Epistatic interactions promote persistence of NS3-Q80K in HCV infection by compensating for protein folding instability

Figures

• 

  Figure 1. Prevalence of epistatic secondary substitutions in DAA-naïveand DAA-experienced patients harboring NS3-Q80K. Comparison of mutant patterns in patients with genotype 1a HCV infection harboring NS3-Q80K between the European HCV database (24), euHCVdb (DAA naïve, n = 100), and DAA-experienced patients harboring NS3-Q80K from the Frankfurt Resistance Database (upon DAA failure, n = 41). Data shown represent percentages (%) of patients from which the respective variants were isolated as major variants before DAA treatment or upon treatment failure. DAA, direct-acting antiviral; HCV, hepatitis C virus.

• 

  Figure 2. Thermal shift of the NS3-4A protease harboring NS3-Q80K and epistatic amino acid substitutions. Data from real-time thermal stability assay using Sypro Orange, a temperature-stable fluorophore that exhibit enhanced fluorescence upon interacting with unfolded proteins. A, thermal stability of the WT protease and the NS3-Q80K mutant, as well as (B–F), mutants harboring epistatic secondary substitutions assessed under increasing incubation temperatures. Graphs showing impact of mutations on protease protein unfolding patterns characterized by fluorescence emission curves.

• 

  Figure 3. Impact of NS3-Q80K and epistatic amino acid substitutions on the protease protein fold.Tm from purified NS3-4A protease WT and mutants are determined by fitting the sigmoidal melt curve as shown in Figure 2 to the Boltzmann equation. Error bars represent the mean ± SD from at least three independent experiments; ∗p ≤ 0.05; ∗∗p ≤ 0.01; and ∗∗∗p ≤ 0.001; by two-sided t test.

• 

  Figure 4. Impact of NS3-Q80K and epistatic amino acid substitutions on the protease enzymatic function and reaction velocity. Reaction velocity and Michaelis–Menten kinetics as assessed from purified protein of WT protease and NS3-Q80K mutants and the natural polyprotein substrate NS4A/4B. A, reaction constants kcat and (B) KM calculated after nonlinear regression curve fitting. Error bars represent the mean ± SD from at least three independent experiments; ∗p ≤ 0.05; ∗∗p ≤ 0.01; and ∗∗∗p ≤ 0.001; by two-sided t test. C, kinetic progress curves with reaction velocities (RFU, relative fluorescence unit) for the NS3-4A protease and the natural polyprotein substrate NS4A/4B as assessed from purified protein of WT protease and NS3-Q80K mutants.

• 

  Figure 5. Protease fold and protein flexibility estimated by root mean square fluctuations (RMSFs).A, protease structure with cocrystallized natural peptide substrate (in orange) (from Protein Data Bank [PDB] ID: 3M5O (62)); the protein backbone given as ribbon model with transparent surface depiction. Residues of interest highlighted as stick models and colored as follows: residue 80, blue; residue 91, light green; and residue 174, dark green. B, impact of NS3-Q80K and epistatic secondary substitutions on the protease protein flexibility estimated by RMSFs over time for different NS3-Q80K mutant patterns. C, protease protein flexibility averaged over time for protease WT (from PDB ID: 1CU1 (48)) and NS3-Q80K mutants represented as tubes. Epistatic secondary substitutions differentially affect the bilobal fold of the protease. We used GROMACS to generate RMSF B-factors added to the PDB file. We then used PyMOL defaults to visualize flexible, stable, and intermediate residues. Highly flexible residues are shown in red, highly stable residues are shown in blue, and the intermediate flexible residues are shown in white. The size of the tubes is proportional to the observed RMSF values in…

• 

  Figure 6. Impact of epistatic amino acid substitutions on RNA replication of NS3-Q80K variants. Medium was collected and replaced at 8, 24, 48, 72, and 96 h after transfection of H77S.3/GLuc2A RNAs carrying the indicated mutations, and GLuc activity determined at 72 h (left panel) and 96 h in time (right panel). Results normalized to the 8 h GLuc activity represent the mean of triplicate samples and are representative of multiple experiments. Error bars represent the mean ± SD from at least three independent experiments; ∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; and ∗∗∗∗p ≤ 0.0001; by two-way ANOVA. GLuc, Gaussia luciferase.

• 

  Figure 7. Localization of NS3-Q80K, epistatic interaction sites, and the T-cell epitope NS31073–1081. Surface representation showing the destabilized N-terminal part of the NS3-Q80K protease structure (orange; residues 7–78, based on RMSF from Fig. 5B) (from Protein Data Bank ID: 3M5O (62)) comprising the immunodominant T-cell epitope NS31073–1081 (red circled; corresponding to protease residues 47–55); residue 80 (blue); and epistatic interaction site 91 (light green) and 174 (dark green). RMSF, root mean square fluctuation.