Study subjects were recruited among PWID who were participating in the Montreal Hepatitis C cohort (HEPCO, study protocol approval number: SL 05.014). All subjects were HIV-negative. HCV reinfection was defined by an HCV-positive RNA test following two consecutive negative tests >30 days apart (Cobas Ampliprep/Cobas TaqMan HCV Qualitative Test, version 2.0; limit of detection: 15 IU/ml). The median between the last negative and first positive HCV RNA test was used to determine the estimated date of reinfection (EDI). Study subjects were considered spontaneous resolvers of reinfection if HCV RNA was negative at 6 months post-EDI, while chronics were defined by a positive test.

HCV-specific T-cell responses were measured using an IFN-γ enzyme-linked immunospot (ELISpot) assay, as previously described (25) with an input of 2 × 10⁵ PBMCs/well against 11 pools of overlapping peptides spanning the entire HCV polyprotein corresponding to genotype (Gt) 1a (H77), 1b (J4), or 3a (K3a/650) sequences (BEI Resources, Manassas, VA, USA). For some subjects, we used the ELISpot Flex Human IFN-γ (ALP) kit (Mabtech, Cincinnati, OH, USA). All assays were performed directly ex vivo on frozen PBMCs. Specific spot-forming cells (SFCs) were calculated as the mean number of spots in test wells minus the mean number of spots in negative control wells and normalized to SFC/10⁶ PBMCs. Early and late acute time points used for this assay differed from the ones mentioned in Supplementary Table 1 for subjects SR/SR-4, SR/SR-6, and SR/CI-2. Pre-reinfection time points of SR/CI-1 and SR/CI-2 were also different.

Biotinylated E2 monomers were generated by the laboratory of Dr. Arash Grakoui (26) and prepared as previously described (17) by five times addition of 3.75 μl phycoerythrin (PE)-labeled ExtrAvidin (Millipore Sigma, Oakville, ON, Canada) or allophycocyanin (APC)-labeled streptavidin (Invitrogen, Carlsbad, CA, USA) to 4.5 μl of E2 monomers (2.29 mg/ml) followed by 10 min of incubation at room temperature (RT) after each addition. For tetramer staining, 0.56 μl of tetramer-PE and 0.56 μl of tetramer-APC were used in 100 μl of RPMI 10 [RPMI medium (Wisent, St-Jean-Baptiste, QC, Canada) supplemented with 10% fetal bovine serum (FBS) (Corning, Corning, NY, USA) and 0.01% sodium azide (Thermo Fisher Scientific, Waltham, MA, USA)].

Cryopreserved PBMCs were thawed and washed twice with RPMI 10 and then washed twice with FACS buffer (PBS 1×, 1% FBS, 0.01% sodium azide). For the identification of E2-specific MBCs, cells were incubated for 10 min at RT with Human BD Fc Block (BD Biosciences, Franklin Lakes, NJ, USA) and then stained with biotinylated E2 tetramers for 30 min at RT. Cells were then washed twice with FACS buffer and stained for 30 min at 4°C with surface markers (see Supplementary Table 2 : Panel #1 for antibodies) and viability dye [LIVE/DEAD fixable aqua dead cell stain kit (Thermo Fisher Scientific)] to identify live cells. Following two washes with FACS buffer, cells were fixed using 1% formaldehyde (Millipore Sigma). For cTfh phenotyping, cells were stained for 30 min at 4°C with surface markers (see Supplementary Table 2 : Panel #2 for antibodies) and viability dye. Cells were washed again and permeabilized using FoxP3 fixation and permeabilization buffer (Thermo Fisher Scientific) for 20 min at 4°C. Intracellular staining was performed for 30 min at 4°C. Cells were then washed twice with PermWash buffer and fixed using 1% formaldehyde. Multiparameter flow cytometry was performed at the flow cytometry core of the CRCHUM using a BD LSRFortessa instrument equipped with five lasers [UV (355 nm), violet (450 nm), blue (488 nm), yellow-green (561 nm), and red (640 nm)] and the FACSDiva version 9.2 (BD Biosciences). FCS data files were analyzed using FlowJo (version 10.8.1 for Mac; BD Biosciences).

Cryopreserved PBMCs were thawed and then washed twice with RPMI 5 [RPMI medium supplemented with 5% human Serum (Wisent) and 1% penicillin/streptomycin (Wisent)]. Cells (1 to 4 × 10⁶) were then incubated in 1 ml of RPMI 5 for 3 h at 37°C, 5% CO₂ in 5 ml of polypropylene push cap tubes (Thermo Fisher Scientific). Antibodies for the chemokine receptors CCR4, CCR6, CCR7, CXCR3, and CXCR5 were added to the culture, and the cells were further incubated for 15 min at 37°C, 5% CO₂, followed by the addition of a CD40 blocking antibody (Miltenyi, Gaithersburg, MD, USA) to prevent the interaction of CD40L with CD40 and its subsequent downregulation. Cells were then stimulated for 18 h at 37°C, 5% CO₂ with 1 µg/ml of HCV overlapping peptide pools showing the highest response in ELISpot. Cells stimulated with 1 µg/ml of staphylococcal enterotoxin B (SEB) (Toxin Technology Inc., Sarasota, FL, USA) and unstimulated cells served as positive and negative controls, respectively. Cells were then collected and transferred to a 96-well V-bottom plate, washed twice with FACS buffer, and stained for 30 min at 4°C with surface markers (see Supplementary Table 2 : Panel #3 for antibodies) and viability dye. Following two washes, cells were fixed using 1% formaldehyde. Multiparameter flow cytometry was performed as described above.

ELISA was performed as described previously (11) with 0.5 μg/ml in-house generated H77-E2 and J6-E2 (26) or NS3 protein [RayBiotech, Peachtree Corners, GA, USA recombinant NS3 (Gt 1a) amino acids 1192 to 1459]. Plasma samples from the subjects were diluted starting at 1:250 with four-fold dilutions up to 1:16,000. All samples were tested in duplicates. The difference in optical densities (OD450–570) was determined by subtracting the mean absorbance at 570 nm (background) from the mean absorbance at 450 nm. Absorbance measurements were performed on a Synergy 4 Microplate Reader (BioTek Instruments, Winooski, VT, USA).

CD81-knockout (KO) human embryonic kidney 293T (HEK 293T) (kindly provided by Drs. Joe Grove, University of Glasgow and Justin Bailey, Johns Hopkins) and human hepatoma Huh-7 (kindly provided by Dr. Charles Rice, The Rockefeller University) cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fisher Scientific) supplemented with 10% FBS and maintained at 37°C and 5% CO₂.

HCV pseudoparticles (HCVpp) were produced as described previously (27). Briefly, CD81-KO HEK293T cells were seeded at 1 × 10⁶ cells/well on six-well plates. The following day, cells were transfected with pNL4–3.Luc.R-E-plasmid containing the env-defective HIV proviral genome, pAdVantage, and HCV E1E2 plasmids (UKNP1.11.6, 1a154, UKNP4.2.2, 1a72, 1b58, UKNP3.1.2, or UKNP1.18.1) (kindly provided by Dr. Justin Bailey) (28) using Lipofectamine 2000 (Thermo Fisher Scientific) following the manufacturer’s instructions. For mock HCVpp, HCV E1E2 plasmid was not added to the transfection mix. Media were changed 24 h later, and supernatants were collected at 48 and 72 h after transfection. Supernatants were filtered through a 0.45-μm filter, separated into aliquots, and stored at −80°C until future use.

Neutralization assays were performed as previously described elsewhere (27). Huh-7 cells were plated on wells of sterile 96-well white plates (Thermo Fisher Scientific) at a density of 1.5 × 10⁴ cells/well. The next day, plasma aliquots were incubated at 56°C for 30 min to inactivate the complement and centrifuged at 1,200×g for 5 min. Fixed plasma dilutions (1:25) were added to an equal volume of either UKNP1.11.6, 1a154, UKNP4.2.2, 1a72, 1b58, UKNP3.1.2, or UKNP1.18.1 E1E2 HCVpp or mock HCVpp. Dilutions of HCV-negative and HCV-positive (chronically infected donor) plasma were included as negative and positive controls, respectively. Plasma-HCVpp complexes (final plasma dilution 1:50) were incubated at 37°C for 1 h. A volume of 100 µl of plasma-HCVpp mixture was added to cell monolayers in duplicate and incubated for 5 to 6 h at 37°C and 5%CO₂, after which the mixture was discarded and replaced with phenol red-free DMEM (Wisent) supplemented with 10% FBS and 4 mM of L-glutamine (Wisent). After 48 h, cells were lysed in 50 μl of 1× Luciferase cell culture lysis buffer (Promega, Madison, WI, USA) at RT. Luminescence was detected by adding 50 µl of luciferase reagent (Promega) and measuring relative light units (RLU) on a Synergy 4 Microplate Reader (BioTek Instruments). Percent neutralization was calculated using the following equation: % neutralization = [1 − (RLU_{infection time point} − RLU_mock)/(RLU_negative − RLU_mock)]. Neutralizing breadth was defined as the number of HCVpp (of the 7 HCVpp used) neutralized with a % neutralization >50%, and neutralizing potency was calculated as the geometric mean of the % neutralization from the assays against 7 HCVpp for a given time point.

Statistical analyses were performed with Prism version 10.1.0 (GraphPad, Boston, MA, USA). Details of the tests (including the number of data points (n) and P-values) are provided in each figure legend. Differences between groups in longitudinal analyses were determined by two-way repeated measure ANOVA with Tukey’s post-hoc test. Comparisons that did not include multiple time points were examined by the two-tailed Mann–Whitney U test. Correlations between variables were examined using Spearman’s test. Non-linear regression was used to calculate antibody titers. For all statistical tests, P-values less than 0.05 were considered significant.

We examined the HCV-specific immune response during documented HCV reinfection episodes after spontaneous resolution in a group of PWID participating in the Montreal Hepatitis C cohort (HEPCO) study as described in the Materials and methods (29, 30) (n = 22; Table 1 ). Of these reinfection cases, 14 subjects spontaneously resolved the reinfection episode, hereinafter termed resolvers or SR/SR, while eight subjects developed chronic infection, hereinafter termed chronics or SR/CI. The detailed demographics and clinical characteristics of the subjects are summarized in Supplementary Table 1 . We analyzed the response at key time points during reinfection: pre-reinfection (variable), early acute (<3 months post-EDI), late acute (3–7 months post-EDI), and follow-up (>7 months post-EDI). Most SR/SR (64.2%) and SR/CI (62.5%) were reinfected with HCV Gt 1, while only 1 of 14 SR/SR and 2 of 8 SR/CI were reinfected with HCV Gt 3a ( Table 1 ). Both groups had pre-existing virus-specific memory T-cell responses that expanded during reinfection as measured in IFN-γ ELISpot assay against overlapping peptide pools covering the entire HCV polyprotein. During early acute reinfection, both groups had comparable total T-cell responses that were sustained during the late acute time point primarily in SR/SR ( Supplementary Figure 1 ).

First, we examined longitudinally the expansion of total activated cTfh identified as CD3⁺CD4⁺CD45RA⁻CXCR5⁺PD1⁺ICOS⁺FoxP3^– ( Figure 1A ) in SR/SR and SR/CI. We observed higher frequencies of total activated cTfh at the early acute time point in SR/SR ( Figure 1B ) compared to SR/CI subjects ( Figure 1C ), although there was no significant difference between the two groups ( Figure 1D ). In both groups, frequencies of activated cTfh diminished at late acute and follow-up. Notably, at early acute, five SR/SR subjects (SR/SR-6, SR/SR-9, SR/SR-11, SR/SR-12, and SR/SR-14) exhibited high frequencies of activated cTfh ranging from 23.4% to 41.7% ( Figure 1B ), whereas in the SR/CI group, activated cTfh did not surpass 19.9% ( Figure 1C ). Given that the time points analyzed are based on EDI, we reasoned that time points may not be all synchronized and that transient expansion of cTfh might have been missed in some subjects. Hence, we examined the correlation between the frequency of activated cTfh and the EDI. Indeed, there was an overall significant negative correlation between these two parameters in all subjects (r = −0.3516, P = 0.0068) ( Figure 1E ), and when stratified into the two groups, this correlation remained significant only in SR/SR (r = −0.3748, P = 0.0243) ( Figure 1F ) but not in SR/CI (r = −0.3123, P = 0.1571) ( Figure 1G ), underscoring the transient nature of this expansion in resolvers and the effect of time after reinfection. We then examined the distribution of distinct subsets of activated cTfh based on the expression of the chemokine markers CXCR3 and CCR6 (31–35) ( Figure 1H ). In both groups, the activated cTfh were predominantly cTfh2 (CXCR3⁻CCR6⁻) (median: SR/SR = 52.39%, SR/CI = 50.82%) prior to reinfection but became predominantly cTfh1 (CXCR3⁺CCR6⁻) (median: SR/SR = 60.34%, SR/CI = 44.89%) at early acute followed by cTfh2 (median: SR/SR = 17.5%, SR/CI = 31.25%), with smaller subsets demonstrating cTfh17 (CXCR3⁻CCR6⁺) (median: SR/SR = 5.6%, SR/CI = 7.19%) and cTfh1/17 (CXCR3⁺CCR6⁺) (median: SR/SR = 12.62%, SR/CI = 4.73%) phenotypes ( Figures 1I, J ). In summary, resolvers exhibited transient expansion of activated cTfh1 cells at the early acute time point, potentially contributing to clearance of HCV reinfection.

To confirm that the expanded cTfh are virus-specific, we performed an activation-induced markers (AIM) assay at early acute in two SR/SR (SR/SR-9 and SR/SR-10) and two SR/CI subjects (SR/CI-3 and SR/CI-5). We used the peptide pools that gave the highest response in the IFN-γ ELISpot assay, in addition to E2 as it is the main target of NAbs that require Tfh help. HCV-specific (AIM⁺) CD4 T cells were identified as CD14⁻CD19⁻CD3⁺CD4⁺CD8⁻CD69⁺ T cells expressing the activation markers CD40L and/or OX40. The representative gating strategy is presented in Figure 2A . SR/SR subjects exhibited significantly higher frequencies of HCV-specific CD4 T cells (P = 0.026) than SR/CI subjects ( Figure 2B ).

Next, we examined the distribution of different CD4 subsets among AIM⁺ T cells. The representative gating strategy is presented in Figure 2C . HCV-specific cells in SR/SR were predominantly Th17 (CXCR5⁻CXCR3⁻CCR6⁺), followed by Th1 (CXCR5⁻CXCR3⁺CCR6⁻) and Tfh (CXCR5⁺PD-1⁺), while Th2 (CXCR5⁻CXCR3⁻CCR6⁻CCR4⁺) and Th1/17 (CXCR5⁻CXCR3⁺CCR6⁺) represented a small percentage of AIM⁺ cells ( Figure 2D ). Moreover, HCV-specific cTfh cells in the two SR/SR subjects had an effector memory phenotype (CD45RA⁻CCR7^–) ( Figure 2E ). The frequency of HCV-specific CD4 T cells in SR/CI was too low to accurately phenotype the cells. In summary, SR/SR showed significantly higher frequencies of effector memory HCV-specific CD4 T cells compared to SR/CI during early acute reinfection. These CD4 T cells were enriched in Th17, Th1, and cTfh phenotypes.

Given the early expansion of activated HCV-specific cTfh observed in the SR/SR group and their essential role in helping B cells, we investigated the longitudinal dynamics of E2-specific, class-switched MBCs in both groups using HCV glycoprotein E2 tetramers (26) as previously described (17). E2-specific MBCs were identified as CD3⁻CD14⁻CD16⁻CD56⁻CD19⁺IgM⁻CD27⁺Tetramer APC⁺ and PE⁺ ( Figure 3A ). We observed early, transient, and significant expansion of E2-specific MBCs in SR/SR subjects as compared to follow-up time points (P = 0.0075) ( Figure 3B ), and the frequency of E2-specific MBCs remained significantly higher than the follow-up time points during the late acute stage of reinfection. During early acute reinfection, 10 out of 14 SR/SR exhibited elevated levels of E2-specific MBCs above the detection threshold. In contrast, expansion of E2-specific MBCs in SR/CI was delayed until the follow-up time points and did not reach the level of significance, with only three subjects surpassing the threshold of detection at early acute reinfection, two of which already had detectable levels at the pre-reinfection time point ( Figure 3C ). Thus, SR/SR exhibited higher and earlier expansion of E2-specific MBCs compared to the SR/CI subjects ( Figure 3D ).

We phenotyped E2-specific MBCs to distinguish between MBCs with a resting phenotype (CD71⁻), an activated phenotype (CD71⁺CD20^hiCD38^int-lo), and antibody-secreting cells (ASCs; CD71⁺CD20⁻CD38^hi) (36) ( Figure 3E ). At early acute reinfection, E2-specific MBCs in both groups exhibited both an activated or resting phenotype (activated median: SR/SR = 38.2%, SR/CI = 61.6%; resting median: SR/SR = 52.9%, SR/CI = 31.3%). At the peak of E2-specific MBC expansion in SR/CI at follow-up, the cells were predominantly of the activated phenotype (median = 87%) ( Figure 3F ). Importantly, during early acute reinfection, SR/SR but not SR/CI showed a subset of ASCs, with a median of 5.9% ( Figure 3F ). The four SR/SR subjects (SR/SR-6, SR/SR-9, SR/SR-11, and SR/SR-14) who exhibited the highest expansion of activated cTfh were those who also had detectable ASCs. The frequency of ASCs correlated with the frequency of activated cTfh ( Figure 3G ). These results suggest that earlier expansion of activated cTfh1 in SR/SR may enhance the expansion of activated HCV-specific MBCs differentiated into ASCs.

To assess whether expansion of HCV-specific B cells is associated with enhanced production of anti-HCV antibodies, we evaluated the levels of IgG in plasma targeting HCV proteins by ELISA. We performed four-fold serial plasma dilutions to determine the antibody titer at 50% binding, defined as the dilution showing 50% binding to H77-E2 Gt 1a ( Supplementary Figure 2A ), J6-E2 Gt 2a ( Supplementary Figure 2B ), and H77-NS3 Gt 1a as a control ( Supplementary Figure 2C ). At early acute reinfection, SR/SR showed higher titers of IgG than SR/CI against H77-E2 and J6-E2 proteins, 3-fold and 2.5-fold, respectively ( Figures 4A, B ). However, these differences were not significant. Titers of anti-E2 Gt 2a in SR/SR decreased significantly (P = 0.0079) from early acute to follow-up time points ( Figure 4B ). In contrast, SR/CI had higher antibody titers against the H77-NS3 protein compared to SR/SR at all times during reinfection ( Figure 4C ). Overall, SR/CI reached the highest antibody titers at follow-up when chronic HCV infection was already established. Together, these results indicate that SR/SR develops earlier and higher antibody titers against E2 protein, the target for NAbs that may contribute to viral clearance.

Based on the antibody titer results ( Figures 4A–C ), we selected the OD450–570 values from 1:1,000 plasma dilution for all subsequent analyses. Longitudinal analysis of plasma antibody responses at this dilution against H77-E2 Gt 1a, J6-E2 Gt 2a, and H77-NS3 Gt 1a in SR/SR ( Supplementary Figure 3A ) and SR/CI ( Supplementary Figure 3B ) was not significantly different at early acute reinfection ( Supplementary Figure 3C ). However, the levels of anti-E2 antibodies at the early acute time point significantly and positively correlated with the frequency of E2-specific MBCs from SR/SR (H77-E2: r = 0.8407, P = 0.0006 and J6-E2: r = 0.7912, P = 0.002) ( Supplementary Figure 4A ) and SR/CI (H77-E2: r = 0.8571, P = 0.0238 and J6-E2: r = 1.000, P = 0.0004) ( Supplementary Figure 4B ). These results suggest that the recall response of E2-specific MBCs leads to increased antibody titers, potentially through the observed rapid differentiation into the ASC population ( Figure 3F ).

Next, we investigated whether the high titers of HCV-specific antibodies were reflected in the plasma neutralization activity. We performed neutralization assays with seven HCVpp using a standardized panel (28, 37) harboring envelope glycoproteins E1E2 from HCV Gt 1a, 1b, 3a and 4a and representing different tiers of neutralization sensitivity (28, 37). The detailed results are presented in Supplementary Figure 5 and summarized in Figure 4D . Seven out of 14 SR/SR neutralized >50% UKNP1.11.6 (Gt 1a) HCVpp from tier 1 (most sensitive to neutralization) compared to only two out of eight SR/CI at early acute reinfection ( Figure 4D ). However, there was no significant difference in the percent neutralization between SR/SR and SR/CI at this time point. The highest percent neutralization against this HCVpp by SR/SR (98.31%) occurred at early acute, while in the SR/CI, the highest neutralization (96.20%) was not achieved until the follow-up time point ( Supplementary Figure 5A ). Neutralization of tier 2 (1a154-H77 Gt 1a) and tier 3 (UKNP4.2.2 Gt 4a, 1a72 Gt 1a, and 1b58 Gt 1b) HCVpp followed the same trend, where SR/SR showed a higher percentage of neutralization at early acute compared to follow-up, while SR/CI reached the highest percent of neutralization only late during reinfection at follow-up time points ( Supplementary Figures 5B–E ). Regarding tier 4, the most resistant to neutralization represented by UKNP1.18.1 Gt 1b and UKNP3.1.2 Gt 3a, only SR/SR-5 neutralized >50% of both tier 4 HCVpp at early acute ( Supplementary Figures 5F, G ). At the same time point, four SR/SR (SR/SR-3, SR/SR-5, SR/SR-11, SR/SR-14) were able to neutralize >50% of the HCVpp Gt 3a versus only one chronic (SR/CI-6) ( Supplementary Figure 5G ).

We further integrated the neutralization results to define neutralization breadth and potency. Neutralization breadth, which is the number of HCVpp neutralized >50%, peaked at early acute in the SR/SR and declined with time ( Figure 4E ). In contrast, neutralization breadth in SR/CI exhibited slower kinetics, where it increased significantly only at the follow-up time points (P = 0.0365 and 0.0263 compared to early and late acute, respectively). Neutralization potency, which is the geometric mean of % neutralization, demonstrated a similar trend ( Figure 4F ), where the neutralization potency in SR/SR peaked at early acute, but only increased significantly in SR/CI much later at follow-up (P = 0.0460 compared to late acute). The early neutralization breadth and potency correlated positively and significantly in SR/SR with the levels anti-E2 Gt 1a ( Supplementary Figure 6A , breadth: r = 0.8742, P = 0.0002 and potency: r = 0.8736, P = 0.0002) and anti-E2 Gt 2a ( Supplementary Figure 6B , breadth: r = 0.7635, P = 0.0037 and potency: r = 0.7473, P = 0.0046) but not in SR/CI ( Supplementary Figures 6C, D , respectively). These findings suggest that anti-E2 antibodies generated early with high titers in resolvers mediate HCVpp neutralization with high breadth and potency, thus potentially contributing to HCV clearance upon reinfection.

Because of the early coordinated expansion of cTfh and E2-specific MBCs along with the elevated plasma neutralization activity in the SR/SR, we examined the correlations between these responses at early acute. Overall, we observed more significant positive associations between the different early acute immune parameters in SR/SR as compared to SR/CI as summarized in Figure 5A and detailed in Supplementary Figures 7 , 8 . Notably, the frequencies of E2-specific MBCs correlated with neutralization breadth and potency in SR/SR but not SR/CI ( Figures 5B, C ). The frequencies of activated cTfh1 (ICOS⁺FoxP3⁻CXCR3⁺CCR6⁻) also correlated positively with plasma neutralizing activity in the SR/SR but not SR/CI ( Figures 5D, E ). Furthermore, at early acute reinfection, the frequencies of total activated cTfh from SR/SR but not SR/CI showed positive but non-significant correlation with anti-E2 antibodies ( Supplementary Figures 7A, B ), E2-specific MBCs ( Supplementary Figures 7C, D ), and neutralization breadth and potency ( Supplementary Figures 7E, F ). Similarly, activated cTfh1 from SR/SR but not SR/CI showed a positive non-significant correlation with anti-E2 antibodies ( Supplementary Figures 8A, B ) and E2-specific MBCs ( Supplementary Figures 8C, D ) at early acute reinfection. Altogether, these results further support that early activation of cTfh with a Th1 phenotype along with early expansion of E2-specific MBCs may drive the production of HCV-specific NAbs leading to the clearance of HCV reinfection.