Fig 1.
Insertion of HVR1 from DH5, SA13 and HK6a into H77 or S52 HVR1 into J4 attenuates the recombinant HCVcc.
(A-D) Adaptation of the HVR1-swapped viruses in Huh7.5 cells during the indicated time period. The cells were transfected with the indicated in vitro transcribed HCV RNA of the indicated recombinants. Viral spread was monitored by immunostainings of infected cells and sequence analysis was performed on amplified envelope protein encoding nucleotide sequences of the cell culture adapted HVR1-swapped recombinants. Spread of a representative parental H77 (A-C) and J4 (D) recombinant is shown in black in all panels. Only coding mutations are depicted. Gray squares indicate cell cultures that were closed due to infection suppression, as evidenced by the absence of antigen-positive cells for at least two weeks. (E) Schematic overview of the envelope proteins with the identified coding mutations color-coded according to which HVR1-swapped recombinant they were identified in (colours correspond to those shown in panels A-D). Amino acid positions are based on H77 abs. ref. (Genbank #AF009606); HVR1 corresponds to positions 384–410.
Fig 2.
HVR1-swapped H77 recombinants require specific substitutions in E1 and HVR1 of E2 to restore infectivity.
Huh7.5 cells were transfected with in vitro transcribed HCV RNA of the indicated recombinants. Supernatants were collected 24, 48, 72 and 96 hours post transfection, and HCV infectivity titers were determined. At each timepoint, infectivity titers are represented as a mean of three technical replicates. Error bars represent standard deviation. Lower level of quantification was 100 FFU/ml. The substitutions are numbered according to H77 abs. ref. (GenBank #AF009606). The data shown is a representative experiment out of at least 2.
Fig 3.
The substitutions I348N and I348T specifically compensate attenuated H77 HVR1-swapped recombinants.
Huh7.5 cells were transfected with in vitro transcribed HCV RNA of the indicated recombinants. Supernatants were collected 24, 48, 72 and 96 hours post transfection, and HCV infectivity titers were determined. At each timepoint, infectivity titers are represented as a mean of three technical replicates. Error bars represent standard deviation. Lower level of quantification was 100 FFU/ml. The substitutions are numbered according to H77 abs. ref. (GenBank #AF009606).
Fig 4.
Infectivity of H77DH5-HVR1 is restored by introducing the H77 residue at the N-terminal HVR1 positions 386, 388 and 393.
(A-C) Huh7.5 cells were transfected with in vitro transcribed HCV RNA of the indicated recombinants. Supernatants were collected 24, 48, 72 and 96 hours post transfection, and HCV infectivity titers were determined. The infectivity titers of the recombinants are shown at 72 hours post transfection as a mean of three technical replicates (complete data sets are shown in S1 Fig). Error bars represent standard deviation. Lower level of quantification was 100 FFU/ml. The substitutions are numbered according to H77 abs. ref. (GenBank #AF009606). (D) Alignment of HVR1 from HCV genotype 1b isolate DH5 against genotype 1a H77 reference genome (GenBank accession no. AF009606; Mega 7). Dots indicate homology with H77. The data shown is a representative experiment out of at least 2. Positions in HVR1 shown to be responsible for viral attenuation in H77 HVR1 swap recombinants are highlighted in grey.
Fig 5.
E1/E2 substitutions identified in HVR1-swapped H77 recombinants increase sensitivity of H77 to AR3A and AR4A. Neutralization by the bNAbs AR3A (A and C) and AR4A (B and D) of the indicated HCV recombinants. The viruses were incubated in four technical replicates with a 5-fold dilution series of the bNAbs, starting at 50 μg/ml, along with eight technical replicates of virus only. 48 hours post-infection the cells were immunostained for HCV antigen, and the number of FFUs/well was counted and normalized to the mean count of wells with virus only. Each dot represents the mean IC50 value of the indicated recombinants. Error bars represent the 95% confidence intervals (CI95). The colored broken line represents the IC50 value of unmodified H77. The data was analyzed using three-parameter dose-response with the top value set to 100 and lower value set to 0, to calculate IC50 and CI95 using GraphPad Prism v8.0.0.
Fig 6.
J4S52-HVR1 requires specific substitutions in E2 to regain infectivity.
Huh7.5 cells were transfected with in vitro transcribed HCV RNA of the indicated recombinants. Supernatants were collected 24, 48, 72 and 96 hours post transfection and HCV infectivity titers were determined. At each timepoint, infectivity titers are represented by the mean of three technical replicates. Error bars represent standard deviation. Lower level of quantification was 100 FFU/ml. The substitutions are numbered according to H77 abs. ref. (GenBank #AF009606). The data shown is a representative experiment out of at least 2.
Fig 7.
The substitution F447L is adaptive for J4S52-HVR1, while it decreases infectivity of parental J4 and J4ΔHVR1 HCVcc.
Huh7.5 cells were transfected with in vitro transcribed HCV RNA of the indicated recombinants. Supernatants were collected 24, 48, 72 and 96 hours post transfection and HCV infectivity titers were determined. At each timepoint, infectivity titers are represented by the mean of three technical replicates. Error bars represent standard deviation. Lower level of quantification was 100 FFU/ml. The substitutions are numbered according to H77 abs. ref. (GenBank #AF009606).
Fig 8.
Envelope protein substitutions increase sensitivity of J4 HCVcc to AR3A with little effect on AR4A sensitivity.
Neutralization by the bNAbs (A) AR3A and (B) AR4A of the indicated recombinants. The viruses were incubated in four technical replicates with a 5-fold dilution series of the bNAbs, starting at 50 μg/ml, along with eight technical replicates of virus only. 48 hours post-infection the cells were immunostained for HCV antigen, and the number of FFUs/well was counted and normalized to the mean count of wells with virus only. Each dot represents the mean IC50 value of the indicated recombinants. Error bars represent the 95% confidence intervals (CI95). The colored broken line represents the IC50 value of unmodified J4. The data was analyzed using three-parameter dose-response with the top value set to 100 and lower value set to 0, to calculate IC50 and CI95 using GraphPad Prism v8.0.0.
Fig 9.
HVR1 swaps and associated adaptive envelope mutations have divergent effects on antibody binding to extracted E1/E2 protein.
(A-B, D-E, G-H and J-K) Binding curves for dilution series of monoclonal antibodies (mAbs) AR3A and AR4A in ELISA to the indicated extracted E1/E2 proteins. Cell lysates containing E1/E2 protein were added to lectin-coated ELISA plates and incubated overnight. mAbs were added in 5-fold dilution series for a 2-hour incubation. The binding was detected with an HRP-conjugated anti-human IgG secondary antibody. TMB Stabilized Chromagen substrate (Thermo Fisher Scientific) was added, and absorbance was measured at 450 nm. Each data point represents a mean of two technical replicates with error bars representing standard derivation. The data was analysed using three-parameter nonlinear regression using Graphpad Prism v8.0.0. The data shown is a representative experiment out of at least 2. (C, F, I and L) Quantification of E2 protein in collected cell lysates. Equivalent volumes of the cell lysates used in panels A-B, D-E, G-H and J-K were run on SDS-PAGE, and E2 protein was visualized in a fluorescent western blot using the mAb, AP33. β-actin was visualized as a loading control. ΔE1E2 is the negative control plasmid without the HCV envelope protein sequence [59].
Fig 10.
HVR1 swaps and associated adaptive envelope mutations have divergent effects on immunoprecipitation of HCVcc particles.
Immunoprecipitation of H77 (A and B) and J4 (C and D) HCVcc recombinants with the indicated substitutions using AR3A and AR4A (B6 is a control for nonspecific binding). Magnetic beads were coated with the indicated mAbs and incubated with 106 IU HCV RNA from infectious cell culture supernatant containing (A and B) H77, H77I348N/T385A or H77DH5-HVR1, I348N/T385A or (C and D) J4, J4F438V/V710L, J4S52-HVR1, F291I/F438V or J4S52-HVR1, F438V/V710L. The amount of immunoprecipitated HCV RNA was measured by HCV RNA extraction of eluted fractions from the beads and reverse-transcription qPCR in duplicates as described in “HCV immunoprecipitation” in the Methods section. The data are shown as mean of technical duplicates with standard deviation (cut-off was 62 IU). The data shown is a representative experiment out of 2.
Fig 11.
HVR1-swap compensatory E1/E2 substitutions do not affect CD81 entry dependency.
HVR1-swap adaptive substitutions effect on H77 (A and B) or J4 (C and D) dependency on CD81. Huh7.5 cells were incubated with a 5-fold dilution series of anti-CD81 mAb JS-81 that specifically blocks the interaction between HCV and CD81 in four technical replicates with eight technical replicates of virus only and four technical replicates of control antibody. 48 hours post-infection the cells were immunostained for HCV antigen, and the number of FFUs/well was counted and normalized to the mean count of wells with virus only. Error bars represent standard deviation. The data were analysed using three-parameter sigmoid dose-response curves using Graphpad Prism v8.0.0 with bottom constraints set to 0.
Fig 12.
HVR1-swap compensatory E1/E2 substitutions typically decrease virus entry dependency on SR-BI.
HVR1-swap adaptive substitutions effect on H77 (A and B) or J4 (C and D) dependency on SR-BI. Huh7.5 cells were incubated with a 5-fold dilution series of anti-SR-BI mAb, C16-71, which specifically blocks the interaction between HCV and SR-BI in four technical replicates with eight technical replicates of virus only and four technical replicates of control antibody D. 48 hours post-infection the cells were immunostained for HCV antigen, and the number of FFUs/well was counted and normalized to the mean count of wells with virus only. Error bars represent standard deviation. The data were analysed using three-parameter sigmoid dose-response curves using Graphpad Prism v8.0.0 with bottom constraints set to 0.