WO2003062408A1 - Virus-like particles - Google Patents

Abstract

The present invention relates to virus-like particles derived from flaviviruses and to methods for generating the same. These particles are useful in diagnostic applications, and as components of vaccines directed at preventing the incidence of disease.

Description

VIRUS-LIKE PARTICLES

The present invention relates to virus-like particles derived from flaviviruses. These particles are useful in diagnostic applications, and as components of vaccines directed at preventing the incidence of disease. Members of the Flaviviridae family of viruses infect an estimated 30-50 million people annually in many regions of the world. They are positive-strand RNA viruses that can be classed according to their associated vectors; mosquito-vectored, tick-vectored or unknown vector. There are about 70 identified members of the Flavivirus genus including many medically important viruses such as West Nile virus (WNV), Japanese encephalitis virus (JE), Tick-borne encephalitis (TBE), Yellow fever (YF) and Dengue viruses.

Current diagnosis of WNV, TBE and other flaviviruses depends primarily upon serological methods, which require specialist and expensive facilities that are not available in many countries (for instance in Africa/ Asia). Where available, commercially available ELISA-based diagnostic kits are based upon purified attenuated or inactivated virus as antigens; consequently such kits are expensive to manufacture. Simple and cheap diagnosis of these diseases would allow the generation of large epidemiological studies that can target treatment, where available, effectively.

In summer 1999, an outbreak of human encephalitis cases occurred in north-eastern United States (New York state). Originally thought to be St. Louis encephalitis (SLE), the causative agent was later identified as West Nile virus (4). A widespread US WNV surveillance programme was initiated during 2000 and showed a geographical expansion of the virus to 12 states as well as increased epizootic activity (22). During 2001 the virus showed further geographical expansion, as far North as Ontario in Canada, where 122 WNV positive bird samples were found as of November 2001 (source- Ontario government statistics), and as far south as Florida where both human and equine infections were reported (2). The epidemiological trend of WNV (NY99) infections in the U.S. is unknown, although based upon the last three years' experience, it is likely to continue its emergent behaviour. Currently there is no approved vaccine or commercially available diagnostic kit for WNV. TBE infects approximately 200 thousand people throughout Europe and Asia. In endemic areas of Central and Eastern Europe, the former USSR and Finland, TBE virus infection represents a severe public health problem (23). Although TBE vaccines are available, one based upon an attenuated virus and the other an inactivated virus, the detection of specific antibodies against TBE using an ELISA (enzyme-linked immunosorbent assay) system is a prerequisite to establish the prevalence of TBE in patients or to prove seroconversion after active immunization (17). Commercially-available TBE virus-specific ELISA kits are available based upon inactivated whole virus (e.g. Immunozym FSME IgG (Baxter AG) and Enzygnost Anti-TBE virus (Dade Behring Marburg GmbH). There are currently six commercially available IgG ELISA kits, although their sensitivity ranges from 73 to 99%, cross-reactivity with other flaviviruses results in significantly lower specificity rates ranging from just 14 to 81% (28). However because the viruses are classified as high-risk pathogens, the specialist facilities required to make the antigen, mean that the kits are very expensive (23). JEV infects hundreds of thousands of people annually throughout Asia. The current JEV vaccines are formalin-inactivated virion fractions purified from JEV-infected mouse brains. The high cost of these vaccines and incidences of allergic responses in humans are causes for concern (19). Several commercially available ELISA kits exist for the diagnosis of JEV, although at a high cost. It has been reported previously that particles lacking the nucleocapsid are often released during flavivirus infections (31). A comparison of four different constructs of JEV containing different lengths of ORF from PrM to NS2B and expressed in vaccinia virus, showed that recombinants containing both E and PrM produced virus-like particles (VLPs) similar to the wild-type virions (24). Similar PrM/E constructs for the flaviviruses TBE, Louping-Ill (LI), JEV and yellow fever (YF) have been recombinantly expressed in various mammalian cell-lines using the vaccinia virus expression system. Each of these constructs were shown to induce some levels of protection against wild-type virus challenge using mice models (8, 13, 19, 30, 38). A vaccinia construct containing WNV PrM and E genes appeared to result in the formation of VLPs, although the efficiency of particle formation was described as a 'rare observation' and that efficiency of formation 'must be very low' for this virus (39). VLPs have also been observed in mammalian cells electroporated with plasmid DNA containing PrM/E constructs under the control of the human cytomegalovirus early gene promoter. VLPs for JEV, TBE and WNV have been produced using these techniques (1, 7). Whilst the VLPs proved to be good candidates as ELISA antigens, they did not use the VLPs in protection studies. Rather they used the plasmid DNA as an immunogen. So called 'DNA vaccines' have been produced using the PrM E gene cassette for several other flaviviruses including SLE, dengue virus and Murray Valley encephalitis (5, 18, 29). All of these DNA vaccines have been reported to protect animals from viral challenge, however concerns over the technology, in particular the potential for recombination to occur when replicating in mammalian cells, means that its use as a potential vaccine for humans is unlikely without extensive testing to assess the risks (27).

Whilst the authors of these and similar studies have alluded to the vaccine potential of VLPs against flaviviruses, mammalian expressed proteins are not practical for producing the large quantities of protein required for a useful vaccine or for large-scale diagnostics. In addition there is genuine concern about the use of these viral vectors as vaccine candidates due to their oncogenic potential and other complications (10, 38). There thus exists a great need for a method for the generation of agents that are useful for both the diagnosis and prevention of flavivirus-mediated disease.

Summary of the invention According to the invention there is provided a method for the production of virus-like particles (VLPs) from a flavivirus, said method comprising the steps of: a) expressing a construct comprising the PrM and envelope (E) genes of a flavivirus, or a variant of the PrM and/or envelope (E) genes, in a baculoviral expression cassette and cloned under the control of a promoter in insect cells, wherein a region of nucleic acid encoding a Furin cleavage site is included at the junction of the M/E genes; b) culturing the insect cells for a sufficient period of time to allow production of baculovirus particles; and c) separating the VLPs from the baculoviral particles and the insect cells. The virus-like particles (VLPs) produced according to the invention have thus been generated using the baculovirus expression system. In contrast to mammalian expressed proteins, proteins produced using the this system have the advantage of high protein yields as well as the scale-up of insect cells in culture being largely perfected allowing the production of large quantities of recombinant protein a reality. These VLPs are suitable for use as diagnostic antigens, particularly in methods such as enzyme linked immunosorbent assay (ELISA) and in lateral-flow rapidtest-type kits. The VLPs may also be used in vaccines.

The inventor has discovered that the baculovirus expression system is advantageous for the production of flaviviral VLPs. This expression system has been used previously for the production of various individual protein antigens (14, 15, 23, 38). However, to date, it was not appreciated that this expression system would be useful or even effective in the expression of a flaviviral PrM/E cassette or in VLP production. The inventor has discovered that the inclusion of a furin cleavage site at the PrM/E junction greatly enhances post-translational modification resulting in greatly increased antigenicity of the VLPs produced.

The RNA genome of a flavivirus is enclosed by a capsid (C) protein that is surrounded by a host-derived lipid membrane containing the glycosylated viral membrane (M) and envelope (E) proteins. Seven non-structural proteins (NS1, NS2a, NS2b, NS3, NS4a, NS4b, NS5) provide the replicative and proteolytic functions for virus replication (26). The E protein is the dominant antigenic determinant of humoral and cellular immune responses in flavivirus infections (12, 38). Earlier attempts at cloning the E gene showed that they did not necessarily exhibit the same activity or conformation as the native protein (3). It is now believed that correct folding of the flavivirus E protein also requires the co-ordinated synthesis of PrM protein (38).

Wild-type virion assembly occurs first by proteolytic cleavage of the polyprotein at the M/E cleavage site by a cellular signalase found in the endoplasmic reticulum, to form immature virions composed of heterodimeric PrM and E proteins (11). PrM in turn is cleaved by a cellular protease (Furin) in acidic particles of the trans-Golgi network that leads to the release of mature particles (34). In a previous study, the expression of WNV PrM/E using plasmid DNA electroporated into COS cells was found to produce VLPs that from Western analysis and total protein staining, generate E protein of a similar size to wild-type virus E protein (7). However, our results show when PrM/E is expressed in insect cells using the baculovirus expression system, although particles are clearly formed, the efficiency of cleavage of the M/E polyprotein is very low, as shown by western blot analysis. The major antigenic species found in our VLPs is a 60 kDa species, that appears to be uncleaved M/E (E is about 50 kDa, PrM is about 30 kDa and M about 10 kDa).

The results of the ELISA assays carried out on the WNVwt VLPs (herein termed 3-11) produced according to the method of the invention suggest that the resulting M/E structures are not nearly as antigenic as the wild-type virions as detected by monoclonal antibodies raised against epitopes of the E protein of the virus. This may be explained by the fact that the signalase recognition site, unlike the Pr/M Furin recognition site, is conformational rather than sequence specific, and that insect cell signalases may not recognise mammalian signalase motifs. Indeed this may provide a clue as to why flaviviruses are replicative-competent in mammalian cells but not in insect cells (38). To get around this problem, the inventor designed a novel baculovirus construct containing the PrM/E gene with an extra Furin cleavage site at the M/E junction (herein termed IF-

Titration of the 1F-1 and 3-11 antigens against mouse polyclonal (TC217) and human Dengue positive sera showed 1F-1 to be a much more sensitive antigen than 3-11 in direct IgG ELISA assays (more than 4 times at a 1 in 800 dilution using TC127). A P/N ratio of greater than 5 was obtained with 1F-1 even at a 1 in 3200 dilution (equivalent to ~60 ng per well) with mouse sera. Values obtained with 1F-1 were greater, at equal protein concentrations, than those obtained with WNV itself. The VLPs were shown to be morphologically similar if not identical to the wild-type virions. The VLPs are secreted and can be purified relatively simply by centrifugation.

VLPs produced according to the method of the invention have been found to be highly sensitive in human sera diagnosis against the presence of flaviviral IgG, and distinguished between low titre YF vaccine IgG and sera from individuals whom had developed a flaviviral disease. The VLPs reacted strongly against the sera from different animals that had been infected with WNV, including mice, rabbits and guinea-pigs, suggesting that the VLPs can be used diagnostically with different species. This is important in the diagnostic surveillance of WNV. The VLPs were also shown to be highly specific for WNV, as demonstrated by testing with a panel of monoclonals raised against the E protein of several flaviviruses.

Furthermore, VLPs generated according to the method of the invention were shown to be stable in dried form for a prolonged period of time at a range of temperatures from -20°C to 40°C. These data suggest that VLPs generated in this manner are eminently suitable candidates as diagnostic antigens suitable for diagnosis in humans, as well as in animals and probably in birds.

VLPs produced according to the invention have also been shown to have vaccine potential, as demonstrated by their ability to protect mice against a lethal dose of the virus. According to the method of the invention, the VLPs may be derived from any member of the flavivirus genus, including but not limited to the following: WNV; TBE; YF; JE; Louping-Ill; Kunjin virus; St. Louis encephalitis virus; Dengue virus or from Murray Valley encephalitis virus. Preferably, the VLPs are derived from the West Nile Virus. The applicability of the said method to produce VLPs in insect cells from all members flavivirus genus is inferred by the observation that the method has been used by the inventor to produce WNV, TBE and JE VLPs.

The PrM and E genes from all these flaviviruses are known in the art. Accordingly, the skilled reader, imbued with the present teaching, will be capable of practising the invention for all these flaviviruses without the need for inventive skill. Sequences of the relevant genes from the above-listed flaviviruses may be found in publicly-available databases such as GeneBank (http://www.ncbi.nlm.nih.gov), EMBL (http://www.ebi.ac.uk) and DDBJ (http://www.ddbj.nig.ac.jp). Preferably, the entire PrM and E genes are used, although fragments of these proteins may be used, provided that intact VLPs are still generated. Furthermore, alterations from the wild type sequence may be allowed in the sequences of the PrM and E genes, including insertions and deletions, particularly if the insertions or deletions only involve a few amino acids, e.g., under thirty, and preferably under ten, and do not remove or displace amino acids which are critical to a functional conformation, e.g., cysteine residues. Substitutions, particularly conservative amino acid substitutions, may also be allowed in the sequences of the proteins. Such altered PrM and E genes are referred to herein as "variants" of the wild type proteins.

It may also be preferable to incorporate sequence from neighbouring genes in the sequence of the viral polyprotein, such as the capsid (C) gene that abuts the PrM gene, and the NS1 gene that abuts the E gene. According to the method of the invention, a cleavage site for the protease Furin should be incorporated into the construct that is used to express the VLPs according to the invention. This may conveniently be achieved by including a region of nucleic acid encoding a Furin cleavage site at the junction of the PrM/E genes. Furin is a cellular protease that acts in acidic particles of the trans-Golgi network, so leading to the release of mature particles. This protease functions efficiently in insect cells to cleave the flavivirus polyprotein at the Pr/M junction. The inclusion of an additional Furin cleavage site at the PrM/E junction, at the position of the signalase recognition site, has been demonstrated by the inventor to lead to efficient cleavage of the polyprotein at this position, so greatly improving both the quantity and fidelity of VLP production. VLPs thus produced are thus extremely suitable for use in methods for vaccination and diagnosis.

Constructs containing an extra Furin cleavage site at this position have been shown to lead to the generation of more antigenic VLPs (1F-1) that equivalent constructs that lack this cleavage site at this position (3-11). The 1F-1 VLPs were consistently shown to be a more sensitive and specific antigen for use in ELISA-based flavivirus diagnostic tests than the unmodified VLPs. The VLPs were also shown to be more effective than the unmodified VLPs when used to inoculate mice prior to challenge with a lethal dose of WNV, increasing both survival rate and average survival time.

Any Furin cleavage site may be included at the PrM/E junction, provided that this cleavage site is recognised by a Furin protease that is contained within the insect cell in which the baculovirus construct of the invention is contained. Preferably, the Furin cleavage site comprises any one of the consensus sequences "RXXR", "RX(R7R)R", RXXXXR" or "RSRRRS" (see Cieplik et al, (1998), Biological Chemistry (379) 1433- 1440). More preferably, the Furin cleavage site comprises the consensus sequence "RSRRRS".

The Furin may be contained naturally within the insect cells that are used for expression of VLPs. Furin is mainly described in the literature as being a mammalian protease, although a Furin enzyme has been isolated and characterised in Sf9 cells (see Cieplik 1998 Biological Chemistry (379) 1433-1440) that also plays a role in 'natural' baculovirus infection of insect cells (Westenberg et al., (2002) Journal of Virology, 76: 178-184). This Furin has identical substrate specificity as the mammalian enzyme, although it has been reported that Sf9 cells have a low cleavage capacity for proteins that are readily cleaved by furin cells in vertebrate cells such as the haemagglutin (HA) of fowlplague virus (Kuroda et al., 1991, Virology, 180: 159-165) and the E protein of HIV (Wells and Compans (1990), Virology (176): 575-586), and that the use of co-expressed furin is recommended for the expression of proteolytically cleaved recombinant proteins in insect cells.

Accordingly, the source of Furin used in the method of the present invention may be exogenously-derived. For example, the insect cells may be modified to incorporate additional copies of Furin proteases (including non-insect Furin proteases such as mammalian Furin proteases) in order to increase the efficiency of cleavage. Techniques for the introduction of gene copies encoding additional Furin proteases will be known to those of skill in the art and include such techniques as calcium phosphate transfection, DEAE-dextran mediated transfection, transvection, microinjection, cationic lipid- mediated transfection, electroporation, transduction, scrape loading, ballistic introduction or infection (see Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 2001; Ausubel et al., 1991 [supra]; Spector, Goldman & Leinwald, 1998). Other methods by which a source of Furin may be provided to the insect cells will be apparent to those of skill in the art. In order to practise the method of the present invention, it will be necessary to generate an insect cell which includes a construct comprising the PrM and envelope (E) genes of a flavivirus, or a variant of the PrM and/or envelope (E) genes expressed under the control of a promoter in a baculoviral expression cassette. The step of generating such insect cells may form a part of the method of the present invention.

Methods for the generation of such constructs will be apparent to those of skill in the art from reading the present specification and using teachings published in the literature. For example, methods and materials for baculovirus/insect cell expression systems are commercially available in kit form from, inter alia, Gibco BRL (the Bac-to Bac system), Invitrogen, San Diego CA (the "MaxBac" kit). These techniques are generally known to those skilled in the art and are described fully in Summers and Smith, Texas Agricultural Experiment Station Bulletin No. 1555 (1987). In brief, in the classical system, linearised baculovirus (Autographa californica) DNA containing the LacZ gene is co-transfected into Spodoptera frugiperda (Sf) insect cells with the shuttle plasmid containing the insert, flanked by polyhedrm promoter sequences. Homologous recombination replaces the LacZ gene with the insert to produce viable viruses. Plaques are screened for the insert by blue/white selection using X-gal as a substrate in plaque assays.

The Bac-to-Bac system is more efficient than the classical system because recombination occurs in bacteria already containing baculoviral DNA (bacmid). This means that the multiple rounds of plaque purification required by the classical system is no longer necessary, so that expression is significantly faster using this system. The Bac-to-Bac system and similar systems that share these advantageous features are thus preferred according to the methods of the present invention.

All these systems use a baculoviral expression cassette that includes a polyhedrin promoter. By "baculoviral expression cassette" is meant a portion of nucleic acid that contains regulatory signals necessary for the transcription of the protein(s) encoded by genes whose transcription is controlled by the regulatory signals in the cassette. It is, however, not essential for the working of the present invention that the polyhedrein promoter is used. Any "late promoter" that is effective to drive transcription in insect cells may be used. Preferably, very late promoters such as the polyhedrin and plO promoters are used. A polyhedrin promoter may be defined as 5' noncoding region of the polyhedrin gene. Proteins under the control of the very late promoters, such as plO and polyhedrin, can account for up to 50% of the total cell mass during baculovirus infection. Foreign gene inserts under control of the polyhedrin promoter can produce high levels of recombinant protein expression.

Particularly suitable host cells for use in this system include insect cells such as Drosophila S2 and Spodoptera Sf9 cells and High5 (Tn5) cells (Invitrogen). High5 cells are particularly preferred for use in the method of the invention, since these cells have been found to lead to higher expression levels of intact, immunogenic VLPs, as used in the method of the present invention.

According to the method of the invention, insect cells in which the VLPs of the invention are expressed should be cultured for a sufficient period of time to allow production of baculovirus particles. This period of time will vary according to the particular system used and, potentially, the particular flavivirus species from which the genes used in the system are derived. This period of time will be apparent to those of skill in the art. If in any doubt, the optimum period of time may be found by incubating the inset cells under various conditions and analysing the quantity and quality of VLPs that are generated. Generally, the culture time will vary between 1 and 10 days and will optimally be around 5 days. After a sufficient period of culturing, the VLPs must be separated from the baculoviral particles and the insect cells in order to allow their subsequent use, such as in the diagnostic and vaccine applications discussed below. Any method may be used that allows the efficient separation of VLPs from baculoviral particles and insect cells. Centrifugation is one preferred method. For example, an initial, relatively gentle centrifugation step (-5,000 g) at a relatively low speed will pellet the insect cells, whilst the baculoviral particles may be separated at a higher velocity of rotation (~54,000 g) for a short period of time (1 hour). The smaller VLPs may then be separated from the resulting supernatant by a longer period of centrifugation at high speed (12 hours or more). If necessary, VLPs may be further purified, for example, using a 20-60% sucrose gradient in accordance with standard methods known to those of skill in the art (see Sambrook, 1989 [supra]).

According to a further aspect of the invention, there is provided a composition of flavivirus VLPs obtained by a method according to any one of the preceding claims. Preferably, such a composition contains WNV VLPs. The invention also provides pharmaceutical compositions comprising a preparation of such VLPs, in combination with a suitable pharmaceutical carrier. A thorough discussion of pharmaceutically acceptable carriers is available in Remington's Pharmaceutical Sciences (Mack Pub. Co., N.J. 1991). VLPs produced according to the methods of the present invention have a large number of applications, including as therapeutic or diagnostic reagents, as vaccines, or as other immunogenic compositions. Particularly preferred applications lie in the fields of diagnosis and vaccination.

Accordingly, a further aspect of the invention provides for the use of a composition according to the above-described aspect of the invention in diagnosis of a flaviviral- mediated disease or in a method of diagnosis. For example, VLPs generated according to the present invention may form a component of a diagnostic kit.

Such assays may also be used to evaluate the efficacy of a particular therapeutic treatment regimen in animal studies, in clinical trials or in monitoring the treatment of an individual patient, or in epidemiological studies.

A number of methods exist for the diagnosis of disease, that utilise recombinant preparations of protein. Such assays generally detect antibody specific for flavivirus proteins, that circulates in patient sera and include methods that utilise recombinant VLPs and a label to detect circulating antibody in human body fluids or in extracts of cells or tissues. All appropriate methodologies may utilise the recombinant VLPs generated by a method according to the present invention.

Assay techniques that can be used to determine levels of a polypeptide of the present invention in a sample derived from a host are well-known to those of skill in the art and include membrane, solution, or chip based technologies for the detection and/or quantification of antibody (particularly IgG and IgM) (see Hampton, R. et al. (1990) Serological Methods, a Laboratory Manual, APS Press, St Paul, MN) and Maddox, D.E. et al. (1983) J. Exp. Med, 158, 1211-1216). Examples include techniques such as radioimmunoassays (RIA), competitive-binding assays, Western Blot analysis, ELISA (such as direct and capture techniques) and FACS assays and include membrane, solution, or chip based technologies for the detection and/or quantification of antibody (see Hampton, R. et al. (1990) Serological Methods, a Laboratory Manual, APS Press, St Paul, MN) and Maddox, D.E. et al. (1983) J. Exp. Med, 158, 1211-1216).

This aspect of the invention thus provides a diagnostic method that comprises the steps of: (a) contacting a VLP preparation as described above with a biological sample under conditions suitable for the formation of a polypeptide-antibody complex; and (b) detecting said complex. In these techniques, body fluids or cell extracts taken from a patient are contacted with recombinant VLPs under conditions suitable for complex formation Complex will only form with antibodies if antibodies are present in the body fluid or cell extract. The amount of standard complex formation may be quantified by various methods, such as by photometric means. Inclusion of appropriate controls ensures the credibility of such systems.

Samples for diagnosis may be obtained from a patient subject's cells, such as from blood, urine, saliva, tissue biopsy or autopsy material. VLPs may be used either with or without modification, and may be labelled by joining them, either covalently or non-covalently, with a reporter molecule to aid detection of complex. A wide variety of reporter molecules known in the art may be used. Examples include Suitable radionucleides, enzymes and fluorescent, chemiluminescent or chromogenic agents as well as substrates, cofactors, inhibitors, magnetic particles, and the like. When unlabelled VLPs are used, in order to detect the presence of antibody molecules in patients, it may be preferable to use labels that are specific for patient antibodies.

A preferred diagnostic method is an ELISA-based method.

According to a still further aspect of the invention, there is provided a diagnostic kit comprising a preparation of recombinant VLPs generated according to any one of the methods of the invention described above. Preferably, such a diagnostic kit will additionally incorporate at least one reagent useful for the detection of a binding reaction between the antibody and the polypeptide. Such kits will be of use in diagnosing a flaviviral-mediated disease. As the skilled reader will understand, a number of different preparations of VLPs, appropriately labelled to allow their respective distinction may be used, in order to detect the presence of flavivirus-specific antibodies of different types. Furthermore, the VLPs of the invention may be used in conjunction with one or more other systems as a combined diagnostic system for the detection of a range of different disorders and/or diseases. The VLPs of the invention may also be used as components of vaccines. Accordingly, this aspect of the invention includes the use of a composition according to the above- described aspect of the invention in a vaccine or in a method of vaccination. In this aspect of the invention, the VLPs are used to raise antibodies against the disease-causing agent (the flavivirus). Vaccines according to this aspect of the invention may either be prophylactic (i.e. to prevent infection) or therapeutic (i.e. to treat disease after infection). Such vaccines will comprise the immunising VLPs, usually in combination with a pharmaceutically- acceptable carrier as described above, which include any carrier that does not itself induce the production of antibodies harmful to the individual receiving the composition. Additionally, these carriers may function as immunostimulating agents ("adjuvants"). Furthermore, the antigen or immunogen may be conjugated to a bacterial toxoid, such as a toxoid from diphtheria, tetanus, cholera, H. pylori, or from another pathogens.

Since polypeptides such as VLPs may be broken down in the stomach, vaccines are preferably administered parenterally (for instance, by subcutaneous, intramuscular, intravenous, or intradermal injection). Formulations suitable for parenteral administration include aqueous and non-aqueous sterile injection solutions that may contain anti- oxidants, buffers, bacteriostats and solutes which render the formulation isotonic with the blood of the recipient, and aqueous and non-aqueous sterile suspensions which may include suspending agents or thickening agents. The vaccine formulations of the invention may be presented in unit-dose or multi-dose containers. For example, sealed ampoules and vials and may be stored in a freeze-dried condition requiring only the addition of the sterile liquid carrier immediately prior to use. The dosage will depend on the specific activity of the vaccine and can be readily determined by routine experimentation. Furthermore, a number of different VLP preparations according to the invention may be administered as a combination vaccine, for example, to target a combination of different flavivirus-mediated diseases. Additionally, vaccine components specific for unrelated disorders might be included in a combination vaccine, for reasons of program management, enhanced efficacy or, more usually, for lowered cost of administration and preparation.

According to a further aspect of the invention, there is provided a nucleotide construct for use in any one of the aspects of the invention described above. Such a nucleotide construct comprises the PrM and envelope (E) genes of a flavivirus, or a variant of the PrM and/or envelope (E) genes, cloned under the control of a promoter in a baculoviral expression cassette, wherein a region of nucleic acid encoding an additional Furin cleavage site is included at the junction of the M/E genes. As discussed above, the promoter used in the construct is preferably a polyhedrin promoter or a plO promoter. The invention also provides a vector comprising such a nucleotide construct and insect host cells comprising such a nucleotide construct or vector. Various aspects and embodiments of the present invention will now be described in more detail by way of example, with particular reference to VLPs generated for West Nile Virus. It will be appreciated that modification of detail may be made without departing from the scope of the invention.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 : Construction of flavivirus VLP constructs showing primer sequences

Figure 2: Western blot of different fractions of TBE VLP construct probed with (Top) a baculovirus specific monoclonal antibody (gp64) and (Bottom) TBE horse polyclonal. BacPacό infected cell lysate is negative control. Figure 3: Coomassie blue (Top) and western blot (Bottom) of overnight (O/N) spin pellet of 3-11 and 1F-1 VLPs. 1F-1 was further fractionated as shown using sucrose gradient (10-60%)). Samples were standardised for protein concentration (30μg). Western blot was immunoprobed with a 1 in 500 dilution of TCI 27, WNV mouse polyclonal sera. Figure 4: Transmission electron micrographs (magnification as shown) of (A) 1F-1 sucrose gradient fraction 1; (B) 3-11 sucrose-gradient fraction 2; (C) purified WNV (NY 99 strain).

Figure 5: Transmission electron micrographs (magnification as shown) of (A) Purified TBE-VLPs; (B) Purified JEV-VLPs. Figure 6: Titration of 3-11, 1F-1 and WNV antigens in ELISA using mouse polyclonal and human Dengue positive sera. Negative mouse and human sera were used as controls. Antigens were diluted to a concentration of 2 μg μl^"1. All sera were used at 1 in 100 dilution. Assays were carried out in duplicate.

Figure 7: Reactivity of panel of anti- WNV polyclonal sera in ELISA assay with plates coated with WNV, TBE, 3-11 (2 μg per well) or 1F-1 protein (0.5 μg per well). All sera were used at a 1 in 100 dilution. Cut-off value of 0.57 is indicated by dotted line.

Figure 8: Reactivity of panel of monoclonal sera in ELISA assay with plates coated with WNV, TBE, 3-11 (2 μg per well) or 1F-1 protein (0.5 μg per well). All sera were used at a 1 in 200 dilution. Cut-off value of 0.181 indicated by dotted line. Figure 9: Reactivity of panel of human sera in ELISA assay with plates coated with WNV, TBE, 3-11 (2 μg per well) or 1F-1 protein (0.5 μg per well). All sera were used at a 1 in 200 dilution. Sera grouped into pre-immune, TBE positive and post-YF (vaccine). Cut-off value of 0.18 indicated by dotted line.

Figure 10: Reactivity of panel of human dengue negative and positive sera in ELISA assay with plates coated with WNV, TBE, (2 μg per well) or 1F-1 protein (0.5 μg per well). All sera were used at a 1 in 200 dilution. Cut-off value of 0.27 is indicated by dotted line. Figure: 11. Stability of dehydrated 1F-1 protein incubated at -20 °C, 4°C, 25 °C or 40°C for 12 days (A), 29 days (B) or 76 days (C). 1F-1 (0.5 μg) and WNV (2 μg) were freshly coated on ELISA plates as controls. A 1 in 200 dilution of sera was used.

Figure 12: Inoculation of mice with 3-11 or 1F-1 protein (50 μg) and subsequent challenge with 10² or 10³ pfu of WNV. Number of surviving mice shown.

EXAMPLES

MATERIALS AND METHODS

Reverse-transcriptase PCR (RT-PCR)

Primers were designed using published sequence data containing 31 amino acids of the capsid (C) gene, the entire PrM and envelope (E) genes and 31 amino acids of the NS1 gene (see Figure 1).

West Nile Virus (New York 1999 strain (WNV NY)) was obtained from R. Shope (CDC, Atlanta, USA). JEV Jagar-01 strain was obtained from Dr. D. Hayasaka (Hokkaido University, Japan). TBE Vasilchenko strain was used (from Russia). The viruses were amplified in groups of suckling mice according to standard methods (23). RNA was isolated from the viruses using RNAgents total RNA extraction kit (Promega) according to manufacturers' instructions. One μg of RNA was used as template for RT-PCR reactions (Reverse-iT One Step, ABgene) in accordance with manufacturers' instructions. A 10 μM (50 μl total volume) concentration of the respective primers were used (Figure 1). PCR products were purified using Qiagen QIAQuick spin columns in accordance with manufacturer's instructions. The PCR products were digested with the relevant restriction enzymes, according to manufacturers' instructions to create cohesive ends for ligation into plasmid vector similarly cut with the relevant restriction enzymes. For the 1F-1 construct the two PCR products cut with Bglll were ligated together using T4 ligase overnight @4°C in supplied buffer (Roche), in a 30 μl reaction volume. The ligated DNA product was then gel purified using QIAQuick columns. Cloning into baculovirus shuttle vectors

The various constructs were ligated into baculovirus shuttle vectors cut with the relevant restriction enzymes using T4 ligase overnight @4°C in supplied buffer (Roche), in a 30 μl reaction volume. The ligated DNA products were then gel purified using QIAQuick columns. The TBE and JEV constructs were ligated into pBacPAK8 vector (Clontech), and the WNV constructs into pFastBacl vector (Gibco BRL).

Purified plasmid DNA was transformed into E. coli DH5α competent cells using standard molecular cloning techniques (32). Colonies were screened for inserts by PCR using vector and construct specific primers. Colonies containing the correct size inserts were used to inoculate o/n cultures (5 ml). Plasmid DNA was prepared from the cultures using Qiagen mini-prep spin columns according to manufacturer's instructions. The plasmid DNA was sequenced using the above primers to check the integrity of the inserts. W. Tyne using an ABIgene type sequencer at CEH Oxford carried out sequencing.

Production of recombinant baculoviruses Two different baculovirus expression systems were used, the Bac-to-Bac system (Gibco BRL) was used for the WNV constructs and the classical method used for the TBE and JEV constructs (16). In both systems Sf9 cells were transfected using lipofectin reagent according to standard conditions (16). After 3 days incubation of the infected cells @28°C, the cells were harvested as primary viral cultures (2 ml). These stocks were titrated using standard plaque assays and positive clones selected by blue/white selection (16). In the case of the TBE and JEV constructs the plaques underwent at least three rounds of plaque purification.

Six clones of each construct were tested for E gene expression using a general flavivirus E gene mAb in immunofluorescence assays (TC813). In brief, 10⁶ Sf9 cells per well of a six-well plate and containing sterile coverslips, were infected with virus using a multiplicity of infection (MOI) of 10 (i.e. 10 viruses per cell). The coverslips were taken out after 2, 3 or 4 days post-infection. The coverslips were washed repeatedly and incubated with the primary sera (TC813), washed again, and incubated with anti-mouse IgG FITC conjugate. After washing, the coverslips are incubated with the fluorescent substrate and examined under a UV microscope.

On this basis one clone from each construct was chosen for protein expression. The primary cultures were amplified according to standard methods by infecting 25 ml of Sf9 cells (5 x 10⁵ cells ml^"1) at an MOI of 0.1, incubated at 28°C for 5 days (16). These virus stocks were again amplified to make working stocks (200 ml). The working stocks were titrated using standard plaque assays.

Protein production

Protein expression was carried out by infecting cells (either Sf9 or High-5(Invitrogen)) at a density of 1 x 10⁶ cells ml^"1 with an MOI of 5, with the relevant serum-free media. The Sf9 and the High-5 cells were harvested at five days post-infection. The optimum MOI was calculated for each clone empirically, there was no increase in yield after an MOI of 5 (data not shown). Levels of protein expression were measured by western blot analysis at different times post-infection, five days was found to be optimum with both cell lines (data not shown) .

The cultures were centrifuged at 5,000 g for 20 mins to spin-down the insect cells. The supernatant was spun at 54,000 g for 1 hr to remove baculovirus particles. This was repeated to further clarify the supernatant. The supernatant was then spun at 19 krpm this fraction was believed to contain the VLPs if present because only the much larger baculovirus would sediment at the shorter spin times. All of the pellets were resuspended in PBS to an approximate concentration of 10 μg μl^" . Where shown the VLP fraction was further purified by putting onto a 20-60% sucrose gradient in accordance with standard methods (32).

SDS-PAGE and Western-blot analysis Gradient (4-12%>) Bis-Tris polyacrylamide gels (NuPage gels from Novex) were used according to the manufacturer's instructions using MOPS (3-morpholinopropanesulfonic acid ) buffer unless otherwise indicated. Samples were denatured by boiling for 5 min in the presence of loading buffer (containing 10 mM di-thiothreitol) before gel loading. SeeBlue Plus2 pre-stained molecular weight markers (Novex) were used throughout. Gels were either Coomassie-blue stained for total protein or western blotted.

For western blotting, the proteins were electrophoretically transferred to a nitrocellulose membrane using standard methods (36). Membranes were blocked overnight (at 4° C) in blocking solution (5% non-fat milk in PBS containing 0.1% TWEEN-20). The membranes were then incubated with primary sera for 1 h. Sera dilutions were calculated empirically from dot blot experiments (data not shown). After incubation, the membranes were washed three times for 10 min in PBS containing 0.1 % TWEEN-20. The membranes were then incubated for 1 h in blocking solution containing a 1 in 3000 dilution of the relevant anti-IgG alkaline phosphatase conjugate (Sigma). The blots were developed using BCIP/NBT alkaline phosphatase substrate (Sigma).

Electron microscopy (EM)

Samples were absorbed onto film-coated grid (Formvar) and allowed to soak for 3 mins. Phosotungstic acid (2%) was added to the samples and allowed to soak for 3 mins. The samples were processed in Joel 1200EXII scanning electron at a magnification of 75 000.

All EM work was done by Dr. A. Patmanidi at Brookes University, Oxford, UK.

Protection experiments

Female BALB/C mice of different ages (3, 4, 5 and 6 weeks-old) (five mice per group) were sub-cutaneously injected with 10⁴ pfu ml^"1 of WNV (NY-99), to assess kill off rate for the protection experiments. All of the mice died by 8-days post-injection. Three- week old mice had all died after 7-days (data not shown).

Three groups of five 4-week female BALB/c mice were inoculated by i.p. injection with 50 μg of recombinant 1F-1, 3-11 or BacPakδ infected cell lysate as a control. ISA was used as a conjugate at a 1:1 dilution. After 10 days the mice were challenged sub- cutaneously with either 10² or 10³ pfu of WNV. The mice were checked daily for signs of the disease. Surviving mice were sacrificed 15 days post-infection.

ELISA based assays

Direct-capture ELISAs were carried out according to standard procedure (6). All the antigens were diluted to a concentration of 2 μg/μl in PBS. For the results shown, WNV and TBE virus (glycerol tartrate gradient purified suckling-mouse brain suspensions) BacPacό (baculovirus control) and 3 11 were used at a 1 in 100 dilution in coating buffer to a volume of 100 μl per well (i.e. 2 μg). IF 1 was used at a 1 in 400 dilution (i.e. 0.5 μg per well). Working dilutions were calculated empirically (data not shown). The antigens were adsorbed overnight at 4°C onto 96-well imunolon plates. The plates were washed three times for 10 mins between incubations in PBS + 0.1 %> TWEEN20. The plates were blocked using BSA (1 mg ml^'1). All blocking and sera incubations were at 37°C for 1 h. Primary sera was used at various dilutions (1 in 100 to 1 in 800), the results shown in the graphs relate to primary sera dilutions of 1 in 200. A 1 in 5000 dilution of the relevant anti-IgG (Fc-specific) HRP conjugates (Sigma) were used throughout. The substrate (OPD) was added to the plates for 30 mins, then stopped using 2M HC1 and read at 492 nm using a plate reader. Control plates (containing baculovirus BacPak 6 vector) were used as background controls.

Sera used for ELISAs All sera obtained from animals or humans were prepared from freshly taken blood. The blood was allowed to clot on ice for 1 hour and the sera prepared by spinning at 14 krpm for 10 mins.

The human sera used in ELISAs fall into four groups; pre-immune sera, post-YF inoculation sera, TBE positive sera and Dengue positive sera. The pre-immune sera were obtained from individuals from CEH Oxford before they underwent an inoculation program prior to working at the institute. It should be noted that the inoculation history of these individuals is unknown, so that it is possible that some already contained flaviviral antibodies. Post-YF sera represent individuals who were bled after a YF inoculation programme. The TBE patient sera came from individuals with known high levels of antibodies against TBE. The Dengue positive/ negative sera were from dengue infected patients and provided by Omega Diagnostics Ltd, Scotland.

Animal polyclonal sera were obtained by infecting animals with WNV (various strains) at sub-viraemic levels and taking blood 14 days post-infection. With the exception of mouse 217 and rabbit poly sera, that was raised in-house, the sera were obtained from Dr. Milan Labuda, Slovakia. Monoclonal anti-sera, unless otherwise stated was produced by Dr. E.A. Gould of CEH Oxord, raised against the E protein of various flaviviruses (9).

Stability testing of VLPs

Antigens were diluted in 50 μl of coating buffer; a 1 in 400 dilution of 1F-1 and 1 in 100 dilution of WNV were used. The liquid was evaporated overnight in a sterile hood. The dried samples (on 96-well plates) were incubated at -20°C, 4°C, 25 °C or 40°C for 12, 29 or 76 days. After incubation the plates were removed and the antigens resuspended by adding 100 μl of water and placing on an orbital shaker for 1 hour and then overnight at 4°C. The results were compared with freshly made antigens in coating buffer as above. All samples were treated as standard ELISA samples described above.

RESULTS

Optimisation of protein purification strategy

In order to test centrifugation procedure for efficiency of baculovirus particles the various fractions obtained (TBE VLPs) were analysed by western blot analysis using a monoclonal antibody (from R. Possee) against a baculovirus protein, p60 (1 in 1000 dilution). The results are shown in Figure 2 (Top). The fractions were standardised for protein concentration (20 μg per lane). The largest quantity of baculovirus protein was observed in the pellet obtained from the 1 hour spin. The least amount of baculovirus protein observed was in band 4 of the sucrose gradient fractions. A gel run with identical samples in parallel was western blotted with a 1 in 2000 dilution of horse anti-TBE polyclonal antiserum. The results are shown in Figure 3 (Bottom). The major antigen found in overnight pellet and band 4 of the sucrose gradient-purified fractions was a 60kDa antigen believed to be unprocessed M+E polyprotein. This fraction also contained the majority of VLPs as revealed by transmission electron microscopy. Figure 3 shows a Coomassie-blue stained gel and a gel western blotted using an anti- WNV antibody (TC127) at a 1 in 500 dilution; these gels compare the overnight pellets of 3-11 and 1F-1 infected insect cells. The gels were run in parallel and standardised for protein loading. As was observed in Figure 2, the major antigenic species found in either 3-11 or 1F-1 overnight pellets was a 60 kDa species believed to be uncleaved M+E protein, however in the 1F-1 samples an extra band corresponding to correct size for the properly processed E protein. This suggested the addition of the Furin cleavage site causes an increase in the efficiency of M/E cleavage. The 1F-1 overnight pellet was loaded onto a sucrose gradient (10-60 % w/w) and 4 bands were obtained. Compared to the starting material, band 1 appears to be the cleanest, although looking at the Coomassie stained gel, it is clear that the band still contains impurities, probably from baculovirus proteins. Unlike the TBE VLPs, the WNV VLPs did not appear to enter the sucrose gradient to the same extent although an identical gradient was used (20-60% w/w). Comparison of Protein Expression Levels in Sf9 and High5 cells

It has been reported that some proteins expressed using the baculovirus expression system, particularly those that are secreted, display improved expression when High5 insect cells are used instead of Sf9 cells (Clontech). Electron microscopy reveals that these cell types have a more extensive complex Golgi and other secretary organelles than Sf9 cells (A. Patmanidi - unpublished results).

A comparison was made between protein expression of 3-11 and 1F-1 in Sf9 and Tn5 (High5) insect cell lines by infecting 100 ml (@ 1 x 10⁶ cells ml^"1) with an MOI of 5. The supernatant was then centrifuged for 1 hour at 19 krpm and the pellet resuspended in PBS (500 μl). The supernatant was then centrifuged for 20 hours (O/N) at 19 krpm and the pellet resuspended in PBS (500 μl). The protein in the clarified supernatant was then polyethylene glycol (PEG-3350) precipitated according to standard methods (32). The precipitated protein was resuspended in 10 ml of 50 mM Tris pH 7.5. The protein concentration was measured as described above. Secreted protein expression was greater in infected Tn5 cells than Sf9 cells in all cases including the 1 hour spin, suggesting that more baculovirus is also produced as well as VLPs (Table 1). Tn5 cells are the preferred cell line for expression of VLPs made by this method.

Electron Microscopy

Sucrose gradient purified VLP-containing fractions of WNV (3-11, 1F-1), TBE or JE recombinant proteins were examined by transmission electron microscopy. The wild-type WNV is also shown for comparison. The results are shown in Figures 4 and 5. In all figures particles of approximately 50nm in diameter can be seen, an iscosehedral shape can clearly be seen in Figure5. Comparison of the WNV virion and VLP structures suggest the VLPs to be morphologically identical.

ELISA results The different constructs, 3-11 and 1F-1, were tested along with WNV at an equal protein concentration (2 μg μl^"l) over a range of antigen dilutions ranging from 1 in 100 to 1 in 12800. A WNV polyclonal mouse antisera (TC127) and a flaviviral human positive (Dengue) were used as positive controls and naϊve balb/c and negative human sera were used as negative controls. A 1 in 200 dilution of sera was used throughout. Figure 6 shows that not only was the 1F-1 protein more reactive against both TCI 27 and the flaviviral positive sera than the 3-11 protein but that at equal dilutions of the antigens, 1F- 1 protein was more reactive than the WNV. As a consequence for subsequent ELISAs a 1 in 100 dilution of 3-11 or WNV was used compared to a 1 in 400 dilution of 1F-1.

To prove that the ELISA technique was useful in detecting the presence of WNV antibodies in different species, sera from WNV infected Balb/c mice, guinea pig or rabbits was tested using 1F-1, 3-11 or WNV as antigens. Naϊve guinea pig, rabbit or mouse sera were used as negative controls. A cut-off value of 0.57 was calculated using the mean negative value (i.e naive sera) + S.D. X 2 (Figure 7). WNV antigen gave positive reactions against all of the sera tested, the highest value obtained with mouse WNV (2) sera. Both 3-11 and 1F-1 reacted strongly against all of the sera with the exception of rabbit WNV (1) sera where both constructs gave values below the cut-off point. 1F-1 was the most highly reactive of the antigens with an average positive/negative ratio (P/N) of 10.6 (2.34/0.22) compared with a P/N of 8.5 for WNV (1.08/0.12).

To address the specificity of the recombinant antigens, the proteins were tested along with WNV against a panel of monoclonal sera raised against E proteins of various flaviviruses (Figure 8). Ascites fluid from clones 5.28, 5.38, 5.46, 5.49 and 5.41 were raised against WNV, 813 against YF, 34.1, 34.2 and 35.56 were raised against louping-Ill (LI) (9). A cut-off value of 0.18 was calculated using the mean negative value (i.e Balb/c sera) + S.D. X 2. WNV, 1F-1 and 3-11 gave positive results against those antibodies raised against WNV, with the exception of 5.28, but showed little reactivity against those raised against YF or LI (Fig. 6). WNV and 1F-1 but not 3-11 were reactive against the T7 monoclonal (raised against TBE (35)), albeit to a lesser extent than the WNV monoclonals.

The recombinant proteins were assessed for their suitability as human diagnostic antigens by testing against a panel of human sera and comparing with WNV and TBE (Vasilchenko strain) viruses. Unfortunately no human WNV positive serum was available for testing. Because cross-reactivity between flaviviral positive patients is well established in ELISA-based flaviviral diagnostic assays, we used TBE and Dengue positive human sera as positive controls in these experiments (21, 25, 28, 33). Pre- immune sera were compared with sera from individuals after YF vaccination and TBE positive sera in Fig. 7. Flaviviral negative and dengue positive sera were tested in Figure 9.

A cut-off value of 0.18 was calculated using the mean negative value (i.e pre-immune sera) + S.D. X 2 for Fig. 7. The recombinant antigens as well as WNV and TBE were positive against sera from Simpson, Wallace and Cecilia, all whom had a history of TBE infection, but were negative against sera from individuals with little or no flaviviral antibodies. WNV mean P/N (pre-immune/TBE positive) was 9.8 (0.08/0.72), TBE was 13.2 (0.06/0.78) and 1F-1 4.8 (0.08/0.39). Sera from individuals containing YF antibodies as the result of an inoculation course were also tested. These sera reacted more slightly strongly than the pre-immune sera (mean YF inoculated/ pre-immune value =1.25 (0.10/0.08)) but gave values less than the cut-off value in all cases.

A similar situation was found with the Dengue positive sera (Figure 10). Both WNV and TBE viruses were positive for those sera that were positive for Dengue antibodies but not the flaviviral negative sera. 1F-1 reacted positively against Dengue IgG (1) but not to the other Dengue positive sera, although Dengue IgG (3) gave a value close to the cut-off value. The mean P/N ratios for this experiment were 6.4 for WNV (0.64/0.10), 5.0 for TBE (0.50/0.10) and 1.9 for 1F-1 (0.23/0.14). The Dengue IgM-positive sera gave negative results with either viruses or 1F-1, albeit a higher reactivity than that of the negative sera. Stability testing of 1F-1 VLPs

0.5 μg of 1F-1 VLPs were coated onto 96-well plates and dried down overnight before being incubated at -20°C, 4°C, 25°C or 40°C. After 12, 29 or 76 days the samples were re-hydrated and tested along with freshly coated 1F-1 and WNV in ELISAs using mouse WNV polyclonal or human flavivirus (TBE) positive sera. Negative human serum (Skol) was used as a negative control.

The results were compared with freshly-coated antigens (Figure 11). The results of the dried-down samples were remarkably similar to those of the freshly prepared antigens using the various sera. The positive sera; rabbit and mouse as well as human (Simpson and Wallace) had high values whilst the negative human sera (Skol) had a low value. These values were similar to the freshly used WNV virus. The data for day 29 was not significantly different from that of day 12. Although 1F-1 antigen incubated at 40°C was stable at day 29, after 76 days the human negative sera (Skol) was nearly as reactive as the positive human sera, suggesting the antigen maybe being degraded. Protection experiments

Forty female Balb/c mice (3-4 weeks old) were inoculated with 50 μg of BacPacό- infected Sf9 cell-lysate, 1F-1 or 3-11 VLPs. The mice were challenged by sub-cutaneous injection with a dose of 10² or 10³ pfu of WNV (NY 99 strain), ten days post-inoculation as shown in Table 2. A greater proportion of the mice inoculated with 1F-1 survived the challenge doses than those inoculated with 3-11. In addition, the average survival time was increased compared to the control inoculated mice only in those mice inoculated with 1F-1 but not 3-11 (Figure 12).

CONCLUSION

The above data relate to the construction of flaviviral VLPs using the baculovirus expression system. This is the first published example of flaviviral VLP formation in insect cells. The method was used to produce VLPs based upon WNV, TBE and JEV. The VLPs were shown to be morphologically similar to the wild-type virions. The VLPs are secreted and can be purified relatively simply by centrifugation. Expression was found to be greatest using the Tn5 (High5) insect cell-line. The constructs based upon the wild-type PrM/E cassette formed VLPs and were shown to be antigenic, however the cleavage of the M/E polyprotein as has been observed to occur in mammalian cells, was incomplete when expressed in insect cells. This was probably the result of the M/E signalase site not being recognised by insect cells. It has been shown that polyprotein cleavage is often a problem of proteins produced in insect cells (20). A WNV construct containing an extra furin site at the M/E junction was produced (1F-1). Not only was the E protein properly processed as shown by western blot analysis, but when compared in an ELISA with WNV virus and the unmodified construct (3-11), 1F-1 was much more antigenic than either when compared at equal protein loading. This construct, named 1-Fl, was shown through ELISA tests to be consistently better in performance as a diagnostic antigen than the unmodified construct (3-11). The 1F-1 VLPs were highly sensitive in human sera diagnosis against the presence of flaviviral IgG, and distinguished between low titre YF vaccine IgG and sera from individuals whom had developed a flaviviral disease. The VLPs reacted strongly against the sera from different animals that had been infected with WNV, including mice, rabbits and guinea-pigs, suggesting the VLPs can be used diagnostically with different species, something that is important in diagnostic surveillance of WNV. 1F-1 VLPs were shown to be highly specific for WNV as shown by testing with a panel of monoclonals raised against the E protein of several flaviviruses. The produced VLPs were shown to be stable in dried form for a prolonged period of time at a range of temperatures from -20°C to 40°C. These data suggest that 1F-1 VLPs are eminently suitable candidates as a WNV diagnostic antigen that may be used for diagnosis in humans as well as animals and probably in birds. The VLPs were shown to have vaccine potential as shown by their ability to protect mice against a lethal dose of the virus. The 1F-1 VLPs increased the survival rate and the average survival time of the inoculated mice compared with those inoculated with the unmodified construct. TABLE 1

Comparison of protein expression of 3-11 and 1F-1 constructs in Sf9 or Tn5 insect cells

Treatment

Construct (cell line) 1.5 hour spin O N spin PEG ppt.¹

3-11 (Sf9) 20 25 320

3-11 (Tn5) 60 60 260

1F-1 (Sf9) 20 20 220

1F-1 (Tn5) 35 50 160

Note: values given are for total protein in mg total protein per litre of culture (1 x 10⁶ cells ml^"1) infected at an MOI of 5. Values given for PEG ppt. (precipitation) already had media only PEG values subtracted. TABLE 2 C Chhaalllleennggee of Balb/c mice with 10² or 10³ pfu of WNV (NY99) inoculated with 50 μg of 3-11 or 1F-1 protein

Inoculation Dose¹ Survival Average Survival Time³

Control (BacPacθ)³ 10² pfu 0/5 8.8 (1.92)

10³pfu 0/5 7.8 (1.09)

3-11 (50 μg) 10 pfu 2/10 7.7 (2.31)

10³pfu 0/5 6.8 (0.44)

1F-1 (50 μg) 10 pfu 3/10 8.7 (2.87)

10³ pfu 1/5 7.5 (0.50)

Note: Viral stock (from mouse brain suspension) had a titre of 7 x 10⁶ pfu ml^"1. ²Those mice surviving after 21 days post-challenge. ³AST was calculated using mice that had died. Standard deviation shown in brackets. ⁴ Sonicated Sf9 cell lysate infected with BacPacό was used as a control (50 μg), VLPs were purified by ultracentrifugation (see materials and methods) REFERENCES

1. Aberie, J. H., S. W. Aberie, S. L. Allison, K. Stiasny, M. Ecker, C. W. Mandl, R. Berger, and F. X. Heinz. 1999. A DNA immnunization model study with constructs expressing the tick-borne encephalitis virus envelope E protein in different physical forms. Journal of Immunology. 163:6756-6761.

2. Anonymous. 2001. West Nile virus activity- Eastern United States, 2001. Morbidity and Mortality Weekly Report. 50:617-619.

3. Brandt, W. E. 1988. Current approach to the development of dengue vaccines and related aspects of the molecular biology flaviviruses. Journal of Infectious Diseases. 157.

4. Briese, T. 1999. Lancet. 354:1261.

5. Colombage, G., R. Hall, M. Pavy, and M. Lobigs. 1998. DNA-based and alphavirus-vectored immunisation with prM and E proteins elicits long-lived and protective immunity against the flavivirus, Murray Valley encephalitis. Virology. 250:151-163.

6. Crowther, J. R. 1995. ELISA- Theory and Practice, vol. 42. Humana Press, Totowa, New Jersey.

7. Davis, B. S., G.-J. Chang, B. Cropp, J. T. Roehrig, D. A. Martin, C. J. Mitchell, R. Bowen, and M. L. Bunning. 2001. West Nile virus recombinant DNA vaccine protects mouse and horse from virus challenge and expresses in vitro a nonifectious recombinant antigen that can be used in enzyme-linked immunosorbent assays. Journal of Virology. 75:4040-4047.

8. Fonseca, B. A., S. Pincus, R. E. Shope, E. Paoletti, and P. W. Mason. 1994. Recombinant vaccina viruses co-expressing dengue- 1 glycoproteins prM and E induce neutralizing antibodies in mice. Vaccine. 12:279-285.

9. Gould, E. A., A. Buckley, S. Higgs, and S. Y. Gaidamovich. 1990. Antigenicity of flaviviruses. In C. H. Calisher (ed.)₅ Archives of Virology, vol. Supplementum 1. Springer- Verlag, New York. 10. Green, M. 1985. Transformation and oncogensis: DNA viruses, p. 183. In Fields (ed.), Virology. Raven Press, New York.

11. Heinz, F. X. 1999. Tick-bome encephalitis virus: Advances in molecular biology and vaccination strategy in the next century. Zent.bl. Bakteriol. 289:506-510. 12. Heinz, F. X., and S. L. Allison. 1996. Epitope mapping of flavivirus glycoproteins. Advances in Virus Research. 31:103-168.

13. Heinz, F. X., S. L. Allison, K. Stiasny, J. H. Schalich, H., C. W. Mandl, and C. Kunz. 1995. Recombinant and viroin-derived soluble and particulate immunogens for vaccination against tick-borne encephalitis. Vaccine. 13:1636- 1642.

14. Kelly, E. P., and J. J. Greene. 2000. Purified dengue 2 virus envelope glycoprotein aggregates produced by baculovirus are immunogenic in mice. Vaccine. 18:2549-2559.

15. Khromkh, A. A., and T. J. Harvey. 1996. Expression and purification of the seven nonstructural proteins of the flavivirus Kunjin in the E. coli and the baculovirus expression systems. Journal of Virological Methods. 61:47-58.

16. King, L. A., and R. D. Possee. 1992. The baculovirus expression system: A laboratory guide. Chapman & Hall, London.

17. Klockman, U., and H. L. Bock. 1991. Humoral immunity against tick-bome encephalitis virus following manifest disease and active immunisation. Vaccine.

9:42-46.

18. Konishi, E., I. Yamaoka, and P. W. Mason. 2000. A DNA vaccine expressing dengue type 2 virus premembrane and envelope genes induces neutrilizing antibody and memory B cells in mice. Vaccine. 18:1133-1139. 19. Konoshi, E., S. Pincus, E. Paoletti, R. E. Shope, T. Burrage, and P. W. Mason. 1992. Mice immunized with a subviral particle containing the Japanese encephalitis virus prM/M and E proteins are protected from lethal JEV infection. Virology. 188:714-720. 20. Laprise, M., F. Grondin, and C. M. Dubois. 1998. Enhanced TGF Bl maturation in high five cells coinfected with recombinant baculovirus encoding the convertase furin pace: Improved technology for the production of recombinant proproteins in insect cells. Biotechnology and bioengineering. 1:85-91.

5 21. Makino, Y., M. Tadano, M. Saito, N. Maneekam, N. Sittisombut, V. Sirisanthana, B. Poneprasert, and P. Fukunaga. 1994. Studies on serological cross-reaction in sequential flavivirus infections. Microbiological Immunology. 38:951-955.

22. Marfϊn, A. A., L. R. Peterson, M. Eidson, J. Miller, J. Hadler, and C. Farello.

10 2001. Widespread West Nile virus activity, eastern United States, 2000. Emerging

Infectious Diseases. 7:730-735.

23. Marx, F., T. S. Gritsun, B. Grubeck-Loebenstein, and E. A. Gould. 2001. Diagnostic immunoassays for tick-bome encephalitis virus based on recombinant baculovirus protein expression. Journal of Virological Methods. 91:75-84.

15 24. Mason, P. W., and S. Pincus. 1991. Japanese encephalitis virus-vaccinia recombinants produce particulate forms of the structural membrane proteins and induce high levels of protection against lethal JEV infection. Virology. 180.

25. Monath, T. P. 1994. Dengue: the risk to developed and developing countries. Proc. Natl. Acad. Sci. USA. 91:2395-2400.

20 26. Monath, T. P. 1997. Epidemiology of yellow fever: current status and speculations on future trends. Factors in the emergence of arbovirus diseases., p. 143-156. In J. F. Saluzzo and B. Doder (ed.), Emerging diseases. Elsevier.

27. Nasci, R. 2001. Mosquito vector associations in the United States; 1999-2000. International Conference on the West Nile Virus, New York.

25 28. Niedrig, M., and D. Vaisviliene. 2001. Comparison of six different commercial IgG-ELISA kits for the detection of TBEV-antibodies. Journal of Clinical Virology. 20:179-182. 29. Phillpotts, R. J., K. Venugopal, and T. Brooks. 1996. Immunisation with DNA polynucleotides protects mice against lethal challenge with St. Louis encephalitis virus. Arch. Virol. 141:743-749.

30. Pincus, S., P. W. Mason, E. Konishi, B. A. Fonseca, R. E. Shope, C. M. Rice, and E. Paoletti. 1992. Recombinant vaccinia virus producing the PrM and E proteins of yellow fever virus protects mice from lethal yellow fever encephalitis. Virology. 187:290-297.

31. Russell, P. K., W. E. Brandt, and J. M. Dalrymple. 1980. Chemical and antigenic structures of flaviviruses, p. 503-529. In R. W. Schlesinger (ed.), The togaviruses. Academic Press, New York.

32. Sambrook, J., E. F. Fritsch , and T. Maniatis . 1989. Molecular Cloning: A Laboratory Manual, 2 ed. Cold Spring Harbor Laboratory Press, New York.

33. Schwartz, E., F. Mileguir, Z. Grossman, and E. Mendelson. 2000. Evaluation of ELISA-based sero-diagnosis of dengue fever in travellers. Journal of Clinical Virology. 19:169-173.

34. Stadler, K., S. L. Allison, J. Schalich, and F. X. Heinz. 1997. Proteolytic activation of tick-bome encephalitis virus by Furin. Journal of Virology. 71:8475- 8481.

35. Stephenson, J. R., J. M. Lee, and P. D. Wilton-Smith. 1984. Antigenic variation among members of the Tick-Bome encephalits complex. Journal of

General Virology. 65:81-89.

36. Towbin, H., T. Stachelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some application. Proceedings of the National Academy of Sciences of the United States of America. 76:4350-4354.

37. Van Lier, F. L. J., J. M. Vlak, and J. Tramper. 1992. Production of baculovirus-expressed proteins from suspension cultures of insect cells. Animal Cell Biotechnology. 5. 38. Venugopal, K. A., and E. A. Gould. 1994. Towards a new generation of flavivirus vaccines. Vaccine. 12:966-975.

39. Yamshchikov, V. F., and R. W. Compans. 1993. Regulation of late events in flavivirus protein processing and maturation. Virology. 192:38-51.

Claims

1) A method for the production of virus-like particles (VLPs) from a flaviviras, said method comprising the steps of: a) expressing a construct comprising the PrM and envelope (E) genes of a flavivirus, or a variant of the PrM and/or envelope (E) genes, in a baculoviral expression cassette and cloned under the control of a promoter in insect cells, wherein a region of nucleic acid encoding a Furin cleavage site is included at the junction of the M/E genes; b) culturing the insect cells for a sufficient period of time to allow production of baculovirus particles; and c) separating the VLPs from the baculoviral particles and the insect cells.

2) A method according to claim 1, wherein the VLPs are derived from the WNV; TBE; YF; JE; Louping-Ill; Kunjin virus; St. Louis encephalitis virus; Dengue virus or from Murray Valley encephalitis virus. 3) A method according to claim 2, wherein the VLPs are derived from the West Nile Virus.

4) A method according to any one of the preceding claims, wherein said insect cells are Sf9 cells or High5 cells.

5) A method according to any one of the preceding claims, wherein said Furin cleavage site comprises the sequence RXXR, RX(K/R)R, RXXXXR, RSRRRS or

RSRRRS.

6) A method according to any one of the preceding claims, wherein nucleotide sequence from a gene proximal to the PrM and/or E genes in the sequence of the viral polyprotein is included in the construct. 7) A method according to claim 6, wherein said nucleotide sequence from the gene proximal to the PrM and/or E genes is derived from the capsid (C) gene that abuts the PrM gene, and/or the NS1 gene that abuts the E gene. 8) A method according to claim 7, wherein step a) of the method uses the construct referred to herein as WNVFur (1F-1).

9) A preparation of flavivirus VLPs obtained by a method according to any one of the preceding claims. 10) A pharmaceutical composition comprising a preparation of flaviviras VLPs according to claim 9.

11) A method of diagnosis of a flaviviras-mediated disease, said method comprising the steps of:

(a) contacting a VLP preparation according to claim 9 with a biological sample under conditions suitable for the formation of a polypeptide-antibody complex; and

(b) detecting said complex.

12) Use of a preparation according to claim 9 or a composition according to claim 10 in a diagnostic kit. 13) A vaccine composition comprising a preparation according to claim 9 or a composition according to claim 10.

14) A method of vaccinating a subject against infection mediated by a flavivirus, comprising administering a vaccine composition according to claim 13 to said subject. 15) Use of a preparation according to claim 9 or a composition according to claim 10 in a vaccine.

16) A nucleotide constract comprising the PrM and envelope (E) genes of a flavivirus, or a variant of the PrM and/or envelope (E) genes, cloned under the control of a promoter in a baculoviral expression cassette, wherein a region of nucleic acid encoding a Furin cleavage site is included at the junction of the M/E genes.

17) A nucleotide construct according to claim 16, wherein said promoter is a polyhedrin promoter or a pi 0 promoter.

18) A vector comprising a nucleotide constract according to claim 16 or claim 17. 9) An insect cell comprising a nucleotide constract according to claim 16 or claim 17, or a vector according to claim 18.