WO2001077319A2 - Genetic manipulation of clostridium difficile - Google Patents

Abstract

A plasmid for transformation of C. difficile is described, together with its use in transformation of C. difficile and also a C. difficile replicon, a C. difficile replication factor and a C. difficile origin of replication. Methods of expression of genes in C. difficile use the plasmid.

Description

Genetic manipulation of Clostridium difficile

The present invention relates to genetic manipulation of Clostridia, in particular genetic manipulation of including transformation of and vectors for transformation of C. difficile.

C. difficile is the commonest cause of nosocomial diarrhoea in UK and elsewhere, with C. diffic/7e-assoc\ated disease (CDAD) accounting for up to 15% of all diarrhoeal disease associated with antibiotic treatment. CDAD is primarily a disease of the elderly, and is increasing in frequency concomitant with an ageing UK population. Thus, over 80% of cases occur in the over 65s and the organism is now by far the commonest enteric pathogen isolated from such individuals. At particular risk are elderly people in hospitals, nursing homes, and other chronic-care facilities, where CDAD outbreaks can be devastating, both in terms of mortality (e.g., the 1991/2 Manchester hospitals' outbreak caused the death of 17 elderly patients and contributed to the death of at least 43 others) and the cost of disease management (e.g., disruption to services, patient isolation to a separate wards, 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. In more serious CDAD, infection can be cleared by the use of oral vancomycin or metronidazole. However, CDAD will recur in one in four patients. Alternative therapies are clearly required. One approach would be to devise a CDAD vaccine. However, whilst some studies have been undertaken on the use of toxoided toxin A, or fragments thereof, its effectiveness, particularly with regard to preventing diarrhoea (a major contributor to disease management and patient/environmental contamination), is debatable.

A major stumbling block to the formulation of effective forms of CDAD therapies is a general lack of understanding of the pathogenesis of C. difficile infection, and advances will not be achieved until a more detailed understanding of the molecular basis of virulence in C. diffici/e s attained.

Gene systems require two essential components, namely (i) a plasmid vehicle capable of being maintained, and selected, within the organism, and (ii) a means of introducing said vehicle into the cell. Plasmid vehicles are most often based on autonomous replicating episomes, most usually a plasmid but also phages, or hybrids thereof. Such elements are ideally sourced from the organism under investigation. Whilst a number of epidemiological studies have noted the presence of plasmids in clinical isolates of C. difficile no plasmid has yet to be characterised except with regard to crude estimates of their molecular weight.

There have been numerous attempts to devise effective means of introducing recombinant vectors into C. difficile. Despite these efforts, the only system developed to date is an extremely inefficient and cumbersome procedure based on the Gram-positive conjugativetransposon, Tn£7έ> which provides a means, albeit cumbersome, of introducing cloned fragments back into C. difficile. The frequency of transfer is, however, extremely low (frequency of transfer of 10^"8 per donor). Clearly, both improvements on this procedure, as well as the development of alternatives means of introducing DNA into C. difficile are required.

It is an object of the present invention to provide means for genetic manipulation of C. diffici/eso that heterologous gene coding sequences can be expressed in C. difficile. A specific object of the present invention to provide means for transformation of C. difficile. Another object is to enable investigation into C. difficile virulence. A further object is to solve or at least ameliorate the problems identified in the art.

The present invention is based upon and around isolation of a C. difficile plasmid, designated pCD6 (6.8 kb), its characterisation through the determination of its entire nucleotide sequence and the identification of those regions of the plasmid required for replication in clostridia, its introduction into C. difficile and expression of coding sequences therefrom.

Accordingly, the present invention provides a plasmid for transformation of C. difficile.

A plasmid described in more detail below, for expression of a heterologous gene in C. difficile, comprises a C. difficile replicon and a restriction endonuclease site to receive the heterologous coding sequence. In use the gene sequence is inserted into the plasmid and may be used for transformation of C. difficile. The plasmid of particular embodiments of the invention is native to C. difficile, that is to say it is obtained from C. difficile, or is derived therefrom by recombinant DNA techniques.

Typically, the replicon comprises an origin of replication and a sequence coding for a replication protein that binds to the origin of replication and enables replication of the plasmid by C. difficile. A plasmid of the invention may comprise the whole ori (origin of replication) and a whole rep (replication protein) or functional fragments thereof, and hence plasmids for transformation of host C. difficile may contain an ori plus a repA gene, an ori plus another rep gene, a part of an ori plus a rep or a part of an ori and part of a rep.

The replicon thus comprises everything that is needed for replication of the plasmid by C. difficile. Typically, plasmids are cloned in a host, such as E. coli, and also comprise a replicon for the host cell, such as an E. coli replicon.

A further aspect of the invention provides a vector for expression of a gene sequence in C. difficile comprising the gene sequence, or a restriction site into which the gene sequence can be inserted, and characterized in that the vector is not digested by C. difficile restriction enzymes.

A method of expressing a particular gene in C. difficile, hence comprises providing a plasmid or vector as described, and introducing this into C. difficile. The plasmid or vector generally includes a selectable marker and the method can comprise selecting for bacteria that express the selectable marker in order to identify bacteria that express the particular gene.

The invention thus advantageously enables expression of heterologous genes in C. difficile. The invention further provides a method of making a plasmid for expression of a heterologous coding sequence in C. difficile, comprising providing a plasmid that is not digested by C. difficile restriction enzymes, said plasmid comprising a C. difficile replicon, and inserting said heterologous coding sequence into the plasmid. The heterologous sequence can be a nucleotide sequence encoding a polypeptide or can also be an anti-sense sequence, designed to hybridize to a C. difficile genetic sequence, such as a mRNA sequence.

A further method of the invention lies in a method of making a plasmid for expression of a heterologous coding sequence in C. difficile, comprising providing a plasmid wherein said plasmid comprises a C. difficile replicon, and said heterologous coding sequence, and subjecting said plasmid to methylation so as to prevent digestion of said plasmid by C. difficile restriction enzymes.

A key to these aspects of the invention is that C. difficile contains enzymes which will potentially destroy a plasmid intended to achieve the transformation, and hence the inventors have managed to devise protection for the plasmids of the invention, enabling those plasmids to avoid destruction by the C. difficile enzymes. The invention has managed to identify those enzymes that methylate sequences in C. difficile. Those methylases are cloned into E. coli, and plasmids prepared in those E. coli are methylated by the cloned enzymes, preventing digestion of the plasmid when introduced into the C. difficile host. In a specific embodiment of the invention, four C. difficile methylase genes are cloned into E. coli, a plasmid is replicated therein and used to transform C. difficile.

A feature of plasmids of the invention is their replicon, and a further aspect of the invention provides a C. difficile replicon.

A typical replicon has a first DNA sequence comprising a C. difficile origin of replication and a second DNA sequence comprising a sequence encoding a replication protein. In use, the origin of replication binds to or is otherwise associated with the replication protein and hence enables replication in C. difficile of a plasmid containing that origin of replication. The invention additionally provides a C. difficile repWcaWon factor and a C. difficile origin of replication. The replication factor and origin of replication are, in a specific embodiment obtained from a plasmid native to C. difficile, or derived therefrom.

A method of expressing a gene in C. difficile is provided in the present invention, comprising making a plasmid containing that gene and transforming C. difficile with the plasmid.

Vectors can also be introduced into a host by other means. For example, it is possible to takes the transformation vector of the invention, such as a vector based on pCD6, and give it an additional fragment of DNA carrying a "origin of transfer" (an example is oriT). The resulting vector can be mobilised into C. difficile from a suitable donor strain. Mobilisation refers to the process when a conjugative element (eg., a large plasmid or a conjugal transposon, such as Tn916) transfers not only itself, but also any autonomous element in the same cell which carries an oriT region. These large conjugal elements typically encode proteins which act at oriT to bring about mobilisation.

Hence a method of introducing a gene into C. difficile includes devising a pCD6-based plasmid, adding an origin of transfer and using the vector obtained for conjugation. The vector can suitably be introduced into B. subtilis carrying Tn916, and then filter mating with C. difficile earned out and the plasmid transferred along with Tn916. Once in the cell it will replicate autonomously. One advantage of this is that when plasmids are transferred from one cell to another during conjugation they do so in a single stranded DNA form. ssDNA cannot be cut by restriction enzymes, so this offers a further means to get around the problem of restriction. The plasmids of the invention therefore also optionally comprise an origin of transfer.

Other factors aside from a replication protein and an origin can be important in vector design. Thus other optional factors may influence replication efficiency and maintenance (stability) of the plasmid. In the case of pCD6, for instance, we have shown in a specific embodiment of the invention that a gene upstream of the replication protein (orfB) can make the shuttle vectors more efficient, in that colonies arise 24 hrs after transformation of clostridia compared to 48 hrs when only the RepA and the ori are present. Hence, the invention also provides a gene or gene sequence which contributes to efficiency of transformation of C. difficile, together with a plasmid containing that gene or gene sequence. A specific example of such a sequence is orfB.

Naturally occurring antisense RNAs are small, untranslated transcripts that pair with target RNA through regions of complementarity to prevent biological function. Artificial antisense molecules have proven to be spectacularly successful in eucaryotic systems, most notably in the generation of tomatoes with a longer shelf life through inhibition of the degradative enzyme polygalacturonase.

The feasibility of using antisense technology to reduce enzyme expression in procaryotes was demonstrated many years ago. The strategy has not, however, been widely adopted in pathogenic bacteria until recently when it has been employed to bring about a 16-fold reduction in toxin production in both Staphylococcus aureus and C. botulinum, and to create histidine auxotroph in Mycobacterium smegmatis. A further element of this invention is therefore to use specifically constructed conjugal cointegrate vectors which may be used to bring about the inactivation of specific genes through the delivery of appropriate anti-sense DNA molecules.

Accordingly, a still further aspect of the invention lies in a method of identifying a vector that integrates into a gram positive bacterial genome, comprising transforming a gram positive bacteria with a plasmid, wherein the plasmid comprises an inducible promoter and replication of the plasmid is dependent upon presence of an inducer of the promoter, wherein the plasmid includes a sequence coding for a selectable marker, and wherein transformation takes place in the presence of the inducer, removing the inducer, and selecting for bacteria expressing the selectable marker. A further such method comprises transforming a gram positive bacteria with a plasmid, wherein the plasmid comprises a suppressible promoter and replication of the plasmid is dependent upon absence of a suppressor of the promoter, wherein the plasmid includes a sequence coding for a selectable marker, and wherein transformation takes place in the absence of the suppressor, adding the suppressor, and selecting for bacteria expressing the selectable marker. Hence, by turning off extrachromosomal replication of the plasmid in this way it is possible to determine which plasmids have integrated - as only those integrating will express the selectable marker enabling their selection at a later stage.

The methods are suitable for identification of a vector that integrates into Clostridia.

Gene inactivation may thus be achieved through the development of effective integrative vectors. The ability to reproducibly direct DNA to the genomes of clostridial species has generally proven to be problematic, particularly with regard to transposon (Tn) delivery. In other bacteria, the most effective Tn- delivery vehicles are conditional for replication, where transposition may be selected by the imposition of non-permissive conditions following the successful introduction of the Tn-bearing plasmid. Plasmids which are temperature sensitive (ts) for replication have found particularly widespread use. Inexplicably, no Gram-positive ts plasmid capable of replication in clostridia has been found to function effectively. A further aspect of this invention is therefore to derive plasmids in which replication is dependent on the addition of an exogenous inducer. The plasmids may therefore only replicate in the presence of the inducer. Upon its removal, the plasmids are unable to replicate. Antibiotic resistance markers on said plasmid is therefore be lost from the cell unless the plasmid integrates into the host genome. Such integration would be brought about through the insertion into the plasmid of a DNA fragment derived from the host genome. This allows the plasmid to integrate through homologous recombination. Furthermore, through the use of fragments which incorporate parts, or mutated derivatives of genes, the integration event leads to the inactivation of gene at which insertion occurs.

Whilst an anti-sense sequence may be used to specifically inactivate expression of an individual gene, its inactivation may lead to affects on the expression of other genes. This is particularly true of regulatory genes the products of which either directly, or indirectly, control the expression of another gene or genes. This opens up the possibility of deliberately targeting a regulatory gene with the explicit purpose of identifying those genes which are under its regulation. Thus, in pathogenic bacteria, the production of toxins and ancillary virulence factors is co-ordinately controlled at the genetic level through the participation of global regulatory systems. These systems allow the bacteria to respond to environmental stimuli both individually and in a co-ordinated manner via cell-to-cell signalling mechanisms. An essential part of this coordinated response are the so-called two-component signal transduction systems, which comprise: a sensor protein and its cognate response regulator. The former traverses the cell membrane from where, following its interaction with specific environmental stimuli, it transmits receipt of the external signal to the internally located response regulator. The response regulator is a transcriptional factor, which then proceeds to switch on expression of discrete sets of virulence genes (Fig 1 ). The participation of dedicated transcriptional factors in the regulation of virulence provides an indirect route to the identification of the virulence genes themselves. Thus, strains in which the gene encoding such a transcriptional factor is mutated

(i.e., by insertional inactivation of the gene or through inhibition of mRNA translation via anti-sense RNA) display down regulation of virulence genes. This effect is visualised through proteome analysis (Fig. 1 ). Conversely, enhanced expression of such a transcriptional factor leads to up regulation of virulence genes.

Although there is no direct evidence that two-component signal transduction systems operate in C. difficile, they undoubtedly exist. It is clear, for instance, that toxin production is growth phase dependent, with toxin production/ gene expression occurring at high cell densities late in the growth cycle. In Gram-negative bacteria it is now well established that many sets of genes, including virulence factors, are regulated in a cell- density-dependent manner via a two-component system involving N-acyl homoserine lactone signal molecules. The equivalent signal molecules in Gram-positive bacteria are small (8-30 amino acids), secreted peptide- pheromones, which interact directly with the sensor components at the cell surface. Furthermore, whilst there are no reported examples in C. difficile, a two-component system (VirR and VirS) has been identified in Clostridium perf ngens and shown to control the production of toxins and other virulence factors in this clostridial species.

A further object of the invention is therefore to identify novel virulence factors through the modulation of the cellular levels of the transcriptional factors responsible for the expression of their encoding genes.

In a further aspect, the invention provides a method of identifying a C. cZ/T^/c/^'/e virulence factor, comprising culturing C. difficile(a) in the absence of, and (b) in the presence of, a regulating factor that promotes expression of C. difficile virulence factors and identifying a putative virulence factor whose expression is reduced in (a) compared with (b).

In a particular embodiment of the invention, the method comprises reducing the activity of a regulating factor that promotes expression of the virulence factors by administering an antisense sequence to the regulating factor. Subsequent analysis of the proteins expressed with and without the regulating factor can enable identification of virulence factors, as these will be differentially produced according to the level of the regulating factor. A useful tool for discovery of virulence factors is thereby provided.

Also within the ambit of the invention is a protein whose differential activity regulates expression of virulence factors in C. difficile and a DNA whose increased expression regulates expression of a virulence factor in C. difficile.

A particular utility lies in developing a means of bringing about the functional inactivation of genes. The ability to inactive genes allows their biological relevance to be determined. Thus, the inactivation of a gene may lead to an organism which becomes attenuated in its ability to cause disease. The product of the gene may therefore be assumed to play a role in virulence and may have potential as a vaccine candidate or represent a suitable target for therapeutic intervention strategies.

One such target is the virR gene, but other two-component transcriptional factors may also be targeted once the C. difficile genome sequencing information becomes available. Two forms of modulation are possible. One is reliant on the application of procedures where the regulatory gene function (e.g., virR) is repressed or entirely abolished (i.e., by insertional inactivation of the gene or through inhibition of mRNA translation via anti-sense RNA, see Fig.2), while the other is reliant on enhancement of regulatory gene function through increased expression. Virulence factors in cells subject to the former will be "down"-regulated. Increased expression of the regulatory factor will lead to "up"-regulation of virulence determinants. These changes in protein abundance will be detected using 2-D gels. A selection of the proteins thus identified will be subjected to microsequence analysis, and their identity confirmed, where possible, by reference to sequence data bases. The availability of the C. difficile genome sequence allows the rapid isolation of the entire structural region of identified genes. This will provide the information to generate gene knock-outs in C. difficile. The effect of insertional inactivation on pathogenesis may then be tested by analysis of the virulence of each mutant in an animal model, with a particular emphasis on persistence (i.e., colonisation). The effect on adherence, using in vitro cell lines, may also be tested. The isolated gene may also be used in proprietary expression systems to produce sufficient recombinant material to test its ability to elicit protection against CDAD in animal models. The effectiveness of a particular candidate may be assessed in comparison to results achieved with recombinant Toxin A polypeptide. Its effectiveness alone and in conjunction with toxin A may also be assessed.

The integration vectors developed may also be employed to deliver transposons to the host's genome as a means of generating random mutations. However, the only transposons know to function in C. difficile are those large conjugative transposons, such asTn916, which integrate at a single point in the hosts chromosome. They are therefore entirely inappropriate as random mutagenesis tools. A further aspect of the invention is therefore the development of effective transposons able to integrate at random in the C. difficile genome. This is obtained by derivatising insertion elements (IS) of a clostridial origin.

A yet further aspect of the invention provides a C. difficile transposon.

Preferably a C. difficile transposon comprises a sequence coding for a selectable marker. More particularly, the invention provides a C. difficile insertion sequence, comprising a sequence encoding a transposase that catalyses transposition of the insertion sequence within the C. difficile genome and, optionally, a sequence encoding a selectable marker. These insertion sequences are of use in investigations to ascribe gene function in C. difficile and carry out signature tagged mutagenesis or other mutagenesis to investigate gene function and properties.

In use, a transposon is cloned (using PCR) and thereafter derivatised through the creation of a unique restriction site within a non-essential region and the subsequent insertion of DNA fragments carrying either a Cm^R or Sp^R gene.

The ability of each element to transpose can then be tested by cloning them into derivatives of transformation plasmids, such as the specific plasmids pCD35EC or pCD35ES made in examples below, or the conjugation proficient derivatives thereof, (IS-Cm into ppCD35ES or IS-Sp into pCD35EC) which are reliant on an inducer (eg., nisin) for their replication, and then introduced into a clostridial host (for example, in the first instance C. bejeirinckii, thereafter C. difficile). Cells may be grown under non-permissive conditions and then plated onto agar media and screened for loss of the plasmid encoded marker. Determination of whether integration has occurred at random may be achieved by appropriate Southern blot and/or PCR analysis of antibiotic resistant clones, and eventual nucleotide sequence determination of the site of integration.

Micro-organisms protect themselves from invasion by foreign DNA elements (e.g., phages) using restriction/methylation systems. Methylase enzymes recognise specific nucleotide sequences (usually palindromes of 4-8 nucleotides in length) in DNA and specifically methylate either a cytosine or adenosine residue. In the absence of this methylation the site is unprotected and is cleaved by the cognate restriction enzyme. Thus, methylation of specific sequences is an organism's method of labelling its own DNA as self, thereby preventing its destruction by its own restriction enzymes. Foreign DNA which has not been appropriately methyated will be degraded by the host's restriction enzymes.

A still further aspect of the invention provides a method of identifying a C. difficile methylase gene, comprising identifying the sequence of a bacterial methylase gene, comparing the sequence of the gene with the genome of a strain of C. difficile, identifying a region of the genome that contains at least 30% sequence identity with the bacterial methylase gene, and expressing that region.

The identified genes may then be cloned into a bacterial host and used for preparation of plasmids of the invention, ensuring they are resistant to restriction by C. difficile restriction enzymes.

The identification of the restriction/methylation specificity of bacterial cells may be determined by a number of means. The traditional approach is to test bacterial cell lysates for the presence of enzyme activity capable of cleaving plasmids of known sequence into discrete fragments. By estimating the size of the fragments generated, and by reference to the sequence, it is sometimes possible to predict the position at which the plasmid has been cut. Further evidence of the validity of this prediction may be obtained by cutting the marker DNA with a commercially available restriction enzyme which cleaves the predicted sequence, and then comparing the fragment profile obtained to that generated by the action of the bacterial lysate. Further confirmation may be obtained by cloning the fragments obtained and then determining their sequence.

Having identified the site at which restriction occurs, it is necessary to identify the specificity of the methylation enzyme required to protect that site from cleavage. This may be determined by subjecting the DNA carrying the restriction site to the activity of a previously characterized methylase enzyme which is known to act at the desired target site. Thus, in the case of C. botulinum it proved possible to shown that the strain under investigation produced a restriction enzyme (designated Cbo\) which cleaved the palindromic sequence 5'-CCGG-3'. This is the same sequence as is recognized by the commercially available enzymes HpaW and Msp\.

However, the cognate methylase enzymes of these two restriction enzymes, M.HpaW and M.Mspl, respectively, methylate dissimilar bases in the sequence CCGG. Thus, M.Mspl methylates the external "C" residue (5'- C^MCGG-3') whereas M.HpaW protects the internal "C" nucleotide (5'- CC^MGG-3'). Accordingly, whilst methylation of 5'-CCGG-3' by M.Mspl protects the DNA from digestion with Msp\, it does not protect against digestion by HpaW. Similarly, M.HpaW will not protect against cleavage by Mspl. In the case of C. botulinum, it proved possible to show that only DNA treated with M.Mspl was protected against restriction by the clostridial Cbo\. This indicated that M. Cbo has the same specificity as

M.Mspl, and therefore effected the methylation of the external C of the 5'- CCGG-3' restriction enzyme site thus 5'-C^MCGG-3'. As a consequence of this analysis, when clostridial plasmid DNA was subsequently prepared in an E. coil host carrying a cloned methylase {M . BSLA) of equivalent specificity to M. Cbol and M.Mspl, transformation of C. botulinum was achieved (10⁵ transformants/g DNA). A similar strategy has proven successful in C. acetobutylicum.

The present invention has identified the presence of restriction/methylation activity in a C. difficile lysate (from strain CD6) equivalent to Mbo\IM.Mbo\ and Sau96\/M. Sau9G\. The availability of the C. difficile genome sequence, however, provides an alternative route by which restriction/methylation specificity may be determined. This is because methylase enzymes have conserved primary amino acid sequences. Thus, by comparing the predictive amino acid sequences of open reading frames found in the genome sequence with known bacterial methylase enzymes it is possibly to identify putative methylase genes of C. difficile. These may be cloned from the genome, following the amplification of fragments encompassing the structural gene using PCR and appropriate oligonucleotide primers. Once cloned in Eco/l their specificity may be determined by testing the ability of the enzyme encoded by the cloned fragment to protect the DNA (both chromosome and plasmid) of the E.coli host from digestion with a range of commercially available restriction enzymes of known specificity. Thus a further aspect of this invention is the determination of the restriction/methylation specificity of C. difficile as a necessary prerequisite to developing a transformation procedure.

Knowledge of the restriction/methylation specificity of C. difficile will allow the creation of an E. coli host producing those methylase activities able to appropriately methylate clostridial shuttle vectors such that they become immune to restriction by C. difficile. Thus, a further aspect of the invention is the creation of such a strain.

There now follows description of specific embodiments of the invention, illustrated by the accompanying drawings in which:-

Fig. 1 shows identification of virulence factors through modulation of the expression of transcriptional factors;

Fig. 2 shows use of an integrative antisense vector to identify virulence genes; Fig. 3 shows schematic representation of the C. difficile plasmid pCD6 (6829 bp);

Fig. 4 shows amino acid alignment of pCD6 orf A (Rep A; 545 aa) with Rep A from the C. perfringens plasmid plP404 (405 aa); Figure 5a-e shows construction of C. difficile : E. co//shuttle vectors pCD25E, pCD35E and pCD35C;

Fig. 6 shows Amino Acid alignment of VirR response regulator proteins from C. difficile (236 aa) and C. perfringens (252 aa);

Fig. 7 shows construction of pMTL540FT; Fig. 8 shows construction of pMTL4Tc;

Fig. 9 shows construction of pMTL4Fd;

Fig. 10 shows construction of EmR and CmR versions of pMTL4Fd (pMTL4FdEm and pMTL4FdCm);

Fig. 1 1 shows construction of pMTL940C; Fig. 12 shows construction of the integration vector pMTL910E;

Fig. 13 shows demonstration of Restriction Enzyme activity in C. difficile strain CD6;

Fig. 14 shows the demonstration of production of "anti-virR" RNA in the transconjugant strain FM18A, and Fig. 15 shows how the recognition sequence of unknown restriction enzymes may be determined.

In more detail, figure 1 shows identification of virulence factors through modulation of the expression of transcriptional factors. Virulence factor production is modulated via a two-component regulatory pathway, consisting of a membrane-bound sensor protein and a cytoplasmic response regulator protein. A specific stimulus from outside the cell induces autophosphorylation of the sensor which then activates the regulator protein, leading to transcription of virulence factor genes. Abolition of the regulator gene leads to non-production of these virulence factors. Those proteins affected can be identified as absent spots on 2-D gel electrophoresis.

Figure 2 shows use of an integrative antisense vector to identify virulence genes. Normally, the response regulator is transcribed into mRNA, then translated into protein, which then activates expression of virulence factors.

An antisense gene is introduced by integration into the chromosome; this is homologous to the anti-sense strand of the response regulator gene, and produces an "anti" mRNA which binds to the normally-produced RNA and it is not translated. Thus no protein is produced and no virulence genes are expressed.

Figure 3 shows schematic representation of the C. difficile plasmid pCD6

(6829 bp). The plasmid encodes five Open Reading Frames (ORFs). Two ORFs have protein sequence homologies to known proteins: ORF A to the C. perfringens plP404 plasmid replication protein (Rep A), and ORF E to autolysins from a variety of Gram-positive species. Distal to the orf A is a region (245 bp) consisting of Direct DNA repeats (Shown below plasmid diagram). The organisation of these repeat units can be arranged in either of two ways: 7 iterations of a 35 bp repeating unit; or an iteration of one 9 bp unit followed by two identical 13 bp units, repeated 7 times.

Figure 4 shows amino acid alignment of pCD6 orf A (Rep A; 545 aa) with

Rep A from the C. perfringens plasmid plP404 (405 aa). The alignment was performed using the Lipman-Pearson Method. Overall similarity is 20.2%.

Figure 5 shows construction of C. difficile : E.co//shuttle vectors pCD25E, pCD35E and pCD35C.

(5a) A 2002 bp HinDW\/Hinc\\ fragment encoding the 3'-end of orf A and the origin of replication from pCD6 was cloned into the HinDW\/EcoRV sites of pMLT23E to create pMTL23E-HH. Equivalent vectors were also made using pMLT22E, to create pMTL22E-HH.

(5b) A 2030 bp fragment encoding the orf B and the 5' end of orf A was generated by PCR using the primers P3 and P5. This fragment was cloned into the PCR cloning vector pCR2.1 -Topo to create pCR2.1 -P3P5.

(5c) The P3P5 fragment was removed from pCR2.1 -P3P5 by EcoR\ and HinD W and cloned into the EcoRl and HinDW\ sites of pMTL23E-HH to create the C. difficile : E.coli shuttle vector pCD35E. In order to generate a CmR version of this vector, the calP gene from pJIR418 was introduced as a 1513 bp coRI-filled in fragment into the EcoR\ site of pCD35E, thus creating pCD35EC. The same fragment was also inserted into pMLT22E-HH to make pCD36EC.

(5d) A 1268 bp fragment encoding the 5'-end of the orf A gene of pCD6 was generated by PCR using the primers P3 and P5. This fragment was cloned into the PCR cloning vector pCR2.1 -Topo to create pCR2.1 -P2P5. (5e) The P2P5 fragment was removed from pCR2.1-P3P5 by EcoRl and HinDW\ and cloned into the EcoR\ and HinDW\ sites of pMTL23E-HH to create the C. difficile : E.co/i shuttle vector pCD25E. Similarly, pCD26E was generated from pMLT22E-HH.

Figure 6 shows Amino Acid alignment of VirR response regulator proteins from C. difficile (236 aa) and C. perfringens (252 aa). The alignment was performed using the Lipman-Pearson Method. Overall similarity is 45.5%.

Figure 7 shows construction of pMTL540FT. A 190 bp Nhe \IXba I fragment encoding the transcriptional terminator of the C. pasteurianumferrodoxin gene was cloned into the Xba\ site of pMTL540F.

Figure 8 shows construction of pMTL4Tc. A 3121 bp fragment encoding the tetM gene was generated by PCR from l 916. This was cloned into pMTL4 at the EcσRV site.

Figure 9 shows construction of pMTL4Fd. An 862 bp Nhe \/Eco RV fragment containing the Fd po: Fd ter cartridge from pMTL540FT was cloned into the corresponding Nhe \/Eco RV sites of pMTL4.

Figure 10 shows construction of EmR and CmR versions of pMTL4Fd (pMTL4FdEm and pMTL4FdCm). Fragments encoding the EmR gene from pMTL540E and the CmR gene from pJIR418 were amplified by PCR and inserted by blunt end ligation into the EcoRM site of pMTL4Fd.

Figure 1 1 shows construction of pMTL940C. A 2320 bp Nhe I fragment from pMTLFdCm (see figure 10 above) encoding the CmR gene and the Fd po: Fd fercartridge was end-filled and ligated into the unique Eco ICRI site of pMTL4Tc (described in figure 8).

Figure 12 shows construction of the integration vector pMTL910E. A 212 bp encoding the nisA promoter (Box below plasmid diagram) was amplified by PCR from L. lactis. This fragment was cloned upstream (5'-) of the pAMβl repD gene of plasmid pMTL20Eβ10 as a Nsp Hl/Ssp I fragment. Figure 13 shows demonstration of Restriction Enzyme activity in C. difficile strain CD6. A crude lysate of C. difficile CD6 was prepared [Autolytic enzyme system of Clostridium botulinum. I. Partial purification and characterisation of an autolysin of Clostridium botulinum type A. Kawata T, Takumi K Jpn J Microbiol 1971 Jan 15: 1 1-10].This was used to digest preparations of plasmids pGK12 and pJIR418, which had been prepared from a Dam+ or Dam- background. Lane A displays digest patterns obtained by CD6 lysate in a Dam+ background where only Sau 96I cuts. Lane C displays digest patterns in a Dam- background where both CD6 Sau 96I and Mbo\ cut. Lane B shows how commercially-sourced Mbo I enzyme alone cuts.

Referring to figure 14, total RNA was prepared from C. difficile wild-type strain (CD630) and transconjugant strain containing the "anti-virR" gene fragment(FM 18A). This was isolated at two stages of growth, early and late log phase. cDNA was generated and PCRs performed on these RNA templates using oligonucleotide primers specific to "sense" or "antisense" virR gene fragments. Lanes 2-5 show that "anti-sense" virR RNA is only expressed in the transconjugant strain FM 18A. Lanes 6-9 show that "sense" virR RNA is expressed in both strains. Lanes 10-13 are controls where no cDNA is generated, lanes 15-18 are PCR controls performed on DNA alone

Referring to figure 15, in step 1 , plasmid DNA of known sequence is incubated with lysates prepared from C. difficile and the fragments generated analysed on agarose gels( in this example 3 fragments are generated, labelled A, B and C. These can be generated by either a single restriction enzyme or 2 or 3 different enzymes). The DNA fragments are blunt-ended by treatment with T4 polymerase, gel purified and (in step 2) ligated to a a blunt-ended fragment of known sequence (the 'Anchor' fragment). Outward facing oligonulceotide primers based on the proximal and diastal end of the

'Anchor' fragment are then used in PCR (step 3) to generate a DNA fragment composed of the ends of the 'Anchor' fragment and the plasmid-derived restriction fragments (A, B & C). The sequence of this amplified fragment is then determined, using either the P1 & P2 primers, or primers based on sequence residing 3' to the P1 & P2 sequence, but within the 'Anchor' region). The determined sequence shows where the original plasmid was cleaved, thereby revealing the identity of the recognition sequence of the enzyme(s) responsible.

Example 1 .

Characterisation of the C. difficile plasmid pCD6

Nucleotide sequence analysis of plasmid pCD6 (Fig. 3) has shown it to be composed of 6829 bp (SEQ ID 1 ). It carries an ORF encoding a protein of equivalent size, and sharing weak homology (20.6% identity) to (Fig.4), RepA of the C. perfringens plasmid plP404. In plP404, the origin of replication

(an extensive repeat region) immediately follows RepA. The 3'-end of the pCD6 orf A is similarly followed by a repeat region (Fig. 3). However, these repeats show no homology to those found in plP404.

Confirmation that this region encompasses the replication functions of pCD6 has been obtained by cloning a 3.994 kb fragment (position 221 to position 4214, SEQ ID 1 ) carrying orf A + repeat region into the replicon cloning vector pMTL23E and showing the resultant plasmid, pCD35E (Fig. 5a,b,c), will transform Clostridium beij^'erinckii NC\MB 8052. A second plasmid was similarly made (Fig. 5a,d,e) by inserting a smaller pCD6-derived fragment (position 983 to 4214, SEQ ID1 ). This plasmid, pCD25E was also shown to be capable of transforming Clostridium beijerinckii NCIMB 8052. However, Em^R colonies did not develop for a further 24 hours. A major difference between the two plasmids pCD35E and pCD25E resides in the fact that the former carries a second gene {orfB) in addition to orfA.

This result may indicate that orfB plays some role in replication, and could, for instance, influence plasmid copy number.

The derived plasmids (pCD25E and pCD35E), therefore, presented a starting point for electroporation experiments. However, C. difficile strain 630 is resistant to erythromycin (Em). A derivative of pCD35E was made, designated pCD35EC (Fig. 5c), in which the catP gene of the C. perfringens plasmid pJIR418 has also been inserted. The catP gene was similarly inserted into pCD25E to create pCD25EC. Strain 630 (the genome of which is being sequenced) is sensitive to Cm, and the related cat gene (that of pC194) has been shown to express in C. difficile. Other antibiotic genes may be similarly employed, such as genes encoding spectinomycin (Sp) sourced from organisms such Enterococcus faecalis or Staphylococcus aureus. C. difficile strain 630 is sensitive to Sp, and genes encoding resistance to this antibiotic have been shown to function as selectable markers in both C. botulinum and C. beijerinckii [unpublished data of the inventors].

Example 2

Delivery of Anti-sense RNA

We have used the information obtained in C. perfringens to isolate the C. difficile virR homologue. Accordingly, oligonucleotide primers were designed based on the sequence motifs in the C. perfringens VwR protein that are conserved amongst bacterial proteins of the two-component system family. A DNA fragment of approximately the correct size was amplified which, upon cloning and nucleotide sequence analysis was found to encompass an open reading frame (orf) encoding a polypeptide with high homology to the C. perfringens VirR protein. This fragment was then employed to isolate a full length genomic clone carrying the entire virR gene and its entire nucleotide sequence determined (SEQ ID2). The final sequence obtained exhibited 46% similarity to the C. perfringens VirR counterpart (Fig. 6)

To construct an antisense RNA for virR, we have isolated a 496 bp fragment

(position 391 to 887, SEQ ID 2) by PCR from the 5'-end of the virRgene and inserted it in the opposite orientation to which it is normally transcribed downstream of the strong ferredoxin promoter (Fd po) of in a derivative (pMTL540FT, Fig. 7) of our previously constructed clostridial expression vector, pMTL540F into which has been inserted the Fd transcriptional terminator. There appear to be no rules as to what is the most effective size of RNA to use. In previous studies, DNA fragments sizes of 73 bp, 600 bp and 292 bp have been employed. One of the benefits of using the expression cartridge of pMTL540FT is that the RNA molecule produced will carry the transcriptional terminator (Fd teή of the ferredoxin gene. This element will bring structural stability to the anti-sense molecule produced. To deliver the Fd po::ant\-virR.:Fd ter cartridge to the C. difficile chromosome we have constructed a purpose built plasmid, pMTL940C. This was derived through the construction of a series of different vectors. The first step was to insert a 3121 bp fragment from the tetracycline resistance gene {tetM) of Tn976 into the EcoRV site of pMTL4 (an Ecoli plasmid based on the ColE1 replicon) to yield the recombinant plasmid pMTL4Tc (Fig. 8). In parallel the Fd pov. ter cartridge was excisied from pMTL540FT and cloned into pMTL4 between its unique EcoRM and Nhel sites. Two derivatives of the plasmid obtained, pMTL4 Fd (Fig. 9) were then made by inserting either the catP gene of pJIR418 (to give pMTL4FdCm, Fig. 10) or the erm gene of pMTL23E (to give pMTL4FdEm, Fig. 10). Finally pMTL40C was generated (Fig. 1 1 ) by excising the fragment from pMTL4FdCm enompassing the Fd pov. ter cartridge and catP gene into the unique EcσlCRI site within the tetM gene of pMTL4Tc. Whilst not made, an equivalent plasmid (pMTL940E) could also be made using the equivalent fragment from pMTL4FdEm.

Having generated pMTL940C, the anti- v/rr?gene fragment was inserted into the multiple cloning site region adjacent to Fd po. The provision of the tetM gene provides the necessary region of homology for the plasmid to undergo homologous recombination with the chromosomally located Tn 916 in the Bacillus subtilis donor. By flanking the Fd po:: anti- virRv.Fd ter cartridge and catP w\th sequences derived from tetM their insertion into the bacillus genome may be brought about by either double cross-over (ie„ replacement of the wild type fefM gene with tetMv.Ψd povanW-virRv.Ψd ter. : catP. : tetM ) or single cross-over (ie., integration of the entire plasmid by a Cambell-like mechanism).

The final plasmid derived has been transformed into a Bacillus subtilis strain carrying Tn976. No extrachromosomal plasmid DNA could be detected in the lysates of the choramphenicol resistant transformants that arose. Furthermore, when an oligonuleotide primer complimentary to a sequence present in Tn916, but not in pMTL940, was used in PCR in conjunction with an oligonucleotide primer complimentary to a sequence present in pMTL940, but not Tn 7 , a fragment of the expected size was generated. The findings were consistent with the integration of the entire plasmid into Tn976 by single cross-over. The entire integrated plasmid has been transferred to Clostridium difficile by conjugation. The Tn916 pMTL940 construct is self-mobilised and transfers across from Bacillus subtilisXo Clostridium difficile. The two organisms were mixed and allowed to mate on a nitrocellulose filter. They were plated out on C. difficile-selective media containing chloramphenicol in order to select for resistant Clostridium difficile colonies, which demonstrate that the integrated pMTL940 plasmid has transferred to C. difficile. A transconjugant generated in such a way was designated FM 18A.

In order to demonstrate that this transconjugant was producing anti-sense virR RNA Reverse Transcriptase PCR (RT-PCR) was performed. We demonstrated both production of "sense" virR (wild-type virR) RNA and "anti-sense" virR RNA. Firstly, total RNA was prepared from wild-type (CD630) and anti-virR transconjugant (FM18A) strains at early and late log phase. Then a cDNA strand was generated from the RNA transcript from the sense and antisense virR using reverse transcriptase and oligonucleotide primers specific to the relative sense and antisense w/7?DNAs. PCRs using these these cDNAs as templates were performed using oligonucleotide primers specific to the "sense" virR gene (i.e. both primers not in the fragment used in the anti-sense virR construct) and "anti-sense" virRgene fragment (i.e. primers specific to the anti-w^'r 7 region of the integrated pMTL940 plasmid in strain FM18A). (Fig. 14). Using this method we demonstrated that sense virR RNA is expressed in both wild-type (CD630) and transconjugant (FM18A) strains, but that anX\-virR RNA is only expressed in the transconjugant strain.

Example 3

Development of Integration Vectors

To generate a plasmid conditional for replication a key component of the plasmid replicon, for example the Rep gene, is placed under the control of an inducible promoter, and we have placed the repA gene of pMTL20Eβ10 under the control of the nisA promoter and have shown that this plasmid

(pMTL910E, Fig. 12) is only able to transform strains of L. lactis which produce nisin, ie., replication only occurs when a functional n is R gene is present. Evidence was also obtained that the plasmid may only replicate in the presence of nisin. Thus despite the fact that the lactococcal host employed produces nisin, transformation was only obtained when the cells were grown in the presence of exogenous nisin (100 ng per ml) prior to electroporation, and thereafter plated on agar media supplemented with exogenous nisin. Moreover, subsequent retention of the plasmid by the host was reliant on the continued supply of exogenous inducer (at 100 ng per ml). Thus, when cells carrying the plasmid are grown in media (MEDIA) lacking exogenous nisin and plated on agar lacking inducer, then none of the 50 colonies picked onto agar media supplemented with erythromycin (ie., the antibiotic to which the plasmid confers resistance) were found to be resistant to the erythromycin. In contrast when the experiment was repeated incorporating nisin (100 ng per ml) in both the culture media and the agar media, 34 (68%) of those colonies which developed were shown to be erythromycin resistant.

It follows that nisin controlled replication of this plasmid may be elicited in the bacterial hosts, such as C. difficile, which do not produce nisin through provision of recombinant copies of nisRK. This is achieved by either cloning nisRK\nto the vector itself, or into a compatible plasmid (eg., pMTL540E).

Thus, the plasmid will be introduced into C. difficile in the presence of nisin. To remove nisin, transformants are subcultured to fresh media lacking the peptide, and grown without antibiotic selection. The cells can be harvested and washed at the subculture stage to remove all traces of nisin. Under these conditions the plasmid replication ceases and, following re- imposition of selection, its selectable marker is retained only if the plasmid integrates into the genome, through provision of homologous DNA (eg., as in mutagenic cloning). An alternative to pAMβl -based plasmids are equivalent plasmids based on the pCD6 replicon.

The system may also be adapted such that the non-permissive conditions for replication are induced by the presence of the inducer, rather than its absence. Two basic approaches may be followed. In the first, the inducible promoter (ie., nisA) are positioned between the repAJorfA and its natural promoter in an orientation such that it transcribes in a direction that is counter to that of the natural rep promoter. Transcription of the inducible promoter, following addition of inducer, will interfere with transcription of W

- 23 - rep, and therefore replication. Alternatively, the replicon may be engineered such that it is flanked by recombinational /ox sites. The ere recombinase gene is then placed under the transcriptional control of the inducible promoter, and inserted into the plasmid. Induction of Cre, through addition of inducer then results in excision of the replication region

Whilst the nisin promoter has been used to exemplify this system, any inducible promoter may be employed to generate the required conditional replication plasmids. Furthermore, the developed system may be used in any Gram-positive bacterium, provided the replication origin functions in that organism. Thus, derivatives of the pPAL5000 vector commonly used in Mycobacterium tuberculosis, for instance, may be constructed in which its replication gene is placed under the control of an inducible promoter. As with clostridia, the integration plasmids available for use in Mycobacterium are of extremely limited utility.

The regulation afforded by inducible elements such as the nisin system are examples of positive regulatory control, where expression of the genes important in replication is promoted by the addition of inducer, ie., nisin. An alternative strategy would be to place the gene encoding elements of the replication machinery (eg., the replication protein) under the transcriptional control of a promoter that is under negative regulation. That is to say, where transcription of the gene is prevented through the interaction of a repressor protein with an operator sequence which prevents RNA polymerase from producing mRNA, and where addition of the inducer negates the activity of the repressor, resulting in expression of the replication-associated element. An example of such a system is the promoter of the E. coli lac operon. The natura l promoter conta i ns a n operator seq uence (5- GGAATTGTGAGCGGATAACAATTCC-3) to which the product of the lacl gene binds, blocking mRNA transcription. Interaction of the Lacl repressor protein with the inducer (the natural inducer allolactose or the gratuitous inducer iso-propyl-thiogalactoside, IPTG) causes a conformational change in the protein which prevents its interaction with the operator sequence.

We have previously derivatised the ferredoxin gene (Fd) promoter of the

Clostridium pasteurianum promoter such that the E. coli lac operator sequence has been incorporated at the +1 position (Minton et al., 1990). We have further shown that expression from this modified Fd promoter (fac) is repressed in E. coli when a functional lacl gene is present, and that transcription may be induced through the addition of exogenous IPTG. We have now shown that a similar degree of regulatory control may be achieved in clostridia when the lacl gene is placed under the regulatory control of a clostridial promoter, such as that responsible for expression of the C. beijerinckii genes encoding phosphotransbutyrylase/ butyrate kinase. Induction of fac transcription requires unexpectedly high levels of IPTG, equating to a final concentration of 500 μg/ml.

This discovery opens up the possibility of constructing an expression cartridge. This element may be employed to control the expression of desirable genes in clostridial species, such as C. difficile, in which it replaces the promoter which controls expression of a plasmid replication gene (eg., repA of pAMβl -based plasmids, or the equivalent gene of pCD6).

The replication of such a derivatised plasmid would, therefore, be absolutely reliant on the presence of exogenous IPTG. The derivatised plasmids may, therefore, be transformed into clostridia and the resultant transformed cells grown in the presence of IPTG. Subsequent removal of inducer (through appropriate serial dilution, or harvesting of cells by centrifugation and resuspension in media lacking IPTG) would then lead to repression of rep gene transcription, and impairment of replicative functions. Under these conditions plasmid replication ceases and, following re-imposition of selection, its selectable marker is retained only if the plasmid integrates into the genome, through provision of homologous DNA (eg., as in mutagenic cloning).

Example 4

Derivation of Effective Transposons

There are a number of reported examples of IS elements in a range of clostridial species, particularly C. perfringens. It is further apparent that the genomes of C. acetobutylicum and C. difficile contain IS elements.

These IS elements are cloned (using PCR) and thereafter derivatised through the creation of a unique restriction site within a non-essential region and the subsequent insertion of DNA fragments carrying either a Cm^R or Sp^R gene. The ability of each element to transpose is then tested by cloning them into derivatives of pCD35EC or pCD35ES, or the conjugation proficient derivatives thereof, (IS-Cm into ppCD35ES or IS-Sp into pCD35EC) which are reliant on an inducer (eg., nisin) for their replication, and then introduced into a clostridial host (in the first instance C. bejeirinckii, thereafter C. difficile). Cells are grown under non-permissive conditions (ie. in the absence of inducer) and then plated onto agar media supplemented with either Cm (IS- Cm) or Sp (IS-Sp), and screened for loss of the plasmid encoded marker (Sp^R, pCD35ES or Cm^R, pCD35EC). Determination of whether integration has occurred at random is achieved by appropriate Southern blot and/or PCR analysis of antibiotic resistant clones, and eventual nucleotide sequence determination of the site of integration.

Example 5

Circumvention of the Restriction Barrier

The analysis of lysates of the C. difficile strain CD6 has shown the presence of an activity which digests pGK12 and pJIR418 plasmid DNA to give DNA fragments consistent with an enzyme which cleaves at the palindromic sequence 5'-GGNCC-3' (Fig. 13). This is the same sequence that is cleaved by the commercially available enzyme Sau96\. The cognate methylase of this enzyme, M.Saϋ96\ methylates this palindrome at the internal "C" residue, 5'-GGNC^MC-3'. When plasmid DNA was prepared in an E. coli host carrying the M.Satv96l gene, the CD6 lysate was shown to be unable to cleave either pGK12 or pJIR418 DNA. Furthermore, when a region of the CD6 genome was amplified by PCR following sodium bisulphite treatment, cloned and sequenced, all of the cytosine residues had been converted to thymidine with the exception of the internal "C" residues of the two 5'-GGNCC-3' palindromes found within this region. These experiments indicate that CD6 produces restriction/ methylation enzymes with equivalent specificity to Sau9β\/ M.Saϋ96\.

In addition, when the plasmid DNA used as the substrate to assay for restriction activity in cell lysates of CD6 was prepared in and E. coli dam- host (ie, the DNA was not methylated at the "A" nucleotide of the palindromic sequence 5'-GATC-3', thus 5'-GA^MTC-3') then a pattern of cleavage was obtained consistent with the cleavage of DNA at the sequence 5'-GATC-3'. As such cleavage did not occur when the plasmid was prepared in a dam - E. coli host, it is clear that strain CD6 also produces a restriction/methylation activity of equivalent specificity to Mbo\/M.Mbo\ (Fig. 13). Also, when plasmid DNA from an E. coli dam- host was incubated with lysate from strain CD6 in the presence of EDTA (which chelates divalent cations, consequently inhibiting nuclease activity, although not interfering with DNA methyltransferase activity) and S-adenosyl- methionine (which acts as a methyl donor for the methylation reaction), the DNA was rendered insensitive to restriction with either Saudβ or Mbo\, indicating that CD6 lysate had methylated the DNA at both of these sites. Additionally, chromosomal DNA isolated from strain CD6 was not restricted by Sau 6 or Mσol although Dpn\ which requires the sequence GATC to be methylated for recognition, could cut this same DNA

Whilst strain CD6 has been experimentally proven to possess the above restriction/methylation enzyme activities, other strains of C. difficile do not necessarily produce the same enzymes. Examination of the partial genome sequence of strain 630 has revealed that 4 separate genes are present which encode proteins sharing primary sequence homology with know bacterial methylase enzymes. Three of these predicted proteins (designated CD630- 1226, CD630-587, CD630-824) share homology with methylase enzymes which methylate cyotosine residues, whilst the remaining protein (CD630-

25) shares homology with adenosine-specific methylase enzymes. DNA fragments encompassing the genes encoding all four of these putative methylases have been amplified from the chromosome of strain 630 and cloned in E.coli n the plasmid pMTL21 . The plasmids obtained have been designated pCD1226, pCD587, pCD824 and pCD25, encoding the methylases CD630-1226 (SEQ ID 4), CD630-587 (SEQ ID 5), CD630-824 (SEQ ID 6) and CD630-25 (SEQ ID 7), respectively.

Analysis of both plasmid pCD587 and host chromosome prepared from the E.coli strain carrying this plasmid demonstrated that both DNAs were immune to restriction by Sa ϋ96\. This indicates that CD630-587 has methylase activity equivalent in specificity to M.S«_?tv96l. The enzyme has therefore been designated M.Cd/587 and its cognate restriction enzyme

This methylase may not be very active in strain 630 as DNA prepared from the strain is partially restricted by Sau9β. Also bisulphite treatment and subsequent PCR and sequencing of a region of the 630 chromosome which contains two Sau96 sites indicated that only one of the two sites was methylated. Analysis of chromosomal DNA from an E. coli strain carrying pCD824 indicated that CD630-824 conferred resistance to S al. This was confirmed by cloning CD630-824 in the vector pMTL21 (which carries a unique Sma\ site⁰ and by introducing a frameshift mutation in the gene. The presence of the intact copy of the gene prevented the plasmid from being cut by Sma\, while the presence of the mutated gene did not provide protection from Smal. This indicates that CD630-824 has methylase activity equivalent in specificity to M.Smal. The enzyme has therefore been designated

M. Cd/824- and its cognate restriction enzyme Co/824. The identity of the third cytosine specific methylase may be similarly determined by screening the DNA content of the ^".oo//strain carrying pCD1226 with all commercially available enzymes. CD630-1226 was used to probe the chromosome of CD6 and determined to be absent. The specificity of CD630-25 may also be determined in this manner. CD630-25 was determined to be present in the chromosome of CD6 by Southern hybridisation. In this instance the enzyme may have the same specificity as the enzyme experimentally identified in strain CD6, ie., methylation of the adenosine residue of the palindrom 5'- GATC-3'. However, DNA from strain 630 is cut by Mσσl but not by Dpn\, indicating that its GATC sites are not methylated. The methylase of 630 therefore would appear to possess a different specificity. Indeed, bacterial adenosine methylases that share the highest homology with CD630-25 are known to methylate palindromic sequences which conform to the consensus 5'-NTCGAN-3', where 'N' is any nucleotide or no nucleotide at all.

In the case of C. botulinum however, we have devised a direct route for identifying methylation specificity, based on the observation that, whilst unmethylated cytosine residues of genomic DNA may be converted to uracil by treatment with sodium bisulphite, methylated bases are not. Selected regions of the genome may therefore be PCR amplified from chromosomal DNA treated with sodium bisulphite, using appropriate mutagenic primers, and the amplified fragment sequenced to determined which cytosine residues are protected. To ensure that the amplified DNA is likely to carry a restriction site(s), the existing conjugative cointegrate procedure to generate a transconjugant in which a clostridial vector such as pCD35EC is integrated into the C. difficile genome. If pCD35EC does not carry target sequences, then the associated restriction enzyme will be of no consequence to the transformation experiments.

Example 6

Creation of specialised E.coli host able to appropriately methylate clostridial shuttle vector DNA

To enable clostridial shuttle vectors to enter C. difficile cells without being restriction a specialised E.coli host is prepared which appropriately methylates the plasmids DNA in vivo. In the case of strain 630, this E. coli strain carries the recombinant genes specifying the four identified methylase genes, CD630-1226 (SEQ ID 4), CD630-587 (SEQ ID 5), CD630-824 (SEQ ID 6) and CD630-184 (SEQ ID 7). However, other methylase genes from other sources encoding enzymes with equivalent specificity may also be employed. These genes are cloned in tandem on a plasmid which is compatible with the E.coli replicon of the clostridial shuttle vector. Expression of each gene is ensured through the provision of E. coli-de ved promoters. Thus, as the replicon of plasmids such as pCD35EC is based on that of ColE1 , the compatible plasmid is based on a plasmid such as pACYCI 84. Alternatively, the four genes may be integrated into the host genome.

Example 7

Methylase Protection of DNA in vitro using C. difficile lysates

Alternatively, modification of DNA to be transformed into C. difficile may be carried out in vitro using lysate from the strain to be transformed. DNA methyltransferases do not require divalent cations for activity: therefore by including EDTA to chelate divalent cations (and consequently inhibit nuclease activity, while not interfering with DNA methyltransferase activity) and S- adenosyl-methionine to act as a methyl donor, in reactions with the DNA to be used in transformation and the lysate from the C. difficile strain to be transformed, the DNA may be protected in vitro.

Specifically, plasmid DNA is incubated with an equal volume of lysate from the appropriate C. difficile strain in the presence of 20mM Tris-HCI, 50 mM potassium acetate, 5 mM disodium EDTA, 1 mM dithiothreitol, and 0.2 mM S-adenosyl methionine at 37°C for approximately 16 h. The DNA is then re-extracted from the solution by addition of an equal volume of phenolchloroformisoamyl alcohol (25241 ), vortexing briefly, centrifuging at 13,000 rpm for 10 min and removing the resulting upper layer. This step is repeated before precipitation of the DNA by addition of 0.1 V of 3M sodium acetate, pH 4.8 and 3 V of 100 ethanol followed by storage at 70°C for I h. After centrifugation at 13,000 rpm for 20 min, the DNA pellet is then rinsed with approx. 500 μl of 70% ethanol, dried and re-suspended in 10 mM Tris- Cl pH 8.5. Lack of susceptibility to restriction by the appropriate restriction enzymes implies that the DNA has been successfully methylated. This protected plasmid is then a suitable candidate for transformation into that strain of C. difficile. It has been demonstrated in the case of CD6 that the lysate protects the plasmid from subsequent diestion by Sau96 and Mbol.

Example 8

Methylase Protection of DNA in vivo using conjugative plasmid transfer

A further method for bringing about the appropriate methylation of plasmid DNA such that it is resistant to restriction digestion by the endogenous enzymes of C. difficile is to prepare the DNA from C. difficile cells carrying the plasmid. This plasmid may be introduced into C. difficile by transformation (Example 12) or by conjugative means. The latter may be achieved in the following manner.

A fragment carrying the σr/7^" region of a conjugative plasmid, such as RK2 is inserted into a suitable site in a clostridial plasmid, such as pCD35EC. Thus, for example, the oriT region of RK2 is amplified by PCR using the primers 5'-GTGCCTTGCTCGTATC-3' and 5'-CCTGCTTCGGGGTCATTATAG- 3', and cloned into plasmid pCRBIunt II TOPO, and thereafter re-isolated by digestion with the restriction enzymes Xho\ and Kpn\ and inserted into the equivalent sites of pCD35EC. The resultant plasmid, pCD35ECor/7^"is then transformed into an E.coli host, such as SM10, which carries a chromosomal insertion ofthe lncP plasmid RP42 (Tc::Mu). This transformed strain is then used as a conjugative donor in a mating with a C. difficile recipient. Thus, E. co// SM10 cells carrying pCD35ECor/T are grown overnight on 2x YT agar with the addition of ampicillin (50 μg/ml) and kanamycin (10 μg/ml) and a single colony transferred to 10 ml 2x YT broth plus ampicillin and kanamycin, and the culture incubated overnight with shaking (200 rpm) at 37oC. The C. difficile recipient strain (eg., strain 630 or CD3 or CD6) is grown on BHI agar anaerobically for 24 hours. A single colony is picked and transferred to 5 ml BHI broth and grow for approx 18 hours.

The conjugative procedure is undertaken as follows: Spin 1 ml of overnight E. coli SM10 cells harbouring pCD35ECσ/77^~at 5K for 1 min. Aspirate off supernatant and gently resuspend in 1 ml of sterile PBS. Re-spin cells and again remove supernatant. Place pelleted E. coliSM culture in anaerobic cabinet and add 10-100 μl of C. difficile, gently resuspend the E. coli pellet and spin as above for 1 min. Gently resuspend mating mix in approx 250 μl of anaerobic sterile PBS and spot cultures onto a well dried BHI agar plate. Incubate plate upright overnight, before harvesting biomass with a sterile loop. Resuspend cells in approx 500 μl PBS, serially dilute as required and plate onto BHI agar containing appropriate antibiotics (ie., Em and Cm for strains CD3 8- CD6, or Cm for strain 630). Incubate plates anaerobically for 24-48 hours before examining for transconjugant colonies.

Plasmid DNA may be prepared from the resulting antibiotic resistant transconjugants. This may be achieved through growth of the C. difficile transconjugant in 400 ml of Brain-Heart infusion (BHI) broth overnight anaerobically at 37°C. Bacterial cell culture is pelleted by centrifugation and resuspended in Tris-based buffer. Cells are lysed by treatment with 10 mg/ml lysosyme, releasing plasmid and chromosomal DNA from the cell. The DNA in solution is denatured by 1 % SDS and 0.2M NaOH, and renatured using 5M Potassium acetate solution, which allows re-annealling of plasmid DNA. The plasmid DNA in the resultant lysate may then be separated from other components by adding I g/ml CsCI to the solution and centrifuging at high speed (36K for 48 hrs), the plasmid separating in the CsCI gradient thus formed according to its bouyant density. Plasmid is extracted from the CsCI gradient using a needle, and dialysed against sterile water to yield a high quality high yielding plasmid preparation.

The plasmid thus obtained may be used to transform, using electroporation (Example 12), cells of C. difficile which do not carry the plasmid, or plasmids based on the same replicon. Strain CD3 and 630 do not carry plasmid pCD6. They are efficiently transformed with plasmids such as pCD35EC. The strain CD6 carries pCD6, and therefore will be relatively inefficiently transformed with pCD35EC. A plasmid-free variant of CD6 maybe be derived from the transconjugant carrying pCD35ECcv/7T Thus, plasmid pCD6 is lost from CD6 upon introduction of pCD35ECcv77~due to incompatibility between the two plasmids. Plasmid pCD35ECor/7^~ is subsequently cured from the CD6 transconjugant through growth of the organism in liquid media lacking antibiotics (ie., Cm and Em) over many generations (50 to 100) of cell division, subsequent plating of the cells onto agar media lacking antibiotics to give single colonies, and then the individual screening of each colony for the presence of cells which have lost antibiotic resistance through their transfer to agar medium containing antibiotic.

Example 9

PCR-based methods to determine site specificity of C. difficile restriction enzymes

In the case of C. difficile strains whose lysates do not produce a recognisable pattern upon incubation with a plasmid DNA marker of known nucleotide sequence, it is possible to determine the end points of the resulting fragments by purifying them, blunt-ending them and ligating them to a blunt-ended DNA 'Anchor' fragment (of at least 500 bp in size) of known sequence and for which outward-reading PCR primers have been designed.

Upon PCR amplification using these primers, a DNA fragment will be generated encompassing the entire plasmid marker-derived fragment and part of the DNA 'Anchor' fragment. Thus, by sequencing across the junction between the DNA 'Anchor' fragment and the plasmid marker-derived fragment it is possible to determine the sequence at the point at which a restriction enzyme activity in the C. difficile lysate cleaved the marker plasmid (Figure 15).

Alternatively, plasmid fragments generated from the plasmid marker DNA, following its incubation with the C. difficile lysate, is subjected to a second digestion with a range of commercially available restriction enzymes, each known to recognise a distinct single site within the plasmid at regular intervals along the length of the sequence. By comparing which lysate- generated restriction fragment is cleaved by which commercially available restriction enzyme, and the size of the subfragments generated, and through reference to the know sequence, it is possible to orientate and order the fragments generated. This allows outward facing PCR primers to be synthesised at the proximal and distal end of each fragment. The DNA is then blunt-ended (through treatment with T4 polymerase), circularised by treatment with DNA ligase and then used as a template in a PCR employing specific pairs of primers. Sequencing of the resultant amplified DNA identifies the sequence at the proximal and distal ends of the fragments prior to circularisation, and thereby the restriction enzyme recognition sequence.

Example 10

Development of transformation

The above allows the generation of vectors which are immune to restriction, ie., plasmids made in an E.coli host carrying cloned copies of the appropriate methylase genes, or plasmids which naturally, or are engineered to, lack identified restriction sites. Thus, using a plasmid guaranteed to replicate and which will not be restricted, the transformation conditions for C. difficile are adapted (eg., modification of cell growth conditions, preparation of cells, and pulse parameters) from other clostridial transformation protocols.

Having achieved transformation further refinements of the vectors may be made. Presently the pCD6 replicon fragment is relatively large. It follows that the minimum fragment required for replication may be identified, through appropriate subcloning, and used in the construction of generally useful cloning vectors incorporating a lacZregion and multiple cloning sites, eg., as in pMTL500E and JIR418 . The stability of constructed vectors and their derivatives may then be determined. This may require the incorporation of additional regions from plasmid pCD6, or other clostridial plasmids. It will also be important to determine which genes are absolutely required for replication (ie., orf A alone or flanking genes) and how they are transcribed, to enable the construction of integrative vectors (Example 3).

Example 1 1

Development of Inducible Systems

A further essential element for effective functional analysis of genomes is the provision of an inducible promoter system. Such an element is of particular utility in studies involving anti-sense gene modulation. The effectiveness of this strategy would be greatly enhanced if expression of the anti-sense RNA could be regulated, ie., where depression of a gene proved detrimental to the cell.

This may be achieved using the nisA promoter. It may also be brought about through the use of a tef-regulated promoter developed in B. subtilis, which has recently been used to regulate anti-sense production in S. aureus. This is particularly attractive, as C. difficile strain 630 is already

Tet resistant. A third option is to place the Fd promoter under Lacl repression. This approach has been successfully applied to B. subtilis, through the introduction of a lac operator site into the promoter utilised and engineering expression of the lacl gene from a constitutive vegetative promoter. This option is aided by the fact that the Fd promoter has already been derivatised to include a lac operator site, and the resultant promoter shown to be IPTG-inducible in E. coli. In all three strategies there would be a requirement to bring about the introduction, and expression, of the relevant transcription factor {nisR) or repressor (tetR or lacl), in C. difficile using the systems described in example 3. Example 1 2

Procedure for the electro-transformation of C. difficile

Having generated vectors that are immune to restriction by engineering to remove the relevant sites or by in vivo methylation in an E. coli host carrying cloned copies of the appropriate methylase genes or in the appropriate strain of C. difficilelnto which the plasmid had been introduced by conjugation or by in vitro methylation using lysate from the appropriate

C. difficile strain, the vector(s) may then be introduced into C. difficileby electroporation. Specifically, the C difficile strain is grown in BHI to mid- late log phase. The cells are harvested by centrifugation, washed in ice-cold 500 mM sucrose and then re-suspended in 0.01 volume of ice-cold 500 mM sucrose. 100 μl cell volumes are then electroporated with between 0.5 to 5 μg of vector DNA at DNA voltages ranging from 7.5 to 12.5 kV/cm in a Bio- Rad Gene Pulser (resistance of 400 Ohms; capacitance of 25 μF). The cell suspension is then immediately re-suspended in 1 ml BHI broth containing 500 mM sucrose and incubating at 37°C for 1 to 3 hr before plating on BHI agar containing the appropriate antibiotics to select for cells that have been transformed with the vector.

The present invention thus provides genetic manipulation of C. difficile.

Claims

1 . A plasmid for transformation of C. difficile.

2. A plasmid according to Claim 1 , for expression of a heterologous gene in C. difficile, comprising a C. difficile replicon and a restriction endonuclease site to receive the heterologous coding sequence.

3. A plasmid according to Claim 1 or 2, wherein the replicon comprises an origin of replication and a sequence coding for a replication protein that binds to the origin of replication and enables replication of the plasmid by C. difficile.

4. A plasmid according to any of Claims 1 to 3, further comprising a replicon for cloning of the plasmid in a host cell, such as an E. coli replicon.

5. A method of expressing a heterologous gene sequence in C. difficile, comprising:-

providing a plasmid containing the heterologous gene sequence;

introducing the plasmid into C. difficile; and,

optionally, wherein the plasmid also containing a gene coding for a selectable marker, selecting for C. diffici/ethat express the selectable marker.

6. A method according to Claim 5 wherein the plasmid is according to Claims 1 to 4.

7. A method of making a plasmid for expression of a heterologous coding sequence in C. difficile, comprising providing a plasmid that is not digested by C. difficile restriction enzymes, said plasmid comprising a C. difficile replicon, and inserting said heterologous coding sequence into the plasmid.

8. A method of making a plasmid for expression of a heterologous coding sequence in C. difficile, comprising providing a plasmid wherein said plasmid comprises a C. difficile replicon, and said heterologous coding sequence, and subjecting said plasmid to methylation so as to prevent digestion of said plasmid by C. difficile restriction enzymes.

9. A C. difficile replicon.

10. A replicon according to Claim 9, comprising a first DNA sequence comprising a C. difficile origin of replication that binds to a replication protein expressed in C. difficile and enables replication in C difficile a plasmid containing that origin of replication.

1 1. A C difficile replication factor.

12. A . difficile origin of replication.

13. A method of expressing a gene in C. difficile comprising making a plasmid according to any of Claims 1 to 4 containing that gene and transforming C. difficile with the plasmid.

14. A method of transformation of C. difficile.

15. A method of identifying a C difficile virulence factor, comprising culturing C. difficile (a) in the absence of, and (b) in the presence of, a regulating factor that promotes expression of C. difficile virulence factors and identifying a putative virulence factor whose expression is reduced in (a) compared with (b).

16. A method according to Claim 15, comprising reducing the activity of a regulating factor that promotes expression of the virulence factors by administering an antisense sequence to the regulating factor.

17. A protein whose differential activity regulates expression of virulence factors in C. difficile.

18. A DNA whose differential expression regulates expression of a virulence factor in C. difficile.

19. A method of identifying a vector that integrates into a gram positive bacterial genome, comprising transforming a gram positive bacteria with a plasmid, wherein the plasmid comprises an inducible promoter and replication of the plasmid is dependent upon presence of an inducer of the promoter, wherein the plasmid includes a sequence coding for a selectable marker, and wherein transformation takes place in the presence of the inducer, removing the inducer, and selecting for bacteria expressing the selectable marker.

20. A method of identifying a vector that integrates into a gram positive bacterial genome, comprising transforming a gram positive bacteria with a plasmid, wherein the plasmid comprises a suppressible promoter and replication of the plasmid is dependent upon absence of a suppressor of the promoter, wherein the plasmid includes a sequence coding for a selectable marker, and wherein transformation takes place in the absence of the suppressor, adding the suppressor, and selecting for bacteria expressing the selectable marker.

21. A method according to Claim 19 or 20, for identification of a vector that integrates into Clostridia.

22. A vector for transformation of a gram positive bacteria that integrates into the bacterial genome.

23. A vector according to Claim 22, that integrates into Clostridia.

24. A method of identifying a C. difficile methylase gene, comprising identifying the sequence of a bacterial methylase gene, comparing the sequence of the gene with the genome of a strain of C. difficile, identifying a region of the genome that contains at least 30% homology with the bacterial methylase gene, and expressing that region.

25. A C. difficile transposon.

26. A C. difficile transposon comprising a sequence coding for a selectable marker.

27. A C. difficile insertion sequence, comprising a sequence encoding transposase enzyme that catalyzes transposition of the insertion sequence and, optionally, a sequence encoding a selectable marker.