary from 4 to 21%, but when outbreaks occur, the rate of patient-to-patient cross-infection can be as high as 32%. Molecular typing methods have shown that many nosocomially-acquired infections are the result of a clonal isolate which is transmitted to new patients from a reservoir of asymptomatic and symptomatic patients, via medical staff, and from inanimate surfaces.

Once C. difficile disease has been resolved, as many as 50% of patients can relapse and suffer multiple illnesses. True relapses, where the same strain is again isolated from the patient, are in the minority, and reinfection with a different strain occurs in 45-60% of relapse cases . This indicates that patients are becoming infected or re- infected from environmental sources and highlights the importance of adhering to strict hygiene regimes in hospitals and aged care facilities. It is possible that a proportion of the "true relapse" cases are also a result of reinfection, but with the same strain. This scenario has been simulated using the hamster model, where animals which were infected with C. difficile only survived if they were removed from the contaminated environment (Larson et al . , 1978).

Approximately 3% of AAD cases progress to fulminant CDAD (Kelly & LaMont, 1998) , which has an associated mortality rate of more than 35% for pseudomembranous colitis and toxic megacolon cases. However, fluoroquinolone-resistant hypervirulent strains of C. difficile have recently emerged, and these have led to epidemics and unusually high mortality rates in Canada, the United States and the United Kingdom. These strains are associated with a radically increased risk of infection (even for people previously considered not to be at risk, and with morbidity, relapse and mortality. One study reported a ten-fold increase in the incidence of CDAD in elderly patients between 1991 and 2003. Similarly, the short-term mortality rate within the same population dramatically increased, with 13.8% of infected patients dying within 30 days of diagnosis in 2003, compared to 4.7% in 1991-1992. This alarming increase in the virulence and transmissibility of these strains is thought to be due to a range of factors in addition to fluoroquinolone resistance and much higher levels of toxin production: the increased level of prescription of antibiotics, the increasing size of the ageing population and a failure to maintain and improve hospital infrastructure and infection control regimes. This situation may have been exacerbated by the recent introduction of an ethanol hand rub for medical staff treating hospital patients . While this ethanol hand rub has decreased the incidence of methicillin-resistant Staphylococus aureus infections , it may not control the spread of CDAD, because the gel kills vegetative cells, but not the bacterial spores which contaminate hands. While the increased virulence of epidemic-causing isolates is largely blamed on a higher level of toxin production, other, as yet unidentified, virulence factors are likely to be involved, since it is proposed that epidemic strains may also have a superior ability to colonise, persist and sporulate.

¹ The economic burden of C. difficile associated diseases has previously been estimated to be in excess of $1.1 billion annually in the U.S.A. A recent report estimated that the total number of cases of CDAD in the U.K. in 2004 was 43,672, at a cost of approximately £240 million, mostly due to increased length of stay in hospital. However, the economic burden is likely to increase significantly in response to the recent and ongoing epidemics .

Current approaches to the treatment of C. difficile-associated disease depend on the severity of the infection. Antibiotic-associated diarrhoea is usually treated by the withdrawal of antibiotic therapy, which leads to repletion of normal flora and cessation of symptoms. For more severe C. difficile associated diarrhoea or mild colitis, treatment with vancomycin or metronidazole is indicated, while treating severe or fulminant colitis may necessitate surgical removal of part or all of the colon. Approaches to preventing relapse of C. difficile diseases include the use of probiotics to replenish the gastrointestinal normal flora and compete with C. difficile in the intestinal tract with varying levels of success . Some success has also been demonstrated in providing passive immunity to hamsters through administration of antibodies to C. difficile surface layer proteins, and it was found that the administration of these antibodies increased phagocytosis of C. difficile. Other methods have involved oral administration of non-digestible oligosaccharides and toxin A-binding resins, both of which have been proposed to assist in resolution of recurrent infection.

Immunity to C. difficile infection is mediated both by innate and active mechanisms. The first barrier to infection by C. difficile is the normal flora which inhabit the human gastrointestinal tract; it is proposed that disruption of this protective barrier is required for C. difficile to successfully colonise the host, because it allows C. difficile access to attachment sites and increased levels of nutrients . Other innate mechanisms which protect against C. difficile infection may include an effective gastric acid barrier. The mucus layer of the intestinal tract, intestinal motility and movement of the intestinal contents may prevent attachment of C. difficile and the establishment of an infection. The diet of the host is also thought to play a role, as well as stress and overall health. Whether or not a patient is capable of resolving the infection appears to correlate with the level of serum IgG and IgA antibodies to toxin A; patients who do not relapse tend to have higher antibody levels than those who do.

Several studies have found that using a toxoid vaccine to stimulate the production of antibodies which neutralise the cytotoxic activity of toxin A provides protective immunity to hamsters and rabbits . Recent work has focussed on the development of a C. difficile toxoid vaccine for humans. Anecdotal evidence suggests that the stimulation of humoral immunity through the administration of a toxin A toxoid vaccine, and the provision of serum immunoglobulin as a form of passive vaccination can be successful in the resolution of recurrent C. difficile infection, and a phase I human trial of a C. difficile toxoid vaccine is in progress.

The toxin A toxoid vaccine currently undergoing clinical trial (Giannasca and Warny, 2004) has both advantages and disadvantages. If the toxoid is effective, it may help to reduce the severity of illness of patients infected with C. difficile, thereby reducing the mortality rate and the length of stay in hospital, and lowering the economic burden of this disease, especially if it prevents relapses. However, there are several points to consider: (i) The toxoid vaccine has only been tested in healthy subjects and in a handful of chronically ill patients. It is possible that the section of the community which is most vulnerable, ie. the elderly and hospitalised patients, may not be able to mount a sufficient immune response to the toxoid for it to be protective, since they tend to have generally reduced immune competence in any case. Furthermore there is no information available on the duration of protective immunity; the vaccine might only protect for a limited period of time. (ii) The toxoid vaccine will only protect against toxin A. Although toxin A is proposed to be the major virulence factor in CDAD, it is certainly not the only virulence factor and some clinical C. difficile isolates from CDAD patients are toxin A-negative. This implies, that other virulence factors can be responsible for morbidity, and in these cases a toxin A toxoid vaccine will not help. (iii) The toxin A toxoid vaccine cannot protect a patient against colonisation by C. difficile, and therefore it will provide no protection against transmission to other patients. In an epidemic situation, a very large proportion of the hospital population could become colonised. It is unlikely that the toxoid vaccine could be provided to every single person in a hospital, so the organism will persist in this population and in the hospital environment. Therefore, it appears that a combination vaccine approach would offer greater benefits. For example, administration of the toxoid vaccine would help chronic CDAD patients who suffer multiple relapses, but the additional oral administration of an attenuated, non- toxigenic C. difficile strain would prevent uninfected patients from being colonised, and would allow all those vaccinated with this strain to develop an immune response to the C. difficile bacterial cells. This should allow the immune system to eliminate C. difficile from the body, and prevent subsequent colonisation and transmission. Such an approach could stop an epidemic very quickly, because infected patients would recover, and uninfected patients would not become colonised.

Definitions

In the description of the invention and in the claims which follow, except where the context requires otherwise due to express language or necessary implication, the word "comprise" or variations such as "comprises" or "comprising" is used in an inclusive sense, ie. to specify the presence of the stated features but not to preclude the presence or addition of further features in various embodiments of the invention.

As used herein, the singular forms "a", "an", and "the" include the corresponding plural reference unless the context clearly dictates otherwise. Thus, for example, a reference to "an enzyme" includes a plurality of such enzymes, and a reference to "an amino acid" is a reference to one or more amino acids .

Where a range of values is expressed, it will be clearly understood that this range encompasses the upper and lower limits of the range, and all values in between these limits.

The term "homologous recombination" refers to the exchange of DNA sequences between two DNA molecules which have a region of sequence homology. In particular, it refers to homologous recombination between a bacterial chromosome and an extrachromosomal element, such as a plasmid, which carries a region with complete or nearly complete sequence complementarity.

The term "insertional inactivation" means interruption of the coding region of a gene by the insertion of exogenous DNA, leading to the loss of gene function. This is widely used in gene technology to permit easy selection of recombinants following transformation.

The term "suicide vector" means a cloning vector used to clone DNA sequences and transfer them to recipient host cells which undergoes homologous recombination with host cell DNA and loses vital function as a result of this process, so that the vector cannot survive in the host cell, and is eliminated. The term "virulence" or "virulent" means the ability of a bacterium to cause disease or the ability of a factor produced by a bacterium to play a role in the process of pathogenesis. A virulent C. difficile cell will usually carry a pathogenicity locus which encodes toxin A and/or toxin B. Virulent strains usually originate from clinical isolates.

The term "avirulent" means a bacterium that is no longer able to cause disease. An avirulent C. difficile cell will usually have a non- functional pathogenicity locus, ie. one in which toxins A and B are either not expressed or in which the toxins are expressed but lack toxicity. The term "heterologous" means a nucleic acid or protein which is not native to C. difficile. Conversely the term "homologous" means a nucleic acid or protein which is native to C. difficile. The term "transformed" refers to a cell into which heterologous DNA has been introduced.

The term "regulator response gene" means a gene encoding a regulatory protein which is capable of acting as a phosphoacceptor and has conserved residues in its N- terminal domain such that it belongs to the response regulator protein family.

The regulator response genes RR17 and RR55 which were disclosed in Australian patent application No. 2005903063, the priority document for this application, have been renamed as rgaR and rgbR respectively.

"Conjugation" or "mating" means the unidirectional transfer of plasmid DNA from a donor bacterium to an acceptor bacterium after random collision between the bacteria and establishment of a conjugation bridge between cells of opposite mating types. A "conjugative plasmid" is one which encodes all the functions necessary for its own intercellular transmission by conjugation.

A "condition associated with C. difficile" is one which is clinically associated with or caused by C. difficile, including but not limited to diarrhoea, antibiotic-associated diarrhoea, colitis, enterocolitis, antibiotic-associated colitis, fulminant colitis, pseudomembranous colitis and toxic megacolon. The term "probiotic" means an agent which stimulates the growth of beneficial types of bacteria within the digestive system of humans or animals .

"Isolated" nucleic acid is nucleic acid which is identified and separated from, or otherwise substantially free from, contaminant nucleic acid encoding other polypeptides. The isolated nucleic acid can be incorporated into a plasmid or expression vector, or can be labeled for diagnostic and probe purposes, using a label as described further.

The term "derived from" means that a specified integer may be obtained from a particular source, albeit not necessarily directly from that source.

"Stringent conditions" for hybridization or annealing of nucleic acid molecules are those that

(1) employ low ionic strength and high temperature for washing, for example 0.015 M NaCl/0.0015 M sodium citrate/0.1% sodium dodecyl sulfate (SDS) at 50⁰C, or

(2) employ during hybridization a denaturing agent such as formamide, for example 50% (vol/vol) formamide with 0.1% bovine serum albumin/O.1% Ficoll/O.1% polyvinylpyrrolidone/50 mM sodium phosphate buffer at pH 6.5 with 750 mM NaCl, 75 mM sodium citrate at 42°C.

Another example is use of 50% formamide, 5 x SSC (0.75 M NaCl, 0.075 M sodium citrate), 50 mM sodium phosphate (pH 6.8), 0.1% sodium pyrophosphate, 5 x Denhardt ' s solution, sonicated salmon sperm DNA (50 μg/mL) , 0.1% SDS, and 10% dextran sulfate at 42°C, with washes at 42°C in 0.2 x SSC and 0.1% SDS.

An "adjuvant" is a substance which augments, stimulates, activates, potentiates, or modulates the immune response at either the cellular or humoral level. An adjuvant may be added to a vaccine, or may be administered before administering an antigen, in order to improve the immune response, so that less vaccine is needed to produce the immune response. Adjuvants include alum, ISCOMs which comprise saponins such as Quil A, liposomes, and agents such as Freund's adjuvant, Bacillus Calmette Guerin,

Corynebacterium parvum or mycobacterial peptides which contain bacterial antigens . Only some of these are currently approved for human or veterinary use,- others are in clinical trial. Some adjuvants are endogenous, such as histamine, interferon, transfer factor, tuftsin and interleukin-1. Their mode of action is either non-specific, resulting in increased immune responsiveness to a wide variety of antigens, or antigen-specific, ie. affecting a restricted type of immune response to a narrow group of antigens .

The term "antibody" includes all classes and subclasses of intact immunoglobulins, and also encompasses antibody fragments. The term "antibody" specifically encompasses monoclonal antibodies, including antibody fragment clones. "Antibody fragments" comprise a portion of an intact antibody which contains the antigen binding or variable region of the intact antibody. Examples of antibody fragments include Fab, Fab', F(ab')₂, and Fv fragments; diabodies; single-chain antibody molecules, including single-chain Fv (scFv) molecules; and multispecific antibodies formed from antibody fragments. The term "subject" generally means a mammal suffering from, or at risk of, a condition associated with C. difficile. The mammal may be a human, or may be a domestic or companion animal. While it is particularly contemplated that the compounds of the invention are suitable for use in medical treatment of humans, they are also applicable to veterinary treatment, including treatment of companion animals such as dogs and cats, and domestic animals such as horses, cattle and sheep, or zoo animals such as felids, canids, bovids, and ungulates. "Treatment" refers to both therapeutic treatment and prophylactic or preventative measures . Those in need of treatment include those already with the disorder, as well as those at risk of or prone to having the disorder or those in which the disorder is to be prevented. For the purposes of this specification, beneficial or desired clinical results include, but are not limited to, alleviation of symptoms, diminishment of extent of disease, stabilized (ie. not worsening) state of disease, delay or slowing of disease progression, amelioration or palliation of the disease state, and partial or total remission,

^' whether detectable or undetectable. "Treatment" can also mean prolonging survival, as compared to expected survival if not receiving treatment.

"Pharmaceutically acceptable" carriers, excipients, or stabilizers are ones which are non-toxic to the cell or mammal being exposed thereto at the dosages and concentrations employed. Often the physiologically acceptable carrier is an aqueous pH buffered solution. Examples of physiologically acceptable carriers include buffers such as phosphate, citrate, and other organic acids; antioxidants including ascorbic acid; low molecular weight (less than about 10 residues) polypeptides; proteins, such as serum albumin, gelatin, or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagine, arginine or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrins; chelating agents such as EDTA; sugar alcohols such as mannitol or sorbitol; salt-forming counterions such as sodium; and/or nonionic surfactants such as Tween, polyethylene glycol (PEG) , and Pluronics™.

Abbreviations used herein are as follows: AAD antibiotic-associated diarrhoea BHI brain-heart infusion Cd Clostridium difficile CDAD C. difficile-associated disease CDS Chromosomal coding sequence CPE cytopathic effect DIG digoxigenin ORF open reading frame PCR polymerase chain reaction

QRT-PCT quantitative real time reverse transcription PCR RR response regulator RT reverse transcriptase RT-PCR reverse transcription-PCR WT wild-type

It is to be clearly understood that this invention is not limited to the particular materials and methods described herein, as these may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and it is not intended to limit the scope of the present invention, which will be limited only by the appended claims .

The present invention may be performed without any undue need for experimentation. Unless otherwise indicated, the present invention employs conventional molecular biological, recombinant DNA, microbiological, and immunological techniques within the capacity of those skilled in the art. Such techniques are well known to the skilled worker, and are explained fully in the literature. See for example Sambrook, Fritsch et al . 1989; Shuler and Kargi 1992; Graves, Martin et al . 1994; Lundblad 1995; Goding 1996; Sambrook and Russell 2001; Madigan, Martinko et al. 2003; DNA Cloning: A Practical Approach, (D. N. Glover, ed. , 1985), IRL Press, Oxford; Oligonucleotide Synthesis: A Practical Approach (M. J. Gait, ed., 1984)

IRL Press, Oxford; Nucleic Acid Hybridization: A Practical Approach (B. D. Hames & S. J. Higgins, eds . , 1985) IRL Press, Oxford; Perbal, B., A Practical Guide to Molecular Cloning (1984); Handbook of Experimental Immunology, VoIs. I-IV (D. M. Weir and C. C. Blackwell, eds., 1986, Blackwell Scientific Publications) .

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although any materials and methods similar or equivalent to those described herein can be used to practice or test the present invention, the preferred materials and methods are described.

This specification includes amino acid sequence information prepared using Patentln Version 3.3, presented herein after the Abstract. Each sequence is identified in the sequence listing by the numerical indicator <210> followed by the sequence identifier (e.g. <210>l, <210>2, etc) . The length of each sequence and the source organism are respectively indicated by information provided in the numerical indicator fields <211> and <213>, respectively. Sequences referred to in the specification are indicated by the term "SEQ ID NO: ", followed by the sequence identifier, eg. SEQ ID NO: 1 refers to the sequence designated as <400>l.

Oligonucleotides for use as hybridization probes or primers may be prepared by any suitable method, such as by purification of a naturally-occurring DNA or RNA, or by in vitro synthesis. For example, oligonucleotides are readily synthesized using various techniques in organic chemistry, such as those described by Narang, et al . , Meth. Enzymol. 68:90-98 (1979); Brown, et al . , Meth. Enzymol . 68:109-151 (1979); Caruther, et al . , Meth. Enzymol. 154:287-313 (1985). The general approach to selecting a suitable hybridization probe or primer is well known,- see Keller, et al., DNA Probes, pp.11-18 (Stockton Press, 1989) . Typically, the hybridization probe or primer will contain 10-25 or more nucleotides, and will include at least 5 nucleotides on either side of the target sequence so as to ensure that the oligonucleotide will hybridize preferentially to the single-stranded DNA template molecule.

Since no effective vaccine for C. difficile is currently available, the methods of the invention provide the tools needed to make careful study of this organism using conventional genetic principles, a process which was previously impossible. This may yield new information regarding the pathogenesis of this organism. Additionally, since we are now able to achieve genetic manipulation of C. difficile, we may be able to construct attenuated strains for vaccine trials. The methods of the invention can be used in the construction of isogenic C. difficile strains, potentially using any parent strain into which DNA can be introduced, particularly in the alteration of genes encoding pathogenic factors such as C. difficile toxins A and B.

I Treatment of infected patients (1) Deletion of toxin A and B in a pathogenic strain

Introduction of avirulent strains to prevent virulent C. difficile infection in hamsters has been attempted with some success, but the ability of avirulent strains to compete effectively in the colon may be affected by strain-specific factors; consequently strains which are inherently avirulent may not compete well. A C. difficile strain which has been genetically modified to delete a gene or genes essential for virulence may compete successfully with the established virulent strain. In particular it may be possible to pretreat at risk patients with the avirulent derivative, thereby preventing virulent isolates from establishing an infection and causing disease. It may also be possible to construct a C. difficile strain with modified tcdA and tcdB genes which produce toxin molecules which have only a binding domain, not a catalytic domain, but which are not internalised, thus blocking access of active toxin to binding sites on target host cells . This would reduce the pathogenic effects of the strain which is responsible for initial infection. (2) Modification of host immune response

Studies of the pathology of C. difficile- associated diseases indicate that the host immune system plays a role in damage to the colon, as the result of an inflammatory response. It may be possible to introduce a C. difficile strain which has been genetically modified to express host immune-modulating factors such as cytokines, or other signalling molecules, which would modify the host immune response in order to avoid the response which contributes to subsequent damage to the colon.

(3) Modification of the bacterial response

C. difficile toxins A and B are the primary virulence factors in infections caused by this organism. Production of these toxins is tightly controlled by bacterial regulatory networks, feedback mechanisms, environmental and bacterial growth phase factors . Quorum sensing, ie. signalling between individual bacterial cells, may be involved in the regulation of toxin production. Modification of quorum sensing genes may lead to a strain which is able to reduce the amount of toxin produced by the resident, toxin-producing strain, via the release of signalling molecules which tell the cells to ^Λturn off toxin production.

II Active immunisation of patients in ^λat risk' categories:

C. difficile is known to persist for long periods of time in an infected patient's intestinal tract, and such patients are at high risk of possible re-infection or relapse. Active immunisation with a genetically modified strain which has truncations in the tcdA and tcdB genes may lead to production of toxins which are immunogenic but not toxic. The use of C. difficile to deliver such a vaccine would mean that the attenuated toxin is introduced into the target environment, expressed and released in the correct manner at the correct time, and could be highly effective in stimulating secretory IgA antibodies, which would be necessary for effective neutralization of C. difficile toxins secreted into the intestine. In addition there is some evidence to, suggest that C. difficile surface proteins elicit an immune response in human hosts. Therefore introduction of a genetically modified strain of C. difficile could also induce antibody production against potential adhesins, and may be effective in blocking the establishment of infection by preventing invading C. difficile cells from binding to attachment sites in the colon. The hamster model of CDAD (Sambol et al, 2001) may be used to confirm that mutant strains according to the invention in which the activity of toxin genes has been disrupted are in fact avirulent, and that prior infection of host animals with such an avirulent strain protects against such an infection with a toxigenic strain.

C. difficile is an intestinal organism which normally colonizes the gut early in life. Even toxigenic strains are unable to cause CDAD in newborns and infants up to 2-4 years of age. We therefore expect that recombinant C. difficile according to the invention which produces a desired antigen is suitable for oral vaccination at any convenient time after birth. Similarly recombinant C. difficile according to the invention may be used to deliver a desired protein to the large intestine.

Oral administration of the transformed C. difficile to a mammal, such as a human, will lead to intestinal colonization, and to production and presentation of the desired polypeptide, particularly in the large intestine, which is the natural site of colonization of C. difficile. The bowel wall is the site of Peyer's patches, which form part of the immune system, and mount immune responses of various types. Large bowel colonization by a C. difficile vaccine or peptide producer strain thus enables a much longer immune stimulus than conventional administration by parenteral injection.

It will be appreciated that the use of the C. difficile of this invention transformed with DNA encoding different heterologous peptides allows the highly efficient production and export of these polypeptides in hypoxic tissues after intravenous administration, or into the gut, particularly the colon, of the orally colonized individual for a variety of prophylactic or therapeutic uses.

In further embodiments of the invention the recombinant gene expression cassette is used to produce proteins in the gut, for example

(i) peptides and proteins, including enzymes, for therapy and prophylaxis of various diseases, e.g. peptides having specific antimicrobial activity, cytokines, and β-lactamases to prevent diarrhea due to antibiotic therapy (ii) (ii) single, fusion or multiple polypeptide antigens of microbial, animal- or mammalian origin for neonatal immune balancing, vaccination against infections, desensitisation against allergy, metabolic or autoimmune disease, cancer, infertility, and drug addiction; (iii) carrier molecules (adjuvants) either separately 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. The amount of the desired protein (s) presented and/or secreted by the transformed strain may be modulated in the body by using

(i) promoters with different strength, (ii) an inducible promoter or regulator which responds to external stimuli, e.g. to a specific carbohydrate normally present in the gut or administered together with the bacterium,

(iii) different dosage regimens, e.g. number of bacteria per dose and doses per time period, or (iv) methods which 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, to adhere to mucosal cells, and to avoid expulsion by local immune response mechanisms.

The transformed C. difficile cells are conveniently produced by fermentation under conditions which are conventionally used for this organism, and purified and recovered by conventional methods, for example by washing and freeze-drying. They may be formulated together with conventional excipients, 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 cells are administered orally or intranasally, as an aqueous, reconstituted suspension of the lyophilized 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. General methods and pharmaceutical carriers for preparation of pharmaceutical compositions are well known in the art, as set out in textbooks such as Remington's Pharmaceutical Sciences, 19th Edition, Mack Publishing Company, Easton, Pennsylvania, USA.

The compounds and compositions of the invention may be administered by any suitable route, and the person skilled in the art will readily be able to determine the most suitable route and dose for the condition to be treated. Dosage will be at the discretion of the attendant physician or veterinarian, and will depend on the nature and state of the condition to be treated, the age and general state of health of the subject to be treated, the route of administration, and any previous treatment which may have been administered.

The carrier or diluent, and other excipients, will depend on the route of administration, and again the person skilled in the art will readily be able to determine the most suitable formulation for each particular case .

The invention will now be described in detail by way of reference only to the following non-limiting examples and drawings . Materials and methods

Plasmids, bacterial strains and growth media

Plasmids and strains used in this study are listed in Table 1.

TABLE 1 Bacterial strains and plasmids

Escherichia coli strains were cultured in 2 X YT (yeast extract, tryptone, and sodium chloride) agar or broth media, or in SOC broth (Sambrook, Fritsch et al . 1989) at 37⁰C, during incubation, broths were aerated using an orbital shaker; where appropriate cultured strains were supplemented with X-gal (Amresco, Solon, Ohio) , ampicillin (100 μg/ml) , chloramphenicol (30 μg/ml) , erythromycin (150 μg/ml) or streptomycin. C. difficile strains were cultured at 37⁰C in an atmosphere of 10% (vol/vol) H₂ and 10% (vol/vol) CO₂ in N₂, in BHIS medium (brain-heart infusion (Oxoid) (Smith, Markowitz et al . 1981), 37 g/L, yeast extract (Oxoid), 5 g/L), supplemented with 2.5 g/L glucose (Astral Scientific) , 1 g/L L-cysteine (Sigma) , and (for agar media) 0.09 g/L FeSO₄, or TPY medium (2% tryptone (Oxoid), 0.5% peptone (Oxoid), 0.5% yeast extract). C. difficile strains were supplemented with thiamphenicol (10 μg/ml) , lincomycin (50 μg/ml) , or cefoxitin (25 μg/ml) when necessary. All antibiotics were obtained from Sigma, except for erythromycin, which was from Amresco.

Molecular techniques

Chemically competent and , electrocompetent E. coli cells were prepared as described previously (Smith et al, 1990; (Inoue, Nojima et al . 1990); Sambrook et al, 2001). Plasmid DNA was isolated from E. coli strains grown overnight in 5 ml 2YT broth with appropriate antibiotic selection, using Qiaprep spin miniprep columns, according to the manufacturer's instructions. PCR amplification was carried out using Taq DNA polymerase (Roche) or Pfu DNA polymerase (Promega), 0.5 μM concentration_. of each oligonucleotide primer was used. Denaturation (94⁰C for 1 min) , annealing (50⁰C - 60⁰C for 2 min) , and extension (72°C for 3-5 min) steps were carried out for 35 cycles. PCR products were purified on Quiaprep PCR purification columns according to the manufacturer's instructions. When DNA sequencing was required, this was achieved using a PRISM Ready Reaction DyeDeoxy Terminator cycle sequencing kit (Applied Biosystems) , according to the manufacturer's instructions . Oligonucleotide primers used during this study are listed in Table 2.

TABLE 2 Oligonucleotide primers

+ represents the sense strand primer; - represents the antisense strand primer

Rescue of recombinant plasmids from C. difficile Plasmid DNA was extracted from C. difficile using

Qiaprep spin miniprep columns as follows: strains were grown overnight in 20 ml BHI broths, with or without antibiotic selection, 10 ml culture samples were centrifuged, and the cell pellets resuspended in 250 μl of buffer Pl supplemented with 10 mg/ml lysozyme (Astral

Scientific) . The suspension was then incubated at 37°C for 15 mins, then 500 μl of buffer P2 was added, followed by 700 μl buffer N3 ; this solution was incubated for five minutes at room temperature, then centrifuged at maximum speed for 10 minutes. The supernatant was then applied to the miniprep column, and centrifuged for 1 minute at maximum speed to allow the plasmid DNA to bind to the column. The column was then washed with 500 μl buffer PB, and the purification proceeded as per the manufacturer's instructions. Eluted C. difficile plasmid DNA was then concentrated from 48 μl to approximately 15 μl in a Speedivac. Two to five μl of concentrated plasmid DNA was then electroporated into electrocompetent E. coli DH12S cells, and the plasmids were extracted, purified and subsequently subjected to restriction analysis to confirm they still maintained the expected restriction profile.

Isolation of genomic DNA from C. difficile

C. difficile genomic DNA was prepared using a method based on that of (Pospiech and Neumann 1995) . C. difficile strains were inoculated into pre-boiled BHIS broths with or without selection, and incubated overnight, then 5-7 ml culture samples were centrifuged, and the cell pellets resuspended in 495 μl SET buffer (75mM NaCl, 25mM EDTA (pH8.0), 2OmM Tris-HCL (pH7.5) in dH₂0) . 50 μl of 10 mg/ml lysozyme/ dH₂0 solution was added, and a small amount of lysostaphin (Sigma) was added to this suspension. The samples were incubated at 37⁰C for one hour, then 50 μl of 10% sodium dodecylsulphate solution and 5 μl 25 mg/ml proteinase K (Amresco) was added, and the samples were incubated at 55°C for 30 minutes - 2 hours until cell lysis occurred. 200 μl of 5M NaCl and 700μl chloroform: isoamyl alcohol was added, and the solution was incubated at room temperature for 30 minutes with frequent inversions. Following this, the samples were centrifuged at maximum speed in a benchtop centrifuge for 30 minutes at room temperature, and then the aqueous phase was transferred to a fresh microcentrifuge tube. 700 μl isopropanol was added to precipitate the DNA, and the samples were centrifuged again at maximum speed for 10 minutes. The supernatant was decanted and the DNA pellet washed in 500 μl cold 70% ethanol, the samples were centrifuged for five minutes, then the ethanol solution was pipetted off and the DNA pellet dried in a Speedivac for seven minutes on medium heat setting. 100 μl dH₂0 was added to each tube to solublize the DNA, and the samples incubated for 10 minutes on ice. Two μl of RNase A (100 μg/ml; Sigma) was then added to each sample, and the preparation was incubated at 37°C for 10 minutes. DNA was quantitated using a NanoDrop Spectrophotometer ND-1000 (NanoDrop Technologies Inc, Rockland, De, USA) , and stored at -2O⁰C.

Toxin production and purification

Strains JIRIlOl (CD37) , JIR8094 (630E), JIR8253, JIR8263, JIR8276 and JIR8278 were grown in 90 ml TY broth (Mani, Lyras et al 2002) for 48-72 hours at 37°C under anaerobic conditions. The cells were then removed by centrifugation (8,000 x g, 10 minutes, 4⁰C) and the supernatant retained. After ammonium sulfate precipitation at 70% saturation followed by centrifugation, the pellet was resuspended in 2 ml of distilled water and dialysed against phosphate-buffered saline, pH 7.2.

Toxin A Western blots The method followed for Western immunoblotting was as described elsewhere (Tang-Feldtnan, Ackermann et al . 2002) . The toxins were analysed by SDS-PAGE using a 4-15% Tris-HCl Ready Gel (Bio Rad) and a Bio Rad Protean III mini gel system at a constant 200 V for 1 hour. Molecular markers (Precision Prestained Protein Standards, Bio Rad) were used as size standards. Proteins were transferred to nitrocellulose membranes (Protran, Whatman) at a constant 100 V for 60 minutes. ■ The membranes were blocked overnight in a solution consisting of 5% skim milk powder in TBS-Tween 20 solution. Blots were incubated for 1 hour at room temperature with 15 μl of a 1:1000 dilution of affinity- purified polyclonal goat antibody to Toxin A (David Lyerly, Techlabs, USA) . The blots were washed three times for 15 minutes in TBS-Tween 20 solution and incubated with a 1:500 dilution of the horseradish peroxidase-conjugated secondary antibody (anti-goat Ig, Chemicon) for 1 hour. The blot was washed in TBS-Tween 20 buffer for 20 minutes, after which it was washed four times in the same buffer for 5 minutes and incubated with the chemiluminescent detection reagents (Kit name, Rennaisance Western Blot chemiluminescent Reagent, Perkin Elmer) in order to detect bound antibodies.

Toxin B Cytotoxicity assays

Vero cells were cultured in minimal essential medium (alpha medium: GIBCO™, Invitrogen, CA, USA) containing 10% heat inactivated foetal calf serum, 100 μg/ml penicillin and 100 μg/ml streptomycin in culture flasks at 37⁰C in 5% CO₂. The cells were grown to a confluent monolayer and subcultured by incubation in 1-2 ml of 0.1% trypsin in 1 mM EDTA solution. The cells were counted and resuspended in fresh medium at a concentration of 2 x 10⁵ cells/ml. One ml of the cell suspension was seeded into each well of a 24 -well plate. The plates were incubated for 20 - 24 hrs and the culture medium was removed prior to inoculation of the culture supernatants . Serial dilutions of the. C. difficile culture supernatants were made in sterile PBS, and 100 μl aliquots of these supernatants were inoculated into the wells .

Construction of recombinant plasmid vectors.

Standard methods for the digestion, modification, - ligation, and analysis of plasmid and genomic DNA and PCR products were used (Sambrook, Fritsch et al . 1989).

Southern blotting analysis of recombinant C. difficile strains

C. difficile genomic DNA was digested with the appropriate restriction endonucleases, subjected to agarose gel electrophoresis, then transferred to a Nylon H⁺ hybond membrane (Amersham; Sambrook et al . , 1989) . Southern, hybridisation analysis was carried out using standard methods (Lyras and Rood, 2000) . DNA was labelled using PCR random primed labelling (Roche) according to the manufacturer's instructions.

RNA extraction from C. difficile

For RNA extraction, C. difficile was grown overnight in 20 ml of TPY broth (2% tryptone, 0.5% peptone and 0.5% yeast extract) and approximately 4 ml transferred to 90 ml of the same medium. The inoculum was adjusted so that all cultures started with a similar turbidity. The cultures were then grown to early stationary phase (approx. 7.5 h) and harvested by centrifugation. C. difficile RNA was extracted as previously described (Lyras and Rood, 2000) except that cell pellets from 40 ml culture samples were harvested at early stationary phase, and, following the first incubation, 20 μl of PCR grade proteinase K solution (Roche) was added to the buffer. The suspension was incubated at 55⁰C for 10 min, the cells were then centrifuged at 1,900 g and the supernatant removed.

Following the first Trizol (Invitrogen) extraction, the RNA was resuspended in diethylpyrocarbonate (DEPC) -treated water and DNaseI digestion was carried out in a solution of 20 units of RNasin (Promega) , 4 units of TURBO DNase (Ambion) and TURBO DNase buffer in a final volume of 100 μl, at 37⁰C for 1 h. After two more cycles of Trizol extraction and DNaseI digestion, the RNA was purified using a QIAGEN RNeasy RNA column purification kit, as per the manufacturer's instructions. RNA samples were eluted from the column in two volumes of 30 μl DEPC-treated water, quantitated using a NanoDrop spectrophotometer and stored at -70⁰C.

Quantitative RT-PCR analysis of differentially expressed genes

Reverse transcriptase (RT) reactions were performed as previously described (Kennan et al . , 2001), using primers designed with Primer Express (Applied Biosystems) (Table 2) . Prior to real time PCR analysis, control rpoA RT reactions were diluted twofold; all other RT reactions were diluted tenfold. Reactions were performed in a final volume of 25 μl with SYBR Green PCR master mix (Applied :

Biosystems), 2 μl cDNA, and 120 nM primers on an ABI PRISM 7700 sequence detector. Triplicate reactions were undertaken in multiple experiments using different RNA preparations, with at least three biological replicates. The data were normalized to C. difficile rpoA RNA levels. Reactions were determined to be the result of single products by dissociation curve analysis.

Construction of rgaR and rgbR overexpression vectors The response regulator genes rgaR and rgbR were identified during our analysis of C. difficile strain 630.

These genes are identified in the priority document for this application as RR17 and RR55 respectively. The genome sequence of C. difficile strain 630 was scanned for response regulator genes (RR) , and each gene was numbered in the order in which it was identified. The strain 630 sequence is available online at http: //www. Sanger .ac.uk/Proj ects/C__difficile/ . The nucleotide sequences of the rgaR and rgbR open reading frames are set out below.

rgaR open reading frame :

ATGTTCAGGGTAGTTATATGTGATGACGAAAAAATACAAAGAAGCATACTAAAAGAGTT TACAGAAAAACTTTTGTCTGAAAGATGTCTAAATTATCAAATACTGGAATTTTCCTGTG GAGAAGATTTAATATCCAAATACCCAGAAAAAATTGATATTATATTTTTAGATATCCAA ATGAAAGATATCAATGGAATTGAGGTGGCTAAAAAAATAAGAAAGTTTGATGAAAAAGT AGAAATTATATTTACAACGGCTTTTTCAGAGTATGCCCCCCAAGGTTATGAGGTTCGTG CTTATAGATATTTAGTAAAGCCTATAGAGTATAGTAACTTTGCAATGGGTGTAAATCTT TGTATTGATAACCTTCAAAGAAAGAAGGAGCGTTATATAGTTTTAAATAGTAAAAAAGG ATTTAATAGAATACTGATAGATTCTATTTTATATGTTGAAACTGTAAAGAGAGATTTAG TCGTACATACAATAGGTAGAGATTATAAAGCTAATATGAGCATGAAACAAGCAGAAAAA TTACTGTGTGAAAATGGCTTTTTTAGATGTCATACAAGTTTTATTGTTAATTTAAGAAG AATAGAAGATCTTAGGGATAATATGATTACTATTAATGAACAATTTATACCTATAAGTA AGTATAGACTTAAAGATTTAAAAGTAGCATTAACAAGCTTGTTATGTGATATAATGTAT TAA (SEQ ID NO: 57) rgbR open reading frame :

ATGATTAAGATAGCAGTTTGTGAAGATGAAAAAGAAACACAACTTTTAATAGAAGATTA CCTTGAGAATATATTAAAAGATATAAGTATAGAATATGAGATACAAAAGTATATATCTG GAGAAGAGTTACTGGAAAGTAATTTAAAAGATATAGATATATTACTTCTTGATATAAAA ATGGAAAAACTAAATGGAATGGATACTGCAAGAAAGATTAGAGAAGTAGACAATGAAAT

GGAAATAATATTTGTAACTTCATTAATAGATTATGTGCAAGAAGGTTATGAGGTAAGAG CTTATAGGTATTTACTAAAGCCAATAGAGTTGGAAGAACTAAAAAAACATGTGCTAACT TGTATCAAAGATATTGAAATAAATAAGGAAAGTCATATTACAATAAAAAATAAGTCTAA TACATATAAGATTTATTTAAATGAAATAAAATATATAGAAGTTCAAAAAAAAGATATGC TAATACACACAATAAACAAGAATTTTGATATAAAATATAGTTTAGGCAAAATAGAAAAA

GAATTAAATCCATATAAATTTATAAGATGCCATAAAAGTTTTATAGTAAATTTAAGATA TGTTGAAAATATAAAGCCTAATACTGCAATACTGGAAAGTGGAGAAGAGGTACCTATTA GTAGATATAGATATAAAGAAGTTAAGGAGAAATTCTTAAAATTTCTTGGTGATACAATA TGCTAA (SEQ ID NO: 58)

The rgaR and rgbR genomic regions were independently amplified from C difficile strain 630 genomic DNA by PCR, using Pfu polymerase and primer pairs JRP1476/1477 (1.45 kb) and JRP1482/1483 (1.54 kb) respectively; the products were purified and digested with

Spiϊl and EcoRI restriction endonucleases, then cloned into the same restriction sites in plasmid pJIR1457. The inserts were then sequenced to confirm that they were correct. This resulted in plasmid pJIR2512 (pJlR1457+ rgaR) and pJIR2515 (pJIR1457+ rgbR) . The rgaR plasmid insert was then subcloned into pJIR1456 as follows: pJIR1456 was digested with Sad and pJIR2512 was digested with EcoRI, and the cohesive ends of each linearised vector were converted using T4 polymerase (Proraega) . Both plasmids were then digested with Sphl, and the rgaR insert from pJIR2512 was directionally cloned into pJIR1456, yielding pJIR2634. The pJIR2515 rgbR insert was subcloned into pJIR1456 as follows: pJIR2515 was digested with Sphl/EcόRI, and the 1.54 Kb fragment was subcloned into the Sphl/Sacl sites of pJIR1456, yielding pJIR2633.

Construction of rgaR and rgbR insertional inactivation vectors The rgaR and rgbR insertional inactivation vectors were based on a previously constructed vector used in the inventors' laboratory. This vector was in turn based on pJIR750 (Bannam and Rood 1993) . The RP4oriT sequence from pJIR1456 was amplified by PCR, using Pfu polymerase and JRP1942/1943 (243 bp) . The resultant blunt- ended product was cloned into the Xmnl site of pJIR750, yielding pJIR2816. For the construction of the insertional inactivation vectors, rgaR and rgbR internal fragments were PCR-amplified from C. difficile strain 630 genomic DNA, using JRP2307/2308 (424 bp) and JRP2305/2306 (379 bp) respectively. The PCR products were digested with Sphl and Xbal restriction endonucleases, then cloned into the same sites in pJIR2816. E. coli DH5α clones were selected, using X-gal and blue-white selection. The plasmids were confirmed by restriction analysis, then the inserts were sequenced. This resulted in pJIR3012 (pJIR2816 + rgbR fragment) and pJIR3013 (pJIR2816 + rgaR fragment) . To prevent the possibility of read-through expression from this vector, a 2064 bp omega cassette fragment from pJIR2224 (Cheung et al. , 2004) was subcloned using Xbal/Asp718 into the same sites in both pJIR3012 and pJIR3013. Recombinant clones were selected on the basis of streptomycin and chloramphenicol resistance, and the vectors were confirmed to be correct using restriction endonuclease digestion analysis. This resulted in the vectors pJIR3015 (pJIR3012 + Ω) and pJIR3014 (pJIR3013 + Ω) .

Construction of C. difficile rgaR and rgbR mutant and overexpression strains

Plasmids were introduced into electrocompetent JE?. coli strain S17-1, which carries the broad host-range plasmid RP4 , and is capable of mobilizing IncP oriT plasmids, such as pJIR1456 and pJIR2816. These plasmid constructs were then transferred by conjugation, using a method modified from that of Mani et al (2002) , to C. difficile JIR8094, an erythromycin-sensitive derivative of the sequenced toxigenic strain 630 isolated in our laboratory, using a method modified from that used by (Mani, Lyras et al . 2002) . Two ml of an overnight culture of JIR8094 in BHIS medium was used to inoculate 90 ml of the same medium, and this mixture was incubated at 37⁰C for approximately six hours, until the culture reached and OvD.goonm of 1.05-1.2. Meanwhile, cultures of E. coli donor strains derived from strain S17-1 were prepared in 2 X YT broth medium supplemented with chloramphenicol (30 μg/ml) , and grown to an O.D.₆₀on_m of approximately 0.35-0.44. A 100 μl sample of each donor and recipient culture was mixed together and spread on a thick (approx 25 ml) BHIS agar plate. Samples of each individual culture were also spread on separate plates of the same medium to serve as controls . The plates were incubated in an atmosphere of 10% H₂-10% CO₂- 80% N₂ for 20 to 24 h at 37⁰C. Then biomass from each plate was resuspended in one ml of BHI diluent (0.37 g BHI medium and 0.01 g sodium thioglycollate per litre), and 100 μl of this mixture was spread on BHIS agar supplemented with thiamphenicol (10 μg/ml) and cefoxitin (25 μg/ml) .

Cefoxitin was used as a counter selection against E. coli. These plates were incubated anaerobically at 37⁰C for at least 72 hours, until colonies could be seen on the plates. These colonies were patched onto the same medium and incubated for a further two to three days. After this, single C. difficile colonies could be seen within some of the patched areas, and these colonies were subcultured twice, then analysed by PCR, Southern blotting, and plasmid rescue experiments. This resulted in the following recombinant C. difficile strains, respectively designated

JIR8149: JIR8094 (pJIR2634) {rgaR overexpression strain) ,

JIR8150: JIR8094 (pJIR2633 ) {rgbR overexpression strain) ,

JIR8218: JIR8094ΩpJIR3014 {rgbR mutant) ,

JIR8223: JIR8094ΩpJIR3015 {rgaR mutant 1), and JIR8226: JIR8094ΩpJIR3015 {rgaR mutant 2).

C. difficile plasmid stability assays

Recombinant C. difficile strains were subcultured on to BHIS agar containing thiamphenicol and cefoxitin, then colonies were inoculated into two 20 ml BHIS broth cultures, either with or without thiamphenicol selection, and incubated anaerobically overnight. The next day, serial dilutions of each culture were carried out in BHI diluent, and dilutions from each broth were plated out on to BHIS agar containing either thiamphenicol and cefoxitin, or cefoxitin alone for a viable count . Plates were incubated for 48 hours, and then the colonies were counted. Stability of the catP-encoded thiamphenicol resistance marker was expressed as the ratio of thiamphenicol resistant colonies to the total viable count for that broth culture .

Construction of rgaR and rgaR complementation vectors

Complementation vectors were based on the E. coli-C. difficile shuttle vector pMTL9301 (Purdy, O'Keeffe et al . 2002). For the rgaR complementation vector, the rgaR genomic region was PCR-amplified with Pfu polymerase and JRP 1476/1477 (1.45 kb product), then the blunt-ended product was cloned into the Fspl site of pMTL9301, in the same direction as the erm(B) gene in this vector; the plasmid was confirmed to be correct by restriction endonuclease digestion analysis, and the rgaR PCR product sequenced. This construct was named pJIR3041. For the construction of the rgbR complementation vector, an approximately 1.74 kb PvuII fragment from pJIR2515, carrying the rgbR gene, was subcloned into the Fspl site of pMTL9301, in the same direction as the erm(B) gene, then the plasmid was confirmed to be correct by restriction endonuclease analysis. This plasmid was designated pJIR3042.

Introduction of complementation vectors into C. difficile rgaR and rgbR mutant strains

Plasmids were introduced into electrocompetent E. coli strain S17-1. These plasmids were then transferred into the C. difficile rgaR and rgbR mutant strains, using a method similar to those previously described (Purdy,

O'Keeffe et al . 2002). A 20 ml BHIS broth culture of the C. difficile recipient strain was grown overnight, as well as a 5 ml 2 X YT broth culture of the appropriate S17-1 donor strain, supplemented with 150 μg/ml erythromycin. A one ml sample of each C. difficile culture and approximately 300 μl of E. coli S17-1 donor strain was independently centrifuged at 4000 rpm in a benchtop centrifuge for three minutes, washed once with BHI diluent, then the cell pellets were each resuspended in one ml of BHI diluent. A 100 μl sample of donor and recipient culture were plated out together on BHIS agar, and were also plated out separately as controls . These plates were incubated anaerobically for 6-7 hours, then the cells were resuspended in 500 μl BHI diluent and plated on to BHIS agar plates containing 50 μg/ml lincomycin (to select for the plasmid in C. difficile) and 25 μg/ml cefoxitin. These plates were incubated for 48-72 hrs, until colonies appeared. C. difficile transconjugants were screened by their antibiotic resistance profile, and only colonies which were resistant to both lincomycin and thiamphenicol were subjected to PCR analysis, Southern blotting, and plasmid rescue experiments. This resulted in the following C. difficile strains:

JIR8234: JIR8218 (pMTL9301) ,

JIR8236 JIR8218 (pJIR3042) ,

JIR8243 JIR8223 (pMTL9301) , JIR8233 JIR8223 (pJIR3041) ,

JIR8245 JIR8226 (pMTL9301) , and

JIR8247 JIR8226(pJIR3041) .

Construction of the C. difficile double crossover plasmid pJIR2755.

A series of vectors was constructed based on the commercial E. coli high-copy cloning vector, pT7-blue-3 (Novagen) . A sample of pT7-blue-3 vector, linearised at the EcoRV site (from the Novagen Perfectly Blunt™ cloning kit) , was ligated and the EcoRV site was eliminated, resulting in pJIR2567. Then pJIR1456 was digested with Sall/Hindlll and the RP4oriT fragment was gel extracted, the ends were filled in using the end-conversion mix from the Novagen pT7-Blue-3 Perfectly Blunt cloning kit, and the resultant DNA fragment was blunt-end cloned into the CIaI site of pJIR2567, resulting in pJIR2591.

The ermQ gene was subcloned from pJIR478 (Berryman et al, 1994) on a 1.1 kb Pstl/EcoRI fragment, and was cloned into the same sites in pBluescript KS

(Stratagene) ; this resulted in pJIR1120. Antibiotic resistance markers were then inserted into the base plasmid pJIR2591; the ern?(Q) gene was PCR amplified using primers JRP1870/1871 from the plasmid pJIR1120, and cloned into pJIR2591 Spel/Sacl, which resulted in pJIR2630. The catP gene was PCR amplified from Tn4453a in E. coli JIR5707, a strain which carries Tn4453a inserted into the chromosome, using primers JRP1076/JRP1077, and cloned into pT7Blue-3 (Novagene) using the Novagene Perfectly Blunt™ cloning kit, this resulted in pJIR2652. Then pJIR2652 was digested with EcoRl and the 1 kb catP fragment was subcloned into the same site in pJIR2591; this resulted in pJIR2653. To construct a plasmid which carried both the ermQ and catP antibiotic resistance markers, the catP gene was PCR amplified from Tn4453a in E. coli JIR5707 using primers JRP1882/JRP1883 , and the resultant 1 kb PCR product was cloned into pJIR2630, which had been partially digested with Smal . The catP gene was inserted into the Smal site, which lies between the oriT and the E. coli origin of replication and is in the same orientation as the ampicillin resistance gene in this plasmid. This resulted in pJIR2715 .

To construct the C. difficile rgaR double crossover plasmid, the DNA region upstream of the rgaR gene was PCR amplified using primers JRP1700/JRP1697, the 979 bp product was digested with MIuI and cloned into the same site in pJIR2653; this resulted in pJIR2674. The DNA region downstream of the rgaR gene was PCR amplified from C. difficile JIR8094 genomic DNA, using primers JRP1699/JRP1698, and cloned into pJIR2674 at the JVhel site, after the ends of this digested plasmid were filled in using T4 polymerase. Each of the cloned C. difficile DNA fragments was orientated in the same direction in order to to facilitate a double crossover homologous recombination event between this region and the homologous region on the C. difficile chromosome, and was also cloned in the same direction as the ampicillin resistance gene. This resulted in pJIR2720. The cloned rgaR upstream DNA region from pJIR2674 was then subcloned on a 1 kb fragment Pstl/BamΑI into the same sites in pJIR2630; this resulted in pJIR2733. The rgaR downstream DNA region from pJIR2720 on a 1 kb Xhol/Notl fragment was cloned into the same sites in the pJIR2733 plasmid; this resulted in pJIR2736. Following this, the 3.6 kb Sphl/Notl DNA fragment from pJIR2736 which encompassed the C. difficile DNA and the ermQ gene was subcloned into the same sites of the 4.5 kb fragment of the pJIR2715 plasmid backbone, resulting in pJIR2755.

To construct the C. difficile rgaR double crossover suicide plasmid, the DNA fragment encompassing the DNA region to facilitate double cross recombination, which includes ermQ, was subcloned from a previously- constructed vector. This vector, pJIR2755, was digested with Sphl/Notl and the 3557 bp fragment encoding the relevant DNA was subcloned into the same sites of the unstable E. coli-C. difficile shuttle vector, pJIR1456. This resulted in plasmid pJIR3011. The double cross-over region included the following C. difficile DNA: upstream of the rgaR gene, including the first 257 bp of the coding region (PCR amplified using primers JRP1700/1697 ; 979 bp) and the genomic region downstream of the rgaR gene (amplified with primers JRP1699/1698 ; 1186 bp) . The PCR product corresponding to the rgaR upstream DNA was cloned on the 5' side flanking the ermQ gene, and the PCR product corresponding to the rgaR downstream region was cloned on the 3' side flanking the ermQ gene, this resulted in the "double crossover region" .

Construction of C. difficile rgaR double crossover/deletion strain

Upon introduction into C. difficile and two recombination events (double crossover) across the homologous DNA regions of pJIR3011, a substantial deletion in the rgaR gene would be generated. The plasmid was first introduced into the E. coli strain HBlOl (pVS520) , which is capable of mobilising plasmids such as pJIR1456, that carry the RP4 oriT region. The vector was then transferred to C. difficile JIR8094 by conjugation. The conjugative method used was similar to that employed for construction of single cross-over strains, with the following exceptions: the ratio of donor and recipient culture volumes used in the matings was 1:1, 1 ml of each donor and recipient culture was centrifuged, as before, then washed twice and resuspended in 1 ml of BHI diluent, then 100 μl volumes of each donor and recipient were plated out together on non-selective media for the mating step, alongside the appropriate controls . The plates were incubated anaerobically, at 37° C for 6.5 hours, then resuspended in 1 ml of BHI diluent and 100 μl volumes were plated onto selective media containing cefoxitin and lincomycin, to select for C. difficile transconjugants . These plates were incubated anaerobically, at 37 'C for six days, then colonies were cross-patched on to selective agar containing either cefoxitin and lincomycin, to select for presence of the double cross over region of the plasmid, and cefoxitin and thiamphenicol, to determine whether the backbone of the suicide vector was still present, indicating that a single crossover had probably occurred. The patches were incubated for two days anaerobically at 37 "C. All transconjugants which carried the ermQ lincomycin resistance marker also carried the catP thiamphenicol resistance marker, indicating that a single crossover event had most likely occurred, and that the rgaR gene was probably not insertionally inactivated.

Generation of a C. difficile rgaR double cross-over from a single crossover strain.

The C. difficile strains carrying the entire pJIR3011 plasmid were subcultured three times on medium containing cefoxitin and lincomycin (from the primary inoculum of each successive culture) , then colonies from four isolates were inoculated into four separate 20 ml BHIS broths, without any selective pressure, and incubated overnight anaerobically at 37 "C. The next day, 2 ml of each overnight culture was inoculated into a 90 ml BHIS broth and incubated for 7 hours anaerobically at 37⁰C. Following this, serial 1 in 10 dilutions were carried out, in BHIS broth and 100 μl volumes from dilutions between 10^"3 and 10^"5 were plated out on to half of an agar plate containing selective medium with either cefoxitin and lincomycin, or cefoxitin and thiamphenicol, as well as on non-selective medium. These plates were incubated at 30⁰C in an anaerobic jar for 3 days, then for a further two days at 37⁰C in an anaerobic chamber. Colonies growing on the plates containing lincomycin were cross-patched on to plates containing cefoxitin and thiamphenicol or lincomycin. One isolate was found to be lincomycin- resistant and thiamphenicol-sensitive; and was designated JIR8292. Four independent isolates which still carried the catP thiamphenicol resistance marker were also stored, and these were designated JIR8293-JIR8296.

Microarray expression analysis of recombinant C. difficile strains

The C. difficile microarray was designed using the approach previously described (Hinds et al . , 2002) to represent all 3,688 chromosomal coding sequences (CDSs) originally predicted in strain 630; a further 92 additional small CDS have been annotated since construction of the microarray and are therefore not represented. For each gene, a single PCR product reporter element which had minimal cross- hybridisation to other genes in the genome was designed. Multiple reporter elements were designed for some genes, including eight for tcdA, seven for tcdB, three for cdtA, four for cdtB and two for each gene involved in S-layer formation. The microarrays were constructed by robotic spotting of the PCR products in duplicate on UltraGaps amino-silane coated glass slides (Corning, USA) using a MicroGrid II (BioRobotics, UK) (Hinds et al . , 2002). The array design is available in BμGΘSbase (Accession number: A-BUGS-20; http://bugs.sgul.ac.uk/A-BUGS-20) and also ArrayExpress (Accession number: A-BUGS-20). Synthesis of cDNA, labelling and hybridisation was carried out using a Genisphere 3DNA Array 900MPX microarray kit

(Genisphere Inc., U.S.A.), optimised as follows. The amount of starting RNA was increased to 6.5-10 μg, the microarray slide was prehybridised in a solution of 25% formamide, 5 X

-1 -1

SSC, 0.1% SDS, lOmg ml bovine serum albumin and lmg ml salmon sperm DNA (Promega) , in a final volume of 30 μl in dH₂0, at 42⁰C for 45 min, and for the probe hybridisation step, the high sensitivity buffer (vial 5) was used in a final hybridisation volume of 44 μl, hybridisation was carried out overnight at 57°C using 22 x 50 mm glass coverslips (HD Scientific) . Following hybridisation, the first wash was carried out at 42⁰C for 20 min, and the next two washes were carried out as per the manufacturer. The Cy3 or Cy5 dyes were attached to the probes in hybridisation buffer 6, in a volume of 44 μl at 55⁰C for 4 h. The first wash was carried out at 55⁰C for 20 min and the following two washes were carried out at room temperature for 20 min. The microarray slides were dried and then scanned using a GMS418 array scanner (Affymetrix) . The scans were acquired using GMS Scanner Software

(V.1.51.0.42) (Affymetrix) and the flu ®orescence intensity information quantitated using ImaGene version 5.1 (BioDiscovery, Inc) . The ImaGene data files were then subjected to statistical analysis using the Limma software package for R (Smyth 2004; Smyth, Michaud et al . 2005) . Briefly, background correction was performed, then Loess normalisation was applied to each print-tip group separately. The arrays were then normalised together by scaling the log-ratios to ensure consistency between arrays. The fold-ratio for each spot was calculated and then a single estimate was made of the correlation between within-array spot replicates under the assumption that the between-replicate correlation for each gene is common across all genes (Smyth et al., 2005) . A moderated t-test was then applied to each gene, where the variance was calculated as a combination of the variance for this gene, the pooled variance and the estimated spot replicate correlation. Each gene was then ranked according to the level of differential expression determined. The P-values were subsequently adjusted for multiple testing using False Discovery Rate (FDR) . This test provides more sensitive analysis when compared to rank-tests, because it allows for deviation between within-array replicate spots when estimating the accuracy of the data for each individual gene. Fully annotated microarray data have been deposited in BμGQSbase (Accession number: E-BUGS-XX; http://bugs.sgul.ac.uk/E-BUGS- 25 XX) and also ArrayExpress (Accession number: E-BUGS-XX) .

Bioinformatic analysis of the C. difficile genome

The search for putative C. difficile VirR boxes was carried out using the following search parameters; the conserved VirR box sequence, (CCAGTT (N₁₅) CCAGTT) was used and a one base-pair mismatch at any of the conserved nucleotides was allowed. The results were then analysed to determine whether the sequences identified were appropriately located upstream of a promoter region, and in the correct orientation to constitute a potentially functional VirR box, in accordance with previous findings (Cheung and Rood 2000; Cheung, Dupuy et al . 2004). BLASTP searches were used to identify the C. difficile VirR homologues and to assign potential functions to C. difficile genes of interest (Altschul, Gish et al . 1990).

Expression and purification of His- tagged VirR-like proteins

His-tagged C. perfringens VirR protein was overexpressed and purified as previously described (Cheung and Rood 2000; Cheung, Dupuy et al . 2004) . The C. difficile rgαR gene was PCR amplified and cloned into the E. coli expression vectors, pET22b⁺ (C-terminal hexa-histidine tag) and pET28a⁺ (N-terminal hexa-histidine tag) resulting in pJIR2259 and pJIR2561, respectively. These plasmids were then introduced into the B. coli expression strain C43(DE3) (Miroux and Walker 1996) . Transformants were grown at 37°C to mid-log phase and then expression of the recombinant protein was induced with ImM IPTG and allowed to progress for three hours. His-tagged C. difficile RgaR protein was purified by Talon chromatography as previously described (Cheung and Rood 2000) . Purified His-tagged RagR was detected by SDS-PAGE and Western blotting analysis, using Penta-His™ Antiserum (QIAGEN), then dialyzed in 30OmM NaCl, 5OmM Tris, 2OmM EDTA, 10% glycerol, pH 7.5 at 4⁰C and stored at -7O⁰C. Purified protein was quantitated using a Bradford assay (Bio-Rad) .

Gel mobility shift analysis

Gel mobility shift assays were carried out using PCR products corresponding to the promoter regions of rpoA, CD0587, CD2098 and CD2750, all of which were amplified from

C. difficile JIR8094 genomic DNA using primers listed in Table 2. The C. perfringens P _foA VirR boxes, and the mutated VirR boxes (CCA to TAG in both VirR boxes; (Cheung and Rood 2000)), were amplified from pJIR1546 and pJIR1821, respectively, using primers JRP589/JRP618. DNA was end- labelled using digoxigenin (DIG) -11-ddUTP and terminal transferase, according to the manufacturer's instructions

(Roche) . Binding reactions were carried out and detected as previously described (Cheung and Rood 2000) , and included 0-30 pmol of VirR or RgaR protein and 15 fmol of DIG- labelled DNA. When required, binding specificity assays were carried out through the addition to the reaction of 5 or 10 pmol of unlabelled DNA competitor of either specific

(same sequence as the labelled target) or non-specific (P_r , or mutated VirR boxes) origin.

Example l(a) Introduction of pJIR1456-based overexpression vectors into C. difficile JIR8094 results in homologous recombination between regions of sequence identity Because of the absence of a reproducible method for the targeted construction of mutants in C. difficile, we initially sought to study the function of individual response regulators by overexpressing them on a C. perfring-ens-E. cσli multi-copy shuttle vector, pJIR1456, which was described in US Patent No. 5,955,368. The methodology needed to introduce this vector into C. difficile had previously been established in our laboratory (Mani et al . , 2002) . Previous work from our own and other laboratories indicates that overexpression of a response regulator is generally equivalent to it being activated by a cognate sensor histidine kinase (Lyristis et al . , 1994; Alsaker et al . , 2004) . This method should have allowed us to study the function of response regulators without being required to know the signal which leads to their activation. Two response regulators, the VirR homologues rgaR and rgbR, were chosen for study using this system.

(b) Construction of C. difficile overexpression strains

Primers were designed to amplify rgaR and rgbR along with upstream and downstream DNA such that any putative promoter and terminator sequences should have been included. These PCR products were then cloned EcoRl/Sphl into pJIR1457, resulting in the vectors pJIR2514 {rgaR) and pJIR2515 {rgbR) , such that each gene was cloned in the same direction as the plasmid-encoded β-galactosidase gene promoter, to attempt to increase expression levels of the response regulators, or to allow for some expression if their promoter regions had been inadvertently deleted. The cloned fragments were sequenced, then the fragments were sub-cloned into pJIR1456, resulting in pJIR2633 (pJIR1456 rgbR*) and pJIR2634 (pJIR1456 rgaR*) and the plasmids electroporated into the B. coli donor strain S17-1. The plasmids were then introduced into strain JIR8094, an erythromycin-sensitive derivative of C. difficile strain 630, previously isolated in our laboratory, by conjugation, and transconjugants were selected on media containing cefoxitin and thiamphenicol . This resulted in the following strains:

JIR8145 : JIR8094 (pJIR1456) ; JIR8149: JIR8094 (pJIR2634) ; and JIR8150: JIR8094 (pJIR2633) .

(c) Recovery of shuttle vectors from recombinant C. difficile strains

The resistant C. difficile transconjugants were subcultured twice to obtain pure cultures, and then we attempted to confirm that these strains were genotypically correct by extracting C. difficile plasmid DNA from overnight broth cultures and introducing a sample of this preparation into E. coli DH12S via. electroporation. E. coli transformants were selected on medium containing 30μg/ml chloramphenicol. Transformants were then analysed by plasmid miniprep and restriction analysis. This experiment yielded surprising results : intact plasmid DNA was recovered from C. difficile harbouring pJIR1456, albeit in small amounts, but plasmid DNA could not be recovered from the C. difficile isolates harbouring the overexpression plasmids pJIR2633 or pJIR2634. It was initially thought that this lack of plasmid recovery might have been due to the absence of antibiotic selection in the C. difficile broth cultures, so the cultures were grown again, as before, this time with the inclusion of lOμg/ml thiamphenicol . Our results indicated that the additional selection pressure had been successful in increasing the efficiency of plasmid recovery from the C. difficile strain carrying pJIR1456, but again plasmids could not be recovered from the C. difficile strains harbouring pJIR2633 or pJIR2634. On the basis of these results, we suspected that the shuttle vector pJIR1456 was unstable in C. difficile JIR8094, and so we conducted stability assays to determine whether this was the case.

(d) Analysis of the stability of pJIR1456-based shuttle vectors in C, difficile JIR8094 In order to determine the relative stability of the pJIR1456-based shuttle vectors in C. difficile, several colonies from each relevant strain were taken from agar containing selection antibiotic and inoculated into 20ml BHIS broths with or without lOμg/ml thiamphenicol . During this experiment we also analysed the previously constructed strain, JIR8176 from our laboratory, which is an avirulent C. difficile strain, CD37, harbouring the shuttle vector pJIR1456. Following overnight incubation in broth culture, serial dilutions were made for each strain, and duplicate aliquots were plated out on to medium which selected for presence of the plasmid, or on to non-selective medium to allow for a viability count. The proportion of resistant colonies in the population was then determined as a percentage of the total viable count and the relative stability of the plasmids was then determined. The results indicated that in the absence of antibiotic selection pJIR1456 was unstable in both C. difficile JIR8094 and CD37. When selection was included in the broth culture, •strains carrying pJIR1456 retained the antibiotic resistance marker at a much higher rate, as summarised in Table 3.

Table 3

Stability of thiamphenicol resistance in JIR8145, JIR8149 and JIR8150

N. B. The stability assay results are an average of two ^'. independent experiments .

Analysis of C. difficile JIR8149 and JIR8150 revealed that these strains remained thiamphenicol- resistant, irrespective of whether antibiotic selection was included or not. PCR analysis was carried out on these strains to confirm whether plasmid-specific genes were present, and it was found that the strains did indeed carry the plasmid-specific antibiotic resistance determinant, and thus were not spontaneous mutants. Together, these results led us to suspect that the overexpression vector constructs, pJIR2633 and pJIR2634, were no longer ; replicating autonomously in the C. difficile transconjugant j strains, and had integrated into the bacterial chromosome. !

(e) Southern blotting of C. difficile strains harbouring pJIR1456-based vectors Genomic DNA was isolated from C. difficile strains JIR8145, JIR8149 and JIR8150, grown in the absence of selective pressure. The DNA was digested with restriction endonuclease enzymes Seal or AsplOO, and separated on a 0.8% agarose gel, then Southern blotting was carried out and the blots were probed with catP, and the rgaR / rgbR genes. The results are shown in Figure 1, and revealed that the pJIR2633/pJIR2634 overexpression vectors had indeed integrated into the chromosome of the parent strain by a single crossover event, and that autonomous vector could not be detected. To confirm that the ^' duplicated response regulators were intact, the region ^! encompassing the single cross-over site was PCR amplified and seguenced; the results indicated that no mutations or rearrangements had occurred. (f) Transcriptional analysis of rgaR and rgbR expression levels in C difficile single crossover strains Having confirmed that the overexpression vectors had integrated into the C. difficile chromosome, it was then necessary to determine whether the duplicated response regulator genes had high enough expression levels to allow detection of the effects of these gene products using microarray analysis, ie. to determine whether each response regulator was expressed highly enough to allow activation of target genes .

Growth curve analysis was carried out under conditions which have been widely reported to promote higher levels of C. difficile toxin production (Ketley,

Mitchell et al . 1986; Kamiya, Ogura et al . 1992; Dupuy and Sonenshein 1998) . The approximate time and optical density at which these strains entered into stationary phase, which was the same for the single crossover strains and the parent strain, was used as the time for culture harvest. RNA extraction was then carried out, and quantitative RT- PCR was carried out to determine the level of rgaR/rgbR expression in each single crossover strain, as compared to the wild-type. The C. difficile rpoA gene transcript was used as the normalization factor for each strain, since expression levels for this gene should not be subject to modification due to regulatory signals.

Due to the low %G+C content of the C. difficile genome, designing appropriate QRT-PCR PCR primers was difficult, and it was not possible to design functional primers for rgbR. The primers designed for rpoA and rgaR worked sufficiently well to allow analysis to proceed on the C. difficile rgaR single crossover mutant. The results indicated that rgaR expression levels in the mutant were twice that of the wild type strain; however, we did not consider this level of expression to be high enough to ensure that differences in expression levels of target genes would be reliably detected by microarray analysis. This result confirmed that JIR8149 now contained two copies of the rgaR gene.

Example 2 (a) Insertional inactivation of rgaR and rgbR in

C. difficile JIR8094 by single crossover The evidence gathered during our attempts to construct C. difficile RR overexpression strains demonstrated that the instability of pJIR1456 in this system could be used for reproducible construction of single crossover strains. Using this information, we designed single crossover vectors based on pJIR2816, a derivative of pJIR1456. Single crossovers were designed to include an internal fragment of the 5 ' end of the rgaR or rgbR genes, such that a single crossover event between this construct and the chromosome would lead to the 5 ' end of the response regulator being expressed and the 3 ' end being inactivated. The constructs were designed in this way because the C-terminal domain of most response regulator proteins is responsible for the DNA binding/activation activity. Thus inactivation of this region of the gene would be expected to make the corresponding protein nonfunctional .

(b) Construction of rgaR and rgbR single crossover mutants

An internal fragment of the 5 ' end of each of rgaR and rgbR was PCR amplified, separately cloned into pJIR2816 Xbal /Sphl , and the cloned regions confirmed by sequencing. The resultant plasmids were designated pJIR3012 {rgaR) and pJIR3013 {rgbR) . The Ω transcriptional terminator cassette was then inserted immediately upstream of this fragment to prevent any chance of read-through from the plasmid into the 3 ' end of the gene once a single crossover had occurred. The resultant plasmids were designated pJIR3015 {rgaR) and pJIR3014 {rgbR) , these vectors are represented schematically in Figure 2. These plasmids were analysed by restriction analysis, and found to have the correct profile; the constructs were then electroporated into the JS?. coli donor strain S17-1, and the resultant strains were then used to introduce the single crossover constructs into C. difficile JIR8094. Transconjugants were selected on agar medium containing cefoxitin and thiamphenicol ; one rgbR mutant, designated JIR8218, was isolated. Two independently- generated rgaR mutants were isolated and named JIR8223 and JIR8226. The strains were subcultured twice, then genomic and plasmid DNA was extracted in the presence and absence of thiamphenicol selection.

(c) Plasmid analysis of C difficile rgaR and rgbR mutants

Plasmid DNA was purified from the transconjugants and electroporated into E. coli DH12S; transformants were selected on agar medium containing 30 μg/ml chloramphenicol . Plasmid DNA was extracted from some of these transformants, and the undigested plasmid profile was compared to that of the plasmid originally introduced; their sizes were found to be the same. There appeared to be a ten-fold reduction in the efficiency of plasmid recovery in the strains into which the single crossover vectors had been introduced, as compared to the shuttle control strain.

A stability assay was carried out in C. difficile, as before, using the rgbR mutant, JIR8218, and the pJIR1456 shuttle control strain, JIR8145. The thiamphenicol resistance marker specific to the single crossover vector was found to be stably maintained, while the pJIR1456 shuttle control strain rapidly lost the thiamphenicol resistance marker. The rgbR mutant and several rgaR mutants were repeatedly subcultured on non- selective medium in an attempt to cure the strains of the autonomously replicating vector. No cured derivatives were obtained. (d) PCR analysis of single cross-over mutants PCR analysis of the extracted genomic DNA from the rgaR mutants indicated that a single crossover event had occurred, since strong products could be detected using a primer firing out of the plasmid and one specific to the chromosome. Under conditions using a large amount of genomic DNA template, trace amounts of PCR product could also be detected for the product (data not shown) , which indicated reconstitution of the gene targeted for inactivation, ie. the plasmid was looping out from the chromosome. Additionally, PCR across the multiple cloning site of the vector, which would only be positive if the plasmid existed autonomously from the chromosome, yielded a faint product. The C. difficile rgbR mutant strain was analysed in the same manner, and the results are shown in Figures 3 and 4.

These results indicated that a single crossover event had inactivated the rgbR and rgaR genes respectively, but that a product corresponding to the intact gene could also be seen, indicating that looping out of the single crossover plasmid occurs at low frequency.

(e) Southern blotting analysis of rgaR and rgbR single crossover mutants

Genomic DNA was purified from the mutant and parent strains and digested with restriction enzymes, such that the enzyme would cleave the integrated suicide vector backbone and the replicating complementation vectors once each, and a unique hybridization profile would be obtained when compared to the wild-type strain. Blots were probed with the response regulator open reading frame, the plasmid-encoded thiamphenicol resistance marker, catP, and the 3 ' end of the insertionally inactivated response regulator (ie. the region which should not be expressed) . Initially, Hpal was used for rgbR mutants, and Scsl/Ecdl for rgaR mutants. In a second experiment Hpal/Ncol for rgbR mutants and Sacl/Ncol for rgaR mutants were used to optimise the fragment sizes.

The results, shown in Figures 5 and 6 {rgaR mutant) and 7 and 8 {rgbR mutant) , indicated that two copies of the 5 ' end of the respective response regulator gene could be detected, compared to one copy in the wild type strain, but that only one copy of the 3' end of the response regulator gene was present. The hybridisation profiles were as expected for insertional inactivation of the 3' end of the response regulator genes. The Southern blots also indicated that some autonomously replicating vector could be detected in the rgaR mutant strains (Figures 5 and 6) , but the reconstituted response regulator gene, previously demonstrated by PCR analysis to exist, could not be detected in the mutants , suggesting that the reconstitution of the response regulator gene occurred at low frequency in the overall bacterial population, so that it could only be detected by very sensitive assays.

Example 3 (a) Complementation of RR insertional inactivation mutants

Since the vector used to construct the single crossover mutant encoded thiamphenicol resistance and carried a pIP404-based replicon, we used the C. difficile shuttle vector pMTL9301 for complementation (Purdy,

O'Keeffe et al. 2002) . This vector carries a replication region isolated from pCD6, a plasmid which was originally isolated in C. difficile strain CDS. Previous work had demonstrated that this replicon allowed stable maintenance of an autonomously-replicating shuttle vector in C. difficile strain 630, the strain from which our wild type strain, JIR8094 was derived (Herbert, O'Keeffe et al . 2003) .

The rgaR complementation fragment was PCR amplified using primers JRP1476/JRP1477, and blunt-end cloned into the pMTL9301 Fspl site. This resulted in pJIR3041, then the resultant clone was sequenced. The rgbR complementation fragment was expressed from pJIR2515 using PvuII, and the blunt fragment was subcloned into the Fspl site of pMTL9301; this resulted in pJIR3042. Both constructs were analysed by restriction digestion, introduced into E. coli SIl-1 and introduced into the appropriate C. difficile response regulator mutant. Transconjugants were selected on agar medium containing cefoxitin and lincomycin, then cross-patched back on to thiamphenicol-containing medium to confirm that the single crossover antibiotic resistance marker was still present.

(b) Analysis of complemented C difficile mutants

Plasmid DNA was extracted from C. difficile overnight broth cultures, grown in the absence of selection, and the preparations electroporated into E. coli DH12S, as described herein. E. coli transformants were selected on agar medium containing erythromycin alone, chloramphenicol alone, or both chloramphenicol and erythromycin, in order to select independently for the single crossover vector and the complementation vector, and to allow for the potential recovery of cointegrates .

The results showed that for all transformations, the complemented mutants and the vector control strains, E. coli transformants were only detected on erythromycin selection alone, indicating that the residual autonomously replicating single crossover plasmid was not detectable. Nor was co-integration of pMTL9301 and pJIR1456-based suicide vectors detected.

(c) PCR analysis of C. difficile complemented mutant strains

In order to demonstrate that the pMTL9301 control vector or the pJIR3041 and pJIR3042 complementation vectors had been successfully introduced into the appropriate C. difficile mutant strains, PCR analysis was carried out on the transconjugants to show that the integrated single crossover plasmid was still present, indicating that the chromosomal copy of the gene remained inactivated. The results are illustrated in Figures 3 {rgaR strains) and 4 {rgbR strains) . A repA PCR was performed to detect the replication region from the introduced complementation and control vectors, and a PCR to show linkage between the plasmid encoded erm(B) gene and the rgaR/rgbR complementation fragment was also carried out. These results are shown in Figure 9. The results indicated that the complemented strain retained the single crossover plasmid, and also harboured the correct complementation or control vector, as appropriate.

(d) Southern blotting analysis of complemented C. difficile mutants

Southern blotting analysis was carried out on the rgaR mutant strains, strains into which pMTL9301 (control vector) and pJTR3041 (rgaR complementation vector) had been introduced. The results are shown in Figure 6. Digestion of the genomic DNA with Sacl and Ncol yielded a unique profile which indicated that the single crossover plasmid remained integrated into the chromosome in both mutant- based strains analysed. In addition the pMTL9301 and pJIR3041 repA region could be detected on a fragment of the expected size (approx. 7.1 kb) , which corresponded to the plasmid control lanes. The strains carrying pJIR3041 were also found to carry an intact rgaR gene, since the rgaR probes hybridized to a band specific to the complementation vector. Genomic DNA from the rgbR mutant strain and from strains carrying pMTL9301 and pJIR3042 control vector was digested with Hpal/Ncol, to confirm that the chromosomal copy of the rgbR gene remained inactivated, and that the complementation and control vectors were present in the appropriate strains. As shown in Figure 8, the chromosomal arrangement of the strains remained unchanged, and the pMTL9301 and pJIR3042 plasmids were present. The rgbR probes also hybridized to bands specific to the rgbR complementation vector, pJIR3042.

Example 4 Microarray and Quantitative Real Time RT-

PCR analysis of C. difficile rgaR and rgbR single crossover mutant strains .

(a) Microarray analysis was carried out in an attempt to deduce the regulatory role of the rgaR and rgbR gene products in C. difficile . The microarrays consisted of gene-specific PCR fragments representing the entire C. difficile genome, and were obtained from the Bacterial Microarray Group, St George's, University of London (http://bugs.sgul.ac.uk). In these experiments, the strains were cultured in broth under conditions that were expected to optimise toxin production, ie. in the absence of glucose, which is known to inhibit toxin production (Dupuy and Sonenshein 1998; Mani, Lyras et al . 2002) . The cultures were . harvested in early stationary phase, when both toxins A and B should be produced and the cellular mRNA should still be sufficiently intact to allow transcriptional analysis (Dupuy and Sonenshein 1998) . Subculture on the appropriate selective medium, at the time that the cultures were harvested, confirmed that the single crossover thiamphenicol antibiotic resistance marker (catP) was stable under these conditions (data not shown) . Similarly, the control pMTL9301 vector was stable in both the rgaR and rgbR mutants, as was the rgbR⁺ complementation vector. However, the equivalent rgaR⁺ complementation plasmid was unstable in the rgaR mutant when grown under these conditions, suggesting that overexpression of the RgaR protein was detrimental to these cells. As a result it was not possible to carry out microarray analysis on the complemented rgaR mutant .

Microarray analysis was carried out using RNA from three independent biological replicates of each strain, as described in Materials and Methods. Analysis of microarrays which compared RNA from the wild-type strain and the rgbR mutant indicated that, under these growth conditions, mutation of the rgbR gene had no detectable effect on gene expression. Furthermore, comparison of the complemented rgbR mutant with the same mutant carrying the vector control plasmid indicated that only one differentially-expressed gene could be detected, namely the rghR gene itself, as expected (4.1 fold higher expression, P = 0.02). This result confirmed that complementation of the rgbR gene was successful, and that there was no other effect on gene expression.

Different results were obtained with the rgaR mutant. Comparison with RNA from the wild-type strain indicated that the rgaR mutation significantly down- regulated the expression of four genes. These results are shown in Table 4.

Table 4: Microarray and QRT-PCR analysis of the C. difficile rgaR mutant

Gene Microarray QRT-PCR a b

Fold ratio P-value Expression P- level value

CD0588 0.053 0.006 0.039 ± 0.018^e 0.0001

CD0590 0.11 0.006 0.083 ± 0.028 0.0002

CD2098 0.31 0.010 0.17 ± 0.091 0.002 αgrB (CD2750) 0.49 0.006 0.34 ± 0.23 0.019

^a Ratio of expression in rgaR mutant relative to wild-type expression level. ^bP-value calculated by moderated students t-test ^c Expressed as a proportion of wild type expression level ^dP-value calculated by one-tailed students t-test ^e Standard deviation

(b) Quantitative Real Time RT-PCR (QRT-PCR) confirmed the microarray expression data, with each of these genes having significantly decreased expression levels in the rgaR mutant. This is also reported in Table 4. The first two genes, CD0588 and CD0590, appeared to be part of an operon which consisted of four ORFs, CD0587-CD0590. Expression of the other genes in the operon also appeared to be down-regulated, but this degree of down-regulation did not meet our stringent statistical significance threshold. BLASTP searches indicated that the putative proteins encoded by this operon had no similarity to any proteins in the GenBank database. The third gene, CD2098, also had no significant matches in the database. The final down-regulated gene, CD2750, had 28% identity to the quorum sensing protein, AgrB, from Staphylococcus aureus, mostly in the N-terminal domain, which is reported to have the highest level of conservation between staphylococcal AgrB homologues (Dufour, Jarraud et al . 2002) . Since the putative C. difficile protein also had the conserved cysteine and histidine residues which are found in putative AgrB proteins from many bacterial genera, and are proposed to be required for the catalytic function of this protein (Qiu, Pei et al . 2005), the protein was provisionally designated as the C. difficile AgrB protein. The putative C. difficile agrB gene appeared to be located in an operon with a putative agrD gene (CD2749A) , which encoded a predicted 45 amino acid protein with an appropriately located cysteine residue, which is known to be critical for the function of the staphylococcal AgrD quorum sensing peptide (Dufour, Jarraud et al . 2002). The small C. difficile agrD gene was not recognised in the original annotation, and therefore was not printed on the microarray slides used in these experiments. The relative expression level of this gene therefore was not determined, but our preliminary evidence indicates that agrB and agrD are transcriptionally linked in C. difficile JIR8094. However, none of the genes in the C. difficile pathogenicity locus, including the tcdA and tcdB genes, which encode toxins A and B, respectively, were up- or down-regulated in either the rgaR or rgbR mutants. Example 5 (a) Analysis of promoter regions of RgaR- regulated genes in C. difficile revealed a conserved binding site with strong similarity to the C. perfringens VirR binding target.

The putative promoter regions of the RgaR- regulated genes were examined for evidence of a consensus binding site. An alignment of the regions upstream of CD0587 (the first gene in the putative CD0587-0590 operon) ,

CD2098 and CD2750 revealed a close match to the C. perfringens P _foA VirR consensus binding site. This is illustrated in Table 5.

Table 5: Alignment of putative C. difficile RgaR binding sites.

Gene 5' -3 ' Sequence

C. perfringens P P_foo_A C CCAGTT ATTCACGATTAAAGC CCAGTT CTGCAC

(SEQ ID NO. 59)

CD0587 T CCAGTT ATGCATTTTCATTGA CCAGTT ACACAG

(SEQ ID NO. 60)

CD2098 A CTAGTT TTACATGTTAAACAA CCAGTT ATGTCA

(SEQ ID NO. 61)

CD2750 (agrB) A CCAGTT TTACATTTTTAACAA CTAGTT TTGTTT

(SEQ ID NO. 62)

CD1667/CD1668 T CTAGTT ATACGTTTTTATTGA CCAGTT ATGCAG intergenic (SEQ ID NO. 63)

CD1511/CD1512 T CCAGTT TTGCGATTATAATAA CCAGTT TTACAT intergenic (SEQ ID NO. 64)

Consensus n CcAGTT nTnCatttttAannA CcAGTT ntgcnn

(SEQ ID NO. 65)

The CD0587 region had a putative VirR binding site in which the core functional nucleotides, CCAGTT (Cheung and Rood 2000; Cheung, Dupuy et al . 2004), were totally conserved in both VirR boxes. The equivalent CD2098 region also had two correctly-spaced VirR boxes, but VirR box I had one change in the conserved region, to CTAGTT, and VirR box II was conserved. Similarly the agrB or CD2750 gene region had a conserved VirR box I sequence and a one base pair change in the conserved region of VirR box II, also to CTAGTT.

Following this discovery, a search of the entire C. difficile genome sequence was undertaken, using a search engine especially written for this purpose. Search parameters were set to identify the VirR-box conserved sequence, CCAGTT (N₁₅) CCAGTT, where N is any nucleotide, and the search allowed a one base-pair mismatch at any of the conserved nucleotides . The results were then analysed to determine whether any other potential VirR boxes were located upstream of potential genes or near likely promoter sequences . This analysis suggested that there were at least two more putative RgaR binding sites in the C. difficile genome, as shown in Table 5. These sites were located in intergenic regions between CD1511/CD1512 and CD1667/CD1668. Although there were no predicted ORFs at these locations, we cannot rule out the possibility that putative RgaR- regulated regulatory RNA molecules may be encoded at these sites. There is a precedent for this suggestion, since the C. perfringens regulatory RNA molecule, VR-RNA, is directly regulated by the binding of VirR to VirR boxes located upstream of its promoter (Shimizu, Ohtani et al. 2002; Cheung, Dupuy et al . 2004) .

(b) The C. difficile RgaR protein binds specifically to the VirR box sequences located upstream of CD0587, CD2098 and agrB. Purified C. difficile RgaR protein was analysed in gel mobility shift assays, using the relevant VirR box regions as the DNA targets. The results, shown in Figure 10, indicated that the RgaR protein was capable of specific binding to all three of the target DNA sequences, appearing to bind most strongly to the VirR box region from CD0587, followed by CD2098 and agrB (CD2750) . This observation was in agreement with the microarray and real-time RT-PCR expression data, which showed that disruption of rgaR had the greatest effect on the expression of the CD0588/CD0590 genes, followed by CD2098 and agrB. On the basis of these results it was concluded that, under the conditions tested, the RgaR protein directly and positively regulates all of these genes .

Example 6 (a) Insertional inactivation of rgaR

C. difficile JIR8094 by double crossover The evidence gathered during our construction and complementation of C. difficile single crossover strains demonstrated that the instability of pJIR1456 in this system could be extended to attempt the construction of double crossover mutants, since we had also established that lincomycin selection could be used to select for the presence of MLS (macrolide-lincosamide-streptogramin) antimicrobial resistance markers. Using this information, we designed a double crossover vector based on pJIR1456, which carried the MLS resistance marker, ermQ. This ermQ marker had previously been tested successfully in our laboratory for its ability to confer lincomycin resistance in C. difficile when located on a shuttle vector based on pJIR1456. A double crossover for rgaR was designed to include genomic DNA from upstream and downstream of the rgaR gene, cloned either side of the ermQ marker, and constructed in such a way that a double crossover between the plasmid and the chromosome would lead to the 5 ' end of the response regulator being expressed and the 3 ' end being deleted entirely. The construct was designed in this way because the C-terminal domain of most response regulator proteins is responsible for the DNA binding/activation activity. Thus deletion of this region of the gene would be expected to make the corresponding protein nonfunctional .

(b) Introduction of a rgaR double crossover plasmid into C. difficile JIR8094. To construct the C. difficile rgaR double crossover suicide plasmid, the double crossover region, including ermQ, was subcloned from a previously-constructed vector, pJIR2755, which was digested with Sphl/Notl, and the 3557 bp fragment encoding the relevant DNA was subcloned into the same sites of the E. coli-C. perfringens shuttle vector, pJIR1456; this resulted in pJIR3011, which is illustrated schematically in Figure 11. The double crossover region included the following C. difficile DNA: upstream of the rgaR gene, including the first 257 bp of the coding region, was PCR amplified using primers JRP1700/1697 (979 bp) and the genomic region downstream of the rgaR gene was amplified with primers JRP1699/1698 (1186 bp) . The PCR product corresponding to the rgaR upstream DNA was cloned upstream of the ermQ gene, and the PCR product corresponding to the rgaR downstream region was cloned downstream of the ermQ gene, this resulted in the "double crossover region" .

The plasmid was analysed by restriction analysis, and found to have the correct profile; the construct was then introduced into the E. coli donor strain HBlOl (pVS520) , which is capable of mobilising plasmids such as pJIR1456 which carry the RP4 oriT region, and the resultant strain was then used to introduce the double crossover construct into C. difficile JIR8094. The conjugative method used was similar to that employed for construction of single crossover strains, with the following exceptions: the ratio of donor and recipient culture volumes used in the matings was 1:1, 1 ml of each donor and recipient culture was centrifuged, as before, then washed twice and resuspended in 1 ml of BHI diluent, then 100 μl volumes of each donor and recipient were plated out together on non-selective media for the mating step, alongside the appropriate controls. The plates were incubated anaerobically at 37° C for 6.5 hours, then resuspended in 1 ml of BHI diluent, and 100 μl volumes were plated on to selective medium containing cefoxitin and lincomycin, to select for C. difficile transconjugants .

These plates were incubated anaerobically at 37° C for six days, then colonies were cross-patched on to selective agar containing either cefoxitin and lincomycin (to select for presence of the double crossover region of the plasmid) or cefoxitin and thiamphenicol (to determine whether the backbone of the suicide vector was still present, indicating that a single crossover had probably occurred) , and the patches were incubated for two days anaerobically at 37 °C. All transconjugants which carried the ermQ lincomycin resistance marker also carried the catP thiamphenicol resistance marker, indicating that a single crossover event had most likely occurred, and that the rgaR gene was probably not insertionally inactivated.

Generation of a C. difficile rgaR double crossover from a single crossover strain.

The C. difficile strains carrying the entire pJIR3011 plasmid were subcultured three times on medium containing cefoxitin and lincomycin (from the primary inoculum of each successive culture) , then colonies from four isolates were inoculated into four separate 20 ml BHIS broths, without any selective pressure, and incubated overnight anaerobically at 37 °C. The next day, 2 ml of each overnight culture was inoculated into a 90 ml BHIS broth and incubated for 7 hours anaerobically at 37° C. Following this, serial 1 in 10 dilutions were carried out, in BHIS broth and 100 μl volumes from dilutions between 10^"3 and 10^"5 were plated out on to half of an agar plate containing selective medium with either cefoxitin and lincomycin, or cefoxitin and thiamphenicol, as well as on non-selective medium. These plates were incubated at 3O⁰C in an anaerobic jar for 3 days, then for a further two days at 37° C in an anaerobic chamber. Colonies growing on the plates containing lincomycin were cross-patched on to plates containing cefoxitin and thiamphenicol or lincomycin. One isolate was found to be lincomycin- resistant and thiamphenicol-sensitive; this was designated JIR8292. Four independent isolates that still carried the catP thiamphenicol resistance marker were also stored, and these were designated JIR8293-JIR8296.

Southern blotting analysis of C. difficile double crossover/rgaR deletant strain

Genomic DNA extracted from JIR8094 and JIR8292 was digested with the Aspl00 restriction endonuclease enzyme, and subjected to Southern blotting analysis. The probes used were DIG- labelled by PCR: the ermQ probe was generated with primers JRP1870/1871, the catP probe was generated as before, and the C. difficile rgaR ORF probe was generated with primers as described in Materials and Methods . The results are shown in Figure 12 , and indicated that in the wild type C. diffiicle strain(JIR8094) , the rgaR gene is located on a 5007 bp DNA fragment, as expected (Figures 9A and D) . In addition the linearised 10,274 bp suicide vector weakly hybridised to the probe, because this plasmid carries 257 bp of the rgaR gene. The JIR8292 mutant strain yielded a fragment of 5940 bp which corresponded to the 257 bp intact region of the rgaR gene and the insertion of ermQ (Figures 9A and D) ; this fragment hybridised weakly to the rgaR probe for the same reason as the suicide vector. In addition, the same fragment hybridised to the ermQ probe, which was also present in the suicide vector but not in the wild type strain, as expected (Figure 12B) . Finally, the loss of the plasmid backbone in the JIR8292 double crossover strain was confirmed by probing the blot with catP, which hybridised to the suicide plasmid, but not to the wild type or mutant C. difficile strains (Figure 9C) .

These results confirmed that JIR8292 was indeed a rgaR mutant, generated by a double crossover homologous recombination event between the suicide vector and the chromosome . Example 7 Construction of tcdA and tcdB mutants of C. difficile

In order to construct the required plasmids for generating disruptions in the C. difficile toxin genes, it was necessary to generate and clone tcdA- and tcdB-specific PCR products . The oligonucleotide primers used for PCR amplification of these fragments were JRP2342 with JRP2343 to amplify tcdA, and JRP2344 with JRP2345 to amplify tcdB. The DNA template was that from strain JIR8094, and both reactions generated products of 567 bp. These DNA products were cloned using the Perfectly Blunt™ Cloning Kit (Novagen) , which generated the tcdA- specific clone pJIR3048 and the tcdB-specific clone pJIR3049.

The ted-specific EcoRl fragments from these clones were subsequently subcloned into the C. perfringens- E. coli shuttle vector pJIR145β. This resulted in the loss of the smaller (approximately 0.8 kb) of the two fragments generated upon EcoRl digestion of pJIR1456. Plasmids pJIR3051 and pJIR3050 resulted from this cloning, and carried the tcdA- and tcdB- specific fragments, respectively. Nucleotide sequencing confirmed that each clone carried the expected tcd-specific fragment.

Plasmids pJIR3050 and pJIR3051 were introduced into strain S17-1 by transformation, after which matings were performed to C. difficile strain JIR8094, selecting for rifampicin and thiamphenicol resistant transconjugants . The mating procedure used was as described above. Putative colonies were subcultured on to the same medium, and then grown in broth with and without selection. DNA was then isolated from these cultures, and Southern blots performed in order to determine if the tcdA and tcdB gene had been disrupted by a single crossover homologous recombination event, incorporating the whole plasmid in the integration site. The probes used for Southern blots were as follows: (1) tcdA internal fragment generated using primers JRP2342 and JRP2343, (2) tcdB internal fragment generated using primers JRP2344 and JRP2345 and

(3) catP fragment generated using primers JRP2142 and JRP2143. The results are shown in Figure 13. This analysis identified two tcdA mutants

(JIR8253 and JIR8263) and five tcdB mutants (JIR8266, JIR8276, JIR8277, JIR8278 and JIR8279) of C. difficile strain JIR8094. Further analysis of the mutants was carried out via. Western blots and cell culture assays. The Western blots, shown in Figure 14, confirmed that strains JIR8253 and JIR8263 no longer carry an intact tcdA gene, since the product from this gene, the TcdA toxin, cannot be detected from supernatants harvested from these strains. Similarly, the cytotoxicity assays, shown in Table 6, confirmed that strains JIR8276 and JIR8278 no longer carry an intact tcdB gene, since the active product from this gene, the TcdB toxin, cannot be detected in supernatants harvested from these strains .

Table 6.

Cytotoxicity of toxins isolated from culture supernatants on Vero cells .

CPE: cytopathic effect The results of these combined assays confirm that the mutants carry inactivated tcdA or tcdB genes. Published research has shown that, of the two toxins, TcdB is about one thousand times more cytotoxic against cultivated cells than TcdA (Burger et al . , 2003). The cytotoxicity assays confirm that strains JIR8276 and JIR8278 have inactivated tcdB genes, since no CPE was observed when supernatant preparations from these strains were added to VERO cells. These data confirm that, for the first time, the C. difficile toxin genes have been successfully inactivated through the use of genetic manipulation, specifically via homologous recombination.

Example 8 Testing of the tcdA and tcdB mutants in an animal model Animal experiments are carried out using the hamster model in order to determine the effect of the gene disruptions in vivo. This model is widely accepted in the art, and is described in numerous publications; see for example Ebright et al, 1981; Borriello et al, 1987; Larson and Borriello, 1990; Larson and Welch, 1993; Sambol et al, 2001. In this model, groups of 10 adult Syrian Golden hamsters per strain are given one dose of clindamycin, in order to clear the normal flora. Five days later, they are given a gastric inoculation of 100 colony forming units of C. difficile spores (Sambol et al . , 2001) . The hamsters are then monitored for disease for up to one month post infection. Further analysis is carried out on animal tissue once the animals are killed.

We are investigating the following questions-. (a) are the mutant strains virulent?

(b) if they are not virulent, can they still colonise the host?

(c) if they are not virulent and can still colonise the host, will a toxin mutant strain serve as a protective vaccine against challenge by other virulent strains?

(d) Does the attenuated strain prevent colonisation with toxigenic isolates, once the vaccine strain has been cleared from the body?

(e) Can it compete out or cure an existing infection? In order to assess this possibility, the animal model as described above is used. To assess the first of these posulates, the hamsters are given a dose of clindamycin to remove normal flora. Five days later they are given a gastric inoculation of C. difficile spores, as above, but these are a mixture of mutant and wild-type C. difficile in a 1:1 ratio. The animals are then monitored to determine the ratio of strains isolated from fecal material obtained from the animals . Finding the mutant strain in significantly larger numbers compared to the wild-type will suggest that it can compete out the wild-type.

To assess the second postulate, the hamsters are given a dose of clindamycin to remove normal flora. Five days later, they are given a dose of the wild-type strain which is then allowed to colonise the animals. Having established an existing infection, the animals are inoculated with the mutant strains and the animals are monitored to determine if the existing infection can be cured. Both of these experiments require comparison to control animals infected with wild-type or mutant alone in order to determine if the presence of the mutant strains has an effect on the wild-type.

Example 9 Use of a C. difficile strain in which toxin

A and B have been inactivated as a vehicle to deliver other peptides to the intestinal mucosa Experiments are conducted to determine whether a

C. difficile strain in which toxin A and B have been inactivated can be used to deliver other peptides to the intestinal mucosa to stimulate mucosal immunity.

An inactivated strain is used to express a small portion of the toxin A or B protein which is immunogenic but not toxic, ie a toxin A/B subunit vaccine, and to deliver it directly to the intestinal mucosa. This gene is expressed from a plasmid location or by being inserted into the C. difficile chromosome via a double crossover homologous recombination event, most likely by replacing the existing toxin-encoding region on the chromosome.

This is will be carried out using a simple in vitro model (Borriello and Barclay, 1986; Drummond et al, 2004) , a triple-stage chemostat model which simulates the human gut (Macfarlane et al, 1998; Freeman et al, 2003; Baines et al, 2005) , or the hamster model referred to in Example 8.

Using the hamster model, following inoculation of hamsters with the modified C. difficile strains, blood is collected and analysed in order to determine if an immune response has been mounted by the animals. The animals are also challenged with the wild-type strain to establish whether they are able to resist infection by this strain, which indicates a good immune response against this organism.

Example 9 Further development of genetic manipulation technology for C difficile

(a) Double crossover mutations in tcdA and tcdB are constructed both in separate strains and together in one strain. This method is expected to be the most likely way to construct a mutant in which both toxin genes are inactivated. Such a strain would then serve as the "delivery platform" or base strain for further vaccine testing.

(b) Unmarked deletions of target genes are constructed. This means that the gene of interest is stably insertionally inactivated or deleted, but no antibiotic resistance cassette is introduced on to the chromosome. Developing this approach is important for two reasons. Firstly, only a limited number of antibiotic selections is available for manipulation of C. difficile, so the construction of unmarked deletions means that we would be able to use a single marker for all our insertional inactivation and complementation experiments . The availability of such unmarked deletions would increase the number of options which could be used for downstream genetic manipulations. Secondly, recombinant strains which are to be used in humans, for example as potential attenuated vaccines, are required by regulatory authorities to have no additional antibiotic resistance genes because these may facilitate the spread of antimicrobial resistance to other bacteria, particularly in a hospital environment, where selective pressure already favours the development of multidrug resistance.

DISCUSSION

In order to overexpress response regulator genes in C. difficile strain JIR8094, we intended to introduce intact copies of the response regulator genes into the parent strain using conjugation. We expected that the plasmid would be maintained as an autonomously replicating, multicopy vector. This principle had been successfully employed during previous work carried out in this laboratory, when a related vector, with the same replication region, was successfully introduced into C. difficile strain CD37, and the plasmid was maintained at a level which enabled it to be recovered for plasmid analysis (Mani, Lyras et al . 2002). Quantitative data from the analysis of these strains indicated that the plasmids were present, and that the genes they carried were being expressed; thus we expected to be able to apply the same principles to analyse response regulator genes in JIR8094. Remarkably, when these plasmids were introduced into C. difficile JIR8094 we found that when there was a region of sequence identity between the plasmid and chromosome, integration of the vector into the chromosome was always detected. This result suggests that, due to a combination of vector instability and selective pressure to maintain the resistance marker, the preferentially selected (or more common) event was integration, rather than autonomous replication of the vector.

This theory is supported by the stability assay results reported in Table 3 for the base vector, pJIR1456, in strain JIR8094 and CD37, which displayed a similar level of instability in both strains in the absence of selection, and indicated that when selection was present and there was no sequence identity between vector and chromosome the vector plasmid was more stably maintained. The previous studies in CD37 involved the introduction of a pIP404 -based vector, pJIR1457, carrying the promoter regions of C. difficile strain VPI10463 toxin genes and the E. coli beta- glucuronidase encoding gene, none of which have any known similarity to the CD37 chromosome; thus in these earlier studies integration did not occur. Analysis of the original C. difficile rgaR overexpression strain, JIR8149, indicated that, as expected, two copies of the response regulator gene on the chromosome resulted in approximately twice as much response regulator transcript when compared to the wild type strain. Although these results suggested that the response regulators were not sufficiently overexpressed to guarantee detection of differential gene expression by microarray analysis, the results of the experiment demonstrated that single crossover events could be readily screened by exploiting the instability of pJIR1456, and that a small, but consistent, increase in the level of response regulator expression could subsequently be detected, suggesting that the integration of the vector was relatively stable in the overall population. Although we were unable to use these strains to investigate the role of rgaR and rgbR in C. difficile, the information gained from these experiments provided proof of the principle that homologous recombination events could be generated in a reproducible, targeted fashion via interaction with the unstable shuttle vector, pJIR1456.

Using this discovery, we were then able to construct rgaR and rgbR mutants by insertional inactivation. Inactivation of rgaR and rgbR was achieved by cloning a short DNA fragment, internal to the response regulator gene, such that a homologous recombination event would lead to an intact 5 ' region and inactivate the 3 ' end of the gene, so that the DNA binding domain was not expressed. Southern blotting of these strains indicated that the response regulator gene was indeed insertionally inactivated, but autonomously replicating vector could still be detected at low levels in some strains . The reconstituted response regulator gene which would result from the * looping out' of this plasmid was only detectable at the PCR level, not by Southern blotting. This result implies that although the plasmid appears to be looping out of the chromosome, this event occurs at low frequency within the population, but once the plasmid began replicating autonomously, it could either continue to exist independently of the chromosome, or could re-integrate. If the plasmid copy number relative to the chromosome was greater than one when it existed autonomously, this would explain the Southern blotting results, which clearly showed the presence of intact shuttle vector and integrated vector in some strains, but not an intact response regulator gene. Since the Southern blotting analysis was carried out on DNA extracted from strains cultured in the absence of selective pressure, if the plasmid was looping out at a high rate, one would expect it to be rapidly lost from the population. If so, one would expect the reconstituted response regulator gene to be detectable in Southern blots, suggesting that the integrated plasmid is relatively stable or maintains a fluid relationship between plasmid and chromosome, such that multiple rounds of insertion and looping out occur, the net result being that the response regulator is inactivated most of the time. This assumption is supported by the stability assay data, which demonstrates that even in the absence of selective pressure, the antibiotic resistance marker carried by the plasmid is maintained in the vast majority of the population, as shown in Table 3.

Transcriptional analysis of the rgaR mutant by microarray and QRT-PCR experiments confirmed that the insertional inactivation of this gene was successful and stable, since a significant and reproducible difference in the expression levels of four genes, CD0588, CD0590, CD2098 and agrB, was detected (Table 4) . In addition, we were able to conclude that these differences in gene expression between the wild type and the mutant strain were a direct consequence of the insertional inactivation of the rgaR gene. This was achieved by the analysis of the genomic region upstream of the down-regulated genes, where putative promoters would be expected to be located. In these ^■ regions regulatory sequences were identified. Very similar regulatory sequences are found in C. perfringens and are known to be bound by the C. perfringens VirR protein, which is 45% identical to its C. difficile RgaR homologue which we insertionally inactivated. Comparison of the regulatory sequences (called 'VirR boxes') found in C. perfringens and C. difficile yielded a putative consensus binding sequence for the C. difficile RgaR protein (Table 5) . DNA binding assays (Figure 7) confirmed that the C. difficile RgaR protein did in fact bind to the C. perfringens VirR boxes, as well as to the C. difficile VirR boxes which we identified from the microarray and QRT-PCR analysis.

The successful construction of the rgaR double crossover /deletant strain is significant, because it provides proof of principle that this next essential step in the development of C. difficile genetic maniptulation systems has been achieved. This new methodology will allow us to construct mutant strains which will not be able to revert to wild-type under any circumstances, regardless of whether selective pressure is applied. This is important, because if the C. difficile mutant strains are to be tested in an animal model for their efficacy as a probiotic or an attenuated vaccine, a strain which cannot revert to wild type will ultimately be required. As a first step toward this aim, single crossover recombination mutants of tcdA and tcdB, the toxin-encoding genes, have already been constructed.

Furthermore, this work now enables us to attempt the construction of strains which carry stable unmarked deletions of genes of interest, e.g. toxin genes. This is important, because it is necessary to minimise the number of antimicrobial resistance markers carried by any bacterial probiotic or attenuated vaccine intended for human consumption, in order to eliminate the possibility that the recombinant bacterial strain will facilitate dissemination of antimicrobial resistance to other bacteria carried by the host .

Thus we have shown that insertional inactivation of genes can be reproducibly and simply achieved in C. difficile, and that complementation analysis can be performed on the resultant mutant strains . This provides a genetic manipulation system for use in C. difficile, so that experiments fundamental to improving our understanding of the virulence of this organism can be carried out.

It will be apparent to the person skilled in the art that while the invention has been described in some detail for the purposes of clarity and understanding, various modifications and alterations to the embodiments and methods described herein may be made without departing from the scope of the inventive concept disclosed in this specification. References cited herein are listed on the following pages, and are incorporated herein by this reference .

REFERENCES

Alsaker, K. V., T. R. Spitzer, et al . (2004). "Transcriptional analysis of spoOA overexpression in Clostridium acetobutylicum and its effect on the cell ' s response to butanol stress." J Bacteriol 186(7): 1959-71.

Altschul, S. F., W. Gish, et al . (1990). "Basic local alignment search tool." J MoI Biol 215(3): 403-10.

Bannam, T. L. and J. I. Rood (1993) . "Clostridium perfringens-Escherichia coli shuttle vectors that carry single antibiotic resistance determinants." Plasmid 29(3): 233-5.

Baines, S. D., Freeman, J., Wilcox, M. H. (2005) . Effects of piperacillin/tazobactam on Clostridium difficile growth and toxin production in a human gut model. J Antimicrob Chemother 55: 974-982

Berryman, D. I., M. Lyristis, and J. I. Rood. 1994. Cloning and sequence analysis of ermQ, the predominant macrolide- lincosamide-streptogramin B resistance gene in Clostridium perfringens. Antimicrobial Agents and Chemotherapy 38:1041- 1046.

Borriello SP, Barclay FE (1986) "An in vitro model of colonisation resistance to Clostridium difficile infection." J. Med. Microbiol 21:299-309

Borriello SP, Ketley JM, Mitchell TJ et al (1987) "Clostridium difficile - a spectrum of virulence and analysis of putative virulence determinants in the hamster model of antibiotic-associated colitis." J. Med. Microbiol 24:53-64

Boyer, H. W. and D. Roulland-Dussoix (1969) . "A complementation analysis of the restriction and modification of DNA in Escherichia coli." J MoI Biol 41(3) : 459-72.

Carter, G. P., D. Purdy, et al. (2005) . "Quorum sensing in Clostridium difficile: analysis of a luxS-type signalling system." J Med Microbiol 54 (Pt 2): 119-27.

Cheung, J. K., B. Dupuy, et al . (2004). "The spatial organization of the VirR boxes is critical for VirR- mediated expression of the perfringolysin 0 gene, pfoA, from Clostridium perfringens . " J Bacteriol 186(11): 3321- 30.

Cheung, J. K. and J. I. Rood (2000) . "The VirR response regulator from Clostridium perfringens binds independently to two imperfect direct repeats located upstream of the pfoA promoter ." J Bacteriol 182(1): 57-66.

Drummond LJ, Smith DEG, Poxton IR. (2004) "Effects of sub- MIC concentrations of antibiotics on growth of and toxin production by Clostridium difficile." J. Med. Microbiol 52:1033-8

Dufour, P., S. Jarraud, et al . (2002). "High Genetic

Variability of the agr Locus in Staphylococcus Species." J. Bacteriol. 184(4): 1180-1186.

Dupuy, B. and A. L. Sonenshein (1998) . "Regulated transcription of Clostridium difficile toxin genes." MoI Microbiol 27(1): 107-20.

Dupuy, B., and Sonenshein, A. L. (1998) Regulated transcription of Clostridium difficile toxin genes. MoI Microbiol 27: 107-120.

Ebright J. R., Fekety R., Silva J. et al . (1981) "Evaluation of eight cephalosporins in hamster colitis model." Antimicrobial Agents and Chemotherapy 19:980-6

Freeman, J., Baines, S. D., Jabes, D., Wilcox, M. H. (2005) . Comparison of the efficacy of ramoplanin and vancomycin in both in vitro and in vivo models of clindamycin-induced Clostridium difficile infection. J Antimicrob Chemother 56: 717-725

Freeman, J., O'Neill, F. J., Wilcox, M. H. (2003) "The effects of cefotaxime and desacetylcefotaxime upon Clostridium difficile proliferation and toxin production in a triple-stage chemostat model of the human gut." J Antimicrob Chemother 52: 96-102

Goding, J. W. (1996) . Monoclonal antibodies : principles and practice : production and application of monoclonal antibodies in cell biology, biochemistry and immunology. London, Academic.

Graves, D. J., B. L. Martin, et al . (1994). Co- and post- translational modification of proteins : chemical principles and biological effects, Oxford University Press Inc, USA.

Haraldsen, J.D. , and Sonenshein, A. L. (2003) Efficient sporulation in Clostridium difficile requires disruption of the sigmaK gene. MoI Microbiol 48: 811-821.

Harris, L. M., N. E. Welker, et al . (2002). "Northern, morphological, and fermentation analysis of spoOA inactivation and overexpression in Clostridium acetobutylicum ATCC 824." J Bacterid 184(13): 3586-97.

Herbert, M., T. A. O'Keeffe, et al . (2003). "Gene transfer into Clostridium difficile CD630 and characterisation of its methylase genes." FEMS Microbiol Lett 229(1): 103-10. Hinds, J., A., A. Witney, et al . (2002). Microarray design for bacterial genomes. Methods in Microbiology: Functional Microbial Genomics. B. W. Wren and N. Dorrell. London, Academic Press: 67-82.

Hinds, J., K. G. Laing, et al . (2002). Glass slide microarrays for bacterial genomes. Methods in Microbiology: Functional Microbial Genomics. B. W. Wren and N. Dorrell. London, Academic Press: 83-99.

Inoue, H., H. Nojima, et al . (1990). "High efficiency transformation of Escherichia coli with plasmids . " Gene 96 (1) : 23-8.

Kamiya, S., H. Ogura, et al . (1992). "Correlation between cytotoxin production and sporulation in Clostridium difficile." J Med Microbiol 37(3): 206-10.

Kennan, R. M., 0. P. Dhungyel, et al . (2001). "The type IV fimbrial subunit gene (fimA) of Dichelobacter nodosus is essential for virulence, protease secretion, and natural competence." J Bacterid 183(15): 4451-8.

Ketley, J. M., T. J. Mitchell, et al. (1986). "Sporogenesis and toxin A production by Clostridium difficile ." J Med Microbiol 22(1): 33-8.

Larson HE, Borriello SP. (1990) "Quantitative study of antibiotic-induced susceptibility to Clostridium difficile enterocecitis in hamsters." Anti microb Agents Chemother 34:1348-53

Larson HE, and Welch A. (1993) "In-vitro and in-vivo characterisation of resistance to colonisation with Clostridium difficile." Journal of Medical Microbiology 38:103-8 Liyanage, H., S. Kashket, et al . (2001). "Clostridium beijerinckii and Clostridium difficile detoxify methylglyoxal by a novel mechanism involving glycerol dehydrogenase." Appl Environ Microbiol 67(5): 2004-10.

Lundblad, R. L. (1995) . Techniques in protein modification. Boca Raton, Fl. USA, CRC Press.

Lyras, D. and J. I. Rood (1998) . "Conjugative transfer of RP4-oriT shuttle vectors from Escherichia coli to Clostridium perfringens . " Plasmid 39(2): 160-4.

Lyras, D. and J. I. Rood (2000) . "Transposition of Tn4451 and Tn4453 involves a circular intermediate that forms a promoter for the large resolvase, TnpX." MoI Microbiol 38 (3) : 588-601.

Lyristis, M., A. E. Bryant, et al . (1994). "Identification and molecular analysis of a locus that regulates extracellular toxin production in Clostridium perfringens . " MoI Microbiol 12(5): 761-77.

Macfarlane GT, Macfarlane S, Gibson GR. (1998) "Validation of a three-stage compound continuous culture system for investigating the effect of retention time on the ecology and metabolism of bacteria in the human colon." Microb Ecol 35:180-7

Madigan, M. T., J. Martinko, M, et al . (2003). Brock,

Biology of Microorganisms. Upper Saddle River, USA, Pearson Education Inc .

Mani, N., and Dupuy, B. (2001) Regulation of toxin synthesis in Clostridium difficile by an alternative RNA polymerase sigma factor. Proc Natl Acad Sci U S A 98: 5844- 5849. Mani, N., Lyras, D., Barroso, L., Howarth, P., Wilkins, T., Rood, J.I., Sonenshein, A. L., and Dupuy, B. (2002) Environmental response and autoregulation of Clostridium difficile TxeR, a sigma factor for toxin gene expression. J Bacterid 184: 5971-5978.

Miroux, B. and J. E. Walker (1996) . "Over-production of Proteins in Escherichia coli: Mutant Hosts that Allow Synthesis of some Membrane Proteins and Globular Proteins at High Levels." Journal of Molecular Biology 260(3): 289- 298.

Mullany, P., M. Wilks, et al , (1994). "Gene cloning in Clostridium difficile using Tn916 as a shuttle conjugative transposon." Plasmid 31(3): 320-3.

Palombo, E. A., K. Yusoff, et al . (1989). "Cloning and genetic analysis of tra cistrons of the Tra 2/Tra 3 region of plasmid RPl." Plasmid 22(1): 59-69.

Pospiech, A. and B. Neumann (1995) . "A versatile quick-prep of genomic DNA from gram-positive bacteria." Trends Genet 11(6) : 217-8.

Purdy, D., T. A. O'Keeffe, et al . (2002). "Conjugative transfer of clostridial shuttle vectors from Escherichia coli to Clostridium difficile through circumvention of the restriction barrier." MoI Microbiol 46(2): 439-52.

Qiu, R., W. Pei, et al . (2005). "Identification of the putative staphylococcal AgrB catalytic residues involving the proteolytic cleavage of AgrD to generate autoinducing peptide." J Biol Chem 280(17): 16695-704.

Sambol, S. P., J. K. Tang, M. M. Merrigan, S. Johnson, and D. N. Gerding. 2001. Infection of hamsters with epidemiologically important strains of Clostridium difficile. Journal of Infectious Diseases 183:1760-1766.

Sambrook, J., E. F. Fritsch, et al . (1989). Molecular cloning : a laboratory manual. Cold Spring Harbor, N. Y. , Cold Spring Harbor Laboratory Press .

Sambrook, J. and D. W. Russell (2001) . Molecular cloning: a laboratory manual. Cold Spring Harbor, N. Y., Cold Spring Harbor Laboratory Press.

Shimizu, T., K. Ohtani, et al. (2002). "Complete genome sequence of Clostridium perfringens, an anaerobic flesh- eater." Proc Natl Acad Sci U S A 99(2) : 996-1001.

Shuler, M. L. and F. Kargi (1992) . Bioprocess Engineering: Basic Concepts. Englewood Cliffs, New Jersey, Prentice-Hall International .

Simon, R., Priefer, U. and Puhler, A (1983). "A broad host range mobilization system for genetic engineering: transposon mutagenesis in gram negative bacteria." Biotechnology 1: 27-45.

Smith, C. J., S. M. Markowitz, et al . (1981). "Transferable tetracycline resistance in Clostridium difficile ." Antimicrob Agents Chemother 19(6): 997-1003.

Smyth, G. K. (2004) . "Linear models and empirical Bayes methods for assessing differential expression in microarray experiments." Statistical Applications in Genetics and Molecular Biology 3 (1) : Article 3.

Smyth, G. K., J. Michaud, et al . (2005). "Use of within- array replicate spots for assessing differential expression in microarray experiments." Bioinformatics 21(9): 2067-75. Tang-Feldman, Y. J., G. Ackermann, et al . (2002). "One-step cloning and expression of Clostridium difficile toxin B gene (tcdB) . " MoI Cell Probes 16(3): 179-83.

Wust, J. and U. Hardegger (1983) . "Transferable resistance to clindamycin, erythromycin, and tetracycline in Clostridium difficile." Antimicrob Agents Chemother 23(5): 784-6.

Claims

1. A method of inactivating a target gene in a Clostridium difficile host cell, comprising the steps of (i) introducing a plasmid which is unstable in C. difficile into the host cell,

(ii) exposing the host cell to conditions which select for cells expressing the selectable marker, thereby providing sufficient time to permit homologous recombination between the C. difficile DNA in the plasmid and C. difficile DNA in the host cell to occur, and (iii) removing the selective conditions, in which the plasmid comprises

(b) a gene encoding a selectable marker which functions in both a Clostridium species and in the non- clostridial donor species,- (c) an origin of replication which is functional in the non-clostridial donor species;

(d) an origin of replication which is functional in a Clostridium species; and

2. A method of inactivation of a target gene in a C. difficile host cell and introducing a nucleic acid molecule of interest into the cell, comprising the steps of

(i) introducing a plasmid which is unstable in C. difficile into the host cell,

(ii) exposing the host cell to selective conditions which select for cells expressing the selectable marker, thereby- providing sufficient time to permit homologous recombination between the C. difficile DNA in the plasmid and C. difficile DNA in the host cell to occur, and (iii) removing the selective conditions, in which the plasmid comprises - Ill -

(d) an origin of replication which is functional in a Clostridium species; e) a fragment of C. difficile DNA which is homologous to a region of DNA in the host cell; and f) a nucleic acid molecule of interest.

3. A method of introducing a nucleic acid molecule of interest into a C. difficile host cell, comprising the steps of

(i) introducing a plasmid into the host cell, (ii) exposing the host cell to selective conditions which select for cells expressing the selectable marker, thereby providing sufficient time to permit homologous recombination between the C. difficile DNA in the plasmid and C. difficile DNA in the host cell to occur, and (iii) removing the selective conditions, in which the plasmid is unstable in C. difficile, and comprises

(d) an origin of replication which is functional in a Clostridium species; e) a fragment of C- difficile DNA which is homologous to a region of non-coding DNA or a repeated region of DNA in the host cell; and f) the nucleic acid molecule of interest.

4. A method according to claim 3, in which the non- coding DNA encodes a ribosomal RNA gene . B. A method according to any one of claims 2 to 4, in which the nucleic acid molecule of interest is homologous .

6. A method according to any one of claims 2 to 4, in which the nucleic acid molecule of interest is heterologous.

7. A method according to any one of claims 1 to 6, in which the origin of conjugative transfer is capable of modulating conjugative transfer of the plasmid from Escherichia coli into a clostridial species. 8. A method according to any one of claims 1 to 7, in which the gene encoding a selectable marker is a gene which functions in both clostridial hosts and in Escherichia coli.

9. A method according to any one of claims 1 to 8 , in which the origin of replication which is functional in the non-clostridial donor species is functional in Escherichia coli.

10. A method according to any one of claims 1 to 9, in which the origin of replication which is functional in a Clostridium species is functional in Clostridium perfringens.

11. A method according to any one of claims 2 to 10, in which the heterologous nucleic acid molecule encodes a peptide, polypeptide or protein, selected from the group consisting of antibodies or fragments thereof; antibodies or fragments thereof coupled to toxins of non- C. difficile origin; peptide or polypeptide ligands, immunomodulatory agents; cytokines; peptide or polypeptide antibiotics or bacteriocins,- peptide or polypeptide anti-cancer agents; antigens of bacterial, viral or cell surface origin; peptide antigens or epitopes; enzymes; regulatory factors; and secretory leader sequences encoding a signal peptide from an exported clostridial N-acetyl-muramoyl-L- alanineamidase-like protein.

12. A method according to claim 11, in which the regulatory factor modulates expression of a non-polypeptide component of the C. difficile cell.

13. A method according to any one of claims 2 to 2 , in which the nucleic acid molecule of interest is under the control of a promoter of prokaryotic origin.

14. A method according to claim 13 , in which the promoter is one which is functional in a Clostridium species .

15. A method according to claim 12 or claim 13, in which the promoter is the tnpX promoter from Clostridium perfringens . 16. A method according to any one of claims 1 to 15, in which the plasmid also comprises a plasmid replication sequence and/or a multiple cloning site

17. A method according to any one of claims 1 to 16, in which the origin of conjugative transfer is RP4 oriT. 18. A method according to any one of claims 1 to 17, in which the gene which is inactivated is one which encodes a protein involved in virulence of C. difficile.

19. A method according to claim 18, in which the gene which is inactivated encodes toxin A, toxin B or a factor which regulates the expression of one or both of these toxins .

20. A method according to claim 19, in which the factor is selected from the group consisting of promoters of the genes for toxin A and toxin B, sigma factors and anti-sigma factors.

21. A method according to any one of claims 18 to 20, in which virulence of the host cell is reduced or abolished.

22. A method according to any one of claims 1 to 21, in which the selectable marker is antibiotic resistance.

23. A method according to any one of claims 1 to 22, in which the host cell is subsequently recovered, and/or subjected to further manipulation.

24. A C. difficile cell in which a target gene has been subjected to insertional inactivation by homologous recombination, or a progeny cell, mutant or derivative of the cell.

25. A cell according to claim 24, in which the gene which is inactivated is one which encodes a protein involved in virulence of C. difficile.

26. A cell according to claim 25, in which the gene which is inactivated encodes toxin A, toxin B or a factor which regulates the expression of one or both of these toxins into which a nucleic acid molecule of interest has been introduced.

27. A cell according to claim 26, in which the factor is selected from the group consisting of promoters of the genes for toxin A and toxin B, sigma factors and anti-sigma factors .

28. A cell according to any one of claims 24 to 27, in which virulence of the host cell is reduced or abolished.

29. A C. difficile cell produced by a method according to any one of claims 1 to 23.

30. , A mobilisable conjugative recombination vector for gene inactivation in C. difficile, comprising (a) an origin of conjugative transfer capable of modulating conjugative transfer of the plasmid from a non- clostridial donor species into a clostridial species;

(c) an origin of replication which is functional in the non-clostridial donor species,-

(d) an origin of replication which is functional in a Clostridium species; (e) at least a fragment of C. difficile RgaR and

RgbR response regulator DNA

31. A conjugative recombination vector according to claim 30, in which the nucleic acid segment which prevents read- through expression from the vector is an Ω transcriptional terminator cassette.

32. A conjugative recombination vector according to claim 30 or claim 31, in which the selectable marker is a macrolide-lincosamide-streptogramin antimicrobial resistance marker.

33. A conjugative recombination vector according to any one of claims 30 to 33, in which the macrolide- lincosamide-streptogramin resistance marker is ermQ.

34. A conjugative recombination vector according to claim 30, which is a plasmid selected from the group consisting of pJIR2512, pJIR2515, pJIR2363, pJIR2364, pJIR2633, pJIR2634, pJIR3012, pJIR3013, pJIR3014 and pJIR3015.

35. A complementation vector for C. difficile, comprising

(b) a gene encoding a selectable marker which functions in both a Clostridium species and in the non- clostridial donor species,-

(d) an origin of replication which is functional in a Clostridium species,-

(e) a C. difficile RgaR or RgbR response regulator DNA, or a fragment thereof;

36. A vector according to claim 35, which is plasmid PJIR34041 or pJIR34042.

37. A vector according to any one of claims 30 to 36, in which the origin of conjugative transfer is capable of modulating conjugative transfer of the plasmid from Escherichia coli into a clostridial species. 38. A vector according to any one of claims 30 to 37, in which the gene encoding a selectable marker is a gene which functions in both clostridial hosts and in Escherichia coli.

39. A vector according to any one of claims 30 to 38, in which the origin of replication which is functional in the non-clostridial donor species is functional in Escherichia coli.

40. A vector according to any one of claims 30 to 39, in which the origin of replication which is functional in a Clostridium species is functional in Clostridium perfringens .

41. A double crossover suicide vector for C. difficile, comprising

(a) a region of DNA upstream of a target gene; (b) a region of DNA downstream of a target gene;

(c) a first antibiotic resistance marker which selects for integration into the chromosome, and

(d) a second antibiotic resistance marker which selects against single crossovers and independently replicating plasmids.

42. A suicide vector according to claim 41, in which the target gene is rgaR.

43. A suicide vector according to claim 42, in which the region upstream of the rgaR gene comprises the first 257 bp of the coding region.

44. A suicide vector according to any one of claims 41 to 43, in which the first antibiotic resistance marker is an ermQ erythromycin resistance cassette.

45. A suicide vector according to any one of claims 41 to 43, in which the second antibiotic resistance marker is catP.

46. A suicide vector according to any one of claims 41 to 43, which is pJIR3011,

47. A recombinant C. difficile strain transformed with or comprising a vector according to any one of claims

30 to 45. 48. A recombinant C. difficile strain according to claim 47, selected from the group consisting of JIR8149,

JTR8150, JIR8218, JIR88223, JIR8226,

JIR8145: JIR8094 (pJIR1456) , JIR8149 : JIR8094 (pJIR2634) ,

JIR8150:JIR8094 (pJIR2633) , JIR8149 : JIR8094 (pJIR2634) , JIR8150: JIR8094 (pJIR2633 ) , JIR8218 : JIR8094ΩpJIR3014 ,

JIR8223: JIR8094ΩpJIR3015 , JIR8226: JIR8094ΩpJIR3015 ,

JIR8234-.JIR8218, JIR8236 : JIR8218 , JIR8243 : JIR8223 ,

JIR8233: JIR8223, JIR8245 : JIR8226 , JIR8247 : JIR8226 , and

JIR8292. 49. A composition comprising a C. difficile cell according to any one of claims 25 to 29, 47 or 48, together with a pharmaceutically-acceptable carrier.

50. A composition according to claim 49, in which virulence of the C. difficile is reduced or abolished. 51. A composition according to claim 50, which is a vaccine composition.

52. A composition according to claim 51, further comprising an adjuvant.

53. A composition according to claim 52, which is a probiotic composition.

54. An isolated C. difficile response regulator nucleic acid sequence, comprising

(a) the nucleic acid sequence set out in SEQ ID NO: 57;

(b) a nucleic acid molecule which is able to hybridise to (c) under stringent conditions; and

(d) a nucleic acid molecule which has at least 75% sequence identity to (a) .

55. An isolated C. difficile response regulator nucleic acid sequence comprising (a) the nucleic acid sequence set out in SEQ ID NO: 58;

(b) a nucleic acid molecule which is able to hybridise to

(c) under stringent conditions; and (d) a nucleic acid molecule which has at least 75% sequence identity to (a) .

56. An isolated nucleic acid molecule comprising

(a) a VirR box region selected from the group consisting of CD0587, CD2098 agrrB (CD2750) , CD1667/CD1668 intergenic,

CD1511/CD1512 intergenic and nCcAGTTnTnCatttttAannACcAGTTntgcnn, which is capable of specific binding to a VirR protein, or

(d) a nucleic acid molecule which has at least 75% sequence identity to (a) .

57. An isolated protein encoded by a nucleic acid according to claim 54. 58. An isolated protein encoded by a nucleic acid according to claim 55.

59. An isolated C. difficile protein which is down- regulated by disruption of virR, and which is encoded by a gene selected from the group consisting of CD0588, CD0590, CD2098 and CD2750, or a mutant or derivative of the gene.

60. An isolated C. difficile AgrB quorum sensing protein encoded by CD2750, or by a mutant or derivative thereof .

61. An isolated C. difficile operon comprising open reading frames CD0587-CD0590.

62. An isolated antibody specific for a C. difficile cell according to any one of claims 25 to 29, 47 or 48, or for a protein according to according to any one of claims 57 to 60. 63. A method of assessing a parameter selected from the group consisting of host immune response, modification of a bacterial response, virulence, gene activation and gene complementation, comprising the use of a C. difficile cell according to any one of claims 25 to 29, 47 or 48. 64. A method of treatment or prophylaxis of a condition associated with C. difficile, comprising the step of administering an effective dose of a composition according to any one of claims 50 to 53 to a subject suffering from or at risk of the condition. 65. Use of a composition according to any one of claims 50 to 53 in the manufacture of a medicament for the treatment or prophylaxis of a condition associated with C. difficile.