(A) Sequence comparison of the NS2 segment 1–100 from Con1 and JFH1 used to design the NS2 mutants. Amino acids are numbered with respect to NS2. The helical segments in the membrane binding domain deduced from NMR analyses of Con1 NS2 peptides ([6] and this study) are shown at the top (h, helical). Identical, highly similar, and similar residues at each position are symbolized by an asterisk, a colon, and a dot, respectively, according to Clustal W convention. Boxes indicate helix segments that were exchanged with the helix-swap mutants (for nomenclature of the mutants see legend to Figure 2). Residues exhibiting strong physico-chemical differences within these boxes are colored violet. Arrows indicate residues that were mutated in this study and point to the substituting residues that are color-coded according to their effect on infectious particle assembly: green, no significant effect; orange and red, medium and strong reduction of infectious virus production, respectively. Green double arrows shown in the bottom refer to the three putative TMS as deduced from the predictions. (B) Membrane association mediated by individual NS2 TMS. U-2 OS cells were transiently transfected with pCMVNS2-GFP or pCMVGFP (upper panel) or with pCMVNS2_1-27-GFP, pCMVNS2_27-59-GFP, pCMVNS2_60-99-GFP or pCMVNS2_1-99-GFP, as indicated in the captions (lower panel). Twenty four to 48 h later cells were fixed and analyzed by confocal laser scanning microscopy. GFP fusion proteins were detected by autofluorescence, ER membrane localization of NS2-GFP was confirmed by indirect immunofluorescence by using PDI as ER marker (upper panel). (C) Ribbon diagrams of representative NMR structures of NS2[27]-[59] and NS2[60-99] (left and right, respectively). Residues are color-coded according to their physicochemical properties. Hydrophobic residues are grey, aromatic residues are dark grey, Gly is light grey, Pro is black. Polar residues are yellow and positively and negatively charged groups of basic and acidic residues are blue and red, respectively. Histidine residues are cyan, the cystein residue is green. Some aa are labelled to provide a visual reference for residue positions in the structures. Figures were generated from structure coordinates (PDB entry: 2KWT and 2KWZ for NS2[27]-[59] and NS2[60-99], respectively) by using the VMD program [67]) and rendered with the POV-Ray software package. (D) Model of NS2 membrane topology. In the left panel TMS 1, 2 and 3 are shown in ribbon representation and colored bronze, green and copper, respectively. The TMS are tentatively positioned in the membrane and the limits of transmembrane helices are given (TMS1, 4–23; TMS2, 27–49; and TMS3, 72–94). The three TMS are represented as separated entities, since their intramolecular and/or intermolecular interactions are not known. The NS2 protease domain is shown in surface representation (side view) with dimer subunits (PDB entry 2HD0) shown in light blue and pink. For simplicity, the MBD for only one protease subunit is shown. The phospholipid bilayer is tentatively and schematically represented. The ribbon model of NS2 MBD in the right panel shows the location of mutated residues analyzed in this study. Side chain atoms of mutated aa are shown as spheres corresponding to van der Waals radius and color-coded as described in panel A to indicate the assembly phenotype. Backbones of residues with low similarity in helix-swap mutants are colored violet.

(A) Huh7.5 cells were transfected with JFH1mut4-6 or derivatives thereof containing single point mutations specified in the bottom. Amounts of infectious particles released into the supernatants of cells 24, 48 and 72 h after transfection were determined by TCID₅₀ assay. (B) Comparison of intra- and extracellular infectivities obtained with Huh7.5 cells 48 h after transfection. Representative results of three independent experiments with standard deviation are shown in panels (A) and (B). Background of the assays was determined by using JFH1-ΔE1E2 (black line and arrow head). (C) Western blot analysis of viral proteins (core, NS2, NS3 and NS5A) in Huh 7.5 cells 48 h p.t. with constructs specified below the panels. As negative control we used an NS2 mutant unable to replicate because of a mutation blocking the active site of the protease (NS2-H952A; [6]). Beta actin detected on the same blots was used as internal loading control. Wt, wild type; tNS2, truncated NS2.

(A) Schematic representation of the JFH1mut4-6 constructs in which the N-terminus of NS2 was fused to either the Flag-tag (F) or two copies of this tag (FF) or a hexahistidin-Flag fusion tag (HisF) or a HA-Flag fusion tag (HAF). The cleavage site between p7 and NS2 is indicated with a dotted arrow and the scissor; aa sequences of the duplicated N-terminal 5 aa residues and the Gly-Ser linkers are displayed. (B) Kinetic of virus production attained with JFH1mut4-6 derivatives depicted in panel (A) was determined by transient transfection of Huh7.5 cells in parallel with the parental genome (wt). A representative result of three independent experiments with standard deviations is shown. Background of the assays was determined by using JFH1-ΔE1E2 (black line and arrow head). (C) Expression of HCV proteins was determined by Western blot analysis 48 h after transfection using antibodies specified in the right. (D) Interaction of NS2 with structural and nonstructural proteins. JFH1mut4-6 without a tagged NS2 (wt) or a derivative thereof with the HA- and Flag-tag fused to the N-terminus of NS2 (HAF-NS2) were transfected into Huh7.5 cells. Seventytwo hours later cells were harvested and lysates were used for immunoprecipitation with an HA-specific antibody. After extensive washing, immunocomplexes were analyzed by electrophoresis into a Tris-Tricine gel and subsequent Western blot using antibodies given in the right of each panel (NS2 IP). Ten percent of the input used for immuno-capture was analyzed in parallel (left panels). Mock-transfected cells served as negative control of Western blot and the wt genome lacking the HA-tag was used as specificity control for immunoprecipitation.

Huh7-Lunet cells were transfected with JFH1mut4-6 HAF-NS2 and 36, 48 and 72 h later, NS2 and LDs were detected by using an HA-specific antibody (red) or BODIPY staining (blue), respectively. (A) Representative images of cells 36 and 72 h p.t. (left and right half, respectively). An overview is shown in the top, a magnification of a single cell from a different view field below. Boxed areas are enlarged in the smaller panels to the left and right, respectively. Enlargements show 3D reconstructed images created with the Volocity 5.3 software. In case of the 72 h value 3 representative structures are displayed. The staining pattern shown for the 36 h time point and in panel 2 for the 72 h time point corresponds to phenotype 1. (B) Huh7-Lunet cells were transfected with the wild type (left panel) or given NS2 mutants (right panel) and analyzed as described above. Shown are the ratios of phenotype 1 and 2. In case of the wild type values determined at different time points after transfection are shown, in case of the mutants only the values measured 72 h p.t. For each construct, 180 cells were analyzed.

(A) Huh7-Lunet cells were transfected with helix-swap mutation CT1.h or its corresponding pseudorevertant (CT1.h-p7) and NS2 colocalization with E2, NS3 and NS5A as well as their accumulation around LDs was determined by immunofluorescence. Cells were analyzed 72 h post transfection as described in Figure 8. The small inserts in the bottom of each panel represent 3D reconstructed enlargements of each corresponding image created with the Volocity 5.3 software. (B) Colocalization of NS2 with E2 or NS3 or NS5A was quantified by determining Pearson's correlation coefficients providing a measure for the relative degree of co-localization of proteins. For each sample 50 cells were quantified by using the ImageJ software. For further details see legend to Figure 8.

(A) A schematic diagram of the cell culture adapted JFH1 genome construct (JFH1mut4-6; [10]) used for functional characterization of the NS2 membrane binding domain (MBD) is shown in the top. The titer enhancing mutations are indicated with asterisks. The membrane topology model of the N-terminal MBD is schematically shown below. Helices in individual TMS are represented by cylinders and numbered. Note that TMS3 contains 3 individual helices (h1–h3). C refers to Con1 that was used to replace the corresponding JFH1 fragment in NS2. For instance, CT1.h refers to the replacement of the helix of JFH1 TMS1 by the one of Con1 etc. (B) Huh7.5 cells were transfected with constructs specified in the bottom and kinetics of release of infectious particles into culture supernatants were quantified at time points given in the top by TCID₅₀ assay. (C) Intracellular and extracellular infectivities were determined 48 h post transfection by using TCID₅₀ assay. In panels (B) and (C) representative results of three independent experiments with standard deviations are shown. Background of the assays was determined by using the JFH1-ΔE1E2 mutant in which the region encoding the envelope glycoproteins was deleted and that does not release infectious particles (black line and arrow head). (D) Western blot analysis of HCV proteins expressed in Huh7.5 cells after transfection with constructs specified in the bottom. Cell lysates were prepared 48 h post transfection and proteins were detected by using antibodies with specificities given to the right side of each panel. Beta-actin detected on the same blot was used as internal loading control.

JFH1mut4-6 mutants carrying either single aa subtitutions (panel A and B) or helix swaps (C) causing a profound reduction of virion production were used for insertion of pseudoreversions that are specified in Table 1. Parental NS2 mutants and corresponding pseudorevertants were transfected into Huh7.5 cells and kinetic of virus production was determined by TCID₅₀ assay at time points specified in the top. A representative result of two independent experiments with standard deviations is shown in each panel. Background of the assays was determined by using JFH1-ΔE1E2 (black line and arrow head). Shown below each panel are the NS2 expression levels of mutants and the corresponding pseudorevertants as determined by Western blot using lysates of Huh7.5 cells prepared 48 h after transfection. A more comprehensive analysis of additional viral proteins is shown in Supplementary Figure S4.

JFHmut4-6 (wild type without tag; wt), JFHmut4-6HAF-NS2 with the HA-Flag double tag or helix swap mutants and double mutants derived from this construct were transfected into Huh7.5 cells and harvested 72 h later. (A) Cell lysates were used for immunoprecipitation (IP) with an HA-specific antibody and immunocomplexes were analyzed by Western blot (WB) as specified to the left and right of each panel. For comparison, 10% of the lysate used for the IP was analyzed in parallel. The wt genome lacking the HA-tag was used as specificity control for immunoprecipitation. Note that in case of the p7 mutant the pseudoreversion (E3D) resides in the epitope recognized by the p7-specific antibody and thus this p7 protein is not detected (labeled with an arrow). (B) Western blot signals of proteins in immunocomplexes from two independent experiments were quantified by densitometry scanning and normalized to the amounts of the respective viral protein present in the cell lysate. Values obtained with the wild type genome were set to 100%. (C) Relative enhancement of co-immunoprecipitation of NS2 with HCV proteins specified in the top by pseudoreversions. Values obtained with the corresponding NS2 mutant in the absence of the pseudoreversion were set to 1 (highlighted with the grey line and the black arrow head) and used for normalization of the values measured with the pseudorevertants.

(A) Huh7-Lunet cells were transfected with JFHmut4-6-HAFNS2 (wild type, wt) and 36 h or 72 h later NS2 was detected with an HA-specific antibody (red). E2, NS3 or NS5A were detected with mono-specific antisera and LDs were labeled with BODIPY (blue). Pictures were deconvolved by using the Huygens Essential 3.5 software and a theoretical point spread function. The small inserts in the bottom of each panel represent 3D reconstructed enlargements of each corresponding image created with the Volocity 5.3 software. (B) The degree of colocalization was quantified by determining Pearson's correlation coefficients. For each sample 50 cells were quantified by using the ImageJ software. The background of the assay (0.25) was determined by analyzing cells co-expressing cytosolic RFP and HA tagged NS2; the maximum of colocalization (0.85) was established by using double staining of E2 with two different antibodies.

The structural models of NS2 (this study), p7 [16], the transmembrane domain of E1 (TM-E1;[68]) and E2 (TM-E2; theoretical model) are represented as ribbon. The phospholipid bilayer is indicated tentatively and schematically. Only the side chains of aa encoded by pseudoreversions are shown with spheres correspondíng to van der Waals radia. Residues likely involved in intra- or inter-molecular interactions are indicated with the same color and connected with double arrows. Interactions within and outside of NS2 are represented by small and large double arrows, respectively. The backbone of aa residues with low similarity identified in helix-swap mutants and potentially involved in inter-helix interactions are colored violet as in Figure 1D. The large gray arrow indicates the region of NS2 likely involved in interaction with NS3. Residues colored in pink-gray (position 45) and light green (position 73) correspond to pseudoreversions at the very same site as the original NS2 mutation. Numbering of altered residues in TM-E1 and TM-E2 correspond to JFH1 and numbers in parenthesis correspond to the H77 reference polyprotein. The molecular models of TM-E1 and TM-E2 were constructed by using the Swiss Model server facilities (http://www.expasy.ch/swissmod/). The figure was generated on the basis of the structure coordinates by using the Visual Molecular Dynamics program and rendered with POV-Ray.