WO2008152652A2 - A dengue envelope domain iii-based tetravalent protein vaccine - Google Patents

Abstract

The present invention relates to a novel recombinant envelop domain - III based tetravalent protein having SEQ ID No.: 1 which elicits protective immune responses against each of the four serotypes of dengue virus, DEN-I, DEN-2, DEN-3 and DEN-4. The invention also relates to a novel polynucleotide sequence having SEQ ID No.: 2 which encodes novel recombinant tetravalent protein and which is codon optimized for expression in eukaryotic expression system. The process for the preparation and purification of novel recombinant tetravalent protein is also disclosed. The process involves chemically synthesizing the novel polynucleotide sequence having SEQ ID No.:2, codon optimizing the sequence, followed by cloning, transforming and purifying the novel recombinant tetravalent protein. The novel recombinant tetravalent protein of the present invention is with and without secretory signal peptide. The present invention also discloses that the novel recombinant tetravalent protein results in inhibition of infectivity of each dengue virus serotype.

Description

A NOVEL DBNGUB ENVELOPE DOMAIN IH-BASED TETRAVALENT PROTEIN VACCINE

FIELD OP THB INVENTION The present invention relates to a novel recombinant envelop domain - III based tetravalent protein which elicits protective immune responses against each of the four serotypes of dengue virus, DEN-I, DEN-2, DEN-3 and DEN-4. In particular, the present invention relates to novel polynucleotide sequence encoding novel recombinant tetravalent protein, and which is codon optimized for expression in eukaryotic expression system. The present invention also relates to a process for the preparation and purification of novel recombinant tetravalent protein. In particular, the present invention relates to a novel recombinant tetravalent protein with and without secretory signal peptide. The present invention also relates to novel recombinant tetravalent protein that results in inhibition of infectivity of each dengue virus serotype.

BACKGROUND OF THE INVENTION

There are four closely related, yet antigenically distinct, serotypes of dengue (DEN) viruses (DEN-I, 2, 3 and 4), which are members of the Flaviviridae family (Lindenbach and Rice, 2001, Field's Virology, 4^th edition. Philadelphia, PA: Lippincott Williams & WiUrins, 991-1041). Infection with any one of these viruses can result in a spectrum of clinical symptoms ranging from inapparent or mild dengue fever (DF), to severe and fatal dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS). About 2.5 billion people, in over a hundred tropical and sub-tropical countries, representing ~40% of the world's population, are at risk of DEN infections. The World Health Organization estimates that there may be -100 million cases of DEN infections worldwide every year (Gubler, 1998, Clin. Microbiol. Rev. 11: 480-496). Infection with any one DEN serotype provides lifelong homologous immunity to that serotype with only transient cross-protection against the remaining three (Innis, 1997, Dengue and Dengue Hemorrhagic Fever, New York, NY: CAB International, 221-243). Epidemiological and laboratory data suggest that cross-reactive antibodies produced during a primary infection, can predispose an individual to potentially fatal DHF and DSS, during a subsequent infection, through antibody dependent enhancement (ADE). This has prompted the view that a DEN vaccine must be tetravalent', affording protection against all four DEN virus serotypes. This, together with the lack of a suitable animal model has made the development of a DEN vaccine a challenging task.

While an overwhelmingly large proportion of recombinant subunit vaccines have so far targeted single serotypes, a few tetravalent vaccine candidates are also being developed. Current efforts to develop tetravalent DEN vaccines, which are in advanced stages of development, are based on live attenuated DEN viruses (Kanesa- thasan et aL, 2001, Vaccine 19: 3179-3188; Edelman et al., 2003, Am. J. Trop. Med. Hyg. 69 (Suppl 6): 48-60; Kitchener et al., 2006, Vaccine 24: 1238-1241) or genetically manipulated chimeric flaviviruses (Guirakhoo et al., 2001, J. Virol. 75: 7290-7304; Guirakhoo et al, 2006, Human Vaccines 2: 60-67). All these vaccine viruses are monovalent in that each one is specific to a single serotype (Hombach et al, 2005, Vaccine 23: 2689-2695). Recent studies have shown that tetravalent formulations, obtained by mixing these monovalent vaccine viruses, corresponding to the four serotypes, are prone to elicit an un-balanced immune response, predominantly to one serotype, due to Viral interference'. This has been observed both in non-human primates as well as human volunteers. This poses a serious risk in the context of the ADE phenomenon. In addition, another reason for concern is that, live flaviviruses have the potential to undergo genetic recombination (Seligman and Gould, 2004, Lancet 363: 2073-2075).

In parallel, several alternative approaches for the development of recombinant DNA- and protein-based subunit vaccines are being explored by many groups. The genetic vaccines being explored utilize either naked plasmid DNA, pox virus (Men et al, Vaccine 18: 3113-3122) or, more recently, adeno-virus vectors Raja et al., 2007, Am. J. Trop. Med. Hyg. 76: 743-751; Holman et al, 2007, Clin. Vac. Immunol. 14: 182-189; Khanam et al., 2006, Vaccine 24: 6513-6525; Khanam et al, 2007, BMC Biotechnol. 7: 10) encoding DEN virus antigens. Recently, two groups have developed tetravalent DNA vaccines, both based on full-length E protein genes. While one is based on mixing four monovalent plasmid vaccines (Konishi et al, 2006, Vaccine 24: 2200-2207), the other is a plasmid encoding a single shuffled E gene chimera, carrying epitopes representative of all four serotypes (Apt et al, 2006, Vaccine 24: 335-344). Yet another group has used a mixture of four EDIII-encoding plasmids as an experimental vaccine (Mota et al, 2005, Vaccine 23: 3469-3476). The protein- based approaches are based on both prokaryotic as well as eukaryotic expression hosts.

The majority of current efforts that seek to develop such recombinant subunit candidate vaccines focus on the major envelope (E) protein, a -500 amino acid (aa) residues long, cysteine-rich, multifunctional protein. Its structure is stabilized by six disulfide (S-S) bridges, and is organized into three discrete domains, a central domain (I)₁ a dimerization domain (Ii) and an immunoglobulin (Ig)-like domain (III) (Modis et al, 2003, Proc. Natl. Acad. ScL USA 100: 6986-6991; Modis et al., 2005, J. Virol. 79: 1223-1231). The E protein binds to host cells through as yet unidentified receptor (s) (Chen et al, 1997, Nature Med. 3: 866-871) contains multiple serotype-specific, conformation-dependent neutralizing epitopes (Megret et al., 1992, Virology 187: 480- 491; Roehrig et al., 1990, Virology 177: 668-675, Roehrig et al, 1998, Virology 246: 317-328), elicits long-lasting antibody response (Churdboonchart et al, 1991, Am. J. Trop. Med. Hyg. 44: 481-493), and most importantly, confers protective immunity (Putnak et al, 1991, Am. J. Trop. Med. Hyg. 45: 159-167; Men et al, Vaccine 18: 3113-3122; Mune et al, 2003, Arch. Virol. 148: 2267-2273; Apt et al, 2006, Vaccine 24: 335-344).

A growing body of evidence, accumulated in recent years by several groups has shown that many of the properties of the E protein, important from a vaccine perspective, are associated with domain III, referred to as EDIII, spanning aa residues 300-400 (Simmons et al, 1998, Am. J. Trop. Med. Hyg. 58: 655-662; Simmons et al, 2001, Am. J. Trop. Med. Hyg. 65: 159-161; Khanam et al., 2006, Am. J. Trop. Med. Hyg. 74: 266-277; Khanam et at, 2006,Vaccine 24: 6513-6525; Khanam et al, 2007, BMC Biotechnol 7: 10; Megret et al., 1992, Virology 187: 480-491; Crill and Roehrig, 2001, J. Virol. 75: 7769-7773; Bhardwaj et al., 2001, J. Virol. 75: 4002-4007; Hung et al, 2004, J. Virol. 78: 378-388; Chin et al, 2007, Microbes and Infection 9: 1-6; Jaiawal et al., 2004, Protein Exp. Purif. 33: 80-91). Its structural and antigenic integrity depends upon a single S-S bond (Roehrig et al, 2004, J. Virol. 78: 2648-2652). The host cell receptor-binding motif has been localized to EDIII (Crill and Roehrig, 2001, J. Virol. 75: 7769-7773; Chin et al., 2007, Microbes and Infection 9: 1-6). Structural studies of intact virions have shown that EDIII is exposed and accessible on the virion surface (Kuhn et al., 2002, Cell 108: 717-725). Consistent with this, it has been demonstrated that recombinant EDIII proteins can block DEN-2 virus infectivity (Chin et al, 2007, Microbes and Infection 9: 1-6; Jaiswal et al., 2004, Protein Exp. Purif. 33: 80-91). Multiple type-and sub-type specific neutralizing epitopes have been mapped to this domain (Mέgret et al., 1992, Virology 187: 480-491). A recent study which analyzed a large panel of anti-E monoclonal antibodies (mAbs) showed that those that bound to EDIII are the most powerful blockers of virus infectivity (Crill and Roehrig, 2001, J. Virol. 75: 7769-7773). Importantly, EDIII has only a very low intrinsic potential for inducing cross- reactive antibodies implicated in the pathogenesis of DHP/DSS (Simmons et al, 1998, Am. J. Trop. Med. Hyg. 58: 655-662; Simmons et al., 2001, Am. J. Trop. Med. Hyg. 65: 159-161). DEN-I, DEN-2 and DEN-3 EDIII have also been expressed proteins using the Neisseria meningitides P64K protein as a carrier (Hermida et al, 2004, Biotechnol. Appl. Biochem. 39: 107-114; Hermida et al, 2004, J. Virol. Methods 115: 41-49; Zulueta et al, 2006, Virus Res. 121: 65-73). These proteins have been only partially purified (~35- 70%), with purity ranging from 35-70%. Virus-neutralizing antibodies induced by these proteins, in mice, are found to vary over a wide range. For example PRNT∞ titers ranged from as low as 1:15 for DEN-3 (Zulueta et al, 2006, Virus Res. 121: 65-73), to 1:640 for DEN-2 (Hermida et al., 2004, J. Virol. Methods 115: 41-49). Taken together, these attributes of the EDIII make it an excellent vaccine candidate. Important from the perspective of recombinant protein expression, the Ig-like flavivirus EDIII has been shown to be an independent folding domain, as evidenced by its release as a discrete fragment upon tryptic digestion of DEN virions (Roehrig et al, 1998, Virology 246: 317- 328; Wang et al., 1999, J. Virol. 73: 2547—2551), and exhibits a very high degree of stability (Bhardwaj et al., 2001, J. Virol. 75: 4002-4007).

In the present invention, a chimeric tetravalent protein is designed by joining the EDIIIs of the four DEN virus serotypes using flexible pentaglycyl peptide linkers. The gene encoding this protein is expressed in the methylotrophic yeast, Pichia pastoris, which combines the advantages of both prokaryotic (high expression levels, easy scale- up, inexpensive growth media) and eukaryotic (capacity to carry out most of the post- translational modifications characteristic of higher eukaryotes) expression systems. In the last several years, the methylotrophic yeast Pichia pastoris has emerged as a powerful and inexpensive heterologous system for the production of high levels of functionally active recombinant proteins of commercial and academic interest (Cereghino et al, 2002, Curr. Opin. Biotechnol. 13: 329-332; Macauley-Patrick et, aL, 2005, Yeast 22: 249-270).The availability of the strong tightly regulated methanol- inducible alcohol oxidase 1 (AOXl) promoter for high-level expression is a distinct advantage for heterologous protein production. Importantly, P. pastoris is well-suited for the expression of the disulfide-rich tetravalent EDIII-based antigen.

Several S-S linked proteins such as recombinant hepatitis B surface antigen Cregg et al., 1987, Biotechnology 5: 479-485; Vassileva et al., Protein Exp. Purif. 21: 71-80) and insulin (Kjeldsen et al, 1999, Biotechnol. Appl. Biochem. 29: 79-86) have been successfully produced in P. pastoris. The strong preference of P. pastoris (unlike S. cerevisiae) for respiratory growth allows it to be cultured at extremely high cell densities of -100g/L dry cell weight or greater. This greatly enhances productivity to gram quantities/L. Since manipulating the carbon source added to the culture medium controls the AOXl promoter, growth and induction can be easily performed at all scales ranging from shake flasks to large fermenters. Finally, P. pastoris is a non-pathogenic organism; recombinant proteins expressed in it will be free of pyrogens (unlike E. coli expressed proteins), toxins and viral inclusions (unlike tissue culture expressed proteins) making them safe for human use.

The present invention discloses a novel recombinant envelop domain - III based tetravalent protein which elicits protective immune responses specific to each of the four DEN virus serotypes. The present invention further discloses a process for the preparation and purification of novel recombinant tetravalent protein: The present invention also discloses a novel recombinant tetravalent protein with and without secretory signal peptide. The present invention further discloses inhibition of infectivity of each dengue virus serotype by novel recombinant tetravalent protein.

OBJECTS OP THE INVENTION

It is an important object of the present invention to provide a novel recombinant envelop domain - III based tetravalent protein haying SEQ ID No.: 1 eliciting protective immune responses against each of the four serotypes of dengue virus, DEN-I, DEN-2, DEN-3 and DEN-4. Another object of the present invention is to provide a novel polynucleotide sequence having SEQ ID No.: 2 encoding novel recombinant tetravalent protein and which is chemically synthesized and codon optimized for expression in eukaryotic expression system. Yet another object of the present invention is to prepare and purify a novel recombinant envelop domain - III based tetravalent protein.

Still another object of the present invention is to provide a novel recombinant tetravalent protein that results in inhibition of infectivity of each dengue virus serotype. Yet another object of the present invention is to provide a novel recombinant tetravalent protein with and without secretory signal peptide.

SUMMARY OP THE INVENTION

The present invention discloses a novel recombinant envelop domain - HI based tetravalent protein having SEQ ID No.:l which elicits protective immune responses against each of the four serotypes of dengue virus, DEN-I, DEN -2, DEN-3 and DEN-4.

In another embodiment of the present invention, a novel polynucleotide sequence having SEQ ID No.:2 which encodes novel recombinant tetravalent protein and which is codon optimized for expression in eukaryotic expression system is disclosed.

In still another embodiment of the present invention, a process for the preparation and purification of novel recombinant tetravalent protein is also disclosed. The process involves chemically synthesizing the novel polynucleotide sequence having SEQ ID No.:2, codon optimizing the sequence, followed by cloning, transforming and purifying the novel recombinant tetravalent protein.

In yet embodiment of the present invention, the novel recombinant tetravalent protein of the present invention is with and without secretory signal peptide.

In another embodiment of the present invention, inhibition of infectivity of each dengue virus serotype of the novel recombinant tetravalent protein is also disclosed.

BRIEP DESCRIPTION OP ACCOMPANYING DRAWINGS

FIGURE 1: Schematic representation of rEDIII-T. rEDIII-T protein in which domain HI of envelope protein from all four serotypes of dengue virus are linked by flexible penta-glycine linkers.

FIGURE IA: Map of plasmid pPIC-EDIII-T. The EDIII-T gene (open box) was inserted into the EcoRI and JVofl sites, in-frame with the S. cereinsiae α factor secretory signal- encoding sequence (S), under the transcriptional control of the methanol-inducible alcohol oxidase 1 (5' AOXl) promoter of the P. pastoris integrative vector, pPIC9K. A 6x His tag-encoding sequence (gray box) was provided at the 3' end of the chimeric gene. The dashed arrow indicates the direction of gene transcription. Other abbreviations: HIS4 denotes the wild type histidinol dehydrogenase gene; TT represents the transcription terminator sequences; and 3' AOXl represents the 3' terminal sequences of the AOXl gene. FIGURE IB: Schematic representation of rEDIII-T protein with and without the signal peptide (indicated by the black box at the left end). EDIII-I, -2, -3 and -4 represent EDIIIs of DEN serotypes 1, 2, 3 and 4, respectively. The gray box at the right end denotes the 6x His tag.

PIGURB 2A: Ni-NTA affinity-chromatography purification profile of rEDIII-T protein from total cell lysate. The entire process of washing and elution was controlled and monitored by connecting the column to an AKTA FPLC system. Protein concentration was monitored by absorbance at 280 nm (solid curve); the profile of the pH gradient used is indicated by the dotted curve. Fraction numbers are shown on the horizontal axis. FIGURE 2B: SDS-PAOE analysis of the purified protein. Peak fractions were analyzed in lanes 1-6. Low molecular weight protein markers were run in lane IA'; their sizes (in kDa) are shown to the left of panel.

FIGURE 2C: Western blot analysis of the purified protein. Aliquots of the pooled peak material (lanes 3 and 6), E. coft-expressed rEDIII-T antigen lacking a secretory signal peptide (lanes 1 and 4) and BSA (lanes 2 and 5) were run on 10% denaturing polyacrylamide gel. The separated proteins were transferred to nitrocellulose and probed with either a polyclonal antiserum raised against the E. co/i-expressed recombinant tetravalent protein (lanes 1-3) or penta-His mAb (lane 4-6). FIGURE 2D: Detection of glycosylation. Aliquots of the pooled peak material (lane 3), E. oofi-expressed rEDIII-T antigen lacking a secretory signal peptide (lane 2) and ovalbumin (lane 1) were electrophoresced and blotted as in C. The blot was probed with Con A-FITC. Pre-stained molecular weight markers were run in lanes ¹M' in panels C and D; their sizes (in kDa) are shown to the left. The arrows on the right of panels B, C and D denote the positions of the precursor (upper arrow) and mature (lower arrow) forms of the rEDIII-T protein.

FIGURE 3: Comparative analysis of serum antibody titres elicited in mice by Pichia pαstoris-expressed recombinant rEDIII-T protein formulated with different adjuvants. Anti-DEN antibodies in murine pre-immune sera (triangles) and immune sera from mice vaccinated with rEDIII-T antigen formulated in Freund's (diamonds), alum (squares), or montanide (circles) adjuvants were determined by ELISA using (A) DEN-I, (B) DEN-2, (C) DEN-3 and (D) DEN-4 as the coating antigen. Sera at each dilution were assayed in duplicates and the data shown represent the average. FIGURE 4: Indirect immunoflouresence analysis of antibodies in sera of mice immunized with Λctøαpαstoris-expressed rEDIII-T protein. BHK cells were infected with each of the four DEN viruses as indicated. One day after infection, the virus-infected cells were fixed and probed with either sera drawn prior to immunization (Pre-imm) or after immunization with rEDIII-T formulated in Freund's adjuvant (Imm).

FIOURB 5: The rEDIII-T protein elicits antibodies that neutralize the infectivity of all four DEN virus serotypes. PRNT was performed by infecting LLCMK2 monolayers separately with (A) DEN-I, (B) DEN-2, (C) DEN-3 and (D) DEN-4 viruses that had been pre-incubated with serial two-fold dilutions of anti-rEDIII-T antiserum. The resultant plaque counts, expressed as percent inhibition of virus infectivity (with reference to the number of plaques generated in the absence of antiserum which was taken to represent 100% infectivity) were plotted as a function of antiserum dilution to determine the antiserum dilution that resulted in 50% neutralization (PRNT50 titers). The dotted line parallel to the horizontal axis, in each panel, represents 50% inhibition of virus infectivity. Each data point shown represents the mean of triplicate assays (the error bars represent standard deviation). Symbols represent Freund's (diamonds), alum (squares), and montanide (circles) adjuvant groups.

FIQURB 6: Analysis of T cell responses elicited by rEDIII-T antigen in mice. Splenocytes were obtained from immunized mice from the alum (open bars), montanide (hatched bars) and Freund's (solid bars) adjuvant groups 10 days after the final immunization and placed in culture. They were either mock-stimulated (no antigen, N) or stimulated in vitro, either with DEN-I (Dl), DEN-2 (D2), DEN-3 (D3) or DEN-4 (D4) viruses (each at 0.03 PFU/cell), for 96 hours for performing the T cell assays. Tritiated [³H] thymidine uptake was determined in a scintillation counter (A). Aliquots of the culture supernatant withdrawn at the indicated time points were assayed for the presence of IFN-γ (B) and IL-4 (C) by solid phase ELISA. Data depicted represent the mean value of three separate determinations (the error bars represent standard deviation).

DETAILED DESCRIPTION OF THE INVENTION The present invention discloses designing and expressing a novel tetravalent antigen by linking together the EDIIIs corresponding to the four serotypes by means of flexible peptide linkers. EDIII contains multiple type-and sub-type specific neutralizing epitopes. It has been shown that the single S-S bond in EDIII is critical for the maintenance of its antigenic integrity. Therefore, to allow proper folding of the EDIII components, it would be necessary to express the tetravalent antigen (with its four S-S bonds), in eukaryotic hosts such as yeast, insect or mammalian cells. Of these, yeast was chosen as an expression host as it combines the advantages of both prokaryotic (high expression levels, easy scale-up, inexpensive growth media) and eukaryotic (capacity to carry out most of the post-translational modifications characteristic of higher eukaryotes) expression systems. Methylotrophic yeast, Pichia pastoris, which is well documented as a eukaryotic expression system was used. The capacity of this yeast to grow to very high cell densities on purely defined media, coupled to the strong, tightly regulated methanol-inducible alcohol oxidase (AOXl) promoter makes it an inexpensive, yet powerful expression system. The present invention presents the design of a novel recombinant teravalent EDIII-based antigen, its expression, purification and a preliminary evaluation of its immunogenic potential in eliciting immune responses specific to each of the four DEN virus serotypes.

A dengue vaccine has been an elusive goal so far. Efforts to develop live attenuated virus vaccines, though promising, are beset with problems of Viral interference' and unbalanced immune response. The occurrence of this phenomenon, which tends to skew the immune response predominantly towards one serotype, emphasizes the limitations and more importantly the risks, associated with mixing four monovalent vaccine viruses to create a tetravalent vaccine. It is likely that the replication defects in one attenuated monovalent vaccine genome may be compensated in trans by mutations in the other components of the tetravalent mix; or the RNA genome, which is inherently unstable, can mutate into more virulent forms either during tissue culture or in the vaccine recipient, or both. Regardless of the mechanism, an unbalanced immune response resulting from viral interference has a risky outcome in the context of the ADE phenomenon. It is this consideration that is driving current efforts to find an 'optimal' tetravalent formulation by empirically varying the relative proportions of the four monovalent vaccines. However, given the inherent genetic instability of the viral RNA genome, there can be no assurance that the Optimal' formulation will continue to be 'optimal' in the vaccine recipient. Under these conditions, investigation of recombinant subunit vaccines is warranted. The present invention is driven by the hypothesis that switching from a strategy reliant on mixing four monovalent attenuated, inherently unstable, RNA viruses to a single tetravalent protein-based, therefore non-replicating, vaccine, may provide a means of circumventing viral interference and the associated risk of ADE. Keeping in mind that a DEN vaccine must be cost-effective, as the major part of the target population lives in the resource-poor regions of the world, the present invention has focused on a novel, hitherto unexplored, strategy. Thus, P. pastoris mediated expression and characterization of a recominant tetravalent antigen, based on a critical domain, domain III, of the E protein of DEN serotypes 1-4 is investigated. The choice of envelope domain III as the precursor for designing the chimeric tetravalent antigen is based on several compelling reasons. Cryoelectron microscopic analysis of DEN virions have revealed that EDIII is exposed on the outer surface of the virion. This implies that it is freely accessible for interaction with host cell receptors. This is consistent with the identification of putative glycosaminoglycan (GAG)-binding motifs in EDIII and the demonstration by several groups that r-EDIII can block flavivirus infectivity. These results suggest that DEN virus infection can be aborted at the level of entry by EDIII- targeting antibodies. That this indeed true is borne out by the observation that murine monoclonal antibodies whose epitopes map to domain III on the E molecule and that polyclonal antibodies raised against recombinant EDIII-based fusion proteins neutralize virus infectivity.

Recombinant EDIII-based monovalent and bivalent proteins elicit serotype- specific neutralizing antibodies in mice. Interestingly, the bivalent protein did not elicit antibodies to DEN-I and DEN-3. This lack of significant cross-reactivity towards DEN-I and DEN-3 suggests that EDIII-based immunogens may eliminate the risk of ADE. Finally, EDIII is highly stable to denaturation and can function as an independently folding domain as evidenced by its release as a discrete fragment upon tryptic digestion of intact DEN virions. This makes it amenable to a beads-on-a-string design in creating a tetravalent antigen. In conclusion, the many attributes of domain III discussed above reinforce the notion that it can serve as a very effective dengue vaccine candidate.

Having chosen EDIII as the precursor for our tetravalent chimeric antigen, P. pastoris is used as the expression host on the basis of the following reasoning: for recombinant EDIII protein to serve as potent antigen capable of eliciting neutralizing antibodies, it must be properly folded in a conformation that maintains the integrity of its neutralizing epitopes. E. coli that has an overall reducing intracellular milieu is not a conducive host for S-S bond formation. In fact, the monovalent and bivalent EDIII-based antigens alluded to above are produced in the form of insoluble inclusion bodies in E. coli, necessitating their solubilization in denaturing solvents followed by empirical refolding. Clearly, the empirical refolding of the E. oo/i-expressed rEDIII-T protein, with its four S-S linkages, from inclusion bodies would be a highly inefficient process. Consistent with this, refolding of the rEDIII-T protein is difficult with high batch-to-batch variability. Thus, it is necessary to express rEDIII-T protein in a eukaryotic host that will permit proper S-S bond formation and folding. Of the available eukaryotic expression hosts, insect and mammalian cell-based systems can be quite expensive, especially on a commercial scale. Yeasts are unique in that they combine the advantages of both prokaryotes (high expression levels, easy scale-up, inexpensive growth media and easy genetic manipulation) and eukaryotes (capacity to carry out most of the eukaryote- specific post-translational modifications such as glycosylation, disulfide bond formation and protein folding). In the last several years, the methylotrophic yeast Pichia pastoris has emerged as a powerful and inexpensive heterologous system for the production of high levels of functionally active recombinant proteins of commercial and academic interest. In the present invention, a novel recombinant Envelope Domain in Tetravalent (rEDIII-T) protein is developed which elicit protective immune response against all four dengue virus serotypes. This novel recombinant tetravalent protein having SEQ. No. 1: MSYVMCTGSFKLEKEVAETQHGTVLVQVKYEGTDAPCKIPFSTQDEKGVTQNRLITANPIVT DKKPVNIETEPPFGESYIWGAGEKAKQWFKKGSSIGKMFEATARGARRMAILGGGGGMSY AMCLNTFVLKKEVSETQHGTILIKVEYKGEDAPCKIPFSTEDGQGKAHNGRLITANPWTKKE EPVNIEAEPPFGESNIVIGIGDKALKINWYRKGSSIGKMFEATARGARRMAILGGGGGMSYM CGKFSGKFSIDKEMREH'QHGTTVVKVKYEGAGAPCKVPIEIRDVNKEKVVGRIISSTPLAENT NSVTNIELERPLDSYIVIGVGNALTLHWFRKGSSIGKMFESTYRGAKRMAILGGGGGMSYSM CTGKFKVVKEIAETQHGTIVIRVQYEGDGSPCKTPFEIMDLEKRHVLGRErrTVNPIVTEKDSP VNIEAEPPFGDSYIIIGVEPGQLKLDWFKKGSSIGQMFETTMRGAKRMAILGGGGGHHHHH HX) does not match with any protein in the protein sequence database. This novel protein is encoded by EDIII-encoding sequences corresponding to the E proteins of all four DEN virus serotypes. Fusing these sequences generates a novel polynucleotide sequence having SEQ. No. 2:

GAATTCACTATGTCCTACGTTATGTGTACTGGTTCCTTCAAGTTGGAGAAGGAAGTTGCTG AAACTCAGCACGGTACTGTTTTGGTTCAGGTTAAGTACGAAGGTACTGACGCTCCATGTAA GATCCCATTCTCCACTCAAGATGAGAAGGGTGTTACTCAGAACAGATTGATCACTGCTAAC CCAATCGTTACTGACAAGAAGCCAGTTAACATCGAAACTGAGCCACCATTCGGTGAATCCT ACATCGTTGTTGGTGCTGGTGAAAAGGCTAAGCAGTGGTTCAAGAAGGGTTCCTCCATCG GTAAGATGTTCGAGGCTACTGCTAGAGGTGCTAGAAGAATGGCTATCTTGGGTGGTGGTG GTGGAATGTCTTACGCTATGTGTTTGAACACΠTTCGTTTTGAAGAAGGAGGTTTCCGAGAC TCAACACGGTA(-TATCNTGATCAAGGTTGAGTACAAGGGTGAAGATGCTCCTTGTAAGATT CCATTCTCAACTGAGGACGGTCAAGGTAAGGCTCATAACGGTAGATTGATTACAGCTAATC CAGTTGTTACTAAGAAGGAGGAGKD(^GTTAACATTGAAGCTGAACCACCTTTCGGAGAGT

CCAACATCGTTATCGGTATCGGTGACAAGCKΠTTGAAGATCAACTGGTACAGAAAGGGAT CTTCTATTGGAAAGATGTTTGAAGCTACAGARAGAGGAGCTAGAAGAATGGCTATTTTGGG AGGTGGTGGAGGAATGTCCTACATGTGTGGTAAGTTCTCCGGAAAATTCTCCATTGACAA GGAGATGAGAGAGACTCAGCACGGAACTACAGTTGTTAAGGTTAAGTATGAGGGTGCTGG TGCTCCATGTAAAGTTCCTATCGAGATCAGAGATGTTAACAAGGAGAAAGTTGTTGGTAGA ATCATCTCCTCCACTCCATTGGCTGAAAACACTAACTCCGTTACAAACATCGAGTTGGAGA GACCATTGGACTCCTACATTGTTATCGGTGTTGGTAACGCTTTGACTTTGCACTGGTTTAG AAAGGGATCATCAATCGGAAAAATGTTCGAGTCCACTTACAGAGGTGCTAAGAGAATGGC TATCITGGGAGGTGGTGGAGGAATGTCTTACTCCATGTGTACTGGAAAGTTCAAAGTTGTT AAGGAGATCGCTGAGACACAGCATGGTACTATCGTTATCAGAGTTCAGTACGAGGGTGAT GGTTCCCCTTGTAAGACTCCATTCGAGATCATGGACTΓGGAGAAGAGACACGTTTTGGGA AGATTGACTACTGTTAACCCTATTGTTACAGAGAAGGACTCCCCTGTTAATATCGAGGCTG AGCCΓCCATTTGGTGACTCTTACATCATCATCGGTGTTGAGCCTGGTCAATTGAAGTTGGA CTGGTTTAAGAAAGGATCTTCCATTGGTCAAATGTTCGAGACTACTATGAGAGGAGCTAAG AGAATGGCTATTTTGGGTGGAGGAGGTGGACATCATCACCATCACCACTAAGAATTC.

The rEDIH-T gene and amino acid (aa) sequences corresponding to envelope domains HI of each dengue virus serotypes are shown below. The EcoRI sites are designed to facilitate cloning of the gene into P. pastoris expression vector. The envelope domain HI of dengue virus serotypes genes are linked together using penta-glycine linkers (black). A 6x His tag at the carboxy-terminus has been added followed by the stop codon.

1/1 31/11

[GAATTq ACT ATG TCC TAC GTT ATG TGT ACT GGT TCC TTC AAG TTG GAG AAG GAA GTT GCT Met ser tyr val met cys thr gly ser phe lys leu glu lys glu val ala 61/21 91/31

GAA ACT CAG CAC GGT ACT GTT TTG GTT CAG GTT AAG TAC GAA GGT ACT GAC GCT CCA TGT glu thr gin his gly thr val leu val gin val lys tyr glu gly thr asp ala pro cys

121/41 151/51

AAG ATC CCA TTC TCC ACT CAA GAT GAG AAG GGT GTT ACT CAG AAC AGA TTG ATC ACT GCT lys ile pro phe ser thr gin asp glu lys gly val thr gin asn arg leu ile thr ala

181/61 211/71

AAC CCA ATC GTT ACT GAC AAG AAG CCA GTT AAC ATC GAA ACT GAG CCA CCA TTC GGT GAA asn pro ile val thr asp lys lys pro val asn ile glu thr glu pro pro phe gly glu

241/81 271/91 TCC TAC ATC GTT GTT GGT GCT GGT GAA AAG GCT AAG CAG TGG TTC AAG AAG GGT TCC TCC ser tyr ile val val gly ala gly glu lys ala lys gin trp phe lys lys gly ser ser

301/101 331/111

ATC GGT AAG ATG TTC GAG GCT ACT GCT AGA GGT GCT AGA AGA ATG GCT ATC TTG GGT GGT ile gly lys met phe glu ala thr ala arg gly ala arg arg met ala ile leu gly gly 361/121 391/131

GOT GGT GGA ATG TCT TAC GCT ATG TGT TTG AAC ACT TTC GTT TTG AAG AAG GAG GTT TCC gly gly gly met ser tyr ala met cys leu asn thr phe val leu lys lys glu val ser

421/141 451/151

GAG ACT CAA CAC GGT ACT ATC TTG ATC AAG GTT GAG TAC AAG GGT GAA GAT GCT CCT TGT glu thr gin his gly thr ile leu ile lys val glu tyr lys gly glu asp ala pro cys

481/161 511/171

AAG ATT CCA TTC TCA ACT GAG GAC GGT CAA GGT AAG GCT CAT AAC GGT AGA TTG ATT ACA lys ile pro phe ser thr glu asp gly gin gly lys ala his asn gly arg leu ile thr 541/181 571/191 GCT AAT CCA GTT GTT ACT AAG AAG GAG GAG CCT GTT AAC ATT GAA GCT GAA CCA CCT TTC ala asn pro val val thr lys lys glu glu pro val asn ile glu ala glu pro pro phe

601/201 631/211

GGA GAG TCC AAC ATC GTT ATC GGT ATC GGT GAC AAG GCT TTG AAG ATC AAC TGG TAC AGA gly glu ser asn ile val ile gly ile gly asp lys ala leu lys ile asn trp tyr arg 661/221 691/231

AAG GGA TCT TCT ATT GGA AAG ATG TTT GAA GCT ACA GCT AGA GGA GCT AGA AGA ATG GCT lys gly ser ser ile gly lys met phe glu ala thr ala arg gly ala arg arg met ala

721/241 751/251

ATT TTG GGA GGT GGT GGA GGA ATG TCC TAC ATG TGT GGT AAG TTC TCC GGA AAA TTC TCC ile leu gly gly gly gly gly met ser tyr met cys gly lys phe ser gly lys phe ser

781/261 811/271

ATT GAC AAG GAG ATG AGA GAG ACT CAG CAC GGA ACT ACA GTT GTT AAG GTT AAG TAT GAG ile asp lys glu met arg glu thr gin his gly thr thr val val lys val lys tyr glu

841/281 871/291 GGT GCT GGT GCT CCA TGT AAA GTT CCT ATC GAG ATC AGA GAT GTT AAC AAG GAG AAA GTT gly ala gly ala pro cys lys val pro ile glu ile arg asp val asn lys glu lys val

901/301 931/311

GTT GGT AGA ATC ATC TCC TCC ACT CCA TTG GCT GAA AAC ACT AAC TCC GTT ACA AAC ATC val gly arg ile ile ser ser thr pro leu ala glu asn thr asn ser val thr asn ile 961/321 991/331

GAG TTG GAG AGA CCA TTG GAC TCC TAC ATT GTT ATC GGT GTT GGT AAC GCT TTG ACT TTG glu leu glu arg pro leu asp ser tyr ile val ile gly val gly asn ala leu thr leu

1021/341 1051/351

CAC TGG TTT AGA AAG GGA TCA TCA ATC GGA AAA ATG TTC GAG TCC ACT TAC AGA GGT GCT his trp phe arg lys gly ser ser ile gly lys met phe glu ser thr tyr arg gly ala

1081/361 1111/371

AAG AGA ATG GCT ATC TTG GGA GGT GGT GGA GGA ATG TCT TAC TCC ATG TGT ACT GGA AAG lys arg met ala ile leu gly gly gly gly gly met ser tyr ser met cys thr gly lys 1141/381 1171/391 TTC AAA GTT GTT AAG GAG ATC GCT GAG ACA CAG CAT GGT ACT ATC GTT ATC AGA GTT CAG phe lys val val lys glu ile ala glu thr gin his gly thr ile val ile arg val gin 1201/401 1231/411 TAC GAG GGT GAT GGT TCC CCT TGT AAG ACT CCA TTC GAG ATC ATG GAC TTG GAG AAG AGA tyr glu gly asp gly ser pro cys lys thr pro phe glu ile met asp leu glu lys arg 1261/421 1291/431 CAC GTT TTG GGA AGA TTG ACT ACT GTT AAC CCT ATT GTT ACA GAG AAG GAC TCC CCT GTT his val leu gly arg leu thr thr val asn pro ile val thr glu lys asp ser pro val 1321/441 1351/451 AAT ATC GAG GCT GAG CCT CCA TTT GGT GAC TCT TAC ATC ATC ATC GGT GTT GAG CCT GGT asn ile glu ala glu pro pro phe gly asp ser tyr ile ile ile gly val glu pro gly 1381/461 1411/471 CAA TTG AAG TTG GAC TGG TTT AAG AAA GGA TCT TCC ATT GGT CAA ATG TTC GAG ACT ACT gin leu lys leu asp trp phe lys lys gly ser ser ile gly gin met phe glu thr thr 1441/481 1471/491 ATG AGA GGA GCT AAG AGA ATG GCT ATT TTG GGT GGA GGA GGT GGA CAT CAT CAC CAT CAC met arg gly ala lys arg met ala ile leu gly gly gly gly gly his his his his his 1501/501-S∞ R I CAC TAA I GAA TTC]

It has been shown that in creating chimeric proteins, glycines (because of its lack of a carbon side chains), are preferred linker amino acid residues. Accordingly, a penta- glycine linker is used to join the four EDIIIs so that they may retain their structural integrity without being subjected to any constraints at the fusion junction. This novel protein is used to address the following questions: 1) would it retain the antigenic identity of its monovalent precursors? 2) would it elicit antibodies specific to each of its constituent serotypes? 3) Would these antibodies be effective in recognizing and neutralizing the infectivity of DEN-I, DEN-2, DEN3 and DEN-4 viruses?

With a view to simplifying downstream processing, an attempt is made to secrete the rEDIII-T protein by inserting the S. cerevisiae α-factor secretion signal at its amino- terminus. However, this is not successful. Therefore, the intracellularly expressed recombinant protein is purified by lysing the cells in presence of a non-ionic detergent to dissociate this protein from the host membranes. Further, it is necessary to purify this protein in presence of urea on Ni-NTA column, as in the absence of urea, this protein failed to bind to Ni-NTA column and remained in the flow-through, suggesting that the C terminal 6x His Tag, is presumably inaccessible to interact with Ni-NTA group, under native conditions during protein purification. However, the purified protein remained soluble in the absence of urea. The yield of rEDIII-T protein from a liter culture of Pichia pastoris is about 40 mg.

In order to evaluate its immunogenicity, the purified rEDIII-T protein is injected ^' intra-peritoneally into BALB/c mice with various adjuvants like FCA, Alum or Montanide. The question at this point is: would the antibodies elicited by rEDIII-T bind to all serotypes of DEN viruses? If so, would such binding block virus adsorption to host cells and neutralize virus infectivity? Using an immunofluorescence approach to visualize cell surface-bound virus, it is observed that the anti-rEDIII-T antisera could detect all the four dengue virus serotypes in infected BHK cells. In another experiment, pre-incubation of dengue viruses with anti-rEDIII-T antisera prevents the binding of all four dengue viruses to the host cell surface. It is likely that because EDIII is involved in host receptor recognition, antibodies elicited by this protein specifically bind to EDIII on the dengue virion surface and thereby preclude its binding to host cell surface. This is borne out by PRNT data, which shows that the infectivity of all four dengue viruses could be effectively neutralized by the anti-rEDIII-T antisera. The PRNT50 titres are approximately 1:160 for all four viruses. The neutralizing titers reported recently, using EDIII-based plasmids are approximately 1:10. In contrast, the EDIII-MBP proteins are reported to elicit much higher levels of neutralizing antibodies. The observed differences are very likely a re- flection of the nature of antigen, route of immunization, and most importantly, differences in the experimental parameters of the PRNT assay. In the absence of a good animal model for dengue, neutralizing antibody titres are widely accepted as surrogate markers of' protective immunity. The PRNTso titers of 1:10 are considered indicative of protective immunity. The results clearly demonstrate the potential of the rEDIII-T protein to elicit neutralizing, and therefore, presumably protective antibodies against all four dengue virus serotypes. As high titre neutralizing antibodies are observed to all four serotypes, it is believed that the ADE phenomenon should not be an issue.

T cell responses are studied by monitoring the magnitude of cell proliferation and production of the cytokines IFN-γ and IL-4 in splenocytes obtained from immunized mice in response to virus stimulation in vitro. Splenocytes from all the three groups of mice displayed pronounced proliferative response upon incubation with any of the four dengue viruses. Of the two cytokines investigated, IL-4 secretion by splenocytes from rEDII-T immunized animals (from all three adjuvant groups) is enhanced significantly in response to in vitro stimulation by each one of the DEN virus serotypes. The IL-4, which is a B-cell stimulatory cytokine, may contribute to the observed dengue virus-neutralizing antibody response. In striking contrast, in vitro stimulated splenocytes manifested barely discernible increase in secreted IFN-γ levels. Overall, the data indicate that the T cell response elicited is predominantly Th2 type.

In the present invention, a novel tetravalent chimeric protein by fusing the BDIIIs of DEN-I, DEN-2, DEN-3 and DEN-4 viruses using a flexible linker is developed. The recombinant protein is expressed in Pichia pastoris and purified to near homogeneity. Our results indicate that the rEDIII-T protein is immunogenic in the presence of all three-tested adjuvants. The polyclonal antibodies raised against rEDIII-T protein, recognized all four dengue viruses equally well, as observed by immunofluorescence analysis of the dengue infected cells. Moreover, these antibodies are able to neutralize all four serotypes of dengue viruses using PRNT assays. This indicated that our tetravalent protein design of host cell receptor binding domain III of each of the four viruses linked to each other through penta-glycine linkers, is able to display individual domain III of each serotype to the host immune system. The results also indicate that rEDIII-T antigen is able to generate T cell response against all four dengue serotypes. The presence of neutralizing antibodies and T cell response to all four serotypes suggests that this molecule is an effective vaccine candidate against all four strains of dengue virus. Coupled with the high expression capacity of the Pichia pastoris system and easy one-step affinity purification, this strategy has the potential to lead to the development of a cost-effective tetravalent dengue vaccine.

The present invention is illustrated and supported by the following examples. These are merely representative examples and optimization details and are not intended to restrict the scope of the present invention in any way.

BtXAMPLB - 1: Creation of recombinant P. pastoris clone expressing EDIII-based tetravalent gene DEN-I (Nauru Island), DEN-2 virus (NGC strain), DEN-3 (H87) and DEN-4

(Dominica) viruses were taken. Fusing EDIII-encoding sequences corresponding to the E proteins of all four DEN virus serotypes generates a tetravalent gene, rEDIII-T. Adjacent EDIIIs were joined using flexible pentaglycine peptide linkers, as depicted in Figure 1. Adjacent EDIIIs were joined using flexible pentaglycine peptide linkers. Each EDIII is composed of - 120 aa residues spanning aa 296-415 of the corresponding E protein. Each of these domains possesses the single S-S bond (Cys 302-Cys 333) that is critical for antigenicity.

A 6x His tag was engineered at the carboxy terminus to aid in the detection and purification of this recombinant protein. The tetravalent antigen is predicted to be ~55 kDa in size. A gene encoding this antigen, codon-optimized for expression in P. pastoris, was chemically synthesized and cloned into the P. pastoris integrative vector pPIC9K as a ~1.5 kilobase (Kb) Bco RI/ Not I restriction fragment. In this construct, the tetravalent antigen-encoding gene was fused in-frame with the S. cerevisiae a factor secretory signal, under the transcriptional control of the strong methanol- inducible AOXl promoter. This resultant plasmid, pPIC-EDIII-T, is shown in Figure IA. A schematic representation of the rEDIII-T protein (precursor and its processed forms), is shown in Figure IB.

The pPIC-EDIII-T plasmid was digested with BgI II to release the tetravalent antigen expression cassette together with the HIS4 and Kan antibiotic selection markers, and transformed into the P. pastoris host strain GS 115, which carries an intact AOXl locus (Mut*). As both ends of this BgI II fragment are homologous to the AOXl region of the GSl 15 genome, it can integrate into the AOXl locus by a double crossover event, with concomitant elimination of the host AOXl gene. Successful replacement of this AOXl gene will generate an aoxl strain (Mut^s) that can grow in the presence of kanamycin, on minimal media lacking histidine.

P. pastoris transformants harboring the inducible tetravalent DEN antigen expression cassette were selected on His- plates and their Mut phenotype determined to be Mut^s. The presence of the tetravalent DEN antigen-encoding gene insert in these His*Mut^s transformants was verified by PCR using specific primers.

EXAMPLE - 2: Expression and purification of recombinant EDIII-T protein

A one-liter culture of the P. pastoris transformant, harboring the rEDIII-T gene growing at logarithmic phase, was spun down and induction initiated by re- suspending the cell pellet in 100 ml of 1% (vol/vol) methanol-containing medium. Cells were harvested 96 hours post-induction. The pellet (-40 grams wet weight) was . washed twice with 500 ml cold distilled water and re-suspended in 200 ml lysis buffer [50 mM phosphate buffer (pH 8.0J/500 mM NaCl/ 1OmM imidazole/6 M guanidine- HCl/ ImM phenyl methyl sulfonyl fluoride], followed by stirring for -90 minutes at room temperature (RT). This suspension was mixed with 30Og of 425-600 micron glass beads (Sigma, catalog # O-8772) and the cells disrupted using a bead mill (Dyno Mill, Multi Lab, WAB, Basel) in four 10-minute cycles, at 4°C. The lysate was collected and kept on ice. The beads were washed twice with - 150 ml cold lysis buffer and pooled with the lysate. The pooled lysate (volume -500 ml) was then clarified by centrifugation in a Sorvall OSA rotor at 10,000 rpm for 1 hour at 4°C. The resultant supernatant was further clarified by passing it through a 0.45μm membrane filter, adjusted to pH 8 (using 1 N NaOH) and then allowed to bind to 20 ml of Ni-NTA superflow resin [50% (wt/vol) slurry] overnight at RT. Lysate/Ni-NTA mixture was loaded into a column and the flow-through was collected. The column was washed with 15 bed- volumes of wash buffer A [50 mM phosphate (pH 6.3) /8 M urea/ 150 mM NaCl/ 10 mM imidazole] followed by 15 bed volumes of wash buffer B [50 mM phosphate (pH 5.9)/8 M urea]. Finally, the bound proteins were eluted using buffer C [50 mM phosphate (pH 4.5J/8 M urea]. A single major homogenous protein peak emerged in the pH 4.5 eluate as shown in Figure 2A. This was dialyzed against Ix PBS (pH -8.5) to eliminate urea and neutralize the acidic pH.

Recombinant 6x-His-tagged EDIII-T protein levels in soluble cell extracts and Ni-NTA column eluates were analyzed using Ni-NTA His-Sorb plates. The protein in sample aliquots (100 μl) was allowed to bind to the wells for -4 hours at 37⁰C. After washinglx PBS-0.1% Twen 20 (Ix PBS-T), bound protein in each well was detected using 100 μl of 1: 4,000 diluted anti- EDIII-T polyclonal murine antiserum (raised against E. coli expressed rEDIII-T protein) in conjunction with HRPO-conjugated secondary antibody and TMB substrate. A summary of the purification is presented in Table 1. Starting from -40 g of induced cell mass, -54 mg of purified tetravalent DEN antigen was obtained, representing a -60 fold purification with a -40% yield.

Table 1: Purification of Pichia pastoris-expressed rEDIII-T protein

Step Total Total Specific Fold Yield • protein* ELISA activity ^c purification ^d (%) (mg) ODs «> (x 10³)

(X 105)

Cell lysate 8030 8.25 0.103 1.0 100 Ni-NTA 54 3.30 6.115 59.4 40 chromatography

^a Total protein was estimated with Bradford reagent using BSA as reference ^b ELISA ODs in appropriately diluted samples were measured at 450 nm using Ni-NTA His Sorb kit. c Represents ELISA OD units per mg protein d Obtained by dividing specific activity at a given step by the specific activity of the crude lysate e Yields were based on ELISA ODs, taking the total ELISA ODs in the cell lysate as ^100%

An SDS-PAGE analysis of the peak fractions is presented in Figure 2B. It is evident that the homogenous peak in fact contains a major band with an apparent molecular weight of -80 kDa and a minor band of -55 kDa. While it could be argued that the -55 kDa protein represents the intended rEDIII-T protein, the identity of the -80 kDa protein is not clear. To investigate this issue further, another version of the rEDIII-T protein, lacking the S. cerevisiae a factor secretory signal peptide, in an E. coli expression system was generated and was used to raise anti-rEDIII-T polyclonal antibodies in mice. EtXAMPLB - 3: Immunoblot assay

The purified, P. pαstoris-expressed rEDIII-T protein was detected in Western blots. The primary antibodies used for detection were anti penta-His mAb (catalog number 34660, from Qiagen) and anti-EDIII-T polyclonal serum (raised in mice using B. ooK-expressed EDIII-T protein). After appropriate dilution with Ix SDS sample disruption buffer, the rEDIII-T protein preps were loaded onto a SDS- 10% polyacrylamide gel. Separated proteins were electro-transferred using Mini Trans-blot Electrophoretic Cell (Bio-Rad) onto a nitrocellulose membrane (Hybond-C, Amersham, England), in the presence of 200 mM glycine/24 mM Tris/20% (vol/vol) methanol at 100V for 1.5 hours, at 4⁰C. Transfer of proteins onto the membrane was visualized by kaleidoscopic markers. The membrane was rinsed briefly in Ix PBS-T and then incubated in blocking solution [1% (wt/vol) PVP (average Mr 40,000)/5% (vol/vol) NGS in Ix PBS] for 4 hours with gentle shaking at RT. The blocking solution was replaced with the primary antibody (anti-penta His mAb at 1:2000 dilution and anti-EDIII-T polyclonal antiserum at 1:6000 dilution) diluted in O.lx blocking buffer. After -2 hours of incubation, the blots were washed 5 times with Ix PBS-T for 10 minutes each, and transferred into alkaline-phosphatase conjugated secondary antibody solution (1:5,000 dilution in blocking solution) and further incubated for 1 hour. The blots were washed again as described above. The protein-antibody complex was developed in 0.1 M Tris- HCl (pH 9.5), 0.1 M NaCl, 5mM MgCb containing 150 μg ml ¹ of Nitro Blue Tetrazolium and 75 μg ml'¹ 5-Bromo-4-chloro-3-indolyl phosphate. The reaction was stopped after ~15 minutes of incubation in a light-proof dish by rinsing the blot in 10 mM EDTA (pH 8.0).To detect glycosylation, blots were probed with Concanavalin A (Con A)-FITC (Fluka, Cat. # 61761) and visualized using Ettan DIGE imager (excitation/emission: 495nm/518nm) from OE Healthcare.

Figure 2C show that both the ~80 and -55 kDa forms of the P. pastoris- expressed proteins possess the C-terminal 6x His Tag, as evidenced by their reactivity to penta-His mAb (lane 6), consistent with their tight binding to the Ni-NTA affinity matrix (Figure 2B). Further, the -55 kDa yeast protein co-migrated with the E. coli- expressed tetravalent antigen (lane 4). These results were mirrored in the immunoblot analysis performed using the polyclonal antiserum raised against the E. coli- expressed tetravalent antigen (lanes 1 and 3).

While these data suggest that the -80 kDa protein is presumably the unprocessed precursor of the -55 kDa protein, it is difficult to reconcile the apparent size difference between these two given that the size of the α factor secretory signal peptide is about -10 kDa. An obvious possibility is that the Pichia pαstoπs-expressed protein may be heavily glycosylated. This indeed turned out to be the case as shown in the experiment in Figure 2D. In this experiment, both the B. coli- and Pichia pαstons-expressed rEDIII-T protein preparations were electrophoresced, blotted onto nitrocellulose and probed with Con A-FITC. This experiment demonstrates that the -80 kDa protein is indeed heavily glycosylated (compare lanes 1 and 3). In contrast, the -55 kDa protein does not bind Con A-FITC (see lanes 2 and 3). As there are no glycosylation sites in EDIII₁ the data suggest that the unprocessed signal peptide of the ~80 kDa protein is presumably the substrate for glycosylation in Pichia pastoris. This suggestion is consistent with the presence of two N-glycosylation sites in the S. oereυisiae α factor secretory signal peptide. Taken together, these data indicate that the two forms present in the protein preparation purified from P. pastoris, indeed represent the rEDIII-T antigen and that signal cleavage is relatively inefficient. The purity of the Ni-NTA purified yeast recombinant tetravalent antigen is judged to be ≥95%.

EtXAMPLB — 4J Immunization of mice Groups of four Balb/c mice (4-6 weeks old) were immunized intra-peritoneally on days 0, 21, 42 and 84 with 20 μg purified P. pαstoris-expressed rEDIII-T protein. The recombinant protein antigen was formulated using alum, Freund's complete adjuvant, and montanide ISA 720. In the case of the Freund's group alone, the booster doses were formulated using Freund's incomplete adjuvant. Sera were collected from the animals ~ 1 week after the final immunization by retro-orbital puncture. Animal experiments were reviewed and approved by the International Centre for Genetic Engineering and Biotechnology Institutional Animal Ethics Committee and adhered to the guidelines of the Government of India. E-XAMPLE - 5; Detection of anti-DEN virus antibodies in immune sera

Antibodies specific for each of the four DEN viruses in the sera of immunized animals were detected by ELISA using tissue culture-derived DEN viruses as coating antigens. Starting from one hundred fold dilution, serial two-fold dilutions of each serum sample was assayed in duplicate against each of the four DEN virus serotypes. Ninety-six-well plates were separately coated, overnight, with of 1:5 diluted

DEN-I, DEN-2, DEN-3 and DEN-4 virus stocks (100 μl/well). Virus coated plates were blocked 2% PVP in Ix PBS for -2-4 hours at 4⁰C, washed 3 times (with Ix PBS-T) and incubated with serial two-fold dilutions (prepared in blocking buffer) of the individual mouse serum samples for ~2 hours at 37⁰C. The wells were washed 5 times (using Ix PBS-T) and incubated with anti-mouse IgG-HRPO conjugate (diluted 1:5,000 in blocking buffer) for 1 hour at 37⁰C, washed again and incubated with 3, 3' 5, 5' tetramethyl benzidine substrate (lOOμl/well) for 20 minutes at 37⁰C. The color reaction was terminated by the addition of IM H2SO4 and the absorbance read at 450 nm. Figure 3 clearly demonstrate that the immune sera did indeed contain antibodies specific to each of the DEN virus serotypes. Among the three adjuvants tested, the use of Freund's adjuvant elicited antibodies that manifested maximal ELISA reactivity towards each serotype, with antibody levels against DEN-I and DEN- 2 being slightly higher than those against DEN-3 and DEN-4. However, the difference was less than two fold. For example, at a 100^:fold serum dilution, anti-DEN-3 and anti-DEN-4 ELISA reactivities were, respectively, -30 and -40% lower. The use of montanide and alum as adjuvants produced comparable, but slightly lower levels of antibodies against all the four serotypes. Again, differences were minimal. In fact, ELISA profiles obtained using DEN-2, DEN-3 and DEN-4, as capture antigens, were practically indistinguishable from each other. ELISA reactivities of the immune sera against DEN-I were slightly higher, but within two-folds, than those against the remaining three serotypes.

EXAMPLE - 6: Immunofluorescence assay The ELISA data were essentially corroborated by the immunofluorecence experiment shown in Figure 4. Baby hamster kidney cells (BHK 21) (American Type Culture Collection, Virginia, USA) were seeded on coverslips ~24 hours prior to infection. Cells were infected separately with each of the four DEN virus serotypes, when they were 80% confluent. At 24 hours post-infection, the growth medium was removed and the cells on cover slips were rinsed three times with Ix PBS and fixed for 15 minutes in 2% formaldehyde, rinsed once with Ix PBS and permeabilized with ice- cold methanol for 15 minutes at 4°C. Cells were rinsed once again and blocked with Ix PBS/2% PVP/0.1% NGS (blocking buffer). Cells were washed three times with Ix PBS-T and incubated with a primary antibody, diluted 1:50 in 0. Ix blocking buffer/ Ix PBS, for 90 minutes at RT. For the positive control experiments, polyclonal murine antisera, raised against E.ooh'-expressed rEDIII proteins were used, as the source of primary antibodies. Cells were washed extensively and incubated further for an hour with fluorescein-conjugated anti-mouse IgG from Calbiochem (1:40 dilution in O.lx blocking buffer/ Ix PBS). After rinsing in Ix PBS-T, the cells were mounted on glass slides in the presence of antifade fluoroguard reagent and visualized under a Nikon microscope equipped for incident illumination with a narrow band filter combination selective for FITC. BHK21 cell line were maintained in Dulbecco's modified Eagle medium (DMEM), supplemented with 10% (v/v) fetal calf serum (FCS), in a 10% CO2 humidified incubator, at 37°C. Pre-immune serum, as expected, did not produce immunofluorescence with any of the four DEN viruses. However, consistent with the ELISA data, immune sera picked up all four DEN viruses as evidenced by immunofluorecsence. Figure 4 shows that while the immune serum used in the experiment was from the Freund's group, essentially similar results were obtained using immune sera from the alum and montanide groups. Positive control experiments were run in parallel, as reported earlier, to ensure that each of the four DEN viruses had successfully infected the BHK cells.

Both ELISA and immunoassay demonstrate that the tetravalent antigen has the potential to induce antibodies that can recognize and bind to each of the four DEN serotypes. The question that arose at that point was that will such binding neutralize the infectivity of these viruses? To address this, each of the four DEN viruses separately with the anti-rEDIII-T immune serum was pre-incubated and then infected BHK cells. Unbound viruses were washed away and bound virus, if any, was visualized by indirect immunofluorescence assay. This experiment revealed that anti- rEDIII-T antiserum, from all three adjuvant groups, could effectively block the binding of all four DEN viruses to BHK cells. This led to the conclusion that immunization with the P. pαstoris-expressed rEDIII-T antigen can induce virus-neutralizing antibodies specific to all four DEN virus serotypes simultaneously.

EXAMPLE - 7: Hague Reduction Neutralization Test fPRNTl In order to quantitate the levels of neutralizing antibodies, PRNT assay was performed. In this, DEN-I, DEN-2, DEN-3 and DEN-4 viruses (about 120 PFU each), prepared from infected tissue culture supernatants by polyethylene glycol precipitation, were separately pre-incubated with serial two-fold dilutions of heat- inactivated (56⁰C/ 10 min) pooled serum (200μl final volume) collected from rEDIII-T- immunized mice, for 1 hour at 37⁰C, and plaqued on the monkey kidney cell line LLCMK2 cells (American Type Culture Collection, Virginia, USA) seeded in 24-well plates (50μl virus-antiserum mix/well) to determine the residual virus titers.

After overnight pre-incubation at 4°C, the virus/serum mixture was diluted to a final volume of 200 μl with DME + 2%ΔFCS and used to infect a single well of 24- well plate. Each serum dilution was assayed in duplicate wells. After adsorption for 2 hours, the inoculum was aspirated off and cells overlaid with 0.8% methylcellulose in DME+6%ΔFCS (1 ml/ well). Appropriate controls were set up in parallel with negative control (mock-infected) wells receiving 200μl DME + 2%ΔFCS and positive control wells receiving DEN-I, DEN-2, DEN-3 or DEN-4 viruses, separately, pre-incubated with plain DME+ 2%ΔFCS, instead of murine serum. The plates were incubated at 37°C in a humidified 5% CO2 incubator. On day 6 post-infection, the overlay was gently decanted and the cells were fixed with 1 ml of 4% formaldehyde solution at RT for 1 hour. Wells were washed with tap water and then stained with 1:40 diluted stock of 2% crystal violet solution in 20% ethanol for half hour. Plaques revealed after staining were counted and the antiserum dilution resulting in 50% reduction in plaque count (with reference to the number of plaques generated by the virus in the absence of antiserum), was expressed as the PRNT50 titer. LLCMK2 cell lines were . maintained in Dulbecco'a modified Eagle medium (DMEM), supplemented with 10% (v/v) fetal calf serum (FCS), in a 10% CO2 humidified incubator, at 37°C. Figure 5 depicts the percent inhibition of infectivity of each DEN virus serotype, as a function of dilution of the immune serum. The figure shows that the recombinant tetravalent antigen does indeed induced neutralizing antibodies simultaneously against all four DEN virus serotypes. Dilutions of immune sera, from the Freund and montanide groups, beyond 1:100 resulted in 50% inhibition of infectivity of DEN-I, DEN-2, DEN-3 and DEN-4. This was true of immune sera from the alum group as well, but only for DEN-I, DEN-2 and DEN-4. In contrast, the capacity of immune serum from the alum group to inhibit DEN-3 was relatively lesser. Geometric mean PRNT50 titers computed from the data shown in Figure 5 are summarized in Table 2. These data show that montanide was a better adjuvant than alum to elicit high titer neutralizing antibodies against all four DEN virus serotypes.

Table 2: Neutralizing antibody responses in rEDiπ-T immunized mice

Adjuvant PRNT50 Titres^a

Group

DEN-I DEN-2 DEN-3 DEN-4

Alum 197 363 47 546

Montanide 153 293 271 588

Freund's 160 118 234 479 ^aQeometric Mean Titers (n=3)

EXABIPLB - 8

Proliferation and cytokine release assays

Spleens were harvested from mice that had been primed 10 days earlier. Splenocytes were seeded in 96-well plates in DME+10%ΔFCS (2.5 x 10⁵ cells in 0.1 ml/well) and either mock-stimulated (no antigen) or stimulated in vitro, separately, with each of the four DEN viruses (0.03 PFU/cell), for four days. Cells were pulsed with [³H]-thymidine (lμCi/well) for 16 hours at the end of the 4-day incubation and harvested for scintillation counting. Levels of interferon-α (IFN-α) and interleukin-4 (IL-4) in the splenocyte culture supematants were determined by solid phase ELISA using the commercially available" Mouse Thl/Th2 ELISA kit (eBiosciences). Figure 6A shows that while DEN-I, DEN-2 and DEN-4 stimulates splenocytes from mice immunized in the presence of Freund's adjuvant manifests a significant proliferative response (≥IO fold over un-stimulated control cells), DEN-3 stimulation results in a very modest response (~4 fold increase). In contrast, when alum was the adjuvant, splenocyte proliferation was significantly stimulated by DEN-2 (-21 fold) and DEN-4 (~11 fold), in comparison to DEN-I and DEN-3 (~4 fold in both cases). When splenocytes from mice immunized with montanide adjuvanted tetravalent antigen were tested, DEN-I, DEN-2 and DEN-3 stimulated more or less comparable levels of thymidine uptake (~7-9 folds over un-stimulated control), while DEN-4 alone produces a relatively lower magnitude of proliferative response (-4 fold).

Cytokine ELISAs were performed on aliquots of culture supematants obtained from the virus-stimulated splenocytes. In vitro stimulated splenocytes manifests barely discernible increase in secreted IFN-γ levels (Figure 6B). When splenocytes were taken from mice that had been immunized with the tetravalent antigen in the presence of Preund's adjuvant, in vitro stimulation with DEN-I and DEN-3 did not affect IFN-γ levels, but DEN-2 and DEN-4 resulted in -2.8 and -1.8 folds, respectively. Similarly, when alum was used, fold enhancements in IFN-γ levels ranged from 1.3- 3.3 folds. Likewise, DEN-I, DEN-2 and DEN-3 produce very modest elevations in the levels of IFN-γ secretion, which ranged from -2-4 folds. The noteworthy exception was DEN-4 stimulated splenocytes from the montanide group. These splenocytes manifests a -9 fold increase in IFN-γ secretion. In contrast, IL-4 secretion was enhanced significantly in response to in vitro stimulation by each one of the DEN virus serotypes (Figure 6C). In most instances, there was ≥IO fold increase in IL-4 secretion, compared to un-stimulated controls. DEN-3 stimulates splenocytes obtained from mice that had been immunized with montanide adjuvanted antigen manifests a slightly lesser increase in IL-4 secretion levels (-7 fold increase).

Claims

We claim:

1. A recombinant envelop domain - III based tetravalent protein with secretory signal peptide eliciting protective immune responses against each of the four serotypes of dengue virus, DEN-I, DEN-2, DEN-3 and DEN-4, the said protein encoded by a polynucleotide sequence codon optimized for expression in eukaryotic expression system.

2. The recombinant tetravalent protein as claimed in claim 1, wherein said polynucleotide sequence has SEQ ID No.: 2.

3. The recombinant tetravalent protein as claimed in claim 2, wherein said protein has SEQ ID No.: 1.

4. The recombinant tetravalent protein as claimed in any of the preceding claims 1 to 3, wherein said protein is rich in disulphide bonds.

5. The recombinant tetravalent protein as claimed in any of the preceding claims 1 to 3, wherein said protein is of about 55 kDa.

6. The recombinant tetravalent protein as claimed in any of the preceding claims 1 to 3, wherein said envelop domain III of the four serotypes of dengue virus are linked by penta-glycine linkers.

7. The recombinant tetravalent protein as claimed in any of the preceding claims 1 to 3, wherein said eukaryotic expression system is selected from the group consisting of Pichia pastoris, adenovirus vectors and adeno-associated virus vectors.

8. The recombinant tetravalent protein as claimed in any of the preceding claims 1 to 7, wherein said protein results in 50% inhibition of infectivity of each of the four serotypes of dengue virus.

9. A process for the preparation and purification of recombinant tetravalent protein as claimed in any of the preceding claims 1 to 8 comprises:

(a) chemically synthesizing a polynucleotide sequence having SEQ ID No.: 2;

(b) designing the polynucleotide sequence obtained by step (a);

(c) digesting the designed polynucleotide sequence obtained by step (b) with restriction enzymes;

(d) cloning the digest obtained by step (c) in a cloning vector;

(e) digesting the clone obtained by step (d) with the restriction enzyme;

(f) transforming the digest obtained by step (e) in an expression host;

(g) selecting and screening the transformants obtained by step (f) by direct colony PCR;

(h) growing and induction initiating the screened transformants obtained by step (g) by re-suspending the cell pellet in 1% (v/v) methanol containing media;

(i) harvesting the induced cells obtained by step (h) after 96 hours post- induction; (j) purifying the pooled lysate obtained by step (i).

10. The process as claimed in claim 9, wherein said polynucleotide sequence obtained by designing of step (b) contains a 6x histidine tag - encoding sequence at the carboxy terminus followed by stop codon.

11. The process as claimed in claim 9, wherein said restriction enzymes of step (c) are EcoEl and Noll.

12. The process as claimed in claim 9, wherein said digest of step (d) is about 1.5 Kb fragment.

13. The process as claimed in claim 9, wherein said cloning vector of step (d) is pPIC9K with Sacchromyces cerevisiae α-factor secretory signal.

14. The process as claimed in claim 9, wherein said restriction enzyme of step (e) is bgUl.

15. The process claimed in claim 9, wherein said transforming of step (f) is in expression host Pichia pastoris GSl 15.

16. The process as claimed in claim 9, wherein said harvesting of step (i) comprises:

(a) washing the pellet of induced cells twice with cold distilled water;

(b) re-suspending the pellet obtained by step (a) in lysis buffer containing 5OmM phosphate of pH 8.0, 50OmM NaCl, 1OmM imidazole, 6M guanidine-HCl and 1 mM phenyl methyl sulfonyl fluoride;

(c) stirring the cells obtained by step (b) for about 90 min. at room temperature;

(d) disrupting the cells obtained by step (c) using a bead mill in four 10- minute cycle at 4°C by mixing the cells with 425-600micron glass beads.

(e) washing the beads obtained by step (d) twice with cold lysis buffer and pooling with the lysate obtained by step (d).

17. The process as claimed in claim 9, wherein said purifying of pooled lysate of step (j) comprises: (a) centrifuging the pooled lysate at 15,000 rpm for 60 min. at 4°C;

(b) passing the lysate obtained by step (a) through a 0.45μm membrane filter adjusted to pH 8.0;

(c) binding the lysate obtained by step (b) to Ni-NTA agarose (50% wt/vol slurry overnight at room temperature; (d) loading lysate-Ni-NTA mixture obtained by step (c) in column followed by washing with 15 bed-volumes of 5OmM phosphate of pH 6.3, 8mM urea, 150 mM NaCl, 10 mM imidazole buffer followed by 15 bed- volumes of 5OmM phosphate of pH 5.9, 8 mM urea buffer; (e) eluting the bound proteins of step (d) with 50 mM phosphate of pH 4.5, 8M urea buffer; (f) dialyzing the elute obtained by step (e) against IX phosphate buffered saline.

18. The process as claimed in claim 17, wherein the purity of Ni-NTA purified protein is at least 95%.

19. The process as claimed in any of the preceding claims 9 to 18, wherein said purified protein has 40% yield.

20. A recombinant envelop domain - III based tetravalent protein, wherein said protein is without secretory signal and is expressed in E. ooli expression system.