WO1988008430A1

WO1988008430A1 - Peptide production by protein engineering

Info

Publication number: WO1988008430A1
Application number: PCT/AU1988/000120
Authority: WO
Inventors: John Stanley Mattick; Philip Anthony; Jennings
Original assignee: Commonwealth Scientific And Industrial Research Or
Priority date: 1987-04-27
Filing date: 1988-04-27
Publication date: 1988-11-03
Also published as: AU616662B2; AU1704988A

Abstract

The present invention relates to a method of peptide production by protein engineering. This method involves the culturing of genetically engineered bacteria which produce as extracellular structures type (4) fimbriae, the peptide being produced in association with these fimbriae. This method is applicable to the production of virtually any peptide or peptides. The engineered bacteria contains the gene encoding the fimbrial subunit of type (4) fimbriae to which has been added at the C-terminal end the nucleic acid sequence encoding the peptide of choice, together with a system for morphogenetic assembly of type (4) fimbriae such that fimbriae in association with the peptide are produced as extracellular structures.

Description

Peptide Production By Protein Engineering

The present invention relates to a method of peptide production by protein engineering. This method involves the culturing of genetically engineered bacteria which produce as extracellular structures type 4 fimbriae, the peptide being produced in association with these fimbriae. This method is applicable to the production of virtually any peptide or peptides. Background of the Invention There is a growing demand for biologically-active peptides, either as homologues or analogues of naturally occurring regulatory, receptor and antigenic sequences. Apart from natural sources, such peptides have to date bee produced by chemical or enzymatic synthesis, or by biosynthesis in recombinant hosts, usually as a fusion product in bacteria. The advent of protein engineering- techniques, which involve a combination of DNA primer-directed mutagenesis with recombinant host-vector cloning and expression systems, has expanded enormously th scope of, and potential for, the design, production and us of such peptides. These techniques combine precision and flexibility, not only to investigate structure-function relationships in existing peptides and proteins, but also for the expression and evaluation of a large repertoire of novel peptide sequences, which is not limited to that already present in nature. However, there are very few systems for the high level expression and particularly export of such peptides from recombinant hosts. What is required is a carrier protein system, preferably from a microbial host (for ease of genetic manipulation and industrial fermentation) , which has the flexibility to accept grafts of exogenous peptide sequences, either as substitutions or additions, and which may be readily recovered and processed (if necessary) to release the active peptide. One potential candidate for such a system is the type 4 fimbria (or common pilus) which occurs in a variety of Gram-negative bacteria, including Bacteroides nodosus, Moraxella bovis, Neisseria gonorrhoeae and Pseudomonas aeruginosa. These fimbriae are long filamentous surface structures, of about 6nm in diameter and ranging up to several urn in length, which are thought to be involved in bacterial colonization of eukaryotic cell surfaces. These fimbriae have a polar or mainly polar location on the cell, and are associated with a phenomenon known as twitching motility. They are comprised primarily of a small structural protein subunit, of approximate molecular weight 16,000, which varies in size (145 to 160 amino acids) in different species and serotypes- The subunits from different species share several features in common, most notably a highly conserved and hydrophobic amino-terminal sequence, which begins with -the modified amino acid N-methylphenylalanine. The subunits are also all preceded by a similar short ( 6 to 7 amino acid) positively-charged leader sequence, which is removed prior to incorporation into the mature fimbrial strand. These regions are thought to contain important signals for the export, assembly and structure of the fimbrial strand. Variation between species and serotypes occurs in the more hydrophilic carboxy-terminal two-thirds of the protein, with most differences clustered in two or three hypervariable regions. A priori it would seem that these regions, due to their confor ational flexibility, are the regions most likely to permit substitution of exogenous peptide sequences. In addition, all subunits end in a series of charged and/or polar amino acid residues, suggesting the carboxy-terminus has a surface location, perhaps suitable for insertion of additional peptide sequences. We have investigated both of these possibilities. For these experiments we used the fimbrial subunit of B.nodosus as the carrier for peptide grafts, Escherichia coli as the host for genetic constructions and P.aeruginos as the host for morphogenetic expression of modified fimbrial subunits. We have determined the primary sequenc of the fimbrial subunit genes representative of all known serogroups existing in the B.nodosus population, allowing close definition of the conserved and variable regions of the protein, and precise modification (Figure 1). B. odosus is an anaerobe and the causative agent of ovine footrot. P.aeruginosa is a genetically well-characterized and easily cultured aerobe, suitable for use in industrial fermentation, and has a compatible type 4 fimbrial system. It has already been demonstrated that high levels of (B.nodosus-type) fimbriae may be obtained from cloned subunit genes under appropriate promoter control in recombinant P.aeruginosa cells, and that this material is suitable for use as a vaccine (Australian Patent Application No. 50154/85).

The peptide used in the protype studies was the 16 amino acid sequence LRGDLQV AQKVARTL, corresponding to residues 144-159 of the surface protein VPl of foot-and-mouth disease (FMD) virus (strain 01-BFS).

This particular sequence was chosen for several (related) reasons. Firstly, and importantly, this peptide includes a well-characterized epitope capable of eliciting neutralizing antibodies against the natural FMD virus, thereby providing a potential route for the development and production of a new vaccine against the disease. Secondly, this peptide appears to contain both this B-cell epitope and a helper T-cell determinant, as well as part of a likely binding site important for viral entry into cells. The latter is related to the sequence RGDL, which is similar to the sequence RGDS which forms part of the cell/platelet binding site identified on the extracellular matrix protein fibronectin. Thirdly, the peptide may possess a pseudo- utonomous structure, proposed to be alpha-helical, which has the advantage that its important structural and functional features may be maintained in a new environment, but which, on the other hand, may present certain difficulties as a graft in a carrier protein. This peptide also has a significant length, comparable to that of many other active peptides. In an exploration of the feasibility and (general) utility of the fimbrial expression system for peptide production, this sequence then has the dual advantage that it not only possesses intrinsic biological and immunological importance, but als represents a relatively demanding test case.

The present inventors have found that construction involving substitution within the hypervariable regions of the type 4 fimbrial subunit of a replacement peptide did not succeed, whilst the alternative construction involving a carboxy-terminal substitution was successful. __#

Disclosure of Invention

Accordingly, in a first aspect the present invention consists in a method^* of producing a peptide comprising culturing bacteria containing the gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae to which has been added at the C-terminal end the nucleic acid sequence encoding the peptide, and an endogenous compatible system for the morphogenetic assembl of_type 4 fimbrial and/or the geaes for the morphogenetic assembly of such fimbriae derived from a type 4 fimbriate species, such that mature type 4 fimbriae in association with the peptide are produced as extracellular structures, harvesting the whole or part of the fimbriae substantially free of the host cells, and optionally separating the peptide from the type 4 fimbriae.

In a second aspect the present invention consists in an antigenic preparation for raising antibodies directed against the surface protein VPl of foot-and-mouth disease virus (FMDV) comprising the VPl peptides either in isolation or in association with type 4 fimbriae, which ha been produced by bacterial host cells containing the gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae to which has been added at the

C-terminus the nucleic acid sequence encoding the protein

VPl, and an endogenous compatible system for the morphogenetic assembly of type 4 fimbriae and/or those genes for the morphogenetic assembly of such fimbriae derived from a type 4 fimbriate species and in which the

VPl peptide has optionally been cleaved from the type 4 fimbriae produced by said host cells.

In a preferred embodiment of the present invention th gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae is derived from a bacteria selected from the group comprising B.nodosus , P.aeruginosa

N.gonorrhoeae, N.meningjtidis , M.nonliquefaciens and

M.bovis. ' .

In a further preferred embodiment the host bacteria i of the genus Pseudomonas and in particular is the species

P.aeruginosa preferably the strain P.aeruginosa PAK/2Pfs

(ATCC No. 53308) .

In another preferred embodiment of the present invention the gene encoding the fimbrial structural subuni to which has been added at the C-terminal end the nucleic acid encoding the peptide is contained within a plasmid or phage.

A preferred embodiment of the present invention is hereinafter described by way of example and with reference to the Figures.

Brief Description of Drawings

Fig. 1 is a schematic representation of the B.nodusus fimbrial subunit showing the sites of substitution or addition of peptide sequences; Fig. 2 shows the engineered insertions and substitutions (DNA and protein sequences) together with the B.nodosus fimbrial subunit gene and protein sequences.

Fig. 3 shows the induction and expression of variant fimbrial subunit genes in E.coli as indicated by gel profiles.

Fig. 4 shows a Western Transfer Analysis of Variant fimbrial subunits using guniea pig anti sera to foot and mouth disease virus.

Fig. 5 shows various gel profiles of the expression in P.aeruginosa of the Variant Fimbrial Subunit containing the C-terminal addition of FMDV VPl peptide 144-159. Detailed Description of Invention Fimbrial subunit gene cloning and template preparation

The B.nodosus fimbrial subunit gene was derived from the serogroup A prototype strain VCS1001 (ATCC 25539), cloned initially on a 5.5. kilobase Hind III genomic DNA restriction fragment (Australian Patent Application No. 34979/84). The gene was then isolated as a 576 base-pair Dra 1 cartridge, beginning 30 nucleotides upstream of the initiation codon and ending 69 nucleotides beyond the termination codon. This cartridge also includes a Shine-Dalgarno sequence for ribosomal docking, and a probable rho-independent transcription termination signal. The relevant features and full sequence are shown in Figures 1 and 2. Sma 1 linker sequences (Amersham Corp.) were ligated onto the ends of the Dra 1 cartridge, to allow later portability of the sequence during subcloning operations required for mutagenesis and expression (see below) . The ligation products were then digested with Sma 1 and the modified cartridge subcloned into the corresponding site of the filamentous phage vector M.13 mp8. Single stranded DNA was prepared from the recombinant phage, as the template for subsequent mutagenesis. The construction was checked by direct DNA sequencing of this template using the dideoxynucleotide chain termination method, which also established that the orientation of the insert was such that the 5' end of the fimbrial subunit gene was located adjacent to the universal priming site of the vector. Oligonucleotide-directed mutagenesis

The introduction of the FMDV VPl 144-159 peptide sequence into the fimbrial subunit gene was carried out using oligonucleotide primers on the single-stranded recombinant M13 template. The oligonucleotides were synthesized using an Applied Biosystems DNA Synthesizer, and purified according to the manufacturer's instructions. The basic design of the primers was similar in all cases, consisting of a central core sequence encoding the FMDV peptide (see below) flanked on either side by 12 nucleotid stretches complementary to defined regions of the fimbrial gene template, which specified the exact position of oligonucleotide substitution or addition (see below; Figure 2). The -core sequence was TTA CGC GGT GAT TTA CAA GTT TTA GCT CAA AAA GTT GCT CGC ACT TTA,^' encoding the peptide LRGDLQVLAQKVARTL, homologous to VPl residues 144-159 of FMDV strain 01-BFS. The oligonucleotide sequence was designed to take into account the codon usage of natural fimbrial subunit genes. In one case (VARl), th first codon ( TA- leucine) was omitted, in order to allow better alignment of the substitute peptide sequence with that originally present in the fimbrial subunit, while still retaining the cysteine residue at position 63, required for the presumptive disulphide loop which spans the two major variant regions in the centre of the protein. (There is also a third more dispersed variant region in the carboxy-terminal third of the protein (see Figure 1) which was not modified). In another case (VAR4), two additional glycine codons (GGTGGT) were added to the 5' end of the core sequence to serve as a flexible "hinge" region for the terminal addition of the FMDV epitope. All told, four mutant genes were constructed and analyzed: VARl, epitope substitution for fimbrial variant region 1, from residues 64-78 inclusive, immediately downstream of cysteine 63; VAR2, epitope substitution for fimbrial variant region 2, from residues 87-102 inclusive, just upstream of cysteine 104; VAR3 , double epitope substitution at variant regions 1 and 2, as described above; and VAR4, epitope addition following the terminal asparagine residue at position 158, including the di-glycine "hinge", introduced to minimize any interference between the FMDV sequence and the fimbrial subunit. In determining the exact position of epitope substitutions in variant regions 1 and 2, two factors were taken into account, (i) optimization of the alignment between the replacement FMDV sequence and the original fimbrial sequence in the region while (ii) limiting the substitution to the segments known to be variant in different serotypes, i.e. to minimize encroachment on conserved regions of the protein. Full details are shown in Figure 2. For mutagenesis, oligonucleotide primers were annealed to the single-stranded recombinant M13-fimbrial gene template, followed by in vitro synthesis of the complementary strand using the Klenow fragment of DNA polymerase I. Strand closure was effected by T4 DNA ligase under standard conditions. Following transfeetion into

E.coli TGI cells and mismatch repair in vivo, mutant clones (plaques) were identified by DNA hybridization screening using ³²P-end-labelled ol^'igonucleotide as the probe. Positive clones were plaque-purified and their (altered) sequence confirmed by DNA sequencing of the purified recombinant phage DNA using the dideoxy chain termination method.

Subcloning and gene expression

Subcloning, of the fimbrial subunit gene for high level morphogenetic expression in P.aeruginosa was carried out by a strategy similar to that described previously (Australian Patent Application No. 50154/85). Double stranded (replicative form) DNA was prepared from cells infected with recombinant phage carrying the various mutan constructions. The modified fimbrial subunit genes containing FMDV epitope coding sequences were then excised by digestion with Sma 1 and subcloned into the Hpa 1 site of the expression vector plasmid pPL-lambda (Pharmacia-P. Biochemicals ) , using as host the temperature-sensitive lambda lysogen E.coli strain N4830 ( Clt_s 857), maintaine at the permissive temperature of 30°C. Recombinant clones were identified by DNA hybridization using the Sma fragment (containing the normal fimbrial subunit gene), P-labelled by nick-translation, as the probe. The entire P_L promoter-modified fimbrial gene construction was then excised on a Bam Hi cartridge and subcloned into the corresponding site of the broad host range plasmid vector pKT240, using E.coli N4830 as the host, as describe above. Positive clones were again identified by DNA ' hybridization screening, using ²P-labelled nick-translated Bam HI segment as probe. These recombinan plasmids are termed pK-VARl, pK-VAR2, pK-VAR3, and pK-VAR4, for each of the constructions, according to the terminolog introduced earlier. E. coli cells containing these plasmids were maintained on media containing 50 ug ampicillin per ml.

Gene expression was checked in the E.coli host cells by diluting an overnight culture 1:4 into fresh medium (Luria broth), growing at 30^CC for 30 minutes, and then shifting to the restrictive temperature of 42^αC for 2 hours. Samples of induced and uninduced cells were analyzed by sodium dodecyl sulphate ( SDS) -urea-gradient polyacrylamide gels and Western transfer immunoblotting, as described in Australian Patent Application No. 50154/85. Control samples of cells containing the equivalent construction with an unmodified fimbrial subunit gene in pKT240 (pJSM202; ATCC40203 - Australian Patent Application No. 50154/85) were also included.

The recombinant pKT240-P, promoter-variant subunit gene constructions were then transformed into the multifimbriate P.aeruginosa strain PAK/2Pfs (ATCC No. 53308) as described in Australian Patent Application No. 50154/85. Control transformations with pKT240 itself, as well as pJSM202, were also carried out. Transformation were selected and maintained on nutrient plates containing 750 ug carbenicillin per ml.

Fimbrial fractions were prepared by standard procedures: P.aeruginosa cells were resuspended in phosphate-buffered saline, subjected to mechanical blending, and following removal of the cells by centrifugation, fimbriae were recovered from the supernatant by either MgCl₂ (0.1M) precipitation or isoelectric precipitation at pH 4.5 with sodium acetate, as described in Australian Patent Application No. 50154/85. Fimbrial fractions were analysed in SDS-urea-gradient polyacrylamide gels and Western transfer, as above. Diagnostic antisera were raised in rabbits against either native B.nodosus fimbriae or denatured B.nodosus fimbriae (disrupted with 1.6% SDS, 1% 2 -mercaptoethanol at 100^CC for 3 minutes, prior to emulsification in Freund's incomplete adjuvant) .

RESULTS The general features of the B.nodosus fimbrial subunit and the strategy for peptide subtraction and insertion are outlined in Figure 1. The fimbrial subunit of B.nodosus exhibits particular regions of major sequence variation of which the most prominent are located at either end of the presumptive disulphide loop, which spans the centre one third of the protein. These represented logical candidates as sites for substitution with a replacement peptide such li ¬

as the FMDV sequence. However (see below) these constructions did not succeed. Since the size of such substitutions at these positions may be limited, and in tw cases proved non-viable, we examined the alternative strategy of a carboxy-terminal fusion of the FMDV sequence. The fusion protein thus generated, retains the complete primary sequence of the fimbrial subunit, which may well be important in morphogenetic expression, and als has several other potentially desirable features in the context of a general expression strategy: the peptide addition is presumably exposed as a tail on the carboxy-terminus of the protein where it should be both accessible for interaction with other molecules and/or the immune system and where there is no a_ priori limit on the size of the peptide. Additionally the peptide should be accessible to cleavage and thus separation from the carrie protein.

The details of the normal and mutant sequences, and the oligonucleotides used, are shown in Fig. 2. The mutated genes were placed under P,_ promoter control and their expression induced in E. coli (Fig. 3). All mutant genes showed good induction and relatively high levels of expression, directed by the P promoter in E. coli . However it is also clear that replacement of the normal sequence with the FMDV sequence had a significant effect o the electrophoretic mobility of the protein although the peptide chain lengths were unchanged by the substitution and the molecular weights were approximately the same. Substitutions at both positions 1 and 2 (Figure^' 1) increased the electrophoretic mobility (i.e. lower apparent molecular weight) with substitutions at both sites showing an additive effect (evidenced by an approximate doubling in the shift in electrophoretic mobility relative to substitutions at a single site). On the other hand .the C-terminal fusion of FMDV sequence showed, as might be expected, a higher molecular weight on gels.

The position of the bands for the B.nodosus fusion proteins was verified and their antigenicity tested by Western transfer analysis using antisera raised against native and denatured B.nodosus fimbriae (Fig. 3) and against foot-and-mouth disease virus (Fig. 4). All modified subunits were recognised by antisera raised against denatured fimbriae whereas, the substitution at position 2 was associated with a marked reduction ( 80%) in the binding of antibodies against native fimbriae. Only a minor effect was seen for the substitution at position 1, with no such effect being apparent for the carboxy-terminal addition. This implies that variable region 2 plays an important role in the antigenic profile of the native protein. It also suggests that the carboxy-terminal fusion may not significantly perturb the overall structure of the protein.

Weste ii transfer analysis of the variant fimbrial ^"constructs shows that all FMDV-peptide grafts are recognised by anti-FMDV antisera (neutralising antisera raised in guinea pigs in Fig. 4) with, however, significant quantitative differences between them. The carboxy-terminal- graft was recognised the most strongly, with VARl also showing a high level of reactivity. VAR2 was only relatively weakly recognised and somewhat surprisingly the tandem construct VAR3 was only weakly recognised as was the substitution at the fimbrial variant region 2 (VAR2), despite VAR 3 it containing the identical graft as the strongly recognised VARl in addition to the substitution at variant region 2. This data strongly indicate local stereochemical effects on antigenicity for this epitope, as is consistent with the synthetic peptide data of other workers and with for a body of data for other continuous epitopes. The modified fimbrial subunit gene - P promoter constructions in the broad host range vector pKT240 were then purified and transformed into P.aeruginosa. Transformants were obtained with the C-terminal construction but not with the internal substitutions, despite repeated attempts. A likely explanation for this is that the internal substitutions result in a lethal accumulation of the engineered subunit, presumably becaus of structural alteration of the protein preventing its normal assembly and export. In this context we note that the normal subunit, under P_L promoter control, while not lethal in P.aeruginosa (which correctly processes and exports the protein) is lethal in E.coli (which lacks this ability). The only mutant fimbrial protein which was expressed in .aeruginosa was the C-terminal addition, which was, however, associated with an impaired growth rat Fimbrial fractions were prepared from the viable subunit P.aeruginosa transformants . Fimbriae may normally be derived from the supernatant of cells, following mechanical blending, by either of the two independent methods of MgCl₂ precipitation or isoelectric precipitation (at pH 4.5) with sodium acetate. Figure 5 shows Coomassie stain and Western-transfer profiles of the cells and derived fimbrial fractions from P. eruginosa containing the pKT240/fimbrial protein VAR4. In both acetate and MgCl, fimbrial fractions there is clearly evident a protein of the expected size, which was confirme by antibody recognition in western transfer analysis. Thi is a similar profile exhibited by normal fimbriate P.aeruginosa cells expressing either the normal endogenous subunit and the cloned wild- type B.nodosus subunit (see Australian Patent Application No. 50154/85), whereas when non-fimb iate cells such as E. coli express this protein there is little or no signal in the supernatant fimbrial fractions (see Figure 5). Thus it appears that the C-terminal fusion of FMDV sequence to the B.nodosus subuni is being assembled and exported on fimbriae in the recombinant host.

DISCUSSION The results presented herein demonstrate the feasibility of engineering the fimbrial subunit (gene) of B.nodosus to act as a carrier protein for the production and export of exogenous peptide sequences. The FMDV 144-159 VPl peptide used here represents a formidable test case because of its length and higher order structure. The observation that this peptide, while covalently fused to the carboxy-terminus of the fimbrial subunit, is viably expressed and exported is indicative of the general utility of the system. While there may be particular problems with the expression of individual peptides (which may need to be assessed on a case-by-case basis), and there may be certain broad limitations on the size and nature of grafted peptides, there is a strong implication that the system is amenable to substantial manipulation, with clear potential for use with a variety of peptides. The successful construction described herein involved the C-terminal addition of the peptide sequence to the fimbrial subunit. Contrary to what might be expected, substitution at either one of the major variant regions within the protein failed to produce a viable outcome. This suggests that there are serious constraints on acceptable substitutions in these regions, presumably related to effects on the overall structure of the protein, with consequential interference in the export/assembly process. There are also other limitations on such substitutions as a vehicle for peptide production, related to both the size of the substitution and its environment. The flanking of the substitution by carrier sequences increases interference problems in terms of protein folding (in both directions), which may (at the least)- then require removal of the peptide sequence to obtain biological activity. The same considerations and problems^' apply to internal additions (i.e. rather than substitutions) of peptide sequences. With terminal additions, on the other hand, there are less likely to be size and structural constraints (as suggested by the present study).

Furthermore, risk of mutual interference is lower and may be further reduced by the addition of neutral linker (e.g. glycine) residues or a spacer arm to the end of the protei (e.g. (glycine-proline) _n -random coil, or a sequence corresponding to a helical folding domain), and more convenient and practical strategies may be designed for peptide release and purification (see below).

It may prove possible to reduce to a minimum, by deletions of regions of its gene, the size of the fimbrial subunit protein and yet retain assembly and export. This would further limit the range of potential interactions between the subunit carrier and the covalently (and _# genetically) coupled peptide/protein, and would increase the relative molar yields of^* the newly expressed peptide/protein compared to the fimbrial subunit carrier. It may also prove possible to utilise the natural propensity for internal cleavage of certain serogroups by grafting such that novel expressed sequences are proximal to the C or N terminii generated by protease cleavage of the fimbrial subunit protein. Accordingly we would envisage experimentally testing these possibilities.

In this prototype study we have expressed the sequenc of an important neutralizing antigenic determinant of an economically important virus, FMDV. An immediate application of this system is therefore the development an production of an epitope-based vaccine against foot and mouth disease. The recombinant material is while clearly antigenic in an in-vivo assay is to be tested for its antigenicity and its immunogenicity in vivo. Recent studies have indicated that another VPl peptide, covering residues 200-213, plays a significant contributory role in eliciting protective immunity in cattle, and this peptide is therefore a target for analogous or (combinatorial) expression in the fimbrial system, as part of the strategy for vaccine development, as are other peptide epitopes (e.g. from malaria parasites, polio viruses, hepatitis viruses) .

As will be recognised by persons skilled in the art there are a wide range of possible applications and modifications of the fimbrial peptide expression system described herein. Peptides that may be amenable to production by this system include:

(1) Peptide hormones, such as luteinizing hormone releasing hormone, somatostatin, growth hormone (or fragments thereof), etc.

(2) Pharmacologically and neurologically active peptides, such as exorphina,. endorphins, peptide neurotransmitters, "toxins" and trophic factors.

(3) Peptide epitopes, of which FMDV, malarial repeat sequences, poliovirus and hepatitis B virus and various T cell epitopes are examples. Also included may be peptide epitopes which mimic more complex conformational structures or peptides which while normally non- immunogenic may be rendered immunogenic via presentation alone or with other peptide sequences __ (e.g. T cell epitopes) or fimbriae as the carrier species.

(4) Peptide inhibitors of biological reactions, such as proteolysis, or of biological processes, such as metastasis.

(5) Novel peptides which represent analogues of naturally occurring species, or new sequences with biological activity.

(6) Whole proteins or folding domains of proteins. It is possible that self-contained folding domains or several intact domains (as in many proteins) may interfere less with fimbrial assembly than peptide sequences which are free to interact with fimbrial sequences. Many antigenic epitopes (e.g. the majorit of epitopes in the influenza viral haemoggutinin) are discontinuous in nature and in some such cases whole proteins or intact folding domains may be required to present such epitopes in engineered protein constructions, whilst in others it may prove possible to mimic the determinant in a relatively short peptid sequence. Examples of proteins which could be tested in the fimbrial system range from Bovine pancreatic trypsin inhibitor (a small protein) to B-galactsidase (a large protein). A range of proteins including structural proteins, enzymes and trophic factors (e.g. trypsin, interleukins , members of the growth hormone family) may prove susceptible to production via the system embodied in this patent. 7. Peptides which are either partially or completely of random generated amino-acid sequence. This approach would serve to test the scopes of the fimbrial expression system and would also form the basis of a recombinant DNA approach to the generator/selection o novel peptide specificities. The system may be modified in several ways. Firstly, more general and flexible cloning vector systems can be developed to improve the speed and efficiency of cloning and mutagenesis operations. One may envisage, for example, the introduction of (multiple) rest iction/cloning sites a the 3' end of the fimbrial subunit gene and the construction of more sophisticated vectors to enable reduction in the number of steps required for the engineering and expression of the (modified) fimbrial subunit gene. Secondly, different promoters may be used for gene expression, including regulatable inducible systems which may have significant advantages in circumstances where the mutant construction affects the viability of the host (which may be any compatible type 4 fimbriate bacteria) , or to suit the particular demands of industrial production. Thirdly, the introduction of coding sequences encoding flexible linker or buffer regions between the fimbrial subunit and the grafted peptide may minimize mutual interference in expression and/or activity. In this context, it may also be desirable to include at such junctions, sequences encoding chemically or proteolytically susceptible cleavage sites, enabling subsequent separation and purification of the graft peptide. Following such cleavage the fimbrial carrier could be conveniently removed by standard and industrially applicable precipitation procedures. The advantage of the C-terminal peptide addition is that the peptide itself should be accessible as produced or amenable to cleavage and purification. Thus there exists a choice in presenting the peptide as a macromolecular polymer (which may for example confer immunological benefits) or as a separate entity.

Dependent on the length and nature of allowable additions, it is also possible to envisage a construction and expression of more complex, multi-component peptide sequences with multi-dimensional function, for example, a peptide immunogen containing multiple epitopes together with appropriate signals for effective interaction with various cellular components of the immune system. These various possibilities and modifications of the system will be explored.

Claims

CLAIMS : -

1. A method of producing a peptide comprising culturing bacteria containing the gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae to which has been added at the C-terminal end the nucleic acid sequence encoding the peptide, and an endogenous compatible system for the morphogenetic assembly of type 4 fimbrial and/or the genes for the morphogenetic assembly of such fimbriae derived from a type 4 fimbriate species, such that mature type 4 fimbriae in association with the peptide are produced as extracellular structures, harvesting the whole or part of the fimbriae substantially free of the host cells.

2. A method of producing a peptide as defined in claim 1 wherein the peptide is separated from the type 4 fimbriae.

3. A method of producing a peptide as defined in Claim 1 wherein the gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae, is derived

• from a bacteria selected from the group comprising Bacteroides nodosus, Pseudomonas aeruginosa, Neisseria gonorrhoeae, Neisseria meningitidis, Moraxella nonliquefaciens and Moraxella bovis.

4. A method of producing a peptide as defined in Claim 1 wherein the host cells are bacteria of the genus Pseudomonas and in particular the species P. aeruginosa preferably P.aeruginosa strain PAK/2Pfs (ATCC No. 53308).

5. A method of producing a peptide as claimed in Claim 1 wherein the gene encoding the fimbrial structural subunit to which has been added at the C-terminal end the nucleic acid encoding the peptide, is contained within a plasmid or phage.

6. A method of producing a peptide as claimed in Claim 1 wherein the peptide is VPl derived from foot and mouth disease virus strain 01-BFS.

7. An antigenic preparation for raising antibodies directed against the surface protein VPl of foot-and-mouth disease virus (FMDV) comprising the VPl peptides either in isolation or in association with type 4 fimbriae, which has been produced by bacterial host cells containing the gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae to which has been added at the C-terminal end the nucleic acid sequence encoding the protein VPl, and an endogenous compatible system for the morphogenetic assembly of type 4 fimbriae and/or those genes for the morphogenetic assembly of such fimbriae derived from a type 4 fimbriate species.

8. An^' antigenic preparation as claimed in claim 7 wherein the VPl peptide has optionally been cleaved from the type 4 fimbriae produced by said host cells.

9. An antigenic preparation as claimed in Claim 7 wherein the gene encoding the fimbrial subunit of a bacteria normally producing type 4 fimbriae is derived from a bacteria selected from the group comprising

B.nodosus, P.aeruginosa, N.gonorrhoeae, N.meningitidis, M. nonli uefaciens and M.bovis.

10. An antigenic preparation as claimed in Claim 7 wherein the host cells are bacteria of the genus Pseudomonas and in particular the species is P.aeruginosa preferably P.aeruginosa strain PAK/2Pfs (ATCC No. 53308).

11. An antigenic preparation as claimed in Claim 7 wherein the gene encoding the fi bri.al structural subunit to which has been added at the C-terminal end the nucleic acid encoding the peptide is contained within a plasmid or phage vector.

12. An antigenic preparation as claimed in claim 7 wherein the peptide is VPl derived from foot and mouth disease virus strain 01-BFS.

13. A method of producing a peptide which is substantially as described in the accompanying example.