1. Field of the invention The invention relates to a form of herpesvirus particle for use in vaccination and to vaccines containing it.

2. Description of the related art

In infectious virions of herpes simplex virus, the virus DNA is contained in an icosahedral nucleocapsid that is in turn enclosed by an amorphous proteinaceous tegument and a glycoprotein-containing envelope (Dargan, 1986).

During purification of a strain of herpes simplex virus type 1 (HSV-1) by velocity gradient centrifugation, two distinct particle bands were observed. In addition to the virion band, there was a band of lighter particles above it. These consist of tegument surrounded by an envelope but lack the viral capsid and DNA, and consequently are non-infectious. They have been termed "L-particles" (Szilagyi and Cunningham, 1991).

The appearance and composition of L-particles suggest that their genesis is related to that of virions although differences in the compositions of the two types of particle indicate that their pathways of assembly are not identical. L-particles offer potential for a non-infectious subunit vaccine but inability to produce them substantially free from contaminating virions was an obstacle and a prejudice against this line of research. Summary of the invention

We have examined the production of virus-related particles using ts 1201, a temperature-sensitive mutant of HSV-1 with a defect in gene UL26 (Preston et al., 1983, Preston e± ϋl. , 1991). At non-permissive temperatures, ts 1201 makes viral DNA and a full spectrum of viral proteins including the late structural species. Capsid assembly takes place at these temperatures but viral DNA is not packaged within the capsid and the empty capsids are retained in the cell nucleus (Preston __t al-, 1983). Thus, mature virions are not produced at the non-permissive temperature.

We have now found that the mutant ts 1201 produces L-particles at the non-permissive temperature in quantities similar to those generated by wild type virus. The ability of this mutant to assemble L-particles shows that their formation can take place without the involvement of capsids. Moreover, for ts 1201, and by inference for other mutants of wild type virus also, assembly of L-particles can be made independent of virion assembly, making it possible to isolate L-particles substantially free of contaminating virions and thereby enable their use as the active ingredient of a vaccine. This is surprising since it might reasonably have been supposed that the pathway to the production of L-particles would involve the ability of the virus-infected cell to produce virions. We have also found that L-particles are also produced by another strain of HSV-1 and by other herpesviruses.

In view of the above findings, the present invention provides virus-related (or virion-like), but non-infectious, L-particles of a herpesvirus, comprising tegument surrounded by an envelope, but which lack a capsid and the viral DNA within the capsid, for use in vaccination against herpesvirus infections.

The term "L-particles", as used herein, is not to be construed as indicative of any particular method used to produce them, but, rather, as a convenient label by which to refer to the particles defined above. Although they have been produced from a wild type virus and from a temperature-sensitive mutant, it will be appreciated that they will more conveniently be obtainable by reco binant DNA means. The findings explained herein indicate that by disabling the processes and signals responsible for the formation or release from the nucleus of a nucleocapsid structure it will be possible to produce

L-particles substantially free of infectious virions.

By using a recombinant herpesvirus containing DNA encoding a foreign protein or peptide, the L-particles can be prepared to contain such a foreign protein or peptide, thus enabling them to

be used in combatting other diseases than herpesvirus infections. The invention therefore also provides virus-related (or virion-like), but non-infectious, L-particles of a herpesvirus, said L-particles incorporating a foreign protein or peptide, both j_er e and for use in vaccination against any disease (which term includes any treatable medical condition) for which the foreign protein or peptide is an immunogen. It will be appreciated that the idea of using recombinants of herpesvirus DNA with foreign DNA was arrived at only after it had been found, by this invention, that the L-particles had any realisable potential for use in vaccines, more especially after it had been shown that the L-particles could be produced essentially free of virions from a mutant strain of HSV-1. Thus, it is considered that the idea did not flow from the Szilagyi and Cunningham (1991) paper.

The invention also includes a virus preparation based on or consisting of or including herpesvirus L-particles and substantially free of virions. The expression "substantially free of virions" indicates, of course, that the preparation has a greater freedom from virions than the L-particle band which is obtained by a separation of H-particles from L-particles, e.g. as in the Szilagyi and Cunningham (1991) paper. It has been found that the L-particle band is contaminated to the extent of an L-particle: virion ratio of 10 ^* ^:1 in that procedure, whereas in the present invention it would be normal to expect a ratio of at least 10 ⁴ :1 and preferably at least 10^:1. The invention includes such a virus preparation both per j_e and for use in a vaccine.

The invention also includes vaccine containing such L-particles or virus preparations of a herpesvirus. The vaccine can contain a vehicle or an immunostimulating agent such as an adjuvant.

Brief description of the drawings In the accompanying drawings:

Fig. 1 is a photograph showing virus particles separated on a gradient, for wild type (A) and the mutant ts 1201 (B); Fig. 2 is a photograph of a silver-stained polyacrylamide gel showing polypeptide bands obtained by PAGE from the gradient-separated viral particles; and

Fig. 3 is a plasmid diagram showing the derivation of a plasmid carrying foreign DNA for insertion into herpesvirus DNA, for preparing L-particles incorporating a foreign protein (see Example 2). Description of the preferred embodiments

When L-particles were first identified, it was not known whether they were peculiar to HSV-1 or to the strain of HSV-1 used, namely strain 17, or of more widespread occurrence. It has now been shown that these particles are also produced by a different HSV-1 strain, Strain F (Ejercito fit al-, 1968) and other, distantly related α-herpesviruses, namely pseudorabies virus strain Ka (Kaplan and Vatter, 1959) and equine herpesvirus 1 strain AB4 (obtained from E. Telford, MRC Virology Unit, Institute of Virology, Glasgow Gil 5JR, Scotland). The diverse nature of these viruses makes it very highly probable that L-particles will be generally obtainable from α-type herpesviruses and from other types of herpesvirus. The present invention makes it possible to produce L-particles substantially free of infectious virions, by using a virus which has undergone an appropriate genetic disablement. In a preferred embodiment, cells are infected with a conditional lethal mutant under conditions non-permissive for viral growth. The term "conditional lethal mutant" refers to a mutant lethal to virus growth under certain conditions. Temperature is a common example of such a condition, cells being infected with a temperature-sensitive mutant, e.g. ts 1201, at a non-permissive temperature. The non-permissive temperature is any temperature at which the wild type virus grows, but the mutant does not and

is usually in the 38-40°C range. The infected cells are incubated for a period sufficient to allow viral DNA replication to take place and late-translated proteins to be produced, thus allowing L-particles to be assembled and accumulate. The period required might be typically 20 hours or longer, but will vary according to the nature of the mutation.

Many other temperature-sensitive mutants can be produced and tested for effects similar to those described in the above-cited literature. Many mutants are "leaky", in the sense that they produce virions at some temperatures within the non-permissive temperature range. It is, therefore, necessary to optimise the temperature to mitigate contamination by virions, but this is a routine matter for those skilled in the art.

An alternative condition to temperature sensitivity which would prevent virion formation is the creation of a virus from which an essential gene has been deleted. For example, if a capsid protein were deleted from the virus genome, then such a virus could only be grown on genetically engineered cell lines which express this deleted capsid protein. Under these conditions, the cell line will provide the capsid protein essential for virion production. When the virus grown on these cell lines is used to infect cells which do not express the capsid protein, no virion production would take place. However, L-particles would be made, since capsid proteins are not an essential component of their production. Under these conditions, the deletion would be lethal to virus growth. There are other more sophisticated means of creating conditional lethal mutants and the term covers many ways of preventing virion production. Any of the genes responsible for the formation of a full capsid can be appropriately disabled, e.g. by deletion, mutation or insertion of nucleotide(s). Generally, a deletion will be preferred since its effects are more readily predictable, but many temperature-sensitive mutants already exist and it is likely that some of these will also suffice to produce L-particles substantially free of infectious virions.

HSV-1 mutant ts 1201 has a mutation in a gene UL26, which codes for a protein involved in DNA packaging. Other genes of HSV-1 which could be disabled to produce L-particles substantially free of virions comprise: UL6, a gene encoding a late protein involved in DNA packaging. Mutations of tsN (Matz fit ____., 1983), tsF18 and tsF43, (Wel er fit al. , 1987) are in this gene.

UL12, another gene encoding a protein, alkaline nuclease, involved in DNA packaging (Weller fit al-, 1990). UL18, a capsid protein gene (Rixon fii l., 1990).

UL19, a capsid protein gene. Mutations of tsG3 and tsG8 (Weller fit ϋl., 1987) are in this gene.

UL28, a gene encoding a protein involved in DNA packaging. A mutation of tsl203 (Addison fit al. , 1990) is in this gene. UL33, another gene involved in producing a full capsid (Al-Kobaisi fit al., 1991).

UL38, a capsid protein gene (Pertuiset fit al-, 1989); Rixon fital-. 1990 ⁾ .

In virus preparations from wild type virus, the numerical ratio of L-particles to infectious virions in the L-particle band would normally be about 10 ³ :1. Using a conditional lethal mutant as proposed above, this ratio would normally be at least 10 ⁴ :1 and preferably lie within the range of 10 ⁴ to 10 ¹² :1, typically being about 10 ⁶ :1 for the mutant ts 1201. Such virus preparations are included in the invention.

Even a small proportion of virions is undesired in most intended uses of vaccination, so in the final commercial vaccine the L-particles might be produced by recombinant DNA technology. Thus, the entire herpesvirus genome, as one or a small number of DNA fragments and appropriately disabled to preclude virion formation, will be cloned and expressed to produce L-particles 'artificially', thus avoiding even slight contamination by virions.

Two methods for cloning HSV DNA to produce L-particles free from virions will now be mentioned by way of illustration. The

first of these uses a bacteriophage cloning system (called the PI cloning system) which allows cloning of the HSV genome as two large pieces of DNA (>50 kilobase pairs). The ability to clone such large segments of DNA is particularly advantageous since most other cloning methods only allow shorter DNA fragments to be cloned. When these large fragments are cloned using the PI system, they will be amenable to manipulation using standard methods. The contemplated types of manipulation of the cloned DNA are those which would prevent virion assembly. They include deletion of one or more structural genes. In order to form L-particles without forming virions, it is expected that it will be sufficient to prevent the formation of a full capsid. A capsid lacking essential structural components will be unable to create virions. Thus, one example of a manipulation is deletion of a capsid protein. Alternatively, the signals located within the genome which are required for DNA packaging could be removed, again preventing production of mature virions. In order to make L-particles, these cloned DNA fragments would be transfected into mammalian cells and recently developed methods give high levels of transfection efficiency.

A second method is to insert signals from yeast DNA into the HSV genome and in this way create a 'yeast artificial chromosome' (YAC). Following transfection into yeast cells, YACs behave in the same way as normal yeast chromosomes and are therefore inherited by daughter cells after cell division. Again the HSV genome would be manipulated in such a way that only L-particles and not virions would be made. DNA made from the yeast cells containing the HSV YAC would then be transfected into mammalian cells and the production of L-particles would be assayed. The production of L-particles in a virus-free system using these particular recombinant DNA techniques is a preferred feature of the present invention.

The L-particles are defined herein as having tegument and outer envelope, but lacking capsid and viral DNA. It is expected that L-particles will be able to stimulate the immune

responses elicited by virions. They will thus be able to stimulate a humoral response. Furthermore, we have demonstrated experimentally that L-particles will bind to host cells, fuse with those cells and release their contents. In this way they will assist in the acquisition of cell-mediated immunity. (The experimental demonstration involved studying the properties of two tegument proteins, namely (1) the HSV-1 or HSV-2 virion host shut-off (vhs) protein and (2) the transcriptional activator α-TIF (V w 65). For these experiments, virions and L-particles were prepared from wild type HSV-1 and HSV-1 mutants. L-particles were found to be as efficient as virions in supplying these proteins in a fully functional state.

These immune responses are expected to provide protection against the type of herpesvirus from which the L-particles are derived. It is intended that the L-particles will provide protection against a herpesvirus infection of the strain which provides the relevant proteins of the L-particles, but there will doubtless, in many cases, be a degree of protection against other strains within a serotype or even against strains of a different serotype.

While the invention thus far described relates to L-particles wholly composed of elements of a single strain of herpesvirus, a wider range of protection could be provided by producing virus recombinants which would have the ability to express foreign DNA in such a way as to incorporate proteins or epitopic peptides from other viral types or other strains of herpesvirus into L-particles. For example, varicella-zoster virus DNA can be inserted in HSV-1 DNA to produce particles containing VZV proteins. Indeed, the foreign (non-native or heterologous) DNA in such constructs could be non-herpesvirus DNA encoding foreign protein(s) or peptide(s). (The term "foreign" herein means foreign to the strain of herpesvirus from which the L-particles are derived).

Where the foreign DNA is foreign only in the sense that it is derived from another strain or type of HSV, it is expected

that a recombinant HSV containing the foreign DNA will express it without difficulty in most cases, so that the expressed protein becomes incorporated within the L-particles, (e.g. in the envelope or in the tegument). Thus, for example, L-particles containing a foreign protein in their structure can be prepared by co-transfecting appropriate cells with DNA comprising (1) DNA of HSV-1 , preferably of a conditional lethal mutant, and (2) a plasmid containing (a) HSV-1 DNA which is inessential to viral growth (so-called non-essential region [NER]), interrupted by (b) the foreign DNA. The cells can be any appropriate to vaccine use and approved by regulatory authorities. They may be, for example, baby hamster or monkey kidney cells. The transfected cells are incubated to grow the virus, and the virus is then purified by selecting plaques and re-growing them. The viral culture is then incubated at the appropriate non-permissive condition, preferably for at least 24 hours, to produce L-particles containing foreign protein or a peptide expressed from the recombinant herpesvirus DNA, substantially free of virions. (Alternatively, wild type virus could be used and H-particles separated from L-particles, but this is not preferred). The supernatant medium is separated from the cellular material as completely as possible and provides a suspension of L-particles. The above description applies mutant mutandis to other starting herpesviruses. Considerable interest centres on the expression of foreign DNA which is not native to HSV, e.g. of another herpesvirus such as human cytomegalovirus (HCMV) or varicella zoster virus (VZV), or of another family of viruses, e.g. the immunodeficiency virus group. It may be necessary to incorporate the foreign DNA into a region of the herpesvirus genome which will provide signals by which the protein expressed is targeted to the L-particle. This will be a matter of trial and error, but it has already been shown in Example 2 hereinafter that the virion host shut-off (vhs) protein gene of HSV-1 is capable of providing the requisite signals. Experimentation with marker genes along the

lines indicated in Example 2 will readily provide appropriate locations for the construction of recombinants. The same procedure of homologous recombination, as described above, can be used to make the recombinant. In appropriate cases, the foreign DNA could instead be introduced in the herpesvirus DNA by direct ligation.

Examples of the foreign proteins and peptides or heterologous antigens which can be introduced into L-particles by procedures as described above are any HSV structural protein, or combination of structural proteins, such as gD, gB, immediate-early 183, Vmw 65, and proteins of other herpesviruses such as HCMV gB, and human immunodeficiency virus proteins or peptides, such as of HIV-1 or HIV-2 gp 120 or gp 160.

The vaccine of the invention may be of any formulation by which L-particles can stimulate formation of antibodies to the relevant proteins. The vaccine will therefore normally contain an immunostimulant (e.g. adjuvant) or vehicle, as well as the L-particles. It can, of course, also contain other conventional additives, such as excipients. The invention includes a live L-particle vaccine, as well as a vaccine of L-particles which have been inactivated, e.g. by formaldehyde or irradiation. Particularly for a live vaccine, it is suggested that whole L-particles are used. If the vaccine is inactivated the L-particles could be used whole or possibly in a disrupted or comminuted form if it is desired to release proteins from the tegument. The vaccine is desirably made up into unit dosage forms. Appropriate doses can be derived from knowledge of the use of herpesvirus vaccines, e.g. of HSV-1 or varicella-zoster virus, but the dose will also depend on whether the L-particles are intended to protect against a herpesvirus infection or are the product of vectors by which an immunogenic foreign protein is produced. In the latter case, dosages appropriate to the foreign protein will have to be calculated.

For administration of the L-particles to the appropriate mammals, any of the conventional routes used in the viral

vaccine field can be used. These will be predominantly subcutaneous or intramuscular injection, but other routes, e.g. oral, intravenal, intravaginal and intranasal, may be more appropriate on occasions. They can be administered for the prophylaxis of herpesvirus or, if foreign DNA is incorporated in the herpesvirus DNA, other diseases to which a protein or peptide expressed from the foreign DNA is immunoprotective or serves to stimulate the immune system to increase an immunoprotective effect. In particular, the invention is expected to be useful in the prophylaxis of AIDS in HIV-susceptible or HIV-positive patients.

Although the invention applies predominantly to human patients, it is also applicable, for example, to diseases in animals, e.g. to equine herpesviruses, and to the L-particles of such a virus for administration to horses.

The following Examples illustrate the invention. "Ficoll" is a Registered Trade Mark. EXAMPLE 1

This Example shows how L-particles can be obtained from a mutant virus, without being accompanied by significant production of infectious virions. METHODS

Preparation of a temperature-sensitive mutant of herpes simplex virus type I. designated ts 1201. A mutant 17tsJC116 was isolated following UV-mutagenesis of HSV-1 strain 17 by Coates (1982). An EcoRI F fragment from a clone of DNA from this mutant was recombined into HSV-1 strain 17 wild type (WT) virus to produce mutant ts 1201. (Preston fit al., 1983 ⁾ . Identification of a mutation

A BamHI/Sall fragment containing the ts 1201 mutation was sequenced and a single base-pair change from the wild type DNA sequence was identified. The change was from an A residue to a

T residue at position 50897 on the sequence of McGeoch et al., (1988). This change results in the substitution of a tyrosine

with a phenylalanine in the amino acid sequence of the UL26 protein (Al-Kobaisi, 1989). Patent deposit of ts 1201

While it is likely that this mutation is sufficient to produce the requisite genomic change to prevent capsid formation and thus generate L-particles, and therefore that the L-particles of this mutant of HSV-1 can be produced by recombinant DNA means, ts 1201 has been deposited, purely by way of precaution, to ensure that the invention can be even more readily carried out. The deposit is under the Budapest Treaty on the deposit of micro-organisms for the purposes of patent procedure at the European Collection of Animal Cell Cultures, PHLS Centre for Applied Microbiology and Research, Porton Down, Salisbury, Wiltshire SP4 OJG, England on 20th June 1991 under Accession No. V91062004.

Growth of viruses and purification of virions

BHK C13 cells were grown as described previously (Rixon and McLauchlan, 1990). Stocks of HSV-1 strain 17 and ts 1201 were obtained by infecting 10 roller bottles of cells at a moi of 1/300 pfu per cell. Following infection at 31°C for 4 days, virus was harvested and purified on 5-15% Ficol1 gradients as described previously in Example 1 above. Bands were collected by side puncture, diluted with Eagle's medium lacking phenol red and pelletted by centrifugation at 21,000 rpm in a Sorvall Tst41 rotor for 2h at 4°C. Pellets were resuspended in Eagle's medium lacking phenol red and stored at -70°C. Titration of purified virions

Purified virion preparations of WT and ts 1201 were titred at 31°C (permissive) and 38.5°C (non-permissive temperature for ts 1201) and their particle/pfu ratios were determined at 31°C to be of the same order of magnitude. Titration of the ts 1201 virion stock confirmed that it was temperature-sensitive under the conditions used, with a reduction in titre over WT of greater than 10^.

Electron Microscopy

Cells to be used for electron microscopy were grown in 30mm petri dishes containing a 13mm glass coverslip. At appropriate times after infection the cells were washed with phosphate buffered saline (PBS) and fixed with 2.5% glutaraldehyde in PBS for lh at 4°C. Cells were then washed in PBS and osmium tetroxide (1% in PBS) was added for lh. Replica Formation

The coverslips were removed to a 24 well tissue culture tray, dehydrated through a graded alcohol series and dried in a critical point drier. The coverslips were shadowed with Pt/Pd for 7s at an angle of 75° in a Balzers shadowing unit, then rotary shadowed with carbon to provide additional support. The shadowed surface was overlayed with 50μl of 0.25% parlodion in amyl acetate and allowed to dry. Each coverslip was broken in half and the shadowed surface was scored into approximately 2mm squares. To release replicas, the coverslips were floated on hydrofluoric acid until they dissolved. The replica squares were picked up on a Pt wire loop and washed on distilled water. Cell material was removed by floating replicas on 66% chromic acid for lh. Clean replicas were washed 3 times on water, then dried onto 400 mesh copper microscope grids. When dry, the parlodion support film was removed by immersing the grids in amyl acetate and the replicas were examined in a JE0L 100S electron microscope.

Thin Sectioning

The remaining cells on each 30mm petri dish were harvested, pelleted in BEEM capsules, dehydrated and embedded in Epon 812 resin as described previously (Preston fit al- , 1983). 130nm thick sections were cut and stained with uranyl acetate and lead citrate. Polvacryl mide Gel Electrophoresis

Proteins were separated on a 5-12% gradient polyacrylamide gel which had been crosslinked with 5% N.N'-methylene bisacrylamide, using the buffer system described by Laemmli

(1970). Proteins were visualised by silver staining as described by McLean fit al- , (1990). Production of L-particles

Five roller bottles of cells were infected with either gradient-purified ts 1201 or WT virus. Following incubation at 38.5°C for 24h, particulate matter present in the supernatant media was pelletted, resuspended and banded on 5-15% Ficoll gradients as described by Szilagyi and Cunningham (1991). RESULTS a. Sampling

Cells were grown on 30mm petri dishes which each contained a 13mm glass cover slip. Following infection with 5 pfu/cell of either WT virus or ts 1201, cells were incubated at 38.5°C. A portion of each inoculum was retained for subsequent titration ⁽ these constituted the zero hour samples). At lh after infection, the input inocula were removed and were also retained ⁽ these constituted the 1 hour samples). The cells were washed with ETC10 and incubation was continued at 38.5°C. At 6, 10 and 24h,, supernatant medium was collected from each plate and retained for titration. The cells from each time point were used for formation of surface replicas, while the remaining cells on the dish were prepared for thin sectioning. b. Virus growth

As expected, the WT virus, after an initial decline in titre during the eclipse phase, showed a rapid increase in virus production to a level about 10-fold higher than the input virus. By contrast, ts 1201 titres declined throughout the period of infection to about 100-fold below that of the input virus, thereby confirming that the ts 1201 used in this experiment was behaving in a fashion typical of this mutant. This ensured that the results obtained from the ts 1201-infected cells were not a result of leak or reversion of the mutant. c. Appearance of particles on the cell surface

Examination of replicas prepared from WT-infected cells revealed a characteristic series of changes to the cell

surface. At lh the cell surface was smooth and showed no evidence of infection compared with mock-infected samples. By 6h, many cells had small numbers of approximately 250nm diameter particles on their surfaces and by lOh large numbers of these particles had accumulated on most cells. The sizes of these particles were consistent with those of virions. At 24h the cells were decidedly rounded and had dense coverings of particles. The distribution of these particles on the cell surface was frequently assymetric with marked concentrations over the area of the nucleus and around regions at the periphery of the cell being typical.

Examination of ts 1201-infected cells revealed a very similar pattern of changes. Thus at lh after infection their surfaces were indistinguishable from those of mock-infected cells apart from the presence of occasional particles which presumably were derived from the inoculum. By 6h, small numbers of particles were present on the cell surfaces and their number had increased at lOh and 24h. The distribution and abundance of particles were comparable to those found with WT virus but they appeared slightly less uniform in size. d. Nature of the particles on the cell surface

To determine the nature of the particles present on the surfaces of infected cells, thin sections prepared from the remainder of each sample were examined. Both virions and particles lacking obvious capsids, which we consider to be

L-particles, were present on WT-infected cells. Similar particles were also found with virions in intracellular vacuoles.

Thin sections of ts 1201-infected cells demonstrated the phenotype typical of this mutant (Preston e_t al-, 1983). Thus, the nuclei contained aggregations of characteristically large cored capsids which lacked DNA. Full, DNA-containing, capsids were not apparent and there was no evidence of mature virions either in the cytoplasm or on the cell surface. The particles that were present on the surface lacked capsids and closely resembled L-particles in appearance. Particles of a similar

type were also frequently seen in cytoplasmic vacuoles. A number of features were also seen which could be considered as L-particles in the process of formation by budding into vacuoles. However, due to the lack of diagnostic capsid, it was impossible to be dogmatic as to the nature of these features. Identification and protein composition of particles released from ts 1201-infected cells

Referring to Fig. 1 of the drawings, the photograph shows bands of pelleted particulate material separated on the gradient, with bands L ascribed to light particles and V to virions. Tube A contains the wild type and B the mutant. The gradient containing the WT virus preparation (Figure 1, lane A) had two bands which corresponded to those described in Example 1: a sharp lower band of virions and a diffuse upper band of L-particles. By contrast, the virion band could not be observed in the ts 1201 gradient (Figure 1, lane B). However, a diffuse band was present which co-migrated with, and was of a similar intensity to, the L-particle band present in the WT virus preparation. When* collected and examined by negative staining, this band was shown to contain material which appeared identical to typical L-particles.

Analysis of the polypeptide composition of the gradient- banded material by PAGE confirmed the above assignment. Referring to Fig. 2 of the drawings, lanes 1 and 2 show the polypeptide profiles of virions and L-particles, respectively, from the WT virus, while lane 3 shows the composition of the particles produced by ts-1201 at the non-permissive temperature. The positions of bands for HSV-encoded proteins V w 175 (diagnostic of L-particles) and Vmw 37 and 155 (diagnostic of virions) are shown. (The Vmw numbers indicate relative molecular masses in kiloDaltons). The compositions of the WT virions and L-particles conformed to the previously described patterns (Example 1). Thus, Vmw 155 (Mr 155000) and Vmw 37 (Mr 37000) capsid proteins, which were abundant in virions, were greatly reduced in L-particles, while Vmwl75 was

apparent only in L-particles. The protein profile of the ts 1201 band was virtually indistinguishible from that of the WT L-particles and the diagnostic Vmwl75 band was clearly visible (Figure 2, lane 3). Interestingly, despite the very low levels of infectious virions in this band, the Vmw 155 (Mr 155000) capsid protein was present in quantities similar to those found in WT L-particles. PREPARATION OF A VACCINE

Although virions are still produced at low levels by ts 1201- infected cells, the numerical ratio of L-particles to virions is typically about 10^:1, compared with about 10 ³ :1 in

WT virion preparations. For the preparation of a vaccine it will be advisable to remove all residual possibility of herpesvirus infection. This can be done by inactivating the preparation, e.g. with formaldehyde or by UV irradiation. A dose of 160 mJ/cm ² of UV light of wavelength 254 nm delivered by a UV Stratalinker Model 180 would normally suffice and probably this dose could be lowered. However, the possibility of using a live vaccine is not excluded, since it is likely that methods of further purifying the L-particles will soon be devised.

EXAMPLE 2

This Example demonstrates incorporation of a foreign gene into HSV-1 and expressing it to produce foreign protein as part of a fusion protein in L-particles. Thus, the L-particles are here serving as a vector. In this Example, we have linked the polypeptide coding sequences of a structural component of the virus (in this case a protein called the virion host shut-off protein, abbreviated to vhs) encoded by gene UL41 to those of a non-structural marker protein (a bacterial enzyme called chloramphenicol acetyltransferase, CAT for short). We have then inserted this chimeric gene into the virus genome and looked for the presence of a vhs-CAT fusion protein in virions and L-particles.

The plasmid which was used as the basis for constructing a vhs-CAT fusion gene was pMFl . pMFl consists of a 3.7kb fragment

containing the vhs gene from HSV-2 strain G inserted into a plasmid, pTKl that contained the thymidine kinase (TK) gene from HSV-1 strain 17 inserted into the well known pAT 153 plasmid. This 3.7kb fragment was inserted into the TK sequence in pTKl in such a way as to inactivate the TK enzyme. In order to link the CAT coding sequences to the HSV-2 vhs coding sequences, the following strategy (illustrated schematically in Fig. 3) was employed:

1) pMFl contains a unique Hindlll restriction enzyme site and this was removed from the plasmid by standard gene manipulation procedures. Thus the resultant plasmid (called pVHSl) did not contain any Hindlll sites.

2) An Aatll restriction enzyme site is located just upstream from the 3' end of the vhs protein coding sequences. A synthetic oligonucleotide was inserted at this Aatll site in pVHSl to create pVHS2. The oligonucleotide has two essential features. Firstly, it contains the sequences between the Aatll site and the 3' end of the vhs open reading frame (ORF), which are present in pVHSl , and hence it will regenerate the complete ORF. Secondly, a Hindlll site is present immediately after the codon specifying the C-terminal amino acid of vhs. This Hindlll site was incorporated into the oligonucleotide in such a way that the CAT coding sequences inserted at this site (see step 3) would be in the same reading frame as the vhs gene. 3) A Hindlll fragment from a plasmid pCATl carrying the CAT coding sequences was then inserted into the Hindlll site in pVHS2 to generate plasmid pVHS6. Thus, the vhs gene in pVHS6 was intact and in-frame with the coding sequences for CAT.

To insert this vhs-CAT chimeric gene into the virus genome, pVHS6 was recombined with HSV-1 strain 17 DNA in tissue culture cells. Since the DNA fragment carrying the vhs-CAT fusion was inserted into the HSV-1 TK polypeptide coding sequence, the incorporation of the vhs-CAT sequences into the virus genome should inactivate the endogenous virus TK gene, resulting in the creation of a TK minus virus. Such viruses can be

preferentially selected by growing the virus population generated from the recombination procedure in the presence of bromodeoxycytidine (BCdR). Therefore, after the recombination had been performed, the resultant pool of viruses was grown twice in the presence of BCdR and thereafter plaques from individual viruses were picked and grown. These procedures, including the TK- selection, are standard methods for producing and purifying recombinant viruses. One of the individual plaques which had been picked and grown was shown to contain CAT DNA sequences that had been inserted in the TK gene in the virus DNA. A large scale preparation of this virus was grown and called vhs-CAT. The virus preparation was banded on Ficoll gradients and virions and L-particles were purified as in Example 1. The proteins present in the virions and L-particles were separated on a 7.5% polyacrylamide gel alongside a L-particle preparation from a wild-type virus which did not contain any CAT sequences. After transfer to a nitrocellulose membrane, the membrane was probed for the presence of the vhs-CAT fusion protein. This antibody is made by 5 Prime -> 3 Prime Inc., 5603 Arapahoe, Boulder, Colorado CO 80303, USA. An autoradiograph showed that this antibody detected a protein of ^• approximately 85,000 to 90,000 molecular weight in the vhs-CAT virions and L-particles but not in the CAT minus, wild-type virus. The size of this protein fits with the predicted size of the vhs-CAT fusion polypeptide (vhs gene ORF 492 codons = 59 kDa; CAT gene ORF 219 codons = 24 kDa; linker between vhs and CAT gene ORFs 14 codons = 1.6 kDa; total = 85 kDa approximately). An additional band appeared only in the vhs-CAT virion sample at a position indicating a m.w. of about 155,000. This is ascribed to a cross-reaction between the CAT antibody and the major capsid protein Vmw 155 present in virions, but not in L-particles, and as such is not significant.

In addition, extracts from cells infected with vhs-CAT and from purified vhs-CAT virions and L-particles were tested for CAT activity in a simple assay. The results show that the

vhs-CAT fusion protein retains CAT activity. To date, therefore, preliminary analyses of the vhs-CAT virus have revealed that the vhs-CAT fusion protein is incorporated into virions and L-particles and the CAT portion of the protein remains enzymically active. This shows that it is possible to insert heterologous, non-structural proteins into virions and L-particles by linking them to known structural genes.

CAT was chosen as an appropriate gene for these studies since it is easily detected and is a bacterial gene with no counterpart in animal cells. However, similar procedures could be used to introduce other proteins, including those significant for vaccine purposes, into L-particles.

REFERENCES ADDISON, C, RIXON, F. 3. and PRESTON, V. G. (1990). Herpes simplex virus type 1 UL28 gene product is important for the formation of mature capsids. 3. Gen. Virol. 21, 2377-2384. AL-KOBAISI, M. F. (1989). Identification and characterisation of herpes simplex virus genes required for encapsidation of DNA. PhD thesis, Glasgow University.

AL-KOBAISI, M. F., RIXON, F. 0., McDOUGALL, I. and PRESTON, V. G. ⁽ 1991). The herpes simplex virus UL33 gene product is required for assembly of full capsids. Virology 180. 380-388.

BROWN, S. M., RITCHIE, D. A. and SUBAK-SHARPE, J. H.

⁽ 1973 ⁾ . Genetic studies with herpes simplex virus type 1. The isolation of temperature-sensitive mutants, their arrangement into complementation groups and recombination analysis leading to a linkage map. 3. Gen. Virol, 1_J, 329-346.

COATES, 3. A. V. (1982). Genetic and biochemical studies with temperature-sensitive mutants of herpes simplex virus type 1. PhD thesis, Glasgow University. DARGAN, D. 3. (1986). The structure and assembly of herpesviruses. In Electron Microscopy of Proteins, vol. 5, Virus Structure, pp. 359-437. Edited by 3. R. Harris and R. W. Home. London: Academic Press.

EJERCIT0, P. M., KIEFF, E. D. and ROIZMAN, B. (1968). Characterisation of herpes simplex virus strains differing in their effect on social behaviour of infected cells. 3. Gen. Virol. I, 357-364.

IRMIERE, A. and GIBSON, W. (1983). Isolation and characterisation of a non-infectious virion-like particle released from cells infected with human strains of cytomegalovirus. Virology JJO, 118-133.

KAPLAN, A. S. and VATTER, A. E. (1959). A comparison of herpes simplex and pseudorabies viruses. Virology 7, 394-407. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature,

London 227, 680-685.

LOWRY, 0. H., R0SEBR0UGH, N. J., FARR, A. L. and RANDALL, R. J., (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem.123, 265-275. MATZ, B., SUBAK-SHARPE, J. H. and PRESTON, V. G. (1983). Physical mapping of temperature-sensitive mutations of herpes simplex virus type 1 using cloned restriction endonuclease fragments. J. Gen. Virol. 64, 2261-2270.

MCGEOCH, D. J., DALRYMPLE, M. A., DAVISON, A. J., DOLAN, A., FRAME, M. C. , MCNAB, D., PERRY, L. J., SCOTT, J. E. , and

TAYLOR, P. (1988). The complete DNA sequence of the long unique region in the genome of herpes simplex virus type 1. J. Gen.

Virol. ____, 1531-1574.

McLEAN, G., RIXON, F. J., LANGELAND, N., HAARR, L., and MARSDEN, H. (1990). Identification and characterisation of the virion protein products of herpes simplex virus type 1 gene UL47. J. Gen. Virol., H, 2953-2960.

ONGRADI, J., CUNNINGHAM, C. and SZILAGYI, J. F. (1985). The role of polypeptides L and NS in the transcription process of vesicular stomatitus virus New Jersey using the temperature- sensitive mutant tsEl. J. Gen. Virol. __6, 1011-1023.

PERTUISET, B., BOCCARA, M., CEBRIAN, J., BERTHELOT, N.,

CHOUSTERMAN, S., PUVION-DUTILLEUL, F. , SISMAN, J. and

SHELDRICK, P. (1989). Physical mapping and nucleotide sequence of a herpes simplex virus type 1 gene required for capsid assembly. J. Virol. __2, 2169-2179.

PRESTON, V. G., C0ATES, J. A. V., and RIXON, F. J. (1983). Identification and characterisation of a herpes simplex virus gene product required for encapisidation of virus DNA. J. Virol. 4_>, 1056-1064.

PRESTON, V. G., RIXON, F. J., McDOUGALL, I. M., McGREGOR, M. and AL K0BAISI, M. F. (1992). Processing of the Herpes Simplex

Virus assembly protein ICP 35 near its carboxy terminal end requires the product of the whole of the UL26 reading frame. Virology ]_i. 87-92.

RIXON, F. J., DAVISON, M. D. and DAVISON, A. J. (1990). Identification of the genes encoding two capsid proteins of herpes simplex virus type 1 by direct amino acid sequencing. 3. Gen. Virol. 21, 1211-1214. SZILAGYI, J. F. and CUNNINGHAM, C. (1991). Identification and characterisation of a novel non-infectious herpes simplex virus-related particle. J. Gen. Virol., 22, 661-668. Publication date: 1st March 1991.

WELLER, S. K., CARMICHAEL, E. P., ASCHMAN, D. P., GOLDSTEIN, D. J. and SCHAFFER, P. A. (1987). Genetic and phenotypic characterisation of mutants in four essential genes that map to the left half of HSV-1 U _L DNA. Virology lfil, 198-210.

WELLER, S. K., SEGHATOLESLAMI, M. R., SHAO, L., ROWSE, D. and CARMICHAEL, E. P. (1990). The herpes simplex virus type 1 alkaline nuclease is not essential for viral DNA synthesis: isolation and characterisation of a lacZ insertion mutant. J.

Gen. Virol.21, 2941-2952.

-2Λ-