Cells were infected with WNV_KUN strain MRM61C at an approximate multiplicity of infection (m.o.i.) of 3 as has been described previously (Westaway et al., 1997a). Vero and BHK cells were maintained in DMEM supplemented with 5% FCS (Lonza, Basel, Switzerland) and penicillin/streptomycin (100 U/mL and 100 μg/mL, respectively, GIBCO-BRL) at 37°C with 5% CO₂.

WNV_KUN specific anti-NS1 (clone 4G4; Macdonald et al., 2005) and anti-NS5 (clone 5H1.1) monoclonal antibodies were generously provided by Dr. Roy Hall (University of Queensland, Brisbane, Australia). WNV_KUN-specific rabbit anti-NS4A polyclonal antisera has been described previously (Mackenzie et al., 1998). Rabbit anti-β-1,4-galactosyltransferase (GalT) polyclonal antibodies (Berger et al., 1993) were generously provided by Dr. Eric Berger (University of Zurich, Zurich, Switzerland). Rabbit anti-Erlin-2 antibodies were purchased from Cell Signaling Technologies and mouse anti-dsRNA (clone J2) antibodies were purchased from English & Scientific Consulting Bt. (Hungry). Alexa Fluor 488- and 594-conjugated anti-rabbit and anti-mouse specific IgG were purchased from Molecular Probes (Invitrogen, Leiden, Netherlands).

Recombinant cDNA plasmids expressing eGFP-tagged erlin-1, erin-2, prohibitin, stomatin, and flotillin were kindly provided by Dr. Stephen Robbins (University of Calgary, Canada). Plasmids were introduced into cells via Lipofectamine delivery following the manufacturer’s instructions.

Vero cell monolayers on coverslips were infected with WNV_KUN and incubated at 37°C for 24 h. The cells were subsequently washed with PBS and fixed with 4% paraformaldehyde (Sigma Aldrich, St. Louis, MO) and permeabilized with 0.1% Triton X-100 as previously described (Malet et al., 2007). Primary and secondary antibodies were incubated within blocking buffer (PBS containing 1% BSA) and washed with PBS containing 0.1% BSA between incubation steps. After a final wash with PBS the coverslips were drained and mounted onto glass slides with a quick dry mounting medium (United Biosciences, Brisbane, Australia) before visualization on a Leica TCS SP2 confocal microscope. Images were collected using a Leica digital camera and Leica 3D software before processing for publication using Adobe Photoshop^TM software. Colocalization was quantified based on the fluorescence microscopy images and was performed using ImageJ software via the colocalization analysis plug-in JACoP. Images were collected from replicate experiments with at least 20 cells counted from each.

Tyrosine to serine (Y/S), at position 28 within NS4A, or lysine to leucine (K/L), at position 35, or double mutations (Y/S + K/L) were generated in the WNV_KUN cDNA infectious clone, FLSDX (Khromykh et al., 1998), using site-directed mutagenesis (Stratagene). All clones were sequenced prior to maxiprep amplification (Invitrogen) and transfection.

All replicon templates were linearized with XhoI (New England BioLabs) and purified using Phenol/Chloroform extraction and ethanol precipitation. In vitro RNAs were transcribed using 1 μg linearized DNA template using established methods previously described (Khromykh and Westaway, 1997; Khromykh et al., 1998), except introduction of the in vitro transcribed RNAs into mammalian cells was performed via Lipofectamine-mediated delivery or electroporation via the Neon^TM transfection system (Invitrogen) following the manufacturers instructions. Briefly, Vero or BHK cells (1.0 × 10⁷) were electroportated using a 100 μL tip, at 1,300 volts, width 20, and 2 pulses. Cells were resuspended in cell culture media, seeded, and incubated for various periods for examination of virus replication.

Vero C1008 cells were seeded in DMEM complete media in 6-well plates and incubated at 37°C overnight. Virus stock was diluted 10-fold in 0.2% BSA/DMEM and cells were infected with 300 μL of stock dilutions (in duplicate) and incubated at 37°C for 60 min. Two milliliter of a semi-solid overlay containing 0.3% w/v low-melting point agarose, 2.5% w/v FCS, 1% Penicillin/Streptomycin, 1% Glutamax, 1% HEPES and 0.1% NaCO₃ was added to cells and solidified at 4°C for 30 min. Cells were incubated at 37°C for 3 days, fixed in 4% v/v formaldehyde (in PBS) for 1 hour and stained in 0.4% crystal violet (with 20% v/v methanol and PBS) at RT for 1 hour. Plaques were manually counted and plaque-forming units per mL (pfu/mL) calculated.

Transfected cells were aspirated in PBS then lysed in SDS lysis buffer (0.5% SDS, 1 mM EDTA, 50 mM Tris-HCl) containing protease inhibitors leupeptin (1 μg/mL) and PMSF (0.5 mM) and phosphatase inhibitors sodium orthovanadate (25 mM), sodium fluoride (25 mM) and β-glycerophosphate (25 mM) (Sigma). Lysates were diluted in LDS loading buffer (Invitrogen), heated at 70°C for 5 min and separated on a 10% Tris-Glycine polyacrylamide gel. Proteins were transferred to Hi-Bond ECL nitrocellulose membrane (Amersham Biosciences) and the membrane was blocked with 5% w/v skim milk (Diploma) in TBS with 0.05% Tween (PBS-T). Primary antibodies were incubated at 4°C with membrane overnight in blocking solution as above. Following primary incubation, the membrane was washed in TBS-T then incubated with secondary antibodies conjugated to Cy5 (Amersham Biosciences), Alexa Fluor 647 or Alexa Flour 488 (Invitrogen) in TBS-T at RT for 2 h. The membrane was washed twice in TBS-T then TBS, and proteins visualized on the Storm Fluorescent scanner (Amersham Biosciences) on either 635 nM or 430 nM emission channel.

Cells were fixed with 3% glutaraldehyde in 0.1 M cacodylate buffer for 2 h at room temperature. Cells were washed several times in 0.1 M cacodylate buffer followed by fixation with 1% OsO4 in 0.1 M cacodylate buffer for 1 h. After washing of the cells in 0.1 M cacodylate buffer, specimens were dehydrated in graded acetones for 10–20 min each. Subsequently, samples were infiltrated with EPON resin and polymerized in molds for 2 days at 60°C. 50–60 nm thin sections were cut on a Leica UC7 ultramicrotome using a Diatome diamond knife and collected on formvar-coated copper mesh grids. Before viewing in a JEOL 2010 transmission electron microscope cells were post-stained with 2% aqueous uranyl acetate (UA) and Reynold’s lead citrate.

Tyrosine to serine (Y/S), at position 28 within NS4A, or lysine to leucine (K/L), at position 35, or double mutations (Y/S + K/L) were transferred from the mutant FLSDX to a pcDNA3.1 NS4A(-2K)-6xHis expression construct. Briefly, the NS4A-2K amplicon was amplified from FLSDX using forward (ATGGGCCCACCATGTCTCAAATAGGT) and reverse (TAT TTCTAGACTAATGGTGATGGTGATGGTGGCGTTGCTTCT CTGGCTCAGG) primers and ligated into pcDNA3.1 using EcoRV and XbaI (Promega). Following propagation and midiprep purification (QIAGEN), 5 μg of DNA was transfected into 293T cells using Lipofectamine 2000 (Life Technologies) as indicated by manufacturers. At 24 h.p.t, cells were lysed on ice in a cholesterol extraction buffer (50 mM Tris pH 8.0, 150 mM NaCl, 0.1% w/v Digitonin, 0.1% NP-40, 0.5% Triton-X 100 and 0.25% sodium deoxycholate) containing a protease inhibitor cocktail (Astral Scientific), and centrifuged at 10,000 rcf to pellet cellular debris. Immunoprecipitation was performed on the lysates using an anti-His6 antibody (Abcam) coupled with Protein A Sepharose (Life Technologies) to pull-down the expressed NS4A-2K proteins. Lysates and IP samples were then assayed for cholesterol content using the Amplex Red cholesterol assay kit as indicated by the manufacturers. Total cholesterol (in μg) was calculated using internal standard curves and error bars indicate ± 1 standard deviation (n = 4).

Data is representative of 3 independent experiments and was analyzed by unpaired Student’s t-test using GraphPad Prism v8.0.

Previously we have shown that WNV_KUN redistributes intracellular cholesterol to the sites of viral RNA replication, the VP (Mackenzie et al., 2007b). In addition, we have observed that the biogenesis of the VP appears to occur on a membrane platform derived from the ER, and recruitment of both host and viral proteins occurs at a pre-Golgi step (Gillespie et al., 2010). To further characterize the interactions that occur during development of the VP we utilized two GFP-tagged protein markers, erlin-1 and erlin-2, originally used to define cholesterol-rich micro-domains within the ER (Browman et al., 2006). Both proteins fall within the growing family of prohibitin domain-containing (PHB) proteins and are selectively targeted into ER micro-domains in a cholesterol dependent manner (Browman et al., 2006). We examined subcellular distribution of ectopically expressed eGFP-tagged erlin-1/2 proteins within the ER membranes in WNV_KUN-infected Vero cells. Immunofluorescence (IF) analysis indicated that a significant pool of erlin-1/2 was confined to the cytoplasmic foci recognized by anti-dsRNA antibody (Figures 1a–h). The specific accumulation of erlin proteins within WNV_KUN replication sites was further confirmed by observation that other PHB family members, with different intracellular targets, did not co-localize with anti-dsRNA antibody stained foci (Figures 1i–t). In agreement with published findings, eGFP-tagged versions of prohibitin-1, flotillin-1 and stomatin-1a, were found to inhabit lipid micro-domains in the mitochondria (Figure 1i) and in endosomes and the plasma membrane (Figures 1m,q), respectively, quite distinct from WNV_KUN dsRNA.

These results suggest that WNV_KUN replication is associated with cholesterol enriched membrane domains within the ER, and co-located with the host proteins erlin-1 and 2.

As we had observed previously and above; cholesterol appears to play a significant role in the biogenesis and establishment of the WNV_KUN RC (Figure 1; Mackenzie et al., 2007b) we aimed to identify whether a specific viral gene product is actively involved in lipid recognition/and redistribution. Our initial studies focused on the WNV_KUN NS4A protein as we, and others, had previously observed that flavivirus NS4A has the capacity to remodel intracellular membranes (Roosendaal et al., 2006; Miller et al., 2007). To this end, we utilized gene mining and revealed that NS4A contains a potential cholesterol recognition/interaction amino acid consensus (CRAC) motif [L/V²⁴-X_(1–5)-Y²⁸-X_(1–5)-R/K³⁵] motif near its’ N terminus (Figure 2A). Sequence alignments revealed that this motif is highly conserved within members of the Japanese Encephalitis subgroup, with limited homology within the other members of the flavivirus genus (Figure 2A). A CRAC motif has been identified in the peripheral-type benzodiazepine receptor (Li and Papadopoulos, 1998) and other proteins known to bind cholesterol, including caveolin-1 (Parton et al., 2006), human immunodeficiency virus (HIV) gp41 (Vishwanathan et al., 2008) and the influenza M2 protein (Schroeder et al., 2005). Although it should be noted that presence of a CRAC motif does not solely confer the ability of a protein to associate with cholesterol (e.g., caveolin and cholesterol binding; Epand et al., 2005). To interrogate the importance of the CRAC domain on virus replication we generated recombinant viruses containing inhibitory mutations within this motif in the NS4A protein. Mutations were made within the WNV_KUN infectious clone (FLSDX) to alter residues within NS4A at position 28 from tyrosine to serine (Y/S), or to change residue at position 35 from lysine to leucine (K/L). Mutation of these amino acids was previously observed to disrupt cholesterol binding and trafficking of the major myelin protein P0 (Saher et al., 2009) and influenza virus M2 protein (Thaa et al., 2011).

Recombinant viruses harboring corresponding single or combined double mutations (Y/S + K/L) exhibited attenuated phenotypes varying in degrees (Figure 2B), of which the double mutation was shown to be extremely attenuated and was difficult to analyze even after subsequent re-infection. Viral replication efficiency was assessed by IF, Western blotting and plaque assay in Vero cells transfected with RNAs transcribed in vitro from FLSDX cDNAs (Figures 2B,D). It was observed that recombinant viruses harboring the K/L mutation displayed a slight delay in growth kinetics and was moderately attenuated, as assessed by plaque assay (Figure 2D). Whereas recombinant viruses harboring the Y/S mutation were more attenuated than the K/L and WT viruses. We failed to recover any of the double mutant in Vero cells. However, we could detect basal virus protein production and infectious virus release in BHK cells that were electroporated with the viral Y/S + K/L RNA (data not shown). Sequencing analysis of the recovered viruses revealed that no compensatory mutations had occurred, even after prolonged incubation of the transfected cells (data not shown).

These results indicated that the potential CRAC motif within the WNV_KUN NS4A protein appeared to play a major role during the replication cycle. The results also indicated that conservation of Tyr at position 28 is critical for efficient replication of the WNV_KUN genome and double mutation of Tyr-28 and Lys-35 is lethal to WNV_KUN replication in Vero cells.

To determine the stage in the replication cycle affected by the Y/S and K/L mutants, we assessed the ability of these constructs to form the WNV_KUN RC by IF analysis. Vero cells were electroporated with the individual CRAC mutants and at 48 h.p.e., localization and distribution of the RC was determined by immuno-staining for dsRNA and NS4A or NS1 and NS4A (Figure 3). For both wild-type (FLSDX) and the K/L mutant, dsRNA and NS1 were detected in individual and often sizable foci scattered throughout the perinuclear/reticular area of the cytoplasm (Figures 3Aa–c,g–i), that largely coincided with NS4A (Figures 3Ad–f,Bj–l). Comparatively in the case of the Y/S mutant, dsRNA and NS1 were further dispersed in smaller foci (Figures 3Cm–o,D) and there was a reduction in observable co-localization between the dsRNA and NS4A (Figures 3Cp–r). These results suggested that RC formation was significantly impaired in the Y/S mutant viruses.

As previously shown, WNV_KUN replication is associated with cholesterol enriched membrane domains in the ER (Figure 1), and recently YFV replication was observed to localize to detergent resistant membranes high in cholesterol content (Yi et al., 2012). Thus, we aimed to further characterize the defect in RC formation by assessing whether the Y/S and K/L mutants had retained the ability to localize to cholesterol-rich domains within the ER, as marked by the host protein erlin-2. Vero cells were electroporated with mutant genome RNAs and the eGFP-tagged erlin-2 plasmid DNA to define cholesterol-rich domains in the ER (Figure 4A). At 48 h.p.e., localization and distribution of the RC was determined by immuno-staining for dsRNA. In transfected and mock-infected cells, eGFP-erlin-2 was observed to localize in a typical reticular ER-like pattern without any noticeable cytoplasmic foci (Figures 4Aa–d). For both FLSDX and the K/L mutant, dsRNA was confined to large reticular/perinuclear cytoplasmic foci where eGFP-erlin-2 was significantly co-located (Figures 4Ae–h,m–p). However, we observed that eGFP-erlin-2 and dsRNA were dispersed in visibly smaller foci throughout entire cytoplasm of the cell and revealed reduced co-localization in cells replicating the Y/S mutant (Figures 4Ai–l). These results suggested that there was an impairment of the WNV_KUN Y/S mutant to localize to lipid microdomains in the ER to establish biogenesis of the RC, an impairment that correlates with attenuation of replication efficiency (Figure 2).

To further investigate the defect in RC formation we assessed whether it coincided with a defect in recruitment of cellular proteins known to localize to the RC (Mackenzie et al., 1999). Vero cells were electroporated with mutant genome RNAs and at 48 h.p.e., localization of the RC and trans-Golgi network glycoproteins was determined by immunostaining for dsRNA and GalT (Figure 4B). For both FLSDX and the K/L mutant, dsRNA and GalT colocalized in foci in the perinuclear/reticular region of the cytoplasm (Figures 4Ba–d,i–l). Comparatively in the case of the Y/S mutant, dsRNA and GalT were dispersed in significantly smaller foci throughout the cytoplasm and there was a reduction in the co-localization between the two (Figures 4Be–h). These results highlighted that there was dissociation between cellular and viral components that are normally found within the RC in the Y/S mutant viruses.

Overall, our results suggest that the CRAC motif within the WNV_KUN NS4A protein appears to play a major role facilitating efficient virus replication. In particular that the Y/S mutation significantly impaired the capacity of the mutant viruses to effectively form the WNV_KUN RC complex, which may be related to the inability of mutant viruses to unite the required viral and cellular replicative components to specific subdomains on the ER membrane.

During WNV_KUN infection ER membranes undergo extensive proliferation and rearrangements such that they can be observed as an intricate network of CM/PC and VP via ultrastructural analysis (Mackenzie et al., 1996a; Westaway et al., 1997b; Mackenzie, 2005). The CM/PC is thought to function during translation and proteolytic maturation of the polyprotein, whereas the VP is known to house the RC (Mackenzie et al., 1996a, 1998, 2007a; Westaway et al., 1997b, 1999). In view of the defect in RC formation in the Y/S mutant observable via IF analysis, we were interested to investigate whether the similar characteristic membrane structures, particularly the VP, were induced in these mutant viruses. Thus, Vero cells were electroporated with the individual CRAC mutants, fixed at 48 h.p.e. and analyzed by electron microscopy (Figure 5). Interestingly, VP formation was decreased but observable in the Y/S mutants in comparison to the FLSDX and K/L mutants. More strikingly, however, was that recombinant viruses harboring the Y/S mutation were observed to be significantly impaired for CM/PC induction; the induced CM/PC structures were abundant and could clearly be defined in the FLSDX and K/L mutants but were almost absent in the Y/S mutants (compare panels in Figure 5). All mutant viruses appeared to produce virus particles, albeit greatly reduced in the Y/S mutant. No evidence of infection was observed for the double mutant.

These results indicated that the CRAC motif within the WNV_KUN NS4A protein appeared to play a major role in facilitating membrane proliferation and remodeling.

During our studies we could not directly observe co-location of NS4A with endogenous erlin-2 due to conflict with the same species of antibodies. As an alternative to understand this we transfected Vero cells with cDNA expression plasmids expressing the individual NS4A fused to a 6xHIS tag and co-stained those cells with anti-erlin-2 antibodies (Figure 6A). As can be observed expression of WT and all single NS4A mutants induced the formation of cytoplasmic foci that co-located strongly with erlin-2 for WT and NS4A K/L and to a lesser degree for NS4A Y/S (quantitation provided in Figure 6B as Manders’ co-efficient). In contrast, the NS4A K/L + Y/S mutant displayed a more diffuse cytoplasmic staining with little co-location with erlin-2 (Figure 6A).

To address whether the CRAC motif within NS4A conferred a biochemical interaction with cellular cholesterol we expressed the WT and mutant NS4A proteins transiently in 293T cells (due to increased transfection efficiency). Our initial analysis to co-localize intracellular cholesterol (using filipin or perfringolysin-O) with NS4A was unsuccessful (data not shown). Therefore, we aimed to isolate NS4A and determine if we could co-purify cholesterol (or not) after immune-precipitation. As observed in Figure 6D we could immune-purify all forms of the NS4A protein and could interact with a similar percent of cholesterol, as determined by the Amplex Red cholesterol assay kit, with the WT and K/L mutant NS4A (Figure 6C). In contrast we observed less cholesterol isolated with the Y/S and K/L + Y/S mutant NS4A proteins, however, these results were not significant even after repeated experiments (Figure 6C; n = 4). At this stage we cannot conclusively determine whether the potential CRAC motif is promoting an association with cholesterol or whether other domains within NS4A or additional protein-protein or protein-lipid interactions are occurring.

Overall, we can conclude that the CRAC motif is important for WNV_KUN replication and appears crucial for the ability of NS4A to co-locate with erlin-2 within microdomains within the ER to remodel intracellular membranes and promote WNV_KUN RC assembly and function. Whether cholesterol is indeed a contributing factor is still unresolved.