Hepatitis C virus hypervariable region 1 antibodies interrupt E2-SR-B1 interaction to suppress viral infection

Summary

Whether hypervariable region 1 (HVR1)-targeting antibodies elicited during natural hepatitis C virus (HCV) infection contribute to virus clearance and what is the mechanism underlying remain unclear. Here, we demonstrated that treatment of HCV-infected hepatoma Huh7.5 cells with the IgGs purified from 2 of 28 (7.1%) chronic hepatitis C (CHC) patients efficiently controlled the infection, for which genotype 1b HVR1 (1bHVR1)-binding antibody was critical. Moreover, we found that 1bHVR1 peptide was superior to 2aHVR1 in rabbit immunization to elicit antibodies neutralizing genotypes 1a, 2a, 3a, and 4a. The neutralization effect of 1bHVR1 IgG could be augmented by HH-1, an antibody constructed from CHC memory B cells but without binding to HVR1 peptide. Mechanistic studies showed that 1bHVR1 antisera and IgGs disrupted the interaction of E2-SR-B1 receptor. This study highlights the neutralizing activity of HVR1 antibody elicited by CHC patients and generated by HVR1-immunization against the established infections of multiple HCV genotypes.

Graphical abstract

Highlights

• •HCV-induced humoral response could abrogate an ongoing infection in vitro
• •HVR1-targeting antibodies clear HCV by blocking cell-free and cell-cell transmissions
• •HVR1-targeting antibodies interfere with E2-SR-B1 interaction

Immunology; Virology

Introduction

Hepatitis C virus (HCV) chronically infects approximately 58 million people worldwide, which increases the risk of developing liver cirrhosis and hepatocellular carcinoma (HCC) and has posed major threats to global health.^1,2 Despite the clinical use of direct-acting antivirals (DAAs) has greatly improved the cure rate of hepatitis C, several major challenges remain, including no HCV vaccine, a large number of patients without being diagnosed, limited or unavailable DAA therapy in underdeveloped countries and regions, the risk of HCC and reinfection, and the emergence of drug resistance, etc.^3,4 Although World Health Organization (WHO) has called for the global elimination of viral hepatitis by 2030 on the basis of highly effective DAA regimens,⁵ it is a consensus that a prophylactic vaccine is indispensable for the control and prevention of HCV infection. Unfortunately, a T cell vaccine based on the priming with a chimpanzee adenovirus expressing HCV nonstructural proteins (ChAd3-NS) and the boost of modified vaccinia (strain Ankara) had no protection, even though it induced a broad HCV-specific CD4⁺ and CD8⁺ T cell responses.^6,7 To date, experimental evidence has convinced that a prophylactic HCV vaccine need to be able to induce broad and potent neutralizing antibodies (nAbs).

About 25% of HCV natural infections are cleared spontaneously in acute phase, and 75% of infections develop chronic hepatitis C (CHC).⁸ Accumulating evidence supports that, apart from T cell immunity,⁹ nAbs are positively involved in HCV clearance, although they become detectable typically at 8–12 weeks post infection.^10,11,12 Moreover, induction of potent nAbs at an early phase of infection is associated with spontaneous resolution of acute infection, whereas a delayed or lack of antibody response has been related to the occurrence of CHC.^13,14,15,16 Spontaneous clearance of HCV has also been reported in CHC patients and correlated with the presence of cross-reacting and potent nAbs.¹⁷ Besides, monoclonal antibodies (mAbs), AR3A, AR3B, and AR4A, have been reported to abrogate an ongoing HCV infection (J6^{5’UTR−NS2}/JFH1 recombinant) in human primary hepatocytes and a mouse model populated with human hepatocytes,¹⁸ thus indicating a therapeutic effect of nAbs. However, CHC patient serum contains both polyclonal nAbs and other antibodies without neutralizing activity, which potentially interfere with the potency of nAbs. As a consequence, the infection may progress with the existence of a bulk of antibodies in chronic phase.^19,20,21 In addition, HCV has evolved multiple strategies to evade nAbs, such as cell-cell transmission,²² high genetic diversity and quasispecies,²³ and glycan shielding of glycoproteins.²⁴ To date, it is unclear whether and which antibody from CHC patients could neutralize an ongoing HCV infection and what is the mechanism underlying.

HCV is an enveloped, positive-stranded RNA virus of the Flaviviridae family with eight major genotypes and multiple subtypes.^25,26,27 Two viral envelope proteins E1 and E2 form heterodimer E1E2 with conformational dynamic states that affect receptor binding and nAbs accessibility, thus playing critical roles in virus entry and antibody neutralization.^28,29 The E2 region is highly diverse especially the hypervariable region 1 (HVR1, amino acids [aa] 384–410). The HVR1 is immunodominant in both natural infection and immunization trials involved in temperature-dependent neutralization enhancement and modulates nAb sensitivity by influencing E1E2 conformational dynamics.^30,31,32 Although HVR1-targeting nAbs have been shown to protect chimpanzees from homologous HCV challenge,³³ they are strain-specific without protection from heterologous infection and instead, they drive the evolution of viral quasispecies as a consequence of immune pressure in CHC patients.^33,34,35,36 Besides, antibodies targeting HVR1 may interfere with nAbs binding and its recognition to adjacent epitopes, such as conserved E2 domain aa412-423.¹⁹ In contrast, HVR1-binding mAb HEPC98 could primarily block E2 binding to scavenger receptor B1 (SR-B1) and heparan sulfate, and its combination with mAb HEPC74 that blocks E2-CD81 binding could synergize and broaden the neutralization against HCV,³⁷ indicating that HVR1-specific antibody could be advantageous. A study also showed that HVR1-deleted E1E2 was not superior to wild-type E1E2 in eliciting nAbs.³⁸ Therefore, the role of HVR1 in protective humoral response seems more complicated than previously thought and warrants further investigation.

In this study, we found that some nAbs from a minority of CHC patients could efficiently abrogate an ongoing HCV infection in hepatoma Huh7.5 cells. HVR1-specific antibodies induced in natural HCV infection or by HVR1 peptide immunization played a pivotal role in inhibiting HCV infection, to which interruption of E2 and SR-B1 interaction had a contribution.

Results

A proportion of CHC patient-derived IgGs inhibited HCV infection and transmission in Huh7.5 cells

To investigate antibody response following HCV infection, we recruited 28 CHC patients with or without DAA therapy in this study. These patients were infected with HCV genotypes 1b (10 out of 28, 10/28, 35.7%), 3a (5/28, 17.9%), 3b (3/28, 10.7%), 6a (5/28, 17.9%), and unknown genotype (5/28, 17.9%) (Table S1). We isolated IgG from the sera of all CHC patients and tested its neutralizing efficacy against HCVcc of genotypes 1a, 2a, 3a, 4a, 5a, and 6a. To this end, the viruses were incubated with IgG (100 μg/mL) before virus infection of Huh7.5 cells. The results showed that incubation of IgGs from 22 CHC patients (Pts.1–22) reduced the infection of three HCV genotypes by >50%, whereas none of the IgGs from Pts. 23–28 inhibited any HCV genotype (<50% reduction of infection) (Figure 1A).

Figure 1.

A proportion of patients-derived nAbs blocked the established HCV infection and transmission pathways

(A) Serum IgG was purified from 28 CHC patients and their neutralization activities against HCVcc of genotypes 1–6 were evaluated at a final concentration of 100 μg/mL. The neutralization effect (percentage of neutralization) relative to normal human IgG control was presented in the heatmap. The details of CHC patients are shown in Table S1.

(B) Treatment of established HCV infection with IgG. Huh7.5 cells in 48-well plates (2.4×10⁴ cells) were infected with 400 FFUs of HCVcc. Patient IgG (100 μg/mL) was diluted in complete culture medium and added on days 2, 5, and 8 post infection (dpi), and the percentage of HCV-positive cells relative to untreated mock control was determined at 9 dpi by an immunofluorescence assay.

(C) IgG treatment (100 μg/mL) of 2a recombinant J6^{5’UTR−NS2}/JFH1. At 9 dpi, the percentage of HCV-positive cells of each group normalized to mock control was shown by mean ± standard error of the mean (SEM) from three independent experiments.

(D) IgG treatment of HCVcc recombinants TNcc(1a), J6cc(2a), S52(3a), ED43(4a), SA13(5a), and HK6a(6a). At 9 dpi, the percentage of HCV-positive cells of each group normalized to mock control was determined.

(E) Effect of patient IgGs on HCV cell-free transmission. Naive Huh7.5-NIrD cells (target cells, 2.5×10⁵ cells) and J6^{5’UTR−NS2}/JFH1-infected Huh7.5 cells (MOI = 0.01, producer cells) were seeded in six-well plates on either side of a transwell insert, and then patient IgG (100 μg/mL) was added to producer cells. At 2 dpi, the percentage of HCV-positive cells in target cells was examined relative to mock control. nAb AP33 was included as a positive control. The mean ± SEM of three independent experiments is shown.

(F) Effect of patient IgGs on HCV cell-cell transmission. J6^{5’UTR−NS2}/JFH1-infected Huh7.5 cells (MOI = 0.01, producer cells) and naive Huh7.5-NIrD cells (target cells) were mixed (1.25×10⁵ cells for each) and seeded in 12-well plates, and AP33 (25 μg/mL) was added to block cell-free transmission. At 2 dpi, mCherry-positive Huh7.5-NIrD target cells were counted, relative to mock control. The mean ± SEM of three independent experiments is shown.

(G) Representative images of target Huh7.5-NIrD cells at 2 dpi in cell-cell transmission assay. Bar, 100 μm. In panels C-G, CTRL IgG was normal human IgG control.

Next, we tested whether patient IgG could block an established HCV infection. To this end, Huh7.5 cells were infected with genotype 2a recombinant J6^{5’UTR−NS2}/JFH1 for 2 days, patient IgGs (100 μg/mL) were added at 2, 5, and 8 dpi, and HCV infection was assessed at 9 dpi using an immunofluorescence assay (Figure 1B). The results showed that IgGs from Pt-3 and Pt-10 efficiently blocked the established HCV infection in Huh7.5 cells, whereas other patient IgGs had no inhibition to the established infection (Figure 1C). We further tested the potency of Pt-3 and Pt-10 IgGs for other genotype viruses and found that both IgGs blocked the established infections of TNcc(1a), J6cc(2a), S52(3a), ED43(4a), SA13(5a), and HK6a(6a) (Figure 1D); other IgGs, selected from Pt-7 and Pt-23 and a healthy subject, had no inhibition to these genotype viruses. Noting that, Pt-3 received sofosbuvir plus daclatasvir treatment, whereas Pt-10 were not treated, and both of them were cured; in contrast, other patients, e.g., Pt-23, remained uncured after a follow-up period.

It is known that HCV infects hepatocytes through two pathways, cell-free and cell-cell transmissions. We further investigated which pathway was affected by IgG treatment using transmission assays. IgGs from Pt-3, Pt-7, and Pt-10 potently inhibited the cell-free transmission of J6^{5’UTR−NS2}/JFH1 comparable to AP33 control antibody, a well-characterized nAb,³⁹ whereas Pt-23 IgG had no inhibition effect (Figure 1E). In cell-cell transmission assay using NIrD cells as target cells,⁴⁰ we demonstrated that IgGs from Pt-3 and Pt-10 efficiently inhibited cell-cell transmission of J6^{5’UTR−NS2}/JFH1, but Pt-7 and Pt-23 IgGs showed no inhibition effect (Figure 1F and 1G). Collectively, these data suggest that nAbs were prevalent (22/28, 78.6%) in CHC patients, of which a proportion of IgGs could block an established HCV infection and viral transmission pathways.

CHC patient-derived HVR1-binding IgGs contributed to the inhibition of established HCV infection in cultured cells

To investigate the binding epitopes of patient IgGs, we first tested HVR1 region, which contains neutralizing epitopes and is involved in viral evasion from protective humoral response.^41,42 Since genotype 1b accounted for 35.7% (10/28) of the recruited CHC patients and is the most prevalent genotype worldwide,²⁶ we synthesized a genotype 1b HVR1 peptide (1bHVR1) identical to NC1 strain aa384-410 (GenBank accession no. CAB53095.1)⁴³ and tested its binding activity with patient IgGs by ELISA (Figure 2A and 2B). We found that 23 of 28 (82.1%) patient IgGs showed binding activity with 1bHVR1 peptide, including most IgGs from patients of genotype 3a, 3b, and 6a (Figure 2B). These results suggest that CHC patients elicited IgGs with cross-reactivity to other genotype HCVs, in line with previous reports.^44,45,46

Figure 2.

CHC patient-derived HVR1-binding IgGs contributed to the inhibition of HCV infection

(A) HVR1 sequences of genotypes 1–6. The HVR1s of genotype 1b (CAB53095.1)⁴³ and J6 (2a) were also used for immunization in this study (see panel 2B and Figure 3A below).

(B) Binding of patient IgGs (100 μg/mL) with genotype 1b HVR1 peptide (1bHVR1) in ELISA. HVR1 antibody (anti-1bHVR1) was generated by New Zealand rabbits immunized with 1bHVR1 peptide (Figure 3A below). CTRL IgG, normal human IgG control. A broken line indicates the cutoff value of ELISA (OD₄₅₀ nm = 0.2). The mean ± SEM of three independent experiments is shown.

(C) Depletion of HVR1-antibody with 1bHVR1 peptide reduced the inhibition effect of Pt-3 and Pt-10 IgGs on HCV infection (100 μg/mL). CTRL IgG, normal human IgG. Anti-SR-B1 and anti-CD81 antibodies (100 μg/mL) were included as controls. The mean ± SEM of three independent experiments is shown. The difference between two groups was determined by Student’s unpaired t-test; ∗∗, p< 0.01; ∗∗∗, p< 0.001.

Next, we investigated whether 1bHVR1-binding antibody played a role in the inhibition of HCV infection. As Pt-3 and Pt-10 inhibited the established HCV infection (Figure 1D), we depleted the 1bHVR1-binding antibody in these two IgGs by incubating them in 1bHVR1-coated ELISA plates (500 ng/well) for 30 min.²¹ Depletion of HVR1-binding antibody reduced the potency of Pt-3 and Pt-10 IgGs against HCV infection, as demonstrated by an increased infection rate after treatment with depleted-IgG compared to respective original IgG (depleted-IgG versus original IgG: Pt-3, 25.1% vs. 3.1%, p<0.01; Pt-10, 32.1% vs. 5.6%, p<0.001) (Figure 2C). However, the inhibition rate of Pt-3 and Pt-10 remained 67.9% and 74.9%, respectively, which indicates that some non-HVR1-binding IgGs also contributed to viral inhibition. The control antibodies anti-SR-B1 and anti-CD81 (25 μg/mL) inhibited HCV infection by 47.1% and 20.4% of infections, respectively (Figure 2C). Together, these results suggest that HVR1-binding IgGs contributed to the inhibition of an established HCV infection.

IgGs generated by 1bHVR1 peptide immunization blocked the established HCV infection

Since HVR1-specific antibody contributed to HCV clearance (Figure 2C), and HVR1 contains nAb epitopes and is directly involved in protecting diverse antibody-specific epitopes on E1 and E2.⁴¹ Immunization of a soluble E2 (aa384-661 of 1b strain Con1) enables to elicit pan-genotypic neutralizing antibodies and intrahepatic T cell responses in mice and nonhuman primates.^47,48 To further study whether 1bHVR1-specific antibody block HCV infection, we generated anti-1bHVR1 and anti-2aHVR1 antisera by immunization of New Zealand rabbits with KLH-conjugated 1bHVR1 or 2aHVR1 peptides using aluminum hydroxide adjuvant (Figures 2A and 3A); both 1b and 2a represent the genotypes prevalent worldwide.²⁶ The titers of anti-1bHVR1 and anti-2aHVR1 antisera increased gradually and reached a plateau at week 6 and maintained this level to week 9, when the rabbits were euthanized (Figure 3A). Anti-1bHVR1-1 antiserum (1:20 dilution) efficiently neutralized TNcc(1a), J6^{5’UTR−NS2}/JFH1(2a), S52(3a), and ED43(4a), marginally affected HK6a(6a), and had no neutralization for SA13(5a). Anti-1bHVR1-2 antiserum (1:20 dilution) neutralized efficiently genotypes 2a, 3a, and 4a, but to a less extent for genotype 1a, and showed no neutralization for SA13(5a) and HK6a(6a). None of the anti-2aHVR1 antisera had neutralizing effect for any of the genotype viruses tested (Figure 3B). The levels of 50% neutralizing titer (NT50) of these antisera were determined by the neutralization effect of serially diluted antisera (Figure 3C). The results showed that anti-1bHVR1-1 and -2 antisera potently neutralized HCVcc of genotypes 1–4 (NT50 ≥ 80) and could not neutralize SA13(5a) and HK6a(6a) (NT50 ≤ 5). Anti-2aHVR1 antisera neutralized genotypes 4a and 6a weakly (NT50 = 20) and hardly affected other genotype viruses (NT50 ≤ 5).

Figure 3.

Antisera and IgGs generated by 1bHVR1 peptide-immunized rabbits inhibited the infections of HCV genotypes 1–4

(A) Titers of anti-1bHVR1 and anti-2aHVR1 antisera following immunization of HVR1 peptides. Two New Zealand rabbits (no. 1 and 2) were immunized with 500 μg of 1bHVR1 or 2aHVR1 peptides per rabbit at weeks 0 and then boosted with 250 μg peptide per rabbit at weeks 2, 4, and 5 (indicated by arrowheads). Rabies virus (RV) peptide, KLH protein, and mock controls were included. The endpoint titers of antisera from each rabbit are shown. A broken line indicates baseline antiserum titer (1.69, log₁₀).

(B) Neutralization of rabbit antisera against HCVcc of genotypes 1–6. The antisera collected from immunized rabbits (panel A) were tested for their neutralizing activity against HCVcc (1:20 dilution; Materials and Methods). The mean ± SEM of two independent experiments is shown.

(C) Half neutralizing titers (NT50) of rabbit antisera against HCVcc of genotypes 1–6. Antisera were diluted into a 2-fold serial dilution starting from 1:5 and tested its neutralization effect on HCV genotypes 1–6. NT50 was defined as the highest dilution of antisera neutralizing 50% of HCV infection.

(D) Neutralization effect of IgGs generated from immunized rabbits. IgGs were purified from anti-1bHVR1 and anti-2aHVR1 antisera (panel A) and diluted into 2-fold serial dilutions (ranging from 3200 μg/mL to 1.56 μg/mL), and their neutralization activities against HCVcc of genotypes 1–6 were evaluated. Neutralization curves were plotted by nonlinear regression and IC50 of IgGs were calculated for each virus that was neutralized.

(E) Anti-1bHVR1-IgG-1 blocked the established HCV infections of genotypes 1–4. Anti-1bHVR1-IgG-1 (100 μg/mL) purified from anti-1bHVR1 antisera and tested for its neutralizing activity against the infections of HCV genotypes 1–6. Control, normal rabbit IgG. The mean ± SEM of two independent experiments is shown.

To exclude the possibility that other factors in antisera may have contributed to the neutralization of HCVcc infection, we purified IgGs from rabbit antisera and determined their neutralizing activity by serial dilutions. The results showed that anti-1bHVR1-IgG-1 and -2 potently neutralized TNcc(1a) (IC50s, 50.39 and 252.9 μg/mL, respectively), J6^{5’UTR−NS2}/JFH1(2a) (IC50s, 23.91 and 48.84 μg/mL), S52(3a) (IC50s, 43.14 and 35.11 μg/mL), and ED43(4a) (IC50s, 25.09 and 9.847 μg/mL). However, these two IgGs poorly neutralized HK6a(6a) and could not neutralize SA13(5a) (Figure 3D). Anti-2aHVR1-IgG-1 and -2 neutralized weakly S52(3a), ED43(4a), and HK6a(6a). Taken together, 1bHVR1 peptide was superior to 2aHVR1 peptide in eliciting IgG with cross-neutralizing activity for other HCV genotypes.

Next, we tested the activity of anti-1bHVR1-IgG-1 in blocking the established HCV infection (Figure 3E) and found that it blocked the established infection of HCV genotypes 1–4, reducing the infection by 70% for TNcc(1a), 79% for J6^{5’UTR−NS2}/JFH1(2a), 76% for S52(3a), and 84% for ED43(4a), but without inhibition for SA13(5a) and HK6a(6a). These results suggest that anti-1bHVR1-IgG-1 inhibited the established infections of multiple HCV genotypes.

Monoclonal antibody constructed from CHC memory B cells augmented the neutralization effect of anti-1bHVR1 antibody

We also constructed a monoclonal antibody, designated HH-1, from CD19⁺IgM^–IgA^–IgD^– memory B cells of Pt-3 that have been co-cultured with stimulatory factors for 13 days using the protocol previously described.⁴⁹ The third complementarity-determining region (CDR3) sequences of HH-1 (“ARDRAVGPRDYEY” for heavy chain and “SSYSHTSDFDYV” for light chain) were determined (Figure 4A), and HH-1 expression was confirmed with silver staining method (Figure 4B). Neutralization experiment revealed that mAb HH-1 efficiently neutralized TNcc(1a) (IC50, 24.5 μg/mL), J6^{5’UTR−NS2}/JFH1(2a) (IC50, 27.1 μg/mL), S52(3a) (IC50, 12.2 μg/mL), and SA13(5a) (IC50, 35.1 μg/mL), but could not inhibit ED43(4a) or HK6a(6a) (Figure 4C). However, HH-1 did not bind to 1bHVR1 and 2aHVR1 peptides (Figure 4D), indicating that HH-1 and anti-1bHVR1-IgGs did not share the same epitope. Furthermore, HH-1 alone (100 μg/mL) could not block the established infection of any HCV genotypes (Figure 4E). However, HH-1 augmented the inhibition effect of anti-1bHVR1-IgG-1 against TNcc(1a), J6^{5’UTR−NS2}/JFH1(2a), and S52(3a), but had no augment effect on 4a, 5a, and 6a viruses (Figure 4E). Together, these results suggest that the neutralization effect of HVR1 antibody could be augmented by mAbs targeting non-HVR1 epitope.

Figure 4.

Monoclonal nAb HH-1 constructed from CHC patient augmented inhibition effect of anti-1bHVR1 IgG

(A) CDR3 regions of heavy and light variable chains of monoclonal nAbs HH-1 are shown, together with their family names and similarity to the corresponding germline ancestor sequences.

(B) Silver staining of PAGE-separated HH-1 that was expressed in 293T cells in serum-free medium.

(C) HH-1 neutralized HCVcc of genotypes 1a, 2a, 3a, and 5a. HH-1 was two-fold serially diluted from 400 μg/mL to 0.2 μg/mL and tested its neutralization activity against HCVcc of genotypes 1–6. Neutralization curves were plotted by nonlinear regression and IC50 was calculated.

(D) HH-1 did not bind to HVR1 peptides of genotypes 1b and 2a. ELISA was performed using HH-1 (100 μg/mL) and 1bHVR1 and 2aHVR1 peptides. Control, normal rabbit IgG. Anti-1bHVR1-IgG-1 and anti-2aHVR1-IgG-1 were included as positive controls. A broken line indicates the cutoff value (OD₄₅₀ nm = 0.2).

(E) Augment effect of HH-1 (100 μg/mL) on the neutralization of anti-1bHVR1-IgG-1 for HCV genotypes 1a, 2a, and 3a.

Anti-1bHVR1 antiserum and IgG interrupted E2-SR-B1 interaction

Previous studies suggested HCV evades neutralization mainly via cell-cell transmission, which is highly dependent on SR-B1 receptor,²² and HVR1 domain interacts with SR-B1 receptor.⁵⁰ To investigate the mechanism by which anti-1bHVR1 antiserum and IgG inhibited an established HCV infection, we examined whether anti-1bHVR1 antibody interferes with the binding of E2 with SR-B1 receptor. To this end, we expressed E2-Flag protein in 293T cells and examined its interaction with cell SR-B1 by IP assay (Figure 5A). The results showed that anti-1bHVR1-1 antiserum (1:20) and anti-1bHVR1-IgG-1 (100 μg/mL) interrupted the binding between SR-B1 and E2-Flag protein, which was detected in normal rabbit IgG and mock controls (Figure 5A). Interruption of E2-SR-B1 interaction may explain its inhibitory effect on HCV cell-cell transmission (Figure 1). In contrast, HH-1 (50 μg/mL and 100 μg/mL) did not disrupt E2-SR-B1 interaction (Figure 5B), supporting the result that HH-1 hardly inhibited the established HCV infection (Figure 4E). These results suggest that nAbs targeting HVR1 disrupted E2-SR-B1 interaction, which may contribute to the blocking activity against an established HCV infection.

Figure 5.

Anti-1bHVR1 antiserum and IgG interfered with E2-SR-B1 interaction

(A) 1bHVR1-IgG-1 or antisera disrupted E2-SR-B1 interaction. Huh7.5 cells (5×10⁶ cells) were incubated with 1 mL of E2-Flag from transfection supernatant at 37 °C for 1.5 h together with normal rabbit IgG, 1bHVR1-IgG-1 (100 μg/mL) or antisera (1:20 dilution), or culture medium (mock). The mixtures were washed with PBS and lysed with IP lysis buffer, 1/10 of the sample was loaded as “input”, and the remaining was incubated with anti-Flag magnetic beads at 4 °C overnight. The immunoprecipitated part was detected by western blot with an anti-SR-B1 antibody.

(B) HH-1 IgG did not block E2-SR-B1 interaction. Huh7.5 cells were incubated with E2-Flag in the presence of HH-1 (50 or 100 μg/mL) and evaluated by IP assay as described above.

Discussion

Previous studies have revealed a remarkable association between viral clearance and potent induction of broad nAbs in acute hepatitis C or CHC patients,^{13,14,15,16,17} but direct evidence supporting the abrogation effect of patient nAbs is lacking. Here, we showed that IgGs from 2 of 28 CHC patients (7.1%) efficiently inhibited the established HCV infections of six major genotypes in Huh7.5 cells (Figure 1C and 1D) as well as cell-cell transmission (Figure 1F), a major pathway for HCV evasion from antibody neutralization. Unlike primary hepatocytes, Huh7.5 hepatoma cells are impaired in interferon signaling pathway,^51,52 thus the viral inhibition could mainly be attributed to the function of nAbs. During natural HCV infection, the nAbs with inhibition effect may collaborate with protective T cell immunity to accelerate viral clearance. Our data also showed that even though broad nAbs were prevalent in CHC patients (22/28, 78.6%), most of them had little effect on the established HCV infection, as the nAbs from the remaining 20 CHC patients did not inhibit HCV infections. These data support the notion that only a small proportion of acute hepatitis C patients (∼25%) could spontaneously clear HCV.⁸ It should be noted that analyzing nAbs from the patients who naturally clear the infection may be more meaningful for the study of nAbs elicited during HCV infection. However, we were unable to collect a group of patients with spontaneous HCV clearance. Therefore, we analyzed the CHC patients available for this study, with only one self-resolved patient (Pt 10).

We found that 23 of 28 (82.1%) patient IgGs had binding with 1bHVR1 peptide, regardless of patient genotypes (Figure 2B), suggesting the immune-dominance and the cross-reactivity characteristics of this epitope, consistent with previous reports.^41,44,45,46 Notably, we only evaluated the antibody reactivity to 1bHVR1 peptide, without excluding the existence of antibodies specific for other HVR1 sequences. HVR1 antibodies are believed to be strain-specific and interfere with antibody neutralization; however, we found that the HVR1 nAbs contained in Pt-3 and Pt-10 IgGs played an important role in inhibiting HCV infection, as 1bHVR1 peptide-depleted IgGs lost the capacity of inhibiting HCV infection to some extent (Figure 2C). Thus, the presence of HVR1 antibodies may not account for the neutralization failure in these patients.

By generation of rabbit anti-1bHVR1 and anti-2aHVR1 antisera, we found that 1bHVR1 peptide was superior to 2aHVR1 peptide in eliciting neutralizing antibodies, as the former induced nAbs with potent neutralizing activity against HCVcc of genotypes 1a, 2a, 3a, and 4a (Figure 3D). Anti-1bHVR1-1 efficiently inhibited the infections of 1a, 2a, 3a, and 4a in Huh7.5 cells at 9 dpi (Figure 3E), providing direct evidence that HVR1 nAbs contributed to the inhibition of ongoing HCV infection, consistent with the results for CHC patients (Figure 2C). Although anti-2aHVR1 antisera recognized 2aHVR1 peptides (Figure 3A), anti-2aHVR1-IgGs conferred very weak protection against HCV infection, even for homologous 2a virus (Figure 3D). Recently, an 1b (strain Con1) E2 was found to be able to elicit broadly neutralizing antibodies and intrahepatic T cell response in animals,^47,48,53 and 1b Con1cc has just been developed,⁵⁴ it would be interesting to include a 1b HCVcc for neutralizing study in the future. Although the distinct neutralizing efficacy of anti-1bHVR1 and anti-2aHVR1 remain unclear, previous studies and our results have suggested that 1b envelope antigens may possess properties advantageous for eliciting broad immune responses.

In this study, we found that the inhibition effect of HVR1-specific antibodies could be augmented by non-HVR1-binding mAb (Figure 4E). mAb HH-1 was constructed from B cells of CHC patient Pt-3, and it potently neutralized TNcc(1a), J6^{5’UTR−NS2}/JFH1(2a), S52(3a), and SA13(5a) in a dose-dependent manner. HH-1 did not bind to 1bHVR1 or 2aHVR1 peptide (Figure 4D) and did not affect the established HCV infection (Figure 4E), suggesting that it recognized epitopes outside these HVR1 sequences. However, HH-1 could augment the inhibition effect of anti-1bHVR1-IgG-1 on genotypes 1a, 2a, and 3a (Figure 4E). Collectively, this observation provides a possibility that a combination of HVR1 nAbs with other nAbs or mAbs may achieve a synergistic effect or confer a neutralizing activity to HVR1 antibodies.

HVR1 is reported to be involved in the interaction between E2 and SR-B1, a key host receptor for HCV cell-free entry and cell-cell transmission, respectively.^{22,41,55,56,57} HCV recombinants with HVR1 deletion were less dependent on SR-B1 and low-density lipoprotein receptors⁵⁸ and HVR1 is also important in genotype differences in neutralization sensitivity.^59,60 In addition, a recent study reported that the HVR-1 safety catch controls the efficiency of virus entry and maintains resistance to nAbs.⁶¹ Here, we demonstrated that the antisera or IgGs generated by 1bHVR1 immunization disrupted the E2-SR-B1 interaction (Figure 5A), whereas monoclonal IgG HH-1 recognizing other epitopes did not block the interaction (Figure 5B).

In summary, we have demonstrated that a minority of IgGs induced during natural HCV infection contributes to the control of an established HCV infection, potentially through interrupting E2-SR-B1 interaction.

Limitations of the study

Although HVR1-targeting nAbs contributed to HCV clearance, it was identified only in a minority of CHC patients. In this study, a limited number of patients were recruited and analyzed, thus we could not conclude some generalized characteristics of the patients who elicited HVR1-specific nAbs. However, it is interesting to screen a greater number of CHC patients, which may be helpful to reach a conclusion. Moreover, construction and characterization of more monoclonal HVR1 nAbs from CHC patients would allow a comprehensive study of their roles in controlling an ongoing HCV infection. Besides, we did not know more about the non-HVR1-binding IgGs that also contributed to viral inhibition.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
HCV Core mAb C7-50	Santa Cruz Biotechnology	Cat# sc-57800; RRID: AB_783742
Goat anti-Mouse IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 488	Thermo Fisher Scientific	Cat# A-11029; RRID: AB_2534088
Scavenging Receptor SR-BI antibody	Abcam	Cat# ab217318
CD81 antibody	Proteintech	Cat# 27855-1-AP; RRID: AB_2880995
FLAG tag (MA4) Mouse Monoclonal Antibody	Ray Antibody Biotech	Cat# RM1002
Human IgG Iso-type Control	Thermo Fisher Scientific	Cat# 02–7102; RRID: AB_2532958
Rabbit IgG control Polyclonal antibody	Proteintech	Cat# 30000-0-AP; RRID: AB_2819035
Goat anti-Rabbit IgG-HRP	Ray Antibody Biotech	Cat# RM3002
Goat anti-Mouse IgG-HRP	Ray Antibody Biotech	Cat# RM3001
Mouse anti-Human IgG Fc Secondary Antibody, HRP	Thermo Fisher Scientific	Cat# 05–4220; RRID: AB_2532922
Bacterial and virus strains
TNcc (1a), GenBank: JX993348	Li et al.62	N/A
J65’UTR−NS2/JFH1 (2a), GenBank: JF343782	Li et al.63	N/A
J8cc (2b), GenBank: JQ745652	Li et al.64	N/A
S52 (3a), GenBank: KF589884	Li et al.65	N/A
ED43 (4a), GenBank: KF589885	Li et al.65	N/A
SA13 (5a), GenBank: KF589886	Li et al.65	N/A
HK6a (6a), GenBank: KF589887	Li et al.65	N/A
Chemicals, peptides, and recombinant proteins
Doxycycline	MedChemExpress (MCE)	HY-N0565
cytokine IL-2	Absin	abs04045
cytokine IL-21	Absin	abs04900
HVR1 peptides of genotype 1b, GenBank: CAB53095.1	Koch, JO et al.43	N-GTYVTGGTMAKNTLGITSL FSPGSSQK-C
HVR1 peptides of genotype 2a, GenBank: AEB71624.1	Li et al.63	N-RTHTVGGSAAQTTGRLTSL FDMGPRQK-C
control peptide	Dietzschold, B et al.66	N-VNLHDFRSDEIE-C
Critical commercial assays
BCA Protein Assay Kit	GenStar	E162-05
Magne™ Protein G Beads	Promega	G747A
ECL Western Blotting Substrate	Pierce	WBULS0100
Experimental models: Cell lines
Huh7.5 cells	Li et al.54	N/A
293T cells	Hao et al.67	N/A
Huh7.5 derived-NIrD cells	Ren et al.40	N/A
Experimental models: Organisms/strains
New Zealand rabbits	GL Biochem	N/A
Oligonucleotides
forward primer for Flag-E2: CCGAAGCTTGCCGCCAC CATGATG CAGGGAGCGTGGGCG	This paper	N/A
reverse primer for Flag-E2: TGCTCTAGACTACTTATC GTCGTCATCCTTGTAATC TTGACTTCTGTCTCTGTC	This paper	N/A
Software and algorithms
GraphPad Prism 8	GraphPad Software	https://www.graphpad.com
Other
CDR3 sequences of HH-1	This paper	Data Figure 4A

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Yi-Ping Li (lyiping@mail.sysu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Experimental model and subject details

Human subjects and ethics statement

The serum samples of CHC patients were obtained from the Fifth Affiliated Hospital, Sun Yat-Sen University, according to national guidelines with the consent of patients and approved by the Ethics Committee of Fifth Affiliated Hospital, Sun Yat-sen University (approval number [2016] K96-1). The age, sex and HCV genotype infected of these CHC patients were provided (Table S1).

All rabbit immunization studies were approved by the Institutional Animal Care and Use Committee of Guangzhou Eighth People’s Hospital and conducted according to institutional guidelines.

In vivo rabbit studies

Eight female 12-week-old New Zealand white rabbits were used for this work, and two rabbits were allocated for each group.

Cells and antibodies

Huh7.5 cells and 293T cells were propagated in Dulbecco’s modified Eagle medium (DMEM, Thermo Fisher, USA) supplemented with 10% fetal bovine serum (Gibco, USA) and 1% nonessential amino acids. Huh7.5 derived-NIrD cells (NS3-4A Inducible rtTA-mediated Dual-reporter) were provided by Prof. Wensheng Wei (Peking University, China).⁴⁰ Upon infection, HCV NS3-4A protease cleaves NIrD rtTA-MAVs on mitochondria and releases rtTA, which enters the nucleus and activates a delta-TK-2A-mCherry expression in the presence of doxycycline (2 μg/mL). An anti-HCV Core mAb C7-50 (sc-57800, Santa Cruz, USA) and goat anti-mouse IgG (H + L) Alexa Fluor 488 (A-11029, Thermo Fisher) were used in the immunofluorescence assay. Other antibodies included anti-SRB1 (ab217318, Abcam, UK), anti-CD81 (27855-1-AP, Proteintech, China), anti-Flag (RM1002, Ray Antibody Biotech, China), normal human IgG (02–7102, Thermo Fisher), normal rabbit IgG (30000-0-AP, Proteintech), goat anti-rabbit IgG-HRP (RM3002, Ray Antibody Biotech), goat anti-mouse IgG-HRP (RM3001, Ray Antibody Biotech), and mouse anti-human IgG-HRP (05–4220, Thermo Fisher). To avoid the potential adverse effect of solvent sodium azide on cell experiment, the commercial antibodies were purified with Magne™ Protein G Beads (G747A, Promega, USA) and dissolved in 2M Tris-HCl buffer (pH 7.5), and the antibody concentration was measured with a BCA kit (E162-05, GenStar, China).

Peptides

HVR1 peptides of genotype 1b (N-GTYVTGGTMAKNTLGITSLFSPGSSQK-C, CAB53095.1)⁴³ and 2a (N-RTHTVGGSAAQTTGRLTSLFDMGPRQK-C, AEB71624.1) and a control peptide (N-VNLHDFRSDEIE-C) from rabies virus (RV) glycoprotein⁶⁶ were synthesized (GL Biochem, China). An extra cysteine residue was added to the C terminal of peptides for conjugation with keyhole limpet hemocyanin (KLH) protein. The purity of peptides was more than 98%, and the sequences were confirmed using mass spectrometry.

Method details

Enzyme-linked immunosorbent assay (ELISA) and affinity depletion

The reactivity of patient sera with HVR1 peptides was evaluated by an indirect ELISA.⁶⁸ Briefly, streptavidin-coated 96-well plates (Pierce, USA) were incubated with 1 μg of biotinylated HVR1 peptide at 4 °C overnight, after washing with phosphate-buffered saline (PBS, pH 7.4), patient IgG (100 μg/mL) was added and incubated at 37 °C for 1 h. After washing three times with PBS, a mouse anti-human IgG-HRP antibody was added (1:1000 dilution) and the optical density value at 450 nm (OD₄₅₀) of each well was taken by a plate reader (TECAN, Switzerland).

For affinity depletion of HVR1-specific antibodies in patient IgG, the IgG was purified from patient serum with Magne™ Protein G Beads (G747A, Promega). ELISA plate was coated with genotype 1b (or 2a) HVR1 peptide (500 ng/well) in Tris-HCl buffer (pH 7.5) containing 0.02% Tween20. After incubation for 30minat room temperature and washing with PBS, a diluted patient IgG was added and left for absorption for 30 min. The depletion course was monitored with ELISA analysis and the unbound IgG was collected.

Immunization of rabbits with HVR1 peptides to generate antibodies

Eight female 12-week-old New Zealand white rabbits were purchased (GL Biochem), and two rabbits were allocated for each group. HVR1 peptides of HCV genotype 1b and 2a and RV control peptide were conjugated with KLH protein with a conjugation kit (Sigma, USA). KLH-conjugated peptides (500 μg) were mixed with the same volume of aluminum hydroxide adjuvant (InvivoGen, USA) and were subcutaneously immunized to multiple sites. KLH protein alone plus adjuvant was also injected as a mock control. Boost injections were done with the same peptides (250 μg) at weeks 2, 4, and 5. The antibody titers elicited were periodically evaluated by ELISA during immunization course. All rabbits were euthanized at week 9 and the antisera were collected for neutralization experiments.

Construction of HH-1 monoclonal antibody from a CHC patient

The antibody isolation was performed as described previously.^49,69 Briefly, CD19⁺IgM^–IgA^–IgD^– memory B cells were sorted from peripheral blood B cells of patient 3 (Pt-3) with flow cytometry and were seeded in 384-well plates (4 cells/well) in the presence of cytokine IL-2 (abs04045, Absin, China), IL-21 (abs04900, Absin) and irradiated 3T3-msCD40L cells (established in our lab) as stimulators. After 13 days of culture, the supernatant of each well was detected for neutralizing activity against HCV recombinants (below), and the wells containing nAbs (neutralizing HCV) were subjected to the molecular cloning of antibody heavy chain and light chains using vectors IgG-AbVec, Igκ-AbVec, and Igλ-AbVec (provided by Prof. Patrick C Wilson, University of Chicago).⁶⁹ 293T cells were transfected with antibody-expressing plasmids and cultured with serum-free medium for 3 days (Sigma). The culture medium was collected, filtered (0.45 μm, Pall, USA), and enriched with Amicon Ultra-15 Centrifugal Filter (30 KD cutoff, Millipore, USA), and IgG was purified with Magne™ Protein G Beads (Promega). The sequence of antibody variable and junctional gene were identified using BLAST tool (http://www.ncbi.nlm.nih.gov/igblast/), and the expression of antibody was confirmed with silver staining method.⁷⁰

Generation of cell culture-derived HCV (HCVcc)

HCVcc of genotypes 1–6 were generated and maintained as described in our previous studies (provided by Dr. Jens Bukh, Hvidovre Hospital and University of Copenhagen, Denmark), including full-length infectious clones (genotypes): TNcc(1a),⁶² JFH1-based J6 5′UTR-NS2 recombinants J6^{5’UTR−NS2}/JFH1(2a) and J6cc(2a),^63,64 as well as isolate-specific 5′UTR-NS5A recombinants S52(3a), ED43(4a), SA13(5a), and HK6a(6a).⁶⁵ The titers of HCVcc were determined by focus forming units (FFUs) assay as our previous studies.^{62,64,65,71,72} Briefly, Huh7.5 cells (7×10³ cells) were seeded in 96-well plate in 100 μL DMEM for 16–18 h. The HCV supernatant was diluted into two-fold serial dilutions using DMEM and incubated with the cells for 48 h. Then, the cells were fixed with methanol (−20°C), incubated with anti-Core C7-50 antibody, visualized by Alexa Fluor 488 goat anti-mouse IgG (H + L) secondary antibody. The number of FFUs was counted manually under a fluorescence microscope (Olympus IX70).

Neutralization assay for HCVcc

For neutralization of HCVcc, 6×10³ Huh7.5 cells were seeded in 96-well plates one day before infection. One hundred FFUs of HCVcc stock were incubated with patient IgG or rabbit antisera/IgG at 37 °C for 1 h and then added to Huh7.5 cells and cultured for 6 h. The mixture was replaced with fresh medium and left for 48 h, and then HCV infection was examined by an indirect immunofluorescence method. The percentage of neutralization was evaluated relative to control.

Cell-free and cell-cell transmission assays of HCV infection

Huh7.5 cells were infected with J6^{5’UTR−NS2/}JFH1 virus stock at a multiplicity of infection (MOI) of 0.01 for 48 h prior to transmission assays. For cell-free transmission assay, 2.5×10⁵ naive Huh7.5 cells (target) and J6^{5’UTR−NS2}/JFH1-infected Huh7.5 cells (producer) were seeded in 6-well plates on either side of a transwell insert (0.1 μm; BD Falcon, USA).²² Mock control or patient IgG (100 μg/mL) was added to the producer cells and cultured for 2 days, and then the infection of target cells was monitored by indirect immunofluorescence assay. For cell-cell transmission assay, J6^{5’UTR−NS2}/JFH1-infected Huh7.5 cells (1.25×10⁵ cells, MOI = 0.01; producer) and the same number of naive NIrD cells (target) were mixed and seeded in 12-well plates for 48 h mAb AP33 (25 μg/mL, Creative BioLabs, USA) was used to block the cell-free transmission.³⁹ The number of HCV-infected target NIrD cells was determined by mCherry expression in the presence of doxycycline (2 μg/mL) using a fluorescence microscope (Olympus IX70).

Antibody treatment of HCV-infected Huh7.5 cells

Huh7.5 cells were seeded in 48-well plates (2.4×10⁴ cells/well) and infected with 400 FFUs of HCVcc for 24 h. The culture medium was changed every 72 h, and the patient IgG (100 μg/mL) or HVR1 peptides of different concentrations were added at 2, 5, 8 days post infection (dpi). HCV infection at 9 dpi was determined with indirect immunofluorescence assay or western blot.

Indirect immunofluorescence assay

At 72 h post infection, Huh7.5 cells were fixed with methanol (−20°C) for 25 min and blocked with 1% bovine serum albumin (BSA) in PBS for 30 min. Anti-HCV Core mAb C7-50 (1:200) and secondary goat anti-mouse IgG (H + L) Alexa Fluor 488 (1:300) were used to detect HCV infection. The fluorescence foci were enumerated with a fluorescence microscope (Olympus IX70).

Western blot

The cells were lysed with immunoprecipitation (IP) lysis buffer (87787, Thermo Fisher) in the presence of protease inhibitor cocktail (87786, Thermo Fisher) according to the manufacturer’s instruction. Total protein (20 μg/lane) was dissolved in 4× loading buffer (0.2 M Tris-HCl [PH 6.5], 0.4 M dithiothreitol, 277 mM sodium dodecyl sulfate [SDS], 6 mM bromophenol blue, and 4.3 M glycerol), separated by a 12% SDS-polyacrylamide gel electrophoresis (PAGE) gel, and transferred to immune-blot PVDF Membranes (16201777, Bio-Rad, USA). The membrane was blocked with 3% BSA in Tris-buffered saline and Tween 20 (TBST) for 1 h and probed with primary C7-50 antibody (1:200) and secondary goat anti-mouse IgG-HRP (1:2000) with 1% BSA in TBST. The signal was detected with ECL Western Blotting Substrate (Pierce).

Immunoprecipitation assay

The sequence of soluble HCV E2 protein of genotype 2a with a FLAG tag at C-terminal (E2-Flag; J6 strain, GenBank accession no. JQ745650; aa364-661 according to H77 reference) was cloned to pcDNA3.1 vector as described previously.⁷³ Briefly, the sequence was amplified with forward primer containing Kozak sequence (GCCGCCACC), CCGAAGCTTGCCGCCACCATGATGCAGGGAGCGTGGGCG and reverse primer containing FLAG tag sequence, TGCTCTAGACTACTTATCGTCGTCATCCTTGTAATCTTGACTTCTGTCTCTGTC. And then the PCR product was cloned to pcDNA3.1 vector using HindIII and XbaI restriction sites. To express E2-Flag, 293T cells grown in a 10-cm culture dish were transfected with 10 μg of plasmid and left for 48 h, and the culture supernatant was collected and designated as E2-Flag protein mix, filtered (0.45 μm, Pall), aliquoted, and stored at −80°C for later use. Huh7.5 cells (5×10⁶ cells) were harvested and incubated with 1 mL of E2-Flag protein mix at 37 °C for 1.5 h, in the presence of control IgG, nAbs, or culture medium (mock control). Then, the cells were centrifuged at 1500 rpm for 3 min, washed with PBS three times, and lysed with IP lysis buffer (87787, Thermo Fisher) for 15 min on ice. Approximately, 1/10 of the cell lysate in each group (input) was analyzed by western blot, and the remaining sample was incubated with anti-Flag magnetic beads (M8823, Sigma) at 4 °C overnight, and E2-immunoprecipitated HCV receptors were detected with western blot.

Quantification and statistical analysis

Student’s paired or unpaired t-tests were performed to compare the differences between two groups with GraphPad Prism 8 software (GraphPad, USA). All p values were two-sided and p< 0.05 was considered to be significant.

Published: March 16, 2023

Data and code availability

• •Data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.