WO1990015145A2

WO1990015145A2 - Poliovirus chimaeras

Info

Publication number: WO1990015145A2
Application number: PCT/GB1990/000841
Authority: WO
Inventors: Jeffrey William Almond; David John Evans; Karen Lesley Burke
Original assignee: Medical Research Council; University Of Reading
Priority date: 1989-05-31
Filing date: 1990-05-31
Publication date: 1990-12-13
Also published as: CA2055613A1; WO1990015145A3; EP0478588A1

Abstract

A cassette vector suitable for use in constructing poliovirus chimaeras, which vector comprises, under the control of a promoter, a full length infectious cDNA of an attenuated strain of type 1 poliovirus having Sal 1 and Dra 1 sites flanking atigenic site 1 of the poliovirus (I), where the numbers represent the numbers of amino acids of the VP1 capsid protein and X represents intervening nucleotides of DNA, present in sufficient numbers to allow the vector to be digested with both Sal 1 and Dra 1, the said Sal 1 and Dra 1 sites being the only Sal 1 and Dra 1 sites in the vector.

Description

POLIOVIRUS CHIMAERAS The present invention relates to cassette vectors suitable for use in the preparation of poliovirus chimaeras. The icosahedral poliovirus particle is composed of sixty copies of each of four capsid proteins, VPl - VP4, which enclose a positive sense single-stranded RNA genome of approximately 7500 nucleotides. Because of their importance in protective immunity the antigenic sites on the capsid proteins of the three poliovirus serotypes have been studied in derail. These studies have revealed the existence of at least four independent antigenic sites, which induce the production of neutralising antibodies. Antigenic site 1 is a continuous epitope, comprised of residues 91 to 102 of capsid protein VPl. Sites 2, 3 and 4 are conformational, being composed of residues from more than one capsid protein. These sites can be readily located on the 3-dimensional crystallographic model of the virus where they form part of the surface topography. As part of a wider study of poliovirus antigenicity relevant to the development of new and improved poliovirus vaccines, we have previously reported the construction of a type 1/type 3 poliovirus chimaera (Burke et al Nature 332 , 81-82, 1988). This virus, which exhibits dual antigenicity, was constructed by the replacement of antigenic site 1 of the Sabin type 1 poliovirus vaccine strain by the corresponding region of a type 3 strain using oligonucleotide-directed mutagenesis (Kramer et al., Nuc. Acids Res. 12., 9441-9456, 1984) on an infectious full-length Sabin 1 cDNA clone (Stanway et al, J. Virol. 57, 1187-1190, 1986) . The virus induced an immune response against both type 1 and type 3 polioviruses in mice, rabbits and primates.

The Sabin strain of type 1 poliovirus has an established safety record as a vaccine. This, coupled with extensive experience of its manufacture and control, make the Sabin 1 vaccine strain a particularly attractive vector for use as a vehicle for the expression of potentially important epitopes from other pathogens. Since polivoirus is able to induce a mucosal as well as a systemic immune response the approach may be of considerable value where the pathogen in question infects via a mucosal surface where secretory antibodies may play a role in protection from infection.

The cassette approach to in vitro mutagenesis has been reported before for poliovirus (Kuhn et al, Proc. Natl. Acad..Sci. USA £5., 519-523, 1988). It has been employed in the construction of antigenic site 1 chimaeras based upon the neurovirulent Mahoney strain of poliovirus (Martin ^t al_f EMBO, J. 1, 2839-2847, 1988; Murray et al, Proc. Natl. Acad. Sci USA ££, 3202-3207, 1988). The cassette approach was also employed in EP-A-0302801 in preparing hybrid type 1 poliovirus in which a heterologous epitope replaces the C3 epitope which is normally exposed on the surface of the capsid of the poliovirus. A cassette vector has now been constructed which allows rapid and extensive modification of antigenic site 1 of the Sabin 1 poliovirus vaccine strain, Pl/LSc 2ab. Unique restriction endonuclease sites flanking antigenic site 1 have been engineered into a full-length infectious Sabin 1 cDNA clone with minimal alteration to the coding sequence. This facilitates replacement of this region by oligonucleotides encoding foreign amino acid sequences. Results indicate that this region is highly flexible in terms of number and sequence of amino acids which can be accommodated. The approach has general applicability to any attenuated strain of poliovirus type 1.

Accordingly, the present invention provides a cassette vector suitable for use in constructing poliovirus chimaeras, which vector comprises, under the control of a promoter, a full length infectious cDNA of an attenuated strain of type 1 poliovirus having Sal 1 and Dra 1 sites flanking antigenic site 1 of the poliovirus as follows 92 93 102 103

GTC GAC - X - TTT AAA Sal 1 Dra 1

where the numbers represent the numbers of amino acids of the VPl capsid protein and X represents intervening nucleotides of DNA, present in sufficient numbers to allow the vector to be digested with both Sal 1 and Dra 1. the said Sal 1 and Dra 1 sites being the only Sal 1 and Dra 1 sites in the vector.

This cassette vector can be employed to present foreign antigenic determinants. Foreign epitopes can be inserted at antigenic site 1 to replace VPl amino acid residues 94 to 102, thereby obtaining poliovirus chimaeras capable of acting as epitope presentation systems, for example vaccines. The cassette vector has the additional advantage that the Sal 1 and Dra 1 sites are unique to the entire vector, allowing replacement of the region flanked by these sites in a single step and thus obviating the need for subcloning steps in the construction of recombinant cDNAs. The amino acid change at position 102 from aspartic acid to phenylalanine, resulting from the creation of the Dra 1 site, does not affect the viability and growth properties of the vector. Preferably the cassette vector comprises an infectious full length cDNA clone of the Sabin strain of poliovirus type 1 into which the Sal 1 and Dra 1 sites have been engineered. In such circumstances, X represents the codons for amino acid residues 94 to 101 of the VPl capsid protein of Sabin type 1. A suitable spacer region X is one which represents a DNA sequence encoding VPl amino acid residues 94 to 101 of the attenuated strain of type 1 poliovirus being used. X can, however, denote a DNA sequence from which one or more of these codons is missing or, indeed, represent a longer sequence. Typically X consists of from 6 to 30 nucleotides, for example from 9 to 24 nucleotides.

The cassette vector is typically a plas id. The plasmid generally comprises an origin of replication, so that it is replicable in a host which harbours it.

Typically the host is a icrobial host such as a strain of bacterium e.g. E. coli. The plasmid also generally comprises a marker gene such as an antibiotic-resistance gene. A preferred plasmid is pCASl. E. coli MC1061 harbouring pCASl has been deposited at the National

Collection of Industrial and Marine Bacteria, Aberdeen, GB on 25th May 1989 under accession number NCIMB 40148.

Cassette vectors according to the present invention are, like pCASl, generally double-stranded. The nucleotide sequence, and amino acid sequence according to the one letter code (Eur. J. Biochem. 138. 9-37, 1984), for pCASl and other type 1 vectors which do not have missing any site 1 codons in the region of antigenic site 1 is: 91 105 T V D N S A S T K N K F K L F ACCGTCGACAAgPC_AGCTTCCACCAAGAATAAGTTTAAACTATTT TGGCΑGCTGTTGAGTCGAAGGTGGTTCOTA'rTCAAATTTGATAAA Sail Dral

A cassette vector according to the invention may be prepared by first engineering the Sal 1 and Dra 1 sites into a full length infectious cDNA of an attenuated strain of type 1 poliovirus. This may be achieved by subcloning a partial fragment of the cDNA into a single-stranded cloning vector such as one of the M13 vectors and creating the Sal 1 and Dra 1 sites by site-directed mutagenesis using appropriate oligonucleotides. The modified fragment is then reintroduced into the cDNA from which it has been derived. The cDNA may be provided with a promoter, for example a T7 promoter. Alternatively some full length cDNAs are infectious in which case a promoter is not strictly necessary. The cDNA is introduced into a vector having no Sal 1 and Dra 1 sites. The vector may be pFBI 2 (Pharmacia) which has been modified to remove its three Dra 1 sites. Alternatively, the Sal 1 and Dra 1 sites can be provided by using a polymerase chain reaction (PCR) method. The published method of "overlapping PCR" (Recombinant PCR by R. Higuchi in "PCR Protocols: A guide to methods and applications", editors Innis, Gelfand, Sninsky and White, Academic Press, 1990) can be employed. This method overcomes the use of a single-stranded vector for site- directed mutagenesis by the application of the PCR to introduce, via degenerate oligonucleotides the Sal 1 and Dra 1 restriction sites.

In order to obtain a cassette vector in which the spacer region X is modified, i.e. which does not include the normal codons for VPl amino acid residues 94 to 101, a cassette vector prepared as just described is digested with Sal 1 and Dra 1 and an appropriate DNA fragment is ligated with the digested vector. Alternatively, such a vector may be obtained by site-directed mutagenesis or by the PCR method. A modified spacer region X can therefore be provided by digesting a cassette vector according to the invention with Sal 1 and Dra 1 and ligating a double- stranded DNA fragment comprising the desired nucleotide sequence X into the digested vector such that the Sal 1 and Dra 1 sites of the vector are retained. Alternatively, the region separating the Sal 1 and Dra 1 sites may be engineered into a full length infectious cDNA of an attenuated strain of type 1 poliovirus. The cDNA is provided with a promoter, for example a T7 promoter, and is introduced into a vector having no Sal 1 and Dra 1 sites. Where it is wished that X comprises one or more further restriction sites unique to the cassette vector, the vector into which the cDNA is introduced must not also contain these sites. A cassette vector according to the invention is preferably an expression vector. The full length infectious poliovirus cDNA is therefore generally provided in a vector with a promoter and other transcriptional and translational control sequences required for expression. The vector into which a cDNA is cloned may be the plasmid pJMl or a similar vector based on pAT153 from which the gene encoding tetracycline resistance has been removed and replaced by a gene cassette encoding kanamycin resistance. This gene cassette may be from the transposon Tn 903. The plasmid therefore carried both kanamycin and ampicillin resistance genes. The latter gene is totally removed and replaced by the poliovirus cDNA above. The extreme 3' end of the vector may be provided with a Mlu 1 restriction site.

A cassette vector can be constructed in which X represents a nucleotide sequence which results in a frameshift for the downstream poliovirus coding sequence. This prevents the recovery of virus in which the region between the Sal 1 and Dra 1 sites has not been modified by the provision of a nucleotide sequence which restores the reading frame.

The nucleotide sequence denoted by X therefore contains 3n-l or 3n-2 nucleotides, in which n is an integer of at least 1. Typically n is an integer from 2 to 10, for example from 3 to 8. A suitable vector is provided in particular when n is 4. X may denote any nucleotide sequence provided the result is a frameshift in the downstream poliovirus coding sequence.

A cassette vector can also be constructed in which X includes at least one restriction site, such as a Sst II , Not I or Mlu 1 site. The or each restriction site should be the only site of that type in the vector or at least in the poliovirus coding sequence. The provision of a unique restriction site within the spacer region X allows the digestion of the ligation reaction (vector plus oligonucleotide/restriction fragment insert) . Self re- ligated vector is digested by the restriction enzyme for which a site is provided within the spacer region X. The recovery of self re-ligated vector can therefore be significantly reduced, if not prevented altogether. A cassette vector can further be constructed in which a stop codon is provided in the spacer region X. This ensures that the protein encoded by the cassette vector is prematurely terminated and can not be recovered. The stop codon can be introduced in conjunction with a frameshift by providing a nucleotide sequence X having a length of 3n-2 nucleotides. The nucleotides of the Dra 1 site then provide the stop codon as follows. *

TT TAA A where * denotes the stop codon. A cassette vector can be provided which incorporates within the spacer region X any combination of a frameshift, a unique restriction site and a stop codon. A particularly preferred plasmid in which all three are present is pCAS7, which is double-stranded. E. coli MC1061 harbouring pCAS7 were deposited at the National Collection of Industrial and Marine Bacteria, Aberdeen, GB on 18 April 1990 under accession number NCIMB 40277. The nucleotide sequence, and amino acid sequence according to the one letter code, for pCAS7 is shown in the accompanying Figure in which:

(a) shows the Sst2 and Mlul sites and * denotes a stop codon; (b) shows the Sail and Dral sites (underlined) and * denotes the stop codon;

(c) shows how the poliovirus coding sequence can be restored to the correct reading frame; and

(d) shows the provision of a Mlul site at the extreme 3* end of the poliovirus coding sequence.

The cassette vectors of the invention can be employed to present foreign antigenic determinants. Foreign epitopes can be inserted at antigenic site 1 to replace VPl amino acid residues 94 to 102, thereby obtaining poliovirus chimaeras capable of acting as epitope presentation systems, e.g. vaccines.

Poliovirus chimaeras which present a foreign amino acid sequence at antigenic site 1 are prepared by a process comprising:

(i) providing a double-stranded DNA fragment which encodes the amino acid sequence and which has a 5'-Sal 1 cohesive end and a 3*-blunt end;

(ii) digesting a cassette vector according to the invention with Sal 1 and Dral and ligating the fragment from step (i) with the digested vector; and (iϋ) obtaining live virus from the modified vector obtained in step (ii) .

Step (i) is generally conducted by constructing a double-stranded DNA fragment by synthesising complementary oligonucleotides and annealing the oligonucleotides. The oligonucleotides may be boiled together for from 2 to 5 minutes, for example for about 3 minutes, and allowed to cool to room temperature. Alternatively, a restriction fragment may be provided.

In step (ii) the DNA fragment, for example the annealed oligonucleotides, is ligated with a cassette vector which has been digested with Sal 1 and Dra 1 to excise the intervening DNA. E. coli may then be transformed with the ligation mix. Where the spacer region X contains a site for a restriction enzyme and the site is unique to the cassette vector, the ligation mix is digested with the restriction enzyme. This is to prevent recovery of the re-ligated parental cassette vector. The ligation mix is screened for the presence of the recombinant vector.

Live virus is recovered from the modified full length cDNA by production of a positive sense RNA. The vector incorporating the foreign DNA fragment is cut by a restriction enzyme outside the Sabin 1 cDNA. The promoter controlling transcription of the cDNA then enables RNA to be obtained. A T7 promoter is particularly suitable for directing transcription in vitro (van der Werf et al, Proc. Natl. Acad. Sci. USA jS3_, 2330-2334, 1986). The recovered RNA may be applied to tissue cultures by standard techniques (Koch, Curr. Top. Microbiol. Immunol. § , 89-138, 1973). For example, the RNA can be used to transfect Hep2C monolayers. After 4 to 6 days incubation, virus can be recovered from the supernatant of the tissue culture.

Any foreign amino acid sequence may be inserted in this way into antigenic site 1 of an attenuated strain of poliovirus type 1. The foreign amino acid sequence may be composed of from 5 to 50 amino acid residues, for example from 6 to 30 residues or from 8 to 20 residues. By

"foreign" is meant that the amino acid sequence is different from the amino acid sequence of antigenic site 1 of the attenuated strain of type 1 poliovirus being employed. Typically the foreign amino acid sequence is not an amino acid sequence derived from a poliovirus.

The foreign amino acid sequence comprises an epitope to which it may be desired to raise monoclonal or polyclonal antibodies. A monoclonal antibody may be produced by any known method. For example, a mouse, rat, rabbit or non-human primate is innoculated with the recombinant virus of the invention. After a sufficient time has elapsed to allow the host animal to mount an immune response, antibody producing cells, e.g. the splenocytes are removed and immortalized by fusion with an immortalizing cell line such as a myeloma cell line. The resulting fusions are screened for antibodies to the foreign amino acid sequence. The antibodies may be, for example, of the IgG or IgM type. Polyclonal antiserum may also be produced using known methods, using, for example, the animals mentioned above.

The epitope may be flanked by one or more spacer amino acid residues at either or each end. From 1 to 4 spacer residues may be provided at either or each end. The spacer residues may be A or G residues or a combination of both. Alternatively residues from the original poliovirus sequence may be employed.

An epitope may be provided to any known or predicted antigenic determinant which is capable of raising neutralising or non-neutralising antibody. The foreign amino acid sequence may for example comprise an antigenic determinant capable of raising neutralising antibody to a pathogenic organism. The epitope may be derived from a virus, bacterium, fungus, yeast or parasite. More especially, the epitope may be derived from a type of human immunodeficiency virus (HIV) such as HIV-1 or HIV-2, hepatitis A or B virus, human rhinovirus such as type 2 or type 14, herpes simplex virus, poliovirus type 2 or 3, foot- and-mQuth disease virus, influenza virus, coxsackie virus, the cell surface antigen CD4, Chlamydia trachomatis, RSV and HPV e.g. HPV 16. The epitope may be the CD4 receptor binding site from HIV, for example from HIV-1 or -2. Examples of specific epitopes are shown in the Examples.

For reasons which are not entirely clear, not all foreign amino acid sequences enable viable viruses to be obtained. Whether or not a viable virus presenting a particular epitope can be obtained can be determined by carrying out the process steps (i) and (ii) and by seeking to obtain viable virus according to step (iii) . Where a foreign sequence does not give rise to viable virus, it may be possible to alter the sequence slightly to that viable virus is produced, e.g. by shortening or lengthening the sequence or by substituting one or more amino acid residue. Alternatively, viable chimaeras may. be recovered following the modification of other regions of the particle, for example surface adjacent residues. A test may therefore need to be undertaken to see if a particular foreign amino acid sequence will give rise to viable virus.

The poliovirus chimaeras that are obtained can be used as vaccines. They may therefore be formulated as pharmaceutical compositions further comprising a pharmaceutically acceptable carrier or diluent. Any carrier or diluent conventionally used in vaccine preparations may be employed. For example, the presently used live attenuated poliovirus strains are stabilised in a solution of 1M MgCl₂. The poliovirus chimaeras may therefore be used to prevent infections and/or diseases in a human or animal.

The chimaeras may also be administered for therapeutic reasons. For this purpose, they may be administered orally, as a nasal spray or parenterally, for example by subcutaneous or intramuscular injection. A dose corresponding to the amount administered for a conventional live poliovirus vaccine, such as from 10⁵ to 10⁶*⁵TCID5Q, may be given although the dose will depend upon a variety of factors including the viability and replicative capacity of the poliovirus chimaera.

The following Examples illustrate the invention. Example 1: Construction of cassette vector pCASl Taking advantage of codon degeneracy, the nucleotide sequence of Sabin 1 cDNA in the region 2740-2800 was searched for sequences at which restriction endonuclease sites unique to the cDNA could be introduced with minimal alteration to the amino acid sequence. It was observed that a Sal 1 site at nucleotide 2753 could be created without alteration to the amino acid sequence and that this site would be unique within the virus sequence. Similarly a unique Dra 1 site could be created at position 2783 resulting in the replacement of aspartic acid (VPl residue 102) by phenylalanine.

The synthetic oligonucleotides 5•-GGAAGCTGAGTTGTCGACGGTTATAATGG-3• and

5'-CACTGTAAATAGTTTAAACTTATTCTTGG-3' (bases inducing changes underlined) were used to create Sal 1 and Dra 1 restriction sites at positions 2753 and 2783 respectively on a 3.6kb Kpn 1 partial fragment (nucleotides 66-3660) of an infectious Sabin 1 cDNA (Stanway et al, J. Virol. 57, 1187-1190 1986) subcloned in Ml3mpl8, using the gapped-duplex mutagenesis technique (Kramer et al, Nuc. Acids Res. 12_, 9441-9456, 1984). The alterations made to the antigenic site were confirmed by dideoxy chain termination sequencing.

The nucleotide and amino acid sequence of poliovirus Sabin 1 illustrating changes introduced in the construction of pCASl are shown below. Nucleotides 2750-2794 of the cDNA sequence of the viral sense strand are shown, together with the location of the introduced restriction sites. The resulting amino acid change to phenylalanine from aspartic acid at position 102 is shown in parenthesis.

ANTIGENIC SITE 1 91 (F) 105

T V D N S A S T K N K D K L F ACC GTG GAT AAC TCA GCT TCC ACC AAG AAC AAG GAT AAG CTA TTT GTC GAC TTT AAA

Sal 1 Dra 1

The mutated fragment was introduced into a full- length 1 cDNA of Sabin type 1 onto which a T7 promoter had previously been engineered at the extreme 5' end. This full-length clone was subsequently transferred into vector pFBI 2 (Pharmacia), which had been modified to remove its 3 Dra 1 sites at positions 2052, 2071 and 2763, by insertion of an Eco Rl linker following Dra 1 digestion. An Eco Rl - Sal 1 fragment carrying this modified full-length poliovirus clone was ligated into Eco Rl-Xho 1 digested pFBI 2-derived vector thereby destroying this Sal 1 site. The resulting plasmid, pCASl, therefore contained a full-length Sabin 1 cDNA under the control of a T7 promoter and in which the introduced Sal 1 and Dra 1 sites were unique.

N Recovery of infectious virus from Nael linearised pCASl was achieved following transfection of Hep 2C onolayers with transcripts produced in vitro by T7 RNA poly erase (Stratagene) as previously described (van der Werf et al, Proc. Natl. Acad. Sci. USA 8_3, 2330-2334, 1986). The genomic sequence of recovered virus was verified by primer extension sequencing of viral RNA (Rico-Hesse e_t al, Virology, 160, 311-322, 1987). The single substitution of aspartic acid for phenylalanine at residue 102 had no apparent effect on virus viability. Furthermore the design of the cassette was such that the altered amino acid would be lost upon insertion of replacement sequences.

Example 2: Construction of a chimaeric poliovirus containing residues 735-752 of the transme brane glycoprotein gp41 of HTLV-IIIB

The residue numbers 735-752 are those as defined by Kennedy et al. Science 231, 1556-1559, 1986. 100 ng each of complementary oligonucleotides encoding the HIV-1 sequence of choice were boiled for three minutes and allowed to cool

SUBSTITUTESHEET to room temperature. The oligonucleotides were: TCGACCGCCCTGAGGGCATCGAGGAAGAGGGCGGTGAGCGCGATCGTGATCGTTCG; and GGCGGGACTCCCGTAGCTCCTTCTCCCGCCACTCGCGCTAGCACTAGCAAGC.

Aliquots of this annealed mix were then ligated with Sal 1 - Dra 1 digested pCASl. Competent E. coli were transformed with the ligation mix and screened for the presence of a recombinant plasmid containing the HIV sequence inserted. The resulting recombinant plasmid, pSl/env/3 was linearised with Nael, which cuts within vector sequences of the construct, and used as a template in a T7 transcription reaction (van der Werf et al, Proc. Natl. Acad. Sci. USA 80, 5080-5084, 1983) prior to transfection of sub-confluent Hep2C monolayers.

After three to four days a cytopathic effect was observed. The RNA sequence of approximately 200 bp spanning antigenic site 1 of the recovered chimaeric virus Sl/env/3 was confirmed by primer directed chain termination sequencing. The nucleotide and amino acid sequence of the region of antigenic site 1 of pCASl and of the corresponding region of pSl/env/3 are shown below. Bg 1 and Pvu 1 restriction sites, introduced into pSl/env/3 to aid in the screening of recombinant plasmids, are shown underlined as are the Sal 1 and Dra 1 sites in pCASl.

" pCASl 91 105

• • T V D N S A S T K N K F K L F ACCGTCGACAACTCAGCTTCCACCAAGAATAAGTTTAAACTATTT Sail Dral

pSl/env/3 T V D R P E G I E E E G G E R D R D R ACCGT GACCGCCCTGAGGGCATCGAGGAAGAGGGCGGTGAGCGCGATCGTGATCGT TGGCAGCTGGCGGGACTCCCGTAGCTCCTTCTCCCGCCACTCGCGCTAGCACTAGCA Ball Pvul

S K L F TCGAAACTATTT AGCTTTGATAAA

Example 3: Characterisation of Sl/env/3

The antigenic properties of Sl/env/3 were investigated. The virus was neutralized by poliovirus type 1 polyclonal antisera, and by monoclonal antibodies directed against antigenic sites 2 and 3 of the Sabin 1 strain (data not shown) . However monoclonal antibodies specific for antigenic site 1 of Sabin 1 did not neutralize the chimaeric virus. The monoclonal antibodies failed to recognize the chimaera in antigen blocking tests (data not shown) , thus confirming that this site was altered in the chimaera. The recent observation that the structure of antigenic site 1 may influence poliovirus host range prompted us to investigate the interaction of Sl/env/3 with the poliovirus receptor. Sl/env/3 infection of Hep2C monolayers was blocked by a MAb specific for the receptor (Minor e_t al. Virus Res. 1, 203-212, 1984), thus demonstrating that the chimaera still used the normal poliovirus cellular receptor for attachment and entry into cells.

The ability of Sl/env/3 to absorb out the HIV-1 neutralizing activity of antipeptide monoclonal antibodies or of human immune sera was investigated. Preincubation of antibodies with sucrose cushion purified Sl/env/3 abrogated the neutralization of HTLV-IIIB by the IgM monoclonal antibodies (Mab) ED6 and LA9 which were raised against the corresponding gp41 peptide, amino acids 735-752 (as defined above). Preincubation with Sl/env/3 also reduced the neutralization of HTLV-IIIB by five of six human immune sera. In contrast the neutralization of HTLV-IIIB by an IgGl MAb (110.3), mapping to the type specific loop in the second conserved region of env (amino acids 307-321) was not inhibited by Sl/env/3. Moreover, preincubation with Sabin type 1 poliovirus had no effect on the neutralization of HTLV-IIIB by ED6, LA9, 110.3 or the human immune sera.

The results are shown in Table 1 below, where they are expressed as the reciprocal of the serum dilution giving >90% reduction in HIV titre (Weiss et al.. Nature 324, 572-575, 1986) following pre-incubation of the MAbs or human immune sera with culture medium (Mock), Sabin 1 or Sl/env/3. Human immune sera numbered 1-6 represent anonymous HIV+ blood donors. 5x10* TCID₅₀ units of sucrose purified Sl/env/3 or Sabin 1 were preincubated with a 1:10 dilution of the ED6, LA9, a 1:100 dilution of 110.3 and a 1:1 dilution of each human serum for 1 hour at 37°C. Residual HIV-neutralizing activity was determined by incubating dilutions of the antibody/virus mixture with 10³ infectious units (TCID₅₀) of HTLV-IIIB for 1 hour at 37°C. 100 μl of medium containing 2x10* C8166 cells were added, and the presence of εyncytia recorded after 48 hours as an indication of HIV infection.

Table 1: Sl/env/3 inhibition of HTLV-IIIB neutralization HTLV-IIIB neutralization titre - pre-incubation with

Mock Sabin 1 Sl/env/3

Monoclonal Antibod;ies

ED6 640 640 0

LA9 640 640 0

110.3 1000 1000 1000

Human Sera

1 160 160 20

2 40 40 10

3 40 40 10

4 80 80 20

5 320 160 40

6 80 80 80

The immunogenic potential of Sl/env/3 was investigated by raising antisera in rabbits. Neutralizing activity against HIV-1 was determined by infectivity inhibition and plaque reduction assays. All antisera showed neutralising activity against HTLV-IIIB confirming that the chimaera has^' immunogenic potential. Antiserum Rl which contained the highest anti-poliovirus activity was further tested against a range of HIV-1 isolates. This antiserum neutralized the entire test panel at various titres, including the African isolates CBL4 (Tanzania) and Z84 (Zaire). The neutralizing titres observed in both neutralization assays used were in good agreement.

Antisera R7 and R8 also neutralized the HIV-I isolates tested (HTLV-IIIRF or Z84), though the titres observed were lower against both HIV-1 and Sl/env/3. Pre-immune sera and hyperimmune Sabin 1 rabbit antisera displayed negligible neutralizing activity against any of the HIV-1 isolates. The HIV-1 neutralizing activity of antiserum Rl was absorbed out by pre-incubation with the chimaera, but not with Sabin 1, confirming that the activity was induced by the gp41 epitope and not by a chance cross- reactive poliovirus epitope. Antiserum Rl was also tested for its ability to inhibit early syncytial formation in a mixture of HTLV-IIIB producing cells and uninfected C8166 cells, a T cell line sensitive to HIV-mediated fusion. Antiserum Rl was found to inhibit HIV-induced cell fusion, though at lower titre than that determined by virus inhibition (data not shown).

Table 2 below shows the results of neutralization of HIV-1 infectivity and inhibition of syncytium formation by Sl/env/3 antisera. Rabbit Rl was immunised intrader ally with 0.1ml (approx. 10* TCID₅₀ ml^"1) of sucrose purified Sl/env/3 in complete Freunds adjuvant, and boosted subcutaneously at two week intervals with the same virus preparation in incomplete Freunds adjuvant. Rabbits R7-R9 (Sl/env/3) and R19-R21 (Sabin 1) were immunised intramuscularly with 0.5ml of tissue culture fluid (approx. 10* TCID₅₀ ml^"1) in complete Freunds adjuvant, and boosted at two week intervals in a similar manner. Neutralization titres were determined by incubating 10/1 of heat inactivated antiserum with 40/1 of virus supernatant containing 10³ infectious units of HIV-1 at 37°C for 1 hour.

Residual HIV-1 infectivity was measured by the infectivity inhibition assay (Weiss e_t al_^, 1986) described in relation to Table 1. Antiserum Rl was also tested for HIV-1 neutralising activity in a plaque reduction assay (Harada e_t a^, Science 229, 563-565, 1985) on the sensitive HT4 cell. line. The results obtained by this independent assay are shown in brackets. Results are expressed as the reciprocal of the serum dilution giving >90% reduction in HIV infectivity or plaque formation. Also shown is the reciprocal neutralisation titre of Sl/env/3 antisera with 100 TCID₅₀ units of the homologous virus, nt - not tested.

Table 2: Neutralization of HIV-1 infectivity and inhibition of syncytium formation by Sl/env/3 antisera

Reciprocal neutralization titre Virus strain

Anti .serum

IIIB IIIRF SF2 SF33 CBL4 Z84 Sl/env/3

Rl pre- ■immune <10 <10 <10 <10 <10 <10 <10 final bleed 80 80 80 40 10 40 >28960

(40) (40) (160) (40) (80) (40)

R7 \ pre- •immune <10 <10 nt nt nt <10 <10 final bleed 20 40 nt nt nt 40 >2560

R8 pre- ■immune <10 <10 nt nt nt nt <10 finatl bleed 40 20 nt nt nt nt >2560

R9 pre- •immune <10 <10 nt nt nt nt <10 final bleed 40 10 nt nt nt nt 1280

R19, R20 and R21 pre- ■immune <10 post :-immune <10

Monoclonal antibodies were raised against Sl/env/3, and their reactivity with Sl/env/3 but not the parental Sabin 1 was demonstrated in antigen blocking tests (data not shown). Four MAbs specific for Sl/env/3 were characterized in terms of their reaction with several HIV-1 isolates. One MAb (1577) displayed neutralizing activity against all HIV-1 isolates tested, including the three African strains CBL4, Z84 and Z129. Monoclonal antibodies 1575 and 1583 displayed a more restricted respone neutralizing only some of the isolates, suggesting that they recognize a defined epitope within the gp41 735-752 region, that is less well conserved. Monoclonal antibody 1578 displayed no HIV-1 neutralizing activity, suggesting that it recognised an epitope formed from both HIV-1 and poliovirus amino acids.

The neutralization of HIV-1 infectivity by Sl/env/3 monoclonal antibodies is shown in Table 3. Murine monoclonal antibodies were raised as described by Fergusson et al, J. Gen. Virol, £5, 197-201 (1984), and screened agaitist Sl/env/3 and Sabin 1 in antigen blocking tests. The Sl/env/3 and HIV-1 neutralizing activity of the ascites from four hybridomas was determined. Results are expressed as the reciprocal of the MAb ascitic fluid dilution giving >90% reduction in HIV-1 titre, or neutralizing 100 TCID₅₀ units of Sl/env/3.

Table 3: Neutralization of HIV-1 infectivity by Sl/env/3 monoclonal antibodies

Monoclonal Antibody - Reciprocal Neutralization titre

Virus Strain

Mab IIIB IIIRF SF33 CBL4 Z84 Z129 Sl/env/3

1575 40 20 40 10 10 <10 2560

1577 40 160 80 20 20 10 640

1578 <10 <10 <10 <10 <10 <10 640

1583 40 40 20 <10 <10 <10 >28960

The specificity of the HIV-1 immune response to Sl/env/3 was demonstrated by Western blotting and by a peptide binding assay. Antiserum Rl reacted with the envelope glycoprotein precursor gpl60 and gp41 in Western blots whereas re-immune sera was negative. All the rabbit antisera bound specifically to a linear synthetic peptide corresponding to the HTLV-IIIB epitope present on the SI/env/3 chimaera whereas they failed to bind to a 15 amino acid peptide derived from the type specific neutralization epitope on gpl20 (residues 307-321). Pre-immune sera and control rabbit hype immune Sabin 1 antisera (R19, 20, 21) displayed no specific binding to either the gp41 or gpl20 derived peptides. The three SI/env/3 monoclonal antibodies which neutralized HIV-1 (1575, 1577, 1583) also reacted with the gp41 735-752 peptide in peptide binding assays (data not shown) .

Example 4: Construction of chimaeric polioviruses containing amino acid sequences from hepatitis A virus, rhinovirus types 2 and 14 and coxsackie B4 virus

Short sequences of hydrophilic amino acids, as shown in Table 4, from the capsid protein VPl of Hepatitis A virus (Najaran e_t al., Proc. Natl. Acad. Sci. USA £[2,2627-2631, 1985), which might constitute potential antigenic sites, were inserted into the Sal 1-Dra 1 digested pCASl vector according to the procedure described in Example 2. Appropriate oligonucleotides were annealed by boiling and allowing to cool to room temperature prior to ligation into Sal 1-Dra 1 digested pCASl. T7 transcripts prepared from the resulting recombinant plasmids were used to transfect Hep 2C monolayers (van der Werf e_t al, 1986). All plasmids containing the Hepatitis A virus inserts gave rise to viable virus 24-36 hours post-transfection. The sequence of each recovered virus was confirmed by sequencing viral RNA (Rico-Hesse et al, 1987).

Similarly oligonucleotides corresponding to known or predicted epitopes from human rhinovirus (HRV) serotypes 2 (Skern et al, Nuc. Acids Res. l , 2111-2126, 1985) and 14 (Stanway e_t al., Nuc. Acids Res. L2, 7859-7875, 1984) and from the capsid protein VPl of Coxsackie B4 virus (Jenkins et al, J. Gen. Virol. 6 , 1835-1848, 1987) were ligated into pCASl. Resulting plasmids were tested for viability as above. Sequence analysis of genomic RNA confirmed that recovered viruses had the expected modifications of antigenic site 1. In one case, however, a recombinant plasmid containing an insert sequence (marked *) from Coxsackie B4 virus did not produce viable virus upon repeated transfection although the DNA sequence of the engineered site was correct.

Table 4: Amino acid sequences inserted into pCASl and their origin Virus Amino acid sequence Location

Sabin 1 NSASTKNKD VPl 94-102

CAS 1 NSASTKNKF VPl 94-102

Sabin 3 NEQPTTRAQ VPl 92-100

Hepatitis A (N)EQNPVD(D) VPl 15-20

KDLKGKANRGKMD VPl 29-41

ELKPGESRHTSD VPl 70-81

TFNSNNKEY VPl 99-108

NSNNKEYT(D) VPl 111-118

(N)ATDVDG(KD) VPl 150-155

NTRRTGN(KD) VPl 191-197

(N)GLGDKTDS VPl 217-224

DPRSEEDKRFE VPl 290-300

Rhinovirus 2 KLEVTLANY VPl 81-89

14 KDATGIDNHREA VPl 85-96

14 (N)MYVPPGAPNP(D) VPl 151-160

14 (N)KLILAYTPPGARGPQD VP3 126-141

Coxsackie B4 (N)SAESNNL(D) VPl 81-87

IYIKYSSAESNNL* VPl 75-87

Residues in parenthesis correspond to amino acids which have been retained from the wild-type Sabin 1 sequence.

Example 5: Construction of further viable poliovirus chimaeras

Following the procedure described in Example 2, viable poliovirus chimaeras were constructed by inserting annealed complementary oligonucleotide encoding the epitopes shown below into the Sal 1 - Dra 1 digested pCASl. HIV-1

Sl/env/1 GGDMRDNWRSELY residues 469-482

Sl/env/4 NMWQEVGKAMYAPPISG residues 423-439 Sl/env/5 AAPKNPRNKA (-ve strand) residues 48-57 (Numbering of residues is according to the sequence of the molecular clone NY5/LAV-1 as referenced in "AIDS and Human Retroviruses 1988" compiled by G. Myers, Los Alamos, US). gp41 residues 598-609 according to Gnann et al_^, J. Virol. 61, 2639-2641, 1987

(D) L G L W G C S G TC GAC CTG GGG TTG TGG GGT TGC TCT GGA

K L I C T T AAG CTT ATT TGC ACC ACT CD4

S1/CD4/1 NQGSFLTKGPSKLND residues 64-78 (Numbering of residues is according to Maddon et al. Cell 42, 93-104, 1986 quoting the unprocessed polypeptide)

Herpes Simplex Virus type 1

Sl/HSV/1 LKMADPNRFRGKDL residues (gD) 9-22

(Numbering of residues is according to Minson et a , J. Gen.

Virol. £7, 1001, 1986)

Influenza Virus

Sl/flu/1 NACKRGPGSGFFS residues (HA) 137-149 (Numbering of residues is according to the sequence of Aichi/2/68/(X-31) reported in Laver e_t al.. Nature 283, 454-457 , 1980 ) .

Haemagglutinin residues 186-200 (site B) according to Wiley e_t al. Nature 28_£, 373-378, 1981

(D) S T N Q E Q T S TC GAC AGT ACT AAC CAG GAA CAG ACC TCA

L Y V Q A S G CTG TAC GTG CAG GCA TCA GGA

Poliovirus type 2 Sabin strain

Sl/2 NDAPTKRAS VPl residues 94-102

(Numbering of residues is according to Minor e_t a^, J. Gen.

Virol. 67, 1283-1291, 1986)

Chlamydia trachomatis serovar A MOMP VAGLEKDPVA MOMP A no. 2

VAGLENDPVA MOMP A no. 1 Serovar B MOMP NNENQTKVSNGAFV MOMP B serovar C MOMP TKTQSSSFNTAKLI (non-viable) serovar L2 NENHATVSDS MOMP L2 (MOMP - major outer membrane protein; epitopes are according to Baehr et al, Proc. Natl. Acad. Sci. USA 8_ , 4000-4004).

Foot-and-mouth disease virus

VPl 40-49 VKVTPQNQIN FMDV5/6

*VP 140-160 AVPNLRGDLQVLAQKVARTLP FMDVABCD

VPl 142-153 NLRGDLQVLAQ FMDVl/2

VPl 147-156 DLQVLAQKVA FMDV3/4

VPl 200-213 RYKQKIIAPAKQLL FMDV7/8 (Numbering of residues is according to the sequence of 0₁K strain as reported by Forss et a_l., Nuc. Acids Res. 12, 6587-6601, 1984 apart from the entry marked * which is an epitope identified by Bittle et l. Nature 7 ^, 30-33, 1982 and Pfaff e_t al, EMBO J. 1, 869-874, 1982).

Human Papillomavirus type 16 (HPV16)

LI residues 269-284 according to Patel e_t al, J. Gen. Virol

J_0, 69-77, 1989

(D) E N V P D D L Y I TC GAC GAA AAC GTG CCA GAC GAC TTG TAC ATT

K G S G S T A AAA GGA TCA GGA TCC ACC GCA

Respiratory Syncytial Virus (RSV)

Fusion glycoprotein (FI) residues 221-229 according to

Trudel et al, J. Gen Virol. 8, 2273-2280, 1987

(D) (N) I E F Q Q K N N R TC GAC AAC ATT GAA TTC CAG CAG AAA AAC AAC AGA.

£^ ? rr~. »^__ Example 6: Construction of cassette vector PCAS7

Cassette vector pCASl is described in Example 1. pCAS7 was obtained from pCASl as follows: (a) The Sail and Dral sites are maintained in the same position as pCASl, as is the remainder of the Sabin 1 poliovirus sequence which is unmodified (see point d) .

(b) The region separating the Sal 1 and Dra 1 sites has been modified in the following way; 1) The poliovirus sequences have been removed and, 2) replaced with oligonucleotides containing two restriction sites unique to Sabin 1 which 3) result in the introduction of a "frameshift" i.e. the polyprotein reading frame is not maintained, thereby preventing the recovery of virus in which the region separating the Sal 1 and Dra 1 has not been modified. To further ensure the latter, we have also introduced a stop codon that will be read immediately following the frameshift, therefore prematurely terminating the polyprotein. (c) The plasmid vector in which the poliovirus sequence is cloned has been changed. The new vector is called pJMl, and is based upon pAT 153, from which the gene encoding tetracycline resistance has been removed, and replaced with a gene cassette encoding kanamycin resistance (from the transposon Tn903) . This vector therefore carries both kanamycin and ampicillin resistance genes, but the latter is totally removed and replaced with the Sabin 1 sequence. This vector has proved to be far more stable than the original one used for pCASl (which was the commercially made pFBI) , and results in faster growing colonies which contain few, if any, deleted plasmids.

(d) The extreme 3' end of the poliovirus sequence has been slightly modified by the addition of a Mlul restriction site (through since this lies outside the poly- A tail it could be considered a vector restriction site into which the poly-A tail is cloned) . The recognition sequence of Mlul is A'CGCGT, and cleaves between the A and the C. We therefore cut recombinant (chimaeric) cDNA's with Mlul prior to producing runoff transcripts for transfection.

Example 7: Construction of viable poliovirus chimaeras using PCAS7

Following the producedure described in Example 2, the following constructs have been generated with pCAS7. All these chimaeras are viable.

Name ^N Sequence amino acid numbers a) Containing sequences derived from HIV-1 gpl20. (Unless otherwise stated all HIV-l sequences are based upon the BRU isolate catalogued by Myers et al. , 1989 (AIDS and Human Retroviruses, 1989. Los Alamos National Laboratory, Los Alamos USA) , and originally sequenced by Wain-Hobson et al 1985 Cell 4J): 9-17. Some of the chimaeras are variations upon particular sequences eg. Sl/env/2a. The only exceptions to this are Sl/env/2b which is based upon the MN isolate sequence, and Sl/env/6 and Sl/env/7 which are respectively based upon the MAL and ELI isolates. All these isolates are catalogued in the Myers et al 1989 reference) .

Sl/env/2a LGIWGCSGKLICTT 598-611

Sl/env/2b LGFWGCSGKLICTT 598-611

Sl/env/3 DRPEGIEEEGGERDRDRS 735-752

Si/env/6 NMVAGRKAIYAPPIERN

Sl/env/7 NTWQGYGQAMYAPPIEG

Sl/env/8 NSASQRGPGRAFTKNKF

Sl/env/9 APPISGQISCSSNID 441-455

Sl/env/10 LPCRIKQFINMWQEVG 421-436

b) Containing sequences derived from HIV-l ?ιef

Sl/nef/1 VRERMRRAEPAADG 16-29

Sl/nef/2 LEAQEEEEVGFPV 58-70

Sl/nef/3 GLEGLIHSQRRQDILD 97-112

Sl/nef/4 VPVEPDKVEEANKGEN 146-161 c) Containing sequences derived from HIV-2 gpllO (Based upon the original sequence by Guyader et al 1986 Nature 326: 662-669, catalogued by Myers et al 1989) Sl/env/11 NTWHKVGRNVYLPPREG

d) Containing sequences derived from SIVmac251 (Based upon the original sequence by Franchini ≤t al 1987 Nature 328: 539-543, catalogued by Myers et al- , 1989)

Sl/SIV/1 PVTIMSGLVFHSQPLTD 324-340 Sl/SIV/2 NTWHKVGKNVYLPPREG

e) Containing sequences derived from human CD4 (Based upon the original sequence by Maddon et al 1986, Cell 12.: 93-104) S1/CD4/3 NQIKELGNQGSFLTKGPSKLNDRAD 32-56 Sl/CD4/3i NQIKILGNQGSFLTRGPSKLNDRAD 32-56

f) Containing sequences from Hepatitis A Virus (Originally sequenced by Emini et al 1985 J. Virol J55: 836-839) HepA VPl constructs

S1/H18 VDTPWVEKESALS 166-178 S1/H19 ITLSSTSNPPHGL 111-123

Example 8: Antigen chimaera of poliovirus induces antibodies against HPV16

The growth characteristics and antigenicity of the HPV16 chimaera of Example 5 were investigated. This chimaera was designated S1/HPV16/L1. Growth characteristics were assessed in Hep-2c cells. Confluent Hep-2c monolayers in 35-mm tissue cultures dishes (Sterilin) were washed twice with phosphate-buffered saline (PBS) and infected at a multiplicity of infection of 10 PFU per cell. Separate culture dishes were infected with poliovirus type 1 Sabin strain and S1/HPV16/L1.

Virus was adsorbed for 13 minutes at room temperature, and then pre-warmed medium was added and the plates were incubated at 34°C. At regular points of time over a period of 15 hours, medium was aspirated and the cells were washed twice in PBS before being scraped from the dish into a 1.5 ml tube. Cells were pelleted and then lysed by being resuspended in 0.1 ml of buffer containing 0.1 M Tris (pH1.5), 0.1 mNaCl, 1.5 mM MgCl₂ and 0.25% (vol/vol) Nonidet P-40 (Sigma Chemical Co, St. Louis, Missouri, US) .

Virus titres in the lysate were determined by plaque assay on Hep-2c cells grown in six-well dishes (Corning Glass Works, Corning, New York, US) . S1/HPV16/L1 replicated slightly less well than the unmodified Sabin 1 in hep-2c cells, producing a titre of 10⁸*⁴ compared with 10⁹ for Sabin 1. The plaques formed by S1/HPV16/L1 were indistinguishable in size from those of Sabin 1. The antigenic character of S1/HPV16/L1 was examined by a standard neutralisation assay (Jenkins et al, J. Virol., March 1990, 54(3), 1203-1206). S1/HPV16/L1 was neutralised by the anti-HPV16 monoclonal antibodies (Mabs) 8C4, ID6 and 5A4, which are specific for peptide 269-284 of L1-HPV16, and by a polyclonal se um raised against a

Ll-HPVl6/0-galactosidase fusion protein. However, the virus was not neutralised by Mabs which recognised other regions of L1-HPV16.

In a radioimmunoprecipitation assay, a 1/400 dilution of Mab 5A4 recognised S1/HPV16/L1, whereas Mab IC6 failed to react at a dilution of 1/10. Neither antibody recognised Sabin 1. This suggests that the antigenicity of the HPV16 sequence expressed on the surface of poliovirus closely resembles that of the papillomavirus capsid antigen found in infected cells, although it was of interest that one Mab specific for peptide 269-284 (3D1) did not neutralise and failed to recognise the chimaera in a radioimmunoprecipitation assay. The chimaera was also neutralised by polyclonal Sabin 1 antiserum and by Mabs raised against Sabin 1 antigenic sites 2 and 3. As expected, the chimaera was not neutralised by a site 1- specific Mab (955) . Infection of Hep-2c-monolayers by S1/HPV16/L1 was blocked by a Mab specific for the cellular receptor of polioviruses (Minor et al. Virus Res. 1, 1984, 203-212). This demonstrated that the chimaera retained its capacity to bind to an enter cells via the normal route used by polioviruses. This observation also suggests that the extensive modification of antigenic site 1 did not confirm novel receptor-binding properties on the virus. The immunogenicity of S1/HPV16/L1 was assessed in rabbits by subcutaneous inoculation in adjuvant. Sequential bleeds from each of three animals showed an increase in antibodies which neutralised the chimaera. The sera also showed an increase in reactivity in ELISAs against the J-galaσtosidase/Ll-HPV16 fusion protein (-amino acids 172 to 375) and the L1-HPV16 peptide 269-284. Similar results were obtained when a different fusion protein of tryptophan E synthatase and L1-HPV16 residues 1 to 505 were used as a solid phase target. No reactivity against L1-HPV16 peptides 299 to 313 and 329 to 343 was observed.

Since HPV16 can not as yet be cultured in vitro, large amounts of purified virions are unavailable for immunological assay. The ability of the rabbit antisera raised against S1/HPV16/L1 to recognise HPV16 virions was therefore tested by immunoperoxidase staining of HPV16- positive human biopsy material. Representative rabbit antisera raised against S1/HPV16/L1 (third bleed serum from a rabbit) contained antibodies which detected HPV16 antigen in human tissue whereas pre-immune rabbit antiserum did not contain such antibody. Immune rabbit antisera to Sabin type 1 did not react with any tissue sections. The invention further relates to cassette vectors generally. A cassette vector is a construct that allows the rapid and extensive modification of a specific region of a protein. The coding sequence for this specific region is delineated by the presence of flanking 5' and 3' restriction sites. Modification of the region is achieved by digestion of the cassette vector at the 5' and 3' sites and ligation of the digested vector with a DNA fragment with suitable complementary ends. This fragment is usually a pair of annealed complementary oligonucleotides but may be a purified restriction fragment or a suitably prepared PCR product.

We have now devised a cassette vector which simplifies the modification of a specific region of a protein. According to the present invention, there is provided a cassette vector which comprises a coding sequence for a polypeptide and in which a pre-selected region of the coding sequence is defined by flanking 5• and 3' restriction sites which are unique to the vector, wherein the said pre-selected region is composed of such a nucleotide sequence that:

(a) recovery can be prevented of self-ligated vector after digestion of the cassette vector with the restriction enzymes which cut at the flanking 5' and 3' restriction sites; and/or

(b) the cassette vector is incapable of expressing the said polypeptide. The cassette vector may encode any polypeptide. The polypeptide may be a viral capsid protein or a therapeutically active protein, for example a protein capable of exhibiting activity in a human or animal. The polypeptide may be a prokaryotic or eucaryotic protein. The cassette vector may comprise a cDNA, for example a poliovirus cDNA such as the cDNA of an attenuated strain of type 1 poliovirus. A pre-selected region of the coding sequence is defined by 5* and 3' restriction sites unique to the vector. These sites may be engineered into the coding sequence by site-directed mutagenesis or by PCR as described previously. In one aspect, the nucleotide sequence of this pre-selected region is constructed so that recovery of self-ligated vector can be prevented after the vector has been cut at the flanking 5• and 3* restriction sites. The use of molecular genetic procedures such as the use of phosphatase to prevent re-ligation of the cut vector, or the gel purification of the cut vector fragment, all theoretically prevent the reformation of the cassette vector. In practice, however, they are not infallible and considerably increase the vector preparation time. Furthermore, they often decrease ligation efficiency for example due to the presence of contaminating nucleases on agarose gels.

According to the invention, recovery of the parental cassette vector can be prevented by providing at least one restriction site within the pre-selected region of the coding sequence and ensuring that this or each such site is unique to the coding sequence and, preferably, unique to the entire cassette vector. The site may be a Sst II. Not I or Mlu I site. The provision of a unique restriction site within the pre-selected region of the coding sequence means that, upon digestion of a ligation mixture of a vector cut at the 5' and 3' restriction sites and of a DNA fragment insert with the enzyme which cuts at the unique restriction site, the pre-selected region is cut. The enzyme should not of course cut the DNA fragment. The recovery of self re-ligated vector can therefore be prevented.

In a second aspect, the nucleotide sequence of the pre-selected region of the coding sequence is constructed so that the polypeptide encoded by the entire coding sequence can not be expressed. The pre-selected region must be modified for the entire coding sequence to be expressed properly. The coding sequence for the polypeptide is not provided as an open reading frame for the full length of the sequence. This can be ensured either by introducing a frameshift in the pre-selected region of the coding sequence or by providing a stop codon within that region. Both a frameshift and a stop codon can be introduced.

As mentioned above, a cassette vector allows modification of a specific region of a protein. In a cassette vector according to the invention, the protein coding sequence on each side of this region must be maintained. However, the coding sequence for the specific region of the protein it is wished to modify should not be maintained completely. Indeed it is irrelevant what the coding sequence is as it will be replaced. The nucleotide sequence spanning the region of the polypeptide coding sequence that is to be replaced can therefore be constructed so that it results in a frameshift in the downstream nucleotide sequence and, in particular, in the polypeptide coding sequence downstream of the flanking 3' restriction site. Alternatively or additionally, a stop codon can be provided within the nucleotide sequence spanning the region of the polypeptide coding sequence that is to be replaced. Whichever approach is adopted, it is then impossible for the full length protein coding sequence to be translated. A protein for which it is desired to modify a specific region can be represented by the formula (I) :

X-Y-Z (I) in which Y represents the amino acid sequence of the region to be modified and X and Z represent the amino acid sequences of the remainder of the protein. The corresponding coding sequence may be represented by the formula (II) : X'-Y'-Z¹ (II) in which Y' represents the nucleotide sequence encoding the region of the protein to be modified and X* and Z' represent the nucleotide sequences encoding the remainder of the protein. A cassette vector according to the invention can comprise the nucleotide sequence of formula (III) :

X'-y-Z' (III) wherein Y" is flanked by unique 5• and 3' restriction sites and: (a_j) Y' and Y'¹ do not contain the same number of nucleotides and the difference in the number of nucleotides is indivisible by 3; and/or

(b-j Y' * comprises a stop codon. The number of nucleotides of the sequence Y^• ' must be sufficient to ensure that the cassette vector can be cut at both the 5' and 3' flanking restriction sites. Where Y' represents an integral number of codons, the nucleotide sequence denoted by therefore must contain 3n-l or 3n-2 nucleotides, in which n is an integer of at least 1.

Typically n is an integer from 2 to 10, for example from 3 to 8. A suitable vector is provided in particular when n is 4.

The 3' flanking restriction site of the cassette vector may be a Dra 1 site. A stop codon can be introduced, typically in conjunction with a frameshift, by providing the Dra 1 site as follows: *

TT TAA A

where * denotes the stop codon.

The cassette vector is typically a plasmid. The plasmid generally comprises an origin of replication, so that it is replicable in a host which harbours it.

Typically the host is a microbial host such as a strain of bacterium e.g. E. coli. The plasmid also generally comprises a marker gene such as an antibiotic-resistance gene. Preferred cassette vectors and, in particular, plasmid cassette vectors are poliovirus cassette vectors as mentioned previously, in particular pCAS7.

A cassette vector according to the invention can be prepared by first engineering flanking 5• and 3• restriction sites into a polypeptide coding sequence, and then cutting at the restriction sites and ligating an appropriate DNA fragment with the digested vector. Alternatively, the vector may be obtained by site-directed mutagenesis or by the PCR method.

The cassette vector is preferably an expression vector. A promoter and other transcriptional and translational control sequences are provided so that a recombinant protein can be expressed. v A recombinant vector which encodes a protein in which a specific region has been replaced by a different amino acid sequence can be prepared by:

(i) providing a double-stranded DNA fragment which encodes the different amino acid sequence and which has 5* and 3• ends compatible with the restriction sites digested in step (ii) ;

(ii) digesting a cassette vector according to the invention with the restriction enzymes which cut at the said flanking 5' and 3¹ restriction sites of the cassette vector; and

(iii) ligating the fragment from step (i) with the digested vector obtained in step (ii) .

SUBSTITUTESHEET In step (ii) , the cassette vector of the invention is digested with the restriction enzymes which cut at the 5• and 3^» restriction sites which flank the portion of the cassette veόtor which is to be replaced. In step (iii) , the DNA fragment from step (i) is ligated with the cut vector. If appropriate, the ligation mixture is digested with the restriction enzyme which cuts at a unique site provided within the nucleotide sequence of the cassette vector flanked by the 5' and 3' restriction sites. This is to prevent recovery of re-ligated cassette vector.

Any different nucleotide sequence can be provided in the cassette vector. The resulting recombinant vector can therefore encode a polypeptide having a different amino acid sequence in a specific region. The different amino acid sequence may be any foreign amino acid sequence as described previously. A chimaeric protein incorporating the different amino acid sequence can be expressed. A recombinant vector capable of expressing the chimaeric protein is introduced into a compatible host. Expression is allowed to occur.

Cells are transformed with the recombinant vector. Any appropriate host-vector system may be employed. The transformed host may be a prokaryotic or eucaryotic host. A bacterial or yeast host may be employed, for example E. coli or j3__. cerevisiae. Cells of a mammalian cell line may be transformed. The transformed host is cultured under such conditions that expression of the chimaeric protein occurs. This protein can be isolated and purified. It may be obtained in biologically pure form. It may be formulated as a pharmaceutical composition also comprising a pharmaceutically acceptable carrier or diluent.

Claims

1. A cassette vector suitable for use in constructing poliovirus chimaeras, which vector comprises, under -the control of a promoter, a full length infectious cDNA of an attenuated strain of type 1 poliovirus having Sal and Dra 1 sites flanking antigenic site 1 of the poliovirus as follows

92 93 102 103

... GTC GAC X TTT AAA gal 1 E_∑a_l where the numbers represent the numbers of amino acids of the VPl capsid protein and X represents intervening nucleotides of DNA, present in sufficient numbers to allow the vector to be digested with both Sal 1 and Dra 1. the said Sal 1 and Dra 1 sites being the only Sal 1 and Dra 1 sites in the vector.

2. A vector according to claim 1, wherein X represents the codons for amino acid residues 94 to 101 of the VPl capsid protein.

3. A vector according to claim 1, wherein X represents a nucleotide sequence which results in a frameshift of the downstream poliovirus coding sequence and/or which comprises a stop codon.

4. A vector according to claim 1, wherein X represents a nucleotide sequence which comprises one or two restriction sites unique to the poliovirus coding sequence.

5. A vector according to claim 4, vherein the or each restriction site is unique to the vector.

6. A vector according to claim 1, which is a plasmid.

7. A vector according to claim 6, which is pCASl or pCAS7.

8. A process for the preparation of a poliovirus chimaera which presents a foreign amino acid sequence, which process comprises:

(i) providing a double-stranded DNA fragment which encodes the amino acid sequence and which has a 5'-Sal X cohesive end and a 3'-blunt end;

(ii) digesting a cassette vector according to claim 1 with Sal l and Dra l and ligating the fragment from step (i) with the digested vector; and

(iii) obtaining live virus from the modified vector obtained in step (ii) .

9. A pharmaceutical formulation comprising a pharmaceutically acceptable carrier or diluent and a poliovirus chimaera prepared by a process as claimed in claim 8.

10. A cassette vector which comprises a coding sequence for a polypeptide and in which a pre-selected region of the coding sequence is defined by flanking 5' and 3' restriction sites which are unique to the vector, wherein the said pre-selected region is composed of such a nucleotide sequence that:

(b) the cassette vector is incapable of expressing the said polypeptide.

11. A vector according to claim 10, wherein the said nucleotide sequence results in a frameshift of the downstream polypeptide coding sequence and/or comprises a stop codon.

12. A vector according to claim 10, vherein the said nucleotide sequence comprises at least one restriction site unique to the polypeptide coding sequence.

13. A vector according to claim 11, wherein the or each said restriction site is unique to the vector.

14. A vector according to claim 10, vhich is a plasmid.

15. A process for the preparation of a recombinant vector, which process comprises: (i) providing a double-stranded DNA fragment vhich encodes a foreign amino acid sequence and vhich has 5' and 3' ends compatible with the restriction sites digested in step (ii) ; (ii) digesting a cassette vector according to claim 10 vith the restriction enzymes vhich cut at the said flanking 5' and 3' restriction sites of the cassette vector; and

(iii) ligating the fragment from step (i) vith the digested vector obtained in step (ii) .

16. A process for the preparation of a chimaeric protein, which process comprises introducing a recombinant vector prepared by a process as claimed in claim 15 into a host and enabling expression of the chimaeric protein encoded by the recombinant vector to occur.

17. A pharmaceutical composition comprising a pharmaceutically acceptable carrier or diluent and a chimaeric protein prepared by a process as claimed in claim 16.