AU2001269305B2 - Expression system - Google Patents

Description

Expression System

The present invention relates to polypeptides which produce an immune response which is protective against infection by Bacillus anthracis, to methods of producing these, to recombinant Escherischia coli cells, useful in the methods, and to nucleic acids and transformation vectors used.

Present systems for expressing PA for vaccine systems use protease deficient Bacillus subtilis as the expression host. Although such systems are acceptable in terms of product quantity and purity, there are significant drawbacks. Firstly, regulatory authorities are generally unfamiliar with this host, and licensing decisions may be delayed as a result. More importantly, the currently used strains of Bacillus subtilis produce thermostable spores which require the use of a dedicated production plant.

WO00/02522 describes in particular VEE virus replicons which express PA or certain immunogenic fragments.

E. coli is well known as an expression system for a range of human vaccines. While the ability to readily ferment E. coli to very high cellular densities makes this bacterium an ideal host for the expression of many proteins, previous attempts to express and purify recombinant PA from E. coli cytosol have been hindered by low protein yields and proteolytic degradation (Singh et al . , J. Biol. Chem. (1989) 264; 11099-11102, Vodkin et al., Cell (1993) 34; 693-697 and Sharma et al., Protein Expr. purif. (1996), 7, 33-38).

A strategy for overexpressing PA as a stable, soluble protein in the E. coli cytosol has been described recently (Willhite et al., Protein and Peptide Letters, (1998), 5; 273-278). The strategy adopted is one of adding an affinity tag sequence to the N terminus of PA, which allows a simple purification system. A problem with this system is that it requires a further downstream processing step in order to remove the tag before the PA can be used.

Codon optimisation is a technique which is now well known and used in the design of synthetic genes. There is a degree of redundancy in the genetic code, in so far as most amino acids are coded for by more than one codon sequence. Different organisms utilise one or other of these different codons preferentially. By optimising codons, it is generally expected that expression levels of the particular protein will be enhanced.

This is generally desirable, except where, as in the case of PA, higher expression levels will result in proteolytic degradation and/or cell toxicity. In such cases, elevating expression levels might be counter-productive and result in significant cell toxicity.

Surprisingly however, the applicants have found that this is not the case in E. coli and that in this system, codon optimisation results in expression of unexpectedly high levels of recombinant PA, irrespective of the presence or absence of proteolytic enzymes within the strain.

Furthermore, it would appear that expression of a protective domain of PA does not inhibit expression in E. coli .

The crystal structure of native PA has been elucidated (Petosa C, et al. Nature 385: 833-838,1997) and shows that PA consists of four distinct and functionally independent domains: domain 1, divided into la, 1-167 amino acids and lb, 168-258 amino acids; domain 2, 259-487 amino acids; domain 3, 488-595 amino acids and domain 4, 596-735 amino acids. The applicants have identified that certain domains appear to produce surprisingly good protective effects when used in isolation, in fusion proteins or in combination with each other.

According to the present invention there is provided an immunogenic reagent which produces an immune response which is protective against Bacillus anthracis, said reagent comprising one or more polypeptides which together represent up to three domains of the full length Protective Antigen (PA) of B. anthracis or variants of these, and at least one of said domains comprises domain 1 or domain 4 of PA or a variant thereof.

Specifically, the reagent will comprise mixtures of polypeptides or fusion peptides wherein individual polypeptides comprise one of more individual domains of PA.

In particular, the reagent comprises polypeptide (s) comprising domain 1 or domain 4 of PA or a variant thereof, in a form other than full length PA. Where present, domains are suitably complete, in particular domain 1 is present in its entirety.

The term "polypeptide" used herein includes proteins and peptides .

As used herein, the expression "variant" refers to sequences of amino acids which differ from the basic sequence in that one or more amino acids within the sequence are deleted or substituted for other amino acids, but which still produce an immune response which is protective against Bacillus anthracis . Amino acid substitutions may be regarded as "conservative" where an amino acid is replaced with a different amino acid with broadly similar properties. Non-conservative substitutions are where amino acids are replaced with amino acids of a different type. Broadly speaking, fewer non-conservative substitutions will be possible without altering the biological activity of the polypeptide. Suitably variants will be at least 60% identical, preferably at least 75% identical, and more preferably at least 90% identical to the PA sequence.

In particular, the identity of a particular variant sequence to the PA sequence may be assessed using the multiple alignment method described by Lipman and Pearson, (Lipman, D.J. & Pearson, W.R. (1985) Rapid and Sensitive Protein Similarity Searches, Science, vol 227, ppl435-1441) . The "optimised" percentage score should be calculated with the following parameters for the Lipman-Pearson algorithm: ktup =1, gap penalty =4 and gap penalty length =12. The sequences for which similarity is to be assessed should be used as the "test sequence" which means that the base sequence for the comparison, (SEQ ID NO 1) , should be entered first into the algorithm.

Preferably, the reagentof the invention includes a polypeptide which has the sequence of domain 1 and/or domain 4 of wild-type PA.

A particularly preferred embodiment of the invention comprises domain 4 of the PA of B. anthracis .

These domains comprise the following sequences shown in the following Table 1. Table 1

Domain Amino acids of full-length PA*

4 596-735

1 1-258

These amino acids numbers refer to the sequence as shown in Welkos et al. Gene 69 (1988) 287-300 and are illustrated hereinafter as SEQ ID NOs 15 (Fig 4) and 3 (Fig 3) respectively.

Domain 1 comprises two regions, designated la and lb. Region la comprises amino acids 1-167 whereas region lb is from amino acid 168-258. It appears that region la is important for the production of a good protective response, and the full domain may be preferred.

In a particularly preferred embodiment, a combination of domains 1 and 4 or protective regions thereof, are used as the immunogenic reagent which gives rise to an immune response protective against B. anthracis. This combination, for example as a fusion peptide, may be expressed using the expression system of the invention as outlined hereinafter.

When domain 1 is employed, it is suitably fused to domain 2 of the PA sequence, and may preferably be fused to domain 2 and domain 3.

Such combinations and their use in prophylaxis or therapy forms a further aspect of the invention.

Suitably the domains described above are part of a fusion protein, preferably with an N-terminal glutathione-s-transferase protein (GST) . The GST not only assists in the purification of the protein, it may also provide an adjuvant effect, possibly as a result of increasing the size.

The polypeptides of the invention are suitably prepared by conventional methods. For example, they may be synthesised or they may be prepared using recombinant DNA technology. In particular, nucleic acids which encode said domains are included in an expression vector, which is used to transform a host cell.

Culture of the host cell followed by isolation of the desired polypeptide can then be carried out using conventional methods.

Nucleic acids, vectors and transformed cells used in these methods form a further aspect of the invention.

Generally speaking, the host cells used will be those that are conventionally used in the preparation of PA, such as Bacillus subtilis . The applicants have found surprisingly that the domains either in isolation or in combination, maybe successfully expressed in E. coli under certain conditions .

Thus, the present invention further provides a method for producing an immunogenic polypeptide which' produces an immune response which is protective against B. anthracis , said method comprising transforming an E. coli host with a nucleic acid which encodes either (a) the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, or (b) a polypeptide comprising at least one protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response as described above, culturing the transformed host and recovering the polypeptide therefrom, provided that where the polypeptide is the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, the percentage of guanidine and cytosine residues within the said nucleic acid is in excess of 35%.

Using these options, high yields of product can be obtained using a favoured expression host.

A table showing codons and the frequency with which they appear in the genomes of Escherichia coli and Bacillus anthracis is shown in Figure 1. It is clear that guanidine and cytosine appear much more frequently in E. coli than B. anthracis . Analysis of the codon usage content reveals the following:

Thus it would appear that codons which are favoured by E. coli are those which include guanidine or cytosine where possible. By increasing the percentage of guanidine and cytosine nucleotides in the sequence used to encode the immunogenic protein over that normally found in the wild-type B. anthracis gene, the codon usage will be such that expression in E. coli is improved.

Suitably the percentage of guanidine and cytosine residues within the coding nucleic acid used in the invention, at least where the polypeptide is the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, is in excess of 40%, preferably in excess of 45% and most preferably from 50-52%.

High levels of expression of protective domains can be achieved, with using the wild-type B. anthracis sequence encoding these units. However, the yields may be improved further by increasing the GC% of the nucleic acid as described above.

In a particular embodiment, the method involves the expression of PA of B . anthracis .

Further according to the present invention, there is provided a recombinant Escherischia coli cell which has been transformed with a nucleic acid which encodes the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, and wherein the percentage of guanidine and cytosine residues within the nucleic acid is in excess of 35%.

As before, suitably the percentage of guanidine and cytosine residues within the coding nucleic acid is in excess of 40%, preferably in excess of 45% and most preferably from 50-52%.

Suitably, the nucleic acid used to transform the E. coli cells of the invention is a synthetic gene. In particular, the nucleic acid is of SEQ ID NO 1 as shown in Figure 2 or a modified form thereof.

The expression "modified form" refers to other nucleic acid sequences which encode PA or fragments or variants thereof which produce a protective immune response but which utilise some different codons, provided the requirement for the percentage GC content in accordance with the invention is met. Suitable modified forms will be at least 80% similar, preferably 90% similar and most preferably at least 95% similar to SEQ ID NO 1. In particular, the nucleic acid comprises SEQ ID NO 1.

In an alternative embodiment, the invention provides a recombinant Escherischia coli cell which has been transformed with a nucleic acid which encodes a protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response.

Preferably, the nucleic acid encodes domain 1 or domain 4 of B. anthracis.

Further according to the invention there is provided a method of producing immunogenic polypeptide which produces an immune response which is protective against B. anthracis, said method comprising culturing a cell as described above and recovering the desired polypeptide from the culture. Such methods are well known in the ar .

In yet a further aspect, the invention provides an E. coli transformation vector comprising a nucleic acid which encodes the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, and wherein the percentage of guanidine and cytosine residues within the nucleic acid is in excess of 35%. A still further aspect of the invention comprises an E. coli transformation vector comprising a nucleic acid which encodes a protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response.

Suitable vectors for use in the transformation of E. coli are well known in the art. For example, the T7 expression system provides good expression levels. However a particularly preferred vector comprises pAG163 obtainable from Avecia (UK) .

A nucleic acid of SEQ ID NO 1 or a variant thereof which encodes PA and which has at 35%, preferably at least 40%, more preferably at least 45% and most preferably from 50-52% GC content form a further aspect of the invention.

If desired, PA of the variants, or domains can be expressed as a fusion to another protein, for example a protein which provides a different immunity, a protein which will assist in purification of the product or a highly expressed protein (e.g. thioredoxin, GST) to ensure good initiation of translation.

Optionally, additional systems will be added such as T7 lysozyme to the expression system, to improve the repression of the system, although, in the case of the invention, the problems associated with cell toxicity have not been noted.

Any suitable E. coli strain can be emplqyed in the process of the invention. Strains which are deficient in a number of proteases (e.g. Ion^", ompT^") are available, which would be expected to minimise proteolysis. However, the applicants have found that there is no need to use such strains to achieve good yields of product and that other known strains such as K12 produce surprisingly high product yields. Fermentation of the strain is generally carried out under conventional conditions as would be understood in the art. For example, fermentations can be carried out as batch cultures, preferably in large shake flasks, using a complex medium containing antibiotics for plasmid maintenance and with addition of IPTG for induction.

Suitably cultures are harvested and cells stored at -20°C until required for purification.

Suitable purification schemes for E. coli PA (or variant or domain) expression can be adapted from those used in B. subtilis expression. The individual purification steps to be used will depend on the physical characteristics of recombinant PA. Typically an ion exchange chromatography separation is carried out under conditions which allow greatest differential binding to the column followed by collection of fractions from a shallow gradient. In some cases, a single chromatographic step may be sufficient to obtain product of the desired specification.

Fractions can be analysed for the presence of the product using SDS PAGE or Western blotting as required.

As illustrated hereinafter, the successful cloning and expression of a panel of fusion proteins representing intact or partial domains of rPA has been achieved. The immunogenicity and protective efficacy of these fusion proteins against STI spore challenge has been assessed in the A/J mouse model.

All the rPA domain proteins were immunogenic in A/J mice and conferred at least partial protection against challenge compared to the GST control immunised mice. The carrier protein, GST attached to the N-terminus of the domain proteins, did not impair the immunogenicity of the fusion proteins either in vivo, shown by the antibody response stimulated in immunised animals, or in vitro as the fusion proteins could be detected with anti- rPA antisera after Western blotting, indicating that the GST tag did not interfere with rPA epitope recognition. Immunisation with the larger fusion proteins produced the highest titres. In particular, mice immunised with the full length GST 1-4 fusion 5 protein produced a mean serum anti-rPA concentration approximately eight times that of the rPA immunised group (Figure 5) . Immunisation of mice with rPA domains 1-4 with the GST cleaved off, produced titres of approximately one half those produced by immunisation with the fusion protein. Why this

10 fusion protein should be much more immunogenic is unclear. It is possible that the increased size of this protein may have an adjuvantising effect on the immune effector cells. It did not stimulate this response to the same extent in the other fusion proteins and any adjuvantising effect of the GST tag did not

15 enhance protection against challenge as the cleaved proteins were similarly protective to their fusion protein counterparts.

Despite having good anti-rPA titres, some breakthrough in protection at the lower challenge level of 10²MLD's, occurred in

20 the groups immunised with GSTl, cleaved 1, GSTlb-2, GSTlb-3 and GST1-3 and immunisation with these proteins did not prolong the survival time of those mice that did succumb to challenge, compared with the GST control immunised mice. This suggests that the immune response had not been appropriately primed by

25. these proteins to achieve full resistance to the infection. As has been shown in other studies in mice and guinea pigs (Little S.F. et al. 1986. Infect. Immun. 52: 509-512, Turnbull P.C.B., et al., 1986. Infect. Immun. 52: 356-363) there is no precise correlation between antibody titre to PA and protection against

30 challenge. However a certain threshold of antibody is required for protection (Cohen S et. al., 2000 Infect. Immun. 68: 4549- 4558), suggesting that cell mediated components of the immune response are also required to be stimulated for protection (Williamson 1989) .

35 GSTl, GSTlb-2 and GST1-2 were the least stable fusion proteins produced, as shown by SDS-Page and Western blotting results, possibly due to the proteins being more susceptible to degradation in the absence of domain 3, and this instability may have resulted in the loss of protective epitopes.

The structural conformation of the proteins may also be important for stimulating a protective immune response. The removal of Domain la from the fusion proteins gave both reduced antibody titres and less protection against challenge, when compared to their intact counterparts GST1-2 and GST1-3. Similarly, mice immunised with GST 1 alone were partially protected against challenge, but when combined with domain 2, as the GST1-2 fusion protein, full protection was seen at the 10² MLD challenge level. However the immune response stimulated by immunisation with the GST1-2 fusion protein was insufficient to provide full protection against the higher 10³ MLD' s challenge level, which again could be due to the loss of protective epitopes due to degradation of the protein.

All groups immunised with truncates containing domain 4, including GST 4 alone, cleaved 4 alone and a mixture of two individually expressed domains, GST 1 and GST 4 were fully protected against challenge with 10³ MLDs of STI spores (Table 1) . Brossier et al showed a decrease in protection in mice immunised with a mutated strain of B. anthracis that expressed PA without domain 4 (Brossier F., et al. 2000. Infect. Immun. 68: 1781-1786) and this was confirmed in this study, where immunisation with GST 1-3 resulted in breakthrough in protection despite good antibody titres. These data indicate that domain 4 is the immunodominant sub-unit of PA. Domain 4 represents the 139 amino acids of the carboxy terminus of the PA polypeptide. It contains the host cell receptor binding region (Little S.F. et al., 1996 Microbiology 142: 707-715), identified as being in and near a small loop located between amino acid residues 679- 693 (Varughese M., et al . 1999 Infect. Immun. 67:1860-1865). Therefore it is essential for host cell intoxication as it has been demonstrated that forms of PA expressed containing mutations (Varughese 1999 supra.) or deletions (Brossier 1999 supra.) in the region of domain 4 are non-toxic. The crystal structure of PA shows domain 4, and in particular a 19 amino acid loop of the domain (703-722), to be more exposed than the other three domains which are closely associated with each other (Petosa 1997 supra.). This structural arrangement may make domain 4 the most prominent epitope for recognition by immune effector cells, and therefore fusion proteins containing domain 4 would elicit the most protective immune response.

This investigation has further elucidated the role of PA in the stimulation of a protective immune response demonstrating that protection against anthrax infection can be attributed to individual domains of PA.

The invention will now be particularly described by way of example, with reference to the accompanying drawings in which:

Figure 1 is a Table of codon frequencies found within E. coli and B. anthracis;

Figure 2 shows the sequence of a nucleic acid according to the invention, which encodes PA of B. subtilis, as published by Welkos et al supra; and

Figure 3 shows SEQ ID NOs 3-14, which are amino acid and DNA sequences used to encode various domains or combinations of domains of PA as detailed hereinafter;

Figure 4 shows SEQ ID NOs 15-16 which are the amino acid and DNA sequences of domain 4 of PA respectively; and

Figure 5 is a table showing anti-rPA IgG concencentration, 37 days post primary immunisation, from A/J mice immunised intramuscularly on days 1 and 28 with lOμg of fusion protein included PA fragment; results shown are mean + sem of samples taken from 5 mice per treatment group.

Example 1

Investigation into expression in E.coli rPA expression plasmid pAG163::rPA has been modified to substitute Km^R marker for original Tc^R jene. This plasmid has been transformed into expression host E. coli BLR (DE3) and expression level and solubility assessed. This strain is deficient in the intracellular protease La (Ion gene product) and the outer membrane protease OmpT.

Expression studies did not however show any improvement in the accumulation of soluble protein in this strain compared to Ion+ K12 host strains (i.e. accumulation is prevented due to excessive proteolysis) . It was concluded that any intracellular proteolysis of rPA was not due to the action of La protease.

Example 2

Fermentation analysis

Further analysis of the fermentation that was done using the K12 strain UT5600 (DE3) pAG163::rPA.

It was found that the rPA in this culture was divided between the soluble and insoluble fractions (estimated 350mg/L insoluble, 650mg/L full length soluble) . The conditions used (37°C, ImM IPTG for induction) had not yielded any detectable soluble rPA in shake flask cultures and given the results described in Example 1 above, the presence of a large amount of soluble rPA is surprising. Nevertheless it appears that manipulation of the fermentation, induction and point of harvest may allow stable accumulation of rPA in E. coli K12 expression strains . Example 3

A sample of rPA was produced from material initially isolated as insoluble inclusion bodies from the UT5600 (DE3) pAG163::rPA fermentation. Inclusion bodies were washed twice with 25mM Tris-HCl pH8 and once with same buffer +2M urea. They were then solubilized in buffer +8M urea and debris pelleted. Urea was removed by dilution into 25mM Tris-HCl pH8 and static incubation overnight at 4°C. Diluted sample was applied to Q sepharose column and protein eluted with NaCl gradient. Fractions containing highest purity rPA were pooled, aliquoted and frozen at -70°C. Testing of this sample using 4-12% MES-SDS NuPAGE gel against a known standard indicated that it is high purity and low in endotoxin contamination.

Example 4

Further Characterisation of the Product

N terminal sequencing of the product showed that the N-terminal sequence consisted of

MEVKQENRLL (SEQ ID NO 2)

This confirmed that the product was as expected with initiator methionine left on.

The material was found to react in Western blot; MALDI -MS on the sample indicated a mass of approx 82 700 (compared to expected mass of 82 915) . Given the high molecular mass and distance from mass standard used (66KDa), this is considered an indication that material does not have significant truncation but does not rule out microheterogeneity within the sample.

Example 5

Testing of Individual domains of PA

Individual domains of PA were produced as recombinant proteins in E. coli as fusion proteins with the carrier protein glutathione-s-transferase (GST) , using the Pharmacia pGEX-6P-3 expression system. The sequences of the various domains and the DNA sequence used to encode them are attached herewith as Figure 3. The respective amino acid and DNA sequences are provided in Table 2 below.

These fusion proteins were used to immunise A/J mice (Harlan Olac) intra-muscularly with lOμg of the respective fusion protein adsorbed to 20% v/v alhydrogel in a total volume of lOOμl.

Animals were immunised on two occasions and their development of protective immunity was determined by challenge with spores of B. anthracis (STI strain) at the indicated dose levels. The table below shows survivors at 14 days post-challenge.

The data shows that a combination of all 4 domains of PA, whether presented as a fusion protein with GST or not, were protective up to a high challenge level. Removal of domain 4, leaving 1+2+3, resulted in breakthrough at the highest challenge level tested, 9xl0⁵. Domains 1+2 were as protective as a combination of domains 1+2+3 at 9xl0⁴ spores. However, removal of domain la to leave a GST fusion with domains lb+2, resulted in breakthrough in protection at the highest challenge level tested (9xl0⁴⁾ which was only slightly improved by adding domain 3.

The data indicates that the protective immunity induced by PA can be attributed to individual domains (intact domain 1 and domain 4) or to combinations of domains taken as permutations from all 4 domains .

The amino acid sequence and a DNA coding sequence for domain 4 is shown in Figure 4 as SEQ ID NOs 15 and 16 respectively.

Example 6

Further Testing of domains as vaccines

DNA encoding the PA domains, amino acids 1-259, 168-488, 1-488, 168-596,1-596, 260-735, 489-735, 597-735 and 1-735 (truncates

GSTl, GSTlb-2, GSTl-2, GSTlb-3, GSTl-3, GST2-4, GST3-4, GST4 and GST1-4 respectively) were PCR amplified from B. anthracis Sterne DNA and cloned in to the XhoI/BamRi sites of the expression vector pGEX-6-P3 (Amersham-Pharmacia) downstream and in frame of the lac promoter. Proteins produced using this system were expressed as fusion proteins with an N-terminal glutathione-s- transferase protein (GST) . Recombinant plasmid DNA harbouring the DNA encoding the PA domains was then transformed in to E. coli BL21 for protein expression studies.

E. coli BL21 harbouring recombinant pGEX-6-P3 plasmids were cultured in L-broth containing 50μg/ml ampicillin, 30μg/ml chloramphenicol and 1% w/v glucose. Cultures were incubated with shaking (170 rev min^-1) at 30°C to an A6oonm 0.4, prior to induction with 0.5mM IPTG. Cultures were incubated for a further 4 hours, followed by harvesting by centrifugation at 10 000 rpm for 15 minutes.

Initial extraction of the PA truncates-fusion proteins indicated that they were produced as inclusion bodies. Cell pellets were resuspended in phosphate buffered saline (PBS) and sonicated 4x20 seconds in an iced water bath. The suspension was centrifuged at 15 000 rpm for 15 minutes and cell pellets were then urea extracted, by suspension in 8M urea with stirring at room temperature for 1 hour. The suspension was centrifuged for 15 minutes at 15000 rpm and the supernatant dialysed against lOOmM Tris pH 8 containing 400mM L-arginine and O.lmM EDTA, prior to dialysis into PBS.

The successful refolding of the PA truncate-fusion proteins allowed them to be purified on a glutathione Sepharose CL-4B affinity column. All extracts (with the exception of truncate GSTlb-2, amino acid residues 168-487) were applied to a 15 ml glutathione Sepharose CL-4B column (Amersham-Parmacia) , previously equilibrated with PBS and incubated, with rolling, overnight at 4°C. The column was washed with PBS and the fusion protein eluted with 50mM Tris pH7, containing 150mM NaCl, ImM EDTA and 20mM reduced glutathione. Fractions containing the PA truncates, identified by SDS-PAGE analysis, were pooled and dialysed against PBS. Protein concentration was determined using BCA (Perbio) .

However truncate GSTlb-2 could not be eluted from the glutathione sepharose CL-4B affinity column using reduced glutathione and was therefore purified using ion exchange chromatography. Specifically, truncate GSTlb-2 was dialysed against 20mM Tris pH8, prior to loading onto a HiTrap Q column (Amersham-Parmacia), equilibrated with the same buffer. Fusion protein was eluted with an increasing NaCl gradient of 0-1M in 20mM Tris pH8. Fractions containing the GST-protein were pooled, concentrated and loaded onto a HiLoad 26/60 Superdex 200 gel filtration column (Amersham-Parmacia) , previously equilibrated with PBS. Fractions containing fusion protein were pooled and the protein concentration determined by BCA (Perbio) . Yields were between 1 and 43mg per litre of culture. The molecular weight of the fragments and their recognition by antibodies to PA was confirmed using SDS PAGE and Western Blotting. Analysis of the rPA truncates by SDS Page and Western blotting showed protein bands of the expected sizes. Some degradation in all of the rPA truncates investigated was apparent showing similarity with recombinant PA expressed in B. subtilis . The rPA truncates GSTl, GSTlb-2 and GSTl-2 were particularly susceptible to degradation in the absence of domain 3. This has similarly been reported for rPA constructs containing mutations in domain 3, that could not be purified from B. anthracis culture supernatants (Brossier 1999) , indicating that domain 3 may stabilise domains 1 and 2.

Female, specific pathogen free A/J mice (Harlan UK) were used in this study as these are a consistent model for anthrax infection (Welkos 1986) . Mice were age matched and seven weeks of age at the start of the study.

A/J mice were immunised on days 1 and 28 of the study with lOμg of fusion protein adsorbed to 20% of 1.3% v/v Alhydrogel (HCI Biosector, Denmark) in a total volume of lOOμl of PBS. Groups immunised with rPA from B. subtilis (Miller 1998), with recombinant GST control protein, or fusion proteins encoding domains 1, 4 and 1-4 which had the GST tag removed, were also included. Immunising doses were administered intramuscularly into two sites on the hind legs. Mice were blood sampled 37 days post primary immunisation for serum antibody analysis by enzyme linked immunosorbant assay (ELISA) .

Microtitre plates (Immulon 2, Dynex Technologies) were coated, overnight at 4° C with 5μg/ml rPA, expressed from B. subtilis (Millerl998) , in PBS except for two rows per plate which were coated with 5μg/ml anti-mouse Fab (Sigma, Poole, Dorset) . Plates were washed with PBS containing 1% v/v Tween 20 (PBS-T) and blocked with 5% w/v skimmed milk powder in PBS (blotto) for 2 hours at 37° C. Serum, double-diluted in 1% blotto, was added to the rPA coated wells and was assayed in duplicate together with murine IgG standard (Sigma) added to the anti-fab coated wells and incubated overnight at 4° C. After washing, horseradish peroxidase conjugated goat anti-mouse IgG (Southern Biotechnology Associates Inc.), diluted 1 in 2000 in PBS, was added to all wells, and incubated for 1 hour at 37° C. Plates were washed again before addition of the substrate 2,2 -Azinobis (3-ethylbenzthiazoline-sulfonic acid) (1.09mM ABTS, Sigma). After 20 minutes incubation at room temperature, the absorbance of the wells at 414nm was measured (Titertek Multiscan, ICN

Flow) . Standard curves were calculated using Titersoft version 3.1c software. Titres were presented as μg IgG per ml serum and group means + standard error of the mean (sem) were calculated. The results are shown in Figure 5.

All the rPA truncates produced were immunogenic and stimulated mean serum anti-rPA IgG concentrations in the A/J mice ranging from 6μg per ml, for the GSTlb-2 truncate immunised group, to 1488μg per ml, in the GST 1-4 truncate immunised group (Figure 5) . The GST control immunised mice had no detectable antibodies to rPA.

Mice were challenged with B. anthracis STI spores on day 70 of the immunisation regimen. Sufficient STI spores for the challenge were removed from stock, washed in sterile distilled water and resuspended in PBS to a concentration of IxlO⁷ and lxlO⁶ spores per ml. Mice were challenged intraperitoneally with 0.1ml volumes containing IxlO⁶ and 1x10^s spores per mouse, respectively, and were monitored for 14 day post challenge to determine their protected status . Humane end-points were strictly observed so that any animal displaying a collection of clinical signs which together indicated it had a lethal infection, was culled. The numbers of immunised mice which survived 14 days post challenge are shown in Table 3. Table 3

Challenge Level MLDs

Domain survivors/no. challenged (%)

10 MLDs 10^J MLDs

GST 1 3/5 (60) 1/5 (20)

GST lb-2 1/5 (20) nd

GST 1-2 5/5 (100) 3/5 (60)

GST lb-3 3/5 (60) nd

GST 1-3 4/5 (80) nd

GST 1-4 nd 5/5 (100)

GST 2-4 nd 5/5 (100)

GST 3-4 nd 5/5 (100)

GST 4 5/5 (100) 5/5 (100)

GST 1+ GST nd 5/5 (100)

Cleaved 1 1/5 (20) 2/5

Cleaved 4 5/5 (100) 5/5

Cleaved 1-4 nd 5/5 PA nd 4/4 (100) cont ^■.rol 0/5 (0) 0/5 (0)

1 MLD = aprox. 1 x 10 ^³ STI spores nd = not done

The groups challenged with 10³ MLD' s of STI spores were all fully protected except for the GSTl, GSTl-2 and cleaved 1 immunised groups in which there was some breakthrough in protection, and the control group immunised with GST only, which all succumbed to infection with a mean time to death (MTTD) of

2.4 + 0.2 days. At the lower challenge level of 10² MLD' s the GSTl-2, GST4 and cleaved 4 - immunised groups were all fully protected, but there was some breakthrough in protection in the other groups. The mice that died in these groups had a MTTD of

4.5 + 0.2 days which was not significantly different from the GST control immunised group which all died with a MTTD of 4 + 0.4 days.

Claims (1)

1. Claims

   1. An immunogenic reagent which produces an immune response which is protective against Bacillus anthracis,- said reagent comprising one or more polypeptides which together represent up to three domains of the full length Protective Antigen (PA) of B. anthracis or variants of these, and at least one of said domains comprises domain 1 or domain 4 of PA or a variant thereof.

   2. An immunogenic reagent according to claim 1 which comprises the sequence of domain 1 and/or domain 4 of wild-type PA.

   3. An immunogenic reagent according to claim 1 or claim 2 which comprises domain 4 of the PA of B. anthracis.

   4. An immunogenic reagent according to any one of the preceding claims which comprises a combination of domains 1 and 4 or protective regions thereof.

   5. An immunogenic reagent according to claim 4 wherein said domains are present in the form of a fusion polypeptide.

   6. An immunogenic reagent according to claim 5 which comprises domain 1 fused to domain 2 of the PA sequence.

   7. An immunogenic reagent according to claim 6 which is fused to domain 3 of the PA sequence.

   8. An immunogenic reagent according to claim 4 which comprises a mixture of a polypeptides, one of which comprises domain 1 and one of which comprises domain 4 of the PA sequence.

   9. An immunogenic reagent according to any one of the preceding claims wherein a polypeptide is fused to a further polypeptide.

   10. An immunogenic reagent according to claim 9 wherein said further peptide is glutathione-S-transferase (GST) .

   11. A nucleic acid which encodes a polypeptide of an immunogenic reagent according to any one of the preceding claims .

   12. An expression vector comprising a nucleic acid according to claim 11.

   13. A cell transformed with a vector according to claim 12.

   14. A method for producing an immunogenic polypeptide which produces an immune response which is protective against B. anthracis, said method comprising transforming an E. coli host with a nucleic acid which encodes either (a) the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, or (b) a protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, culturing the transformed host and recovering the polypeptide therefrom, provided that where the polypeptide is the protective antigen (PA) of Bacillus anthracis a variant thereof which can produce a protective immune response, the percentage of guanidine and cytosine residues within the said nucleic acid is in excess of 35%.

   15. A method according to claim 14 wherein the said nucleic acid encodes the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response.

   16. A method according to claim 15 wherein the percentage of guanidine and cytosine residues within the said nucleic acid is in excess of 45%.

   17. A method according to claim 16 wherein the percentage of guanidine and cytosine residues within the said nucleic acid is from 50-52%.

   18. A method according to claim 14 wherein the said nucleic acid encodes a protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response.

   19. A method according to claim 18 wherein the domain is domain 1 and/or domain 4 of PA of B. anthracis.

   20. A recombinant Escherischia coli cell which has been transformed with a nucleic acid which encodes the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, and wherein the percentage of guanidine and cytosine residues within the nucleic acid is in excess of 35%.

   21. A recombinant Escherischia coli cell according to claim 20 wherein the percentage of guanidine and cytosine residues within the said nucleic acid is in excess of 45%.

   22. A recombinant Escherischia coli cell according to claim 21 wherein the percentage of guanidine and cytosine residues within the said nucleic acid is from 50%-52%.

   23. A recombinant E. coli cell according to claim 20 wherein said nucleic acid is of SEQ ID NO 1 as shown in Figure 2 or a modified form thereof.

   24. A recombinant E. coli cell according to claim 23 wherein said nucleic acid is of SEQ ID NO 1.

   25. A recombinant Escherischia coli cell which has been transformed with a nucleic acid which encodes a protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response.

   26. A recombinant cell according to claim 25 wherein the nucleic acid encodes domain 1 or domain 4 of PA of B. anthracis.

   27. A method of producing a polypeptide which produces an immune response which is protective against B. anthracis, said method comprising culturing a cell according to any one of claims 20 to 26 and recovering the protective polypeptide from the culture.

   28. An E. coli transformation vector comprising a nucleic acid which encodes the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response, and wherein the percentage of guanidine and cytosine residues within the nucleic acid is in excess of 35%.

   29. An E. coli transformation vector comprising a nucleic acid which encodes a protective domain of the protective antigen (PA) of Bacillus anthracis or a variant thereof which can produce a protective immune response.

   30. A nucleic acid of SEQ ID NO 1 or a modified form thereof which encodes PA or a variant thereof which produces a protective immune response and which has at least 35% GC content.

   31. A nucleic acid according to claim 30 which is at least 90% identical to SEQ ID NO 1.

   32. A nucleic acid according to claim 31 which comprises SEQ ID NO 1.

   34. A method of preventing or treating infection by B. anthracis, said method comprising administering to a mammal in need thereof, a sufficient amount of an immunogenic reagent according to any one of claims 1 to 10.

   35. The use of an immunogenic reagent according to any one of claims 1 to 10 in the preparation of a medicament for the prophylaxis or treatment of B. anthracis infection.