AU2004203141B2 - Methods and reagents for vaccination which generate CD8 T cell immune response - Google Patents

Description

P00ool Section 29 Regulation 3.2(2)

AUSTRALIA

Patents Act 1990 COMPLETE SPECIFICATION STANDARD PATENT Application Number: Lodged: Invention Title: Methods and reagents for vaccination which generate a CD8 T cell immune response The following statement is a full description of this invention, including the best method of performing it known to us: METHODS AND REAGENTS FOR VACCINATION WHICH GENERATE A CD8 T CELL IMMUNE RESPONSE This invention relates to generation of a protective CD8+ T cell immune response against target antigens using different primer and s booster compositions as sources of CD8+ T cell epitopes.

Introduction A general problem in vaccinology has been an inability to generate high levels of CD8 T cells by immunisation. This has impeded io the development of vaccines against several diseases including malaria.

Plasmodium falciparum malaria causes hundreds of millions of malaria infections each year and is responsible for 1-2 million deaths annually. The development of an effective vaccine against malaria is thus a major, priority for global public health. A considerable body of immunological research over the last twenty years had led to the identification both of candidate vaccine antigens from the parasite and immunological mechanisms on the host that are likely to protect against infection and disease. However, despite this progress there is still no means of vaccinating against malaria infection which has been shown to be effective in field trials.

A major problem has been the identification of a means of inducing a sufficiently strong immune response in vaccinated individuals to protect against infection and disease. So, although many malaria antigens are known that might be useful in vaccinating against malaria the problem has been how to deliver such antigens or fragments of them known as epitopes, which are recognised by cells of the immune system, in a way that induces a sufficiently strong immune response of a particular type.

It has been known for many years that it is possible to protect individuals by immunising them with very large doses of irradiated malaria sporozoites given by bites from infected mosquitoes. Although this is a wholly impractical method of mass vaccination it has provided a model for analysing the immune responses that might be mediating protective immunity against sporozoite infection (Nardin and Nussenzweig 1993).

A considerable amount of research over the last decade or more has indicated that a major protective immune response against the early pre-erythrocytic stage of P. falciparum malaria is mediated by T lymphocytes of the CD8+ ve type (CD8+ T cells). Such cells have been shown to mediate protection directly in mouse models of malaria infection (Nardin and Nussenzweig 1993). Such T cells have also been identified in io individuals naturally exposed to malaria and in volunteers immunised with irradiated sporozoites (Hill et a. 1991; Aidoo et al. 1995; Wizel et al. 1995).

There is much indirect evidence that such CD8+ T cells are protective against malaria infection and disease in humans (Lalvani et al. 1994).

CD8+ T cells may function in more than one way. The best known function is the killing or lysis of target cells bearing peptide antigen in the context of an MHC class I molecule. Hence these cells are often termed cytotoxic T lymphocytes (CTL). However, another function, perhaps of greater protective relevance in malaria infections is the ability of CD8+ T cells to secrete interferon gamma (IFN-y). Thus assays of lytic activity and of IFN-y release are both of value in measuring a CD8+ T cell immune response. In malaria these CD8+ve cells can protect by killing the parasite at the early intrahepatic stage of malaria infection before any symptoms of disease are produced (Seguin et al. 1994).

The agent of fatal human malaria, P. falciparum infects a restricted number of host species: humans, chimpanzees and some species of New World monkey. The best non-human model of malaria is the chimpanzee because this species is closely related to humans and liver-stage infection is observed consistently unlike in the monkey hosts (Thomas et al. 1994). Because of the expense and limited availability of chimpanzees most laboratory studies of malaria are performed in mice, using the rodent malaria species P. berghei or P. yoelii. These latter two models are well studied and it has been shown in both that CD8+ve lymphocytes play a key role in protective immunity against sporozoite challenge.

s Previous studies have assessed a large variety of means of inducing CD8+ T cell responses against malaria. Several of these have shown some level of CD8+ T cell response and partial protection against malaria infection in the rodent models Li et al. 1993; Sedegah et al 1994; Lanar et al. 1996). However, an effective means of immunising with subunit vaccines by the induction of sufficiently high levels of CD8+ T lymphocytes to protect effectively against malaria sporozoite infection has not previously been demonstrated.

In recent years improved immune responses generated to potential vaccines have been sought by varying the vectors used to deliver Is the antigen. There is evidence that in some instances antibody responses j are improved by using two different vectors administered sequentially as prime and boost. A variety of combinations of prime and boost have been tested in different potential vaccine regimes.

Leong et al. (Vaccines 1995, 327-331) describe immunising mice firstly to DNA expressing the influenza haemagglutinin (HA) antigen and then with a recombinant fowlpox vector expressing HA. An enhanced antibody response was obtained following boosting.

Richmond et al. (Virology 1997, 230: 265-274) describe attempts to raise neutralising antibodies against HIV-1 env using DNA priming and recombinant vaccinia virus boosting. Only low levels of antibody responses were observed with this prime boost regime and the results were considered disappointing.

Fuller et al. (Vaccine 1997, 15:924-926 and Immunol Cell Biol 1997, 75:389-396) describe an enhancement of antibody responses to DNA immunisation of macaques by using a booster immunisation with replicating recombinant vaccinia viruses. However, this did not translate into enhanced protective efficacy as a greater reduction in viral burden and attenuation of CD4 T cell loss was seen in the DNA primed and boosted animals.

s Hodge et al (Vaccine 1997, 15: 759-768) describe the induction of lymphoproliferative T cell responses in a mouse model for cancer using human carcinoembryonic antigen (CEA) expressed in a recombinant fowl pox virus (ALVAC). The authors primed an immune response with CEA-recombinant replication competent vaccinia viruses of to the Wyeth or WR strain and boosted the response with CEA-recombinant ALVAC. This led to an increase in T cell proliferation but did not result in enhanced protective efficacy if compared to three wild type recombinant immunisations (100% protection), three recombinant

ALVAC-CEA

immunisations (70% protection) or WR prime followed by two ALVAC-CEA immunisations (63% protection).

Thus some studies of heterologous prime-boost combination have found some enhancement of antibody and lymphoproliferative responses but no significant effect on protective efficacy in an animal model. CD8 T cells were not measured in these studies. The limited enhancement of antibody response probably simply reflects the fact that antibodies to the priming immunogen will often reduce the immunogenicity of a second immunisation with the same immunogen, while boosting with a different carrier will in part overcome this problem. This mechanism would not be expected to be significantly affected by the order of immunisation.

Evidence that a heterologous prime boost immunisation regirfie might affect CD8 T cell responses was provided byLi et al. (1993).

They described partial protective efficacy induced in mice against malaria sporozoite challenge by administering two live viral vectors, a recombinant replicating influenza virus followed by a recombinant replicating vaccinia virus encoding a malaria epitope. Reversing the order of immunisation led to loss of all protective efficacy and the authors suggested that this might be related to infection of liver cells by vaccinia, resulting in localisation of CTLs in the liver to protect against the hepatocytic stages of malaria parasites.

Rodrigues et al. Immunol. 1994, 4636-4648) describe immunising mice with repeated doses of a recombinant influenza virus expressing an immunodominant B cell epitope of the malarial circumsporozoite (CS) protein followed by a recombinant vaccinia virus booster. The use of a wild type vaccinia strain and an attenuated but to replication-competent vaccinia strain in the booster yielded very similar levels of partial protection. However the attenuated but replication competent strain was slightly less immunogenic for priming CD8 T cells than the wild type vaccinia strain.

Murata et al. (Cell. Immunol. 1996, 173: 96-107) reported is enhanced CD8 T cell responses after priming with replicating recombinant influenza viruses and boosting with a replicating strain of vaccinia virus and suggested that the partial protection observed in the two earlier studies was attributable to this enhanced CD8 T cell induction.

Thus these three studies together provide evidence that a booster immunisation with a replicating recombinant vaccinia virus may enhance to some degree CD8 T cell induction following priming with a replicating recombinant influenza virus. However, there are two limitations to these findings in terms of their potential usefulness. Firstly, the immunogenicity induced was only sufficient to achieve partial protection against malaria and even this was dependent on a highly immunogenic priming immunisation with-an unusual replicating .recombinant influenza virus. Secondly, because of the potential and documented side-effects of using these replicating viruses as immunogens these recombinant vectors are not suitable for general human use as vaccines.

Modified vaccinia virus Ankara (MVA) is a strain of vaccinia virus which does not replicate in most cell types, including normal human tissues. MVA was derived by serial passage >500 times in chick embryo fibroblasts (CEF) of material derived from a pox lesion on a horse in Ankara, Turkey (Mayr et al. 1975). It was shown to be replication-impaired yet able to induce protective immunity against veterinary poxvirus infections (Mayr 1976). MVA was used as a human vaccine in the final stages of the smallpox eradication campaign, being administered by intracutaneous, subcutaneous and intramuscular routes to >120,000 subjects in southern Germany. No significant side effects were recorded, despite the deliberate targeting of vaccination to high risk groups such as those with eczema (Mayr et al. 1978; Stickl et al. 1974; Mahnel et al.

1994;). The safety of MVA reflects the avirulence of the virus in animal models, including irradiated mice and following intracranial administration to neonatal mice. The non-replication of MVA has been correlated with the production of proliferative white plaques on chick chorioallantoic membrane, abortive infection of non-avian cells, and the presence of six genomic deletions totalling approximately 30 kb (Meyer et al. 1991). The avirulence of MVA has been ascribed partially to deletions affecting host range genes K1L and C7L, although limited viral replication still occurs on human TK-143 cells and African Green Monkey CV-1 cells (Altenburger et al. 1989). Restoration of the K1L gene only partially restores MVA host range (Sutter et al. 1994). The host range restriction appears to occur during viral particle maturation, with only immature virions being observed in human HeLa cells on electron microscopy (Sutter et al. 1992). The late block in viral replication does not preveirt efficient expression of recombinant genes in MVA. Recombinant MVA expressing influenza nucleoprotein, influenza haemagglutinin, and SIV proteins have proved to be immunogenic and provide varying degrees of protection in animal models, although this has never been ascribed to CD8+ T lymphocytes alone (Sutter et al. 1994, Hirsch et al. 1995; Hirsch et al. 1996).

Recombinant MVA is considered a promising human vaccine candidate because of these properties of safety and immunogenicity (Moss et al.

1995). Recombinant MVA containing DNA which codes for foreign antigens is described in US 5,185,146 (Altenburger).

Poxviruses have evolved strategies for evasion of the host immune response that include the production of secreted proteins that function as soluble receptors for tumour necrosis factor, IL-1p, interferon (IFN)-a/P and IFN-y, which normally have sequence similarity to the extracellular domain of cellular cytokine receptors (Symons et al. 1995; Alcami et al. 1995; Alcami et al. 1992). The most recently described receptor of this nature is a chemokine receptor (Graham et al. 1997).

These viral receptors generally inhibit or subvert an appropriate host immune response, and their presence is associated with increased pathogenicity. The 11-1P receptor is an exception: its presence diminishes the host febrile response and enhances host survival in the face of infection (Alcami et al. 1996). We have discovered that MVA lacks functional cytokine receptors for interferon y, interferon ap, Tumour Necrosis Factor and CC chemokines, but it does possess the potentially beneficial IL-1 P receptor. MVA is the only known strain of vaccinia to possess this cytokine receptor profile, which theoretically renders it safer and more immunogenic than other poxviruses. Another replicationimpaired and safe strain of vaccinia known as NYVAC is fully described in Tartaglia et al.(Virology 1992, 188: 217-232).

It has long been recognised that live viruses have some afttractive features as recombinant vaccine vectors including a high capacity for foreign antigens and fairly good immunogenicity for cellular immune responses (Ellis 1988 new technologies for making vaccines. In: Vaccines. Editors: Plotkin S A and Mortimer E A. W B Saunders, Philadelphia, page 568; Woodrow G C 1977. In: New Generation 8 Vacciness 2nd Edition. Editors: Levine M M, Woodrow G C, Kaper J B, Cobon G, page 33). This has led to attempts to attenuate the virulence of such live vectors in various ways including reducing their replication capacity (Tartaglia J et al. 1992 Virology 188: 217-232). However such a reduction in replication reduces the amount of antigen produced by the virus and thereby would be expected to reduce vaccine immunogenicity.

Indeed attenuation of replicating vaccinia strains has previously been shown to lead to some substantial reductions in antibody responses (Lee M S et al, 1992 J Virology 66: 2617-2630). Similarly the non-replicating to fowlpox vector was found to be less immunogenic for antibody production and less protective than a replicating wild-type vaccinia strain in a rabies study (Taylor J et al. 1991 Vaccine 9: 190-193).

It has now been discovered that non-replicating and replication-impaired strains of poxvirus provide vectors which give an extremely good boosting effect to a primed CTL response. Remarkably, this effect is significantly stronger than a boosting effect by wild type poxviruses. The effect is observed with malarial and other antigens such as viral and tumour antigens, and is protective as shown in mice and nonhuman primate challenge experiments. Complete rather than partial protection from sporozoite challenge has been observed with the novel immunisation regime.

It is an aim of this invention to identify an effective means of immunising against malaria. It is a further aim of this invention to identify means of immunising against other diseases in which CD8+ T cell responses play a protective role. Such diseases include but are not limited to infection and disease caused by the viruses HIV, herpes simplex, herpes zoster, hepatitis C, hepatitis B, influenza, Epstein-Barr virus, measles, dengue and HTLV-1; by the bacteria Mycobacterium tuberculosis and Listeria sp.; and by the protozoan parasites Toxoplasma and Trypanosoma; and certain forms of cancer e.g. melanoma, cancer of the breast and cancer of the colon.

We describe here a novel method of immunising that generated very high levels of CD8+ T cells and was found to be capable of inducing unprecedented complete protection against P. berghei sporozoite challenge. The same approach was tested in higher primates and found to be highly immunogenic in this species also, and was found to induce partial protection against P. falciparum challenge. Induction of protective immune responses has also been demonstrated in two additional mouse models of viral infection and cancer.

We show further than the novel immunisation regime that is described here is also effective in generating strong CD8+ T cell responses against HIV epitopes. Considerable evidence indicates that the generation of such CD8+ T cell responses can be expected to be of value in prophylactic or therapeutic immunisation against this viral infection and disease (Gallimore et al 1995; Ada 1996). We demonstrate that strong CD8+T cell responses may be generated against epitopes from both HIV and malaria using an epitope string with sequences from both of these micro-organisms. The success in generating enhanced immunogenicity against both HIV and malaria epitopes, and also against influenza and tumour epitopes, indicates that this novel immunisation regime can be effective generally against many infectious pathogens and also in noninfectious diseases where the generation of a strong CD8+ T cell response may be of value.

A surprising feature of the current invention is the finding of the very high efficacy of non-replicating agents in both priming and particularly in boosting a CD8+ T cell response. In general the immunogenicity of CD8+ T cell induction by live replicating viral vectors has previously been found to be higher than for non-replicating agents or replication-impaired vectors. This is as would be expected from the greater O amount of antigen produced by agents that can replicate in the host. Here c however we find that the greatest immunogenicity and protective efficacy is surprisingly observed with non-replicating vectors. The latter have an added advantage for vaccination in that they are in general safer fcor use in humans than replicating vectors.

The present invention provides in one aspect a medicament when used for boosting a naturally primed CD8+ T cell response against at least one target antigen in an individual naturally exposed to the target antigen, comprising a source of one or more CD8+ T cell epitopes of the target antigen, wherein the source of CD8+ T cell epitopes is a non-replicating or a replication-impaired recombinant poxvirus vector, and a pharmaceutically acceptable carrier.

Preferably, the medicament according to the invention boosts a naturally primed CD8+ T cell response against at least one target antigen, wherein the target antigen is a malaria antigen.

Preferably, the medicament according to the invention boosts a naturally primed CD8+ T cell response, wherein the CD8+ T cell response is a primate CD8+ T cell response.

In another aspect the invention provides a method of boosting a naturally primed CD8+ T cell immune response in an individual naturally exposed to the target antigen, which method comprises administering a medicament according to the invention.

In another aspect of the invention a recombinant non-replicating or replication-impaired pox virus vector encoding one or more CD8+ T cell epitopes of a target antigen is used in the manufacture of a medicament for boosting a naturally primed CD8+ T cell immune response against the target antigen.

In another aspect the invention provides a recombinant epitope string comprising the CD8+ T cell epitopes of the amino acid sequences of SEQ ID NOs 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 30, 32 and 34, and further comprising the epitopes of the amino acid sequences of SEQ ID NOs 26, 28, 36, 38 and In a further aspect the invention provides a replication-impaired pox virus encoding the P. falciparum antigen TRAP, for immunising an animal against malaria.

In one preferred embodiment of the invention, the source of CD8+ T cell epitopes in the priming composition is a nucleic acid, which may be DNA or RNA, in particular a recombinant DNA plasmid. The DNA or RNA may be packaged, for example in a lysosome, or it may be in free form.

In another preferred embodiment of the invention, the source of CD8+ T cell epitopes in the priming composition is a peptide, polypeptide, protein, polyprotein or particle comprising two or more CD8+ T cell epitopes, present in a recombinant string of CD8+ T cell epitopes or in a target antigen. Polyproteins include two or more proteins which may be the same, or preferably different, linked together. Particularly preferred in this embodiment is a recombinant proteinaceous particle such as a Ty virus-like particle (VLP) (Burns et al. Molec.

Biotechnol. 1994, 1: 137-145).

Preferably, the source of CD8+ T cell epitopes in the boosting composition is a vaccinia virus vector such as MVA or NYVAC. Most preferred is the vaccinia strain modified virus ankara (MVA) or a strain derived therefrom. Alternatives to vaccinia vectors include avipox vectors such as fowl pox or canarypox vectors.

Particularly suitable as an avipox vector is a strain of canarypox known as ALVAC (commercially available as Kanapox), and strains derived therefrom.

Poxvirus genomes can carry a large amount of heterologous genetic information. Other requirements for viral vectors for use in vaccines include good immunogenicity and safety. MVA is a replication-impaired vaccinia strain with a good safety record. In most cell types and normal human tissues, MVA does not replicate; limited replication of MVA is observed in a few transformed cell types such as BHK21 cells. It has now been shown, by the results described herein, that recombinant MVA N:\2\16646\au\01\20040610 Amended speci.doc\\ and other non-replicating or replication-impaired strains are surprisingly and significantly better than conventional recombinant vaccinia vectors at generating a protective CD8+ T cell response, when administered in a boosting composition following priming with a DNA plasmid, a recombinant Ty-VLP or a recombinant adenovirus.

It will be evident that vaccinia virus strains derived from MVA, or independently developed strains having the features of MVA which make MVA particularly suitable for use in a vaccine, will also be suitable for use in the invention.

MVA containing an inserted string of epitopes

(MVA-HM,

which is described in the Examples) has been deposited at the European Collection of Animal Cell Cultures, CAMR, Salisbury, Wiltshire SP4 OJG, UK under accession no. V97060511 on 5 June 1997.

The term "non-replicating" or "replication-impaired" as used herein means not capable of replication to any significant extent in the majority of normal mammalian cells or normal human cells. Viruses which are non-replicating or replication-impaired may have become so naturally they may be isolated as such from nature) or artificially e.g. by breeding in vitro or by genetic manipulation, for example deletion of a gene which is critical for replication. There will generally be one or a few cell types in which the viruses can be grown, such as CEF cells for MVA.

Replication of a virus is generally measured in two ways: 1) DNA synthesis and 2) viral titre. More precisely, the term "nonreplicating or replication-impaired" as used herein and as it applies to poxviruses means viruses which satisfy either or both of the following criteria: 1) exhibit a 1 log (10 fold) reduction in DNA synthesis compared to the Copenhagen strain of vaccinia virus in MRC-5 cells (a human cell line); 2) exhibit a 2 log reduction in viral titre in HELA cells (a human cell line) compared to the Copenhagen strain of vaccinia virus.

Examples of poxviruses which fall within this definition are MVA, NYVAC and avipox viruses, while a virus which falls outside the definition is the attenuated vaccinia strain M7.

Alternative preferred viral vectors for use in the priming composition according to the invention include a variety of different viruses, genetically disabled so as to be non-replicating or replication-impaired.

Such viruses include for example non-replicating adenoviruses such as El deletion mutants. Genetic disabling of viruses to produce non-replicating or replication-impaired vectors has been widely described in the literature McLean et al. 1994).

Other suitable viral vectors for use in the priming composition are vectors based on herpes virus and Venezuelan equine encephalitis virus (VEE) (Davies et al. 1996). Suitable bacterial vectors for priming include recombinant BCG and recombinant Salmonella and Salmonella transformed with plasmid DNA (Darji A et al. 1997 Cell 91: 765-775).

Alternative suitable non-viral vectors for use in the priming composition include lipid-tailed peptides known as lipopeptides, peptides fused to carrier proteins such as KLH either as fusion proteins or by chemical linkage, whole antigens with adjuvant, and other similar systems.

Adjuvants such as QS21 or SBAS2 (Stoute J A et al. 1997 N Engl J Medicine 226: 86-91) may be used with proteins, peptides or nucleic acids to enhance the induction of T cell responses. These systems are sometimes referred to as "immunogens rather than "vectors", but they are vectors herein in the sense that they carry the relevant CD8+ T cell epitopes.

There is no reason why the priming and boosting compositions should not be identical in that they may both contain the 14 priming source of CD8+ T cell epitopes as defined in above and the boosting source of CD8+ T cell epitopes as defined in (ii) above. A single formulation which can be used as a primer and as a booster will simplify administration. The important thing is that the primer contains at least the priming source of epitopes as defined in above and the booster contains at least the boosting source of epitopes as defined in (ii) above.

The CD8+ T cell epitopes either present in, or encoded by the priming and boosting compositions, may be provided in a variety of different forms', such as a recombinant string of one or two or more epitopes, or in the context of the native target antigen, or a combination of both of these. CD8+ T cell epitopes have been identified and can be found in the literature, for many different diseases. It is possible to design epitope strings to generate a CD8+ T cell response against any chosen antigen that contains such epitopes. Advantageously, the epitopes in a string of multiple epitopes are linked together without intervening sequences so that unnecessary nucleic acid and/or amino acid material is avoided. In addition to the CD8+ T cell epitopes, it may be preferable to include one or more epitopes recognised by T.helper cells, to augment the immune response generated by the epitope string. Particularly suitable T helper cell epitopes are ones which are active in individuals of different HLA types, for example T helper epitopes from tetanus (against which most individuals will already be primed). A useful combination of three T helper epitopes is employed in the examples described herein. It may also be useful to include B cell epitopes for stimulating B cell responses and antibody production.

The priming and boosting compositions described-may advantageously comprise an adjuvant. In particular, a priming composition comprising a DNA plasmid vector may also comprise granulocyte macrophage-colony stimulating factor (GM-CSF), or a plasmid encoding it, to act as an adjuvant; beneficial effects are seen using GM- CSF in polypeptide form.

The compositions described herein may be employed as therapeutic or prophylactic vaccines. Whether prophylactic or therapeutic s immunisation is the more appropriate will usually depend upon the nature of the disease. For example, it is anticipated that cancer will be immunised against therapeutically rather than before it has been diagnosed, while anti-malaria vaccines will preferably, though not necessarily be used as a prophylactic.

The compositions according to the invention may be administered via a variety of different routes. Certain routes may be favoured for certain compositions, as resulting in the generation of a more effective response, or as being less likely to induce side effects, or as being easier for administration. The present invention has been shown to is be effective with gene gun delivery, either on gold beads or as a powder.

In further aspects, the invention provides: a method for generating a protective CD8+ T cell immune response against a pathogen or tumour, which method comprises administering at least one dose of a recombinant DNA plasmid encoding at least one CD8+ T cell epitope or antigen of the pathogen or cancer, followed by at least one dose of a non-replicating or replication-impaired recombinant pox virus encoding the same epitope or antigen; a method for generating a protective CD8+ T cell immune response against a pathogen or tumour, which method comprises administering at least one dose of a recombinant protein or.particle Scomprising at least one epitope or antigen of the pathogen or cancer, followed by at least one dose of a recombinant MVA vector encoding the same epitope or antigen; the use of a recombinant non-replicating or replication-impaired pox virus vector in the manufacture of a medicament for boosting a CD8+ T cell immune response; the use of an MVA vector in the manufacture of a medicament for boosting a CD8+ T cell immune response; and the priming and/or boosting compositions described herein, in particulate form suitable for delivery by a gene gun; and methods of immunisation comprising delivering the compositions by means of a gene gun.

The terms "comprise", "comprises", "comprised" and "comprising" when used in this specification are taken to specify the presence of stated features, integers, steps or components but does not preclude the presence or addition of one or more other features, integers, steps, components or groups thereof.

Example Formulations and Immunisation Protocols Formulation 1 Priming Composition: DNA plasmid 1 mg/ml in PBS Boosting Composition: Recombinant MVA, 108 flu in PBS Protocol: Administer two doses of 1 mg of priming composition, at 0 and 3 weeks followed by two doses of booster intradermally at 6 and 9 weeks.

Formulation 2 Priming Composition: Boosting Composition: Ty-VLP 500pg in PBS MVA, 10 ffu in PBS Protocol: Administer two doses of priming composition, at 0 and 3 o0 weeks, then 2 doses of booster at 6 and 9 weeks. For tumour treatment, MVA is given i.v. as one of most effective routes.

Formulation 3 Priming Composition: Boosting.Composition: Protein 500p.g adjuvant (QS-21) Recombinant MVA, 108 ffu in PBS Protocol: Administer two doses of priming composition at 0 and 3 weeks and 2 doses of booster i.d. at 6 and 9 weeks.

Formulation 4 Priming Composition: Boosting Composition: Adenovirus vector, 109 pfu in PBS Recombinant MVA, 108 ffu in PBS Protocol: Administer one or two doses of priming composition intradermally at 0 and 3 weeks and two doses of booster i.d. at 6 and 9 weeks... The above doses and protocols may be varied to optimise protection.

Doses may be given between for example, 1 to 8 weeks apart rather than 2 weeks apart.

The invention will now be further described in the examples which follow.

EXAMPLES

EXAMPLE 1 Materials and Methods Generation of the epitope strings.

The malaria epitope string was made up of a series of cassettes each encoding three epitopes as shown in Table 1, with restriction enzyme sites at each end of the cassette. Each cassette was constructed from four synthetic oligonucleotides which were annealed together, ligated into a cloning vector and then sequenced to check that no errors had been introduced. Individual cassettes were then joined together as required. The BamHI site at the 3' end of cassette C was fused to the is Bgll/ site at the 5' end of cassette A, destroying both restriction enzyme sites and encoding a two amino acid spacer (GS) between the two cassettes. Cassettes B, D and H were then joined to the string in the same manner. A longer string containing CABDHFE was also constructed in the same way.

Table 1. CTL epitopes of the malaria string Cassette Epitope Amino acid DNA sequence Type HLA Sequence restrictio n A Ls8 KPNDKSLY AAGCCGAACGACAAGTCCTTGTAT CTL 835 Cp26 KPKDELDY AAACCTAAGGACGAATTGGACTAC CTL Ls6 KPIVQYDNF AAGCCAATCGTTCAATACGACAACTTC CTL 853 B Tr42/43 ASKNKEKALII GCCTCCAAGAACAAGGAAAAGGCTTTGATCA CTL 88

TC

Tr39 GIAGGLALL GGTATCGCTGGTGGTFTGGCCTTGTTG CTL A2.1 Cp6 MNPNDPNRN ATGAACCCTAATGACCCAAACAGAAACGTC CTL B7

V

C St8 MINAYLDKL ATGATCAACGCCTAC17GGACAAGTTG CTL A2.2 ISKYEDEI ATCTCCAAGTACGAAGACGAAATC CTL B1 7 Pb9 SYIPSAEKI TCCTACATCCCATCTGCCGAAAAGATC CTL mouse H2-Kd D Tr26 HLGNVKYLV CACTTGGGTAACGTTAAGTACTTGGTT CTL A2. 1 Ls53 KSLYDEHI AAGTCTTTGTACGATGAACACATC CTL 858 Tr29 LLMDCSGSI TTATTGATGGACTGTTCTGGTTCTATT CTL A2.2 E NANP NANPNANPN AACGCTAATCCAAACGCAAATCCGAACGCCA B cell ANPNANP ATCCTAACGCGAATCCC TRAP DEWSPCSVT GACGAATGGTCTCCATGTTCTGTCACTTGTG Heparin AM CGKGTRSRK GTAAGGGTACTCGCTCTAGAAAGAGAGAA binding RE motif F Cp39 YLNKIONSL TACTTGAACAAAATTCAAAACTCTTTG CTL A2.1 La72 MEKLKELEK ATGGAAAAGTTGAAAGAATTGGAAAAG CTL 8 ex23 ATSVLAGL GCTACTTCTGTCTTGGCTGGTTTG CTL B58 H CS -P DPNANPNVD GACCCAAACGCTAACCCAAACGTTGACCCA T helper Universal PNANPNV AACGCCAACCCAAACGTC BCG QVHFQPLPP CAAGTrCACTCcAACCATUGcCTCCGGCCG T helper epitopes AWVKL TTGTCAAGTTG TT aFIKANSKFI CAATTCATCAAGGCCAACTCTAAGTTCATCG T helper .GITE GTATCACCGAAII Table 1 Sequences included in the malaria epitope string. Each cassette consists of the epitopes shown above, in the order shown, with no additional sequence between epitopes within a cassette. A Bgll site was added at the 5' end and a BamHI site at the 3' end, such that between cassettes in an epitope string the BamHl/Bglll junction encodes GS. All epitopes are from P. falciparum antigens except for pb9 berghei), BCG tuberculosis) and TT (Tetanus). The amino acid and DNA sequences show n in the table have SEQ ID NOS. 1 to 40 in the order in which they appear.

Figure 1 shows the construct used to express Ty-VLP with the malaria epitope cassette CABDHFE. CTL epitopes are from P.

falciparum STAERP (sporozoite threonine- and asparagine-rich protein) (st), LSA-1 (liver stage antigen 1) CSP (circumnsporozoite protein) (cp), TRAP (thrombospondin-related adhesive protein) LSA-3 (liver stage antigen 3) (1a) and Exp-1 (exported protein 1) Helper epitopes are from the P. falciparum CS protein, the M. tuberculosis 38Kd antigen and Tetanus Toxoid. NANP is the antibody epitope from CS and AM is the adhesion motif from P. falciparum TRAP (Muller et al 1993). The length of the complete string is 229 amino acids as shown in the table 1 legend, with the amino acid sequence:-

MINAYLDKLISKYEDEISYIPSAEKIGSKPNDKSLYKPKDELDYKPIVQYDN

FGSASKN KE KALI IG IAGGLALLM NPN DPN RNVGS HLGNVKYLVKS LYIDE i0 H ILLMDCSG SI GSDPNAN PNVDPNAN PNVQVHFQ PLP PAVKLQFI KNS KFI GITEG SYLN KIQ NSLM EKLKELE KATSVLAG LGS NAN PNANPNAN PNA NPDEWSPCSVTCGKGTRSRKREGSGK [SEQ ID NO: 41].

The HIV epitope string was also synthesised by annealing oligonucleotides. Finally the HIV and malaria epitope strings were fused by joining the BamHI site at the 3' end of the HIV epitopes to the BgIII site at the 5' end of cassettes CAB to form the HM string (Table 2) Table 2 CTL epitopes of the HIVISI1V epitope string Epitope

YLKDQQLL

ERYLKDQQL

EITPIGLAP

PPI PVGEIY

GEIYKRWII

KRWIILGLNK

II LGLNKIVR

LGLNKIVRMY

YNLTMKCR

RGPGRAFVTI

GRAFVTIGK

TPYDINQML

CTPYDINQM

Restriction A24, B8 B14 Mamu-B*01 B35 B8 B*2705 A33 Bw62 Mamu-A*02 A2, H-2Dd 8*2705 B53 Mamu-A*O 1 Origin HIV-1 gp4l HIV-1 gp4l SIVen HIV-1 p24 HIV-1 p24 HIV-1 p24 H!V-1 p 24 HIV-1 p24 SIV env H!V-1 gpl2O HIV-1 gpl2O HIV-2 gag Sly gag RPQVPLRPMTY 851 HIV-1 nef QVPLRPMTYK A*0301, All HIV-1 nef VPLRPMTY B35 HIV-1 nef AVDLSHFLK All HIV-1 nef DLSHFLKEK A*0301 HIV-1 nef FLKEKGGL B8 HIV-1 nef ILKEPVHGV A*0201 HIV-1 pol ILKEPVHGVY Bw62 HIV-1 pol HPDIVIYQY B35 HIV-1 pol VIYQYMDDL A*0201 HIV-1 pol Table 2 Sequences of epitopes from HIV or SIV contained in the H epitope string and assembled as shown in figure 2. The amino acids in the table have SEQ ID NOS: 42 to 64 in the order in which they appear.

Figure 2 shows a schematic outline of the H, M and HM proteins. The bar patterns on the schematic representations of the polyepitope proteins indicate the origin of the sequences (see tables 1 and The positions of individual epitopes and their MHC restrictions are depicted above and below the proteins. Pb is the only epitope derived from the protein of P. berghei. All other epitopes in the M protein originate from proteins of P. falciparum: cs circumsporozoite protein, st STARP, Is LSA-1 and tr TRAP. BCG 38 kDa protein of M. tuberculosis; TT tetanus toxin.

For the anti-tumour vaccine an epitope string containing CTL epitopes was generated, similar to the malaria and HIV epitope string. In this tumour epitope string published murine CTL epitopes were fused together to create the tumour epitope string with the amino acid sequence:

MLPYLGWLVF-AQHPNAELL-KHYLFRNL-SPSYVYHQF-IPNPLLGLD

[SEQ ID NO: 65]. CTL epitopes shown here were fused together. The first amino acid methionine was introduced to initiate translation.

Ty virus-like particles (VLPs).

The epitope string containing cassette CABDH was introduced into a yeast expression vector to make a C-terminal in-frame fusion to the TyA protein. When TyA or TyA fusion proteins are expressed in yeast from this vector, the protein spontaneously forms virus like particles which can be purified from the cytoplasm of the yeast by sucrose gradient centrifugation. Recombinant Ty-VLPs were prepared in this manner and dialysed against PBS to remove the sucrose before injection Layton et al. 1996).

Adenoviruses Replication-defective recombinant Adenovirus with a deletion of the El genes was used in this study (McGrory et al, 1988). The Adenovirus expressed E. coli P-galactosidase under the control of a CMV IE promoter. For immunisations, 10' pfu of virus were administered intradermally into the ear lobe.

Peptides Peptides were purchased from Research Genetics (USA), dissolved at 10 mg/ml in DMSO (Sigma) and further diluted in PBS to 1 mg/ml. Peptides comprising CTL epitopes that were used in the experiments described herein are listed in table 3 Table 3 Sequence of CTL peptide epitopes sequence Antigen

MHC

restriction LPYLGWLVF P1A tumour antigen Ld SYIPSAEKI P. berghei CSP Kd

RGPGRAFVTI

HIV gag TPHPARIGL E. coli P-galactosidase Ld TYQRTRALV Influenza A virus NP Kd SDYEGRLI Influenza A virus NP Kk ASNENMETM Influenza A virus NP Db INVAFNRFL P. falciparum TRAP Kb The amino acid sequences in Table 3 have SEQ ID NOS: 66 to 73, in the order in which they appear in the Table.

Plasmid DNA constructs A number of different vectors were used for constructing DNA vaccines. Plasmid pTH contains the CMV IE promoter with intron A, followed by a polylinker to allow the introduction of antigen coding sequences and the bovine growth hormone transcription.termination sequence. The plasmid carries the ampicillin resistance gene and is capable of replication in E. coli but not mammalian cells. This was used to make DNA vaccines expressing each of the following antigens: P. berghei TRAP, P. berghei CS, P. falciparum TRAP, P. falciparum LSA-1 (278 amino acids of the C terminus only), the epitope string containing cassettes CABDH and the HM epitope string (HIV epitopes followed by cassettes CAB). Plasmid pSG2 is similar to pTH except for the antibiotic resistance gene. In pSG2 the ampicillin resistance gene of pTH has been replaced by a kanamycin resistance gene. pSG2 was used to to make DNA vaccines expressing the following antigens: P. berghei PbCSP, a mouse tumour epitope string, the epitope string containing cassettes CABDH and the HM epitope string. Plasmid V1J-NP expresses influenza nucleoprotein under the control of a CMV IE promoter. Plasmids CMV-TRAP and CMV-LSA-1 are similar to pTH.TRAP and pTH. LSA-1 but do not contain intron A of the CMV promoter. Plasmids RSV.TRAP and RSV.LSA-1 contain the RSV promoter, SV40 transcription termination sequence and are tetracycline resistant. For induction of p-galactosidase-specific CTL plasmid pcDNA3/His/LacZ (Invitrogen) was used. All DNA vaccines were prepared from E. coli strain DH5a using Qiagen plasmid purification columns.

Generation of recombinant vaccinia viruses Recombinant MVAs were made by first cloning the antigen sequence into a shuttle vector with a viral promoter such as the plasmid pSC11 (Chakrabarti et al. 1985; Morrison et al. 1989). P. berghei CS and P. falciparum TRAP, influenza nucleoprotein and the HM and mouse io tumour epitope polyepitope string were each expressed using the promoter (Mackett et al. 1984), and P. berghei TRAP was expressed using the strong synthetic promoter (SSP; Carroll et al. 1995). The shuttle vectors, pSC11 or pMCO3 were then used to transform cells infected with wild-type MVA so that viral sequences flanking the promoter, antigen coding sequence and marker gene could recombine with the MVA and produce recombinants. Recombinant viruses express the marker gene (1 glucuronidase or 0 galactosidase) allowing identification of plaques containing recombinant virus. Recombinants were repeatedly plaque purified before use in immunisations. The recombinant NYVAC-PbCSP vaccinia was previously described (Lanar et al. 1996). The wild type or Western Reserve (WR) strain of recombinant vaccinia encoding PbCSP was described previously (Satchidanandam et al. 1991).

Cells and culture medium Murine cells and Epstein-Barr virus transformed chimpanzee and macaque B cells (BCL) were cultured in RPMI supplemented with heat inactivated fetal calf serum (FCS). Splenocytes were restimulated with the peptides indicated (final concentration 1 .g/ml) in MEM medium with 10% FCS, 2mM glutamine, 50U/ml penicillin, 50 pM 2mercaptoethanol and 10mM Hepes pH7.2 (Gibco, UK).

Animals Mice of the strains indicated, 6-8 weeks old were purchased from Harlan Olac (Shaws Farm, Blackthorn, UK). Chimpanzees H1 and H2 were studied at the Biomedical Primate Research Centre at Rijswick, The Netherlands. Macaques were studied at the University of Oxford.

Immunisations Plasmid DNA immunisations of mice were performed by io intramuscular immunisation of the DNA into the musculus tibialis under anaesthesia. Mouse muscle was sometimes pre-treated with 50 pi of 1mM cardiotoxin (Latoxan, France) 5-9 days prior to immunisation as described by Davis et al (1993), but the presence or absence of such pre-treatment was not found to have any significant effect on immunogenicity or protective efficacy. MVA immunisation of mice was performed by either intramuscular intravenous (into the lateral tail vein) intradermal intraperitoneal or subcutaneous immunisation.

Plasmid DNA and MVA immunisation of the chimpanzees H1 and H2 was performed under anaesthesia by intramuscular immunisation of leg muscles. For these chimpanzee immunisations the plasmid DNA was coadministered with 15 micrograms of human GM-CSF as an adjuvant.

Recombinant MVA administration to the chimpanzees was by intramuscular immunisation under veterinary supervision. Recombinant human GM-CSF was purchased from Sandoz (Camberley, UK). For plasmid DNA immunisations using a gene gun, DNA was precipitated onto gold particles. For intradermal delivery, two different types of gene guns were used, the Acell and the Oxford Bioscience device (PowderJect Pharmaceuticals, Oxford, UK).

EL/SPOT assays CD8+ T cells were quantified in the spleens of immunised mice without in vitro restimulation using the peptide epitopes indicated and the ELISPOT assay as described by Miyahara et al (1993). Briefly, 96-well nitrocellulose plates (Miliscreen MAHA, Millipore, Bedford UK) were coated with 15 pg/ml of the anti-mouse interferon-y monoclonal antibody R4 (EACC) in 50 .il of phosphate-buffered saline (PBS). After overnight incubation at 40C the wells were washed once with PBS and blocked for 1 hour at room temperature with 100 pil RPMI with 10% FCS. Splenocytes to from immunised mice were resuspended to 1 x 10 7 cells/mi and placed in duplicate in the antibody coated wells and serially diluted. Peptide was added to each well to a final concentration of 1 pg/ml. Additional wells without peptide were used as a control for peptide-dependence of interferon-y secretion. After incubation at 370C in 5%CO, for 12-18 hours the plates were washed 6 times with PBS and water. The wells were then incubated for 3 hours at room temperature with a solution of 1 pig/ml of biotinylated anti-mouse interferon-y monoclonal antibody XMG1.2 (Pharmingen, CA, USA) in PBS. After further washes with PBS, 50 il of a 1 pg/ml solution of streptavidin-alkaline-phosphatase polymer (Sigma) was added for 2 hours at room temperature. The spots were developed by adding 50 .l of an alkaline phosphatase conjugate substrate solution (Biorad, Hercules, CA, USA). After the appearance of spots the reaction was stopped by washing with water. The number of spots was determined with the aid of a stereomicroscope.

ELISPOT assays on the chimpanzee peripheral blood lymphocytes were performed using a very similar method employing the assay and reagents developed to detect human CD8 T cells (Mabtech, Stockholm).

CTL assays CTL assays were performed using chromium labelled target cells as indicated and cultured mouse spleen cells as effector cells as described by Allsopp et al. (1996). CTL assays using chimpanzee or macaque cells were performed as described for the detection of human CTL by Hill et al. (1992) using EBV-transformed autologous chimpanzee chimpanzee or macaque B cell lines as target cells.

P. berghei challenge io Mice were challenged with 2000 (BALB/c) or 200 (C57BL/6) sporozoites of the P. bergheiANKA strain in 200 l RPMI by intravenous inoculation as described (Lanar et al. 1996). These sporozoites were dissected from the salivary glands of Anopheles stephensi mosquitoes maintained at 18°C for 20-25 days after feeding on infected mice. Bloodstage malaria infection, indicating a failure of the immunisation, was detected by observing the appearance of ring forms of P. berghei in Giemsa-stained blood smears taken at 5-12 days post-challenge.

P. falciparum challenge The chimpanzees were challenged with 20,000 P. falciparum sporozoites of the NF54 strain dissected from the salivary glands of Anopheles gambiae mosquitoes, by intravenous inoculation under anaesthesia. Blood samples from these chimpanzees were examined daily from day 5 after challenge by microscopy and parasite culture, in order to detect the appearance of low levels of P. falciparum parasites in the peripheral blood.

P815 tumour challenges Mice were challenged with 1 x 105 P815 cells in 200 ld of PBS by intravenous inoculation. Animals were monitored for survival.

Influenza virus challenges Mice were challenged with 100 haemagglutinating units (HA) of influenza virus A/PR/8/34 by intranasal inoculation. Following challenge the animals were weighed daily and monitored for survival.

Determining peptide specific CTL using tetramers Tetrameric complexes consisting of Mamu-A*01-heavy chain and p3-microglobulin were made as described by Ogg et al (1998). DNA coding for the leaderless extracellular portion of the Mamu-A*01 MHC class I heavy chain was PCR-amplified from cDNA using MamuNdel: 5'-CCT GAC TCA GAC CAT ATG GGC TCT CAC TCC ATG [SEQ ID NO: 74] and 3' primer: 5'-GTG ATA AGC TTA ACG ATG ATT CCA CAC CAT TTT CTG TGC ATC CAG AAT ATG ATG CAG GGA TCC CTC CCA TCT CAG GGT GAG GGG C [SEQ ID NO: 75]. The former primer contained a Nde/ restriction site, the latter included a Hindll/ site and encoded for the bioinylation enzyme BirA substrate peptide. PCR products were digested with Nde/ and Hindll/ and ligated into the same sites of the polylinker of bacterial expression vector pGMT7. The rhesus monkey gene encoding a leaderless p 2 -microglobulin was PCR amplifed from a cDNA clone using primers B2MBACK: 5'-TCA GAC CAT ATG TCT CGC TCC GTG GCC [SEQ ID NO: 76] and B2MFOR: 5'-TCA GAC AAG CTT TTA CAT GTC TCG ATC CCA C [SEQ ID NO: 77] and likewise cloned into the Ndel and Hind//ll sites of pGMT7. Both chains were expressed in E. coli strain BL-21, purified from inclusion bodies, refolded in the presence of peptide CTPYDINQM [SEQ ID NO: 54], biotinylated using the BirA enzyme (Avidity) and purified with FPLC and monoQ ion exchange columns. The amount of biotinylated refolded MHC-peptide complexes was estimated in an ELISA assay, whereby monomeric complexes were first captured by conformation sensitive monoclonal antibody W6/32 and detected by alkaline phosphatase (AP) -conjugated streptavidin (Sigma) followed by colorimetric substrate for AP. The formation of tetrameric complexes was induced by addition of phycoerythrin (PE)-conjugated streptavidin (ExtrAvidin; Sigma) to the refolded biotinylated monomers at a molar ratio of MHC-peptide PEstreptavidin of 4 1. The complexes were stored in the dark at 4 0 C. These tetramers were used to analyse the frequency of Mamu-A*01/gag-specific CD8+ T cells in peripherial blood lymphocytes (PBL) of immunised macaques.

EXAMPLE 2 Immunogenicity Studies in Mice Previous studies of the induction of CTL against epitopes in the circumsporozoite (CS) protein of Plasmodium berghei and Plasmodium yoelii have shown variable levels of CTL induction with different delivery systems. Partial protection has been reported with plasmid DNA (Sedegah et al. 1994), influenza virus boosted by replicating vaccinia virus (Li et al.

1991), adenovirus (Rodrigues et al 1997) and particle delivery systems (Schodel et al. 1994). Immunisation of mice intramuscularly with micrograms of a plasmid encoding the CS protein produced moderate levels of CD8+ cells and CTL activity in the spleens of these mice after a single injection (Figures 3, 4).

For comparison groups of BALB/c mice (n 5) were injected intravenously with 106 ffu/pfu of recombinant vaccinia viruses of different strains (WR, NYVAC and MVA) all expressing P. berghei CSP. The frequencies of peptide-specific CD8+ T cells were measured 10 days later in an ELISPOT assay.. MVA.PbCSP induced 181 48, NYVAC 221+/- 27 and WR 19 (mean standard deviation) peptide-specific CD8+ T cells per million splenocytes. These results show that surprisingly replication-irnpaired vaccinia viruses are superior to replicating strains in priming a CD8+ T cell response. We then attempted to boost these moderate CD8+ T cell responses induced by priming with either plasmid DNA or MVA using homologous or heterologous vectors. A low level of CD8+ T cells was observed after two immunisations with CS recombinant DNA vaccine alone, the recombinant MVA vaccine alone or the recombinant MVA followed by recombinant DNA (Figure A very much higher level of CD8+ T cells was observed by boosting the DNA-primed immune response with recombinant MVA. In a second experiment using ten mice per group the enhanced immunogenicity of the DNA/MVA sequence was confirmed: DNA/MVA 856 201; MVA/DNA 168 72; MVA/MVA 345+/- 90; DNA/DNA 92 46. Therefore the sequence of a first immunisation with a recombinant plasmid encoding the CS protein followed by a second immunisation with the recombinant MVA virus yielded the highest levels of CD8+ T lymphocyte response after immunisation.

Figure 3 shows malaria CD8 T cell ELISPOT data following different immunisation regimes. Results are shown as the number of peptide-specific T cells per million splenocytes. Mice were immunised either with the PbCSP-plasmid DNA or the PbCSP-MVA virus or combinations of the two as shown on the X axis, at two week intervals and the number of splenocytes specific for the pb9 malaria epitope assayed two weeks after the last immunisation. Each point represents the number of spot-forming cells (SFCs) measured in an individual mouse. The highest level of CD8+ T cells was induced by priming with the plasmid DNA and boosting with the recombinant MVA virus. This was more immunogenic than the reverse order of immunisation (MVA/DNA), two DNA immunisations (DNA/DNA) or two MVA immunisations (MVAIMVA). It was also more immunogenic than the DNA and MVA immunisations given simultaneously (DNA MVA 2w), than one DNA immunisation (DNA 4w) or one MVA immunisation given at the earlier or later time point (MVA 2w and MVA 4w).

Figure 4 shows that malaria CD8 T cell ELISPOT and CTL levels are substantially boosted by a recombinant MVA immunisation following priming with a plasmid DNA encoding the same antigen. A and C. CD8+ T cell responses were measured in BALB/c mice using the yinterferon ELISPOT assay on fresh splenocytes incubated for 18 h with the Kd restricted peptide SYIPSAEKI [SEQ ID NO: 67] from P. berghei CSP and the Ld restricted peptide TPHPARIGL [SEQ ID NO: 69] from E. coli pgalactosidase. Note that the ELISPOT counts are presented on a logarithmic scale. B and D. Splenocytes from the same mice were also to assayed in conventional S'Cr-release assays at an effector: target ration of 100:1 after 6 days of in vitro restimulation with the same peptides (1 pg/ml).

The mice were immunised with plasmid DNA expressing either P. berghei CSP and TRAP, PbCSP alone, the malaria epitope cassette including the P. berghei CTL epitope (labelled pTH.M), or Pgalactosidase. ELISPOT and CTL levels measured in mice 23 days after one DNA immunisation are shown in A and B respectively. The same assays were performed with animals that received additionally 1x 107 ffu of recombinant MVA expressing the same antigen(s) two weeks after the primary immunisation. The ELISPOT and CTL levels in these animals are shown in C and D respectively. Each bar represents data from an individual animal.

Studies were also undertaken of the immunogenicity of the epitope string HM comprising both HIV and malaria epitopes in tandem.

Using this epitope string again the highest levels of CD8+ T cells and CTL were generated in the spleen when using an immunisation with DNA vaccine followed by an immunisation with a recombinant MVA vaccine (Table 4, Figure Table 4 Immunogenicity of various DNAIMVA combinations as determined by ELISPOT assays Immunisation 1 Immunisation 2 HIV epitope Malaria epitope DNA-HM DNA-HM 56 26 4 4 MVA-HM MVA-HM 786 334 238 ±106 MVA-HM DNA-HM 306 78 58 18 DNA-HM MVA-HM 1000 487 748 446 None DNA-HM 70 60 100 None MVA-HM 422 128 212 94 Table 4 shows the results of ELISPOT assays performed to measure the levels of specific CD8+ T cells to HIV and malaria epitopes following different immunisation regimes of plasmid DNA and MVA as indicated. The numbers are spot-forming cells per million splenocytes. The- HM epitope string is illustrated in figure 2. BALB/c mice were used in all cases. The malaria epitope was pb9 as in figures 2 and 3. The HIV epitope was RGPGRAFVTI [SEQ ID NO: 51]. The immunisation doses were 50 pg of plasmid DNA or 10' focus-forming units (ffu) of recombinant MVA. All immunisations were intramuscular. The interval between immunisations 1 and 2 was from 14-21 days in all cases.

Figure. 5 shows the CTL responses induced in BALB/c mice to malaria and HIV epitopes by various immunisation regimes employing plasmid DNA and recombinant MVA. Mice were immunised intramuscularly as described in the legend to table 3 and in methods. High levels of CTL specific lysis at effector/target ration of 25:1) were observed to both the malaria and HIV epitopes only after priming with plasmid DNA and boosting with the recombinant MVA. The antigen used in this experiment is the HIV-malaria epitope string. The recombinant

MVA

is denoted MVA.HM and the plasmid DNA expressing this epitope string is denoted pTH.HM. Levels of specific lysis at various effector to target ratios are shown. These were determined after 5 days in vitro restimulation of splenocytes with the two peptides pb9 and RGPGRAFVTI [SEQ ID NO: 51].

Comparison of numerous delivery systems for the induction of CTL was reported by Allsopp et al. (1996). Recombinant Ty-virus like 0o particles (Ty-VLPs) and lipid-tailed malaria peptides both gave good CTL induction but Ty-VLPs were better in that they required only a single immunising dose for good CTL induction. However, as shown here even two doses of Ty particles fail to induce significant protection against sporozoite challenge (Table 7, line Immunisation with a recombinant modified vaccinia Ankara virus encoding the circumsporozoite protein of P.

berghei also generates good levels of CTL. However, a much higher level of CD8+ T cell response is achieved by a first immunisation with the Ty- VLP followed by a second immunisation with the MVA CS vaccine (Table Table 5 Immunogenicity of various Ty-VLP/MVA combinations as determined by ELISPOT and CTL assays Immunisation 1 Immunisation 2 ELISPOT No %Specific Lysis Ty-CABDH Ty- CABDH 75 MVA.PbCSP MVA.PbCSP 38 Ty-CABDH MVA.PbCSP 225 42 Ty- CABDH MVA.HM 1930 nd Table 5 Results of ELISPOT and CTL assays performed to measure the levels of specific CD8+ T cells to the malaria epitope pb9 following different immunisation regimes of Ty-VLPs and recombinant MVA virus as indicated. The CTL and ELISPOT data are from different experiments.

The ELISPOT levels (spots per million splenocytes) are measured on unrestimulated cells and the CTL activity, indicated as specific lysis at an effector to target ratio of 40:1, on cells restimulated with pb9 peptide in vitro for 5-7 days. Both represent mean levels of three mice. BALB/c mice were used in all cases. The.immunisation doses were 50 pg of Ty-VLP or 107 ffu (foci forming units) of recombinant MVA. All immunisations were intramuscular. The interval between immunisations 1 and 2 was from 14- 21 days. MVA.HM includes cassettes CAB.

Priming of an immune response with DNA delivered by a gene gun is and boosting with recombinant MVA Immunogenicity and challenge.

The use of a gene gun to deliver plasmid DNA intradermally and thereby prime an immune response that could be boosted with recombinant MVA was investigated. Groups of BALB/c mice were immunised with the following regimen: I) Three gene gun immunisations with pTH.PbCSP (4 mg per immunisation) at two week intervals II) Two gene gun immunisations followed by MVA i.v. two weeks later III) One intramuscular DNA immunisation followed by MVA i.v. two weeks later.

The immunogenicity of the three immunisation regimens was analysed using ELISPOT assays. The highest frequency of specific T cells was observed with two gene gun immunisations followed by an MVA i.v.

boost and the intramuscular DNA injection followed an MVA i.v. boost (Figure 6).

Figure 6 shows the results of ELISPOT assays performed to measure the levels of specific CD8+ T cells to the malaria epitope pb9 following different immunisation regimes. Groups of BALB/c mice 3) were immunised as indicated gene gun). The time between all immunisations was 14 days. ELISPOT assays were done two weeks after the last immunisation.

CTL induction to the same antigen in different mouse strains To address the question whether the boosting effect described above in BALB/c mice with two CTL epitopes SYIPSAEKI [SEQ ID NO: 67] derived from P. berghei CSP and RGPGRAFVTI [SEQ ID NO: 68] derived from HIV) is a universal phenomenon, two sets of experiments were carried out. CTL responses to the influenza nucleoprotein were studied in five inbred mouse strains. In a first experiment three published murine CTL epitopes derived from the influenza nucleoprotein were studied (see Table Mice of three different H-2 haplotypes, BALB/c and DBA/2 (H-2d), C57BL/6 and 129 CBA/J were used. One set of animals was immunised twice at two week intervals with the plasmid V1J-NP encoding the influenza nucleoprotein. Another set of identical animals was primed with V1J-NP and two weeks later boosted intravenously with 106 ffu of MVA.NP, expressing influenza virus NP. The levels of CTL in individual mice were determined in a s'Cr-release assay with peptide re-stimulated splenocytes. As shown in Figure 7, the DNA priming/MVA boosting immunisation regimen induced higher levels of lysis in all the mouse strains analysed and is superior to two DNA injections.

Figure 7 shows the CTL responses against influenza NP in different mouse strains. Mice of different strains were immunised twice two weeks apart with a DNA vaccine V1J-NP encoding for the influenza nucleoprotein (open circles) or primed with the same DNA vaccine and two weeks later boosted with recombinant MVA expressing influenza virus nucleoprotein (closed circles). Two weeks after the last immunisation splenocytes were restimulated in vitro with the respective peptides (Table The CTL activity was determined in a standard S'Cr-release assay with MHC class I-matched target cells.

CTL induction to different antigens in different mouse strains The effect of MVA boosting on plasmid DNA-primed immune responses was further investigated using different antigens and different inbred mouse strains. Mice of different strains were immunised with different antigens using two DNA immunisations and compared with DNA/MVA immunisations. The antigens used were E. coli -galactosidase, the malaria/HIV epitope string, a murine tumour epitope string and P.

falciparum TRAP. Compared with two DNA immunisations the DNApriming/MVA-boosting regimen induced higher levels of CTL in all the different mouse strains and antigen combinations tested (Figure 8).

Figure 8 shows CTL responses against different antigens induced in different inbred mouse strains. Mice were immunised with two DNA vaccine immunisations two weeks apart (open circles) or primed with a DNA vaccine and two weeks later boosted with a recombinant MVA expressing the same antigen (closed circles). The strains and antigens were: C57BL/6; P. falciparum TRAP in A. DBA/2; E .coli p-galactosidase in B. BALB/c; HM epitope string CTL activity against malaria peptide (pb9) in C. DBA/2; HM epitope string CTL activity against pb9 in D. BALB/c; HM epitope string CTL activity against HIV peptide in E. DBA/2; HM epitope string CTL activity against HIV peptide in F. BALB/c; tumour epitope string CTL activity against P1A-derived peptide in G. DBA/2; tumour epitope string CTL activity against PlA-derived peptide in H. Sequences of peptide epitopes are shown in table 3. Each curve shows the data for an individual mouse.

Sporozoites can efficiently prime an immune response that is boostable by MVA Humans living in malaria endemic areas are continuously exposed to sporozoite inoculations. Malaria-specific CTL are found in these naturally exposed individuals at low levels. To address the question whether low levels of sporozoite induced CTL responses can be boosted by MVA, BALB/c mice were immunised with irradiated (to prevent malaria infection) P. berghei sporozoites and boosted with MVA. Two weeks after the last immunisation splenocytes were re-stimulated and tested for lytic activity. Two injections with 50 or 300 500 sporozoites induced very low or undetectable levels of lysis. Boosting with MVA induced high levels of peptide specific CTL. MVA alone induced only moderate levels of lysis (Figure 9).

Figure 9 shows sporozoite-primed CTL responses are substantially boosted by MVA. Mice were immunised with two low doses 50) of irradiated sporozoites in A. two high doses (300 +500) of sporozoites in B; mice were boosted with MVA.PbCSP following low-dose sporozoite priming in D; high dose sporozoite priming in E. CTL responses following immunisation with MVA.PbCSP are shown in C.

Recombinant adenoviruses as priming agent The prime-boost immunisation regimen has been exemplified using plasmid DNA and recombinant Ty-VLP as priming agent. Here an example using non-replicating adenoviruses as the priming agent is provided. Replication-deficient recombinant Adenovirus expressing E. coli p-galactosidase (Adeno-GAL) was used. Groups of BALB/c mice were immunised with plasmid DNA followed by MVA or with Adenovirus followed by MVA. All antigen delivery systems used encoded E. coli 3galactosidase. Priming a CTL response with plasmid DNA or Adenovirus and boosting with MVA induces similar levels of CTL (Figure Figure 10 shows CTL responses primed by plasmid DNA or recombinant Adenovirus and boosted with MVA. Groups of BALB/c mice were primed with plasmid DNA A or recombinant Adenovirus expressing p-galactosidase B. Plasmid DNA was administered intramuscularly, MVA intravenously and Adenovirus intradermally.

Splenocytes were restimulated with peptide TPHPARIGL [SEQ ID NO: 69] two weeks after the last immunisation. CTL activity was tested with peptide-pulsed P815 cells.

Immunogenicity of the DNA prime vaccinia boost regimen depends on the replication competence of the strain of vaccinia virus used The prime boosting strategy was tested using different strains of recombinant vaccina viruses to determine whether the different strains with strains differing in their replication competence may differ in their ability to boost a DNA-primed CTL response. Boosting with replicationdefective recombinant vaccinia viruses such as MVA and NYVAC resulted in the induction of stronger CTL responses compared to CTL responses following boosting with the same dose of replication competent WR vaccinia virus (Figure 11).

Figure 11 shows CTL responses in BALB/c mice primed with plasmid DNA followed by boosting with different recombinant vaccinia viruses. Animals were primed with pTH.PbCSP 50 pg/mouse i.m. and two weeks later boosted with different strains of recombinant vaccina viruses (106 pfu per mouse expressing PbCSP. The different recombinant vaccinia virus strains were MVA in A; NYVAC in B and WR in C. The sulpriority-of replication-impaired vaccinia strains over replicating strains was found in a further experiment. Groups of BALB/c mice (n 6) were primed with 50tg/animal of pSG2.PbCSP and 10 days later boosted i.v. with 106 ffu/pfu of recombinant MVA, NYVAC and WR expressing PbCSP. The frequencies of peptide-specific CD8+ T cells were determined using the ELISPOT assay. The frequencies were: MVA 1103 438, NYVAC 826 249 and WR 468 135. Thus using both CTL assays and ELISPOT assays as a measure of CD8 T cell immunogenicity a surprising substantially greater immunogenicity of the replicationimpaired vaccinia strains was observed compared to the replication competent strain.

The use of recombinant canary or fowl pox viruses for boosting CD8+ T cell responses Recombinant canary pox virus (rCPV) or fowl pox virus (rFVP) are made using shuttle vectors described previously (Taylor et al.

Virology 1992, 187: 321-328 and Taylor et al. Vaccine 1988, 6: 504-508).

The strategy for these shuttle vectors is to insert the gene encoding the protein of interest preceded by a vaccinia-specific promoter between two flanking regions comprised of sequences derived from the CPV or FPV genome. These flanking sequences are chosen to avoid insertion into essential viral genes. Recombinant CPV or FPV are generated by in vivo recombination in permissive avian cell lines i.e. primary chicken embryo fibroblasts. Any protein sequence of antigens or epitope strings can be expressed using fowl pox or canary pox virus. Recombinant CPV or FPV is characterised for expression of the protein of interest using antigenspecific antibodies or including an antibody epitope into the recombinant gene. Recombinant viruses are grown on primary CEF. An immune response is primed using plasmid DNA as described in Materials and Methods. This plasmid DNA primed immune response is boosted using 107 ffu/pfu of rCPV or rFPV inoculated intravenously, intradermally or intramuscularly. CD8+ T cell responses are monitored and challenges are performed as described herein.

EXAMPLE 3 Malaria Challenge Studies in Mice To assess the protective efficacy of the induced levels of CD8+ T cell response immunised BALB/c or C57BL/6 mice were challenged by intravenous injection with 2000 or 200 P. berghei sporozoites. This leads to infection of liver cells by the sporozoites.

However, in the presence of a sufficiently strong T lymphocyte response against the intrahepatic parasite no viable parasite will leave the liver and no blood-stage parasites will be detectable. Blood films from challenged 0o mice were therefore assessed for parasites by microscopy 5-12 days following challenge.

BALB/c mice immunised twice with a mixture of two plasmid DNAs encoding the CS protein and the TRAP antigen, respectively, of P.

berghei were not protected against sporozoite challenge. Mice immunised twice with a mixture of recombinant MVA viruses encoding the same two antigens were not protected against sporozoite challenge. Mice immunised first with the two recombinant MVAs and secondly with the two recombinant plasmids were also not protected against sporozoite challenge. However, all 15 mice immunised first with the two plasmid DNAs and secondly with the two recombinant MVA viruses were completely resistant to sporozoite challenge (Table 6 A and B).

To assess whether the observed protection was due to an immune response to the CS antigen or to TRAP or to both, groups of mice were then immunised with each antigen separately (Table 6 All mice immunised first with the CS plasmid DNA and secondly with the CS MVA virus were completely protected against sporozoite challenge.

Fourteen out of 16 mice immunised first with the TRAP plasmid DNA vaccine and secondly with the TRAP MVA virus were protected against sporozoite challenge. Therefore the CS antigen alone is fully protective 41 when the above immunisation regime is employed and the TRAP antigen is substantially protective with the same regime.

The good correlation between the induced level of CD8+ T lymphocyte response and the degree of protection observed strongly suggests that the CD8+ response is responsible for the observed protection. In previous adoptive transfer experiments it has been demonstrated that CD8+ T lymphocyte clones against the major CD8+ T cell epitope in the P. berghei CS protein can protect against sporozoite challenge. To determine whether the induced protection was indeed mediated by CD8+ T cells to this epitope we then employed a plasmid DNA and a recombinant MVA encoding only this nine amino acid sequence from P. berghei as a part of a string of epitopes (Table 6 (All the other epitopes were from micro-organisms other than P. berghei).

Immunisation of 10 mice first with a plasmid encoding such an epitope string and secondly with a recombinant MVA also encoding an epitope string with the P. berghei CTL epitope led to complete protection from sporozoite challenge (Table 6 Hence the induced protective immune response must be the CTL response that targets this nonamer peptide sequence.

Table 6 Results of mouse challenge experiments using different combinations of DNA and MVA vaccine Immunisation 1 Immunisation 2 No. Infected/ No. challenged %Protection

A.

DNA

MVA

DNA

MVA

Antigens used: PbCSP PbTRAP

DNA

MVA

MVA

DNA

5/5 9/10 0/5 5/5 0% 100% 0%

DNA

MVA

DNA

MVA

Control mice immunised with 1-galactosidase DNA MVA MVA DNA

B.

DNA (CSP +TRAP) DNA (CSP) DNA (TRAP) DNA (epitope) DNA (beta-gal) none MVA (CSP +TRAP) MVA (CSP) MVA (TRAP) MVA (epitope) MVA (beta-gal) none 0/10 0/10 2/16 0/11 6/7 9/10 100% 100% 88% 100% 14% Table 6 Results of two challenge experiments (A and B) using different immunisation regimes of plasmid DNA and MVA as indicated.

BALB/c mice were used in all cases. The immunisation doses were 50 pg of plasmid DNA or 106 ffu of recombinant MVA. The interval between immunisations 1 and 2 was from 14-21 days in all cases. Challenges were performed at 18-29 days after the last immunisation by i.v. injection of 2000 P. berghei sporozoites and blood films assessed at 5, 8 and 10 days post challenge. CSP and TRAP indicate the entire P. berghei antigen and epitope' indicates the cassettes of epitopes shown in table 1 containing 43 only a single P. berghei Kd-restricted nonamer CTL epitope. Note that in experiment B immunisation with the epitope string alone yields 100% protection Mice immunised twice with recombinant Ty-VLPs encoding pb9 were fully susceptible to infection. Similarly mice immunised twice with the recombinant MVA encoding the full CS protein were fully susceptible to infection. However, the mice immunised once with the Ty-VLP and subsequently once with the recombinant MVA showed an 85% reduction in malaria incidence when boosted with MVA expressing the full length CS protein, and 95% when MVA expressing the HM epitope string which includes pb9 was used to boost (Table 7).

Table 7 Results of challenge experiments using different immunisation regimes of Ty-VLPs and MVA Immunisation 1 Immunisation 2 No. Infected/No.challenged %Protection Ty-CABDHFE Ty- CABDHFE 7/8 13% Ty-CABDH MVA.PbCSP 2/13 Ty- CABDHFE MVA-NP 5/5 0% MVA.PbCSP MVA.PbCSP 6/6 0% MVA.HM Ty- CABDHFE 14/14 0% Ty- CABDHFE MVA.HM 1/21 none MVA.HM 8/8 0% none none 11/12 9% Table 7 Results of challenge experiments using different immunisation regimes of Ty-VLPs and MVA as indicated. BALB/c mice were used in all cases. Immunisations were of 50 ug of Ty-VLP or 107 ffu of recombinant MVA administered intravenously. The interval between immunisations 1 and 2 was from 14-21 days in all cases. Challenges were performed at 18-29 days after the last immunisation by i.v. injection of 2000 P: berghei sporozoites and blood films assessed at 5, 8 and 10 days .post challenge. CSP indicates the entire P. berghei antigen. Ty-VLPs carried epitope cassettes CABDH or CABDHFE as described in table 1.

MVA.HM includes cassettes CAB.

To determine whether the enhanced immunogenicity and protective efficacy observed by boosting with a recombinant MVA is unique to this particular vaccinia virus strain or is shared by other recombinant vaccinias the following experiment was performed. Mice were immunised with the DNA vaccine encoding P. berghei CS protein and boosted with either recombinant MVA encoding this antigen; (ii) recombinant wild-type vaccinia virus (Western Reserve strain) encoding the same antigen (Satchidanandam et al. 1991), or (iii) recombinant NYVAC (COPAK) virus (Lanar et al. 1996) encoding the same malaria antigen. The highest degree of protection was observed with boosting by the MVA recombinant, 80% (Table A very low level of protection was observed by boosting with the wild-type recombinant vaccinia virus and a significant level of protection, 60%, by boosting with the NYVAC recombinant. Hence the prime-boost regime we describe induces protective efficacy with any non-replicating vaccinia virus strain. Both the MVA recombinant and NYVAC were significantly (P <0.05 for each) better than the WR strain recombinant.

Table 8 Challenge data results for DNA boosted with various vaccinia strain recombinants.

Immunisation 1 Immunisation 2 No. Infected/No. challenged %Protection DNA-beta gal. MVA.NP 8/8 0% DNA-CSP MVA-CSP 2/10 DNA-CSP WR-CSP 9/10 DNA-CSP NYVAC-CSP 4/10 Table 8 Results of a challenge experiment using different immunisation regimes of plasmid DNA and various vaccinia recombinants as indicated. BALB/c mice were used in all cases. The immunisation doses were 50 pg of plasmid DNA or 106 ffu/pfu of recombinant MVA or 104 ffu/pfu of recombinant wild type (WR) vaccinia or 10' ffu/pfu of recombinant NYVAC. Because the WR strain will replicate in the host and the other strains will not, in this experiment a lower dose of WR was used.

The interval between immunisations 1 and 2 was 23 days. Challenges were performed at 28 days after the last immunisation by i.v. injection of 0o 2000 P. berghei sporozoites and blood films assessed at 7, 9 and 11 days post challenge. pbCSP indicates the entire P. berghei antigen and NP the nucleoprotein antigen of influenza virus (used as a control antigen). The first immunisation of group A mice was with the plasmid DNA vector expressing beta galactosidase but no malaria antigen.

In a further experiment shown in Table 8, mice were immunised with the DNA vaccine encoding P. berghei CS protein and boosted with either recombinant MVA encoding this antigen; (ii) recombinant WR vaccinia virus encoding the same antigen or (iii) recombinant NYVAC (COPAK) virus encoding the same malaria antigen, all at 106 ffu/pfu. A high and statistically significant degree of protection was observed with boosting with recombinant NYVAC or recombinant MVA A low and non-significant level of protection was observed by boosting with the WR recombinant vaccinia virus (Table MVA and NYVAC boosting each gave significantly more protection than WR boosting (P 0.03 and P 0.001 respectively). These data re-emphasise that non-replicating pox virus strains are better boosting agents for inducing high levels of protection.

Table 9 Influence of different recombinant vaccinia strains on protection.

Immunisation 1 Immunisation No. inf./ DNA 2 No. chall. protection CSP MVA.PbCSP 5/15 66 CSP NYVAC.PbCSP 2/15 CSP WR.PbCSP 11/15 26 p- galactosidase MVA.NP 8/8 0 Table 9 Results of challenge experiments using different immunisation regimes of plasmid DNA and replication incompetent vaccinia recombinants as boosting immunisation. BALB/c mice were used in all cases. The immunisation doses were 50 p.g of plasmid DNA or 106 io ffu/pfu of recombinant MVA or recombinant wild type (WR) vaccinia or recombinant NYVAC. The interval between immunisations 1 and 2 was 23 days. Challenges were performed at 28 days after the last immunisation by i.v. injection of 2000 P. berghei sporozoites and blood films assessed at 7, 9 and 11 days post challenge. PbCSP indicates the entire P. berghei antigen and NP the nucleoprotein antigen of influenza virus (used as a control antigen). The control immunisation was with a plasmid DNA vector expressing p-galactosidase followed by MVA.NP.

Alternative routes for boosting immune responses with recombinant

MVA

Intravenous injection of recombinant MVA is not a preferred route for immunising humans and not feasible in mass immunisations.

Therefore different routes of MVA boosting were tested for their immunogenicity and protective efficacy.

Mice were primed with plasmid DNA i.m. Two weeks later they were boosted with MVA administered via the following routes: intravenous subcutaneous intraperitoneal intramuscular and intradermal Two weeks after this boost peptide-specific CD8+ T cells were determined in an ELISPOT assay. The most effective route which induced the highest levels were i.v. and i.d inoculation of MVA. The other routes gave moderate to poor responses (Figure 12).

Figure 12 shows frequencies of peptide-specific CD8+ T cells following different routes of MVA boosting. Results are shown as the number of spot-forming cells (SFC) per one million splenocytes. Mice were primed with plasmid DNA and two weeks later boosted with MVA via the indicated routes. The number of splenocytes specific for the SYIPSAEKI [SEQ ID NO: 67] peptide was determined in INF-y ELISPOT assays two weeks after the last immunisation. Each bar represents the mean number of SFCs from three mice assayed individually.

Boosting via the i.v. route was compared with the i.d. and i.m route in a challenge experiment. The i.d route gave high levels of protection (80% protection). In the group of animals that were boosted via the i.m. route, 50% of the animals were protected. Complete protection was achieved with MVA boost administered i.v. (Table Table 10 Influence of the route of MVA administration on protective efficacy Immunisation Immunisation No. infected/ 1 2 No. protection DNA MVA challenged CSP CSP i.v. *0/20 100 CSP CSP i.d 2/10 CSP CSP i.m. 5/10 Epitope epitope i.v. 1/10 NP NP i.v. 10/10 0 *culminative data from two independent experiments Table 10 Results from challenge experiments using different routes of MVA boosting immunisation. Animals were primed by intramuscular plasmid DNA injection and two weeks later boosted with the indicated recombinant MVA (106 ffu/mouse) administered via the routes indicated.

The mice were challenged 16 days after the last immunisation with 2000 P.

berghei sporozoites and screened for blood stage parasitemia at day 8 and post challenge. Epitope indicates the polypeptide string HM.

Alternative routes of DNA priming: The use of a gene gun to prime peptide specific CD8+ T cells Gene gun delivery is described in detail in for example in Eisenbraun et al. DNA Cell Biol. 1993, 12: 791-797 and Degano et al.

Vaccine 1998, 16: 394-398.

The mouse malaria challenge experiments described so far using plasmid DNA to prime an immune response used intramuscular injection of plasmid DNA. Intradermal delivery of plasmid DNA using a biolistic device is another route to prime specific CTL responses. Plasmid DNA is coated onto gold particles and delivered intradermally with a gene gun. Groups of mice (n=10) were immunised three times at two weeks intervals with the gene gun alone (4 .g/immunisation), immunised two times with the gene gun followed by an intravenous MVA.PbCSP boost or immunised intramuscularly with 50 u.g of pTH.PbCSP and two weeks later boosted with MVA.PbCSP intravenously. Two weeks after the last immunisation the animals were challenged with 2000 sporozoites to assess protective efficacy of each immunisation regimen. In the group that received the intravenous MVA boost following two gene gun immunisations one out of ten animals developed blood stage parasitemia protection). Complete protection was observed with intramuscular DNA priming followed by MVA i.v boosting. Seven out of 10 animals that were immunised three times with the gene gun were infected. (30% protection) (Table 11).

Immunisation I Immunisation 2 Immunisation 3 No. inf./ DNA No. protection chall.

gene gun DNA gene gun DNA gene gun DNA 7/10 gene gun DNA gene gun DNA MVA.PbCSP 1/10 DNA i.m MVA.PbCSP 0/10 100 NaYve 10/10 0 Table 11 Results of challenge experiments comparing different routes o0 of DNA priming (intradermally by gene gun versus intramuscular needle injection). Groups of BALB/c mice (n=10) were immunised as indicated.

Each gene gun immunisation delivered 4 pg of plasmid DNA intraepidermally. For i.m. immunisations 50 pg of plasmid DNA were injected. Twenty days after the last immunisation mice were challenged as described previously.

Highly susceptible C57BU6 mice are protected C57BL/6 mice are very susceptible to P. berghei sporozoite challenge.

C57BL/6 mice were immunised using the DNA-MVA prime boost regime with both pre-erythrocytic antigens PbCSP and PbTRAP, and challenged with either 200 or 1000 infectious sporozoites per mouse. (Two hundred sporozoites corresponds to more than twice the dose required to induce infection in this strain). All ten mice challenged with 200 sporozoites showed sterile immunity. Even the group challenged with 1000 sporozoites, 60% of the mice were protected (Table 12). All the naive C57BL/6 mice were infected after challenge.

Table 12 Protection of C57BL/6 mice from sporozoite challenge No. animals inf./ No. challenged protection 1000 sporozoites DNA followed by MVA 4/10 Naive 5/5 0 I0 200 sporozoites DNA followed by MVA 0/10 100 Naive 5/5 0 Table 12 Results of a challenge experiment using C57BL/6 mice.

Animals were immunised with PbCSP and PbTRAP using the DNA followed by MVA prime boost regime. Fourteen days later the mice were challenged with P. berghei sporozoites as indicated.

EXAMPLE 4 Protective efficacy of the DNA-priminq/MVA-boostinq regimen in two further disease models in mice Following immunogenicity studies, the protective efficacy of the DNA-priming MVA-boosting regimen was tested in two additional murine challenge models. The two challenge models were the P815 tumour model and the-influenza A virus challenge model. In both model systems CTL have been shown to mediate protection.

P815 tumour challenges: Groups (n 10) of DBA/2 mice were immunised with a combination of DNA followed by MVA expressing a tumour epitope string or the HM epitope string. Two weeks after the last immunisation the mice were challenged intravenously with 105 P815 cells. Following this challenge the mice were monitored regularly for the development of tumour-related signs and survival.

Figure 13 shows the survival rate of the two groups of mice.

Sixty days.after challenge eight out of ten mice were alive in the group immunised with the tumour epitopes string. In the group immunised with the HM epitope string only 2 animals survived. This result is statistically significant: 2/10 vs 8/10 chi-squared 7.2. P 0.007. The onset of death in the groups of animals immunised with the tumour epitope string is delayed compared to the groups immunised with the HM epitope string.

Influenza virus challenges: Groups of BALB/c mice were immunised with three gene gun s1 immunisations with plasmid DNA, two intramuscular plasmid DNA injections, one i.m. DNA injection followed by one MVA.NP boost i.v. or two gene gun immunisations followed by one MVA.NP boost i.v. Plasmid DNA and recombinant MVA expressed the influenza virus nucleoprotein. Two weeks after the last immunisation the mice were challenged intranasally with 100 HA of influenza A/PR/8/34 virus. The animals were monitored for survival daily after challenge.

Complete protection was observed in the following groups of animals two DNA gene gun immunisations followed by one MVA.NP boost i.v., one i.m. DNA injection followed by one MVA.NP boost i.v.

two i.m. DNA injections.

In the group of animals immunised three times with the gene gun 71% of the animals survived and this difference from the control group was not significant statistically (P 0.05). In the naive group 25% of the animals survived (Figure 14) and this group differed significantly (P 0.05) for the two completely protected groups.

Figure 14 shows results of an influenza virus challenge experiment. BALB/c mice were immunised as indicated. GG gene gun immunisations, im intramuscular injection, iv intravenous injection.

Survival of the animals was monitored daily after challenge.

In a second experiment groups of 10 BALB/c mice were immunised with MVA.NP i.v. alone, three times with the gene gun, two times with the gene gun followed by one MVA.NP boost i.v. and two i.m injections of V1J-NP followed by one MVA.NP boost. Two weeks after the last immunisation the mice were challenged with 100 HA units of influenza 0o A/PR/8/34 virus.

Complete and statistically significant protection was observed in the following groups of animals: two gene gun immunisations followed by one MVA.NP boost, two i.m injections of V1J-NP followed by one MVA.NP boost.

In the group receiving one MVA.NP 30% (3 out of 10) of animals survived. In the group immunised with a DNA vaccine delivered by the gene gun three times, 70% of the animals were protected but this protection was not significantly different from the naTve controls. In this challenge experiment 40% (4 out of 10) of the naive animals survived the challenge.

EXAMPLE Immunogenicity studies in non-human primates Immunogenicity and protective efficacy of the prime boost regimen in non-human primates.

In order to show that the strong immunogenicity of the DNA priming/MVA boosting regime observed in mice translates into strong immunogenicity in primates, the regimen was tested in macaques. The vaccine consisted of a string of CTL epitopes derived from HIV and SIV sequences (Figure in plasmid DNA or MVA, denoted DNA.H and MVA.H respectively. The use of defined CTL epitopes in a polyepitope string allows testing for SIV specific CTL in macaques. Due to the MHC class I restriction of the antigenic peptides, macaques were screened for their MHC class I haplotype and Mamu-A*01-positive animals were selected for the experiments described.

Three animals (CYD, DI and DORIS) were immunised following this immunisation regimen: week 0 DNA (8pg, gene gun) week 8 DNA (8p.g, gene gun) week 17 MVA (5 x 108pfu, i.d.) week 22 MVA (5 x 108 pfu, i.d.) Blood from each animal was drawn at weeks 0, 2, 5, 8, 11, 17, 18, 19, 21, 22, 23, 24 and 25 of the experiment. The animals were monitored for induction of CTL using two different methods. PBMC isolated from each bleed were re-stimulated in vitro with a peptide encoded in the epitope string and tested for their ability to recognise autologous peptideloaded target cells in a chromium release cytotoxicity assay. Additionally, freshly isolated PBMC were stained for antigen specific CD8+ T cells using tetramers.

Following two gene gun immunisations very low levels of CTL were detected using tetramer staining (Figure 15). Two weeks after the first MVA boosting, all three animals developed peptide specific CTL as detected by tetramer staining (Figure 15). This was reflected by the detection of moderate CTL responses following in vitro restimulation (Figure 16, week 19). The second boost with MVA.H induced very high levels of CD8+, antigen specific T cells (Figure 15) and also very high levels of peptide specific cytotoxic T cells (Figure 16, week 23).

Figure 15 shows detection of SIV-specific MHC class Irestricted CD8+ T cells using tetramers. Three Mamu-A*A01-positive macaques were immunised with plasmid DNA (gene gun) followed by MVA boosting as indicated. Frequencies of Mamu-A*A01/CD8 double-positive T cells were identified following FACS analysis. Each bar represents the percentage of CD8+ T cells specific for the Mamu-A*01/gag epitope at the indicated time point. One percent of CD8 T cells corresponds to about 5000/106 peripheral blood lymphocytes. Thus the levels of epitope-specific CD8 T cells in the peripheral blood of these macaques are at least as high as the levels obvserved in the spleens of immunised and protected mice in the malaria studies.

Figure 16 shows CTL induction in macaques following DNA/MVA immunisation. PBMC from three different macaques (CYD, DI and DORIS) were isolated at week 18, 19 and 23 and were restimulated with peptide CTPYDINQM [SEQ ID NO: 54] in vitro. After two restimulations with peptide CTPYDINQM [SEQ ID NO: 54] the cultures were tested for their lytic activity on peptide-pulsed autologous target cells.

Strong CTL activity was observed.

EXAMPLE 6 Immunoqenicity and Challenge Studies in Chimpanzees To show that a similar regime of initial immunisation with plasmid DNA and subsequent immunisation with recombinant MVA can be effective against Plasmodium falciparum malaria in higher primates an immunisation and challenge study was performed with two chimpanzees.

Chimp H1 received an initial immunisation with 500 p[g of a plasmid expressing Plasmodium falciparum TRAP from the CMV promoter without intron A, CMV-TRAP. Chimp H2 received the same dose of CMV-LSA-1, which expresses the C-terminal portion of the LSA-1 gene of P. falciparum.

Both chimps received three more immunisations over the next 2 months, but with three plasmids at each immunisation. H1 received CMV-TRAP as before, plus pTH-TRAP, which expresses TRAP using the CMV promoter with intron A, leading to a higher expression level. H1 also received RSV- LSA-1, which expresses the C-terminal portion of LSA-1 from the RSV promoter. H2 received CMV-LSA-1, pTH-LSA-1 and RSV-TRAP at the second, third and fourth immunisations. The dose was always 500 -Lg of each plasmid.

It was subsequently discovered that the RSV plasmids did not express the antigens contained within them, so H1 was only immunised with plasmids expressing TRAP, and H2 with plasmids expressing LSA-1.

Between and following these DNA immunisations assays of cellular immune responses were performed at several time points, the last assay being performed at three months following the fourth DNA immunisation, but no malaria-specific T cells were detectable in either ELISPOT assays or CTL assays for CD8+ T cells.

Both animals were subsequently immunised with three doses of 108 ffu of a recombinant MVA virus encoding the P. falciparum TRAP antigen over a 6 week period. Just before and also following the third recombinant MVA immunisation T cell responses to the TRAP antigen were detectable in both chimpanzees using an ELISPOT assay to whole TRAP protein bound to latex beads. This assay detects both CD4+ and CD8+ T cell responses. Specific CD8+ T responses were searched for with a series of short 8-11 amino acid peptides in both immunised chimpanzees. Such analysis for CD8+ T cell responses indicated that CD8+ T cells were detectable only in the chimpanzee H1. The target epitope of these CD8+ T lymphocytes was an 11 amino acid peptide from TRAP, tr57, of sequence KTASCGVWDEW [SEQ ID NO: 78]. These CD8+ T cells from H1 had lytic activity, against autologous target cells pulsed with the tr57 peptide and against autologous target cells infected with the recombinant PfTRAP-MVA virus. A high precursor frequency of these specific CD8+ T cells of about 1 per 500 lymphocytes was detected in the peripheral blood of this chimpanzee H1 using an ELISPOT assay two months following the final MVA immunisation. No specific CD8+ T cell response was clearly detected in the chimpanzee H2, which was not primed with a plasmid DNA expressing TRAP.

Two months after the third PfTRAP-MVA immunisation challenge of H1 and H2 was performed with 20,000 sporozoites, a number that has previously been found to yield reliably detectable blood stage infection in chimpanzees 7 days after challenge (Thomas et al. 1994 and unpublished data). The challenge was performed with the NF54 strain of Plasmodium falciparum. This is of importance because the TRAP sequence in the plasmid DNA and in the recombinant MVA is from the T9/96 strain of P. falciparum which has numerous amino acid differences to the NF54 TRAP allele (Robson et al. 1990). Thus, this sporozoite challenge was performed with a heterologous rather than homologous strain of parasite. In the chimpanzee H2 parasites were detectable in peripheral blood as expected 7 days after sporozoite challenge using in vitro parasite culture detection. However, in H1 the appearance of blood stage parasites in culture from the day 7 blood samples was delayed by three days consistent with some immune protective effect against the liverstage infection. In studies of previous candidate malaria vaccines in humans a delay in the appearance of parasites in the peripheral blood has been estimated to correspond to a substantial reduction in parasite density in the liver (Davis et al. 1989). Thus the chimpanzee HI, immunised first with P. falciparum TRAP plasmid DNA and subsequently with the same antigen expressed by a recombinant MVA virus showed a strong CD8+ T lymphocyte response and evidence of some protection from heterologous sporozoite challenge.

DISCUSSION

These examples demonstrate a novel regime for immunisation against malaria which induces high levels of protective CD8+ T cells in rodent models of human malaria infection. Also demonstrated is an unprecedented complete protection against sporozoite challenge using subunit vaccines (36 out of 36 mice protected in Table 6 using DNA priming and MVA boosting with the CS epitope containing vaccines).

Induction of protective immune responses using the DNA priming/MVA boosting regimen was demonstrated in two additional mouse models of viral infection influenza A model and cancer (P815 tumour model). More importantly for vaccines for use in humans this immunisation regimen is also highly immunogenic for CD8+ T cells in primates. Strong SIV-gagspecific CTL were induced in 3 out of 3 macaques with plasmid DNA and MVA expressing epitope strings. The levels induced are comparable to those found in SIV-infected animals. The data from the chimpanzee studies indicate that the same immunisation regime can induce a strong CD8+ T lymphocyte response against P. falciparum in higher primates with some evidence of protection against P. falciparum sporozoite challenge.

Ty-VLPs have previously been reported to induce good levels of CD8+ T cell responses against the P. berghei rodent malaria (Allsopp et al. 1995) but alone this construct is not protective. It has now been found that subsequent immunisation with recombinant MVA boosts the CD8+ T cell response very substantially and generates a high level of protection (Table 7).

Recombinant MVA viruses have not been assessed for efficacy as malaria vaccines previously. Recombinant MVA alone was not significantly protective, nor was priming with recombinant MVA followed by a second immunisation with recombinant plasmid DNA. However, a second immunisation with the recombinant MVA following an initial immunisation with either Ty-VLPs or plasmid DNA yielded impressive levels of protection. Non-recombinant MVA virus has been safely used to vaccinate thousands of human against smallpox and appears to have an excellent safety profile. The molecular basis of the increased safety and immunogenicity of this strain of vaccinia virus is being elucidated by detailed molecular studies (Meyer et al. 1991; Sutter at al. 1994).

Plasmid DNA has previously been tested as a malaria vaccine for the P. yoelii rodent malaria. High levels of, but not complete, protection is seen in some strains but in other strains of mice little or no protection was observed even after multiple immunisations (Doolan et al.

0o 1996). Although plasmid DNA has been proposed as a method of immunisation against P. falciparum, success has not previously been achieved. The evidence provided here is the first evidence to show that plasmid DNA may be used in an immunisation regime to induce protective immunity against the human malaria parasite P. falciparum.

A similar regime of immunisation to the regime demonstrated herein can be expected to induce useful protective immunity against P.

falciparum in humans. It should be noted that five of the vaccine constructs employed in these studies to induce protective immunity in rodents or chimpanzees contain P. falciparum sequences and could therefore be used for human immunisation against P. falciparum. These are: 1. The P. falciparum TRAP plasmid DNA vaccine. 2. The P.

falciparum TRAP recombinant MVA virus. 3. The Ty-VLP encoding an epitope string of numerous P falciparum epitopes, as well as the single P.

berghei CTL epitope. 4. The plasmid DNA encoding the same epitope string as 3. 5. The recombinant MVA encoding the longer HM epitope string including many of the malaria epitopes in 3 and 4. Similarly the plasmid DNAs and MVA encoding HIV epitopes for human class I molecules could be used in either prophylactic or therapeutic immunisation against HIV infection.

These studies have provided clear evidence that a novel sequential immunisation regime employing a non-replicating or replicationimpaired pox virus as a boost is capable of inducing a strong protective CD8+ T cell response against the malaria parasite. The examples demonstrate clearly a surprising and substantial enhancement of CD8+ T cell responses and protection compared to replicating strains of pox viruses. Because there is no reason to believe that the immunogenicity of CD8+ T cell epitopes from the malaria parasite should differ substantially from CD8+ T cell epitopes in other antigens it is expected that the immunisation regime described herein will prove effective at generating CD8+ T cell responses of value against other diseases. The critical step in this immunisation regimen is the use of non-replicating or replicationimpaired recombinant poxviruses to boost a pre-existing CTL response.

We have shown that CTL responses can be primed using different antigen delivery systems such as a DNA vaccine i.d. and i.m, a recombinant Ty- VLP, a recombinant adenovirus and irradiated sporozoites. This is supported by the data presented on the generation of a CD8+ T cell response against HIV, influenza virus and tumours. Amongst several known examples of other diseases against which a CD8+ T cell immune response is important are the following: infection and disease caused by the viruses HIV, herpes simplex, herpes zoster, hepatitis C, hepatitis B, influenza, Epstein-Barr virus, measles, dengue and HTLV-1; by the bacteria Mycobacterium tuberculosis and Listeria sp.; and by the protozoan parasites Toxoplasma and Trypanosoma. Induction of protective CTL responses against influenza A virus has been demonstrated in Figure 14.

Furthermore, the immunisation regime described herein is expected to be of value in immunising against forms of cancer where CD8+ T cell responses plays a protective role. The induction of protective CTL responses using the DNA prime MVA boost regime against tumours is shown in Figure 13. Specific examples in humans include melanoma, cancer of the breast and cancer of the colon.

References 1. Nardin EH Nussenzweig RS. T cell responses to pre-erythrocytic stages of malaria: role in protection and vaccine development against pre-erythrocytic stages. Annual Review of Immunology (1993) 11: 687-727.

io 2. Hill AVS, Allsopp, CEM, Kwiatkowski D, Anstey NM, Twumasi P, Rowe PA, Bennett S, Brewster D, McMichael AJ, Greenwood BM. (1991) Common West African HLA antigens are associated with protection from severe malaria. Nature 352: 595-600 3. Aidoo M, Lalvani A, Allsopp CEM, et al. Identification of conserved antigenic components for a cytotoxic T lymphocyte-inducing vaccine against malaria. The Lancet. (1995) 345: 1003-1007.

4. Wizel B, Houghten RA, Parker KC, Coligan JE, Church P, Gordon DM, Ballou WR, Hoffman-SL. Irradiated sporozoite vaccine induces HLA- B8-restricted cytotoxic T lymphocyte responses against two overlapping epitopes of the Plasmodium falciparum sporozoite surface protein 2. J. Exp Med. (1995) 182: 1435-45.

Lalvani A, Aidoo M, Allsopp CE, Plebanski M, Whittle HC, Hill AV. An HLA-based approach to the design of a CTL-inducing vaccine against Plasmodium falciparum. Research in Immunology (1994) 145: 461-8.

6. Seguin MC, Klotz FW, Schneider I, Weir JP, Goodbary M, Slayter M, Raney JJ, Aniagolu JU, Green SJ. Induction of nitric oxide synthase protects against malaria in mice exposed to irradiated Plasmodium berghei infected mosquitoes: involvement of interferon gamma and CD8+ T cells. J. Exp. Med. (1994) 180: 353-8.

7. Thomas AW, Slierendregt B, Mons B, Druilhe P. Chimpanzees and supporting models in the study of malaria pre-erythrocytic stages.

Mem. Inst. Oswaldo Cruz. (1994) 89 Suppl 2: 111-4.

8. Sedegah M, Hedstrom R, Hobart P, Hoffman SL. Protection against malaria by immunization with plasmid DNA encoding circumsporozoite protein. Proc. Natl. Acad. Sci USA. (1994) 91: 9866-70 9. Li S, Rodrigues M, Rodriguez D, Rodriguez JR, Esteban M, Palese P, Nussenzweig RS, Zavala F. Priming with recombinant influenza virus followed by administration of recombinant vaccinia virus induces CD8+ T-cell-mediated protective immunity against malaria.

Proc. Natl. Acad. Sci. USA. (1993) 90: 5214-8.

Lanar DE, Tine JA, de-Taisne C, Seguin MC, Cox WI, Winslow JP, Ware LA, Kauffman EB, Gordon D, Ballou WR, Paoletti E, Sadoff JC. Attenuated vaccinia virus-circumsporozoite protein recombinants confer protection against rodent malaria. Infection and Immunity (1996) 64: 1666-71.

11.Ogg GS, Jin X, Bonhoeffer S, Dunbar PR, Nowak MA, Monard S, Segal JP, Cao Y, Rowland-Jones SL, Cerundolo et al.

Quantitation of HIV-1-specific cytotoxic T lymphocytes and plasma load of viral RNA. Science (1998) 279: 2103-6 12. Ada G. Do cytotoxic T lymphocytes clear some HIV/SIV infections? Journal of Medical Primatology (1996) 25: 158-62.

13. Gallimore A, Cranage M, Cook N, Almond N, Bootman J, Rud E, Silvera P, Dennis M, Corcoran T, Stott J et-al Early suppression of SIV replication by CD8+ nef-specific cytotoxic T cells in vaccinated macaques. Nature Medicine (1995) 1: 1167-73.

14. Mayr A, Hochstein-Mintzel V, Stickl H. Infection (1975) 33: 6-14.

Mayr A, Zentralbl Veterinarmed B (1976) 23: 417-30.

16. Mayr A, Sticki Muller HK, Danner K, Singer H. Zentralbl Bakteriol B.

(1978) 167: 375- 17. Stickl H, Hochstein-Mintzel V, Mayr A, Huber HC, Schafer H, Holzner A. Dtsch Med Wochenschr. (1974) 99: 23866-922.

S18. Mahnel H, Mayr A. Berl Munch Tierarztl Wochenschr(1994) 107: 253- 6.

19. Meyer H, Sutter G, Mayr A. J Gen Virol. (1991) 72: 1031-8.

Altenburger W, Suter CP, Altenburger J. Arch Virol. (1989) 105: 15-27.

21. Sutter G, Ramsey-Ewing A, Rosales R, Moss B. J. Virol. (1994) 68: io 4109-16.

22. Sutter G, Moss B. Proc Natl Acad Sci U S A (1992) 89: 10847-51.

23. Sutter GW, Wyatt LS, Foley PL, Bennink JR, Moss B. Vaccine (1994).

12: 1032-40.

24. Hirsch VM, Goldstein S, Channock R, et al. Channock R. ed Vaccines 95. Cold Spring Harbor Laboratory Press, (1995) 195-200.

Hirsch VM, Fuerst TR, Sutter G, et al. J. Virol. (1996) 70: 3741-52.

26. Moss B, Carroll MW, Wyatt L, et al. In: Anonymous, ed. Vaccines: Novel strategies in design and production. Plenum Pres (1995).

27. Symons JA, Alcami A, Smith GL. Cell. (1995) 81: 551-60.

28. Alcami A, Smith GL. J. Virol. (1995) 69: 4633-9.

29. Alcami A, Smith GL. Cell. (1992) 71: 153-67.

Graham KA et al. Virology (1997) 229: 12-24.

31. Alcami A, Smith GL. Proc. Natl. Acad. Sci. USA (1996) 93:11029-34.

32. Blanchard TJ, Rowland-Jones S, Gotch F, McMichael AJ, Smith GL.

Lancet (1997) 348: 1741 33. McLean CS, Erturk M, Jennings R, Challanain DN, Minson AC, Duncan I, Boursnell ME, Inglis SC. J. Infect. Dis. (1994) 170(5): 1100-9.

34. Davis NL, Brown KW, Johnston RE, J. Virol, (1996) 70(6): 3781-7.

Layton GT, Harris SJ, Myhan J, West D, Gotch F, Hill-Perkins M, Cole JS, Meyers N, Woodrow S, French TJ, Adams SE, Kingsman AJ.

Induction of single and dual cytotoxic T-lymphocyte responses to viral proteins in mice using recombinant hybrid Ty-virus-like particles. Immunology (1996) 87: 171-8.

36. McGrory-WJ; Bautista-DS; Graham-FL A simple technique for the rescue of early region I mutations into infectious human adenovirus type 5. Virology 19 8 8 163: 614-7 37. Rodrigues EG, Zavala F, Eichinger D, Wilson JM, Tsuji M. Single immunizing dose of recombinant adenovirus efficiently induces CD8+ T cell-mediated protective immunity against malaria. Journal of Immunology (1997) 158: 1268-74.

38. Davis HL, Michel ML, Whalen RG. DNA-based immunization induces continuous secretion of hepatitis B surface antigen and high levels of circulating antibody. Human Molecular Genetics (1993) 2: 1847- 51.

39. Miyahira Y, Murata K, Rodriguez D et al. Quantification of antigen specific CD8+ T cells using an ELISPOT assay. J Immunol Methods (1995) 18: 45-54.

Allsopp CEM, Plebanski M, Gilbert S et al. (1996) Comparison of numerous delivery systems for the induction of cytotoxic T lymphocytes. European Journal of Immunology (1996) 26:1951- 1959.

41. Rodrigues EG, Zavala F, Eichinger D, Wilson JM, Tsuji M. Single immunizing dose of recombinant adenovirus efficiently induces CD8+ T cell-mediated protective immunity against malaria. J. Immunol.

(1997) 158: 1268-74.

42. Schodel F, Wirtz R, Peterson D, Hughes J, Warren R, Sadoff J, Milich D.

Immunity to malaria elicited by hybrid hepatitis B virus core particles carrying circumsporozoite protein epitopes. J. Exp. Med. (1994) 180: 1037-46.

43. Satchidanandam V, Zavala F, Moss B. Studies using a recombinant vaccinia virus expressing the circumsporozoite protein of Plasmodium berghei. Mol. Biochem. Parasitol. (1991) 48: 89-99.

44. Davis JR; Murphy JR; Baqar S; Clyde DF; Herrington DA; Levine MM.

Estimate of anti-Plasmodium falciparum sporozoite activity in humans vaccinated with synthetic circumsporozoite protein (NANP)3. Transactions of the Royal Society for Tropical Medicine and Hygiene. (1989) 83: 748-50.

Meyer, Sutter, G. and Mayr, A. (1991). Mapping of deletions in the 0o genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J. Gen. Virol. 72: 1031-38.

46. Sutter G, Wyatt LS, Foley PL, Bennink JR, Moss B. A recombinant vector derived from the host range-restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine (1994) 12: 1032-40.

47. Doolan DL, Sedegah M, Hedstrom RC, Hobart P, Charoenvit Y, Hoffman, SL Circumventing genetic restriction of protection against malaria with multigene DNA immunization: CD8+ cell-, interferon gamma-, and nitric oxide-dependent immunity. J. Exp. Med. (1996) 183: 1739-46.

48. Muller HM, Reckmann I, Hollingdale MR, Bujard H, Robson KJ, Crisanti A. Thrombospondin related anonymous protein (TRAP) of Plasmodium falciparum binds specifically to sulfated glycoconjugates and to HepG2 hepatoma cells suggesting a role for this molecule in sporozoite invasion of hepatocytes. EMBO-J. (1993) 12: 2881-9.

49. Chakrabarti S et al. Mol. Cell. Biol. (1995) 5: 3403-9.

Morrison HG, Bauer SP, Lange JV, Esposito JJ, McCormick JB, Auperin DD Protection of guinea pigs from Lassa fever by vaccinia virus recombinants expressing the nucleoprotein or the envelope glycoproteins of Lassa virus. Virology. (1989) 171: 179-88.

51. Mackett M et al. J. Virol. (1984) 49: 857-864.

52. Carroll MW and Moss BE, Biotechnology (1995) 19: 352-4.

65/1 INDICATIONS RELATING TO A DEPOSITED MICROORGANISM (PCT Rule 13bis) A. The indications made below relate to the microorganism referred to in the description on page 12 ,line 10 13 B. IDENTIFICATION OF DEPOSIT Further deposits are identified on an additional sheet .J Name of depositary institution European Collection of Animal Cell Cultures (CAMR) Address of depositary institution (including postal code and country) Salisbury Wiltshire SP4 OJG United Kingdom Date of deposit Accession Number June 1997 V97060511 C. ADDITIONAL INDICATIONS (leave blank if not applicable) This information is continued on an additional sheet fX In respect of all designated States to which such action is possible and to the extent that it is legal.ly permissible under the law of the designated State, it is requested that a sample of the: deposited microorganism be'made available only by the issue thereof to an independent expert, in accordance with the relevant patent legislation, e.g. EPC Rule 28(4), UK Patent Rules 1995, Schedule 2, Paragraph 3, Australian Regulation 3.25(3), Danish D. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE (if the indications are nofor all designated Stater) E. SEPARATE FURNISHING OF INDICATIONS (leave blank if not applicable) The indications listed below will be submitted to the International Bureau later (specify thegeneal nature oftheindications 'Accession Number of Deposit) For receiving Office use only S- This sheet was received with the international application Authorized officer Form PCT/RO/134 (July 1992) SFor International Bureau use only L This sheet was received by the International Bureau on: 07 EfIcEr IS_ Authorized officer 65/2 INDICATIONS RELATING TO A DEPOSITED MICROORGANISM C. ADDITIONAL INDICATIONS (continued) Patents Act Sections 22 and 33(3), Icelandic Patents Act Sections 22 and 33(3), Norwegian Patents Act Sections 22 and 33(3) and generally similar provisions mutatis mutandis for any other designated State.

Claims (8)

1. A medicament when used for boosting a naturally primed CD8+ T cell response against at least one target antigen in an individual naturally exposed to the target antigen, comprising a source of one or more CD8+ T cell epitopes of the target antigen, wherein the source of CD8+ T cell epitopes is a non-replicating or a replication-impaired recombinant poxvirus vector, and a pharmaceutically acceptable carrier.

2. The medicament according to claim 1, wherein the vector is an avipox virus vector.

3. The medicament according to claim 2, wherein the vector is a fowlpox virus vector.

4. The medicament according to claim 1, wherein the vector is a vaccinia virus vector such as MVA. The medicament according to any one of claims 1 to 4, wherein the target antigen is a malaria antigen.

6. The medicament according to any one of claims 1 to 5, wherein the individual is a primate.

7. A method of boosting a naturally primed CD8+ T cell immune response in an individual naturally exposed to a target antigen, which method comprises administering a medicament according to any one of claims 1 to 6.

8. The use of a recombinant non-replicating or replication-impaired pox virus vector encoding one or more CD8+ T cell epitopes of a target antigen in the manufacture of a medicament according to any one of claims 1 to 6.

9. The use of an MVA encoding one or more CD8+ T cell epitopes of a target antigen in the manufacture of a medicament when used for boosting a naturally primed CD8+ T cell immune response against the target antigen. The use according to claim 8 or claim 9 wherein the CD8+ T cell immune response is a primate CD8+ T cell immune response. OXXON THERAPEUTICS LIMITED WATERMARK PATENT TRADE MARK ATTORNEYS P16646AU02