AU764620B2 - Expression system - Google Patents

Description

N 4 S&F Ref: 424760D1

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT

ORIGINAL

Name and Address of Applicant: Actual Inventor(s): Address for Service: Akzo Nobel N.V.

Velperweg 76 6824 BM Amhem The Netherlands Ruud Philip Antoon Maria Segers, Nicolas Robin Waterfield, Peer Lyng Frandsen, Jeremy Mark Wells Spruson Ferguson St Martins Tower,Level 31 Market Street Sydney NSW 2000 (CCN 3710000177) Expression System Invention Title: The following statement is a full description of this invention, including the best method of performing it known to me/us:- 5845c Expression System The present invention refers to an expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens P-toxin. The present invention also refers to a method for the preparation of native Clostridium perfringens p-toxin which uses an expression system of the invention, typically by culturing a transgenic Gram-positive bacterium, not being Clostridium, under conditions promoting production of said p-toxin, wherein the bacterium comprises an expression system of the invention.

Clostridium perfringens (also known as C. welchii) is a species of the large genus Clostridium. All bacteria belonging to this genus are spore-forming anaerobic Gram-positive bacilli. The species C.

perfringens can be subdivided in subspecies. Five subspecies have been described. These subspecies are generally known as "type" A-E. All subspecies produce several toxins, both major and minor toxins. The four major toxins are the a, P, E and i toxin. All C. perfringens types produce the atoxin. The p-toxin is produced by C. perfringens types B and C. In addition, a range of minor toxins is produced by all C. perfringens types. It is mainly due to the presence of one or more of these various toxins in the five C. perfringens types, that all C. perfringens species are pathogenic.

Type A is known to be pathogenic for both man and pigs. Type B is mainly pathogenic for lamb, sheep and goat, and causes "lamb dysenteria" and haemorrhagic enteritis. Type C is pathogenic for man, sheep, calf, lamb, pig, and bird. It is the cause of "struck", haemorrhagic enteritis, necrotic enteritis and enterotoxaemia.

20 As mentioned above, both C. perfringens type B and type C are known to produce the P-toxin.

The p-toxin is known to play the major role in the pathogenesis of necrotic enteritis in both man and animal. In man, this disease has been termed pigbel (Johnson et al.: Lancet ii: 496-500, (1987)). In :animals, necrotic enteritis has been described in calves, lambs and pigs (Hauschild, A.H.W. in S.

Kadis, T.C.Montie, Microbial toxins, p. 159-188. Academic press, Inc, New York (1971) and Willis, T.A. Anaerobic bacteriology: clinical and laboratory practice, 3rd ed. Butterworths, London (1977).

The p-toxin from Clostridium perfringens has been isolated in pure form (Sakurai et al., Infect. Immun. 18: 741-745 (1977), Sakurai et. al., Toxicon 25: 1301-1310 (1987)). Much is still unknown about the biophysical properties of the toxin and thus about its mode of action. Due however to the 30 fact that it could be obtained in a purified form the toxicity of p-toxin could be clearly demonstrated in animals. The lethal toxicity of purified p-toxin for eg. mice and guinea-pigs was shown by Sakurai et.

al. (Infect. Immun. 21: 678-680 (1978), Sakurai et. al. Toxicon 25:1301-1310 (1987)).

Recently, the nucleic acid sequence of the Clostridium perfringens P-toxin has been elucidated by Hunter et al. (Infect. Immun. 61: 3958-3965 (1993)). The nucleic acid sequence revealed the size of the P-toxin protein to be about Due to the fact that the role of Clostridium perfringens P-toxin from type B and C in pathogenesis is of such paramount importance, much effort has been put in the development of immunity against this toxin. Immunity against the p-toxin is sufficient to protect against Clostridium perfringens type B and type C infection. The only way of inducing immunity against the P-toxin is, to administer the toxin to the animal to be protected. It is however obvious that the toxin must be given [R:LIBC]424760d1.doc:mef in a detoxified form, since otherwise administration would lead to severe illness or death of the animal to be vaccinated.

Vaccines based on detoxified 1-toxin, also called 1-toxoid, were available already around 1960 (eg. GB. 901 433 and US 3 288 680).

All currently available vaccines based on inactivated Clostridium perfringens 1-toxin have however several important drawbacks.

In the first place, all -toxin-based vaccines comprise chemically detoxified, mainly formalindetoxified, -toxin and it was shown through the years that it is very difficult to standardise these chemical detoxification processes. The classical chemical methods for detoxification of proteins have the disadvantage that they alter the overall structure of the protein in a fairly random manner. And as a result, during the process of chemical detoxification the immunogenic properties of the 1-toxin also rapidly decrease. This can be seen eg. from graph 1 and 2: in graph 2 it is shown, that at least 1% formalin, a very commonly used inactivation-compound, is necessary to detoxify the 1-toxin under certain standard conditions. From graph 1 and 2, see Fig. 5, it can be seen however, that the antigenicity titre decreases dramatically with an increasing amount of formalin. A full titre can not be obtained anyway, because even the lowest amount of formalin needed for detoxification (graph 2) already gives a decrease of the titre. This implicates that there is only a very narrow band, in which both detoxification and a reasonable titre and thus immunogenicity can be obtained. Given this very *narrow band, and the fact that there are at least three variables; time of detoxification, temperature 20 and precise formalin-concentration, it is clear that it is very difficult to reproducibly produce detoxified 1-toxin. Another approach for the detoxification of 1-toxin is therefore highly wanted. No acceptable alternative has been found until now, for the following reason: the delicate balance between a sufficient level of detoxification and remaining immunogenicity implicates a close link between, on the one hand, the structure of the protein and, on the other hand, the biological properties of the protein.

It could therefore not be expected to be possible to change the protein structure thereby detoxifying .i the protein, without at the same time significantly impairing the immunogenic characteristics of the protein.

It is one of the merits of the present invention that it discloses for the first time a way to avoid the above-mentioned problem: using genetic manipulation techniques, specific and relatively large 30 amino acid regions were surprisingly found, that can be mutated to provide the desired non-toxic derivatives of 1-toxin without significantly impairing its necessary immunogenic properties.

An immunogenic fragment thereof is understood to be a fragment that, although not comprising the full length amino acid sequence of the derivative of the 1-toxin, still comprises those regions of the derivative that are capable of inducing a protective immune response in the host animal. A mutation is understood to be a change in the amino acid sequence of the derivative 1-toxin in comparison to the amino acid sequence of the wild-type 1-toxin.

The mutation is a substitution, deletion, insertion or inversion, or a combination thereof. A mutation can eg. be such that one or more amino acids of the p-toxin are replaced by other amino acids, with different characteristics. A suitable replacement is eg. the replacement of a negatively charged amino acid by a positively charged one, such as the replacement of Aspartate by Lysine.

[R:LIBC]424760d1 .doc:mef Another suitable replacement is that of an aromatic amino acid by an aliphatic one: eg. Tryptophan- Glycine. Also the replacement of a hydrophobic amino acid by a hydrophilic one, such as eg.

Isoleucine by Aspartate is suitable. Therefore it is preferred that derivatives of p-toxin employed in the invention, wherein at least one mutation is a replacement mutation. It is clear, that next to replacement, mutations involving the insertion and/or deletion of one or more amino acids providing non-toxic derivatives of P-toxin, still capable of inducing immunological response are also contemplated. Therefore in an equally preferred form, derivatives of P-toxin, wherein at least one mutation is a deletion or insertion. When'two or more mutations are made, combinations of replacement and deletion/insertion mutations are equally possible. A derivative of p-toxin is considered non-toxic if its LD50 (lethal dose killing 50% of all animals in an experiment) in mice is at least ten times higher than the LD5o of native P-toxin. The LD50 of native P-toxin is about 0.15 W/mouse.

As mentioned, it is one of the merits of this invention that, contrary to what was assumed in the art, it was found to be possible to genetically alter the amino acid sequence of the p-toxin such that a non-toxic and still immunogenic toxoid was obtained. Not all mutations will however lead to the i desired non-toxic and still immunogenic derivatives of p-toxin. Therefore, a test for rapid screening and selection of mutants producing a non-toxic derivative of P-toxin that is still immunogenic has been developed. This test allows screening of large numbers of mutants producing non-toxic and still immunogenic derivatives of P-toxin. This test will be disclosed below.

:i 20 It was also found that preferably those regions that form a transition domain between neutral and hydrophilic parts of the p-toxin are suitable as target regions for introducing mutations giving a non-toxic derivatives of p-toxin with the desired characteristics. These regions can easily be traced by applying the Hopp-Woods algorithm to the sequence of the p-toxin (Hopp and Woods; Proc. Natl.

S Acad. Sci. 78: 38248-3828 (1981)). Therefore, derivatives of P-toxin having a mutation that is located in a transition domain between neutral and hydrophilic parts of the P-toxin are preferred.

Regions that are very suitable as target regions for mutations are located at position 62, at position 182, at position 197, between amino acid number 80-103, 145-147, 281-291, 295-299 relative to the peptide leader methionine and the region downstream of the unique cysteine at amino acid position 292. Therefore derivatives of P-toxin having a mutation that is located in at least one of the S 30 regions at position 62, 182, 197 and between amino acid number 80-103, 145-147, 281-291, 295-299 and/or the region downstream of amino acid position 292 relative to the peptide leader methionine are even more preferred. The mutations are located at position 62, at positions 80-82, at positions 95-97, at positions 101-103, at position 145-147, at position 182, at position 197, at position 287-291, at position 295-299 are still even more preferred.

Non-toxic immunogenic derivatives of p-toxin can be made by chemically replacing or modifying amino acids in the protein. They can also be made by introducing mutations in the gene encoding the p-toxin. Methods for introducing mutations in DNA fragments are described below. The mutated DNA fragments can then eg. be cloned in a nucleotide sequence such as a suitable expression plasmid and subsequently be expressed.

[R:LIBC424760d1.doc:mef I 4 Therefore, nucleotide sequences comprising a mutated DNA fragment that has as a characteristic that it encodes a genetically detoxified immunogenic derivative of Clostridium perfringens P-toxin or an immunogenic fragment thereof can be provided.

As mentioned above, derivatives of p-toxin having a mutation that is located in a transition domain between neutral and hydrophilic parts of the P-toxin are preferred. Such a derivative of p-toxin can be made by expressing a DNA fragment encoding such a derivative of p-toxin. Therefore the mutated DNA fragment encoding the derivatives of P-toxin has a mutation in at least one DNA region encoding a transition domain between neutral and hydrophilic parts of the derivatives of p-toxin is preferred.

The detoxified immunogenic derivatives of Clostridium perfringens P-toxin can eg. be obtained by means of random mutagenesis of the gene encoding the P-toxin. Many ways of inducing random mutagenesis are known in the art, such as eg. chemically induced mutagenesis using mutagenising chemical agents, mutagenesis by U.V.-radiation or y-radiation or random mutagenesis using recombinant DNA technology.

Mutations at DNA-level can also be made at specific sites: site-directed mutagenesis. This is done with the help of known genetic engineering techniques. It is eg. possible to use restriction enzymes to cut out DNA fragments in or encompassing the target-regions, and to replace these fragments by synthetic fragments having a mutated sequence. Site-directed mutagenesis is also a very convenient technique for introducing mutations in the target-regions. If the mutation concerns a 20 replacement mutation, the reading frame will of course remain unaltered. A combination of deletion and/or insertion mutations and replacement mutations is also possible.

The DNA mutation techniques described above are well-known in the art and have been .described eg. in Maniatis, Molecular Cloning, Cold Spring Harbor Laboratory Press, ISBN 0-87969- 309-6(1989).

Several additional problems are however encountered during the isolation of both toxic and S'.i genetically altered non-toxic derivatives of Clostridium perfringens p-toxin from Clostridium perfringens. First of all, Clostridium is a dangerous bacterium to grow, seen from a workers health point of view. Growth of Clostridium species for the production of P-toxin must take place under high safety conditions which makes large scale production difficult. Secondly as mentioned before, 30 together with the P-toxin several other major/minor toxins are made. Isolation and purification of the p-toxin amongst these other toxins, as is eg. necessary for vaccine production, is difficult and timeconsuming. Thirdly, Clostridium species are spore-forming bacteria. It is necessary but difficult to eliminate all spores from the P-toxin preparation and at the same time retaining the immunogenic characteristics of the 0-toxin, since these spores are highly resistant against heat and chemicals.

Therefore, there clearly is a need for methods to provide p-toxin free from other Clostridium-toxins and free from Clostridium spores.

Non-Clostridium expression-systems based on the Gram-negative bacterium E. coli for the expression of Clostridium P-toxin have been described before. E. coli systems have been used by Hunter et al. (Infect. Immun. 61: 3958-3965 (1993)) for the expression of the P-toxin gene. Only a minor band corresponding to the expected 34kD protein was found whereas by far most of the P-toxin [R:LIBCJ424760d1.doc:mef protein was found in large 118kD multimeric forms. Steinporsdottir et al. (FEMS Microbiology Letters 130: 273-278 (1995)) tried to express the p-toxin gene in E. coli as a fusion protein fused to gluthatione S-transferase. They also, like Hunter, found only high molecular weight non-natural p-toxin. It was concluded in this paper, that the protein produced in E. coli had a conformation different from the native p-toxin. It therefore seems likely that other Clostridium-encoded proteins play an additional role in the conformational folding and excretion of the native p-toxin. This is known to be the case for many other excreted bacterial proteins. For this reason, since as mentioned above there is a delicate balance between structure and immunogenicity, p-toxin obtained from other sources than Clostridium perfringens and thus in a non-native form can not be expected to be an efficient basis for a vaccine, regardless the expression system used. It was surprisingly found now, that expression of the native p-toxin gene in Gram-positive bacteria other than Clostridium perfringens provides a native p-toxin with all the biological and biophysical characteristics and activities of the native p-toxin as produced in Clostridium perfringens, but having several highly desirable advantages over the Ptoxin produced in Clostridium perfringens or E. coli: a) it is not contaminated with Clostridium perfringens spores, b) it is produced free of contaminating lipopolysaccharides, c) it is produced free of other major/minor Clostridium toxins and d) it is in its native conformation. A thus obtained p-toxin can be inactivated with formalin S 20 more advantageously than the p-toxin obtained from Clostridium, since the only biological material to be taken into consideration during inactivation is the p-toxin itself: spores, lipopolysaccharides and other major/minor Clostridium toxins need not be taken care of, since they are absentper se.

'*Therefore, in still another embodiment, the invention relates to Gram-positive S 25 bacterial expression systems wherein the Gram-positive bacterium is not Clostridium perfringens, the said expression system comprising the gene encoding wild-type Clostridium perfringens p-toxin.

According to one embodiment of this invention there is provided an expression Yt' system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens 0..13 p-toxin.

According to another embodiment of this invention there is provided a method for the preparation of native Clostridium perfringens P-toxin characterised in that a nucleotide sequence comprising a DNA fragment encoding said p-toxin is expressed in a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium.

9 9 999.

9 i [I:\DayLib\LIBXX]04264.doc:KOB According to a further embodiment of this invention there is provided an expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens P-toxin.

According to yet a further embodiment of this invention there is provided Expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens p-toxin, said system being substantially as hereinbefore described with reference to any one of the examples.

According to yet another embodiment of this invention there is provided a transgenic Gram-positive bacterium comprising an expression system based on a Grampositive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens p-toxin, said transgenic bacterium being substantially as hereinbefore described with reference to any one of the examples.

The p-toxin gene can be under the control of its native promotor. It may also be placed under the control of a heterologous promotor. Such a promotor can be eg. the Lacpromotor (Chang et al., Nature 275: 615 (1978)) or the Trp-promotor (Goeddel et al., NAR. 8: 4057 (1980)). Such a promotor may be a strong promotor, leading to over- S 20 expression of the p-toxin, or it may be an inducible promotor. Both kinds of promotors *are known in the art.

Several expression systems for Gram-positive bacteria have been described, and versatile expression plasmids for use in several families of Gram-positive bacteria are known in the art. As an example may serve expression systems based upon the Enterococcal broad Gram-positive host range replicon of pAM 1l described by Simon and Chopin (Biochimie 70: 559-567 (1988)). Derivatives of this plasmid can be used for expression of foreign genes in eg. Streptococcus, Lactobacillus, and Bacillus species.

Within the group of non-Clostridium Gram-positive bacteria, those bacteria of which the I i genome has a relatively high AT-content are most suitable for cloning and expression of S 30 genes from the genus Clostridium. Therefore, in a more preferred form, the Grampositive bacterium used for the preparation of the native p-toxin or the derivative of Ptoxin according to the present invention is selected from the group of Lactococcus, S Lactobacillus, Leuconostoc, Pediococcus, Streptococcus, Enterococcus, Staphylococcus, Bacillus, Sarcina, Rurninococcus or Listeria.

[I:\DayLib\LIBXX]04264.doc:KOB It is of course even more advantageous to express a mutated P-toxin gene according to the invention in a Gram-positive bacterium other than Clostridium: in that case the detoxification treatment with formalin can be completely omitted. The so-obtained non-toxic derivative of P-toxin has, next to being free from spores, free from lipopolysaccharides, free of other major/minor Clostridium toxins and being in a native form, the additional advantage that it is non-toxic per se. Therefore, in a more preferred form, the invention relates to Gram-positive bacterial expression system, said Gram-positive bacterium not being Clostridium perfringens, comprising a nucleotide sequence encoding a derivative p-toxin gene according to the invention or an immunogenic fragment thereof.

There are disclosed herein vaccines for combating Clostridium perfringens infections, based upon genetically detoxified immunogenic derivatives of Clostridium perfringens p-toxin. Such vaccines can be made eg. by admixing an immunologically sufficient amount of detoxified immunogenic derivatives of Clostridium perfringens P-toxin according to the invention and a physiologically acceptable carrier. An immunologically sufficient amount is understood to be the amount of detoxified immunogenic Clostridium perfringens P-toxin that is capable of inducing a protective immune response in a host animal. Thus still further disclosed are vaccines for combating Clostridium perfringens infection that comprise a derivative of Clostridium perfringens P-toxoid according to the invention or an immunogenic fragment thereof, and a physiologically acceptable carrier.

A physiologically acceptable carrier is eg. water or a physiological salt solution. Often, the vaccine is additionally mixed with stabilisers, eg. to protect degradation-prone polypeptides from being degraded, to enhance the shelf-life of the vaccine, or to improve freeze-drying efficiency. Useful stabilisers are eg. SPGA (Bovarnik et al; J. Bacteriology 59: 509 (1950)), carbohydrates eg. sorbitol, mannitol, trehalose, starch, sucrose, dextran or glucose, proteins such as albumin or casein or degradation products thereof, and buffers, such as alkali metal phosphates. It goes without saying, that other ways to stabilise the vaccine by adding compounds are also embodied in the present invention.

Optionally, one or more compounds having adjuvant activity may be added to the vaccine.

Adjuvantia are non-specific stimulators of the immune system. They enhance the immune response of the host to the administered immunogens. Therefore, in a more preferred form, the vaccines according to the present invention comprises an adjuvant. Examples of adjuvantia known in the art are Freunds Complete and Incomplete adjuvant, vitamin E, non-ionic block polymers, muramyldipeptides, ISCOMs (immune stimulating complexes, cf. for instance EP 109942), Saponins, mineral oil, vegetable oil, and Carbopol (a homopolymer). Adjuvantia, specially suitable for mucosal application are eg. the E. coli heat-labile toxin (LT) or Cholera toxin Other suitable adjuvants are for example aluminium hydroxide, phosphate or oxide, oil-emulsions (eg. of Bayol F® or Marcol 52®, saponins or vitamin-E solubilisate.

The vaccine can be kept in storage using methods known in the art for storing live vaccines.

Storage can eg. be done at sub-zero temperature.

Freeze-drying also is a known and suitable method for the conservation of vaccines. Freezedrying has the advantage, that it stabilises the vaccine so that it can be kept in stock at temperatures [R:LIBC]424760d1.doc:mef well above those necessary to keep non-freeze-dried stocks. The vaccine according to the present invention can be freeze-dried very efficiently, especially when it is mixed with stabilisers such as those mentioned above before freeze-drying.

The vaccine can be administered to all hosts sensitive to Clostridium perfringens type B or C infection, such as man, lamb, sheep, goat, pig, bird and calf. The vaccine can be administered as early as day of birth, but can also be given at any later stage. Vaccine doses are preferably between 1 and 1001g of the derivative per animal, depending partially on the size of the animal. For eg. pigs, a very suitable dose ranges between 20 and 80 g.

Although in principle all common routes of administration can be used, the vaccine is preferably administered intraperitoneally, intranasally, intramuscularly, subcutaneously, orally or intradermally.

Still further disclosed is an alternative way of making a vaccine for combating Clostridium perfringens infections. Attenuated or non-pathogenic live bacteria and viruses having a Clostridium perfringens type B or C sensitive human or animal as a host, can be used as carriers of the mutated p-toxin gene according to the invention as described above. These so-called live recombinant carrier vaccines can be safely administrated to the host-animal together with a physiologically acceptable carrier: they behave non-pathogenic and they express a non-toxic derivative of the p-toxin. The live recombinant carrier-based vaccines have the advantage that they produce and or present derivatives of the Clostridium perfringens p-toxin directly in the host. Animals vaccinated with such a recombinant bacterium or virus will produce an immunogenic response not only against the immunogens of the 20 vector, but also against the immunogenic parts of the polypeptide(s) for which the genetic code is additionally cloned into the recombinant carrier. This has the advantage that by administrating such a recombinant carrier protection against two or more diseases can be obtained. Many live recombinant carrier viruses and bacteria are currently known in the art. Suitable live recombinant carrier bacteria are eg. Salmonella species and E. coli species. Viruses frequently used in the art as live recombinant carriers are adenoviruses, retroviruses, vaccinia viruses and herpes viruses. Also suitable as a 'i specific orally applicable carrier virus is Porcine Parvovirus. Another virus that is useful as a live recombinant carrier virus for carrying the gene encoding the p-toxoid is the herpes virus Pseudorabies virus (PRV) (eg. as described in EP 606,437). Such a PRV carrying the P-toxoid gene would protect pigs against infection with both PRV and Clostridium perfringens type C. Thus, vaccines for combating Clostridium perfringens infection that comprise a live recombinant carrier organism carrying the mutated P-toxin gene are also disclosed.

It is clear that the native p-toxin is an important, if not the most important virulence factor in Clostridium perfringens. Therefore, a Clostridium strain in which the native p-toxin gene is replaced by a mutant p-toxin gene according to the invention as described above has therefore lost an important virulence factor. Such a strain can be used as a live attenuated vaccine, since the p-toxin is not made in its toxic form but in the form of a non-toxic derivative. The advantage of such a live attenuated vaccine is that it produces a non-toxic but still immunogenic form of the p-toxin, whereas it has in addition all the other immunogenic antigens of the native Clostridium perfringens strain.

Therefore, vaccines comprising a live attenuated Clostridium perfringens strain in which the native ptoxin gene is replaced by a mutated p-toxin gene are also disclosed.

[R:\LIBC424760d1.doc:mef Also disclosed are vaccines that comprise, in addition to a p-toxin derivative according to the invention, immunogens from other pathogens. This allows to vaccinate against various diseases in one vaccine administration step. A more preferred vaccine comprises additional immunogens, selected from the group consisting of Actinobacillus pleuropneumoniae, Pseudorabies virus, Porcine Influenza virus, Porcine Parvovirus, Transmissible Gastroenteritisvirus, rotavirus, Escherichia coli, Erysipelothrix rhusiopathiae, Pasteurella multocida, Bordetella bronchiseptica, Salmonella species, Mycoplasma hyopneumoniae, Haemophilus parasuis and Helicobacter-like bacteria.

The present invention also relates to methods for the preparation of native Clostridium perfringens p-toxin, which methods are based upon the expression of a nucleotide sequence comprising a DNA fragment encoding said p-toxin in Gram-positive bacteria other than Clostridium perfringens.

Furthermore, methods for the preparation of non-toxic derivatives of Clostridium perfringens 3toxin are also disclosed. These methods are based upon the expression of nucleotide sequences comprising a DNA fragment encoding a derivative of p-toxin, in a Gram-positive bacterium.

Also, disclosed are methods for the preparation of a vaccine for combating Clostridium perfringens infection. Such methods eg. comprise admixing a derivative of Clostridium perfringens Ptoxin according to the invention and a physiologically acceptable carrier.

Example 1: Construction of expression plasmids pKS90, pTREXIA, pTREX7 and pTREX14.

20 The pKS90 plasmid is a high-copy number (40-80 per cell) theta-replicating gram positive plasmid based on the pTREX1 plasmid, which is itself a derivative of the previously published plL253 plasmid. plL253 incorporates the broad Gram-positive host range replicon of pAMb1 (Simon, and A. Chopin. 1988. Construction of a vector plasmid family and its use for molecular cloning in Streptococcus lactis. Biochimie. 70: 59-567.) and is non-mobilisable by the L lactis sex-factor. plL253 also lacks the tra function which is necessary for transfer or efficient mobilisation by conjugative parent plasmids exemplified by plL501. The Enterococcal pAMb1 replicon has previously been transferred to various species including Streptococcus, Lactobacillus and Bacillus species as well as Clostridium acetobutylicum, (Gibson, N.M. Chace, S.B. London and J. London. 1979 J. Bacteriol 137: 614-619. LeBlanc, R.J. Hawley, L.N. Lee and E.J. St. Martin. 1978. Proc. Natl. Acad. Sci.

30 USA. 75:3484-3487. Oultram, and M. Young. 1985. FEMS Micro. Lett. 27: 129-134) indicating the potential broad host range utility. While pTREX1 (and pTREX1A) represents a basic constitutive transcription vector, the derivative pKS90 utilises translational coupling to a lactococcal translational initiation region in order to increase expression level. Construction details of pKS90 and its immediate parent pTREX1 are given below.

pTREX1: An artificial DNA fragment containing a putative RNA stabilising sequence, a translation initiation region (TIR), a multiple cloning site for insertion of the target genes and a transcription terminator was created by annealing 2 complementary oligonucleotides and extending with Tfl DNA polymerase. The sense and antisense oligonucleotides contained the recognition sites for Nhel and BamHI at their 5' ends respectively to facilitate cloning. This fragment was cloned between the Xbal [R:AIBC]424760d1.doc:mef 9 and BamHI sites in pUC19NT7, a derivative of pUC19 which contains the T7 expression cassette from pLET1 (Wells, P.W. Wilson and R.W.F. Le Page. 1993. J. Appl. Bact. 74:629-636.) cloned between the EcoRI and Hindlll sites. The resulting construct was designated pUCLEX. The complete expression cassette of pUCLEX was then removed by cutting with Hindll and blunting followed by cutting with EcoRI before cloning into EcoRI and Sad (blunted) sites of plL253 to generate the vector pTREX. The putative RNA stabilising sequence and TIR are derived from the Escherichia coli T7 bacteriophage sequence and modified at one nucleotide position to enhance the complementarity of the Shine Dalgarno (SD) motif to the ribosomal 16s RNA of Lactococcus lactis. A Lactococcus lactis MG 1363 chromosomal promoter designated P1 was cloned between the EcoRI and Bglll sites present in the expression cassette, creating pTREX1. This promoter had been previously isolated using the promoter probe vector pSB292 and characterised by primer extension and DNA sequencing analysis (Nick. R. Waterfield, R. W. F. Le Page, P.W. Wilson, and J.M. Wells. Gene. 165. 1995. 9- The promoter fragment was amplified by PCR using the Vent DNA polymerase. The PCR fragment included all of the DNA sequence upstream of the cloned promoter and up to 15 bases 3' to the transcription start site. The Shine-Dalgarno (SD) sequence present downstream of the transcription start site of this promoter was deliberately excluded from the PCR amplified promoter fragment to prevent translation initiation at sites other than the start codon indicated in the expression cassette. Two similar vectors designated pTREX7 and pTREX14 have also been created using the weaker P7 and P14 promoters, cloned between the EcoRI and Bglll sites in place of P1. The PCR 20 primers used to create these promoter fragments are illustrated in figure 1.d. The pTREX1A plasmid represents an altemative to pTREX1, in which the Xbal bounded expression cassette is replaced by a Xbal-Spel T7-expression cassette fragment from pET3A. The SD sequence is therefore not optimised for Lactococcus. The complete pTREX1A sequence and schematics illustrating the expression cassette are illustrated in Figure 1.

S: The pKS90 plasmid is a derivative of pTREX1, which while utilising the same transcription promoter P1, has a modified TIR containing the SD and first 20 codons of a lactococcal chromosomally derived resolvase gene, P11, (Nick. R. Waterfield, R.W.F. Le Page, P.W. Wilson, and J.M. Wells. Gene. 165. 1995. This was achieved by cloning an artificial DNA sequence, 30 containing the new TIR, between the Bglll and BamHI sites of pTREX1. The oligonucleotide sequences used in this procedure are presented in figure 2. Through the careful choice of PCR primer sequence, a PCR derived target gene can be translationally coupled to this TIR, which has been shown to increase expression level when compared to that obtained from the pTREX1 parent plasmid. The complete pKS90 sequence and a schematic illustrating the expression cassette are illustrated in Figure 2.

Construction of recombinant strains of Lactococcus lactis which secrete active 3 toxin: The pJF2000 DNA (as supplied by J. Frey, Institute for veterinary bacteriology, University of Berne, CH-3012 Berne, Switzerland) comprising a gene encoding the p-toxin as published by Hunter et al. (Infect. Immun. 61: 3958-3965 (1993)) was electroporated into E. coli SURE. Plasmid DNA from six independent transformants was isolated, examined by restriction digest and pooled. These [R:LIBCj424760d1.doc:mef strains were stored at -70C in 15% glycerol. The cpb gene with associated translation initiation region (and upstream RNA stabilising stem-loop) was excised from this pooled preparation of plasmid DNA by digestion with Xbal/BamHI or Smal/BamHI. The purified cpb gene fragments were ligated into the multiple cloning sites of the expression vectors pTREX1A, pTREX14 and pTREX7, which direct high, medium and low levels of constitutive expression respectively. Xbal/BamHI ended cpb fragments were ligated into Xbal/BamHI digested pTREX1ANEW while the Smal/BamHI ended cpb fragment was ligated into BamHI/Bglll (blunted) pTREX14 and pTREX7 DNA. These constructs were successfully electroporated into the specially disabled strain Lactococcus lactis pep5- acmA-, which lacks a chromosomally encoded peptidase, pep5, which enables the organism to utilise peptides liberated from casein. In addition it also lacks a cell wall associated autolysin, acmA. Any other Lactococcus lactis strain is however suitable for this purpose. Ten correct isolates of each type were identified using restriction digest analysis and have been stored at -70oC in 15% glycerol. These strain types are designated L. lactis pep5- acmA- [pTREX1Acpb], L. lactis pep5- acmA- [pTREX14cpb] and L. lactis pep5-acmA- [pTREX7cpb]. In addition, the cpb gene was amplified by PCR using Vent DNA polymerase and BamHI ended oligonucleotides from plasmid pJF2000. This PCR product was cloned into the BamHI site of the TIR coupling vector Total protein extracted from mid-log cells and supernatant was resolved using SDS-PAGE before transferring to nitrocellulose filters. These filters were probed with both monoclonal and polyclonal antibodies to Cpb, using MAB (monoclonal antibody) purified beta-toxin as a control. A *i 20 single protein product of the correct size (approximately 30kDA) was detected in the culture supernatants from all four strains, by western blot and Coomasie Blue staining of SDS gels. The amount of beta toxin secreted by each strain was consistent with the relative strengths of the promoters in the expression vectors used. No intracellular toxin could be detected, suggesting that the clostridial export sequence functions efficiently in Lactococcus. Filtered supernantants from the pTREX1A expressor, L. lactis MG1363 pep5- acmA- and the pKS90 expressor L. lactis SMG1363 pep5- acmA- [p5-123], were shown to contain active toxin, providing a positive control for the recombinant p-toxin.

The cpb mutant genes.

Mutant cpb genes were created by in vitro gene construction using PCR fragments amplified by 30 Vent DNA polymerase from the original template cpb gene (cloned from C. perfringens type C) present in plasmid pJF2000 (as supplied by J. Frey). Primers used in this construction, and an "example of the construction procedure used, are given in figure 3. The inclusion of BamHI sites on the 5' ends of the terminal cpbF2 and cpbR PCR primers allowed these mutants to be cloned into the BamHI site of the pKS90 expression plasmid (see above). The choice of PCR primers incorporating a BamHI site suitably spaced to the cpb ATG start codon, allowed the mutant cpb genes to be translationally coupled to the P11 TIR in pKS90. This method was used to generate a series of mutant expression strains. The exact sequences of these mutant genes (after cloning into were determined by automatic sequencing with 3-4 fold redundancy, which confirmed both the gene sequences and location in the expression plasmids. Cloning the mutant genes in these vectors allowed secretion of the gene product into the culture supernatant by virtue of the native cpb N- [R:lBC]424760d1.doc:mef 1I 11 terminal export leader. The culture supematant was filter-sterilised and used directly in animal bioassays.

Table 1: mutant gene constructs.

Construct Mutation pM1-4 D 8 oDK AsoAA pM1-10 D80DK AsoAA

Y

182

H

182 pM2 T 95 GF D 95

DD

pM3-3 KloiKE AlolAA

T

197

A

197 pM3-69 K 1 ioKE AioiAA pM4-3 P 1 4 5 KN D 1 45

ED

pM4-18 V 6 2 162

P

1 4 5 KN D 1 45

ED

pM6-1 C29 2

A

2 9 2 (pM6-7) pM6-16 C292 -A292 A'29 2

NWNGA

pM7-15 E 2 87RER Q28 7

QQQ

Mutant expression strains: The pKS90-based toxoid expression plasmids are propagated in the specially disabled Lactococcal host strain, L. lactis MG1363 pep5- acmA-, but other Lactococcus strains can also be used. This strain represents a prophage and plasmid cured strain that is highly auxotrophic and S. unable to survive in the natural milk environment. In addition it grows at a slower rate than other lactococcal strains. Non-toxic expression strains constructed as indicated above are represented in Table 1. In this table, it is indicated which amino acids were mutated and where the mutations are located. The toxoid expression strains can be grown in M17 medium (Difco cat. number 1856-1 7) supplemented with 0.5% w/v glucose and 5g/mL erythromycin. Aeration is not required, indeed anoxic conditions are preferred although not essential. Optimal growth temperature is 30*C. The various toxoid proteins produced by all these strains have been shown to accumulate in the culture medium to levels less than 5g/mL. Western blot analysis using both monoclonal and polyclonal antibodies confirmed that protein monomer of the correct size is secreted into the supernatant by all the named expressor strains listed above. N-terminal sequencing of an example mutant protein, M4(4-3), confirmed that the leader peptide is correctly cleaved during export from the lactococcal 2. cytoplasm.

20 Protocol for random mutagenesis of the cpb gene: Typical PCR reaction mix is set up as follows: Tris-HCI (pH8.7 at 250C) KCI, 51g/mL bovine serum albumin M of each of the PCR primers (cpbg2; 5' -ggaggatccaaatgaagaaaaaattatttcatattagttatag-3') (cpbR. 5' -atggatccgtctaaatagctgttactttgtgag-3') 4mM MgCl2 [R:LIBC1424760di.doc:mef MnCl2 3.4mM forcing dNTP and 0.2mM other 3 dNTPs (dNTP concentration may be ranged between 0.1 and 600pM of template DNA (pJF2000 cpb clone or other cpb derivative) 2U Taq DNA polymerase Thermocycle conditions were as follows (94*C for 30s, 50 0 C for 60s, 72 0 C for 180s) 30 cycles, followed by incubation at 72 0 C for 600s.

This reaction was performed using each of the 4 dNTPs as the forcing nucleotide and then pooled.

This gives a suitably random mix of substitutions. The pooled PCR fragments were then BamHI digested and ligated into BamHI digested pKS90. This ligation mix was transformed into Lactococus lactis MG1363. Transformants were grown in liquid culture and the filtered supernatants tested in the in vitro beta-toxin assay described in Example 3.

Example 2: Culture of Lactococcus lactis: L. lactis producing gene modified p-toxin was grown in M17 medium (eg. Difco cat. 1856-17), supplemented with 0.5% w/v glucose and 5.g/mL erythromycin. Aeration is not required, anoxic conditions are preferable although not essential. Optimal growth temperature is 30 0 C. Addition of glucose is necessary during the exponential growth.

Antigen preparations: The culture supernatant was separated from the cells by means of centrifugation. After sterile filtration of the supernatant the derivatives were precipitated with ammonium sulfate The precipitate was resuspended in PBS. As an alternative the culture supernatant can be concentrated on a filter <10 000 Da.

Mouse lethality: Derivatives of P-toxin produced in L. lactis were initially tested by injection of sterile filtered culture supernatant, intraperitoneally, 0.5mL in NRMI mice weighing approx. 20g. Derivatives pM 6-1, 6-5, and 6-7 were lethal at a level similar to that of the native toxin. The derivatives pM 6-1, 6-5, and 6-7 have been modified in amino acid pos. Cys292 Ala292.

Table 2: Mouse lethality test Derivative dead/total (approx. in ig/mL) Sterile filtered Supernatant (NH 4 2

SO

4 conc. Supernatant pM 1-4 0/5 (5W/mL) 2/2 (20 /mL) pM 1-10 0/5 (5 2/2 (20 i/mL) pM 2 0/5 (1 a/mL) 1/2 g/mL) pM 3-3 0/5 (5 g/mL) 1/2 (30 /mL) pM 3-64 0/5 (5 /mL) 2/2 (30 g/mL) pM4-3 0/5 (50/mL) 1/2 (30 g/mL) pM4-18 0/5 (1.5 /mL) 0/2 (301g/mL) pM 6-1 5/5 (0.5W/mL) 2/2 (50 /mL) pM 6-5 5/5 (0.5 /mL) 2/2 (50 /mL) pM 6-7 5/5 (5 /mL) 2/2 (50 p/mL) pM 6-16 0/5 (5 0/2 (50 p/mL) M 7-15 0/5 (5 ig/mL) 2/2 (40 m/mL) [R:LIBCJ424760d1.doc:mef I 13 In subsequent experiments the supernatants containing derivatives were precipitated with ammonium sulfate. Each derivative-containing sediment was resuspended in PBS in a 10x smaller volume. Each derivative was injected intraperitoneally mice. The derivatives pM 1-4, pM 1-10, pM 3- 69 and pM 7-15 are all non-toxic at levels at which the wild-type toxin is lethal. They only become toxic at very high concentrations: 20-40pg/mL. The derivatives pM 4-18 and pM 6-16 are also nontoxic at levels at which the wild-type toxin is lethal. They are not even toxic at even higher concentrations: 30-50 Ig/mL. The remaining 3 derivatives pM 2, pM 3-3 and pM 4-3 were slightly toxic.

(The LDso of the wild-type toxin is 0.15 g/mouse). See also Table 2.

Dermonecrosis in guinea pigs: Albino guinea pigs of own breed, 300g, were depilated in an area of 4 x 8cm on both sides.

0.2mL of extract was injected intradermally at 4 sites on each side of the animal. The animals were examined after 24, 48 and 72h and the lesion at the injection sites was scored. Obvious necrosis at the injection site severe erythema erythema reddening no reaction The derivatives pM 1-4, 1-10, 3-3, 4-3, 6-16 and 7-15 are non-dermonecrotic even at very high concentrations. The derivatives pM 6-1, 6-5, and 6-7 were dermonecrotic at a level of 51g/mL. They were not dermonecrotic at 0.51/mL. After concentration the derivative pM 3-69 turned out to be dermonecrotic in dosages of 6.g/mL, which corresponds to at least 60x less dermonecrotic than the native toxin. Derivative pM 4-18 has a 15x less dermonecrotic potential than that of the native toxin.

Derivative pM 2 has not shown any dermonecrotic action by the tested concentrations of 1 /.gmL and 20 2g/mL respectively. Therefore pM2 is at least 20x less toxic than the native toxin.

See also Table 3.

Vaccine preparation: The antigen preparations were adjuvanted with Freund's incomplete adjuvant or Alhydrogen respectively.

Table 3: dermonecrosis in pigs.

C

C. Derivative Sterile filtered supernatant (NH 4 2

SO

4 concentr. Supernatant 1-4 0 (5 .g/mL) 0 0 (4 p./mL) 1-10 0 (5 g/mL) 0 (103/mL) 0 (4g/mL) 2 0 (1 W/mL) 0 (2W/mL) (1 g/mL) 3-3 0 (5.g/mL) 0 (151W/mL) 0 (61/mL) 3-69 0 (5 .mL) (6 ./mL) 4-3 0 (5g/mL) 0 (5 .g/mL) 0 (5 W/mL) 4-18 0 (1.5 /mL) (151/mL) 0 (6 g/mL) 6-1 (5/.gmL) n.d.

0 (5p./mL) n.d.

0 (0.5 /mL) 6-7 (5 /mL) (10 /mL) (2.5 I.mL) (R:LIBC]424760d1.doc:mef

I

Derivative Sterile filtered supernatant (NH 4 2 S04 concentr. Supernatant 6-16 0 (0.5pg/mL) 0 0 (20 /mL) 7-15 0 (5 0 (20 /mL) 0 (5.g/mL) 0 (8g/mL) CpC3 L/mL) 0 no reaction reddening erythema severe erythema obvious necrosis at injection site CpC p Clostridium perfringens type C P-toxin Immunisation of pigs: The derivative 3-69 was prepared in a concentration of approx. 40[g/mL and adjuvanted to a final concentration of 20/glmL. The pigs were given 2mL intramuscularly at 2 weeks interval. Blood samples were taken before 1st vaccination, before 2 n d vaccination and 2 weeks after 2 nd vaccination.

The pigs were observed for side effects 24h after 1 st vaccination. No side effects were observed.

Example 3: Quick screening test for the detection of non-toxic derivatives of p-toxin obtained from randomly mutagenised p-toxin-genes: Procedure for BT-cell test: A culture of trypsinated BT (Bovine Turbinate cells) cells, 1.5x105 cells/mL, in Eagles medium with 5% calf serum, was transferred to a 96-wells microtiter plate, 1001L/well. The plates were incubated for 24h at 37°C and 3%CO 2 Samples of 0-toxin to be tested were diluted in 20°C PBS 2 pH7.0 in dilution plates. The medium was cautiously discarded from the cells and 1001- of p-toxin 20 dilutions were applied per well. The cells were then incubated for 30min at 37 0 C and 3%C02. The samples were cautiously discarded and fresh Eagles medium with 5% calf serum tempered to 37°C was added. The cells were incubated for three days under the same conditions as mentioned above.

Then the medium was discarded from the microtiter plates and Giemsa staining solution S5/well was added. After 10min the staining agent was discarded and the plate was washed 3x with water and 25 allowed to stand for at least 20min for drying. All wells were checked for the presence of multinucleate cells and irregularities in the monolayer. Lack of these signs in a specific well is indicative for the presence in that well of a non-toxic derivative p-toxin. Those 0-toxin derivatives that proved non-toxic were tested for their immunogenicity in the semi-quantitative ELISA test described below.

30 Semi-quantitative ELISA: Highly specific antibodies raised against native Clostridium perfringens p-toxin in rabbits were purified on a protein A column and ELISA plates were coated with approximately 2g/mL of these specific antibodies. The supernatants containing the p-toxin derivatives were added to the plates in 3fold dilutions. After incubation for 1h, horseradish-peroxidase conjugated polyclonal rabbit anti-C.p.ptoxin antibodies were used for detection. Those P-toxin derivatives that reacted in the ELISA to a titre not less than 1000 times lower than the native p-toxin were considered to be the desired non-toxic and still immunogenic derivatives of p-toxin according to the invention.

[R:lBC]424760d1.doc:mef Example 4: Vaccination experiments in pigs.

Preparation of vaccine.

Derivative pM 1-4 was produced by Lactococcus lactis strain pM 1-4. After fermentation the cells were removed by centrifugation (100000xg for 35min). Then, 0.31g ammonium sulfate was added to each g of supernatant. The mixture was placed at 4°C for 3h for precipitation. After centrifugation at 100000xg for 45min the precipitate was dissolved in 1/30 of the original volume and dialysed against PBS. The final content of derivative pM 1-4 was quantified by SDS-PAGE relative to BSA standards. Finally the derivative solution was mixed with Diluvac Forte® in a ratio of 1:1. The final concentration of pM 1-4 in the vaccine was 80 pg/mL.

Vaccination Five pigs aged 5 weeks were used. The pigs were vaccinated with 2mL intramuscularly on days 0 and 21.

Observations The animals' general condition was recorded during the initial three hours after vaccination according to the following scale: 0 no symptoms 1 mildly depressed, rough coated (for less than 1 hour) 2 depressed, shivering (less than 1 hour) 20 3 depressed, shivering, lying down and limited water intake (less than 1 hour) 4 vomiting, severely depressed (less than 1 hour) 5 depressed for more than one hour and does not eat the following feeding Blood sampling and determination of titres Blood samples were taken before 1 st vaccination (day before 2nd vaccination (day 21) and S 25 two weeks after 2 n d vaccination (day 35) and analysed for anti-p-toxin antibodies by competitive ELISA. Briefly described, monoclonal antibody raised against native p-toxin was coated in an ELISA plate. Serial dilution of sera was allowed to react with a fixed concentration of the p-toxin on a separate plate. The mixtures of toxin and antibodies containing sera were transferred to the Mab coated plate and remaining native p-toxin contents were determined. The titres were calculated as S 30 50% inhibition of uninhibited p-toxin, relative to WHO antitoxin standard 1959.

Results: Observations The pigs did not show any clinical reaction due to any of the vaccinations. See table 4 Serology 35 One pig (no. 52) responded to 1 st vaccination. After the second vaccination all four pigs seroconverted and the average anti-p-toxin titre was 66i.u. measured in ELISA. Individual titres appear from Table 4.

Table 4: Anti-p-toxin antibody response in pigs Anti-p1toxin antibody response in pigs measured in ELISA in i.u. and score in test for nonspecific toxicity.

Pig No. Pre-immune sera After 1stVaccination After 2n d vaccination Non-spec. tox score 51 0 0 159.6 0 52 0 2.4 55.1 0 54 0 0 32.2 0 0 0 17.3 0 (R:.IBC]424760d1.doc:mef Conclusion All pigs responded to vaccination with genetically modified p-toxin by producing p-toxininhibiting anti-p-toxin antibodies.

Legend to the figures: Figure 1.a: structure of plasmid pTREX1ANEW. The coding sequence of the MLS antibiotic resistance marker is shown as a shaded box. Ori pAMB1: origin of replication from the plasmid pAMB1, MLS: macrolides, lincosamindes, streptogramin B antibiotic resistance gene. The structure of the expression cassette cloned between the EcoR1 and blunted Sad sites of plL253 is shown below.

Different functional regions of the cassette are shown as hatched boxes. Unique restriction sites are indicated. SD: Shine Dalgarno sequence complementary to L. lactis 16S rRNA, ATG: translation initiation codon served by SD.

Figure 1.b: the complete pTREX1A sequence, including restriction enzyme positions.

Figure 1.c: features of the highly similar pTREX1 expression cassette.

Figure 1.d: oligonucleotides used to create the PCR derived promotor fragments used in pTREX7, -14 and -1A (shown Figure 2.a: the TIR coupling vector pKS90, schematic representation of the pKS90 expression cassette.

Figure 2.b: the complete pKS90 sequence, including restriction enzyme positions.

Figure 2.c: oligonucleotides used to create the artificial DNA sequence forming the TIR region of 20 Figure 3.a: primers used in the construction of the Cpp-mutants Figure 3.b: example of the construction procedure as used for CpP M1.

Figure 4 Amino acid sequence of Clostridium perfringens p-toxin.

i* o• [R:ULIBC424760dl.doc:mef

Claims (14)

1. Expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens p-toxin.

2. Method for the preparation of native Clostridium perfringens p-toxin characterised in that a nucleotide sequence comprising a DNA fragment encoding said 3- toxin is expressed in a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium.

3. Method according to claim 2, characterised in that the Grampositive o0 bacterium is selected from the group of Lactococcus, Lactobacillus, Leuconostoc, Pediococcus, Streptococcus, Enterococcus, Staphylococcus, Bacillus, Sarcina, Ruminococcus or Listeria.

4. Expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type Clostridium perfringens p-toxin.

Expression system according to claim 4, wherein the Clostridium perfringens p-toxin has the amino acid sequence as depicted in Figure 4.

6. Expression system according to claim 4 or claim 5, wherein the produced Clostridium perfringens P-toxin is an active form. 20

7. An expression system according to any one of claims 4 to 6, wherein said Gram-positive bacterium is selected from the group comprising Lactococcus, *Lactobacillus, Leuconostoc, Pediococcus, Streptococcus, Enterococcus, Staphylococcus, Bacillus, Sarcina, Ruminococcus or Listeria. .9

8. Expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium perfringens, comprising the gene encoding wild-type *o Clostridium perfringens p-toxin, said system being substantially as hereinbefore described with reference to any one of the examples.

9. A transgenic Gram-positive bacterium comprising an expression system according to any one of claims 4 to 8. t* 30

10. A transgenic bacterium according to claim 9, wherein said Gram-positive bacterium is selected from the group comprising Lactococcus, Lactobacillus, Leuconostoc, Pediococcus, Streptococcus, Enterococcus, Staphylococcus, Bacillus, Sarcina, Ruminococcus or Listeria.

11. A transgenic Gram-positive bacterium comprising an expression system based on a Gram-positive bacterium, said Gram-positive bacterium not being Clostridium [l:\DayLib\LIBXX]04264.doc:KOB 18 perfringens, comprising the gene encoding wild-type Clostridium perfringens p-toxin, said transgenic bacterium being substantially as hereinbefore described with reference to any one of the examples.

12. Method for the preparation of wild-type Clostridium perfringens p-toxin, wherein an expression system according to any one of claims 4 to 8 is used.

13. Method for the preparation of wild-type Clostridium perfringens p-toxin, comprising culturing a transgenic bacterium according to any one of claims 9 to 11 under conditions promoting production of said p-toxin.

14. Method for the preparation of wild-type Clostridium perfringens p-toxin, said method being substantially as hereinbefore described with reference to any one of the examples. Wild-type Clostridium perfringens p-toxin prepared by a method according to any one of claims 12 to 14. Dated 13 June, 2003 Akzo Nobel N.V. 00 .4 *0 *se. *1 9 9 *9 0 4 9 90 9 9 9. *9 S 9 Patent Attorneys for the Applicant/Nominated Person SPRUSON FERGUSON 9: :o 9 9 9 *9o [I:\DayLib\LIBXX]04264.doc:KOB .o o 0 o to