-
West Nile virus (WNV) belongs to the Flavivirus genus, which is composed of a single-strand, positive-sense RNA approximately 11, 000 nucleotides in length (Mukhopadhyay S, 2005) that encodes three structural proteins, capsid (C), membrane (prM/M), and envelope (E), and seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5). The virus protease is encoded in the N-terminal portion of NS3 (Fig. 1A) and is responsible for processing the polyprotein precursor into individual functional proteins in the presence of the host proteases furin and secretase (Stadler K, 1997; Thomas G, 2002). NS3 protease cleaves the junctions at the NS2A/NS2B, NS2B/NS3, NS3/NS4A, and NS4B/NS5sites, as well as internal sites within C, NS3, and NS4A (Chappell K J, 2005; Nall T A, 2004). These cleavages are necessary for viral replication and assembly; therefore, the NS3 protease represents a potential therapeutic target (D'Arcy A, 2006; Jia F, 2010; Loughlin W A, 2004; Nall T A, 2004; Tyndall J D, 2005).
Figure 1. Schematic of WNV polyprotein. A: The WNV genome is flanked by 5' and 3' UTRs, which encode three structural proteins (C, prM, E) and seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5). B: The protein sequences of wild-type (WNVNS2BCF44NS3) and the mutants used in this study. (E: envelope, M: membrane, UTR: untranslated region, H: hydrophilic)
NS3 has a classical catalytic triad (H51, D75, S135), and the mutation of any one of these residues will result in the loss of protease activity (Ryan M D, 1998). Furthermore, the role of NS2B as a cofactor for NS3 protease activation has been investigated in many flaviviruses, including WNV, dengue virus (DENV), yellow fever virus (YFV), and Japanese encephalitis virus (JEV) (Chambers T J, 1993; Droll D A, 2000; Falgout B, 1991; Falgout B, 1993; Lin C W, 2007; Nall T A, 2004). The results indicate that the hydrophilic central region of NS2B 48 amino acid (aa) is essential (Leung D, 2001), and some residues are important for promoting the functional configuration of NS2B-NS3 protease and maintaining critical interactions with protease catalysis substrates (Chappell K J, 2008; Ciota A T, 2007; Erbel P, 2006; Mueller N H, 2007; Radichev I, 2008).
The L79DDDGNF85 sequence in WNV NS2B is required for the productive interaction of NS2B with NS3 and for appropriate NS2B-NS3 protease viral enzyme catalytic activity (Johnston PA, 2007). The sequence alignment of NS2B in WNV, JEV, Kunjin virus (KUNV), Murray Valley encephalitis virus (MVEV), St. Louis encephalitis virus (SLEV), Zika virus (ZKV), DENV (serotypes 1, 2, 3, and 4), and YFV reveals that the G83 residue is conserved, but D80DD was not highly conserved (Fig. 2A). The D80DD region of NS2B forms a negatively charged patch that is important for maintaining the negative charge at the S2 pocket, and the substitution of D80DD with AAA produced an inactive NS2B-NS3 protease, suggesting that the negative charge of D80DD is important for protease activity (Radichev I, 2008). The crystal structure of the WNV NS2B-NS3 protease complex with inhibitor indicates that the NS2B (D82-F85) directly interacts with NS3 protease active sites. Specifically, the G83 in NS2B interacts with N152 in NS3 (Erbel P, 2006); G83 makes a hydrogen bond with the Lys residue at the P2 position of the substrate (Mueller N H, 2007). Chappell et al. reported that replacing G83 with Ala decreased NS2B-NS3 enzyme activity to less than 5% of that observed for wild-type NS2B-NS3 (Chappell K J, 2008). However, more details regarding the effects of D80DD and G83 on protease activity and the viral life cycle are needed.
Figure 2. Site-directed mutagenesis and alignment of partial flaviviral NS2B aa. A: Sequence alignment of partial flaviviral NS2B cofactor region. WNV, West Nile virus; KUNV, Kunjin virus; JEV, Japanese encephalitis virus; MVEV, Murray Valley encephalitis virus; SLEV, St. Louis encephalitis virus; ZKV, Zika virus; DENV1–4, dengue virus subtype 1–4; YFV, yellow fever virus. B: Mutation of D80DD and G83. D80DD/EEE (no change in charge), D80DD/KKK (change into positive charge) and D80DD/AAA (loss of charge and bulk), G83K (change to positive charge), G83D (change to negative charge), G83A G83F and G83S (no change in charge but different side chain)
In the present study, we constructed a series of mutant NB2B-NS3 proteins and WNV replicons (D80DD/E80EE, D80DD/K80KK, D80DD/A80AA, G83F, G83S, G83D, G83K, and G83A) to investigate the effects of these residues on NS3 activity and viral replication. Our results confirmed that D80-82 and G83 residues in NS2B are essential for protease activity, which is consistent with previous reports. Furthermore, we demonstrated that the negatively charged patch of D80DD is not absolutely required for providing the negative charge at the S2 pocket to promote substrate–protease interaction, but the side chain of the conserved residue G83 is critical for maintaining NS2B function. In addition, we found that the mutagenesis of D80DD and G83 prevent viral RNA replication.
HTML
-
Fluorogenic Peptide Substrate pERTKR-AMC was obtained from R & D Systems (Minneapolis, MN, USA). His-Bind Resin was purchased from Novagen (Madison, WI, USA). Plasmid pET28a-WNV-NS2BCF44NS3 (provided by Dr. Pei-Yong Shi, Wadsworth Center, New York State Department of Health, USA), contained the central portion of NS2B composed of 44 aa residues (cNS2B, residues 1424–1467 of the WNV precursor) and the NS3pro composed of 184 aa residues (cNS3, residues 1506–1689 of the WNV precursor), two of which were linked by a flexible GGGGSGGG linker. The plasmid WNV replicon cDNA encoding the NS2B–NS3 cDNA sequence from West Nile virus (strain 3356, GenBank accession No. AF404756) was also obtained from Dr. Pei-Yong Shi, Wadsworth Center, New York State Department of Health, USA. To obtain the cNS2B sequence, PCR was carried out using the primers cNS2B-F' and cNS2B-R'. To obtain the cNS3 sequence, PCR was carried out using the primers cNS3-F' and cNS3-R'. The NS2B-NS3 chimeric sequence was generated from an overlap PCR reaction with these two PCR products, digested with BamH I/EcoR I and cloned into the same sites in pET28a. BHK21 cells were maintained in Dulbecco modified Eagle medium (DMEM) with 10% or 2% fetal bovine serum (FBS) in 5% CO2 at 37℃.
-
Primers used for the site-directed mutagenesis of NS2B and NS3 were listed in Table 1. Plasmid pET28a-WNVNS2BCF44NS3 was used as template and the mutant amplifications were performed by overlap PCR reactions. The amplified PCR products were further purified separately and employed as templates in the second round PCR with the primer pairs WNVNS2BCF44NS3-F' and WNVNS2BCF44NS3-R'. These PCR products were digested with BamH I and EcoR I and then cloned into plasmid pET28a, which was C-terminally tagged with the His × 6 tag. Nine mutants (D80DD/E80EE, D80DD/K80KK, D80DD/A80AA, G83F, G83S, G83D, G83K, G83A and D75A) were obtained and all mutant fragments were verified by DNA sequencing. NS3 D75A was designated as negative control (Fig. 1B).
Table 1. Primers used in this study
-
Plasmids containing wild-type and mutant NS2B-NS3 recombinant DNA fragments were transformed into Escherichia coli BL21 (DE3), respectively. The recombinant E. coli were selected and cultured in 1 L Luria-Bertani (LB) broth supplemented with 30 mg/mL kanamycin and grown at 37℃ with shaking at 220 rpm. After being induced by 0.5 mmol/L isopropyl-β-D-thiogalactopyranoside (IPTG) for 16 h at 16℃, cells were collected by centrifugation (3 000 rpm, 10 min) and stored at -70℃for 1 h, then the pellets were thawed and resuspended in 20 mL 1×Binding Buffer (20 mmol/L Tris-HCl, 0.5 mol/L NaCl, 5 mmol/L imidazole, pH 7.9) and shaken at 4℃ for 1 h. Cells were disrupted by sonication on an ultrasound disintegrator. The pellet was removed by centrifugation (12 000 rpm, 30 min) at 4 ℃ and the wild-type and mutant NS2B-NS3 recombinant proteases were purified by using His-Bind Kits (Novagen) according to the procedure provided by the manufacture. The concentrations of purified proteins were determined by using BCA Protein Assay Kit (Beyotime Institute of Biotechnology, Haimen, China) and the samples were stored at -70℃ until use. All proteases were analyzed by 12% SDS-PAGE.
-
Wild-type and mutant proteases (D80DD/E80EE, D80DD/K80KK, D80DD/A80AA, G83F, G83S, G83D, G83K, G83A and D75A) in each sample were resolved on a 12% SDS-PAGE and electro-transferred to PVDF Immobilon-P membranes (Millipore, Billerica, MA, USA), blocked with 5% skim milk in TBST, and then reacted with anti-His monoclonal antibody (1:1000 dilution; Qiagen, Valencia, CA) for 1.5 h at room temperature. The secondaryanti-mouse antibody conjugated to Horseradish peroxidase (1:5000 dilution; Roche, Shanghai, CN) was applied to the blots for 1 h at room temperature after 5 times washing with TBST. Then the blots were developed with NBT/BCIP (Roche) after 5 times (10 min/one time) washing with TBST.
-
The ability of wild-type and mutant NS2B-NS3 proteins to make enzymatic cleavages was tested against pentapeptide substrate pERTKR-AMC (Shiryaev S A, 2006). The total reaction volume of the assay was 0.1 mL, and the assay was carried out in 10 mmol/L Tris-HCl (pH 8.0) with 20% (v/v) glycerol. After preincubation of buffered protease and substrate in separate wells at ambient temperature (21–22 ℃) for 5 min, the substrate was mixed with enzyme-buffer solution by automatic shaking for 5 s. Wild-type and mutant protease activities and initial reaction velocities were monitored at an excitation wavelength of λex=380 nm and an emission wavelength of λem=460 nm using a Synergy HT Multi-Detection Microplate Reader (BioTek Instruments, Winooski, VT, USA). Kinetic parameters were calculated by nonlinear regression of initial velocities using nine different substrate concentrations ranging from 0 to 350 μmol/L (350, 175, 87.5, 43.75, 21.88, 10.94, 5.47, 2.73, and 0 μmol/L) assuming Michaelis–Menten kinetics using the equation =Vmax[S]/([S]+Km), where is the initial velocity of substrate hydrolysis, [S] is the concentration of substrate, Vmax is the maximum rate of hydrolysis, and Km is the Michaelis–Menten constant. All assays were conducted in triplicate. GraphPad Prism version 5.0 software (La Jolla, CA, USA) was used for statistical analyses, and the kinetic parameters are reported as mean ± standard error (S.E.).
-
Mutant WNV replicons were constructed based on a WNV replicon with a renilla luciferase-reporter plasmid (Lo M K, 2003) using overlap PCR reactions (primer pairs are shown in Table 1). The first round of PCR products were separately purified and served as templates for the second round of PCR amplification with the primer pairs Replicon-F' and Replicon-R', respectively. The mutated DNA fragments were purified, digested by Sph I and Bsiw I, and engineered into WNV replicon at the Sph I and Bsiw I sites, respectively. All mutants were verified by DNA sequencing.
-
A transient replicon assay was used to quantify viral RNA translation and replication. Briefly, all replicon cDNA templates were linearized with Xba I and purified using phenol-chloroform extraction and ethanol precipitation. Replicon RNA was transcribed with a mMESSAGE mMACHINE kit (Ambion, Austin, TX, USA) according to the manufacturer's protocols. All RNA transcripts (5 μg each) were electroporated into 8×106 BHK21 cells using a GenePulser apparatus (Bio-Rad, Hercules, CA, USA). After a 10-min recovery at room temperature, the transfected cells were resuspended in 25 mL Dulbecco's modified Eagle medium containing 10% fetal bovine serum. Upon transfection of BHK21 cells with wild-type and mutant-type Replicon RNA, 0.5 mL transfected cells were seeded per well in 12-well plates for assaying luciferase activities at 2, 4, 6, 24, 48, and 72 h post-transfection (h p.t.). Triplicate wells were seeded for each data point. Renilla luciferase (Rluc) activity was determined by using an assay kit according to the manufacturer's protocol.
Reagents
Construction of mutant NS2B-NS3
Expression and purification of NS2B-NS3
SDS-PAGE and Western blot
Enzyme characterization
Construction of mutant WNV replicon
In vitro transcription, RNA transfection, and luciferase assay
-
Nine mutants (G83F, G83S, G83D, G83K, G83A, D80DD/E80EE, D80DD/K80KK, D80DD/A80AA, and D75A) were obtained, and all mutant fragments were verified by DNA sequencing. Plasmids containing wild-type or mutant NS2B-NS3 recombinant DNA fragments were transformed into Escherichia coli BL21 (DE3). The recombinant E. coli strains could produce the wild-type and mutant proteases during bacterial growth after isopropylthio-β-galacoside-induction. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) revealed that the purified proteases had a molecular weight of about 36 kD (Fig. 3A), and could be specifically recognized by an anti-His monoclonal antibody on westernblot (Fig. 3B). The migration ratesofD80DD/E80EE and D80DD/K80KK were slightly faster than the other mutants and wild type. We did not detect any proteolysis fraction by SDS-PAGE or western blot.
Figure 3. Analysis of purified wild-type and mutant proteases. A: All proteins were purified from E. coli by Ni2+ metal-chelating chromatography, and protein purity was confirmed by 12% SDS-PAGE. B: All purified proteins were analyzed by western blot using a monoclonal antibody against His. C: Wild-type and mutant protease activities were tested by using substrate pERTKR-AMC. Lanes 1 to 10 represent D80DD/E80EE, D80DD/K80KK, D80DD/A80AA, G83F, G83S, G83D, G83K, G83A, NS3 D75A, and wild-type, respectively
-
Protease activities and kinetic parameters of wild-type and mutant proteases are shown in Table 2. All mutant proteases lost activity (Fig. 3C). While all proteases exhibited variations in Km, Kcat, and Kcat/Km, the Km of the G83F mutation was most similar to the wild-type. The D80DD/E80EE and D80DD/K80KK mutations retained similar Km, Kcat, and Kcat/Km values. The mutagenesis of D80DD/E80EE caused a large decrease in substrate affinity (11-fold increase in Km) and an 18-fold decrease in Kcat/Km. D80DD/A80AA mutagenesis resulted in a 7-fold increase in Km and a 2-fold decrease in Kcat to produce a 12-fold decrease in Kcat/Km. The G83S, G83D, G83K, and G83A mutations reduced Kcat and Kcat/Km, and caused minor increases in Km, whereas G83F produced a similar Km and a 2-fold decrease in Kcat, which resulted in a 3-fold decrease in catalytic efficiency (Kcat/Km). The NS3 D75A protease (negative control) produced an 11-fold decrease in Kcat/Km due to a 5-fold increase in Km and a 2-fold decrease in Kcat versus wild-type, which was consistent with previously published results (D75 in NS3 is a protease active site) (Ryan M D, 1998).
Table 2. Kinetic parameters of wild-type and mutant protease
-
We used a transient replicon system to investigate whether these mutants inhibit viral translation or viral RNA synthesis. Tilgner and Shi reported that BHK21 cells transfected with WNV replicon RNA produced two distinct Rluc signal peaks, one at 2.5 to 10h (representing viral translation) and another at 22.5 h p.t. (representing viral RNA synthesis) (Tilgner M, 2004). The results showed that both the wild-type and the mutants have equivalent luciferase activity at 2, 4, and 6 h p.t., but all mutants suppressed the luciferase signals at 24, 48, 72 h p.t. (Fig. 4). Taken together, the results suggest that D80DD and G83 residues are required for viral RNA synthesis.
Figure 4. The effects of D80DD and G83 based on viral translation and RNA synthesis based on a transient replicon assay. The luciferase signals of all mutants are equivalent with wild-type at 2, 4, and 6 h p.t. but lower than wild-type at 24, 48, and 72 h p.t. All assays were performed in triplicate