BIOTINYLATED PROTEIN

Abstract

Disclosed are methods of screening a library of molecules to identify or select one or more molecules which selectively bind to a fused protein, or fragment thereof. The fused protein comprises a moiety and a protein selected from a group consisting of flavivirus structural and non-structural (NS) proteins. The method comprises contacting the library of molecules with the fused protein and detecting binding of one or more molecules to the fused protein.

Claims

1. (canceled)

2. A method of screening a library of molecules to identify or select one or more molecules thereof which selectively bind to a fused protein, or fragment thereof, wherein the fused protein comprises a moiety and a protein selected from a group consisting of flavivirus structural and non-structural (NS) proteins, the method comprising: (a) contacting the library of molecules with the fused protein; and (b) detecting binding of one or more molecules to the fused protein.

3. The method according to claim 2, further comprising labeling the molecules with a biotin binding agent and binding is detected by detecting the one or more labeled molecules.

4. The method according to claim 2, wherein the molecules are selected from a group consisting of antibodies and aptamers.

5. The method according to claim 3, wherein the molecules are selected from a group consisting of antibodies and aptamers.

6. The method according to claim 2, wherein the flavivirus structural protein is a flavivirus capsid protein or a flavivirus envelope protein.

7. The method according to claim 2, wherein the moiety is a readily detectable moiety.

8. The method according to claim 7, wherein the moiety is a biotin acceptor signal peptide.

9. The fused protein according to claim 2, wherein the flavivirus is a Dengue virus or a West Nile virus.

10. The fused protein according to claim 8, wherein the biotin acceptor signal peptide is fused with a West Nile envelope Domain III protein.

11. The fused protein according to claim 8, wherein the biotin acceptor signal peptide is fused at the N-terminus of a Dengue virus capsid protein.

12. The fused protein according to claim 2, wherein the fused protein is full-length and non-truncated.

13. The fused protein according to claim 3, wherein the fused protein is full-length and non-truncated.

14. The fused protein according to claim 4, wherein the fused protein is full-length and non-truncated.

15. The fused protein according to claim 5, wherein the fused protein is full-length and non-truncated.

16. The fused protein according to claim 6, wherein the fused protein is full-length and non-truncated.

17. The fused protein according to claim 3, wherein the flavivirus is a Dengue virus or a West Nile virus.

18. The fused protein according to claim 4, wherein the flavivirus is a Dengue virus or a West Nile virus.

19. The fused protein according to claim 5, wherein the flavivirus is a Dengue virus or a West Nile virus.

20. The fused protein according to claim 6, wherein the flavivirus is a Dengue virus or a West Nile virus.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0072] FIG. 1A. Bioinformatics analysis of flavivirus C protein. Multiple sequence alignment of mosquito-borne flavivirus C protein sequences using ClustalW method in MegAlign, DNASTAR Lasergene 7.2 software. Sequences with the highest consensus strength among the nine flavivirus capsids are highlighted.

[0073] FIG. 1B. Bioinformatics analysis of flavivirus C protein. Sequence distances analysis using ClustalV method shows that the protein sequence similarity among the nine capsids are mostly lower than 50%.

[0074] FIG. 1C. Bioinformatics analysis of flavivirus C protein. Antigenic plot of DENV2 C protein generated by Protean, DNASTAR Lasergene 7.2 software, using Jameson-Wolf methodology. The higher the antigenic index, the more likely the region will be recognized by immune system. The secondary structures are also shown in parallel with the antigenic plot. Alpha helixes are shown in rounded rectangles while beta sheets are shown in normal rectangles. The high antigenicity of the first 20 amino acids are highlighted in dotted box. (DENV1-4Dengue virus serotype 1-4; JEVJapanese Encephalitis virus; MVEVMurray Valley encephalitis virus; UVUsutu virus; WNVWest Nile virus; YFVYellow Fever virus; Accession number of the protein sequence is shown in the bracket.)

[0075] FIG. 2A. Cloning strategies. Secondary structure prediction of biotinylated C (BNC) and unbiotinylated C (UBNC) proteins show that the BAP will not affect the overall secondary structure formation when it is added to the N-terminal of C protein. Alpha helixes are shown in gray, rounded rectangles while beta sheets are shown in black, normal rectangles.

[0076] FIG. 2B. Cloning strategies. Schematic diagram showing the overlap extension PCR (OE-PCR) technique. Fragment A is designed such that its 3 overhang is complementary to the 5 overhang of Fragment B. As such, primer B and primer C are complementary to each other. Both fragments are joined together with the complementary sequence and primers A and D.

[0077] FIG. 2C. Cloning strategies. Plasmid map of pET28aDENVBioCap construct.

[0078] FIG. 2D. Cloning strategies. Final construct of recombinant protein generated. 6His tag is at the N-terminal followed by thrombin cleavage site and biotin acceptor peptide (BAP). Enterokinase cleavage site is included at the downstream of BAP. 6His tag is used for affinity purification while BAP is the signal peptide for biotinylation.

[0079] FIG. 3Ai and FIG. 3Aii. Expression screening of bacterial clones with recombinant DENV C protein and detection of biotinylation. Two expression competent cells are tested, namely BL-21 (DE3) and BL-21-CodonPlus. The presence of C protein band in the IPTG-induced bacterial cell lysate, as indicated by arrow, is detected via SDS-PAGE stained with Coomassie blue (FIG. 3Ai) and Western blot using anti-His antibody (FIG. 3Aii). The expression of C protein is much higher in CodonPlus competent cells as the band intensity is much higher than normal BL-21 (DE) competent cells.

[0080] FIG. 3Bi and FIG. 3Bii. Expression screening of bacterial clones with recombinant DENV C protein and detection of biotinylation. FIG. 3Bi. The presence of His tag was detected using monoclonal mouse anti-His antibody and goat anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibody. Bands can be observed in both DENV C proteins with and without biotin acceptor peptide (BAP). The molecular weight of C protein with BAP is higher than C protein without BAP. FIG. 3Bii. When Western-blot is performed using streptavidin-HRP secondary antibody, band is only observed in the DENV C and WNV DIII proteins lanes with BAP. Likewise, band is only observed in the biotinylated (BN) maltose-binding protein (MBP) and not unbiotinylated (UBN) MBP.

[0081] FIG. 4A. Purification of biotinylated DENV C protein. IPTG-induced bacteria are lyzed in lysis buffer containing 8 M urea and the lysate (L) is incubated with nickel-charged resin overnight. Next, the resin is packed in a column and the flow through (FT) is kept for SDS-PAGE. 20 mM imidazole is used to wash away the unbound proteins (W) and 500 mM imidazole is used to elute the C protein in ten fractions (E1-10). SDS-PAGE is carried out and stained with Coomassie blue. Fraction E2 shows the most intense band but there are also many contaminants in the eluates.

[0082] FIG. 4B. Purification of biotinylated DENV C protein. Western blot is also performed for the samples using anti-His antibody. C protein-corresponding bands can be observed from fraction E2 until E10. The band intensity is the highest in fraction E2 and decreases until fraction E10. A faint C protein dimer-corresponding band can also be observed in fraction E2 as detected by anti-His antibody.

[0083] FIG. 4C. Purification of biotinylated DENV C protein. Western blot is also performed for the samples using streptavidin-HRP secondary antibody. C protein-corresponding bands can be observed from fraction E2 until E10. The band intensity is the highest in fraction E2 and decreases until fraction E10.

[0084] FIG. 5Ai. Sequential purification of DENV C protein using fast protein liquid chromatography (FPLC) system after first round of affinity his-tag purification. Dialysed and refolded DENV C protein from eluates E2 to E10 are injected into Resource MonoQ ion exchange chromatography column. Bound proteins are eluted out using increasing concentration of sodium chloride until final NaCl concentration of 1 M. One high peak can be observed when the sodium chloride concentration reaches 350 mM.

[0085] FIG. 5A(ii). Sequential purification of DENV C protein. The identity of the peak is confirmed to be C protein as detected by ELISA using streptavidin-HRP antibody.

[0086] FIG. 5B(i). Sequential purification of DENV C protein. The eluates from the peak are combined and inserted into size exclusion chromatography for further purification. One high peak is observed in the elution profile.

[0087] FIG. 5B(ii). Sequential purification of DENV C protein. The peak observed in FIG. 5B(i) is identified to be C protein as confirmed by ELISA using streptavidin-HRP antibody.

[0088] FIG. 6Ai and FIG. 6Aii. Biotinylation efficiency and functional analysis of purified full-length DENV C protein. FIG. 6Ai ELISA screening of 30 ng biotinylated (BN) and unbiotinylated (UBN) proteins. All three BN maltose-binding protein (MBP), DENV C and WNV DIII proteins show high absorbance at 450 nm as compared to the protein without biotin acceptor peptide (BAP). FIG. 6Aii Biotin-streptavidin binding assay also shows that only BN proteins bind to streptavidin-magnetic beads and are detected in the eluates.

[0089] FIG. 6B. Biotinylation efficiency and functional analysis of purified full-length DENV C protein. The functionality of DENV C protein is examined via binding assay with Sec3 protein which is known to interact with C protein in host cells. Pure Sec3 protein is coated onto ELISA plate and purified full-length DENV C protein is added into the well for binding. Bovine serum albumin (BSA) is used as negative control. Significant absorbance is detected using streptavidin-horseradish peroxidase antibody in the wells with Sec3 protein coated (Sec3-DENVC) but not with BSA (BSA-DENVC). *p-value<0.05

[0090] FIG. 7. FPLC-Gel Filteration Chromatography spectrum corresponding to the West nile virus envelope protein domain III (WNV DIII) and Biotinylated West nile virus envelope protein domain III (BN-WNV DIII).

[0091] FIG. 8. ELISA screening for unbiotinylated and Biotinylated proteins. ELISA data for the control proteins (Unbiotinylated and Biotinylated Maltose Binding protein), Unbiotinylated Dengue Virus2 EDIII, Dengue virus3 EDIII, West Nile Virus Envelope DIII, biotinylated (BN) West Nile Virus Envelope DIII (Batch 1 and 2).

[0092] FIG. 9. Stepwise representation for the production of biotinylated Dengue virus Capsid protein and West Nile virus Envelope protein DIII.

[0093] FIG. 10. Cellular distribution of potential interacting partners of DENV C protein. Total proteins in each cellular compartment are calculated and charted. The diagrams on the top illustrate the cellular localization of each individual protein and some of them may localize in more than one cellular compartment. Majority of the DENV C protein-interacting partners localize in the nucleus followed by cytoplasm. Six proteins are related to cytoskeleton while five are found in the nucleolus and plasma membrane.

[0094] FIG. 11. Categorization of the interacting partners of DENV C protein. The function of each interacting protein is analysed and classified. Cell cycle and regulation of transcription and translation are the two major categories of DENV C protein-interacting partners.

[0095] FIG. 12 is a graph showing that there were 578 potential DENV C protein-binding partners detected to have z-score value above 1.

[0096] FIG. 13 is a graph showing the verification of the binding between potential interaction partners and Dengue virus capsid protein via ELISA.

[0097] FIG. 14 is a graph showing the epitope score of nine 10-residue fragments.

[0098] FIG. 15 is a graph showing that purified full-length recombinant DENV C protein is recognized by monoclonal antibodies against DENV C peptides.

[0099] Supplementary FIG. 1A. Cation exchange purification of DENV C protein using Resource MonoS column. Dialysed and refolded DENV C proteins after affinity chromatography are injected into Resource MonoS ion exchange chromatography column. Bound proteins are eluted out using increasing concentration of sodium chloride until final concentration of 1 M. However, there is no obvious peak during the elution steps except a small increase of UV absorbance when the concentration of NaCl reaches 40 mM (40%).

[0100] Supplementary FIG. 1B. Cation exchange purification of DENV C protein using Resource MonoS column. The presence of C protein is detected by ELISA using streptavidin-HRP antibody. An increasing amount of biotinylated DENV C protein is detected during elution and it reaches plateau at approximately 70% of NaCl (700 mM NaCl).

[0101] Supplementary Table 1. DNA (SEQ ID NO: 17) and protein (SEQ ID NO: 18) sequences of full-length DENV C protein

[0102] Supplementary Table 2. Rare codon analysis of full-length DENV C protein. Rare codons are in BOLD and underlined. The full-length DENV C DNA sequence is SEQ ID NO: 14.

[0103] Supplementary Table 3. Competent cells identified through BLAST analysis against BirA gene.

[0104] Supplementary Table 4. MALDI-TOF mass spectrometry analysis of purified DENV C protein, including FSLGMLQGR (SEQ ID NO: 15) and FLTIPPTAGILKR (SEQ ID NO: 16).

[0105] The present invention relates to a fused protein (in vivo site specific biotinylation of a viral (of the Flavivirus genus) protein), a method of producing the protein and application of the protein. The general steps for the production of fused proteinin an embodiment of the present invention, the fused protein may be a biotinylated Dengue virus Capsid protein or a West Nile virus Envelope protein DIIIare shown in FIG. 9.

EXAMPLE 1

Material and Methods

Construction of pET28aDENVBioCap Plasmid

[0106] DENV C gene was amplified from cDNA synthesized from DENV-2 (NGC strain; accession number: M29095; nucleotide 100-438) RNA using SuperScript III First-Strand Synthesis System (Life Technologies, USA). Primers Biotin_F (5-CTAGCTAGCTCCGGCCTGAACGAC-3) (SEQ ID NO: 1), Biotin_C_F (5-GACGACGACAAGAGCATGAATAACCAA-3) (SEQ ID NO: 2), Biotin_C_R (5-TTGGTTATTCATGCTCTTGTCGTCGTC-3) (SEQ ID NO: 3), and C_R (5-CCGCTCGAGTTACGCCATCACTGT-3) (SEQ ID NO: 4) were used to join biotin acceptor peptide gene (Cull and Schatz, 2000) containing an enterokinase cleavage site with DENV C gene via overlap extension PCR (OE-PCR). Gel-purified PCR products containing the joined fragments were subsequently inserted into expression vector, pET28a (Novagen, Germany) via NheI and XhoI cut sites. 6His tag and thrombin cleavage site are at the N-terminus of signal peptide followed by enterokinase cleavage site and DENV C protein. DNA sequences of the constructs were confirmed by sequencing.

Competent Cell Strain Screening

[0107] Transformed bacterial colonies of each strain [BL-21 (DE3) and BL-21-CodonPlus] (Agilent Technologies, USA) were picked and grown in 20 ml Luria-Bertani (LB) broth with 30 g/ml kanamycin antibiotic. When the bacteria absorbance OD.sub.600nm reached 0.65, protein expression was induced with 1 mM isopropyl -D-thiogalactoside (IPTG) overnight at 28 C. After IPTG induction, the bacteria absorbance OD.sub.600nm was measured again and 200 l of equal bacteria density for both strains were used for expression level screening. Bacterial cells were pelleted down with centrifugation at 8,000 rpm for 15 min at 4 C. and resuspended with 100 l 1protein sample buffer containing -mercaptoethanol. The samples were boiled for 10 min with constant vortexing at 1 min interval. Insoluble substances were pelleted down with centrifugation for 2 min at 20,000 g.

Protein Expression and Extraction

[0108] pET28aDENVBioCap plasmid was transformed into BL-21-CodonPlus expression competent cells (Agilent Technologies, USA) and grown on LB agar containing 30 g/ml kanamycin. Selected clones were cultured in 1 L LB broth (30 g/ml kanamycin) at 30 C. until absorbance OD.sub.600nm of 0.65. Expression of DENV C protein was induced with 1 mM IPTG overnight at 28 C. Bacterial cells were pelleted down with centrifugation at 8,000 rpm for 15 min at 4 C. Pellet was then resuspended in 10 ml resuspension buffer (20 mM Tris, 300 mM NaCl, 0.2% Triton-X, pH 8.0) and pelleted down again at 8,000 rpm for 15 min. The pellet was washed with wash buffer (20 mM Tris, 300 mM NaCl, pH 8.0) before it was resuspended in 30 ml lysis buffer (8 M urea, 20 mM Tris, 300 mM NaCl, 10 mM Imidazole, pH 8.0) containing EDTA-free protease inhibitor (Roche, Switzerland). The mixture was incubated at room temperature for 30 min and the lysate was subsequently clarified by centrifugation at 13,200 rpm for 20 min.

Nickel-Nitriloacetic Acid (Ni-NTA) Affinity Chromatography

[0109] Bacterial lysate containing denatured DENV C protein was incubated with nickel-nitrilotriacetic acid (Ni-NTA) resin (Bio-Rad, USA) for binding in a chromatography column overnight at 4 C. Ten column volume of wash buffer (8 M urea, 20 mM Tris, 300 mM NaCl, 20 mM Imidazole, pH 8.0) was used to wash away non-specific binding proteins. DENV C protein was eventually eluted out with elution buffer (8 M urea, 20 mM Tris, 300 mM NaCl, 500 mM Imidazole, pH 8.0) in ten fractions. Next, all eluates were combined for refolding and dialysis to remove 8 M urea. Briefly, all eluates were pooled into a SnakeSkin dialysis membrane tubing, 3.5 k MWCO, (Thermo Scientific, USA) and 0.5% of Tween-20 was added into the samples. The dialysis tubing was incubated in 1 L of 6 M urea for 6-12 hr at 4 C. and 250 ml of 25 mM Tris (pH 8.0) was added into the solution at every 6-12 hr interval. When the final volume reached 3 L, the dialysis tubing was transferred into 2 L of 20 mM Tris (pH 8.0) for 6 hr. Refolded DENV C proteins were collected from the dialysis tubing.

Ion Exchange Chromatography

[0110] Ion exchange chromatography column, Resource Q/Z 1 ml-packed size column (GE Healthcare, UK), was connected to fast protein liquid chromatography (FPLC) system. The column was equilibrated with ten column volumes of 20 mM Tris (pH 8.0) until the UV baseline and conductivity are stable. Refolded DENV C protein was injected into the column and the flow rate was set to 0.5 ml/min. After the sample passed through the column, ten column volumes of 20 mM Tris (pH 8.0) was used to wash away all the unbound proteins. Ionically-bound proteins were eluted out with increasing concentration of sodium chloride to a final concentration of 1 M (100%). All the fractions were analyzed via ELISA to determine the presence of biotinylated C protein.

Size Exclusion Chromatography

[0111] Superdex 75 10/300 GL chromatographic separation column (GE Healthcare, UK) was connected to FPLC system. The column was washed with five column volumes of MilliQ water and then equilibrated with two column volumes of 1 phosphate buffered saline (PBS) (pH 7.2). Samples were injected into the column and the flow rate was set to 0.25 ml/min. All the eluates were analyzed via ELISA to determine the presence of biotinylated C protein.

Enzyme Linked Immunosorbent Assay (ELISA)

[0112] To determine whether the expressed protein is biotinylated or not, 50 l of samples were added into the wells of MaxiSorp plate (eBioscience, USA) in triplicate for coating overnight at 4 C. After washing with 10.1% PBST, 150 l blocking buffer (4% bovine serum albumin) was added into each well and incubated for another hour at room temperature. Next, 150 l streptavidin-HRP enzyme conjugates (1:5000 dilution) was added and incubated for 1 hr. The plate was washed with 1PBST three times to remove unbound conjugates and then 100 l substrate solution, tetramethyl benzidine [(TMB) (Promega, USA)], was added for development. To stop the reaction, 50 l of 0.5 M H.sub.2SO.sub.4 solution was added. The absorbance was measured at 450 nm.

Product Analysis

[0113] Samples collected from flow through, wash, and eluates were analyzed by SDS-PAGE and Western blot. Twelve % Tris-tricine polyacrylamide denaturing gel was used to separate proteins in the samples and subsequently stained with Coomassie blue for detection. The presence of biotinylated DENV C protein was confirmed by Western blot via two different approaches. First, the identity of DENV C protein was determined with anti-His antibody. Briefly, separated proteins were transferred from polyacrylamide gel onto a PVDF membrane using iBlot Dry Blotting System (Life Technologies, USA). Blocking was done with 5% skimmed milk for 1 hr at room temperature. Next, the membrane was incubated with 0.1 g/ml mouse anti-His antibody (Qiagen, Germany) overnight at 4 C. The membrane was then washed with 1TBST and incubated with 0.1 g/ml goat anti-mouse secondary antibody conjugated with HRP (Thermo Scientific, USA) for 1 hr at room temperature. After washing with 1TBST, the membrane was developed using SuperSignal West Pico/Dura/Femto chemiluminescent substrate (Thermo Scientific, USA).

[0114] For the second approach, DENV C protein was detected directly using streptavidin conjugated with HRP. After transferring the samples onto a PVDF membrane, it was blocked with 4% bovine serum albumin (BSA) for 1 hr at room temperature. The membrane was then incubated with HRP-conjugated streptavidin (Millipore, USA) for another hour at room temperature. Subsequently, the membrane was washed thoroughly with 1PBST for 1 hr at room temperature and developed with chemiluminescent substrate.

Sample Preparation for Mass Spectrometry

[0115] Purified protein was electrophoresed through SDS-PAGE using 12% Tris-tricine polyacrylamide denaturing gel and stained with Coomassie blue. The background of Coomassie-stained gel was removed with destaining solution (40% methanol, 10% glacial acetic acid, 50% distilled H2O). DENV C protein-corresponding band was excised from the gel and kept in eppendorf tube containing distilled water. Samples were submitted to Dr. Lim Yong Pin (Department of Biochemistry, NUS) for matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry analysis.

Biotinylated Protein Binding Assay

[0116] The binding affinity of purified biotinylated DENV C protein was tested using streptavidin magnetic beads (GE Healthcare, UK) according to the manufacturer's protocol. Briefly, samples were mixed with the streptavidin magnetic beads and incubated for 1 hr with gentle mixing at 4 C. Unbound proteins were removed with wash buffer while biotinylated proteins were eluted out with elution buffer provided in the kit. Eluted proteins were analyzed by ELISA to confirm the biotinylation.

DENV C Protein Functional Assay

[0117] Pure Sec3 protein [(Abnova, Taiwan) (known interacting partner of flavivirus C protein (Bhuvanakantham, Li, Tan and Ng, 2010))] was added into the wells of MaxiSorp plate (eBioscience, USA) for coating overnight at 4 C. After washing with 1PBST, blocking buffer (4% bovine serum albumin) was added into each well and incubated for another hour. Purified DENV C protein was added into the well for binding at 37 C. for 1 hr. Next, streptavidin-HRP enzyme conjugates was added and incubated for another hour at 37 C. After washing three times with 1PBS, 100 l TMB substrate (Promega, USA) was added for development and 50 l 0.5 M H.sub.2SO.sub.4 solution was added to stop the reaction when necessary. The absorbance was measured at 450 nm.

Results

Engineering of Biotin Acceptor Peptide (BAP) into Full-Length DENV Capsid (C) Plasmid, pET28aDENVBioCap

[0118] To avoid non-specific biotinylation, we engineered a biotin acceptor peptide (BAP) at the N-terminus of DENV C protein. Biotin is a very small molecule (molecular weight =244.31) so it is unlikely to affect the structure and function of the protein. To ensure that the BAP will not result in any steric hindrance to the protein conformation, a secondary structure prediction was performed. As shown in FIG. 2A, the predicted secondary structures of DENV C protein with and without BAP do not differ from each other.

[0119] To clone the full-length DENV C protein, nucleotides 100 to 438 from DENV-2 genome (GeneBank accession number: M29095) was amplified (Fragment B in FIG. 2B), whereas the BAP sequence (Cull and Schatz, 2000) was synthesized together with an enterokinase cleavage site at the C-terminus (Fragment A in FIG. 2B). Both fragments were joined together through overlapping extension-PCR (OE-PCR) method as illustrated in FIG. 2B. The schematic representation of the final construct is shown in FIGS. 2C and 2D.

[0120] The final ligated product was then cloned into bacterial expression vector, pET28 which consisted of a 6His tag and a thrombin cleavage site in the upstream of the multiple cloning sites. Hence, the recombinant full-length DENV C construct (pET28aDENVBioCap) contained 2 tags (6His tag and BAP) at the N-terminus and 2 different enzyme cleavage sites (thrombin cleavage site and enterokinase cleavage site) for tag removal when necessary.

[0121] Five successfully transformed bacterial colonies were picked for colony PCR screening and DNA sequencing was performed to verify the constructs. The final DNA and protein sequence were shown in Suppl. Table 1. To further support our site-specific biotinylation strategy, WNV domain III (DIII) protein which has similar molecular weight as the C protein, was also engineered with a BAP concurrently via the same procedure. DENV C and WNV DIII proteins without BAP were also constructed for comparison purposes.

Optimal Expression Competent Bacterial Strain for DENV C Protein

[0122] To express DENV C protein, an optimal expression competent bacterial strain is requisite. We detected 14 rare codons in the full-length DENV C protein DNA sequence (Suppl. Table 2). This raised a concern for protein expression because 12% of the total 115 codons are rare codons. To screen for the optimal bacterial expression competent cell, pET28aDENVBioCap plasmid was transformed into BL-21 (DE3) and BL-21-CodonPlus. BL-21 (DE3) is a common bacterial strain for high expression of recombinant protein while BL-21-CodonPlus is a bacterial strain specifically engineered for expression of protein with rare codons.

[0123] The expression of DENV C protein was indeed much higher in BL-21-CodonPlus strain. As indicated in FIG. 3, an obvious band corresponding to the recombinant full-length DENV C protein was detected in the lysate of IPTG-induced CodonPlus strain but not in the BL-21 (DE3) strain (FIG. 3Ai). Western blotting with anti-His antibody revealed that DENV C protein was expressed in both BL-21 (DE3) and BL-21-CodonPlus strains because C protein-corresponding bands were detected in the lysates of both strains (FIG. 3Aii). Nonetheless, the observed bands were much thicker in BL-21-CodonPlus strain as compared to BL-21 (DE3) strain. This demonstrated that DENV C protein expression level was much higher in BL-21-CodonPlus strain under the same condition. As for WNV DIII protein, the expression was good enough in normal BL-21 (DE3) strain (data not shown).

Discovery of Endogenous Biotinylation in BL-1 Strains

[0124] As comparing DENV C protein with and without BAP via Western-blot using anti-His antibody, bands were observed in all the lanes for DENV C and WNV DIII proteins (FIG. 3Bi). As shown in FIG. 3Bi, recombinant proteins with BAP have higher molecular weight than those without BAP. This also indicated that all the DENV C and WNV DIII proteins with and without BAP were expressing well in the BL-21 strains.

[0125] Besides detecting the proteins of interest using anti-His antibody, streptavidin-horseradish peroxidase (HRP) was also employed to detect the presence of biotinylated protein. Surprisingly, we found that our proteins of interest were biotinylated even before any in vitro biotinylation process. As shown in FIG. 3Bii, thick bands can be observed in the BAP-containing DENV C and WNV DIII proteins when streptavidin-HRP antibody was used. Similar bands were not detected in DENV C and WNV DIII proteins without BAP. Commercially-available biotinylated (BN) and unbiotinylated (UBN) maltose-binding protein [(MBP) (GeneCopoeia, USA)] were used as positive and negative controls for biotinylation, respectively. Similarly, band was only seen in the biotinylated MBP Lane. This result suggested that DENV C and WNV DIII proteins with BAP were biotinylated endogenously in BL-21 strains.

[0126] We postulated that bacterial BL-21 strains may contain biotin holoenzyme synthetase BirA gene that encodes for biotin ligase protein in their genome. To unveil the mystery as to why biotinylation could occur endogenously, BirA gene sequence was analyzed using BLAST software and the blasting result showed that BL-21 strains indeed possess BirA gene in their genome (Supplementary Table 3). It was also reported that biotin molecules were present in the Luria-Bertani (LB) broth (Tolaymat and Mock, 1989). As a result, proteins engineered with BAP could be directly expressed and biotinylated in BL-21 strains without the extra in vitro enzymatic biotinylation step.

Optimized Sequential Purification Protocol for Full-Length C Proteins

[0127] After confirming the expression and biotinylation of the protein of interest, large scale production of bacterial culture was carried out. Initial attempt was to perform all the extraction and purification steps in native form so that the protein structure could be preserved without the need of refolding. However, the extraction of DENV C protein from the bacterial cells was not effective in native condition. There were still considerably huge amount of DENV C protein trapped in the pellet (data not shown). Therefore, 8 M urea was used to lyze the bacterial cells and affinity his-tag chromatography purification was first performed under denaturing condition.

[0128] The bacterial cell lysate with biotinylated DENV C protein was incubated with nickel-nitrilotriacetic acid (Ni-NTA) resin to purify His-tagged proteins. Recombinant full-length DENV C protein contained 6His tag at the N-terminus bound to the resin and unbound proteins were removed during washing with 20 mM imidazole. To ensure most of the unbound proteins were washed away, ten column volumes of wash buffer were used. The bound proteins were eluted out with 500 mM imidazole. However, many non-specific bands were still observed in the eluates of DENV C protein (FIG. 4A), especially in the second eluate fraction. Nonetheless, Western-blot result confirmed that most of the DENV C proteins were eluted out starting from fraction E2 until E10 (FIGS. 4B and 4C).

[0129] Partially purified DENV C proteins from eluates E2 to E10 were pooled together for step-wise dialysis and concurrent refolding. Due to the hydrophobic C-terminus of DENV C protein, the protein easily aggregated resulting in major loss during refolding and dialysis. To prevent protein aggregation and to enhance efficient refolding, 0.05% of Tween-20 was added to the samples (Krupakar et al., 2012). With this addition of detergent in the samples, protein aggregation was reduced significantly. No obvious precipitation was observed and the solution was clear after dialysis and refolding. This partially purified DENV C protein was then subjected to second round of purification using ion exchange chromatography-fast protein liquid chromatography system (IEX-FPLC).

[0130] Theoretically, DENV C protein is a positively-charged protein so cation-exchanger column, Resource MonoS, should be used. However, we found that biotinylated DENV C protein was not eluted out at a specific concentration of sodium chloride (NaCl). Miniature amount of the protein was eluted out slowly in an increasing gradient manner and it reached plateau at 70 mM of NaCl (Suppl. FIG. 1). This indicated that Resource MonoS column is not suitable for separating DENV C protein from the contaminants.

[0131] Counter intuitively, we were able to obtain better separation when we used anion-exchanger, Resource MonoQ, column (FIG. 5Ai). This could be due to the presence of BAP in the DENV C protein conferring its ability to bind to positively charged beads. As shown in FIG. 5A, biotinylated full-length DENV C protein bound perfectly to Resource MonoQ column as no biotinylated proteins were detected by ELISA in the flow through although there was high UV absorbance detected in those fractions. During elution, one high peak was detected when the NaCl concentration reached approximately 350 mM (35%) while there was another small peak detected at the elution concentration of 100 mM of NaCl (10%). These peaks were confirmed to be biotinylated proteins as detected by ELISA using streptavidin-HRP antibody. This result showed that the second purification step managed to further separate most of the contaminants from DENV C protein.

[0132] To avoid any other contaminants that may have similar charge as biotinylated DENV C protein, eluates corresponding to the high UV absorbance peak after IEX-FPLC were injected into size-exclusion chromatography (SEC), Superdex 75 10/300 GL column. It was previously reported that SEC could not be used to estimate the molecular weight of DENV C protein because there were many unspecific interaction between the gel matrix and C protein (Jones, Ma, Burgner, Groesch, Post and Kuhn, 2003). Nevertheless, it is still a good method to further isolate the proteins of interest from other possible residual contaminants based on the molecular weight differences. After size-exclusion chromatography, three peaks were observed in the chromatogram (FIG. 5Bi). The first peak was identified to be the biotinylated DENV C protein as the fractions detected with high absorbance in ELISA using streptavidin-HRP antibody coincided with the first peak in the elution profile (FIG. 5Bii). The identity of this highly purified protein was further validated with matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry analysis (Suppl. 351 Table 4).

[0133] With this optimized sequential purification protocol, we were able to obtain 0.8-1.0 mg of highly purified biotinylated non-truncated, full-length DENV C protein from 1 L of IPTG-induced bacterial culture repeatedly.

High In Vivo Biotinylation Efficiency in BL-21 Strains

[0134] To compare the biotinylation efficiency of our strategy, the same amount of purified biotinylated DENV C, WNV DIII and MBP proteins were coated on the ELISA plate and streptavidin-HRP antibody was used for detection. Assuming that the coating efficiencies are the same among all three proteins, the absorbance detected via direct ELISA should be proportional to the number of coated proteins and inversely proportional to the power of of its molecular size. The larger the protein, the lesser the proteins can be coated on the plate.

[00001]$\mathrm{Absorbance}\frac{\mathrm{Protein}.Math..Math.\mathrm{Concentration}}{{\left(\mathrm{Protein}.Math..Math.\mathrm{Size}\right)}^{2/3}}$

[0135] The molecular size of biotinylated MBP is about two times larger than biotinylated DENV C and WNV DIII proteins. Assuming that one protein only carries one biotin which will bind to one streptavidin-HRP only, the absorbance of MBP should be two times lower than our purified proteins for the same protein concentration. As shown in FIG. 6i, the absorbance of purified biotinylated DENV C and WNV DIII proteins indeed showed approximately two times higher the absorbance as compared to commercial biotinylated MBP. The same phenomenon was also observed when the biotinylated proteins were streptavidin-captured by streptavidin-magnetic beads and detected via direct ELISA (FIG. 6Aii). The absorbance of eluted biotinylated DENV C protein and WNV DIII protein were two times higher than that of eluted biotinylated MBP. No significant absorbance was detected for unbiotinylated C and DIII proteins.

[0136] This result further supported that engineering an additional BAP on a protein could result in site-specific endogenous biotinylation with high efficiency in Escherichia coli BL-21 strains, without the need of an extra in vitro enzymatic or chemical conjugation step. Our strategy could produce biotinylated proteins that were as good quality as the manufactured biotinylated MBP.

Purified Non-Truncated, Full-Length DENV C Protein is Functional

[0137] It is essential to ensure that the purified proteins are functional before they are used for any further studies. According to our recent findings, WNV and DENV C proteins were found to interact with human Sec3 exocyst protein (Bhuvanakantham, Li, Tan and Ng, 2010). Thus, we examined the functionality of the purified DENV C protein by revisiting the interaction between biotinylated DENV C protein and Sec3 protein using ELISA. Pure Sec3 protein was coated overnight on ELISA plate and purified biotinylated DENV C protein was used as the probe. Bound DENV C proteins were detected with streptavidin-HRP antibody. As shown in FIG. 6B, statistically significant absorbance was detected in Sec3 protein-coated wells but not in BSA-coated wells. Thus, our purified DENV C protein did interact with Sec3 protein. This corroborated that our purified biotinylated full-length DENV C protein is functional and can be used for structural and molecular studies.

High-Throughput Screening of the Interacting Partners of Dengue Virus Capsid Protein

[0138] Dengue virus (DENV) is a positive-stranded RNA virus. Based on the current understanding of flaviviral replication cycle, host cell nucleus is not involved during the transcription, translation and assembly of virus. This traditional notion is challenged when capsid (C) protein is found to localize in the infected cell nucleus. Since then, the role of DENV C protein in the nucleus has been an intriguing mystery for researchers to unveil. Understanding the roles of C protein in the nucleus will not only completes the missing puzzle of flaviviral replication cycle, but also provides insights for novel anti-viral strategies development.

[0139] Using the purified biotinylated full-length DENV C protein of the present invention, high-throughput screening (HTS) of novel interacting partners of DENV C protein is made possible. ProtoArray Human Protein Microarray PPI Complete Kit (Life Technologies, USA) was employed for HTS study of novel DENV C protein-interacting partners. This protein microarray contains 9400 purified human proteins printed in duplicate on the nitrocellulose-coated glass slide. The purified human proteins encompass various cellular kinases, enzymes, nuclear proteins and signalling proteins tagged with glutathione S transferase (GST) or 6His. Purified biotinylated DENV C protein was used as probe and streptavidin-Alexa Fluor 647 antibody is chosen as the detection reagent as it yields very good signal to noise ratio.

[0140] There were 578 potential DENV C protein-binding partners detected to have z-score value above 1 (FIG. 12).

[0141] Z-score value of 3 was chosen to be the cut-off point for statistically significant potential interactors. Z-score indicates how far and in what direction the sample's value deviates from the distribution's mean. After filtering some of the proteins without known functions and with biotin-binding property, 31 significant potential interacting partners were identified for DENV C protein as shown in the table below.

Significant Potential Interacting Partners of DENV C Protein

[0142] Scanned images of the microarray are analyzed using GenePix Pro software to determine the pixel intensity and the data acquired are further analyzed via Protoarray Prospector software to identify potential interacting partners of DENV C protein. Z-score cut-off value of 3 is chosen to be the statistically significant interactors.

TABLE-US-00001 Signal Identity Symbol Database ID Intensity CV Z-score Function Cortactin CTTN NM_138565.1 3076.5 0.03517 13.41837 Regulate and organize actin cytoskeleton SLAIN motif family, SLAIN2 BC031691.2 2569.5 0.04981 11.08491 Control microtubule growth and member 2 organization Immunoglobulin IGBP1 NM_001551.1 1938 0.02043 8.17845 Binds to surface IgM receptor (CD79A) binding and may involve in the protein 1 activation signal transduction pathway Choline kinase CHKA NM_001277.1 1738 0.03743 7.25795 Catalyzes the phosphorylation of alpha ethanolamine and plays a role in phosphatidylcholine biosynthesis. DIM1 DIMT1 NM_014473.2 1666.5 0.26943 6.92887 Dimethylates 18 S rRNA in the dimethyladenosine 40 S particle transferase 1 homolog (S. cerevisiae) Ubiquitin- UBE2S BC004236.2 1353.5 0.00366 5.4883 Catalyzes Lys-11-linked conjugating polyubiquitination on the enzyme E2S anaphase promoting complex/ cyclosome (APC/C) substrates AF4/FMR2 family, AFF4 BC025700.1 1251 0.06896 5.01654 Regulates transcriptional activity member 4 WAS/WASL WIPF1 BC002914.1 1159 0.02562 4.59312 Control the organization of actin interacting protein cytoskeleton family, member 1 Aurora kinase A AURKA BC006423.1 1158 0.04519 4.58851 Cell cycle-regulated kinase regulates microtubules formation and stabilizes spindle poles during mitosis RAD51 associated RAD51AP1 BC016330.1 1143 0.00990 4.51948 Involved in common DNA protein 1 damage response pathway Additional sex ASXL1 BC064984.1 1072 0.04485 4.1927 Regulates transcriptional activity combs like 1 (Drosophila) CDKN2A interacting CDKN2AIP/ BC022270.1 1031.5 0.13642 4.0063 Activates p 53/TP53 via protein CARF CDKN2A-dependent and CDKN2A-independent pathways Calcium channel, CACNB1 NM_000723.3 1017.5 0.12162 3.94187 Increases peak calcium current, voltage-dependent, modulates G protein inhibition, beta 1 subunit and shifts the voltage- dependent activation and inactivation. Spermidine/ SAT2 NM_133491.2 1008.5 0.00351 3.90044 Catalyzes the acetylation of spermine Ni- polyamines acetyltransferase family member 2 Eukaryotic EIF1AX NM _001412.2 997 0.15461 3.84751 Enchances ribosome translation initiation dissociation into subunits and factor 1 A, X-linked stabilizes the binding of Met- tRNA(I) to 40 S ribosomal subunits Signal recognition SRP19 BC010947.1/ 971.5/ 0.34573/ 3.73015/ Mediates the assembly of signal particle 19kDa NM_003135.1 818 0.16770 3.02367 recognition particle Calcium/ CAMK2A NM_171825.2 969 0.11238 3.71865 Involved in hippocampal long- calmodulin- term potentiation by switching dependent protein calmodulin-dependent activity kinase II alpha to calmodulin independent Piccolo (presynaptic PCLO BC001304.1 929 0.02740 3.53455 Involved in the organization of cytomatrix protein) synaptic active zone and synaptic vesicle trafficking Nuclear speckle NSRP1 NM_032141.1 927.5 0.16544 3.52764 Mediates pre-mRNA alternative splicing regulatory splicing regulation protein 1 NIMA (never in NEK7 NM_133494.1 918.5 0.19169 3.48622 Controls initiation of mitosis mitosis gene a)- related kinase 7 Checkpoint CHES1/ NM_005197.2 918 0.27730 3.48392 Transcriptional repressor suppressor 1/ FOXN3 involved in DNA damage- Forkhead box N3 inducible cell cycle arrests at G1 and G2 phases Rtf1, Paf1/RNA RTF1 NM_015138.2 882.5 0.02484 3.32053 Regulates transcriptional activity polymerase II complex component, homolog (S. cerevisiae) Microtubule- MAP2 NM_031845.1 875 0.03879 3.28601 Function unclear. Thought to be associated protein involved in microtubule 2 assembly and stabilizing the microtubules against depolymerization Potassium channel KCTD18 NM_152387.2 856 0.02313 3.19857 Molecular determinants for tetramerisation subfamily-specific assembly of domain containing alpha-subunits into functional 18 tetrameric voltage-gated potassium channels Cyclin B3 CCNB3 NM_033671.1 853.5 0.05385 3.18706 Involved in cell cycle control Eukaryotic EIF1AY BC005248.1 852.5 0.12691 3.18246 Enchances ribosome translation initiation dissociation into subunits and factor 1 A, Y-linked stabilizes the binding of Met- tRNA(I) to 40 S ribosomal subunits Vaccinia related VRK1 NM_003384.1 845 0.17406 3.14794 Serine/Threonine kinase that is kinase 1 involved in regulating cell proliferation and prevent the interaction between p 53/ TP53 and MDM2 Protein kinase C, PRKCI NM_002740.1 841 0.00504 3.12953 Calcium-independent and iota phospholipid-dependent serine/threonine kinase involved various cellular processes WD repeat domain 5 WDR5 NM_017588.1 840.5 0.01598 3.12723 Involved in histone modification and acetylation of histone H 3 p 53-regulated DDA3 PSRC1 NM_032636.2 822 0.02925 3.04208 Regulates mitotic spindle and (DDA3)/proline/ involved in p 53/TP53- serine-rich regulated growth suppression coiled-coil 1 Elongation factor 1 ELOF1 NM_032377.2 816.5 0.00260 3.01677 Maintains proper chromatin homolog (ELF1, structure in actively S. cerevisiae) transcribed regions

[0143] The list of potential interacting partners for DENV C protein was quite well dispersed throughout the cell nucleus and cytoplasm. The total proteins in each cellular compartment are calculated and charted. FIG. 10 illustrates the cellular localization of each individual protein and some of them may localize in more than one cellular compartment. Majority of the DENV C protein-interacting partners localize in the nucleus followed by cytoplasm. Six proteins are related to cytoskeleton while five are found in the nucleolus and plasma membrane. As shown in FIG. 10, approximately 55% of the identified proteins localized in the nucleus while the other half was distributed in the cytoplasm. This was exciting because a comprehensive list of DENV C protein-nuclear interactors involved in the virus replication was identified from this ProtoArray HTS study.

[0144] From the list of potential interacting partners, it appeared that DENV C protein may be involved primarily in cell cycle control and regulation of transcriptional and translational activities (FIG. 11). Five proteins (AFF4, ASXL1, RTF1, WDRS, and ELOF1) were involved in the regulation of transcription and four proteins (DIMT1, EIF1AX, EIF1AY, and SRP19) participated in ribosomal biogenesis and protein synthesis. This is a thrilling result because it unravelled potential roles of DENV C protein in the nucleus which is a million-dollar question in flavivirus research. Investigating these novel non-structural roles of flavivirus C protein will provide better understanding of the underlying molecular mechanism of flavivirus infection. This knowledge is essential for more refinement of anti-viral drug designs and development.

Binding Assay of the Interacting Partners of Dengue Virus Capsid Protein via ELISA Platform

[0145] The purified biotinylated full-length Dengue virus capsid protein of the present invention can also be used to verify the binding between DENV C protein and its interacting partners in ELISA platform. To validate the identified potential interacting partners from the protein microarray study, commercially-available pure proteins were purchased and coated on ELISA plate. Biotinylated full-length DENV C protein was used as probe for binding and streptavidin-HRP antibody was used to detect the bound proteins. Four proteins (DIMT1, EIF1AX, EIF1AY, and SRP19) that are related to transcriptional and translational machinery and five proteins pertaining to cell cycle control (UBE2S, CARF, CHES1, CCNB3, and VRK1) were selected for further validation of DENV C protein binding. As shown in FIG. 13, all nine proteins showed significant higher absorbance as compared to bovine albumin serum (BSA) control, implicating that these nine potential interacting partners bound to full-length DENV C protein in an ELISA-based platform.

[0146] FIG. 13 shows the verification of the binding between potential interaction partners and Dengue virus capsid protein via ELISA. Commercially-available proteins of the identified interactors of DENV C protein are coated on MaxiSorp 96-well plate in triplicates. Purified biotinylated DENV C proteins are added into the wells for binding and bound proteins are detected via streptavidin-horseradish peroxidase (HRP) secondary antibody. The absorbance is measured at 450 nm. Bovine serum albumin (BSA) is used as negative control for binding.

[0147] Therefore, the biotinylated full-length DENV C protein of the present invention is useful to flavivirus laboratories that are studying protein-protein, protein-RNA, and protein-DNA interactions.

Development of Antibody against Dengue Virus Capsid Protein

[0148] One of the challenges in Dengue virus research is that antibodies against the viral proteins are not available commercially. Most of the laboratories need to raise the antibodies on their own before carrying out any further downstream experiments (Puttikhunt et al., 2009, Wang et al., 2002). Anti-DENV C antibody is one of the missing antibodies in the market. Hence both full-length DENV C protein and anti-DENV C antibody are of commercial values.

[0149] Before the present inventors raised anti-C antibody using pure DENV C protein, an antibody generation package from Abmart Inc. was purchased to produce monoclonal antibodies against DENV C peptides. Abmart Inc. was only able to produce monoclonal antibodies against two peptides (MNNQRKKARN (SEQ ID NO: 5) and ERNRVSTVQQ (SEQ ID NO: 7)). Below is a list of predicted epitopes for antibody production:

TABLE-US-00002 EpitopesSequences 1 MNNQRKKARN 4 ARNTPFNMLK 7 PPTAGILKRW (SEQIDNO:5) (SEQIDNO:8) (SEQIDNO:11) 2 KRWGTIKKSK 5 KSKAINVLRG 8 RVSTVQQLTK (SEQIDNO:6) (SEQIDNO:9) (SEQIDNO:12) 3 ERNRVSTVQQ 6 RFSLGMLQGR 9 RGFRKEIGRM (SEQIDNO:7) (SEQIDNO:10) (SEQIDNO:13)

[0150] The epitopes are predicted by Abmart Inc. using their propriety algorithm. An epitope score is assigned to each residue based on criteria such as structural features, sequence conservation, hydrophobicity, and solvent exposure. Nine 10-residue fragments with highest overall scores are chosen as epitopes to generate monoclonal antibodies.

[0151] The first peptide (MNNQRKKARN) (SEQ ID NO: 5) is the first ten residues from the N-terminus of C protein whereas the second peptide (ERNRVSTVQQ) (SEQ ID NO: 7) is from the 19.sup.th residue to the 28.sup.th residue. Both epitopes are from the N-terminal region before the first helix (1) structure. Monoclonal antibodies against these two peptides were tested to be able to recognize purified full-length recombinant DENV C protein as shown in FIG. 10. Further characterization of the antibodies is still on-going.

[0152] FIG. 15 shows that purified full-length recombinant DENV C protein is recognized by monoclonal antibodies against DENV C peptides. C348 and C193 are monoclonal antibodies against MNNQRKKARN (SEQ ID NO: 5) and ERNRVSTVQQ (SEQ ID NO: 7) peptides, respectively. Purified DENV C protein (100 ng) is coated on the MaxiSorp well in triplicates. Antibodies C348 and C193 are diluted 10,000 and bound antibodies are detected by horseradish peroxidase (HRP)-conjugated anti-mouse secondary antibody. Both monoclonal antibodies can recognize recombinant DENV C protein but not bovine serum albumin (BSA).

Discussion

[0153] To produce biotinylated proteins, we opted for site-specific biotinylation by engineering a biotin acceptor signal peptide (BAP) in the upstream of our proteins. To attach a biotin molecule onto BAP either in vivo or in vitro, bacterial biotin ligase BirA enzyme is required (Chapman-Smith and Cronan, 1999, Cull and Schatz, 2000, Tan et al., 2004, Yang et al., 2004). Instead of performing this biotinylation step in vitro, we found that biotinylation of the BAP-containing proteins occurred endogenously in Escherichia coli BL-21 strains with high efficiency.

[0154] The purification of non-truncated, full-length DENV C protein was indeed challenging. Unlike DIII protein (Tan et al., 2010), one single purification step was simply not adequate to produce pure DENV C protein. The presence of rare codons in the C protein sequence necessitates unique expression competent strain for its optimal protein expression. Due to the hydrophobic C-terminus, aggregation easily occurred and most of the expressed proteins were trapped in the inclusion bodies. As such, we had to use 8 M urea to denature the proteins and perform affinity chromatography under denaturing condition. We also managed to solve the aggregation problem during dialysis and refolding by using non-ionic detergent. To obtain highly purified DENV C protein, two more purification steps were included in our protocol, namely IEX and SEC. This sequential purification strategy can produce approximately 1 mg of purified DENV C protein from 1 L of bacterial culture.

[0155] This purified non-truncated, full-length DENV C protein is useful for various molecular and structural studies. For instance, it can be used to study the interaction of NS2B-NS3 serine proteinase and C protein. In order to produce mature C protein for encapsidation and assembly, full-length C protein has to be cleaved by NS2B-NS3 protein complex. Thus, it is crucial to identify the binding sides of NS2B-NS3 complex to full-length C protein, which is still unknown. It will in turn lead to the development of novel anti-viral drugs targeting the assembly of virus particles. Besides, it is also interesting to examine whether the cleaved C-terminus plays any unidentified role in the pathogenesis. The crystal structure of DENV NS2B-NS3 protein complex has been resolved (Erbel et al., 2006). However, the crystal structure of full-length DENV C protein is still not available yet, although the NMR structure of partial DENV C protein was obtained (Ma, Jones, Groesch, Kuhn and Post, 2004). We are now taking up the challenge to resolve the 3D structure of full-length DENV C protein using our purified protein.

[0156] Ma and co-workers (2004) proposed a model for the interaction between C protein and its viral genome using the partial DENV C protein. However, this model is still not validated and the exact mechanism of encapsidation is still unclear until to date. Investigation of how DENV C protein interacts with its RNA and oligomerizes into nucleocapsid during maturation is vital for better understanding of the pathogenesis. Recent publications also demonstrated that N-terminus of DENV C protein was antigenic in mice study and the first 18 amino acid residues were involved in the virus assembly during maturation (Puttikhunt, Ong-Ajchaowlerd, Prommool, Sangiambut, Netsawang, Limjindaporn, Malasit and Kasinrerk, 2009, Samsa, Mondotte, Caramelo and Gamarnik, 2012). As a result, our purified non-truncated DENV C protein is absolutely desirable for delineating the whole process of encapsidation and virus assembly.

[0157] Another difficulty of studying DENV C protein is that anti-C antibody is yet to be available commercially. Most of the laboratories need to raise this antibody on their own before any further downstream experiments (Puttikhunt, Ong-Ajchaowlerd, Prommool, Sangiambut, Netsawang, Limjindaporn, Malasit and Kasinrerk, 2009, Wang et al., 2002). Hence, pure full-length DENV C protein can be obtained simply by enterokinase cleavage and subsequently used to raise antibody in mice or rabbit for laboratory research usage and diagnostic tool. Both full-length DENV C protein and anti-DENV C antibody will be of commercial values.

[0158] In conclusion, we have developed an optimized protocol to engineer, express and purify the first biotinylated, full-length DENV C protein. This purified protein is useful for various molecular studies to further understand the underlying mechanism of C protein in pathogenesis. Only by stitching all these missing pieces of the puzzle together, novel antiviral strategies can be rationally designed.

[0159] Whilst there has been described in the foregoing description preferred embodiments of the present invention, it will be understood by those skilled in the technology concerned that many variations or modifications in details of design or construction may be made without departing from the present invention.

[0160] All references, including any patents or patent applications, cited in this specification are hereby incorporated by reference. No admission is made that any reference constitutes prior art. The discussion of the references states what their authors assert, and the applicants reserve the right to challenge the accuracy and pertinency of the cited documents. It will be clearly understood that, although a number of prior art publications are referred to herein, this reference does not constitute an admission that any of these documents forms part of the common general knowledge in the art.

[0161] For the purposes of this specification, it will be clearly understood that the word comprising means including but not limited to, and that the word comprises has a corresponding meaning.

REFERENCES

[0162] 1. Amberg S. M. and Rice C. M. (1999) J Virol, 73, 8083-8094.

[0163] 2. Bayer E. A. and Wilchek M. (1990) Methods Enzymol, 184, 174-187.

[0164] 3. Bayer E. A. and Wilchek M. (1990) Methods Enzymol, 184, 138-160.

[0165] 4. Bhuvanakantham R., Chong M. K. and Ng M. L. (2009) Biochem Bioph Res Co, 389, 63-69. Bhuvanakantham R., Li J., Tan T. T. T. and Ng M. L. (2010) Cellular Microbiology, 12, 453-472.

[0166] 5. Chapman-Smith A. and Cronan J. E., Jr. (1999) Trends Biochem Sci, 24, 359-363.

[0167] 6. Chapman-Smith A. and Cronan J. E., Jr. (1999) Biomol Eng, 16, 119-125.

[0168] 7. Cull M. G. and Schatz P. J. (2000) Methods Enzymol, 326, 430-440.

[0169] 8. de Boer E., Rodriguez P., Bonte E., Krijgsveld J., Katsantoni E., Heck A., Grosveld F. And Strouboulis J. (2003) Proc Natl Acad Sci USA, 100, 7480-7485.

[0170] 9. Erbel P., Schiering N., D'Arcy A., Renatus M., Kroemer M., Lim S. P., Yin Z., Keller T. H., Vasudevan S. G. and Hommel U. (2006) Nat Struct Mol Biol, 13, 372-373.

[0171] 10. Gubler D. J. (2002) Arch Med Res, 33, 330-342.

[0172] 11. Hase T., Summers P. L., Eckels K. H. and Baze W. B. (1987) Archives of Virology, 92, 273-291.

[0173] 12. He Y. F., Bao H. M., Xiao X. F., Zuo S., Du R. Y., Tang S. W., Yang P. Y. and Chen X. (2009) Proteomics, 9, 5414-5424.

[0174] 13. Howarth M. and Ting A. Y. (2008) Nat Protoc, 3, 534-545.

[0175] 14. Jones C. T., Ma L. X., Burgner J. W., Groesch T. D., Post C. B. and Kuhn R. J. (2003) J Virol, 77, 7143-7149.

[0176] 15. Krupakar P., Ngo A. M. L. and Ng M. L. (2012) In Miguel Benitez and Aguirre V. (eds) Protein Purification. Nova Science, New York, pp. 147-170.

[0177] 16. Ma L., Jones C. T., Groesch T. D., Kuhn R. J. and Post C. B. (2004) Proc Natl Acad Sci USA, 101, 3414-3419.

[0178] 17. Markham K., Bai Y. and Schmitt-Ulms G. (2007) Anal Bioanal Chem, 389, 461-473.

[0179] 18. Markoff L. (1989) J Virol, 63, 3345-3352.

[0180] 19. Moreland N. J., Tay M. Y., Lim E., Paradkar P. N., Doan D. N., Yau Y. H., Geifman Shochat S. and Vasudevan S. G. (2010) PLoS Negl Trop Dis, 4, e881.

[0181] 20. Postel A., Letzel T., Muller F., Ehricht R., Pourquier P., Dauber M., Grund C., Beer M. and Harder T. C. (2011) Anal Biochem, 411, 22-31.

[0182] 21. Puttikhunt C., Ong-Ajchaowlerd P., Prommool T., Sangiambut S., Netsawang J.,

[0183] 22. Limjindaporn T., Malasit P. and Kasinrerk W. (2009) Arch Virol, 154, 1211-1221

[0184] 23. Qi Y. and Katagiri F. (2011) Methods Mol Biol, 712, 21-30.

[0185] 24. Samsa M. M., Mondotte J. A., Caramelo J. J. and Gamarnik A. V. (2012) J Virol, 86, 1046-1058.

[0186] 25. Sueda S., Yoneda S. and Hayashi H. (2011) Chembiochem, 12, 1367-1375.

[0187] 26. Tadano M., Makino Y., Fukunaga T., Okuno Y. and Fukai K. (1989) J Gen Virol, 70 (Pt 6), 1409-1415.

[0188] 27. Tan L. C., Chua A. J., Goh L. S., Pua S. M., Cheong Y. K. and Ng M. L. (2010) Protein Expr Purif, 74, 129-137.

[0189] 28. Tan L. P., Lue R. Y., Chen G. Y. and Yao S. Q. (2004) Bioorg Med Chem Left, 14, 6067-6070.

[0190] 29. Tolaymat N. and Mock D. M. (1989) J Nutr, 119, 1357-1360.

[0191] 30. Wang S. H., Syu W. J., Huang K. J., Lei H. Y., Yao C. W., King C. C. and Hu S. T. (2002) J Gen Virol, 83, 3093-3102.

[0192] 31. Yang J., Jaramillo A., Shi R., Kwok W. W. and Mohanakumar T. (2004) Hum Immunol, 65, 692-699.

[0193] 32. Yang M. R., Lee S. R., Oh W., Lee E. W., Yeh J. Y., Nah J. J., Joo Y. S., Shin J., Lee H. W., Pyo S. et al. (2008) Cell Microbiol, 10, 165-176.

[0194] 33. Zhang W., Chipman P. R., Corver J., Johnson P. R., Zhang Y., Mukhopadhyay S., Baker T. S., Strauss J. H., Rossmann M. G. and Kuhn R. J. (2003) Nat Struct Biol, 10, 907-912.