GB2195340A - Anti-viral vaccine - Google Patents

Abstract

The vaccine comprises an antigen derived from at least partly undenatured or partially structured fusion-protein expressed by a bacterium which has been transformed with a plasmid comprising a gene fragment from the virus against which the vaccine is intended to act. Anti-viral serum may be generated from the vaccine by immunising rabbits or mice with the fusion-protein. The VP7c neutralisation antigen of bovine rotavirus is produced as a fusion protein with beta -galactosidase under the control of the temperature inducible lambda PR promoter. The plasmid may be established in E. coli which can be used as the vaccinating agent.

Description

SPECIFICATION Anti-viral vaccine The present invention relates to anti-viral vaccine and a method for its production utilizing recombinant DNA technology. The invention is more particularly concerned with vaccines against rotaviruses and more specifically bovine rotavirus. However aspects of the invention may have application generally in anti-viral vaccines.

Rotaviruses are responsible for gastroenteritis in a wide range of animals including birds and man. It is estimated that the damage caused in Europe in animal husbandry amounts to about 30,000,00 Ibs (wt) a year. In humans, the problem is generally confined to underdeveloped countries where small children die principally from dehydration.

The particular description which follows describes the production of a vaccine and serum against at least one serotype of bovine rotavirus and its testing in vitro. Experiments conducted as part of and subsequent to this work indicated a failure of serum in an early form to kill live virus from animals. This failure prompted detailed studies in a number of directions, e.g. utilizing alternative fragments from the virus gene. Simultaneously experiments were conducted with other forms of sera derived from the fragment under consideration.

It was then discovered that the initial precedure used to prepare the viral antigen, which involved denaturation of that antigen prior to use in immunising animals, was at fault andresponsible for the lack of efficacy of anti-viral antiserum generated using it.

Subsequently serum was produced against both purified expressed fusion protein that was at least partially re-natured and also whole killed bacteria that had expressed the fusion protein.

These sera when incubated with live virus were able to successfully neutalise virus infectivity in an in vitro assay. The results on partially renatured protein are described hereafter in detail in connection with Fig. 5.

The invention is therefore based on the utilization of a bacterially synthesised viral fusion protein in a form in which it is able to generate an anti-viral serum capable of "killing" live virus.

The expression product is obtained utilising recombinant DNA technology from a fragment of the viral gene which encodes the "neutralisation antigen" of the virus. It includes the immunogenic region necessary to generate neutralising antibodies but when expressed in the bacterium is sufficiently non-toxic to allow production of large amounts of it, e.g. under temperature induction.

Accordingly an aspect of the present invention lies in an anti-viral vaccine comprising an antigen derived from at least partly undenatured or partially structured fusion-protein expressed from a bacterium utilizing a plasmid having an insert comprising of a viral gene fragment.

The invention includes an anti-viral serum generated from the vaccine and effective against live virus.

Preferably the vaccine or serum is effective against at least one serotype of bovine rotavirus and the fragment is a fragment of the bovine rotavirus gene e.g. U.K. tissue culture adapted bovine rotavirus strains as set out under "Bacterial Strains and Reagents" hereafter.

Preferably the fragment is from 300 to 500 bp and is predominantly on the 5'- end of the viral gene. The particularly successful fragment comprises approximately 40% of the aminoterminal coding sequence extending from the Clal site at nucleotide 90 (AA15) to the Hinfi site at nucleotide 481 (AA151).

The particular bacterium utilized in the embodiment is E.coli and the expression is preferably temperature induced e.g. under the regulating control of a bacteriophage promoter.

It is envisaged however that there may be used as the bacterium a bacterium capable of establishing itself as part of the gut flora. Thus the vaccine may be part of a vaccine composition capable of expressing a number of immunising antigens to protect against a plurality of pathogens.

The following particular description illustrates the invention.

Abstract Sequences from genomic RNA segment 8 of the UK tissue culture adapted bovine rotavirus encoding a major viral neutralisation antigen VP7c have been expressed in E.coli. Expression under the regulated control of the bacteriophage lambda PR promoter was as a carboxy-terminal extension to E.coli fi-galactosidase. Following temperature induction high levels of the fusion protein were synthesised and accumulated in induced cells, making up 5%-15% of total bacterial cell protein after two hours of induction. Immunisation of seronegative rabbits and mice with gelpurified fusion-protein raised antibodies which gave specific immunofluorescence with virus infected cells and were able to immunoprecipitate proteins of the VP7 complex from such cells.

Hyperimmune sera also gave a virus type-specific reaction in a solid phase ELISA assay and neutralised virus infectivity in standard plaque reduction assays.

Introduction: In the last ten years rotaviruses have emerged as major medical and veterinary pathogens with the recognition that they are the predominant etiological agents of acute viral gastroenteritis in humans and the main species of domestic livestock (1-3). During the period of overt disease which is usually 3-5 days very large quantities of virus are excreted in the faeces of infected organisms. However, despite their ability to grow well in infected gut tissue, rotaviruses have proved very difficult to adapt to routine growth in tissue culture, with only a few virus strains growing to high titre in a very restricted range of cell lines (4-6).This problem of adaptation of virus isolates to tissue culture growth has severely hampered both the biochemical characterisation of rotaviruses and attempts to produce effective live attenuated or killed virus vaccines by the traditional empirical routes.

The exploitation of recombinant DNA technology to facilitate the production of large quantities of the relevant viral antigens in heterologous expression systems provides an alternative potentially less empirical route towards effective vaccines that is being actively pursued for a range of viral pathogens (7-11). However, the practicalities of this alternative route are such that a prerequisite to its success is a fairly detailed biochemical characterisation of the virus system in question. Two strains of rotavirus, the simian virus SA 11 and the UK tissue culture adapted bovine virus grow to sufficiently high titres to permit their detailed study. Consequently the majority of biochemical data currently available has been obtained by studying these two strains.

The icosahedral virus particle which is approximately 80 nni in diameter has two concentric capsid shells giving rotavirus its characteristic 'wheel-like' appearance in the electron microscope (12). Early serological studies employing 'rough' virus particle preparations from which the outer capsid shell had been removed showed that the virus group specific antigen (VP6) is the major protein of the inner capsid shell and that the type specific neutralising antigen(s) lie in the outer capsid shell (13). Detailed protein analyses revealed that one of the major outer capsid polypeptides, VP7c is a glycoprotein and by analogy with other virus system this was a strong candidate for the virus neutralisation antigen (14).Genetic studies on reassortants between the UK tissue culture (t.c.) adapted bovine virus and various serotypically distinct human virus isolates confirmed VP7c as a major neutralisation antigen (15). In vitro translation of isolated segments of denatured genomic double-stranded (ds) RNA has allowed the genomic RNA segment encoding this important virus protein to be identified as segment 8 in the case of the UK t.c. adapted virus (16). Full length c-DNA clones of this RNA segment have been isolated using a novel cloning strategy (17) and its complete nucleotide sequence determined (18).

The information and basic biological reagents necessary to begin vaccine development using a recombinant DNA based approach are therefore now available. As a first step in this process we have undertaken the expression of the major neutralisation antigen, VP7c, at high levels in E.coli to give an abundant supply of this antigen for potential use in a subunit virus vaccine.

Materials and Methods: 1. Bacterial Strains and Reagents:- Expression studies were carried out using either the strain K12tHlAtrp (20) or MC1O6lpC1 which had a much higher transformation efficiency. The latter was obtained by transforming the E.coli strain MC1061 to kanamycin resistance with the plasmid pC, (19). This piasmid is a derivative of pACYC184 and is therefore compatible with the pBr based expression (pEX) replicons; it carries a temperature sensitive (C,857) C1 gene to allow regulated control of transcription from PR using temperature shift.

The Zubay coupled procaryote transcription-translation system was purchased from Amersham.

2. Construction of expression plasmid and induction of fusion polypeptide: The pEX expression plasmids of Stanley and Luzio (20) were empioyed to achieve the stable expression of rotavirus gene 8 product as a carboxy terminal extension of ssgalactosidase.

Plasmid pCV8 carrying a full length c-DNA copy of genome segment 8 of the UK t.c. adapted bovine rotavirus inserted by G:C tailing into the Pst I site of pAT153 was digested with Pst I and the released viral insert purified on a 1.5% agarose gel (21). The purified insert was digested with CIa I (single cut at nucieotide 90) and Hinf 1 (single cut at nucleotide 481) and the staggered ends filled using Klenow (20). The expression plasmid pEX-3 was digested with Bam H1 which gives a single cut within the polylinker at the end of the ss-galactosidase gene, the Bam H1 ends were filled with Klenow and blunt-end ligated to the prepared viral insert (20). The ligation mixture was used to transform MC1061pC, and transformants selected by growth at 30 on agar plates containing ampicillin (100,ugiml) and kanamycin (10 g/ml). Transformants containing viral sequences were identified by Grunstein-Hogness filter hybridisation (22). If the Clal-Hinfl fragment of viral gene 8 is blunt-end ligated to the filled Bam H1 site in the correct orientation then the Bam H1 site should be reconstructed; therefore clones containing viral inserts were checked using small scale plasmid preparation (23) and digestion of the plasmids with Bam H1 and Xba- 1 (see Fig. 1) to identify those having the correct construction.For induction of fusion polypeptide synthesis fresh overnight cultures grown at 30 were diluted 100-fold into L-broth containing antibodies and allowed to grow at 30 until the Or590 reached 0.3. The culture was then shifted to 42" and shaken vigorously for 90 minutes to 2 hours. The fusion polypeptide was labelled with 35S methionine by pulsing cells after 60 minutes growth at 42" for 10 minutes with 50 ,llCi/ml of 35S methionine.

3. Large scale preparation of fusion polypeptide and production of antisera: Large scale induction of the ss-gal-VP7c fusion protein was carried out on the 1 litre scale using the procedure described above. Induced cells were collected by centrifugation, freeze thawed (3 x), disrupted in gel sample buffer (2% SDS-5% fi-mercaptoethanol) and bacterial polypeptides fractionated on a 7.5% polyacrylaminde gel using the Laemmli buffer system. The large fusion polypeptide was located by straining side strips of the gel. The protein band was then excised with a scalpel and eluted from the polyacrylamide electrophoretically (24).The eluted protein was dialysed against 50mM NH4HCO3, freeze dried and resuspended at approximately 500 Ag/ml in phosphate buffered saline.

For antiserum production Balb-c mice or specific pathogen free New Zealand white rabbits were pre-bled and screened for the presence of rotavirus antibodies using indirect immunofluorescence on virus infected cells. Animals negative at a 1/10 dilution of pre-bleed sera were immunised with approximately 150 pig (rabbits) or approximately 40pig (mice) of purified fusion polypeptide in Freund's complete adjuvant. After 14 and 28 days the animals were given further injections in Freund's incomplete adjuvant and the animals bled and serum screened 10-14 days after the third injection.

4. Immunoprecipitation studies: Radioactivity labelled extracts of virally infected cells were made at 5 hours post infection as previously described (14). Immunoprecipitations were performed using a 100,000 g supernatant of infected cells and in some cases a tight complex of VP7 and VP8 was disrupted prior to precipitation by boiling the sample in gel sample buffer for two minutes and then diluting it ten fold before immunoprecipitation. Immunoprecipitation and washing of the precipitates collected using protein A-sepharose were all carried out by the procedure of Watson et al. (25). Precipitates were fractionated on 5.11% gradient polyacrylamide gels as previously described (14).

5. ELISA and Virus-plaque reduction assays using anti-fusion protein sera:- Solid phase ELISA assays were carried out using purified virions as antigen fixed to 96 well microtitre trays. 50 pil of serial 2-fold dilutions of a 1/5 dilution of rabbit or mouse anti-fusion polypeptide sera was added to wells in duplicate and incubated for 1 hour at 37". The antisera was removed, wells washed 4x with phosphate buffered saline and the plate developed using horseradish conjugated goat anti-rabbit or anti-mouse sera and diaminobenzidine. Plates were read using an ELISA reader and wells scored as positive if the absorbance was greater than 3 x the background control.

For the plaque-reduction assay a solution (0.1 ml) of virus containing 100-200 pfu was mixed with serial 2-fold dilutions of a 1/5 dilution of rabbit or mouse anti-fusion polypeptide sera and incubated for 1 hour at 370. Residual virus was then titred by plaque assay on BSC-1 cells and the assay end point taken as the highest serum dilution giving at least a 60% reduction in virus infectivity.

Results 1. Construction of expression plasmid: The first objective bf this project was to obtain stable expression of a full-length non-fused VP7c in E.coli. Unfortunately our attempts to achieve this have proved unsuccessful, probably due to the instability of this heterologous protein in E.coli. We therefore attempted to generate constructions designed to produce a ss-galactosidase-VP7c fusion in which the viral protein should have been synthesised with a short ( < 15 amino acids) amino terminal extension. This type of construct was produced using the pCL vectors of Zabeau and Stanley (26); it was able to direct the synthesis of the predicted fusion protein when plasmid DNA was used to pro gramme a Zubay coupled in vitro transcription-translation system (27).However when E.coli carrying this expression plasmid were heat induced, stable accumulation of the viral fusion protein in vivo was not apparent. The primary reason for attempting the expression of this protein in bacteria was to give an abundant source of VP7c immunogen for use as a potential subunit vaccine and therefore high level stable expression of the fusion protein in vivo was essential. Expression of the influenza NP protein as a carboxy-terminal extension to ssgalactosi- dase has been shown to give a stable fusion protein in vivo when expression of the same gene as an amino terminal fusion gave no detectable product (Jones personal communication). Con struction of similar carboxy-terminal fusion genes for VP7 sequences was therefore attempted using the pEX expression vectors of Stanley and Luzio (20).Various constructions using different regions of the VP7c coding sequence were made, but of these only that shown in Fig. 1 gave the desired results. Approximately 40% of the amino-terminal coding sequence extending from the Cla 1 site at nucleotide 90 (AA 15) to the Hinf 1 site at 481 (AA 151) was inserted into the pEX3 vector at the carboxy-terminal end of fi-galactosidase. The authenticity of the construction was checked in vitro using the Zubay transcription-translation system programmed (27) with the expression plasmid DNA after digestion with various restriction enzymes (Fig. 1).

2. Expression ofp-gal-VP7e fusion in vivo: Heat induction of E.coli carrying the VP7c expression construct resulted in the high level synthesis of a ss-galactosidase fusion protein larger than the ss-galactosidase made by cells harbouring just the pEX3 expression vector (Fig. 2A). Interestingly in this pulse-labelling experiment at 1 hour post induction some production of unextended fi-galactosidase is clearly evident from the viral gene 8 construct but is not seen or is present in greatly reduced amounts in similar fusion constructions using viral sequences from genes 9 and 10. This probably reflects the accumulation in the cell population of cells carrying mutated plasmids that no longer express the viral sequences.Such cells would be at a stong selective advantage due to the high toxicity in E.coli of viral gene 8 product (VP7c). In line with this interpretation is the observation that if the plasmid carrying the ss-gal-VP7c construct is used to transform E.coli MC1061 not harbouring the C, gene containing plasmid and which therefore should express the fusion protein constituctively, then all the bacterial clones isolated express only ssgalactosidase without the VP7c extension. Despite this apparent high toxicity of VP7c sequences in E.coli Fig. 2B shows that in a Coomassie stained gel in which the product accumulated over the full ninety minutes of the induction the full fusion product is in vast excess over the truncated product observed in pulselabelling experiments.Densitometry of Coomassie stained gels showed that depending on the E.coli strain used for the fusion protein represents approximately 5%-15% of total bacterial cell protein after 90 minutes of induction.

3. Production of antiserum to the fusion and immunoprecipitation studies: To confirm the authenticity of the ss-gal-VP7 fusion and to check its immunogenicity it was used to raise hyperimmune sera in specific pathogen free rabbits and mice which were confirmed seronegative for rotavirus antibodies prior to immunisation. The expressed fusion protein was fractionated on polyacrylamide gels, the relevant protein excised, eluted and used for immunisation. The initial screening of the hyperimmune sera produced by three injections at 10-14 day intervals was by indirect immunofluorescence using fixed infected cells as the antigen source. This showed (data not shown) that the hyper-immune sera gave specific immunofluorescence with rotavirus infected cells.The sera then used in immunoprecipitation studies to show that they were able to specifica!ly precipitate VP7, VP7c, and vpr7 (14) from virus infected cells.

Fig. 3 tracks C-F shows that the hyperimmune serum was able to immunoprecipitate the appropriate viral proteins from a 100,000 g supernatent of infected cells. In these reactions there was however very significant precipitation of a second viral protein VP8. That this precipitation of VP8 was adventitious co-precipitation due solely to the presence in infected cells of a tight complex between VP7 and VP8 was shown by carrying out precipitations on infected cell lysates that had been boiled in gel sample buffer to disaggregate complexes before use. Immunoprecipitation of boiled samples showed that proteins of the VP7 complex were still specifically precipitated but the background co-precipitation of VP8 was greatly reduced (Fig. 3 tracks G-l).

4. Anti-viral activity of fusion protein serum: The immunoprecipitation studies demonstrated that the VP7 fragment of the fusion protein was to some degree immunogenic. Two major questions remained however. Firstly, did the fusion protein contain viral epitopes which would illicit antibodies able to react with complete virus particles, and secondly, was antisera raised against it able to neutralise virus infectivity? The first of these two questions was addressed using a solid-phase ELISA assay employing three different strains of rotavirus known to be in distinct serotypes as the fixed antigen. The results (Fig. 4) showed that the fusion protein which contained VP7 sequences derived from the UK t.c. adapted bovine rotavirus produced a virus type specific immune response as measured in this ELISA assay. This in turn indicated that one or more of the virus serotype specific epitopes was being expressed in this fusion protein.

The second question which was of far greater practical significance was investigated using a standard plaque reduction assay. This showed a greater than 60-fold increase in the neutralising titre of hyperimmune sera compared with the pre-bleed control (Fig. 5), clearly demonstrating that not only did the hyperimmune anti-VP7 sera react with virus in a serotype specific manner but it was able to neutralise virus infectivity. As an additional control in these two assays a sera raised against a fl-gal-VPi0 fusion (see Fig. 2A) was used. Despite the ability of this sera to give specific immunofluorescence with infected cells and immunoprecipitate proteins of the VP10-VP12 complex (14) as expected, sera to this non-structural glycoprotein gave no reaction in the ELISA or plaque reduction assay (Fig. 4 and 5).

Discussion: There has been some progress towards the production of effective vaccines to combat rotaviruses using the traditional empirical approaches. In humans two candidate live attenuated vaccine strains have been developed. The first was a tissue culture adapted bovine virus (RIT 4237) and this gave good levels of protection when used in a trial in human infants in Finland (28). However, the vaccine dose of virus required was very large making the vaccine potentially very expensive to produce. Also when used in developing countries levels of protection were unfortunately much lower, the situation being exacerbated by an apparent interference by prior poliovirus vaccination on the rotavirus 'take' levels (29). The second candidate human vaccine has been derived from a tissue culture adapted rhesus monkey rotarvirus (30).This strain can be used at much lower dose levels making it more economic but it has caused appreciable and occasionally sever fabrile responses in children immunised at the original doseage (31, 32). Todate work in domestic livestock has focused largely on passive protection rather than on active immunisation. This approach involves immunisation of the dam to raise colstral antibody levels to a point where the young animal is passively protected through the time of maximum sensitivity to infection which seems to be pre-weaning in calves and piglets. Such success has been achieved particularly in calves (33) using this procedure and-a virus vaccine for this purpose has recently entered use in the United Kingdom. In human infants, however, the time of maximum sensitivity to infection is 6-24 months and therefore in many countries this passive protection approach would not be practical.A further problem with al! of the above vaccines is that none of them addresses the problem of serotypic variation in rotaviruses and the genetic potential that exists for the rapid emergence of a new serotype in a particular species as the result of genome segment reassortment between viruses normally infecting different species. A number of distinct virus serotypes have already been defined in humans and various domestic livestock species (34) using in vitro neutralisation assays and animal studies have shown a lack of cross protection between some of the serotypes defined in this manner (35). At present not all of the serotypes have proved adaptable to high titre grown in tissue culture and this is a major technical hurdle to the development of completely effective live attenuated or killed vaccine using the traditional routes.

The recombinant DNA based approach to generating large amounts of viral antigen for use as potential vaccines is inherently less empirical than traditional procedures. In addition, it aims to produce a single protein as opposed to a whole virus vaccine which has potential advantages in safety terms. The work reported here clearly represents only the first steps in the realisation of this type of vaccine for rotaviruses and many questions still remain to be answered. Since rotavirus infections are confined solely to the gastrointestinal tract with no evidence of viraemic spread to or from other parts of the organism, it seems reasonable to speculate that the generation of localised immune responses (possible IgA) at gut mucosal surfaces will be of major importance in achieving protective immunity.Therefore an oral route of vaccine administration aimed at stimulating the localised gut responses in preference to a systemic IgG response may prove more effective. A further interesting possibility in this direction is that of incorporating expression constructs of the type reported here into attenuated oral vaccine strains of Salmonella or Shigella. These bacterial strains, used as vaccines in their own right, are able to establish themselves as part of the gut flora. Therefore use of this 'piggy back' vaccine concept would provide one route to producing a 'quasi-live' rotavirus vaccine with all its intrinsic advantages in a manner analogous to that being actively explored using recombinant vaccinia virus vectors (36).

There remains some uncertainty as to the role of a second external virion shall protein (VP3/4) in protection aginst infection. Recent studies (37, 38) have shown that in addition to VP7c this second protein is also involved in elliciting virus neutralising antibody. The relative contributions of these two proteins to the overall generation of protection still remains to be fully established and it may be that to achieve the most efficient vaccine expression of both these viral genes will be required.

Despite the large body of work that remains before realising our goal of an effective recombinant DNA based rotavirus vaccine, the successful expression at high levels in bacteria of a virus antigen able to induce virus neutralising antibody in immunised animals represents a significant step forward.

Acknowledgements This work was supported by grants from the Medical Research Council and Lister Institute.

One of us, MAM, is a Lister Institute Research Fellow.

References: 1. Flewett, T. H. and Woode, G. N. (1978). Arch Virol 57, 1-25.

2. Holmes, I. H. (1983) in The Roeviridae ed. W. K. Joklik Plenum Press 359-423.

3. Du-pont, H. L. (1984). J. Infect Dis 149, 663-666.

4. Estes, M. K., Graham, D. Y., Gerba, C. P. and Smith, E. M. (1979) J. Virol.

5. Smith, E. M., Estes, M. K., Graham, D. Y. and Gerba C. P. (1979). J. Gen Virol 31, 810-815.

6. Sato, K., Inaba, Y., Shinozaki, T., Fuju, R. and Matsumoto, M. (1981). Arch Virol 69, 155-160.

7. Denniston, K. J., Yoneyama, T., Hoyer, B. H. and Gerin, J. L. (1984). Gene 32, 357-368.

8. Kleid, D. G., Yansura, D. Small, B., Dowbenko, D., Moore, D. M., Grubman, M. J., McKercher, P. D., Morgan, D. O., Robertson, B. H. and Bachrach, H. L. (1981). Science 214, 1125-29.

9. Hardy, K., Stahl, S. and Kupper, H. (1981). Nature 293, 481-83.

10. Morairty, A. M., Hoyer, B. H., Shih, J. W. K., Gerin, J. L. and Hamer, D. H. (1981) Proc Natl Acad Sci USA 78, 2606-i 0.

11. Smith, G. L., Murphy, B. R. and Moss, B. (1983). Proc Natl Acad Sci USA 80, 7155-59.

12. Flewett, T. H., Bryden, A. S., Davies, H., Woode, G. N., Bridger, J. C. and Derrick, J. M.

(1974). Lancet 2, 61-63.

13. Woode, G. N. Bridger, J. C., Jones, J. M., Flewett, T. H., Bryden, A. S., Davies, H. A.

and White; G. B. B. (1976). Infect Immun 14, 804-810.

14. McCrae, M. A. and Faulkner-Valle, G. P. (1981). J. Virol 39, 490-496.

15. Kalica, A. R., Greenberg, H. B., Wyatt, R. G., Flores, J., Sereno, M. M., Kapikian, A. Z.

and Chanock, R. M. (1981). Virology 112, 385-390.

16. McCrae, M. A. and McCorquodale J. G. (1982). Virology 117, 435-443.

17. McCrae, M. A. and McCorquodale J. G. (1982). J. Virol 44, 1076-1079.

18. Ellerman, T. C., Hoyne, R. A. Dyall-Smith, M. L., Holmes, i. H. and Azad, A. A. (1983).

Nudl Acids Res 11, 4689-4701.

19. Remaut, E., Tsao, H. and Fiers, W. (1983). Gene 22, 103-113.

20. Stanley, K. K. and Luzio, J. P. (1984). E.M.B.O. J. 3, 1429-1434.

21. Maniatis, T., Fritsch, E. F. and Sambrook, J. (1982). Molecular Cloning, A laboratory manual C.S.H. press.

22. Grunstein, M. and Hogness, D. (1975). Proc Natl Acad Sci USA 72, 3961-3965.

23. Birnboim, H. C. and Doly, J. (1979). Nucleic Acids Res 7, 1513-1520.

24. McCrae, M. A. and Jiklik, W. K. (1978). Virology 89, 578-593.

25. Watson, R. J., Weis, J., Salstrom, J. S. and Enquist, L. W. (1982). Science 218, 381-384.

26. Zabeau, M. and Stanley, K. K. (1982). E.M.B.O. J. 1, 1217-1224.

27. Zubay, G. (1980). Meth Enzymol 65, 856-877.

28. Vesikari, T., Isolauri, E., D'Hondt, E., Delem, A., Andre, F. E. and Zissis, G. (1984).

Lancet, 1, 977-981.

29. De Mol, P., Zissis, G., Butzler, J. P., Mutwewingabo, A., Andre, F. E. (1986). Lancet 2, 108.

30. Stuker, G., Oshiro, L. S. and Schmidt, N. J. (1980). J. Clin Microbiol 11, 202-203.

31. Flewcett, T. H. (1986). Arch Dis Childhood 61, 211-212.

32. Kapikan, A. Z., Midthun, K., Hoshino, Y., Flores, J., Wyatt, R. G., Glass, R. I., Askaa, J., Nakagomi, O., Nakagomi, T., Chanock, R. M., Levine, M. M., Clements, M. L., Dolin, R., Wright, P. F., Belshe, R. B., Anderson, E. L. and Potash, L. (1985). Vaccine 85, 357-367. Cold Spring Harbour Press.

33. Snodgrass, D. R., Fahey, K. J., Wells, P. W., Campbell, I. and Whitelaw, A. (1980).

Infect Immun 28, 344-349.

34. Hoshino, Y., Wyatt, R. G., Scott, F. W. and Appel, M. J. (1982). Arch Virol 72, 113-125.

35. Gaul, S. K., Simpson, T. F., Woode, G. N. and Fulton, R. W. (1982). J. Clin Micro 16, 495-503.

36. Blancou, J., Kiery, M. P., Lathe, R., Lecocq, J. P., Pasturet, P. P., Soulebot, J. P. and Desmettre, P. (1986). Nature 322, 373-375.

37. Greenberg, H. B., Valdesuso, J., Van Wyke, K., Midthun, K., Walsh, M., McAuliffe, V., Wyatt, R. G., Kalica, A. R., Flores, J. and Hoshino, Y. (1983). J. Virol 47, 267-275.

38. Hoshino, Y., Sereno, M. Z., Midthun, K., Flores, J., Kapikian, A. Z. and Chanock, R. M.

(1985). Proc Natl Acad Sci USA 82, 8701-8704.

Fig. 1 Expression plasmid (pEX3/8/5) construction and analysis of translation products from it in a Zubay in vitro transcription-translation system.

The upper portion of the figure shows the essential regions of the ss-gal-VP7 fusion construct (pEX3/8/5) used. The lower part of the figure shows polyacrylamide gel analysis of the protein products synthesised in a coupled in vitro transcription-translation system (20) in response to plasmid DNA digested with various restriction enzymes.

Track A:-Translation products from pBr322 (major product fi-lactamase).

Track B:-products from pBr322 digested with Pst I.

Tracks C and H:-Products from the expression vector pEX3 (major product fi-galactosidase).

Track D:-Products from pEX3 digested with Hind III.

Track E:-Products from the VP7 expression construct pEX/3/8/5.

Track F:-Products from pEX3/8/5 after digestion with Bam H1 (during the construction the Bam Hi site at the point of fusion is reconstructed).

Track G:-Products from pEX3/8/5 digested with Xba 1.

Fig. 2 Expression of the ss galactosidase-VP7 in vivo.

A: 35S methionine labelled protein products synthesised in a 10 minute pulse of uninduced cells (30 ) or induced cells (42 ) at 1 hr post induction. Samples were analysed on gradient polyacrylamide gels as described in Materials and Methods.

VT tracks:-Products synthesised in MC 1061pCl cells carrying the expression vector pEX3.

Sg 8 tracks:-Products synthesised in MC 1061 pC1 cells carrying the RNA segment 8(VP7) expression construct pEX3/8/5.

Sg 9 tracks:-Products synthesised in Mc 1061pC, cells carrying an RNA segment 9 expression construct.

Sg 10 tracks-Products synthesised in MC 1061pC1 cells carrying an RNA segment 10 expression construct.

B:Coomassie-blue stained gel of proteins made by K12SHAtrp cells in the repressed (30 ) state or following ninety minutes of induction (42"). Samples were analysed as for A.

Track A:-Proteins of cells carrying the expression vector pEX3 grown at 300.

Track B:-Proteins of cells carrying the expression vector pEX3 after ninety minutes growth at 42".

Track C:-Proteins of cells carrying the expression construct pEX3/8/5 grown at 30".

Track D:-Proteins of cells carrying the expression construct pEX3/8/5 after ninety minutes growth at 420.

Fig. 3 Immunoprecipitation of virus polypeptides using anti ss-gal-VP7 fusion protein sera Infected cell samples were generated, immunoprecipitated and the precipitates analysed on 5-11% gradient poiyacrylamide gels as described in Materials and Methods. The protein nomenclature used is detailed in (14).

Track A:-Proteins synthesised in a 15 minute pulse-labelling of virus infected cells at 5 hours post infection.

Track B:-Proteins present at the end of a 2 hour chase in 100 x unlabelled methionine of a sample such as that shown in A.

Track C:-lmmunoprecipitate from pulse labelled cells precipitated with pre-immune serum.

Track D:-lmmunoprecipitate from pulse labelled cells precipitated with anti ss-gal-VP7 fusion protein serum.

Track E:-lmmunoprecipitate from pulse labelled cells maintained throughout infection and labelling in the presence of 5 ,ug/ml tunicamycin to block glycosylation precipitated with anti ss- gai-VP7 serum.

Tracks G, H and I are the same as D, E and F except that the cell samples were boiled in gel sample buffer to dissociate tight aggregates before immunoprecipitation.

Track J:-Proteins synthesised in a 15 minute pulse labelling of virus infected cells at 5 hours past infection that had been maintained in 5 pig/ml tunicamycin to block glycosylation.

Track K:-lmmunoprecipitate from pulse-labelled cells precipitated with convalescent bovine anti-virus serum.

Fig. 4 Solid-phase ELISA assay of fusion protein sera using viruses from different serotypes as antigen The ELISA assay was set up and read as described in Materials and Methods. The three viruses used were the homologous UK t.c. adapted bovine virus (UK), the OSU strain of porcine rotavirus (OSU) and the Gottfried strain of porcise rotavirus.

Pre-immune sera.

Anti ss-gal-VP10 sera.

Anti ss-gal-VP7 sera.

Fig. 5 Assay of virus neutralisation by plaque reduction using fusion protein sera.

A standard virus plaque reduction assay was set up, the end point being taken as the highest serum dilution giving at least a 60% reduction in the number of virus plaques. Sera against both a ss-gal-VP7 and. a ss-gal-VP10 (non-structural glycoprotein) were tested.

pre-immune sera.

Hyper-immune fusion protein sera.

Claims (13)

1. An anti-viral vaccine comprising an antigen derived from at least partly undenatured or partially structured fusion-protein expressed by a bacterium transformed with a plasmid which includes an insert comprising a gene fragment from the virus.

2. A vaccine according to claim 1 wherein the gene fragment codes for a polypeptide which includes at least an antigenic portion of a viral protein.

3. A vaccine according to claim 2 wherein the gene fragment codes for a polypeptide consisting of a part only of a viral protein.

4. A vaccine according to any preceding claim wherein the fragment consists of from 300 to 500 base pairs.

5. A vaccine according to any preceding claim wherein the fragment is predominantly taken from the 5'-end of the viral gene.

6. A vaccine according to any preceding claim wherein the vaccine is effective against rotaviruses.

7. A vaccine according to claim 6 wherein the vaccine is effective against bovine rotavirus.

8. A vaccine according to claim 7 wherein the fragment comprises approximately 40% of the amino-terminal coding sequence extending from the Clal site at nucleotide 90 (AA15) to the Hinf 1 site at nucleotide 481 (AA151).

9. A vaccine according to any preceding claim wherein the bacterium is capable of establishing itself as part of the gut flora.

10. A vaccine according to any preceding claim wherein the bacterium is E.coli.

11. A vaccine according to any preceding claim wherein the expression of the fusion-protein is temperature induced.

12. A vaccine according to claim 1 substantially as described herein.

13. An anti-viral serum generated from a vaccine according to any preceding claim effective against live virus.