WO1994008011A1 - Recombinant mutant pasteurella multocida protein and process for preparing the same - Google Patents

Abstract

Mutant recombinant proteins capable of binding to antibodies raised against Pasteurella multocida toxin and their use as vaccines conferring protection against diseases caused by Pasteurella multocida.

Description

RECOMBINANT MUTANT PASTEURELLA MULTOCIDA PROTEIN AND PROCESS FOR

PREPARING THE SAME

FIELD OF INVENTION

The present invention provides novel immunogenic proteins encoded by DNA sequences derived from Pasteurella multocida genes coding for toxins involved i.a in the development of Progressive Atrophic Rhinitis (PAR) in animals. The claimed proteins have low toxicity, they are stable and readily purifiable and are useful in vaccines for the protection of animals against the pathological effects of the Pasteurella multocida toxin (PMT).

TECHNICAL BACKGROUND AND PRIOR ART PAR is a widespread, highly contagious infectious disease in animals, in particular pigs where the infection affects the normal bone structure of the snout as a result of nasal bone resorption causing deformation of the snout, sneezing, nasal discharge and bleeding from the nasal mucosa. In the pig industry the disease causes substantial economic losses due to growth retardation and increased susceptibility to other infectious diseases.

The development of PAR is associated with the presence in the nasal cavity of the animals of strains of the gram-negative bacterium Pasteurella multocida which produce an osteolytically active toxic protein comprising 1285 amino acid residues and having a calculated molecular weight of about 146.5 kD.

In addition to its osteolytic activity, the native

Pasteurella multocida toxin (PMT) has other toxic effects including dermonecrotic effect, cytopathic effect, and mitogenic (proliferative) effect on cells. Based on conventional acute toxicity testing such as e.g. measurement of the dosage which will kill 50% of small laboratory animals

(LD₅₀), the PMT is highly toxic, the LD₅₀ value in mice being reported to be in the range of 10-70 ng.

The Pasteurella multocida toxin is immunogenic and accordingly, the administration of the protein to an animal will provoke the formation of antibodies, which may protect the animal against the pathological effect of this toxin-producing organism. However, due to the high toxicity of the native toxin, it can not be utilized as such as a vaccine component. Accordingly, known vaccines for the immunization of animals, including pigs, against the effects of Pasteurella multocida infections, comprise as the active component killed toxin-producing Pasteurella multocida cells and/or inactivated (detoxified) toxin-containing extracts of toxin-producing Pasteurella multocida. Such inactivated vaccines are commercially available. The inactivation of PMT in such vaccines is typically produced by heat treatment or by the treatment with an aldehyde, including glutaraldehyde and formaldehyde, both of which treatments cause a denaturation of the proteinaceous toxin.

However, it is known that heat or aldehyde detoxification of immunogenic proteins may, as an undesired side effect, result in a reduced or altered immunogenicity presumably due to largely unpredictable modifications of the structure of the protein, including conformational modifications of one or several epitopes. Notably, such structure modifications of the PMT may result in a preferential stimulation of production of antibodies with other specificities and possibly, in an increased susceptibility to proteolytic degradation.

Furthermore, aldehyde detoxification of PMT results in a residual toxicity which is typically about 1%. Therefore, attempts have been made to provide detoxified PMT- based vaccines, which have not been subjected to the above traditional detoxification treatments. In WO89/09617 is disclosed recombinantly detoxified but still immunogenic derivatives of PMT. Essentially, these PMT derivatives are prepared by isolating a PMT-encoding gene and deleting a relatively large fragment of the gene by treating the DNA with a restriction enzyme, inserting the thus truncated gene into a suitable host cell and having it expressed as a truncated form of the native PMT. Several PMT-derivatives of a molecular weight in the range of 53-136 kD have been produced which show a reduced toxicity and at the same time have preserved a substantial immunogenicity. Typically, the truncated protein comprises 50 - 500 amino acid residues less than the native PMT. One preferred PMT-derivative disclosed in this document is derivative O.

An important characteristic of a suitable immunogenic component of a vaccine is its stability to degradation of the primary, secondary and/or tertiary structure, respectively. Degradation of an immunogen may be produced by chemical and/or enzymatic hydrolysis, or by physical or chemical factors present during the manufacturing or storage of the vaccine and may affect the primary, secondary and/or tertiary structure in such a manner that the immunogenicity of the component is reduced or altered momentaneously or gradually.

In addition to its adverse effect on the immunogenicity of a protein intended for use as a vaccine, the above degradation may also be industrially undesirable due to lack of uniformity of the protein molecules present in a vaccine composition comprising the immunogen having been exposed to degradation. Such a lack of molecule uniformity makes it difficult to produce standardized vaccine compositions.

A further major requirement in cost-effective Pasteurella multocida vaccine production is that the immunogenic component is readily purifiable from the culture of cells pro ducing the component. It is particularly desirable that the component can be isolated in a substantially pure form by an uncomplicated one-step purification procedure such as a conventional anion-exchange chromatography step. It has now been found that highly immunogenic PMT toxin- related proteins having a low toxicity can be provided by site-directed (site-specific) mutagenesis of PMT-encoding DNA sequences such as e.g. the toxA gene, including the toxA gene coding for Pasteurella multocida ssp multocida 45/78 toxin, which proteins are more resistant to degradation and more readily purifiable on a commercial scale than the known recombinantly detoxified PMT derivatives.

SUMMARY OF THE INVENTION

Accordingly, in one aspect the present invention relates to a recombinant mutant protein encoded at least in part by a

DNA sequence which is derivable from the toxA gene, including the toxA gene coding for Pasteurella multocida ssp multocida 45/78 toxin (PMT), by substitution, deletion or insertion of one or more codons, said protein being capable of binding to antibodies raised against PMT encoded by Pasteurella mul tocida ssp mul tocida 45/78, the protein having a calculated molecular weight of at least 140 kD.

In another aspect, there is provided a process of preparing a recombinant mutant protein capable of binding to antibodies raised against the Pasteurella multocida ssp mul tocida 45/78 toxin and having a molecular weight of at least 140 kD, comprising the steps of:

(i) isolating a DNA sequence comprising a toxA gene coding for a Pasteurella multocida osteolytic toxin

(PMT), including the toxA gene coding for Pasteurella mul tocida ssp multocida 45/78 PMT, or a DNA sequence comprising a sequence, the gene product of which is reactive with an antibody raised against Pasteurella multocida ssp multocida 45/78 PMT,

(ii) subjecting said DNA sequence to a mutagenization treatment causing substitution, deletion or insertion of one or more codons to obtain a mutated DNA sequence coding for the recombinant mutant protein,

(iii) inserting the resulting mutated DNA sequence into a replicon,

(iv) transforming with the replicon a cell in which said replicon is replicated and in which the mutated DNA sequence coding for the recombinant mutant protein is expressible,

(v) culturing the transformed cell under conditions where the mutated DNA sequence is expressed, and (vi) harvesting the mutant recombinant protein from the culture.

The present invention relates in a further aspect to a DNA sequence comprising a gene coding for a protein having a molecular weight of at least 140 kD which is reactive with an antibody reacting with the Pasteurella mul tocida spp multocida 45/78 toxin (PMT) encoded by the toxA gene, said DNA sequence being derived from a replicon comprising the toxA gene, by substitution, deletion or insertion of one or more codons of said toxA gene or by in-frame fusion of a truncated or mutated toxA gene to another gene in a manner that results in the expression of a fusion protein gene product as defined herein.

In still further aspects, the invention provides a replicon harbouring the DNA sequence as defined above, and a cell which is transformed with such a replicon in which cell said replicon is replicated. In an interesting aspect, the present invention relates to the use of the mutant recombinant protein as defined herein as a vaccine for the protection against Pasteurella mul tocida infections.

DETAILED DISCLOSURE OF THE INVENTION

The complete nucleotide sequence of the Pasteurella multocida toxin (PMT) gene has been disclosed in WO 89/09617 (supra) and by Petersen (Molecular Microbiology, 1990, 4 , 821-830). The toxin-encoding gene, designated as the toxA gene, was shown to be chromosomal, located on a DNA fragment (stretch) larger than 20 kb and unique to the toxin-producing strains of Pasteurella multocida. A nucleotide sequence of 4381 bp of Pasteurella multocida DNA encompassing the entire toxin-encoding region and the native promoter is shown in Figure 2 in Petersen ( supra) and this sequence will appear in the EMBL/GenBank/DDBJ Nucleotide Sequence Databases under the accession number X51512.

The native toxA gene codes for a PMT protein of 1285 amino acids with a calculated molecular weight of 146,553 daltons which is in accordance with the apparent molecular weight of the protein (143 kD).

As mentioned above, the toxA gene is found in toxin-producing strains of Pasteurella multocida and accordingly, the present invention provides a recombinant mutant protein encoded by a DNA sequence which is derivable from such a toxA gene, including the toxA gene isolated from Pasteurella mul tocida ssp. multocida 45/78. In the following, the abbreviation "rmPMT" is used to designate the recombinant mutant PMT protein as defined herein. The rmPMT-encoding DNA sequence may be derived from a native chromosomal toxA gene, or from another replicon comprising the gene such as a recombinant plasmid. Such plasmids which are useful as starting materials in the construction of DNA sequences encoding rmPMT include as examples pSPE312, pSPE680 and pSPE1003 as shown in Fig. 1. The pSPE1003 plasmid having a size of about 7 kb has a unique EcoRI restriction site which facilitates the construction of rmPMT-encoding DNA sequences, in which a modification relative to the native PMT-encoding sequence by the substitution, deletion or insertion of one or more codons, is carried out downstream of this unique restriction site. Accordingly, in one preferred embodiment, the present invention provides a rmPMT which is encoded by a DNA sequence which is derivable from the toxA gene by substitution, deletion or insertion of one or more codons downstream of a unique EcoRI restriction site in the gene or by in-frame fusion of a truncated or mutated toxA gene to a second gene coding for an immunogenic protein, in a manner that results in the expression of a fusion protein gene product as defined herein.

Although it has been disclosed in WO 89/09617 that truncated PMT-derivatives may be isolated which have a calculated molecular weight in the range of 53-139.337 kD and which have preserved the capability of eliciting an immune response in animals protecting against Pasteurella multocida-related diseases, including Progressive Atrophic Rhinitis, it is contemplated that vaccines may be produced which are more advantageous with regard to immunogenicity, stability and readiness of purification, by the incorporation herein of immunogenic non-denatured PMT-related substances which have a calculated molecular weight close to that of the native PMT such as at least 140 kD.

In accordance herewith, the present invention provides in one preferred embodiment a rmPMT which has a calculated molecular weight of at least 141 kD. A rmPMT of a molecular weight of at least 142 kD may be even more preferable, such as one of a calculated molecular weight of at least 143 kD, e.g. of at least 144 kD or particularly one of at least 145 kD.

One significant feature of the recombinant mutant proteins provided herein is the readiness with which they can be recovered and/or purified from suspensions in which they are present. Such suspensions include culturing media in which host cells comprising the toxA-derived gene encoding the protein have been grown under conditions where the gene is expressed, and extracts of host cells in which such a toxA- derived gene has been expressed.

In the present context, the term "readiness" implies that a commercially significant proportion such as e.g. at least 50% of rmPMT present in a suspension can be recovered and/or purified by inexpensive, conventional industrial purification procedures, preferably in a single purification step which results in a rmPMT preparation which substantially does not contain other proteins than the rmPMT in a substantially non- degraded form, or any undesired non-protein substances.

Preferably, however, a higher proportion of the rmPMT present in the suspension should be recoverable such as at least 75% and more preferably at least 90%. Accordingly, any suitable cost-effective conventional purification procedure, which results in a high recovery rate of a substantially pure rmPMT preparation as defined above, may be selected. It is particularly important that the purification procedure selected is one which results in a rmPMT preparation essentially without the presence of proteolytically active enzymes e.g. originating from the host cells in which the rmPMT-encoding gene is expressed. If such enzymes are present in the preparation, it may be prone to enzymatic degradation of rmPMT molecules which are degradable by the enzymes, the implication hereof being that the preparation may be

unstable, i.e. a gradual degradation by time of the originally expressed protein molecules will occur, assumingly resulting in reduced or altered immunogenicity of the rmPMT and, if present in a vaccine composition, in a lack of vaccine standardization.

Preferably, a rmPMT preparation resulting from the selected purification procedure is one which, when stored at a temperature below 10°C, has a stability against degradation which over a period of time of at least 3 months, results in the preservation of at least 90% of the rmPMT molecules as expressed in the host cell. More preferably, at least 95% of molecules are preserved, even more preferably at least 97% and in particular at least 99%.

In preferred embodiments of the invention, the rmPMT is one which is at least 80% purifiable. In this context, the term "purifiable" is used to indicate that it is possible by a given purification procedure to obtain a preparation of the rmPMT which contains this protein in the indicated proportion, calculated on the total content of protein in the preparation resulting from the purification procedure. In accordance with the above definition of the readiness with which the rmPMT is purifiable, a preferred protein is one which is at least 90% purifiable such as at least 94% or even more preferred at least 95% purifiable.

As one example of a suitable purification procedure which may result in a desired degree of rmPMT purity, a purification procedure comprising anion-exchange chromatography using Q-Sepharose Fast Flow (Pharmacia) in a 2 cm² × 25 cm chromatography column, applying to the column about 200 mg of total bacterial protein in the form of an extract of bacteria expressing the protein, and the flow rate being about 30 cm/h, and eluting absorbed protein with a linear gradient of buffer containing 0 to 1000 mM of NaCl, if necessary followed by hydrophobic interaction chromatography of eluted fractions using a column of Phenyl Sepharose CL-4B (Pharmacia) having the dimensions of 2 cm² × 20 cm and at a flow rate of 30 cm/h, eluting absorbed protein, and measuring the amount hereof, may be used. As mentioned above, the native PMT comprises 1285 amino acids. In one useful embodiment, the present invention provides a rmPMT as defined herein, which is a protein encoded by a DNA sequence derived from the toxA gene by substituting at least one codon coding for an amino acid with a

codon/codons coding for a different/different amino acid(s) or by deleting or inserting at least one codon. In the present context, the term "substituting a codon" may designate the exchange of one, two or all three base(s) in a codon present in the native toxA gene. It will also be understood that the term "a codon" as used herein may be more than one particular triplet of nucleotides as long as the triplet codes for the same amino acid. In accordance with the invention, substitution(s) of codons may be carried out at any amino acid position which result(s) in a rmPMT as defined herein. In the present context, the term "amino acid position" is to be understood as the codon corresponding to the position of a particular amino acid. In the following, the term "amino acid" is occasionally used as a designation also denoting the commonly used term "amino acid residue". In certain preferred embodiments of the invention, the codon substitutio(n(s) is/are carried out at the 3' end of the gene encoding the protein e.g. resulting in change(s) at an amino acid position or positions between position 1131 and 1285, such as at a position/positions between positions 1175 and 1285, e.g. between positions 1200 and 1230.

In order to obtain a sufficiently reduced toxicity of the resulting recombinant mutant PMT, it may be advantageous to substitute two or more codons of the toxA gene. In useful preferred embodiments, at least three codons are substituted such as e.g. four codons. It will be understood that a substitution as defined above, comprising two or more codons, may imply the substitution in the rmPMT of an identical amino acid at two or more positions or the substitution of several different amino acid residues. Furthermore, multiple amino acid substitutions may be in the form of one continuous stretch of amino acids, or in the form of several spatially dispersed substitutions.

In addition to the obtainment of a low toxicity rmPMT, a multiple substitution of codons in the toxA gene may result in a more safe immunogenic PMT-related protein, since the probability to have a mutational reversion to the wild-type toxic protein is reduced exponentially with increasing number of substitutions. In the following, a rmPMT obtained by codon substitution is designated by one or several indications comprising the one- letter code of the amino acid residue being substituted and of the replacing amino acid and an indication of the amino acid position where the substitution has taken place. Thus, as examples a protein designated SV1201 indicates a rmPMT where a serine residue is replaced by a valine residue at the position 1201 and a rmPMT designation of

HL1202HY1205HY1223HY1228 is used to indicate a recombinant mutant PMT in which histidine residues occurring in the wild-type PMT have been substituted (replaced) by a leucine residue at position 1202 and by a tyrosine residue at the positions 1205, 1223 and 1228, respectively (in the following the HL1202HY1205HY1223HY1228 rmPMT is also being referred to as 4HX). In accordance with the invention, the rmPMT may comprise any amino acid residue substitution(s) relative to the native PMT which result(s) in a rmPMT as defined herein. Thus as

examples, useful rmPMTs may be obtained by substituting a codon or codons coding for an amino acid selected from serine, histidine, glutamine and threonine. The substituting amino acid(s) may be any amino acid(s) resulting in a rmPMT as defined herein, such as e.g. valine, leucine or tyrosine.

Besides being obtained by amino acid residue substitution as defined above, the recombinant mutant PMT may, in accordance with the invention, be one which is encoded by a DNA sequence derived from the PMT-encoding toxA gene by the deletion of at least one codon coding for an amino acid. Deletion(s) of a codon/codons may be carried out at any amino acid position which result(s) in a rmPMT as defined herein.

However, in certain preferred embodiments of the invention, the codon deletion(s) is/are carried out at the 3' end of the toxA gene encoding the native PMT protein e.g. at an amino acid residue position or positions between position 1131 and 1285, such as at a position/ positions between positions 1175 and 1285, e.g. between positions 1215 and 1285.

Although rmPMTs having a sufficiently reduced toxicity may be obtained by deleting one codon such as the rmPMT ΔH1223 as defined herein, it may be advantageous to have two or more codons deleted to further reduce the toxicity of the rmPMT. In useful preferred embodiments, at least three codons are deleted such as e.g. at least four codons. It may even be preferred to use a rmPMT in which at least seven amino acids are deleted relative to the native PMT. In addition to the obtainment of low toxicity rmPMT, a multiple deletion of basepairs in the toxA gene may result in more safe

immunogenic PMT-related proteins since the probability to have a mutational reversion to the wild-type toxin-encoding gene is reduced exponentially with increasing numbers of basepair deletions.

In the following, a rmPMT obtained by codon deletion(s) is designated by the symbol "Δ" followed by the amino acid position(s) of the native PMT where an amino acid or amino acids have been deleted. Thus, as an example, a protein designated Δ1215-1221 indicates a rmPMT encoded by a toxA-derived gene wherein codons coding for the amino acids at positions from 1215 to 1221 of the native toxA gene product inclusive, are deleted. The rmPMT according to the present invention may comprise any amino acid deletion(s) relative to the native PMT which result(s) in a rmPMT as defined herein. Thus as examples, useful rmPMTs may be obtained by deleting from the toxA gene a codon or codons coding for an amino acid selected from histidine, lysine, glutamic acid, phenylalanine, alanine, valine and aspartic acid. It will be understood that a deletion as defined above, comprising two or more codons, may imply the deletion in the rmPMT of an identical amino acid at two or more positions or the deletion of several different amino acids. Furthermore, multiple amino acid deletions may be in the form of one continuous stretch of amino acids, or in the form of several spatially dispersed deletions.

In addition to being obtained by amino acid substitution or deletion as defined above, the recombinant mutant PMT may in accordance with the invention also be one, which is encoded by a DNA sequence derived from the PMT-encoding toxA gene by the insertion of at least one codon coding for an amino acid residue or amino acid residues. Insertion(s) of a codon or codons may be carried out downstream of any codon corresponding to any amino acid position(s) which result(s) in a rmPMT as defined herein.

However, in certain preferred embodiments of the invention, the codon insertion(s) is/are carried out at the 3' end of the toxA gene e.g. downstream of an amino acid position or positions between position 1131 and 1285, such as downstream of a position/ positions between positions 1175 and 1285, e.g. downstream of a position between positions 1200 and 1230. In order to obtain rmPMTs having a sufficiently reduced toxicity, it may be advantageous to have two or more codons inserted, relative to the native PMT-encoding gene. In useful embodiments of the invention, at least three codons are inserted such as e.g. at least four codons. In addition to the obtainment of low toxicity rmPMT, a multiple insertion of codons in the toxA gene may result in more safe immunogenic PMT-related protein, since it is contemplated that the probability of having a mutational reversion to the wild-type toxic protein is reduced exponentially with the number of inserted basepairs.

In the following, a rmPMT obtained by codon insertion (s) is designated by the symbol "+" followed by the one-letter code for the amino acid residue (s) encoded by the inserted codon (s) and the position of the native PMT immediately upstream of which an amino acid or amino acids have been inserted. Thus, as examples a protein designated +G1203 indicates a rmPMT encoded by a toxA-derived gene wherein a codon coding for glycine has been inserted immediately upstream of the codon coding for an amino acid at position 1203 of the native PMT, and the rmPMT designation +GG1203 indicates a recombinant mutant PMT encoded by a toxA-derived gene wherein two codons coding for glycine have been inserted immediately upstream of the codon coding for an amino acid at position 1203 of native PMT. In accordance with the present invention, the rmPMT may comprise any amino acid insertion (s) relative to the native PMT which result (s) in an rmPMT as defined herein. It will be understood that an insertion as defined above, comprising two or more codons, may imply the insertion in the rmPMT of an identical/identical amino acid(s) at two or more positions, or the insertion of two or more different amino acids. Furthermore, multiple amino acid insertions may be in the form of adjacent insertions or in the form of a multiplicity of spatially dispersed insertions. In interesting embodiments, the present invention pertains to a protein as defined herein which is a fusion protein encoded by a DNA sequence, the sequence comprising a first gene coding for an immunogenic protein which is reactive with an antibody reacting with the Pasteurella multocida spp multocida 45/78 toxin (PMT) encoded by the toxA gene, said gene being derived from a replicon comprising the toxA gene by substitution, deletion or insertion of one or more codons of said toxA gene, and a second gene coding for a second immunogenic protein, the first and the second protein being expressible as a fusion protein. The fusion may be carried out at any position of the second gene resulting in amino- terminal or carboxy-terminal fusion to the toxA-derived first gene.

The fusion protein may comprise as the second immunogenic protein, one which is inherently expressed by a pathogenic organism, or one which is an immunogenic fragment or derivative of such an inherently expressed protein. Such fusion proteins are highly useful as multiple vaccines, since they, when administered to a human or an animal such as a mammal, a bird or a fish, confer protection, not only against diseases caused by toxin-producing Pasteurella multocida, but also against diseases caused by the pathogenic organism, wherefrom the gene coding for the second immunogenic protein, is derived. It is contemplated that a fusion protein according to the invention may comprise further immunogenic proteins. The gene coding for the second immunogenic protein may be isolated from any pathogenic virus, bacterium, protozoan, fungus, yeast or parasitic organism. In this connection, genes derived from porcine pathogens are particularly interesting.

In accordance with the invention, a fusion protein as defined above may comprise as the first immunogenic protein, a protein having a molecular weight which is less than 139.337 kD, or it may in other useful embodiments be a protein which has a calculated molecular weight of at least 139.338 kD, such as at least 140 kD or at least 143 kD. One example of a fusion protein according to the invention is the fusion protein 1132" β-gal comprising as the second immunogenic protein, β-galactosidase. This fusion protein is expressed by the plasmid pSPE1134 as described in the following and it has a calculated molecular weight of about 245 kD. It has been found that an immunogenic PMT-related protein having a molecular weight of less than 140 kD such as e.g. about 126 kD or less, may be unstable during storage, possibly resulting in loss of immunogenicity, or may not be purifiable to the desired extent as defined herein for the protein according to the invention. It has been found that the purifiability of the fusion protein is improved as compared to that of the low molecular weight protein per se.

As it has been mentioned above, the toxicity of the native PMT is so high that it cannot be utilized as such as an immunogen in vaccine compositions and furthermore, its pronounced mitogenic effect would require strict environmental precautions during manufacturing. However, the present invention provides highly immunogenic toxA-related rmPMT which, relative to the native PMT protein, has a reduced toxicity, allowing the direct use hereof as the active component of vaccines protecting against Pasteurella multocida infections.

The term "toxicity" as used herein may describe any detectable adverse effect on living organisms including prokaryotic and eukaryotic cell cultures and also living animals including laboratory animals and farm animals. Thus, toxicity of PMT and PMT-related molecules may be measured by the exposure of such cells and animals to varying amounts of the molecules and measuring the adverse effect hereof in order to obtain a quantitative measure of the toxicity. By including, in defined amounts, the native PMT, optionally expressed in another organism than Pasteurella multocida (rPMT), as a reference toxic substance in a toxicity test and comparing the adverse effect hereof with amounts of a given rmPMT as defined herein, resulting in an equivalent effect, the relative toxicity of the rmPMT in question may be calculated.

Any suitable conventional toxicity test may be applied to verify the reduction in toxic effect of a rmPMT as defined herein. Such a suitable and convenient method is the measurement of the mitogenic potency as measured by the proliferative effect on mammalian cell cultures, including as examples NIH or Swiss 3T3 fibroblast cell cultures, as it is described in details in the below Examples.

In accordance with the present invention there is, in one useful embodiment, provided a rmPMT which has a mitogenic potency as determined by measuring the proliferative effect (PE) of the protein on NIH 3T3 cells, which is at the most 75%, said measuring comprising in a first step the preparation of a standard curve based on PE measurements of serial dilutions of a suspension containing rPMT and calculating the PE according to the formula

where N_x is the mean cell number in 3 dishes with sample dilution x, N_pmt is the mean cell number in 3 dishes with whole-cell extract of E. coli SPE1036 harbouring pSPE680 encoding for rPMT, said extract containing a final concentration of 20 ng/ml rPMT, and N₀ is the mean cell number in 3 dishes with E. coli DH5α whole-cell extract, said dishes having a confluent monolayer of NIH3T3 cells, and in a second step measuring the PE of a suspension of the protein at an amount in the range of 1-1000 ng/ml, and calculating the relative rmPMT concentration (as compared to the rPMT concentration) giving a certain PE, using said standard curve (cf. Figures 2-4).

In preferred embodiments of the invention, the mitogenic potency as defined above is at the most 50% relative to that of the native PMT, more preferably at the most 35%, even more preferably at the most 25%, most preferably at the most 10% such as at the most 5%. In certain advantageous embodiments, it may be preferred that the mitogenic potency is as low as at the most 2% relative to that of the native PMT, such as at the most 1% or even at the most 1 per mille.

Another conventional method of determining the toxic effect of a substance is the measurement of the above-defined LD₅₀ value in animals e.g. in mice of a body weight of about 20 g. As mentioned above, the native PMT has an LD₅₀ value being in range of 10-70 ng as measured by intraperitoneal administration to mice weighing about 20 g, of increasing amounts of the toxin, and recording the number of surviving animals. In order for a rmPMT as defined herein to be used as a safe vaccine component, the LD₅₀ value should preferably be reduced relative to a native PMT or rPMT having an LD₅₀ value of e.g. 10 ng as determined by administering the protein intraperitoneally to mice, by a factor of at least 100, more preferably of at least 500, even more preferably of at least 1000, most preferably of at least 1500 and in particular of at least 2000, which implies that a rmPMT being particularly useful in accordance with the present invention is one which has an LD₅₀ when determined by the same method, which is at least 50 μg such as e.g. at least 51.2 μg . It is contemplated that it may in accordance with the present invention even be possible to produce rmPMTs in which the LD₅₀ value as defined above, is reduced by a factor of at least 10,000, or at least 100,000. The rmPMTs as defined herein are useful as active immunogenic components of vaccine compositions for the protection of animals, particularly pigs, against the above-defined toxic effects of native Pasteurella mul tocida toxins. However, it has been found that vaccination of mice constitutes a useful model system for the testing of the immunogenicity of PMT-related proteins (Petersen et al., Infect. Irπmun., 1991, 59:1387-1393). Accordingly, such a model system may be applied as a screening system to identify recombinant mutant PMT proteins suitable for use in vaccine compositions. A modification of this system, wherein CFlxBALB/c mice were used instead of inbred BALB/c mice, was used to test immunogenicity of the present rmPMTs.

Accordingly, the present invention provides in one useful embodiment, a protein as defined herein which, when administered to mice twice at an interval of two weeks in a vaccine preparation comprising at least 1.0 μg/ml of the protein absorbed to a aluminium hydroxide gel, protects at least 80% of the mice against the lethal effect of 100 ng rPMT injected intraperitoneally two weeks subsequent to the second administration of the vaccine. When administered in a vaccine preparation as defined above, comprising 5.0 μg/ml of the

absorbed protein, the rmPMT as defined herein should preferably protect 100% of the mice.

One example of rmPMT protein which was constructed, is one having four amino acid substitutions and designated

HL1202HY1205HY1223HY1228 (in the following also referred to as 4HX). When tested for toxicity, this recombinant mutant protein had an LD₅₀ value as defined above of at least 102.4 μg and had the immunogenic effect as defined above. The protein is encoded by plasmid pSPE1038 which was deposited on 1 September 1992 with the DSM-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH under the accession number DSM 7221 in the form of a culture of E. coli DH5α strain SPE1038 transformed with the above plasmid. Other examples of rmPMTs include ΔH1223 encoded by pSPE1234, +GG1203 encoded by pSPE1020 and 1132"β-gal encoded by

pSPE1134. pSPE1234, pSPE1020 and pSPE1134 were deposited on 9 September 1993 with the DSM-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH under the accession numbers DSM 8547, DSM 8548 and DSM 8546, respectively in the form of cultures of E. coli DH5α transformed with these plasmids, except pSPE1134 which was deposited in E. coli K12 strain MC1000. As mentioned above, the present invention relates in a further aspect to a process of preparing an immunogenic recombinant mutant PMT protein having a reduced toxicity as defined above, said protein being capable of binding to antibodies raised against the Pasteurella multocida ssp mul tocida 45/78 toxin (PMT) and having a molecular weight of at least 140 kD. In certain preferred embodiments, the rmPMT may be a protein having a molecular weight of at least 141 kD such as at least 143 kD. In a first step, said process comprises the isolation of a DNA sequence comprising a toxA gene coding for a Pasteurella multocida osteolytic toxin (PMT) e.g. the gene coding for Pasteurella multocida ssp multocida 45/78 PMT, or a functional equivalent hereof, which is reactive with an antibody against PMT. The identification and the isolation of the toxA gene may conveniently be carried out as described by Petersen, 1990 (supra) . The gene may be isolated from the chromosome of a recognized toxin-producing Pasteurella multocida strain e.g. a strain isolated from an infected animal. The gene may also be isolated from a plasmid harbouring a DNA sequence comprising the gene such as from a plasmid selected from pSPE308, pSPE312, pSPE481, pSPE525, pSPE680, pSPE716 (Petersen, 1990, supra) or pSPE1003 (Fig.l), or it may be constructed by a conventional polymerase chain reaction method, or synthesized.

In accordance with the invention, the above process comprises in a second step the subjecting of the isolated PMT-encoding gene to a mutagenization treatment, optionally comprising more steps of mutagenization, causing substitution, deletion or insertion of one or more codons. One particularly suitable method of mutagenization is site-directed or site-specific mutagenesis e.g. using a modification of the Thermus aquati cus polymerase chain reaction (PCR) method as it is described by Nelson et al., 1989 (Analytical Biochemistry, 180,

147-151) and in the following Example 1. The method as used herein involves the use of four primers, one of which is mutation specific and three of which are designed to enable specific amplification of the mutated DNA. The unique

elements of this method are two related oligodeoxyribonucleotide primers: (i) a hybrid primer composed of a 3' segment complementary to a region of the DNA to be amplified, and a 5' segment whose sequence complements neither target DNA strand, and (ii) an oligonucleotide primer identical to the 5' segment of the hybrid primer. The oligoribonucleotide primers used for site-directed mutagenesis are shown in Table I in Example 1 below.

Accordingly, in one preferred embodiment, the process comprises as step (ii) a mutagenization treatment of the isolated DNA sequence which is a site-directed or site-specific mutagenesis, comprising constructing mutation specific primer sequences capable of hybridizing to the DNA sequence isolated in step (i), amplifying a DNA fragment comprising said primer sequences by a polymerase chain reaction (PCR) procedure, followed by the replacement of a fragment of the isolated DNA sequence with said amplified DNA fragment or a part thereof. By selecting appropriate primer sequences, the site-specific mutagenesis may be directed so as to obtain recombinant mutant PMT-encoding sequences derived from the toxA gene, in which gene one or more specific, preselected codon(s) is/are changed by substitution, deleted or inserted, respectively. Thus, as one typical example, the mutagenization strategy may be selected so as to result in substitution, deletion or insertion of one or more codons downstream of a certain restriction site of the toxA gene such as a unique EcoRI restriction site. In one specific embodiment, the mutagenization treatment of step (ii) of the process causes substitution of at least one codon coding for an amino acid of position between position 1131 and 1285 with a codon coding for a different amino acid. In specific useful embodiments of the invention, the mutagenization treatment causes substitution(s) of at least one codon of the toxA gene coding for an amino acid of position between position 1175 and 1285 such as of position between position between position 1200 and 1230.

Accordingly, useful rmPMTs may be obtained by selecting a mutagenization treatment in step (ii) whereby at least one step of mutagenization of the isolated PMT-encoding gene causes substitution of at least two codons such as at least three codons. Useful proteins may also be obtained by using a mutagenization treatment of step (ii) which results in the substitution of at least four codons. Using the designation nomenclature as defined above, typical examples of site- specific mutation (s) in the toxA gene produced by substitution of one or more codons include SV1201, HL1202, QL1203, HY1205, TV1206, TV1207, HY1223, HY1228, HL1202HY1205,

HL1202HY1223, HL1202HY1228, HY1205HY1223, HY1205HY1228, HY1223HY1228 and 4HX (HL1202HY1205HY1223HY1228).

The selection of codons to be substituted may be based on known characteristics of individual amino acids such as their hydrophobic/hydrophilic characteristics or their charge. By selecting substitution of one or more amino acids with an amino acid or amino acids having specifically useful physical characteristics, the rmPMTs encoded by such mutated DNA sequences may e.g. become particularly easy to purify or recover in downstream processes. A further significant selection criterion for substitution of one or more codons is the degree of reduction in rmPMT toxicity as defined above, which results from specific amino acid substitutions as well as the structural similarity between the amino acid being replaced and the replacing amino acid. Thus, as useful examples, the codon (s) being substituted may be a codon or codons coding for an amino acid/amino acids selected from serine, histidine, glutamine and threonine and the substituting codon (s) may be a codon or codons coding for an amino acid or amino acids coding for valine, leucine or tyrosine. It is contemplated that useful rmPMTs may be pro duced by having a hydrophilic amino acid replaced by a hydrophobic amino acid.

In a further useful embodiment, the mutagenization treatment of step (ii) may be selected so as to cause deletion of at least one codon coding for an amino acid of position between position 1131 and 1285. Preferably, the deletion(s) concern(s) at least one codon coding for an amino acid of position between position 1175 and 1285 such as e.g. deletion of at least one codon coding for an amino acid of position between position 1215 and 1285.

In accordance with the invention, useful rmPMTs may, based on the same selection criteria as mentioned above, be obtained by selecting a mutagenization treatment of step (ii) of the process as defined herein, causing deletion of at least two codons such as at least three codons. Useful rmPMT proteins may also be obtained by using a mutagenization treatment of step (ii) which results in the deletion of at least four codons such as e.g. at least seven codons. In preferred embodiments of the invention, the codon(s) being deleted is/are a codon or codons coding for an amino acid or amino acids selected from histidine, lysine, glumatic acid,

phenylalanine, alanine, valine and aspartic acid. Examples of useful rmPMTs obtained by deletion of one or more amino acid(s) of the PMT-encoding toxA gene include ΔH1223 and Δ1215-1221, cf. the above-defined designation nomenclature.

During the experimentation leading to the obtainment of rmPMT by deletion of one or more codons, it was found that during the construction of site-specific deletion mutant DNA

sequences, frame-shifts may occur which may have the same effect as direct deletion of codons. Thus, the rmPMT which in Table II is designated Δ1257-1285 carries an 11 amino acid residue out-of-frame extension after amino acid position 1256 as a result of deletion of thymidine residue number 3986 of the toxA gene, i.e. the third base in the codon coding for amino acid residue no. 1256. In a further useful embodiment of the invention, the mutagenization treatment of step (ii) causes insertion of at least one codon downstream of a codon coding for an amino acid of position between position 1131 and 1285, preferably downstream of a codon coding for an amino acid of position between position 1175 and 1285 such as downstream of a codon coding for an amino acid of position between position 1200 and 1230. The rmPMTs obtained by insertion of one or more codon(s) may typically comprise insertion of at least two codons such as at least three codons, or may more preferably comprise the insertion of at least four codons. The codon(s) being

inserted may advantageously be a codon coding for glycine. Examples of such useful rmPMTs include +G1203 and +GG1203.

As also mentioned above, the present process comprises as a step (iii) the insertion of the mutated DNA sequence resulting from step (ii) into a suitable replicon, including as useful examples, prokaryotic and eukaryotic cell chromosomes, plasmids, mitochondria, viruses and chloroplasts, the insertion being carried out by using standard recombination methods.

In this connection, a plasmid may be a convenient replicon. The selection of a suitable plasmid may be directed by well-known considerations, including the capability of plasmids to replicate in a wide range of host organisms, or the ability to occur in a host cell in a high copy number. A particularly useful type of plasmids may be plasmids showing "runaway" replication behaviour. Examples of plasmids which in accordance with the invention may be useful, include a plasmid selected from pSPE680, pSPE888, pSPE900, pSPE1003, pSPE1020, pSPE1038, pSPE1134 and pSPE 1234.

In a subsequent step (iv) of the process as defined herein, the replicon resulting from step (iii) of the process is used to transform a cell in which the replicon is capable of being replicated and in which the DNA sequence comprising the toxA gene mutated as described above, is expressible. A suitable host cell allowing the replication of the replicon and the expression of the mutated gene encoding the rmPMT may be selected from prokaryotic cells including bacteria, and from eukaryotic cells including yeast, fungi, mammalian cells such as human cell cultures, and plant cells. In this connection, a convenient host cell may be selected from genera of gram- positive bacteria such as e.g. Bacillus spp, Staphylococcus spp, Streptococcus spp, Lactococcus spp, Lactobacillus spp, or from a gram-negative bacterium including species selected from Etateroibacteriaceae, including as an example Escherichia coli , and species selected from Pseudomonadaceae and Vibrio-naceae.

In order to have the rmPMT produced, the transformed cell is cultured under conditions, where the mutated toxA gene is expressed, and the recombinant mutant protein harvested from the culture. The selection of culturing conditions such as culture medium composition, culturing temperature, aeration and batch sizes depends i.a. on the specific culturing condition requirements of the selected host cell and the nature of DNA sequences regulating the expression of the toxA gene.

The transformed cells may be ones which, when the mutant toxA gene is expressed, excrete the rmPMT protein into the culturing medium, in which case the protein may be harvested by directly recovering it from the medium, optionally after removing the cells e.g. by a centrifugation or a filtration step. Recovering of the protein from the medium may be carried out by using any conventional protein recovery procedure. Alternatively, the host cells may be ones in which the expressed rmPMT is accumulated periplasmically or

intracellularly. When using such host cells, the rmPMT harvesting step includes a treatment, whereby the cells are ruptured to an extent where the accumulated rmPMT protein is released from the cell. Such a treatment may e.g. be selected from a mechanical or physical cell disintegration treatment, an osmotic treatment or a treatment with a cell wall and/or cell membrane degrading enzyme.

In one convenient embodiment of the process according to the invention, the step of harvesting the rmPMT includes

separating the cells from the culture medium and subjecting the separated cells to a sonication treatment and removing the cell debris. In this manner a crude cell extract comprising the protein, is obtained. However, such a crude cell extract will contain i.a. a mixture of contaminating native bacterial proteins and the expressed rmPMT. It will be understood that such a crude rmPMT preparation will be less suitable for using directly as a vaccine component. Accordingly, the present process preferably comprises, as a further step, a purification of the crude cell extract.

In accordance with the present invention, a suitable specific protein purification may include any selective protein purification procedure, whereby the protein to be purified is isolated from contaminating proteins and preferably other substances. As examples of such well-known procedures may be mentioned affinity chromatography methods, wherein the specific protein is bound to a substance selectively binding the protein. Such a substance may e.g. be a monoclonal or polyclonal antibody or a suitable polymer including a Sepharose. The binding forces in affinity chromatography may e.g. be forces involved in the formation of hydrogen bonds,

hydrophobic interaction, formation of covalent bonds or the formation of antigen-antibody complexes. In one preferred embodiment, the purification of the rmPMT comprises an anion-exchange chromatography, including the procedure as exemplified in the following. Such a chromatography procedure normally involves the use of an elution reagent, typically being a salt, such as e.g. NaCl, dissolved in a buffer.

In suitable embodiments of the present invention, the process as defined above is an anion-exchange chromatography e.g. including an elution step using a buffer with NaCl concentration increasing from 0 to 1000 mM, and a hydrophobic interaction chromatography which may include an elution step using a buffer with decreasing concentration of ammonium sulphate. Typically, such eluates having increasing or decreasing ionic strengths are analyzed for their content of the rmPMT relative to the total content of protein (i.e. the purity of the rmPMT) and accordingly, the ionic strength of the elution reagent resulting in the highest content of rmPMT may be selected. When e.g. using an anion-exchange chromatography procedure as exemplified herein, an eluate collected at about 700 mM NaCl results in the highest purity.

Preferably, the purification is carried out in one step such as an anion-exchange chromatography including a suitable elution. When a particularly high degree of purity of the recombinant mutant PMT protein is desirable, it may be required to use two or more purification procedures e.g.

being selected from an anion-exchange chromatography, a hydrophobic interaction chromatography and an antibody affinity chromatography. Accordingly, a suitable purification step may comprise an initial anion-exchange chromatography, including an elution as defined above, followed by a further step comprising hydrophobic interaction chromatography of the eluates resulting from the preceding step and having a high purity with respect to rmPMT.

As mentioned above, it is desirable to obtain a rmPMT preparation of a high purity. Accordingly, it is preferred to obtain eluates from the purification procedure(s) which have a purity of at least 50 wt%, calculated on the total protein, more preferably of at least 75 wt%, even more preferably of at least 90 wt%, most preferably of at least 94 wt% and in particular of at least 98 wt%.

From an industrial production cost point of view it is of significant importance that the yield of the recombinant mutant rmPMT protein, which is obtained from the manufac turing process including the downstream processes, is high. In this context, the yield of rmPMT is defined as the amount of pure rmPMT protein obtained from the process relative to the total protein content of a crude extract of host cells expressing said protein. Accordingly, the present process is preferably one which results in a yield of mutant recombinant protein which is at least 1 wt%, calculated on the total protein content of the crude cell extract, more preferably at least 2 wt% such as at least 3.5 wt%, even more preferably at least 5 wt%, most preferably at least 7.5 wt%, in particular preferably at least 10 wt% such as at least 12.5 wt% and most particularly, at least 15 wt%.

As mentioned above, the present invention relates in a still further aspect to a DNA sequence coding for the recombinant mutant PMT protein as defined herein, said DNA sequence being derived from a replicon comprising a Pasteurella multocida toxA gene, by substitution, deletion or insertion of one or more codons in said toxA gene in accordance with the process as also defined herein. When the recombinant mutant PMT is a fusion protein as defined above, the DNA sequence may further comprise a sequence coding for a second immunogenic protein, said sequence being associated with the toxA-derived gene by in-frame fusion of a truncated or mutated toxA gene to the sequence coding for the second protein, in a manner that results in the expression of a fusion protein gene product as defined herein.

The DNA sequence may also comprise a promoter sequence and a sequence which regulates the expression of the site-specifically mutated toxA gene. The promoter sequence may suitably be a promoter natively associated with the native toxA gene, but it may also be a promoter sequence which is not natively associated with said gene.

The expression of the mutated rmPMT-encoding gene may be regulated at the transcriptional level or at the translational level. A regulation at the transcriptional level may e.g. occur in the form of a transcriptional repressor substance assumingly including the trans-acting TxaR repressor protein which is expressed in wild-type strains of PMT-producing Pasteurella multocida . A DNA sequence encoding a rmPMT-regulating repressor substance may optionally be regulated by a regulatory sequence operably linked to the gene coding for the repressor. Such a regulatory sequence may be a temperature sensitive sequence such as e.g. a λ cI sequence.

The rmPMT-encoding gene may furthermore be regulated at the post-transcriptional level, e.g. by the presence of an antisense mRNA hybridizing to the rmPMT mRNA.

As mentioned previously, the present invention provides in a further aspect a replicon harbouring the DNA sequence as defined above and prepared according to the process as described above, a typical useful example hereof being a replicon which is selected from the plasmid pSPE1038 encoding the mutant recombinant protein HL1202HY1205HY1223HY1228

(4HX), as deposited on 1 September 1992 with the DSM-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH under the accession number DSM 7221, and pSPE1020 encoding the rmPMT +GG1203, pSPE1134 encoding 1132"β-gal and pSPE1234 encoding the rmPMT ΔH1223 as deposited with the DSM-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH on 9 September 1993 under the accession numbers DSM 8548, DSM 8546 and DSM 8547, respectively. pSPE1038, pSPE1020, pSPE1234 and pSPE1134 were also deposited with the China Center for Type Culture Collection (CCTCC), Wuhan University, Wuhan, The People's Republic of China.

In other aspects, the present invention relates to a cell, which, in accordance with the process as defined herein, is transformed with the herein-defined replicon and in which cell said replicon is replicated, and to the use of the mutant recombinant protein as defined herein as a vaccine for the protection against diseases caused by toxin-producing Pasteurella multocida. When used as a vaccine, the rmPMT is typically present in a vaccine composition optionally comprising an immunologically acceptable carrier and optionally further components such as adjuvants including Freund's incomplete or complete adjuvant and aluminium hydroxide e.g. in the form of a gel, preservative agent(s) and buffer salt(s). A vaccine composition comprising the rmPMT may be an aqueous suspension or it may be a lyophilized or freeze-dried composition. A rmPMT-comprising composition contains the rmPMT in an amount which results in an immunologically effective amount hereof in a normally used vaccine dosage, the amount typically depending on the particular animal to be vaccinated.

The invention is further illustrated in the following non- limiting Examples and by the Figures in which: Fig. 1 is a schematic representation of plasmids pSPE312, pSPE680, and pSPE1003 used for site-directed mutagenesis of the Pasteurella multocida toxA gene. The hatched areas indicate the toxA gene, the filled-in areas indicate vector DNA, and other Pasteurella multocida DNA is represented by open areas;

Fig. 2 shows a standard curve for proliferative effect (PE) of recombinant Pasteurella multocida toxin (rPMT) (i.e. PMT which is encoded by the toxA gene in SPE1036) on NIH3T3 cells, showing the proliferative effect (PE) of different concentrations of rPMT. SPE1036 whole-cell extract containing rPMT was diluted and added to subconfluent NIH3T3 cells in different amounts. The PE was calculated as described in Example 2. Purified rPMT (cf. Example 3) was used for determination of the PE of 20 μg/ml rPMT; Fig. 3 illustrates the proliferative effect of recombinant mutant PMT proteins (rmPMTs) derived from toxA by a single site-directed mutagenesis causing amino acid substitution, insertion or deletion, or by fusion of a subsequence of toxA to lacZ. Whole-cell extracts containing different rmPMTs were assayed for their PE on NIH3T3 cells;

Fig. 4 summarizes the PE of recombinant mutant PMT proteins (rmPMTs) derived from toxA by consecutive site-directed mutagenesis causing multiple amino acid substitutions;

Fig. 5 shows a Coomassie Brilliant Blue stained SDS-PAGE gel showing protein profiles of 4HX-, derivative O-, and rPMT producing E. coli MC1000 (dam). Culture samples were taken in late exponential (lx), between late exponential and stationary (lxs), in early stationary (es), and late stationary (Is) growth phase. The relative amount of rmPMT or rPMT in the sample is indicated below each lane;

Fig. 6 shows a chromatogram of a Q-Sepharose anion-exchange chromatography of crude E. coli extract containing rPMT. The hatched area indicates the fractions containing rPMT of high purity;

Fig 7. shows the SDS-PAGE analysis of fractions collected during Q-Sepharose chromatography of crude E. coli extract containing rPMT. The gels containing 10% polyacrylamide were silver stained. Panel A, lane 1 and panel B, lane 1: crude bacterial extract; panel A, lanes 2-11: fractions 9, 13, 29, 33, 35, 37, 39, 41, 43 and 45, respectively; panel B, lanes 2-11: fractions 47, 49, 51, 53, 55, 57, 59, 61, 63 and 67, respectively. The position of rPMT is indicated by arrows; Fig. 8A illustrates purification of rPMT from crude E. coli extract by anion-exchange chromatography (step 1) and

hydrophobic interaction chromatography (step 2) . The 10% SDS-PAGE gel was silver stained. Lane 1: molecular weight

markers; lane 2: crude bacterial extract; lane 3: effluent from step (1); lane 4: pool from step (1); lane 5: effluent from step (2); lane 6: pool from step (2), i.e. final product resulting from the purification. The position of rPMT is indicated by an arrow. Fig. 8B illustrates purification of 4HX (lanes 2-4) and rPMT (lanes 5-7) by anion-exchange chromatography (step 1) and hydrophobic interaction chromatography (step 2). The 10% SDS- PAGE gel was silver stained. Lane 1: molecular weight

markers; lanes 2 and 5: crude bacterial extract containing 4HX or rPMT, respectively; lanes 3 and 6: pools from step (1); lanes 4 and 7: pools from step (2) i.e. final products resulting from the purification. The position of 4HX is indicated by an arrow. Fig. 9 illustrates purification of rPMT. Two-fold dilutions of a preparation of purified rPMT were analyzed by SDS-PAGE (10% gel, silver staining). Lane 1: molecular weight markers; lanes 2-8: purified rPMT diluted 2x, 4x, 8x, 16x 32x, 64x and 128x, respectively. The position of rPMT is indicated by an arrow.

Fig.10 shows a chromatogram of a Q-Sepharose anion-exchange chromatography of crude E. coli extract containing derivative O (dO). The hatched area indicates the fractions that were enriched for dO (according to SDS-PAGE of the fractions, cf. Fig. 11) and were pooled for further purification attempts;

Fig. 11 shows SDS-PAGE analysis of fractions collected during Q-Sepharose chromatography of crude E. coli dO-containing extract. The 10% polyacrylamide gels were silver stained.

Panel A, lane 1 and panel B, lane 11: crude bacterial

extract; panel A, lanes 2-11: fractions 8, 12, 30, 32, 34, 36, 38, 40, 42, and 44, respectively; panel B, lanes 1-10: fractions 46, 48, 50, 52, 54, 56, 58, 60, 64 and 66, respectively. The position of dO is indicated by arrows;

Fig. 12 illustrates the results of attempts to purify derivative O from crude E. coli extract by anion-exchange chromatography (step one) and hydrophobic interaction chromatography (step two). The 10% SDS-PAGE gel was silver stained. Lanes 1 and 5: molecular weight markers; lane 2: crude bacterial extract; lane 3: pool from step one; lane 4: pool from step two. The position of derivative O is indicated by an arrow.

Fig. 13 shows a chromatogram of a Q-Sepharose anion-exchange chromatography of crude E. coli extract containing 4HX. The hatched area indicates the fractions to be pooled and containing 4HX of high purity (according to SDS-PAGE analysis of the fractions, cf. Fig. 14);

Fig. 14 shows SDS-PAGE analysis of fractions collected during Q-Sepharose chromatography of crude E. coli extract containing the 4HX protein. The polyacrylamide gels were silver stained. Panel A, lane 1 and panel B, lane 1: crude bacterial extract; panel A, lanes 2-11: fractions 9, 12, 31, 34, 36, 39, 41, 43, 45 and 47, respectively; panel B, lanes 2-11:

fractions 49, 51, 53, 55, 57, 63, 66, 69 and 71, respectively. The position of 4HX is indicated by arrows;

Fig. 15 is a silver stained 10% SDS-PAGE gel illustrating purification of 4HX from crude E. coli extract by anionexchange chromatography (step 1) and hydrophobic interaction chromatography (step 2) . Lane 1: molecular weight markers; lane 2: crude bacterial extract; lane 3: effluent from step 1; lane 4: pool from step 1; lane 5: effluent from step 2; lane 6: pool from step 2, i.e. final product of the purification. The position of 4HX is indicated by an arrow.

Fig. 16 shows SDS-PAGE analysis of purified 4HX. Two-fold dilutions of a preparation of purified 4HX were analyzed. The 10% SDS-PAGE gel was silver stained. Lane 1: molecular weight markers; lanes 2-8: purified 4HX diluted 2x, 4x, 8x, 16x, 32x, 64x and 128x, respectively. The position of 4HX is indicated by an arrow. Fig. 17 illustrates preparative scale anion-exchange chromatography on Q-Sepharose of crude E. coli extract containing 4HX. The hatched area indicates the fractions that contained 4HX of high purity (according to SDS-PAGE analysis of collected fractions) and which were pooled;

Fig. 18 shows the results of preparative scale purification of 4HX by anion-exchange chromatography (step 1) and

hydrophobic interaction chromatography (step 2). The polyacrylamide gel was silver stained. Lanes 1 and 6: molecular weight markers; lane 2: crude bacterial extract containing 4HX; lane 3: effluent from step 1; lane 4: pool from step 1; lane 5: pool from step 2, i.e. final product. The position of 4HX is indicated by an arrow.

Fig. 19 shows SDS-PAGE analysis of two-fold dilutions of purified 4HX obtained by two-step preparative-scale purification as illustrated in Fig. 18. The 10% SDS-PAGE gel was stained with Coomassie Brilliant Blue. Lane 1: molecular weight markers; lanes 2-8: purified 4HX diluted 2x, 4x, 8x, 16x, 32x, 64x and 128x, respectively. The position of 4HX is indicated by an arrow.

Fig. 20 shows SDS-PAGE analysis and immunoblotting of

purified 4HX. Samples were run on a 10% gel and transferred to nitrocellulose. The blot was incubated with a polyclonal rabbit antibody against PMT followed by incubation with alkaline phosphatase-conjugated porcine anti-rabbit immunoglobulins (DAKO, Denmark, code no. D-306) and staining for alkaline phosphatase. Lane 1: purified rPMT (0.15 μg, positive control); lane 2: laboratory scale-purified 4HX (0.3 μg): lane 3: preparative scale-purified 4HX (1.1 μg); lane 4: recombinant nuclease produced in E. coli , M_r 30,000 (3 μg, negative control). Molecular weights are indicated to the right. Fig. 21 shows the stability of rPMT, derivative O and 4HX during incubation at 37°C. Purified rPMT, derivative O and 4HX was incubated at 37°C. Samples taken out at time zero (control) and after 24 and 48 h were analyzed by SDS-PAGE on a gel containing 10% polyacrylamide. The gel was silver stained. Lanes 1 and 11: Molecular weight markers; lanes 2-4: rPMT; lanes 5-7: derivative O; lanes 8-10: 4HX. Lanes 2, 5 and 8: control; lanes 3, 6 and 9: 24 h; lanes 4, 7 and 10: 48 h. Fig. 22 shows the stability of rPMT during incubation with trypsin. Aliquots of purified rPMT were mixed with 1, 2 or 5 wt% of trypsin and incubated at 37°C for 24 h. Samples were analyzed by SDS-PAGE as described in the above legend to Fig. 21. Lanes 1 and 6: molecular weight markers; lane 2: purified rPMT without trypsin addition and incubation at 37°C (control); lanes 3, 4 and 5: 1, 2 and 5% trypsin, respectively. The position of rPMT is indicated by an arrow.

Fig. 23 shows the stability of 4HX during incubation with trypsin. Aliquots of purified 4HX were mixed with 1, 2 or 5 wt% of trypsin and incubated at 37°C for 24 h. Samples were analyzed by SDS-PAGE as described in the above legend to Fig. 21. Lanes 1 and 6: molecular weight markers; lane 2: purified 4HX without trypsin addition and incubation at 37°C (control); lanes 3, 4 and 5: 1, 2 and 5% trypsin, respectively. The position of 4HX is indicated by an arrow.

Fig. 24 shows the stability during incubation with trypsin of derivative O purified by immunoaffinity chromatography.

Aliquots of purified derivative O were mixed with 1, 2 or 5 wt% of trypsin and incubated at 37°C for 24 h. Samples were analyzed by SDS-PAGE as described in the above legend to Fig. 21. Lanes 1 and 6: molecular weight markers; lane 2: purified derivative O without trypsin addition and incubation at 37°C (control); lanes 3, 4 and 5: 1, 2 and 5% trypsin, respectively. The position of derivative O is indicated by an arrow. Fig. 25 shows the stability of derivative O enriched by Q- Sepharose and Phenyl Sepharose chromatography, during incubation with trypsin. Aliquots of enriched derivative O were mixed with 1, 2 or 5 wt% of trypsin and incubated at 37°C for 24 h. Samples were analyzed by SDS-PAGE as described in the above legend to Fig. 21. Lanes 1 and 6: molecular weight markers; lane 2: enriched derivative O without trypsin addition and incubation at 37°C (control); lanes 3, 4 and 5: 1, 2 and 5% trypsin, respectively. The position of derivative O is indicated by an arrow.

Fig. 26 shows a chromatogram of a Q-Sepharose anion-exchange chromatography of crude E. coli extract containing +GG1203. The hatched area indicates the fractions containing +GG1203 of high purity. Fig. 27 shows the SDS-PAGE analysis of a selection of fractions collected during Q-Sepharose chromatography of crude E. coli extract containing +GG1203. The gel containing 10% polyacrylamide was silver stained. Lanes 1 and 13: molecular weight markers; lane 2: crude bacterial extract; lanes 3 -12: fractions 55, 56, 57, 58, 59, 60, 61, 62, 63 and 64, respectively. The position of +GG1203 is indicated by an arrow.

Fig. 28 illustrates purification of +GG1203 from crude E.

coli extract by anion-exchange chromatography (step 1) and hydrophobic interaction chromatography (step 2). The 10% SDS-PAGE gel was silver stained. Lanes 1 and 5: molecular weight markers; lane 2: crude bacterial extract; lane 3: pool from step 1; lane 4: pool from step 2, i.e. final product of the purification.

Fig. 29 shows SDS-PAGE analysis of purified +GG1203. Two-fold dilutions of a preparation of purified +GG1203 were analyzed by SDS-PAGE (10% gel, Coomassie Brilliant Blue staining).

Lanes 1 and 8: molecular weight markers; lanes 2-7: purified +GG1203 diluted 2x, 4x, 8x, 16x, 32 x and 64x, respectively.

Fig. 30 shows a chromatogram of a Q-Sepharose anion-exchange chromatography of crude E. coli extract containing ΔH1223.

The hatched area indicates the fractions containing ΔH1223 of high purity. Fig. 31 illustrates purification of ΔH1223 from crude E. coli extract by anion-exchange chromatography (step 1) and

hydrophobic interaction chromatography (step 2). The 10% SDS- PAGE gel was silver stained. Lanes 1 and 6: molecular weight markers; lanes 2 and 3 : crude bacterial extract diluted 2x and 4x respectively; lane 4: pool from step 1; lane 5: pool from step 2, i.e. final product of the purification.

Fig. 32 shows SDS-PAGE analysis of purified ΔH1223. Two-fold dilutions of a preparation of purified ΔH1223 were analyzed by SDS-PAGE (10% gel, silver staining). Lanes 1 and 8: molecular weight markers; lanes 2-7: purified ΔH1223 diluted 2x, 4x, 8x, 16x, 32x and 64x, respectively.

Fig. 33 illustrates purification of 1132"β-gal. Lane 1:

molecular weight markers; lanes 2, 3 and 4: crude bacterial extract diluted 2x, 4x and 8x, respectively; lane 5: APTG affinity purified 1132"β-gal. The position of 1132"β-gal is indicated by an arrow.

Fig. 34 shows the resistance of +GG1203 to trypsin. Aliquots of purified +GG1203 were mixed with 1, 2 or 5 wt% trypsin and incubated at 37°C for 24 h. A control sample was incubated at 37°C for 24 h without trypsin. After the incubation, samples were analyzed by SDS-PAGE on a gel containing 10% polyacrylamide. The gel was silver stained. Lanes 1 and 7: molecular weight markers; lane 2: purified +GG1203; lane 3: control; lanes 4, 5 and 6: 1, 2 and 5 wt% trypsin, respectively.

Fig. 35 shows the resistance of ΔH1223 to trypsin. Aliquots of purified ΔH1223 were mixed with 1, 2 or 5 wt% trypsin and incubated at 37°C for 24 h. A control sample was incubated at 37°C for 24 h without trypsin. After the incubation, samples were analyzed by SDS-PAGE as described in the above legend to Fig. 34. Lanes 1 and 7: molecular weight markers; lane 2:

purified ΔH1223; lane 3: control; lanes 4, 5 and 6: 1, 2 and 5 wt% trypsin, respectively. Fig. 36 shows the sensitivity to trypsin of 1132" β-gal.

Aliquots of purified 1132"β-gal were mixed with 1, 2 or 5 wt% trypsin and incubated at 37°C for 24 h. A control sample was incubated at 37°C for 24 h without trypsin. After the incubation, samples were analyzed by SDS-PAGE as described in the above legend to Fig. 34. Lanes 1 and 7: molecular weight markers; lane 2: purified 1132"β-gal; lane 3: control; lanes 4, 5 and 6: 1, 2 and 5 wt% trypsin, respectively.

Fig. 37 shows serum titres against rPMT for animals immunized with frPMT and 4HX. Log₂ titres are plotted against immunization doses.

Fig. 38 shows serum titres against rPMT for animals immunized with 4HX, +GG1203 and 1132"β-gal. Log₂ titres are plotted against immunization doses. Fig. 39 shows serum titres against rPMT for animals immunized with 4HX and ΔH1223. Log₂ titres are plotted against immunization doses.

EXAMPLE 1

Introduction of site-specific mutations in the 3' end of the Pasteurella multocida toxA gene

Site-directed mutations were introduced in the 3' end of the toxA gene downstream of the unique EcoRI restriction site. In each case, a DNA fragment with the wanted mutation was obtained by a polymerase chain reaction (PCR) amplification procedure.

The E. coli strain DH5α (supE44, ΔlacU169 (φ80lacZM15),hsdR17, recA1, endA1, gyrA96, thi - 1 , relA1) (Hanahan, D.

1983. Studies on transformation of Escherichia coli with plasmid. J. Mol. Biol. 166:557) was used as a host for the various plasmids. The toxin-encoding plasmids pSPE312 (Petersen S.K. et al. 1991. Recombinant derivatives of Pasteurella multocida toxin: Candidates for a vaccine against Progressive Atrophic

Rhinitis. Infect. Immun. 59:1387-1393), pSPE680 (ibid. ) , and pSPE1003, were used to clone PCR fragments (cf. Fig. 1).

pSPE1003 is a derivative of pSPE680 devoid of the toxA extragenic EcoRI restriction enzyme recognition site. The toxA intragenic EcoRI site of this plasmid is therefore unique. pSPE1003 was constructed by partial restriction of pSPE680 with EcoRI, isolation of the linearized plasmid from agarose gel, treatment of the isolated fragment with T4 DNA polymerase in the presence of all four deoxyribonucleotides, ligation with T4 DNA ligase, and transformation of DH5α.

Restriction enzymes, T4 DNA ligase, and T4 DNA polymerase were purchased from New England Biolabs, Inc. MA, and used as recommended by the manufacturer. DNA was isolated from agarose gel by the Geneclean (BIO101 Inc., CA) procedure. In short, this procedure is used for 1) dissolving the DNA-containing agarose with 4 M NaI, and 2) extracting the DNA by use of its affinity for a silica matrix.

Oligodeoxyribonucleotides were synthesized on a Cyclone DNA synthesizer (Biosearch Inc., CA). Table I lists the primer DNA sequences used to introduce mutations (position numbers as in Petersen, S. K. 1990. The complete nucleotide sequence of the Pasteurella multocida toxin gene and evidence for a transcriptional repressor, TxaR. Mol. Microbiol. 4:821-830).

Table I. Oligodeoxyribonucleotides used for site-directed mutagenesis.

The oligonucleotides are shown below the relevant wild-type DNA and amino acid sequences (position 3801 to 3938 and 1195 to 1240, respectively) (numbering as in Petersen et al.

1990). The name of each oligonucleotide and the mutation it introduces, is underlined.

The mutation specific primers were named after the amino acid change they introduce (see below and Table II). The NsiIcw primer hybridizes to position 2109 to 2129 around the first NsiI restriction site in the toxA sequence. The Pstlccw primer is identical to the pBR322 PstI primer PII from Pharmacia LKB Biotechnology, Sweden. In the AB primer, the first half (the A part) is identical to the A primer, which is a 20 base randomly chosen sequence and the second half (the B part) is identical to position 3580 to 3600 of toxA. The sequences of primers C and D are identical to positions 4282 to 4257, and 4105 to 4084, respectively, of toxA.

Table II below lists the rmPMTs and the summarizes site-directed mutagenesis parameters, and shows for each recombinant mutant PMT protein (rmPMT): 1) the amino acid change, 2) the strain number of the recombinant mutant, 3) the PCR primers used for mutagenesis and amplification, 4) the corresponding PCR template, 5) the restriction enzymes used to digest both the amplified DNA fragment and the vector DNA, and 6) the name of the cloning vector. The names of the rmPMTs refer to the wildtype and mutant amino acid encoded by the codon in question, using the standard one-letter code; as an example, SV1201 is produced by a mutant with a mutation changing serine codon number 1201 of the published toxA sequence to a valine codon. Similarly, +G1203 is produced by a mutant with an insertion mutation introducing a glycine codon as codon number 1203, and Δ1215-1221 is produced by a mutant with a deletion of codons 1215 to 1221. The only exception from this nomenclature is Δ1257-1285; this protein is encoded by a plasmid, harbouring a frame-shift which was made and isolated by chance during the construction of the other plasmids. Δ1257-1285 carries an 11 amino acid residue nonsense extension after residue number 1256.

All mutant strains were constructed by recombinant cloning of PCR amplified DNA fragments as summarized in Table II. Except for SPE888, SPE900, and SPE1038 they were constructed by a procedure essentially as described by Nelson and Long

(Nelson, R. M., and G. L. Long, 1989. A general method of site-specific mutagenesis using a modification of the Thermus aquaticus polymerase chain reaction. Anal. Biochem.

180:147-151). In short, this procedure involves the use of 4 primers one of which is mutation specific and the other three are designed to enable specific amplification of the mutated DNA. This is accomplished by the procedure described below.

PCR was carried out in a Techne Programmable Dri-Block PHC-1 (Techne Ltd., UK) or a Hybaid thermal reactor (Hybaid Ltd., UK) in 100 μl reaction buffer (10 mM Tris-HCl pH 8.3, 50 mM KCl, 2.5 mM MgCl₂, 0.01% (w/v) gelatine, 200 μM of each dNTP with 2.5 units Thermus aquaticus (Taq) polymerase (Ampli Taq from Perkin Elmer, CT) and covered with 20 μl mineral oil M3516 (Sigma Chemical Co., MO). Each cycle included

denaturation (2 min at 95°C), annealing (2 min at 37°C), and polymerization (3 min at 72°C). In all cases, the mutation specific primer (100 pmol) and the AB primer (100 pmol) were used in the first PCR amplification (30 cycles) on 1 fmol of template DNA. About 0,6 pmol of the PCR product and 1 fmol template was then run in a single cycle of 5 min at 92°C, 2 min at 37°C, and 10 min at 72°C. This last step results in elongation in the clockwise direction only, on the plasmid template, since the A part of the AB primer has no template homology and its complementary sequence thus cannot prime DNA synthesis in the counterclockwise direction. This elongation product was then used as template in a further PCR amplification where primers A and C or Pstccw (100 pmol of each) were added to the reaction mixture, and 30 more cycles were run to achieve a specific amplification of the mutated DNA. This product was cloned by restriction of vector and PCR product with the indicated restriction enzymes (Table II), ligation of a mixture of these DNA's with T4 DNA ligase, and transformation of strain DH5o;, using the method of Hanahan (Hanahan, D. 1983. Studies on transformation of Escheri chia coli with plasmid. J. Mol. Biol. 166: 557) .

For the construction of SPE888, SPE900, and SPE 1038, different procedures were used, since the mutation specific primers L1202 and L1202Y1205 introduce useful restriction sites (a NsiI- and a PstI site, respectively). SPE888 and SPE900 (encoding HL1202 or HL1202HY1205, respectively) were constructed by simple PCR amplifications: 2 μg template, 0.9 μg of each primer, 4 cycles of denaturation (2.5 min at 95°C), annealing (2.5 min at 52C), and polymerization (10 min at 72°C); and cloning of the amplified fragments as above.

SPE1038 was constructed in a 3 step procedure: first, primers L1202Y1205 and Y1223Y1228 (10 μg each) were used as primers in a PCR amplification on 50 ng template pSPE900 DNA cut with HindIII: 20 cycles of denaturation (2.5 min at 95°C), annealing (2.5 min at 60°C), and polymerization (10 min at 72°C); secondly, the PCR product was elongated using 0.5 μg pSPE900 DNA cut with HindIII as a template: 1 cycle (5 min at 95°C, 2 min at 30°C, and 10 min at 70C); and finally, 0.9 μg of

Pstccw primer was added, and 20 further cycles (2 min at 95°C, 2 min at 35°C, and 2 min at 72°C) were run. The product was cloned as described above.

The DNA sequences of the entire cloned PCR amplified fragments were determined using double-stranded plasmid DNA and the sequenase version 2.0 procedure (United States Biochemical Corporation, Ohio). The sequencing primers used corresponded to the following nucleotide numbers of toxA: 2089 to 2107, 2400 to 2417, 2683 to 2701, 3004 to 3024, 3303 to 3321, 3580 to 3600, and 4105 to 4084 (primer D in Table I). No mutant strain used in the following studies harboured other mutations in the cloned DNA fragment than those indicated in Table II, except for silent third position mutations in 2 strains. A toxA"lacZ fusion gene was constructed, in which the toxA promoter, ribosome binding site and the first 1132 codons of the gene were fused to a functional lacZ structural gene. This construction involved the toxA-harbouring plasmid pSPE1003 (Fig.l) and the lacZ-harbouring plasmid pNM482 (Minton, 1984. Gene, 31:269-273). The toxA"IacZ-harbouring plasmid pSPE1134 was constructed by cloning the 3.6 kb EcoRI fragment from pSPE1003 into the unique EcoRI restriction site of pNM482. The pSPE1134-encoded fusion protein comprises the first 1132 amino acids of the PMT, fused to a functional β- galactosidase. This fusion protein is designated 1132 "β-gal. pSPE1134 was deposited with the DSM-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH on 9 September 1993 under the accession number DSM 8546 and with the China Center for Type Culture Collection (CCTCC), Wuhan University, Wuhan, The People's Republic of China.

EXAMPLE 2

Mitogenic potency of recombinant mutant PMT toxins

In order to test the activity of recombinant mutant PMT proteins (rmPMTs), whole-cell extracts were prepared in the following manner: the plasmid-harbouring strains of Example 1 were grown at 30°C in LB medium (Miller, J. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.) containing 50 μg/ml ampicillin. Stationary phase cultures were harvested by centrifugation, resuspended in 0°C water, sonicated for several 0.5 min periods with a Branson Sonifier 250 (Branson Sonic Power Co., Conn.) and sterile filtered using Millex-GV filters with a 0.45 μm pore size (Millipore S.A., France). The rmPMT content of each whole-cell extract was then evaluated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), [Laemmli, U.K. (1970). Cleavage of Structural Proteins during the Assembly of the Head of

Bacteriophage T4. Nature (London) 227, 680-685] using 7.5% (wt/vol) acrylamide gels (acrylamide/bisacrylamide ratio of 40:1), followed by Coomassie Brilliant Blue R250 staining and destaining of the gels in 5% acetic acid-50% ethanol. Using the purified HL1202 protein (cf. Example 3) as a concentration standard, all whole-cell extracts were diluted to a rmPMT concentration of about 20 μg/ml, as determined by one- dimensional laser densitometry scanning (UltroScan XL Laser Densitometer, LKB/Pharmacia, Sweden) of Coomassie Brilliant Blue stained SDS-PAGE gels.

The diluted extracts were used in the mitogenicity assay, which was carried out essentially as described by Rozengurt et al. (Rozengurt, E., T. Higgins, N. Chanter, A. J. Lax, and J. M. Staddon. 1990. Pasteurella multocida toxin: potent mitogen for cultured fibroblasts. Proc. Natl. Acad. Sci. USA 87:123-127) as follows: about 10⁵ NIH3T3 cells (ATCC CRL1658) were seeded onto 35-mm petri dishes in 2 ml of Dulbecco modified Eagle medium supplemented with 10% newborn calf serum, 2mM L-glutamine, penicillin (100 IU/ml), and streptomycin (10 μg/ml). Fresh medium containing purified protein, whole-cell extract with rmPMT or rPMT, or whole-cell extract of strain DH5α, was added to three equal dishes 1 and 4 days after plating, and cells were counted 7 days after plating, at which time the negative control culture had been confluent for at least 4 days.

The proliferative effect (PE) of a given sample (x) was calculated as

where N_x is the mean cell number in 3 dishes with sample x, N_pmt is the mean cell number in 3 dishes with SPE1036 whole- cell extract containing a final concentration of 20 ng/ml rPMT, and N_O is the mean cell number in 3 dishes with DH5α whole-cell extract, i.e. dishes with a confluent monolayer of NIH3T3 cells. A standard curve for rPMT mitogenicity was made using SPE1036 whole-cell extract in different concentrations. SPE1036 is E. coli DH5θ! harbouring pSPE680 (Fig. 1) . As shown in Fig. 2, the concentration of rPMT resulting in a 50% PE is about 4-5 ng/ml and the maximum PE (100%) is obtained at about 20 ng/ml rPMT which is in accordance with the literature (Rozengurt et al.). A 1000-fold higher concentration of rPMT did not result in a significantly different PE. No toxic effects of DH5α extract were seen even when it was added in a 1000-fold excess compared to the 20 ng/ml rPMT-containing extract (data not shown). In different experiments 0% PE and 100% PE always corresponded to 4-6.5 × 10⁵ and 2.5-4 × 10⁶ NIH3T3 cells per ml, respectively.

The PEs of rmPMTs with single amino acid substitutions, insertions or deletions were then assayed. All whole-cell extracts were applied in amounts corresponding to >5 ng/ml of the mutant protein, preferably in amounts resulting in 20-80% PE, corresponding to points on the steepest part of the protein concentration/PE curve (cf. Fig. 2). Each experiment included a positive (SPE1036) and a negative (DH5α;) control extract. The results are summarized in Fig. 3. Three extracts were assayed twice in independent experiments to ensure that the measured PEs were constant between experiments. Addition of extract to a final concentration of 8 ng/ml TV1206

resulted in PEs of 37.2±6.7% and 31,7±10.2%; 40 ng/ml SV1201 resulted in PEs of 51.8±5.6% and 48.4±5.8%; and 80 ng/ml HY1205 resulted in PEs of 43.4±8.1% and 40.0±9.6%. It was concluded that PEs obtained in independent experiments could be directly compared. In order to compare the mitogenic effect of rmPMTs with that of rPMT, the mitogenic potency (MP) of a given rmPMT was defined as the concentration of rPMT resulting in a given PE relative to the concentration of the rmPMT resulting in the same PE. This definition is based on the assumption that the protein concentration/PE curves for rPMT and rmPMTs, if different, are identical in shape and only displaced in parallel. The rmPMTs HY1202, QL1203, TV1207, and Δ1257-1285 had retained MP's of more than 50%, and were not significantly different from PMT in this assay; the MP's of SV1201, TV1206, and HY1228 were between 10 and 35%; the MP's of

HL1202, and HY1205 were about 5%; and the MP's of +G1203,

+GG1203, Δ1215-1221, HY1223, ΔH1223 and 1132"jS-gal were less than 5%.

It was concluded 1) that the polypeptide stretch from amino acid 1201 to 1228 including 4 histidine residues is of vital importance for the MP of rPMT; 2) that this importance is reflected in the importance of some single amino acids in this stretch (most notably Histidine-1223, but also

Serine-1201, Histidine-1202, Histidine-1205, Threonine-1206 and to some degree Histidine-1228) but apparently not all (i.e. Glutamine-1203 and Threonine-1207); and 3) that (a) structure (s) involving amino acid residues between residue 1200 and 1229 is/are specifically important for the functional activity, since insertions and deletions in this region (resulting in proteins +GG1203, ΔH1223 and Δ1215-1221) have a profound negative effect on the MP, whereas a deletion outside this stretch (of amino acid residues 1257-1285) does not affect the MP. The available evidence does not exclude that other residues either between position 1201 and 1228 or elsewhere are involved in the same structure(s). The nature and effects of the mutations might further indicate (but do not prove) that the polarity of residue 1202 is important, that the polarity or hydroxyl group of both residue 1201 and 1206 is important, and that the imidazole ring of both residue 1205 and 1223 is highly important for the functionality of PMT. The activities of the double mutant rmPMTs described in Example 1 were tested likewise, as was the activity of a potential vaccine candidate with 4 independent mutations: HL1202HY1205HY1223HY1228 also designated 4HX (Fig. 4). The above vaccine candidate proteins 4HX, +GG1203, ΔH1223 and 1132"β--al were also tested for mitogenic potency after purification as described in Example 3. The MP values for these purified rmPMTs were in accordance with the above data for rmPMT-containing whole-cell extracts.

EXAMPLE 3

Production, purification and stability of selected recombinant mutant PMT proteins a. Production

The production profiles of rPMT, i.e. PMT encoded by the toxA gene and expressed in E. coli , the previously published PMT deletion derivative O (Petersen. S.K. et al. 1991. Recombinant derivatives of Pasteurella multocida toxin: candidates for a vaccine against Progressive Atrophic Rhinitis. Infect. Immun. 59:1387-1393 ), and one selected new vaccine candidate 4HX, were compared in shake flask experiments in the following way: E. coli K-12, MC1000 (dam), kindly provided by

Henrik ørum, harbouring plasmids pSPE680 (Petersen. S.K. et al. 1991. Recombinant derivatives of Pasteurella multocida toxin: candidates for a vaccine against Progressive Atrophic Rhinitis. Infect. Immun. 59:1387-1393); pSPE670 (ibid. ) which is a derivative O-producing pSPE680 analogue; or pSPE1038 (see Fig. 1), were grown in LB medium (Miller, J. 1972.

Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.) containing 50 μg/ml

ampicillin at 30°C. Growth was followed by measurement of the optical density of the cultures at 450 nm wavelength (OD450). After exponential growth for more than 20 generations during which the three cultures were diluted to equal optical densities, the cultures were allowed to enter the stationary phase.

1 ml culture samples were collected in the late exponential growth phase (2 samples of each culture (lxl and 1×2)) at OD₄₅₀ = 1.0 and at OD₄₅₀ = 2.0; between late exponential and stationary phase (1 sample (lxs)) at OD₄₅₀ = 4.0; in the early stationary phase (1 sample (es) at OD₄₅₀ = 6.0); and 13 h later in the late stationary phase (1 sample (ls) at OD₄₅₀ = 6.0). Each culture sample was centrifuged to pellet the bacteria, which were then resuspended in different amounts of SDS-sample buffer [62 mM Tris-HCl pH 6.8, 10% (v/v) glycerol, 2% (w/v) sodium dodecyl sulfate (SDS),

0.00125% (w/v) bromphenolblue, 0.025% (w/v) pyronine Y, 0.1 mM dithiothreitol], to normalise to OD₄₅₀ of the culture. These samples were subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using 10% (wt/vol) acrylamide gels (acrylamide/bisacrylamide ratio of 40:1), followed by Coomassie Brilliant Blue staining and destaining of the gels in 5% acetic acid - 50% ethanol (cf. Fig. 5).

Subsequently, the gel was subjected to one-dimensional laser densitometry scanning, as described in Example 2, to give the relative amounts of rPMT, derivative O or 4HX in each sample. The results are shown in Fig. 6.

In the late exponential growth phase, the production levels of the three proteins were similar. However, when the cultures entered the stationary phase an apparent preferential breakdown of derivative O resulted in lower levels of this protein per OD₄₅₀ as compared to the levels of PMT and 4HX.

This apparent breakdown is reflected in a lower maximum level of derivative O, and in a 33% loss of derivative O between early and late stationary phase. In the same time span, the amounts of the two other proteins remained fairly constant at a 300% higher level. In conclusion, the tested new amino acid-substituted mutant has a production profile which is indistinguishable from that of rPMT, and it is thus produced in higher amounts and is apparently more resistant to degradation in E. coli than the known vaccine candidate designated derivative O, which is a deletion derivative of rPMT lacking 121 amino acid residues (see also below for further stability data). rPMT and rmPMTs were produced for purification purposes, by growing the plasmid-harbouring strains of Example 1 in Luria Broth (Miller, J. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.) in the presence of ampicillin (50 μg/ml). 400 ml stationary phase cultures grown at 30°C were harvested and resuspended in 10 ml H₂O, each, sonicated and sterile-filtered as

described in Example 2. b. Purification of rPMT rPMT was purified from a crude bacterial extract of E. coli harbouring plasmid pSPE680 by anion-exchange chromatography followed by hydrophobic interaction chromatography. All steps in the purification were performed at 4°C. Sonicated, sterile-filtered bacterial extract from a 400 ml culture was dialysed overnight against 2 × 1 1 of 25 mM Tris-HCl pH 8.0 containing 1 mM PMSF. The dialysed sample was diluted to 100 ml with buffer A: 25 mM Tris-HCl pH 8.0 and applied to a column of Q-Sepharose Fast Flow (Pharmacia) which had been equilibrated in the same buffer. The total protein content in the applied material was about 200 mg. The dimensions of the column were 2 cm² × 25 cm and the flow rate was 60 ml/h

(30 cm/h). The column was washed with 2.5 bedvolumes of buffer A and adsorbed proteins were then eluted with a linear gradient from buffer A to buffer B: Buffer A + 1 M NaCl over 10 bedvolumes. The elution profile is shown in Fig. 6.

Fractions of 10 ml were collected and analyzed by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), cf. Fig. 7. The gels containing 10% acrylamide were run in a Mini-PROTEAN II apparatus (Bio-Rad) and silver stained [Morrissey, J. H. (1981). Silver Stain for Proteins in Polyacrylamide Gels: A Modified Procedure with Enhanced Uniform Sensitivity. Anal. Biochem. 117, 307-310]. During elution with a sodium chloride gradient, rPMT of high purity eluted at about 700 mM NaCl, see Figs. 6 and 7. rPMT eluted later than most other proteins in the crude extract, indicating that it is more negatively charged at pH 8 than most of the other proteins present in the extract. Fractions containing pure rPMT were pooled and stored at -20°C until the second chromatographic step described below. After each chromatographic run, the column was cleaned by washing with 0.5 M NaOH. rPMT was further purified by hydrophobic interaction chromatography. The pooled fractions from step one were adjusted to 0.72 M ammonium sulphate and pH was adjusted to 7.0. The sample was applied to a column of Phenyl Sepharose CL-4B (Pharmacia) which had previously been equilibrated in buffer one: 25 mM sodium phosphate, 0.72 M ammonium sulphate, pH

7.0. The dimensions of the column were 2 cm² × 20 cm and the flow rate was 60 ml/h (30 cm/h).

After sample application, the column was washed with two bedvolumes of buffer one. Adsorbed proteins were eluted with a linear gradient from buffer one to H₂O over 6 bedvolvimes followed by 4 bedvolumes of H₂O. Fractions of 10 ml were collected and analyzed by SDS-PAGE as above. Fractions containing pure PMT were pooled and stored at -20°C. In this step, however, only minor residual contaminants had to be removed, cf. Figs. 8A and 8B. The final product resulting from the purification was analyzed by SDS-PAGE of two-fold serial dilutions of this preparation (Fig. 9). This analysis showed that the purity of the obtained rPMT was ≥94%. c. Attempts to purify derivative O (dO) by conventional chromatographic procedures

A crude bacterial extract containing the previously described derivative O (dO) deletion protein was subjected to anion exchange chromatography as described above, cf. Fig. 10. In this case, however, most of the dO emerged in the run-through fraction and the remaining part of it eluted at about 500 mM sodium chloride as part of a mixture containing several other proteins, cf. Fig. 11. Fractions of optimal purity were, however, pooled and subjected to hydrophobic interaction chromatography as described above. The obtained dO was not of an acceptable purity, cf. Fig. 12. Therefore, it was attempted to optimize the purification procedure for dO, e.g. by changing pH during ion-exchange chromatography. These attempts, however, were unsuccessful and it was concluded that the dO cannot readily be purified by conventional column chromatography as was the case for rPMT which suggests that major structural changes have resulted from deletion of 121 amino acid residues. This is in accordance with earlier published work which indicated that the retained amino acid stretches in the N-terminal 1/4 of dO are folded in a manner different from that in native PMT. d. Laboratory-scale purification of the rmPMT 4HX protein

Crude bacterial extract containing the 4HX rmPMT protein was subjected to anion-exchange chromatography as described above for rPMT. The elution profile obtained was similar, if not identical, to that of rPMT, cf. Fig. 13. SDS-PAGE analysis of the fractions collected during the fractionation showed that 4HX of high purity eluted at about 700 mM sodium chloride, cf. Fig. 14. This similarity between the elution positions of rPMT and 4HX indicates that the chromatographic properties of rPMT are conserved in the mutant protein 4HX, and it suggests that only minor structural changes, (yet functionally important, cf. Examples 2, 4 and 5) have been introduced by the amino acid substitutions in 4HX. 4HX was further purified by hydrophobic interaction chromatography as described above. As for rPMT this step was only required to remove minor residual contaminants, see Fig.

3.11. The final product of the purification was analyzed by SDS-PAGE of two-fold serial dilutions of the preparation of 4HX, cf. Fig. 16. Similar to rPMT, the purity was ≥94%. The yield of purified 4HX was generally 3.5% (w/w) of total protein in the starting material. For example, crude extract containing 200 mg of total protein resulted in 7 mg of purified 4HX. Protein concentrations were determined according to Bradford (1976) (A Rapid and Sensitive Method for the Quantitation of Microgram Quantities of Proteins Utilizing the Principle of Protein-Dye Binding. Anal. Biochem. 72, 248-254). It was not possible to accurately determine the recovery of 4HX during the purification. However, the recovery was estimated to about 50% by determination of protein concentration in crude extract and final product and comparison of SDS-PAGE lanes with dilution series of crude extract and purified 4HX protein.

e. Preparative-scale purification of the 4HX protein

In order to test the performance of the purification procedure described above, at a larger scale, 4HX was purified from crude extract of 5 1 of an overnight culture of the 4HX producing strain. Cells from 5 1 of overnight culture were harvested, resuspended in H₂O, sonicated and sterile filtered as described above (final volume 100 ml) . All steps in the purification were performed at room temperature (20-22°C).

The bacterial extract was dialysed overnight against 2 × 5 1 of 25 mM Tris-HCl pH 8.0 containing 1 mM PMSF. The dialysed sample was diluted to 1750 ml with buffer A: 25 mM Tris-HCl pH 8.0 and applied to a column of Q-Sepharose Fast Flow which had been equilibrated in the same buffer. The dimensions of the column were 20 cm² × 25 cm. The flow rate was 600 ml/h (30 cm/h). The column was washed with 2.5 bedvolumes of buffer A and adsorbed proteins were then eluted with a linear gradient from buffer A to buffer B: Buffer A + 1 M NaCl over 10 bedvolumes.

After each chromatographic run, the column was cleaned by washing with 0.5 M NaOH. The elution profile obtained during anion-exchange chromatography on Q-Sepharose (Fig. 17) was similar to that obtained in laboratory-scale purification (Fig. 13). Fractions of 170 ml were collected during the gradient elution and analyzed by SDS-PAGE as described above. This analysis showed that 4HX of high purity eluted in fractions 19-20 corresponding to 650-700 mM sodium chloride.

Fractions 19-20 were pooled and subjected to further purification by hydrophobic interaction chromatography.

The pooled fractions from step one were adjusted to 0.72 M ammonium sulphate and pH was adjusted to 7.0. The sample was then applied to a column of Phenyl Sepharose CL-4B which had previously been equilibrated in buffer one: 25 mM sodium phosphate, 0.72 M ammonium sulphate, pH 7.0. The dimensions of the column were 20 cm² × 20 cm and the flow rate was

600 ml/h (30 cm/h). After sample application the column was washed with 2 bedvolumes of buffer one. Adsorbed proteins were eluted with a linear gradient from buffer one to H₂O over 6 bedvolumes followed by washing with 4 bedvolumes of H₂O. Fractions of 170 ml were collected during the elution and analyzed by SDS-PAGE as above. Fractions containing pure 4HX were pooled and stored at -20°C. SDS-PAGE analysis of crude extract, product of step one and product of step two (Fig. 18) indicated that no significant further purification was achieved during step two. Thus, the purification protocol for 4HX may well be simplified to consist only of an ion-exchange step. The purity of the product (after step two) was assessed by SDS-PAGE of two-fold dilutions of the purified 4HX preparation, cf. Fig. 19. This analysis showed a purity of ≥91%.

Similar to the yield obtained in laboratory-scale purification, 3.5% of the total protein was obtained as purified 4HX. (225 mg from 6475 mg of total protein in crude extract) f. Testing for identity of 4HX preparations

In order to verify the identity of the purified preparations of 4HX resulting from laboratory-scale as well as preparative-scale purification, the preparations described above were analyzed by SDS-PAGE followed by immunoblotting with a polyclonal rabbit antibody raised against rPMT, cf. Fig. 20. rPMT was included as a positive control (lane 1)

and another recombinant E. coli protein, nuclease from

Serratia marcescens (M_r 30,000) was included as a negative control (lane 4). As seen in Fig. 20, laboratory-scale purified as well as preparative-scale purified 4HX exhibited staining only at a M_r identical to the M_r of rPMT. g. Purification of +GG1203 Crude bacterial extract (strain SPE1020) containing the

+GG1203 rmPMT protein was subjected to anion-exchange chromatography as described above for rPMT. The elution profile obtained (cf. Fig. 26) was similar to the profiles seen during purification of rPMT (Fig. 6) and 4HX (Fig. 13). SDS-PAGE analysis of the fractions collected during the

fractionation showed that +GG1203 of high purity eluted at 700-800 mM sodium chloride, cf. Fig. 27. This elution range is similar to the ones observed for rPMT and 4HX, as

described above. The similarity among the elution positions of rPMT, 4HX and +GG1203 indicates that the chromatographic properties of rPMT are conserved in the mutant proteins 4HX and +GG1203. +GG1203 was further purified by hydrophobic interaction chromatography as described above for rPMT, in order to remove residual contaminating proteins, cf. Fig. 28. The final product of the purification was analyzed by SDS- PAGE of two-fold serial dilutions of the preparation of

+GG1203, cf. Fig. 29. The purity was estimated to be about 90%. The yield of purified +GG1203 was 2.5 % (10 mg of purified +GG1203 obtained from 400 mg of total protein in crude bacterial extract). h. Purification of ΔH1223

Crude bacterial extract (strain SPE1234) containing the

ΔH1223 rmPMT protein was subjected to anion-exchange chromatography as described above for rPMT. The elution profile obtained (Fig. 30) was similar, but not identical, to the profiles seen during purification of rPMT (Fig. 6), 4HX (Fig. 13) and +GG1203 (Fig. 26). SDS-PAGE analysis of the fractions collected during the fractionation showed that ΔH1223 of high purity eluted at about 700 mM sodium chloride, similar to rPMT, 4HX and +GG1203.

ΔH1223 was further purified by hydrophobic interaction chromatography as described above for rPMT, cf. Fig. 31. The final product of the purification was analyzed by SDS-PAGE of two-fold serial dilutions of the preparation of ΔH1223, cf. Fig. 32. This analysis showed that the purity of the purified ΔH1223 was 95 %. The yield of purified ΔH1223 was 0.6% (3 mg of purified ΔH1223 obtained from 485 mg of total protein in crude bacterial extract). i. Purification of 1132"β-gal

This fusion protein was purified from crude bacterial extract (strain SPE1134) by ammonium sulphate precipitation and affinity chromatography on the substrate analogue for β- galactosidase p-aminophenyl-β-D-thiogalactopyranosid coupled to agarose (APTG-agarose). The purification was carried out essentially as described by Germino et al. (1983), Proc.

Natl. Acad. Sci. USA, 80, 6848-6852. The crude bacterial extract was clarified by centrifugation and mixed with one volume of 80% saturated ammonium sulphate (room temperature). The precipitate was collected by centrifugation. The

supernatant, containing only 1% of the total β-galactosidase activity was discarded. The pellet was resuspended in 3 ml of 10 mM Tris-Cl, 250 mM NaCl, 10 mM MgCl₂, 1 mM EDTA, 0.1% Triton X-100, pH 7.6 (= "buffer") and dialysed against 500 ml of buffer overnight at 4°C. The dialysed sample was applied on a 1 ml column of APTG agarose (Sigma Chemicals, # A-8648) which had been equilibrated in buffer. The chromatography was performed at room temperature. Fractions of 1 ml were collected throughout the purification. The column was washed with 10 ml of buffer followed by 5 ml of buffer without

Triton X-100. Adsorbed proteins were eluted with 1) 4 ml of 0.1 M Tris-Cl, pH 10.0, 2) 4 ml of Tris, pH 10.8, 3) 4 ml of 0.1 M sodium carbonate, pH 11.0 and 4) 4 ml of 6 M urea in buffer. The fractions were neutralized with 1 M sodium acetate, pH 5 immediately after collection, and analyzed by SDS- PAGE and β-galactosidase assay. These analyses showed that most of the 1132"β-gal protein was eluted by 0.1 M sodium carbonate pH 11.0, whereas only a minor fraction was eluted at pH 10.0 and 10.8. The β-galactosidase activity, however, was relatively higher in the fractions collected at pH 10, probably because β-galactosidase is inactivated at a pH around 11. Fractions containing 1132"β-gal of high purity were selected among the fractions eluted at pH 10.0, 10.8 and 11.0 and pooled. The overall recovery of β-galactosidase activity was 13%. The low recovery of activity is probably due to (partial) inactivation of β-galactosidase at the strongly alkaline pH which was required to elute this β-gal fusion protein from the APTG column, β-galactosidase fusion proteins are generally reported to elute from the APTG matrix at pH 10, but this was obviously not possible in this case. The crude extract used for the purification and the purified fusion protein 1132"β-gal were analyzed by SDS-PAGE, cf. Fig. 33. In addition to the main component having a M_r of about 200,000 dalton, the purified preparation exhibited a band co migrating with β-galactosidase at M_r 116,000. A band at a M_r of about 70,000 was also observed. j. Resistance to possible residual protease activity and resistance to trypsin of dO, rPMT, 4HX, +GG1203, ΔH1223 and 1132"β-gal

4HX, +GG1203, ΔH1223 and 1132"β-gal were compared with the derivative O regarding susceptibility to proteolytic degradation, and the rPMT was included as a reference protein in these experiments. Preparations of purified rmPMTs and rPMT were obtained as described above (laboratory scale preparation). A preparation of purified dO was obtained by immunoaffinity chromatography using monoclonal antibodies coupled to agarose as described by Petersen et al., 1991 (Infection and Immunity, 59, 1387-1393). In the first experiment, the preparations of 4HX, dO and rPMT were diluted to the same protein concentration (0.1 mg/ml); 25mM Tris-HCl, pH 7.5 was added in order to secure the same pH in the samples, and the samples were subsequently incubated at 37°C. Samples were collected at time zero (control), after 24 h, and after 48 h and analyzed by SDS-PAGE, cf. Fig. 21. No significant degradation was observed for any of the three tested proteins during the incubation. The intensity of the bands constituting rPMT, derivative O or 4HX, respectively, remained constant after 24 and 48 h. No significant appearance of low molecular weight bands indicating degradation, was observed. It was concluded from these experiments that 4HX, dO and rmPMT exhibit identical properties regarding stability under these incubation conditions.

In a further experiment, the stability of rPMT, derivative O and 4HX in the presence of the proteolytic enzyme trypsin was compared. Samples of the three proteins were adjusted to the same protein concentration and pH as above and incubated at 37°C for 24 h in the presence of 1, 2 or 5 wt% trypsin, respectively. A preparation of derivative O which had been enriched during purification attempts by chromatography on Q- Sepharose and Phenyl Sepharose was included in the comparison. The trypsin-treated samples were compared by SDS-PAGE analysis with samples collected before the addition of trypsin, cf. Fig. 22 (rPMT), Fig. 23 (4HX), Fig. 24 (purified derivative O) and Fig. 25 (enriched derivative O). No significant degradation was observed with 1, 2, or 5 wt% trypsin in samples containing rPMT (cf. Fig. 22) or 4HX (cf. Fig.

23). The derivative O, however, was highly sensitive to the trypsin treatment. SDS-PAGE of samples containing immunoaffinity chromatography-purified derivative O showed no residual bands after trypsin treatment, cf. Fig. 24. In the samples containing enriched derivative O, the derivative O band at 133 kD had disappeared after trypsin treatment. Other components of lower molecular weight, however, were still present after the trypsin treatment, cf. Fig. 25. These bands assumingly represent trypsin resistant E. coli proteins contaminating this preparation enriched for derivative O.

From the above experiments it was concluded that 4HX

(purified as described above) and the derivative O (purified by immunoaffinity chromatography) are stable during incubation at 37°C for 24 or 48 h in the absence of proteolytic enzyme. However, in the presence of trypsin it is evident that the derivative O is susceptible to proteolytic degradation, whereas the 4HX rmPMT (and rPMT) is stable under the same conditions. Accordingly, the trypsin resistance of rPMT appears to be preserved in 4HX, suggesting that the amino acid substitutions in this rmPMT have not introduced any significant structural changes in the tertiary structure of this protein, irrespective of the functional significance of the amino acid substitutions in the rmPMT, cf. Example 4 below.

In a third experiment, the susceptibility to proteolytic degradation of +GG1203, ΔH1223 and 1132"β-gal was examined by incubation of the purified preparations at 37°C for 24 h without proteolytic enzyme added and with addition of 1, 2 or 5 wt% trypsin, as described above for rPMT, dO and 4HX. After incubation, samples were analyzed by SDS-PAGE, cf. Fig. 34 (+GG1203), Fig. 35 (ΔH1223) and Fig. 36 (1132"β-Gal). dO and 4HX were included as controls in this experiment and performed as described above (data not shown).

No degradation of +GG1203 was observed after 24 h at 37°C without trypsin. In the presence of trypsin the intensity of the +GG1203 band was not significantly reduced, but two faint bands of lower molecular weight appeared, cf. Fig. 34. Thus, +GG1203 was essentially resistant to degradation by the proteolytic enzyme. Similarly, ΔH1223 was stable during incubation at 37°C for 24 h without trypsin. In the presence of proteolytic enzyme, however, numerous, tightly spaced, faint bands of lower molecular weight appeared, and the intensity of the ΔH1223 band was slightly reduced, cf. Fig. 35. The fusion protein 1132"β-gal was also stable during incubation at 37°C for 24 h without trypsin. This protein, however, was completely degraded during incubation with trypsin, since neither the 1132"β-gal band nor any degradation products were observed, cf. Fig. 36.

From this experiment, it was concluded that +GG1203 is resistant to degradation by the proteolytic enzyme trypsin, whereas ΔH1223 is slightly sensitive. In contrast, the 1132"β-gal fusion protein, like the dO protein, is susceptible to proteolytic degradation. k. Conclusions

To summarize, it was shown that the inactivated rmPMTs 4HX, +GG1203, ΔH1223, and 1132"0-gal can be purified to 90% purity in milligram amounts from 400 ml shake flask cultures, by simple and inexpensive chromatographic procedures. Very similar elution profiles during anion-exchange chromatography of rPMT, 4HX, +GG1203, and ΔH1223 indicated that these rmPMTs have maintained the overall structure of rPMT, and thus of native PMT. Moreover, the purified rmPMT preparations were devoid of residual protease activity, since no degradation was detected after incubation at 37C for 24 h. Finally, 4HX and +GG1203 are highly resistant to trypsin, ΔH1223 is slightly sensitive, whereas 1132"β-gal and dO are highly susceptible to tryptic digestion.

EXAMPLE 4

Toxicity of a selected recombinant mutant PMT protein

Earlier studies have indicated that the function of inactivated PMT was equally affected in different biological assays (Petersen S.K. et al. 1991. Recombinant derivatives of

Pasteurella multocida toxin: Candidates for a vaccine against Progressive Atrophic Rhinitis. Infect. Immun. 55:1387-1393). The absence of any detectable mitogenic effect of 4HX therefore suggested a general lack of biological function of this protein. This issue was addressed by assaying the toxicity of 4HX towards mice, essentially as described by Pedersen (Pedersen, K.B. 1983. Cultural and serological diagnosis of atrophic rhinitis in pigs, p. 22-31. In K. B. Pedersen and N. C. Nielsen (eds.), Atrophic rhinitis in pigs. CEC Agriculture EUR, 8643 EN, EEC, Luxembourg): adult CFlxBALB/c mice

received intraperitoneal injections with various concentrations of rPMT (positive control), bovine serum albumin (BSA)- (negative control), or 4HX, in 200 μl sterile-filtered phosphate-buffered saline with 0.02% normal mouse serum (DAKO, Denmark). The concentrations of rPMT and rmPMT samples were determined by amino acid analysis (Barkholt, V. and A. L.

Jensen. 1989. Amino acid analysis: Determination of cysteine plus half-cysteine in proteins after hydrochloric acid hydrolysis with a disulfide compound as additive. Anal. Biochem. 177:318-322).

The mice were divided into 12 groups of 3 mice, all three receiving the same protein dose and formulation, and the number of dead mice in each group within 5 days after chal lenge was recorded. Table III summarizes the results of this experiment.

The data indicate an LD50 of rPMT of 4-16 ng which is in accordance with the LD₅₀ values disclosed in WO 89/09617 (supra).

Table III also shows that no mice subjected to 4HX died, although amounts up to about 5000 times the LD50 of rPMT were used. In conclusion, 4HX was as atoxic as BSA under the conditions tested. In similar experiments, the toxicities of +GG1203 and 1132"- β-gal, and of ΔH1223, respectively, were compared to that of rPMT. The experiments were performed exactly as described above for 4HX, and the resulting data are summarized in tables IV and V.

EXAMPLE 5

Immunogenicity of selected vaccine candidates

Earlier studies have shown that induction of anti-PMT antibodies in pigs is closely correlated to protection against Progressive Atrophic Rhinitis (Nielsen et al., 1991, Can. J. Vet. Res., 55, 128-138). A suitable model system for testing the immunogenicity of inactivated PMT in pigs involves vaccination of mice (Petersen, S. K., N. T. Foged, A Bording, J. P. Nielsen, H. K. Riemann, and P. L. Frandsen, 1991. Recombinant derivatives of Pasteurella multocida toxin: candidates for a vaccine against Progressive Atrophic rhinitis. Infect. Immun. 59:1387-1393). A modification of this model, using CFlxBALB/c mice instead of inbred BALB/c mice was used to verify the immunogenicity of purified 4HX, and accordingly its usefulness as the active component of vaccines. The concentrations of rPMT and rmPMT samples were determined by amino acid analysis according to the method as defined above.

The vaccine preparations tested comprised different concentrations of formaldehyde inactivated purified recombinant PMT (frPMT) (positive control), or the genetically inactivated, purified 4HX protein, in 250 μl phosphate-buffered saline (2 mM sodium phosphate, pH 7.2, 6 mM NaCl) (PBS) . Purification of the vaccine candidate 4HX and of rPMT was carried out as described in Example 2 and formaldehyde treatment of rPMT was performed as described by Foged (Foged, N. T. 1991. Detection of stable epitopes on formaldehyde-detoxified Pasteurella mul tocida toxin by monoclonal antibodies. Vaccine 9:817-824): purified rPMT (0.5 mg/ml, 1 ml) was dialysed against 1 1 PBS, pH 7.6 with 0.37% (w/v) formaldehyde, 0.1% methanol at 20°C. After 72 h lysine-HCl was added to the dialysis buffer to a final concentration of 0.1% (w/v) , and dialysis was continued for another 8 h at 4°C. This was followed by several rounds of dialysis against PBS pH 7.2 at 4°C to give the final frPMT preparation. All vaccine preparations were adsorbed to 0.26% aluminium hydroxide gel (Alhydrogel grade A; Superfos, Denmark). Adult CFlxBALB/c mice were vaccinated twice at two weeks interval. Two weeks after the second vaccination the mice were challenged by intraperitoneal injection of 100 ng rPMT in 200 μl PBS with 0.02% normal mouse serum (Dako, Denmark). The day before the challenge blood samples were taken from all mice, and subsequently analyzed for anti-rPMT antibodies using the following ELISA method: Microtiter plates (96 well High Capacity, Nunc, Denmark) were coated with rPMT diluted to 3 μg/ml in PBS coating buffer (25 nM NaH₂PO₄, 75 mM

NaH₂PO₄, 145 mM NaCl, 0.1% Tween 20, pH 7.2), the wells were washed 3 times with washing buffer (PBS + 0.5 M NaCl, 0.1% Tween 20, pH 7.2) and incubated with sera diluted in washing buffer for 2 h at 20°C. After 3 more washings as described above, secondary antibody (Horse Radish Peroxidase conjugated Rabbit anti-Mouse IgG antibody, DAKO, Denmark) diluted 1:5000 in washing buffer was added and allowed to react for 2 h at 20°C. After 3 further washings as above, a colour reaction was allowed to occur by the addition of 0.67 mg/ml orthophenylenediamine (OPD), 0.0125% H₂O₂, 34.7 mM citric acid, 66.7 mM NaH₂PO₄. The reaction was stopped by the addition of 1 M H₂SO₄. The absorption at 492 nm was subsequently determined on a Titertek Multiskan II Plus ELISA reader (Labsystems OY, Finland), and the anti-rPMT antibody titre of each serum sample was determined as the dilution giving an absorbance at 492 nm of 0.5. The results of the vaccination experiment are summarized in Tables VI and VII and Fig. 37 below:

The above data show that 4HX is as efficient as frPMT (and consequently as efficient as the above-mentioned derivative O as disclosed in WO 89/09617) in its protection of mice in this model system.

The anti-rPMT serum titre data for this experiment are presented in Fig. 37. Immunization with 25 or 5 μg/ml of frPMT or rmPMT resulted in serum conversion of all animals. Immunization with 1 μg/ml also resulted in serum conversion of all 5 animals immunized with frPMT, whereas 4 animals were positive in the group immunized with the rmPMT; the animal in which seroconversion did not occur, died upon challenge. No animals seroconverted after immunization with doses of 0.2 or 0.08 μg/ml. Accordingly, a 100% correlation between seroconversion and protection against challenge was found with rmPMT. There was no significant difference (at the 5% significance level) between the serum anti-rPMT titres of animals vaccinated with frPMT and rmPMT, respectively, when these data were analyzed by the Student's t-test, indicating that these vaccine compositions give rise to similar anti-rPMT antibody titres in the vaccinated animals and to similar dose-response relationships. In a similar experiment, the immunogenicities of +GG1203 and 1132"β-gal were assessed by comparison to 4HX. Purification of rmPMTs, vaccination, blood sampling, and challenge was performed as described above, with the exception that the challenge dose was changed to 200 ng. The results of this experiment are shown in Tables VIII and IX, and Fig. 38.

The data in Table VIII, IX and Fig. 38 demonstrate that +GG1203 is as efficient as 4HX in its protection of mice against challenge with a lethal dose of rPMT and in its induction of anti-rPMT serum titres in vaccinated mice in this model system. Immunization with 5 and 25 μg/ml +GG1203 or 2.5 and 12.5 μg/ml 4HX resulted in serum conversion of all animals. This correspondingly is reflected in low standard deviations between 0.7 and 1.3. Immunization with 1 μg/ml +GG1203 gave serum conversion in all five animals in the group whereas immunization with 0.5 μg/ml 4HX resulted in serum conversion of four of the five animals in the group and the mean titre 6.1 in contrast to 7.5 for +GG1203. Accordingly, the standard deviation is low (0.7) for +GG1203 and high (3.4) for 4HX. With higher doses, the resulting titres for 4HX and +GG1203 are comparable.

In contrast to 4HX and +GG1203, ll32"β-gal was not able to elicit an immune response with doses up to 2 μg/ml. Immunization with 10 μg/ml 1132"β-gal resulted in serum conversion of two of the five animals in the group with the mean titre 3.6 which is considerably lower than the resulting titres for immunization with similar doses of 4HX and +GG1203. In conclusion, +GG1203, 4HX and 1132"β-gal all gave protecting antibodies against rPMT. +GG1203 and 4HX were comparable in the ability to raise antibodies reacting with rPMT whereas 1132"β-gal needed higher doses in order to raise an antibody response comparable to the responses for 4HX and +GG1203. There was no significant difference (at the 5% significance level) between 4HX and +GG1203 in their ability to cause serum titre's against rPMT, indicating that these vaccine compositions give rise to similar anti-PMT antibody titres in the vaccinated animals and to similar dose-response relationships. For all three vaccine candidates, there was a high degree of correlation between serum anti-rPMT antibody titre and protection against challenge with rPMT. However, one +GG1203 vaccinated mouse seroconverted, but was not protected against challenge.

In a third experiment, the immunogenicity of ΔH1223 was assessed by comparison to 4HX. Purification of rmPMTs, vaccination, blood sampling, and challenge was performed as described above, the challenge dose was 200 ng. The results of this experiment are shown in Tables X and XI, and Fig. 39.

The data in Table X, XI, and Fig. 39 demonstrate that ΔH1223 is as at least as efficient as 4HX in its protection of mice against rPMT in this model system. The lowest immunization dose gave no serum response with either of the immunogens. At 1 μg/ml (0.5 μg/ml) and higher doses, all animals serum converted. The titres were similar at the 5% significance level when analyzed by the Student's t-test. This indicated that these two vaccine compositions give rise to similar anti-PMT antibody titres in the vaccinated animals and to similar dose-response relationships. Nevertheless, two of the mice vaccinated with 0.2 μg/ml ΔH1223 survived challenge with 200 ng of rPMT.

Claims

Conclusion Vaccination with purified substitution, deletion, insertion, or fusion rmPMTs was shown to protect mice against challenge with a lethal dose of purified rPMT. In most cases this protection correlated with induction of serum antibodies against rPMT. However, as compared to the other rmPMTs and frPMT, the selected fusion protein required immunization with higher doses in order to induce a similar serum response and protection. The selected substitution, insertion, and deletion rmPMTs were equally efficient in inducing serum titres, and comparable in their ability to cause protection against challenge with rPMT, although ΔH1223 in low amounts gave protection of two vaccinated mice which did not seroconvert, and +GG1203 in one case did not protect a seroconverted mouse against challenge. The selected substitution, deletion, and insertion rmPMTs were as efficient as frPMT in inducing serum anti-rPMT antibody titres in mice and protection of mice against challenge with rPMT. In WO 89/9617, however, it is disclosed that derivative O-vaccinated animals show development of significantly lower serum titres against PMT as compared to animals vaccinated with fPMT. It is contemplated that this difference is due to the expected smaller change in structure of these rmPMTs as compared to the truncated derivative O rrther than to the recombinant origin of the PMT used in this study, since rPMT and PMT in all known biological systems seem to be indiscriminable. The improved immunogenicity, the higher production yield, and improved stability of the rmPMTs as compared to the previously described restriction enzyme-deletion O-derivative and the retained PMT-associated physical properties of the rmPMTs as reflected by the readiness of its purification in a simple conventional purification procedure, make the rmPMTs as defined herein commercially interesting candidates for a vaccine component for efficient protection diseases caused by Pasteurella multocida infections such as Progressive Atrophic Rhinitis. The selected substitution, insertion and deletion mutant genes all comprise at least three basepair mutations and accordingly, the risk of reversion to wildtype toxA is therefore in these cases negligible. CLAIMS

1. A recombinant mutant protein encoded at least in part by a DNA sequence which is derivable from the toxA gene including the toxA gene coding for Pasteurella multocida ssp multocida 45/78 toxin (PMT), by substitution, deletion or insertion of one or more codons, said protein being capable of binding to antibodies raised against PMT encoded by Pasteurella multocida ssp multocida 45/78, the protein having a calculated molecular weight of at least 140 kD.

2. A protein according to claim 1 which is encoded by a DNA sequence which is derivable from the toxA gene coding for Pasteurella mul tocida ssp mul tocida 45/78 toxin by substitution, deletion or insertion of one or more codons hereof downstream of a unique EcoRI restriction site in said gene.

3. A protein according to claim 1 which has a molecular weight of at least 141 kD.

4. A protein according to claim 3 which has a molecular weight of at least 143 kD.

5. A protein according to claim 1 which is at least 80% purifiable by a purification procedure comprising anion-exchange chromatography using Q-Sepharose Fast Flow (Pharmacia) in a 2 cm² × 25 cm chromatography column when applied hereto in the form of an extract of bacteria expressing the protein, the amount of total bacterial protein applied to the column being about 200 mg and the flow rate being about 30 cm/h, and eluting absorbed protein with a linear gradient of buffer containing 0 to 1000 mM of ΝaCl, if necessary followed by hydrophobic interaction chromatography of eluted fractions using a column of Phenyl Sepharose CL-4B (Pharmacia) having the dimensions of 2 cm² × 20 cm and at a flow rate of 30 cm/h, eluting absorbed protein, and measuring the amount hereof.

6. A protein according to claim 5 which is at least 94% purifiable by the purification procedure as defined.

7. A protein according to claim 1 which is a protein encoded by a DNA sequence derived from the toxA gene by substituting at least one codon coding for an amino acid or amino acids of position(s) between position 1131 and 1285 with a

codon/codons coding for a different/different amino acid(s).

8. A protein according to claim 7 which is encoded by a DNA sequence in which at least two codons are substituted such as at least three codons.

9. A protein according to claim 8 which is encoded by a DNA sequence in which at least four codons are substituted.

10. A protein according to claim 7 which is encoded by a DNA sequence in which the codon(s) being substituted is/are codon(s) coding for an amino acid selected from serine, histidine, glutamine and threonine.

11. A protein according to claim 7 which is encoded by a DNA sequence in which the substituted codon (s) is/are a

codon/codons of position(s) between position 1175 and 1285.

12. A protein according to claim 11 which is encoded by a DNA sequence in which the substituted codon(s) is/are a

codon/codons of position(s) between 1200 and 1230.

13. A protein according to claim 1 which is a protein encoded by a DNA sequence derived from the toxA gene by deleting at least one codon coding for an amino acid or amino acids of position (s) between position 1131 and 1285.

14. A protein according to claim 13 which is encoded by a DNA sequence in which at least two codons are deleted such as at least three codons.

15. A protein according to claim 14 which is encoded by a DNA sequence in which at least four codons are deleted.

16. A protein according to claim 15 which is encoded by a DNA sequence in which at least seven codons are deleted.

17. A protein according to claim 13 which is encoded by a DNA sequence in which the codon (s) being deleted is/are codon (s) coding for an amino acid selected from lysine, glutamic acid, phenylalanine, alanine, valine and aspartic acid.

18. A protein according to claim 13 which is encoded by a DNA sequence in which the deleted codon (s) is/are a codon/codons of position (s) between position 1175 and 1285.

19. A protein according to claim 18 which is encoded by a DNA sequence in which the deleted codon (s) is/are a codon/codons of position (s) between 1215 and 1285.

20. A protein according to claim 1 which is a protein encoded by a DNA sequence derived from the toxA gene by inserting at least one codon downstream of a codon coding for an amino acid or amino acids at amino acid position (s) between position 1131 and 1285.

21. A protein according to claim 20 which is encoded by a DNA sequence in which at least two codons are inserted such as at least three codons.

22. A protein according to claim 21 which is encoded by a DNA sequence in which at least four codons are inserted.

23. A protein according to claim 20 which is encoded by a DNA sequence in which the codon being inserted is a codon coding for glycine.

24. A protein according to claim 23 which is encoded by a DNA sequence in which the inserted codon(s) is/are inserted at position(s) between position 1175 and 1285.

25. A protein according to claim 24 which is encoded by a DNA sequence in which the inserted codon(s) is/are inserted at position(s) between 1200 and 1230.

26. A protein according to claim 1 which has a mitogenic potency as determined by measuring the proliferative effect (PE) of the protein on NIH3T3 cells, which is at the most 75% relative to that of rPMT, said measuring comprising in a first step the preparation of a standard curve based on PE measurements of serial dilutions of a suspension containing about 20 ng/ml of PMT and calculating the PE according to the formula

where N_x is the mean cell number in 3 dishes with sample dilution x, N_pmt is the mean cell number in 3 dishes with whole-cell extract of E. coli SPE1036 harbouring pSPE680 encoding for PMT, said extract containing a final concentration of 20 ng/ml PMT, and N_O is the mean cell number in 3 dishes with E. coli DH5α whole-cell extract, said dishes having a confluent monolayer of NIH3T3 cells, and in a second step measuring the PE of a suspension of the protein at an amount in the range of 1-100 ng/ml, and calculating the PE relative to the PMT using said standard curve.

27. A protein according to claim 26 which has a mitogenic potency calculated as defined in claim 26 of at the most 50%, preferably at the most 35%, more preferably at the most 25%, most preferably at the most 10% and in particular at the most 5%.

28. A protein according to claim 1 which is at least 90% purifiable by the purification procedure as defined in claim 5.

29. A protein according to claim 28 which is at least 95% purifiable.

30. A protein according to claim 1 which in the anion- exchange purification step as defined in claim 5 elutes at about 700 mM NaCl.

31. A protein according to claim 1 which has an LD₅₀ of at least 51.2 μg as determined by administering the protein intraperitoneally to mice.

32. A protein according to claim 1 which when administered to mice twice at an interval of two weeks in a vaccine preparation comprising at least 1.0 μg/ml of the protein adsorbed to an aluminium hydroxide gel, protects at least 80% of the mice against the lethal effect of 250 ng PMT injected intraperitoneally two weeks subsequent to the second administration of the vaccine, an anti-PMT titre being induced in the protected mice.

33. A protein according to claim 32 which when administered in a vaccine preparation as defined in claim 31 and comprising 5.0 μg/ml of the adsorbed protein protects 100% of the mice.

34. A protein according to claim 1 which is selected from the group consisting of 4HX, ΔH1223, +GG1203 and 1132"β-gal.

35. A protein according to claim 1 which when administered to mice in the vaccine composition as defined in Example 5 comprising 1.0 μg/ml gives rise to an anti-rPMT titre determined as defined in Example 5 of at least 6.2.

36. A protein according to claim 35 which when administered to mice in the vaccine composition as defined in Example 5 comprising 1.0 μg/ml gives rise to an anti-rPMT titre determined as defined in Example 5 of at least 9.

37. A protein according to claim 1 which is a fusion protein encoded by a DNA sequence, the sequence comprising a first gene coding for an immunogenic protein which is reactive with an antibody reacting with the Pasteurella multocida spp mul tocida 45/78 toxin (PMT) encoded by the toxA gene, said gene being derived from a replicon comprising the toxA gene by substitution, deletion or insertion of one or more codons of said toxA gene, and a second gene coding for a second immunogenic protein, the first and the second protein being expressible as a fusion protein.

38. A protein according to claim 37 wherein the first

immunogenic protein has a calculated molecular weight of at least 140 kD.

39. A protein according to claim 37 wherein the second immunogenic protein is one expressed by a pathogenic organism, the protein conferring protection against diseases caused by Pasteurella multocida and said pathogenic organism.

40. A process of preparing a recombinant mutant protein capable of binding to antibodies raised against the Pasteurella multocida ssp multocida 45/78 toxin (PMT) and having a molecular weight of at least 140 kD, comprising the steps of:

(i) isolating a DNA sequence comprising a toxA gene coding for a Pasteurella multocida osteolytic toxin (PMT), including the gene coding for Pasteurella multocida ssp multocida 45/78 PMT, or a DNA sequence comprising a sequence, the gene product of which is reactive with an antibody raised against Pasteurella multocida ssp multocida 45/78 PMT, (ii) subjecting said DNA sequence to a mutagenization treatment causing substitution, deletion or insertion of one or more codons to obtain a mutated DNA sequence coding for the recombinant mutant protein, (iii) inserting the resulting mutated DNA sequence into a replicon,

(iv) transforming with the replicon a cell in which said replicon is replicated and in which the mutated DNA sequence coding for the recombinant mutant protein is expressible, (v) culturing the transformed cell under conditions where the mutated DNA sequence is expressed, and

(vi) harvesting the mutant recombinant protein from the culture.

41. A process according to claim 40 wherein the DNA sequence being isolated in step (i) is the toxA gene coding for

Pasteurella multocida spp multocida 45/78 toxin (PMT).

42. A process according to claim 41 wherein the mutagenization treatment of step (ii) causes substitution, deletion or insertion of one or more codons downstream of a unique EcoRI site in said gene.

43. A process according to claim 40 wherein the mutated DNA sequence obtained in step (ii) is a sequence coding for a recombinant mutant protein having a molecular weight of at least 141 kD.

44. A process according to claim 43 wherein the mutated DNA sequence obtained in step (ii) is a sequence coding for a recombinant mutant protein having a molecular weight of at least 143 kD.

45. A process according to claim 41 wherein the mutagenization treatment of step(ii) causes substitution of at least one codon coding for an amino acid of position between position 1131 and 1285 with a codon coding for a different amino acid.

46. A process according to claim 45 wherein the mutagenization treatment of step (ii) causes substitution of at least two codons such as at least three codons.

47. A process according to claim 46 wherein the mutagenization treatment of step (ii) causes substitution of at least four codons.

48. A process according to claim 45 wherein the codon(s) being substituted is/are a codon/codons coding for an amino acid/amino acids selected from serine, histidine, glutamine and threonine.

49. A process according to claim 45 wherein the mutagenization treatment of step (ii) causes substitution of at least one codon coding for an amino acid of position between position 1175 and 1285 with a codon coding for a different amino acid.

50. A process according to claim 49 wherein the mutagenization treatment of step (ii) causes substitution of at least one codon coding for an amino acid of position between position 1200 and 1230.

51. A process according to claim 41 wherein the mutagenization treatment of step (ii) causes deletion of at least one codon coding for an amino acid of position between position 1131 and 1285.

52. A process according to claim 51 wherein the mutagenization treatment of step (ii) causes deletion of at least two codons such as at least three codons.

53. A process according to claim 52 wherein the mutagenization treatment of step (ii) causes deletion of at least four codons.

54. A process according to claim 53 wherein the mutagenization treatment of step (ii) causes deletion of at least seven codons.

55. A process according to claim 51 wherein the codon(s) being deleted is/are codon (s) coding for an amino acid selected from lysine, glumatic acid, phenylalanine, alanine, valine and aspartic acid.

56. A process according to claim 51 wherein the mutagenization treatment of step (ii) causes deletion of at least one codon coding for an amino acid of position between position 1175 and 1285.

57. A process according to claim 56 wherein the mutagenization treatment of step (ii) causes deletion of at least one codon coding for an amino acid of position between position 1215 and 1285.

58. A process according to claim 41 wherein the mutagenization treatment of step (ii) causes insertion of at least one codon downstream of a codon coding for an amino acid of position between position 1131 and 1285.

59. A process according to claim 58 wherein the mutagenization treatment of step (ii) causes insertion of at least two codons such as at least three codons.

60. A process according to claim 59 wherein the mutagenization treatment of step (ii) causes insertion of at least four codons.

61. A process according to claim 58 wherein at least one codon being inserted is a codon coding for glycine.

62. A process according to claim 58 wherein the mutagenization treatment of step (ii) causes insertion of at least one codon downstream of a codon coding for an amino acid of position between position 1175 and 1285.

63. A process according to claim 62 wherein the mutagenization treatment of step (ii) causes insertion of at least one codon downstream of a codon coding for an amino acid of position between position 1200 and 1230.

64. A process according to claim 40 wherein the mutagenization treatment of step (ii) is site-directed mutagenesis comprising constructing mutation specific primer sequences capable of hybridizing to the DNA sequence isolated in step (i), amplifying said primer sequences by a polymerase chain reaction (PCR) procedure followed by the excision of a fragment of the isolated DNA sequence and replacing it with an amplified primer sequence.

65. A process according to claim 40 wherein the replicon into which the mutated DNA sequence is inserted in step (iii) is a plasmid.

66. A process according to claim 65 wherein the plasmid is one selected from pSPE680, pSPE888, pSPE900, pSPE1003 and pSPE1038.

67. A process according to claim 40 wherein the cell being transformed in step (iv) is a bacterium.

68. A process according to claim 67 wherein the bacterium is a gram-negative bacterium.

69. A process according to claim 68 wherein the bacterium is Escherichia coli .

70. A process according to claim 40 wherein the mutant recombinant protein is harvested by separating the cells and subjecting the separated cells to a sonication treatment and removing the cell debris to obtain a crude cell extract comprising the protein.

71. A process according to claim 70 which as a further step comprises a purification of the crude cell extract.

72. A process according to claim 71 wherein the purification comprises an anion-exchange chromatography including an elution step using an increasing gradient buffer containing 0 to 1000 mM NaCl.

73. A process according to claim 72 wherein the purification as a further step comprises hydrophobic interaction chromatography including an elution step using an increasing gradient buffer containing 0 to 1000 mM NaCl.

74. A process according to claim 72 or 73 wherein the eluates obtained from the elution step has a content of the mutant recombinant protein which is at least 50 wt%, calculated on the total protein content of the starting material.

75. A process according to claim 74 wherein the content of mutant recombinant protein is at least 75 wt%.

76. A process according to claim 75 wherein the content of mutant recombinant protein is at least 90 wt%.

77. A process according to claim 76 wherein the content of mutant recombinant protein is at least 94 wt%.

78. A process according to claim 77 wherein the content of mutant recombinant protein is at least 98 wt%.

79. A process according to claim 71 wherein the yield of mutant recombinant protein is at least 1 wt%, calculated on the total protein content of the crude cell extract.

80. A process according to claim 79 wherein the yield of mutant recombinant protein is at least 2 wt%, calculated on the total protein content of the crude cell extract.

81. A process according to claim 80 wherein the yield of mutant recombinant protein is at least 3.5 wt%, calculated on the total protein content of the crude cell extract.

82. A DNA sequence comprising a gene coding for a protein having a molecular weight of at least 140 kD which is reactive with an antibody reacting with the Pasteurella mul tocida spp multocida 45/78 toxin (PMT) encoded by the toxA gene, said DNA sequence being derived from a replicon comprising the toxA gene, by substitution, deletion or insertion of one or more codons of said toxA gene.

83. A DNA sequence according to claim 82 which is derived from the toxA gene-comprising replicon by substitution, deletion or insertion of one or more codons downstream of a unique EcoRI site in said gene.

84. A DNA sequence according to claim 82 which is derived from the toxA gene-comprising replicon by substituting at least one codon coding for an amino acid or amino acids between amino acid position 1135 and 1285 of said toxA gene with a codon/codons coding for a different/different amino acids.

85. A DNA sequence according to claim 84 which is derived from the ϋoxA gene-comprising replicon by substituting at least two codons such as at least three codons.

86. A DNA sequence according to claim 85 which is derived from the toxA gene-comprising replicon by substituting at least four codons.

87. A DNA sequence according to claim 84 which is derived from the toxA gene-comprising replicon by substituting at least one codon coding for an amino acid selected from serine, histidine, glutamine and threonine.

88. A DNA sequence according to claim 84 which is derived from the toxA gene- comprising replicon by substituting at least one codon coding for an amino acid or amino acids between amino acid position 1175 and 1285 of said toxA gene.

89. A DNA sequence according to claim 84 which is derived from the toxA gene-comprising replicon by substituting at least one codon coding for an amino acid or amino acids between amino acid position 1200 and 1285 of said toxA gene.

90. A DNA sequence according to claim 82 which is derived from the toxA gene-comprising replicon by deleting at least one codon coding for an amino acid or amino acids of position(s) between position 1131 and 1285.

91. A DNA sequence according to claim 90 which is derived from the toxA gene-comprising replicon by deleting at least two codons such as at least three codons.

92. A DNA sequence according to claim 91 which is derived from the toxA gene-comprising replicon by deleting at least four codons.

93. A DNA sequence according to claim 92 which is derived form the toxA gene-comprising replicon by deleting at least seven codons.

94. A DNA sequence according to claim 90 which is derived from the toxA gene-comprising replicon by deleting at least one codon coding for an amino acid selected from lysine, glutamic acid, phenylalanine, alanine, valine and aspartic acid.

95. A DNA sequence according to claim 90 which is derived from the toxA gene-comprising replicon by deleting at least one codon coding for an amino acid or amino acids of position(s) between position 1175 and 1285.

96. A DNA sequence according to claim 95 which is derived from the toxA gene-comprising replicon by deleting at least one codon coding for an amino acid or amino acids of position(s) between position 1215 and 1285.

97. A DNA sequence according to claim 82 which is derived from the toxA gene-comprising replicon by inserting at least one codon downstream of a codon coding for an amino acid or amino acids at amino acid position(s) between position 1131 and 1285.

98. A DNA sequence according to claim 97 which is derived from the toxA gene-comprising replicon by inserting at least two codons such as at least three codons.

99. A DNA sequence according to claim 98 which is derived from the toxA gene-comprising replicon by inserting at least four codons.

100. A DNA sequence according to claim 97 in which at least one codon being inserted is a codon coding for glycine.

101. A DNA sequence according to claim 97 which is derived from the toxA gene-comprising replicon by inserting at least one codon downstream of a codon coding for an amino acid or amino acids at amino acid position(s) between position 1175 and 1285.

102. A DNA sequence according to claim 101 which is derived from the toxA gene-comprising replicon by inserting at least one codon downstream of a codon coding for an amino acid or amino acids at amino acid position(s) between position 1200 and 1230.

103. A replicon harbouring the DNA sequence of claim 82.

104. A replicon according to claim 103 which is a plasmid.

105. A replicon according to claim 104 which is a plasmid selected from pSPE680, pSPE888, pSPE900, pSPE1003, pSPE1038, pSPE1020, pSPE1134 and pSPE1234.

106. A replicon according to claim 105 which is plasmid pSPE 1038 encoding the mutant recombinant protein 4HX, as

deposited with the DSM-Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH under the accession number DSM 7221.

107. A cell which is transformed with the replicon of claim 103 in which cell said replicon is replicated.

108. A cell according to claim 107 which is a bacterium.

109. A cell according to claim 108 which is a gram-negative bacterium.

110. A cell according to claim 109 which is Escherichia coli .

111. Use of the mutant recombinant protein of claim 1 as a vaccine for the protection against diseases caused by

Pasteurella multocida.