Targeted Removal of HCV E2 N2 N-Glycan Is Associated with Improved Immune Responses in Mice

TL;DR

Removing a specific sugar from a hepatitis C virus protein improves immune responses in mice, suggesting a better vaccine strategy.

Contribution

Selective removal of the N2 glycan on HCV E2 improves immune responses while preserving antigen integrity.

Findings

N2 mutant (sE2-N2) showed higher anti-E2 titers and better inhibition of HCV genotype 1a.

N2 immunization increased CD8+ T cell activity and tumor growth inhibition in mice.

Therapeutic vaccination with sE2-N2 reduced HCV RNA and liver signals in infected mice.

Abstract

Hepatitis C virus (HCV) still lacks a licensed vaccine. The envelope glycoprotein E2 is a key neutralizing target, but its dense N-glycan shield can hinder epitope exposure. In this study, we revisit E2 glycan editing and examine whether single-site deletion preserves antigen integrity while improving immune responses in mice under a DNA immunization setting. Using a secreted E2 ectodomain (sE2384–661), we generated five N to D mutants at conserved sites (N1, N2, N4, N6, and N11) and evaluated them in a unified DNA immunization model with identical CpG content and delivery conditions across groups. The N2 mutant (N423, sE2-N2) maintained expression, secretion, and ER localization; furthermore, in mice, it was associated with higher anti-E2 titers and greater inhibition of H77 (genotype 1a) HCVcc at the tested dilutions, with limited activity against Con1 (1b). Cellular analyses showed increased IFN-γ ELISPOT counts and higher frequencies of granzyme B+/perforin+ CD8+ T cells after N2 immunization, while IL-4 remained low. Functionally, N2 elicited stronger specific lysis of CT26-sE2 targets in vitro and slowed CT26-sE2 tumor growth in vivo. In HCV-infected ICR4R+ mice, therapeutic vaccination with sE2-N2 reduced blood HCV RNA and hepatic readouts compared with sE2. A monoclonal antibody isolated from sE2-N2-immunized mice (1C1) neutralized HCVcc in vitro and, after passive transfer, lowered viremia and liver signals in infected mice. Collectively, these findings indicate that selective removal of the N2 glycan preserves antigen properties and is associated with improved humoral and cellular immunity and measurable in vivo activity, supporting targeted glycan editing as a practical strategy to refine E2-based HCV vaccines.

Figures

Full text

1. Introduction

Hepatitis C virus (HCV) infection remains a major global health concern [1,2]. Although direct-acting antivirals (DAAs) can cure most treated individuals, barriers to access, persistent reinfection in high-incidence populations, and the absence of herd immunity continue to sustain transmission and disease burden [3,4,5]. Developing a prophylactic vaccine is therefore still a critical goal for HCV elimination [6,7,8,9].

Among HCV antigens, the envelope glycoprotein E2 is the primary target of broadly neutralizing antibodies (bnAbs) that block viral attachment and entry via host receptors such as CD81 and SR-BI [10,11,12]. Structural studies of the E2 core and the E1E2 complex have revealed a compact, β-sandwich-like scaffold decorated with a dense array of N-linked glycans, with conserved neutralizing epitopes (e.g., AR3/”neutralizing face”) located near receptor-binding regions [13,14,15]. While essential for protein folding and virion assembly, this glycan shield can occlude bnAb epitopes, dampen B-cell recognition, and facilitate immune escape through glycan shifting or remodeling [16,17,18,19]. Consistent with this, bnAbs that overcome or bypass glycan shielding can neutralize diverse genotypes and confer protection in vivo, underscoring E2’s potential as a vaccine immunogen if key epitopes are effectively presented [20,21,22].

Accumulating virological and biochemical data indicate that individual E2 glycans make non-redundant contributions to entry, receptor engagement, and antigenicity [13,23,24]. Mutagenesis of selected N-glycosylation sequons alters CD81 binding, neutralization sensitivity, and epitope exposure, suggesting that rational “glycan editing” could improve the quality of vaccine-elicited antibody responses without destabilizing the antigen [13,25,26]. Recent engineering efforts using modified E2 constructs support this premise by enhancing neutralization sensitivity and inducing broader cross-genotype responses in animals [27,28,29]. However, a systematic evaluation of which conserved E2 glycans can be deleted to maximize immunogenicity while preserving secretion and a native-like conformation remains limited.

In this study, we focus on a secreted E2 ectodomain (sE2) platform and test the hypothesis that targeted removal of specific conserved glycans can “unmask” bnAb epitopes and potentiate both humoral and cellular immunity. Our central finding is that, under the conditions tested in this study, the N2-site mutant (sE2-N2; position N423) is associated with enhanced immune responses and improved functional readouts compared with the wild-type sE2 and other single-site mutants. We aim to define the immunological and mechanistic basis for this improvement by characterizing how E2 glycan deletion reshapes antigenicity and T-cell responses and by comparing the protective efficacy of sE2-N2 versus wild-type sE2 in a humanized HCV challenge model. We also plan to map the neutralization specificity, epitope targets of sera, and representative monoclonal antibodies elicited by sE2-N2.

2. Materials and Methods

2.1. Ethical Approval

All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Wuhan University School of Medicine and complied with national regulations (S01317012R, 25 July 2024). HCVcc experiments were performed in BSL-2+ facilities under approved SOPs. Humane endpoints were predefined and strictly observed.

2.2. Animals and Housing

Female BALB/c mice (6–8 weeks, GemPharmatech, Changzhou, China) were used for immunogenicity, ELISPOT/flow cytometry, CTL, and tumor studies. ICR^4R+^ humanized mice (ICR background expressing human SR-BI, CD81, CLDN1, and OCLN) were used for HCV infection, therapeutic vaccination, and passive transfer studies; ICR parental mice served as negative controls where indicated. Mice were specific-pathogen-free and maintained on a 12 h light/dark cycle with ad libitum access to food and water.

2.3. Cell Lines and Culture

HepG2 (IFA, CVCL_0027), HEK293T (CVCL_0063, heterologous E2 expression), Huh7.5.1 (CVCL_E049, HCVcc infection), and CT26 (murine colon carcinoma) were authenticated and mycoplasma-free. Cells used in this study were obtained from our in-house laboratory cell bank (cryopreserved stocks maintained in our laboratory). Cells were cultured in DMEM supplemented with 10% FBS, 2 mM L-glutamine, and penicillin/streptomycin at 37 °C with 5% CO_2_.

2.4. Plasmids, Site-Directed Mutagenesis, and Sequence Verification

A secreted E2 ectodomain (sE2 384–661, genotype 1a H77) was cloned into pcDNA3.1-myc-his (cell expression) and into pGL3 (SV40 promoter) for DNA immunization and in vivo luciferase imaging. A CpG motif (CACGTT) was retained upstream of the expression cassette. Conserved N-glycosylation sites (N1, N2, N4, N6, and N11) were individually mutated by substituting asparagine with aspartate within the N-X-S/T sequon (N to D). All coding regions were confirmed by Sanger sequencing. The specific primers are listed in Supplementary Table S1. For genotype-diverse binding assays, E2 from genotypes 1a/1b/2a/3/4/5/6 was cloned into pCMV-tag2A.

2.5. Transfection and Recombinant Antigen Preparation

Cells were transfected with Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s instructions. Conditioned media and cell lysates were collected at 48–72 h, clarified (3000× g, 10 min), and used for immunoblotting or as antigen sources.

2.6. RT–qPCR Analysis

Total RNA was extracted from transfected cells 48 h post-transfection using TRIzol reagent according to the manufacturer’s protocol. RNA was reverse-transcribed into cDNA, and quantitative PCR was performed with SYBR Green Master Mix. Relative mRNA levels of sE2 constructs were calculated using the 2^−ΔΔCt^ method and normalized to GAPDH. The primer sequences were 5′-ATTCAGCTGACCCAGTCTC-3′ and 5′-TCGGACCTGTCCCTGTCGTC-3′.

2.7. Immunoblotting and Glycosidase Treatment

Samples were resolved using SDS–PAGE and transferred to PVDF. Primary antibodies included anti-myc (E2 tag), anti-HCV NS3 (infection readout), and anti-β-actin (loading), and HRP-conjugated secondaries with chemiluminescent detection were used. All the antibodies were obtained from Invitrogen. For glycan assessment, denatured proteins were treated with Endo H or PNGase F before electrophoresis.

2.8. Immunofluorescence Assay (IFA)

HepG2 cells on coverslips were transfected with pcDNA3.1-myc-his-sE2-WT or -sE2 N2. At 24–36 h, cells were fixed (4% paraformaldehyde), permeabilized (0.1% Triton X-100), blocked (5% BSA), and stained with anti-myc (for E2 detections, Invitrogen) and ER antibody-anti-calnexin (Cell Signaling Technology, Danvers, MA, USA). Nuclei were counterstained with Hoechst. Confocal images were acquired under matched settings.

2.9. DNA Electroporation and In Vivo Luciferase Imaging

Mice received intramuscular pGL3-sE2-WT or pGL3-sE2 N2 followed by electroporation (plate electrodes; identical parameters for all groups). Three days later, D-luciferin was injected i.p., and bioluminescence was imaged on an IVIS Lumina II (Caliper Life Science, Hopkinton, MA, USA). Regions of interest were placed at the injection site and analyzed with the Living Image software 4.0 (Caliper Life Science).

2.10. Vaccination Regimen

DNA vaccines (sE2-WT, sE2-N1, sE2-N2, sE2-N4, sE2-N6, and sE2-N11) were administered intramuscularly with electroporation on days 0, 14, and 28. Sera were collected 10–14 days after each dose and at endpoints, and splenocytes were harvested one month after the final immunization.

2.11. ELISA for Total IgG Titers and Isotypes

Plates were coated with purified sE2 or with GNA lectin (to perform an ELISA). After blocking, serial serum dilutions were added in duplicate. HRP-anti-mouse IgG (total) or isotype-specific secondaries (IgG1, IgG2a, and IgG2b) were used with TMB substrate. The above antibodies were purchased from Frdbio (Wuhan, China). The endpoint titer was the highest dilution with OD450 ≥ mean blank + 3 SD. The IgG2a:IgG1 ratio was calculated from linear-range OD values.

2.12. GNA Capture ELISA for Genotype-Diverse E2

GNA-coated plates captured E2 from 293T lysates expressing genotypes 1a/1b/2a/3/4/5/6. After washing and blocking, mouse sera (or mAb 1C1) were added and detected with HRP-conjugated secondaries. Background-subtracted OD450 values were compared across groups.

2.13. HCVcc Production and Infection/Neutralization Assays

HCVcc representing H77 (1a) and Con1 (1b) were generated by RNA transfection of producer cells; supernatants were collected and stored at −80 °C. For neutralization, Huh7.5.1 cells were infected in the presence of immune sera or purified antibodies (pre-incubation or co-incubation; identical conditions for all groups). After 48–72 h, cells were analyzed using RT-qPCR for intracellular HCV RNA and using immunoblot for NS3.

2.14. RT-qPCR for HCV RNA

Viral RNA was extracted from blood samples using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions., Reverse transcription was carried out using random hexamers. HCV RNA was quantified with 5′-UTR primers/probe. For blood and liver, absolute copies were derived from standard curves; for cell assays, values were normalized to housekeeping transcripts as appropriate.

2.15. ELISPOT

Splenocytes were seeded in PVDF plates coated with anti-IFN-γ or anti-IL-4. Cells were restimulated with HCVcc (MOI defined in pilot tests) or left unstimulated for 30 h. After incubation with biotinylated detection antibodies and streptavidin-HRP, plates were developed and read on an automated ELISPOT reader. Background-subtracted spots per 10^6^ cells are reported.

2.16. Flow Cytometry for CD8+ Cytotoxic Markers

Splenocytes were stimulated ex vivo with HCVcc or peptides in the presence of Golgi inhibitors. CD3 and CD8 were used for surface staining, while granzyme B and perforin were used for intracellular staining. All antibodies were obtained from Biolegend (San Diego, CA, USA). Data were acquired on a multicolor cytometer and analyzed with a fixed gating strategy (lymphocytes–singlets–live–CD3^+^CD8^+^–intracellular markers).

2.17. CT26-sE2 Target Cell Generation

CT26 cells were transduced/transfected with an sE2 secretion construct and selected to establish a stable line (CT26-sE2). E2 expression in lysates and supernatants was confirmed via immunoblot.

2.18. LDH-Release Cytotoxicity Assay

Splenocytes (effectors) and CT26-sE2 (targets) were co-cultured at graded effector-to-target (E:T) ratios for 4–6 h. LDH in supernatants was measured colorimetrically (Biolegend). Specific lysis (%) was calculated as follows: (experimental − effector spontaneous − target spontaneous)/(target maximum − target spontaneous) × 100. Triplicate wells were run for each condition.

2.19. CT26-sE2 Tumor Challenge

Immunized BALB/c mice (same DNA regimen as Section 2.9) were injected s.c. with 1 × 10^6^ CT26-sE2 cells. Tumor length and width were measured every 3 days by a blinded operator; volumes were computed as V = L × W^2^/2.

2.20. HCV Infection and Therapeutic Vaccination in ICR4R+ Mice

ICR^4R+^ mice were infected with HCV. For therapy, mice received sE2-WT or sE2-N2 DNA on the post-infection schedule. Blood was sampled at indicated time points for RNA analysis. At endpoint, livers were harvested for RNA and histology.

2.21. Serum Infectivity Assay

Post-treatment sera from infected mice were used to inoculate Huh7.5.1 cultures. After 15 days of passaging, cellular HCV RNA was quantified using RT-qPCR.

2.22. Liver Histology

Formalin-fixed, paraffin-embedded liver sections were stained with hematoxylin and eosin. A pathologist blinded to group identity assessed inflammatory foci and hepatocellular injury.

2.23. Hybridoma Generation and Monoclonal Antibody Purification

Spleens from sE2-N2-immunized mice were fused with myeloma partners using PEG. Hybridoma supernatants were screened through RT-qPCR neutralization against HCVcc-infected Huh7.5.1 cells. Positive clones were subcloned, and antibodies were purified from culture supernatants by protein A chromatography and buffer-exchanged into PBS. The lead clone was designated 1C1.

2.24. Binding Assays

For Kd estimation and isotyping of 1C1, GNA-captured HCVcc plates were incubated with serial 1C1 dilutions. Bound IgG was detected with HRP-anti-mouse IgG; data were fit to a one-site model in GraphPad Prism 9 (Boston, MA, USA) to estimate apparent Kd. The isotype was determined through an ELISA using HRP-anti-mouse IgG1/IgG2a/IgG2b secondaries.

2.25. Passive Transfer of 1C1 in ICR4R+ Mice

HCV-infected ICR^4R+^ mice were randomized to receive 1C1 or isotype/vehicle. Baseline blood was collected prior to dosing. End-point PCR and RT-qPCR were used to quantify blood HCV RNA at serial time points. At endpoint, livers were imaged with a small-animal system under identical exposure; regions of interest were analyzed quantitatively.

2.26. Statistical Analysis

Continuous variables are presented as the mean ± SEM. For two-group comparisons, an unpaired two-tailed Student’s t-test was performed (or the Mann–Whitney test, if assumptions were not met). For multi-group comparisons, a one-way ANOVA with Tukey’s post hoc test was performed (or the Kruskal–Wallis test with Dunn’s correction). Time-course data (viremia and tumor growth) were analyzed using a repeated-measures ANOVA or mixed-effects models. Seroconversion proportions were analyzed via Fisher’s exact test. Significance was set at p < 0.05. Exact n, dilutions, exclusion criteria, and the test applied are provided in the figure legends or text.

3. Results

3.1. Rational N-glycan Editing Identifies N2 Deletion as a Tractable Immunogen with Preserved Expression and In Vivo Producibility in Mice

To assess whether glycan deletion affects functional antibody activity, we constructed sE2384–661WT and the five single-site mutants (N1, N2, N4, N6, and N11). These five sites—N1 (N417), N2 (N423), N4 (N448), N6 (N532), and N11 (N645)—have been repeatedly highlighted as high-impact positions for improving E2 antigenicity, and several reports describe stronger vaccine responses after targeted deglycosylation or related E2 engineering at these positions [13,24].

Our results show that the N2 mutant preserves comparable E2 transcript levels (RT-qPCR; Figure S1A), detectable intracellular expression and secreted signal under matched production conditions, and ER localization in vitro and supports clear, localized in vivo expression after intramuscular electroporation of pGL3-sE2 constructs. This indicates that it has intact antigen integrity and practical producibility for downstream immunogenicity studies (Figure 1A). Under matched production conditions, the N2 mutant (sE2-N2) showed intracellular expression comparable to the wild-type (WT) after densitometric quantification and normalization to β-actin, and a similar secreted sE2 signal was observed in supernatants loaded at equal volumes without additional degradation bands (Figure 1B,C). To assess subcellular trafficking, an immunofluorescence assay (IFA) in HepG2 cells transfected with pcDNA3.1-myc-his-sE2WT or -sE2-N2 showed punctate perinuclear staining for E2 (anti-myc) with clear co-localization to the endoplasmic reticulum marker calnexin (anti-CNX), and similar patterns were observed between the WT and N2, indicating no gross misfolding or ER retention caused by N2 deletion (Figure S1B). Finally, to confirm in vivo producibility of our DNA immunogens, we cloned sE2WT or sE2-N2 into a pGL3 reporter cassette and delivered the plasmids through intramuscular electroporation. Three days later, D-luciferin administration yielded strong, localized bioluminescent signals at the injection site in both groups on an IVIS Lumina II system, demonstrating clear, localized in vivo expression (Figure S1C). ROI photon flux quantification further summarized the bioluminescent signal at the injection site and showed no significant difference between sE2-WT and sE2-N2 under the same delivery conditions (Figure S1C).

To assess whether glycan deletion affects functional antibody activity, we immunized mice with sE2384–661WT and the five single-site mutants (N1, N2, N4, N6, and N11). After the final boost, sera were tested against cell culture HCV (HCVcc). Seroconversion was observed in most vaccine groups (WT 6/6; N1 5/6; N2 6/6; N4 6/6; N6 4/6; N11 6/6; and vector 0/6). The mean total anti-E2 IgG titers varied by construct: N2 was the highest (55,466), followed by the WT and N4 (both 28,800), N11 (17,066), N1 (16,000), and N6 (12,800). Isotype profiles included IgG1 with variable IgG2a/IgG2b across groups. The IgG2a:IgG1 ratio—used here only as an indicator—was greater in N2 (1.2) than in the WT (0.75) and N11 (0.5); N6 was near zero. In the HCVcc neutralization column in Table 1, mean titers were detectable for N2 and N11 (both 3200) and lower for N1 (1600) and N6 (800), while the WT and N4 were undetectable under these conditions. Compared with sE2-WT, sE2-N2 elicited markedly higher total anti-E2 IgG titers (55,466 vs. 28,800; ~1.9-fold), a more Th1-skewed isotype distribution (IgG2a: IgG1 = 1.2 vs. 0.75), and detectable HCVcc neutralization (mean 1:3200 vs. 0), indicating that N2 enhances antibody magnitude and quality. In the RT-qPCR assay, several groups reduced intracellular HCV RNA compared with PBS and vector controls. The N2 group showed the largest decrease under the tested conditions (Figure 1D,E). Western blotting for NS3 gave a similar pattern: NS3 signals were weaker in immune sera than in controls, with the strongest reduction observed in the N2 group (Figure 1F). We next examined binding to E2 proteins from multiple genotypes using GNA-capture ELISA. Immune sera bound E2 from genotypes 1a, 1b, 2a, 3, 4, 5, and 6 with variable efficiency. Binding from the N2 group was consistently detectable and was higher for 1a, with modest or heterogeneous signals for other genotypes (Figure 1G). Finally, we conducted a head-to-head comparison of sE2-WT and sE2-N2 using an HCVcc RT-qPCR assay. Under the same dosing and time points, N2 anti-sera inhibited H77 (1a) infection more than the WT and showed limited but measurable inhibition of Con1 (1b). Differences that met the predefined threshold are marked in the figure (p < 0.05 vs. WT and controls) (Figure 1H). Overall, within this study design, glycan deletion at N2 was associated with stronger neutralization against 1a and detectable activity against 1b, while other mutants showed smaller or variable effects.

3.2. N2 and, to a Lesser Extent, N1 Vaccination Enhance E2-Specific Cytotoxicity and Slow CT26-sE2 Tumor Growth

To evaluate post-immunization cellular immunity, splenocytes were collected and restimulated with HCVcc for 30 h (or left unstimulated). IFN-γ ELISPOT showed higher spot numbers in vaccine groups than in vector. Under these conditions, N2 exceeded WT and N1, and N1 was above the WT (group means ± SEM from four experiments); representative plates are shown (Figure 2A,B). IL-4 ELISPOT remained low across groups with only small differences (Figure 2C,D). Guided by the IFN-γ results, we analyzed cytotoxic markers in the higher IFN-γ groups. Flow cytometry of CD3^+^ CD8^+^ T cells showed greater frequencies of granzyme B^+^ and perforin^+^ cells in N1 and N2 than in the WT and vector (Figure S2A). Taken together, within this protocol, N1 and especially N2 were associated with stronger Th1-leaning cellular responses compared with the WT.

To test whether immunization induces E2-specific cytotoxic activity and confers tumor control, we established a CT26 line stably expressing secreted E2 (CT26-sE2); Western blot confirmed robust E2 expression in cell lysates and supernatants (Figure 2E). Mice were vaccinated exactly as shown in Figure 2 (same dosing, schedule, and plasmids). One month after the last dose, splenocytes were collected for CTL assays, and a parallel cohort was used for tumor challenge. Splenocytes harvested 1 month after the last dose served as effectors in an LDH-release assay against CT26-sE2 targets (n = 3 per condition). Specific lysis increased with higher effector-to-target (E:T) ratios across vaccine groups and was minimal in vector controls. Under the same E:T ratios, sE2-N2 showed the greatest lysis, sE2-N1 was intermediate, and sE2-WT was lower. The separation between N2 and WT was most evident at upper E:T ratios, while N1 consistently exceeded the WT but by a smaller margin (Figure 2F–J). Vaccinated mice (n = 6/group) received 1 × 10^6^ CT26-sE2 cells subcutaneously, and tumor volumes were measured every 3 days. Tumors in vector and sE2-WT mice grew steadily. Growth in the sE2-N2 group was slower throughout the observation period, with smaller mean volumes than the WT at multiple time points. The sE2-N1 group showed a modest delay relative to the WT (Figure 2K). In this model, N2 immunization produced the strongest E2-specific cytotoxic activity in vitro and the most pronounced slowing of CT26-sE2 tumor growth in vivo; N1 showed a consistent but smaller effect. Taken together, the data above led us to select the sE2-N2 construct for the candidate vaccine.

3.3. Splenic Transcriptomics Show N2-Skewed Enhancement of IFN/T-Cell Programs with Preserved Humoral Modules

To understand how the sE2-N2 immunogen reshapes the splenic immune program relative to sE2WT, we profiled spleen transcriptomes after immunization (Figure 3). The volcano plot (Figure 3A) shows a coordinated shift toward genes involved in antigen presentation and T-cell/IFN signaling in N2. Canonical markers—including Ifnγ, Tcf7, Tbx21, Cxcl10, Gzmb, Irf1, H2-Ab1, B2m, and Psmb8/9/10—were increased in N2, consistent with enhanced helper/cytotoxic T-cell activity and improved antigen processing. Conversely, multiple B-lineage/plasma-cell and complement components (Pax5, Prdm1, Xbp1, Igj, Fcgr2b, and C1q/C3) tended to lie on the opposite side of the plot.

Functional enrichment supported these observations (Figure 3B). N2-upregulated genes were significantly enriched for response to interferon-γ/α, T-cell activation, antigen processing and presentation, and leukocyte chemotaxis. Terms preferentially represented among genes lower in N2 (relative to WT) included humoral immune response, B-cell activation, germinal center formation, and complement activation. A targeted view of immune markers (Figure 3C) highlighted higher scores for antigen-processing machinery and key IFN/T-cell effectors in N2, while selected B/plasma markers remained detectable. Hallmark GSEA provided an orthogonal confirmation (Figure 3D): IFN-γ and IFN-α responses, TNF/NF-κB, IL6–JAK–STAT3, and allograft rejection signatures were positively enriched toward N2, whereas complement was negatively enriched (depleted in N2). These transcriptomic data dovetail with our serology in which N2 elicited higher total anti-E2 IgG titer levels than the WT (geometric mean 55,466 vs. 28,800) and a more Th1-skewed isotype (higher IgG2a: IgG1), together with functional neutralization. Taken together, the results indicate that sE2-N2 amplifies IFN-driven cellular immunity and antigen presentation pathways while maintaining sustained humoral responses, providing a mechanistic explanation for the enhanced antibody quality and overall antiviral profile observed for sE2-N2 in our experimental settings, relative to sE2WT.

3.4. Therapeutic Comparison of sE2 and sE2-N2 in HCV-Infected ICR4R+ Mice

To test the therapeutic effect of the vaccines after infection, ICR^4R+^ mice were challenged with HCV and then treated according to the therapeutic schedule shown in Figure 4A. RT-qPCR of blood at the indicated time points showed lower HCV RNA in the sE2-N2 group than in sE2 and vector controls under the conditions tested (Figure 4B). Sera from treated mice were used to infect Huh7.5.1 cells and passaged for 15 days. Cultures exposed to sE2-N2 sera contained less HCV RNA than those exposed to sE2 or control sera (Figure 4C). Hepatic HCV RNA was reduced in the sE2-N2 group relative to sE2 and controls (Figure 4D). H&E staining showed fewer inflammatory foci and less hepatocellular injury in livers from sE2-N2-treated mice; sE2 showed a milder effect (Figure 4E). Taken together, sE2-N2 was associated with lower viremia, reduced serum infectivity, and decreased liver viral RNA and pathology compared with sE2.

3.5. Selection and Characterization of a Monoclonal Antibody from sE2-N2-Immunized Mice

To isolate a representative antibody induced by sE2-N2, we generated hybridomas from sE2-N2-immunized mice and screened culture supernatants for anti-HCV activity. In RT-qPCR assays of HCVcc-infected Huh7.5.1 cells, several candidates reduced intracellular viral RNA; clone 1C1 showed the strongest reduction among the tested supernatants (Figure 5A). Protein-A-purified antibodies were then evaluated using immunoblot for NS3, which confirmed greater suppression of viral protein in the 1C1 group compared with other candidates (Figure 5B). To assess binding, we measured 1C1 interaction with GNA-captured HCVcc using an ELISA across serial concentrations. The binding curve was well fit by a one-site model and yielded an apparent KD (GraphPad Prism), indicating specific recognition of virions by 1C1 (Figure 5C). Isotyping through an ELISA identified the IgG subclass of 1C1 (Figure 5D). Together, these data show that a monoclonal antibody derived from the sE2-N2 regimen—1C1—exhibits the most pronounced in vitro neutralizing activity among candidates tested and binds HCVcc with saturable, specific kinetics, supporting its use for subsequent in vivo evaluation.

3.6. Passive Transfer of mAb 1C1 Reduces Viremia and Hepatic Signals in HCV-Challenged ICR4R+ Mice

We used the ICR4R^+^ humanized mouse model, which expresses human SR-BI, CD81, CLDN1, and OCLN on an ICR background and supports HCV infection under immunocompetent conditions. For context, parental ICR mice do not sustain infection; in some studies, a dual-receptor ICR^2R+^ strain (CD81/OCLN) was used as a reference [30]. To test the in vivo activity of the sE2-N2-derived antibody, ICR^4R+^ mice were infected with HCV and then given mAb 1C1 according to the schedule shown in Figure 6A. Baseline blood samples confirmed comparable RNA levels across groups before the first antibody dose. At each post-treatment time point, end-point PCR showed fewer RNA-positive samples in the 1C1 group than in isotype/vehicle controls. Quantitatively, RT-qPCR curves separated after the first dose and remained lower for 1C1 through subsequent collections (Figure 6B). An AF680-labeled ZE18 aptamer (an HCV E2–specific probe) was used to visualize intrahepatic infection in vivo [31]. Reductions were observed in most animals within the 1C1 cohort and were reproducible across independent runs performed under the same dosing and sampling windows. Exact group sizes, routes, doses, and statistical outputs are provided in the Section 2. At the study endpoint, whole-organ imaging revealed diminished hepatic signal intensity in 1C1-treated mice relative to the controls (Figure 6C). Group means (±SEM) were lower in the 1C1 cohort, consistent with the reduced intrahepatic viral burden measured by blood RNA. Together, these data indicate that passive transfer of the sE2-N2-elicited mAb 1C1 confers measurable antiviral activity in vivo in ICR^4R+^ mice—evidenced by fewer PCR-positive samples, persistently lower blood HCV RNA, and reduced liver imaging signals under the conditions tested.

4. Discussion

This study revisits E2 glycan editing as a route to improve HCV vaccine performance. We focused on conserved N-linked sites in the E2 ectodomain and found that deleting the N2 glycan (N423) was associated with the most consistent benefits across assays. The sE2-N2 construct preserved detectable intracellular expression and secreted signal and showed ER localization under matched conditions, indicating that the modification did not grossly disrupt folding under our conditions. These observations are consistent with the possibility that targeted deglycosylation can increase epitope accessibility while maintaining antigen integrity, which is compatible with reports that E2 glycans modulate antigenicity and receptor engagement [13,24,32]. However, we cannot exclude that removal of the conserved N423 glycan may also induce subtle conformational changes in sE2 relative to native virion-associated E1E2, potentially biasing responses toward non-native epitopes. Therefore, future studies are needed to directly validate native-like conformation and epitope presentation; for example, through high-resolution structural approaches (e.g., cryo-EM/cryo-MS) and binding/competition analyses using well-characterized conformational neutralizing antibody panels, including AR3-region bnAbs.

Humoral measurements favored sE2-N2 over the wild-type in our mouse model. ELISA titers were higher, and sera from sE2-N2-immunized mice reduced HCVcc RNA and NS3 to a greater extent at the tested dilutions. Binding to E2 from multiple genotypes was detectable and strongest for genotype 1a, with more variable signals observed for other genotypes. These results point to improved antibody induction by N2 deletion under the current dosing and schedules. They are directionally consistent with engineering studies showing that removal or repositioning of specific E2 glycans can increase sensitivity to neutralization and broaden recognition [33]. At the same time, the extent of cross-genotype activity remained limited in this experiment and warrants broader panels and titration.

Cellular assays indicated a predominance of Th1-type responses after vaccination, with the highest IFN-γ ELISPOT values observed in the N2 group and low IL-4 levels across arms. CD8^+^ T cells from N1/N2 groups expressed more granzyme B and perforin than the wild-type. Although E2 is classically treated as a B-cell antigen, these findings may reflect platform- and processing-related effects (e.g., CpG/DNA sensing, expression kinetics, and antigen processing/presentation) in addition to glycan changes, thereby influencing T-cell responses [34,35,36]. Therefore, mechanistic interpretation should be considered correlative at this stage. Direct assays of antigen presentation and cytokine milieu would help define the mechanisms.

Functional consequences were further explored in two settings. First, splenocytes from immunized mice lysed CT26-sE2 targets in vitro, with a rank order of N2 > N1 > wild-type under matched E:T ratios. Second, in a tumor challenge using CT26-sE2, N2 vaccination slowed tumor growth compared with controls, and N1 showed a smaller delay. These models do not recapitulate liver infection but provide orthogonal evidence that the immune response elicited by N2 deletion has cytotoxic potential against E2-expressing cells.

We then examined therapeutic use in an HCV challenge model. Using the same plasmid immunogens as in the immunization experiments, sE2-N2 lowered blood HCV RNA and hepatic measures compared with sE2. The differences were observable at several time points and at the study endpoint. The magnitude of effect depended on schedule and dose, and the model has known constraints, yet the pattern again favored N2. Finally, a monoclonal antibody derived from the sE2-N2 regimen (1C1) showed in vitro neutralization and, after passive transfer, reduced viremia and liver signals in infected ICR^4R+^ mice. These observations suggest that the sE2-N2 regimen can elicit functional antibody specificities; however, we cannot yet assign a specific epitope class for 1C1. Defining the 1C1 epitope and its relationship to known neutralizing surfaces on E2, including the AR3 region, will be informative. Future work will include epitope-mapping strategies such as binding/competition profiling with well-characterized bnAb panels (including AR3-region antibodies), targeted mutagenesis or mutational scanning, and/or selection of escape variants to delineate the 1C1 footprint and clarify its relationship to established E2 neutralizing epitopes.

This study has several limitations. We evaluated single-site deletions, and combinatorial edits were not tested. Neutralization breadth was evaluated using a limited set of HCVcc strains, and several assays were performed at fixed serum dilutions rather than full titration series. In addition, the sample sizes and follow-up duration limited our ability to assess potency and durability, and the tumor/therapeutic models used do not fully recapitulate natural HCV infection. Because glycan occupancy and neutralization sensitivity vary across HCV genotypes, the advantages observed for N423 (N2) deletion in this study may not fully generalize to all circulating strains or exposure scenarios. Although the ICR4R^+^ model supports infection in vivo, it may not capture key aspects of human B- and T-cell immunity; therefore, broader genotype panels and more translationally predictive models will be required in future work. Secretion and ER co-localization argue against gross misfolding under our conditions but do not substitute for high-resolution structural validation. In addition, we did not systematically quantify and normalize antigen expression/secretion efficiency or in vivo antigen load across all glycan mutants; therefore, differences in antigen dose, expression kinetics, or stability may contribute to some of the observed immune differences. Finally, while WT and mutants were compared under identical DNA delivery conditions, the DNA platform and CpG content may shape the immune profile, and we cannot exclude contributions from differences in innate sensing, in vivo expression kinetics/antigen dose–time profiles, or antigen processing/presentation that may accompany sequence changes. Future studies comparing alternative vaccine platforms and/or directly assessing antigen dose–time profiles and antigen presentation will be required to strengthen glycan-specific causal inference.

Despite these constraints, the data outline a practical path. N2 deletion produced gains in antibody binding and neutralization, increased IFN-γ-biased cellular responses, improved CTL activity, slowed the growth of E2-expressing tumors, and enhanced therapeutic performance in an infection model while enabling isolation of a monoclonal antibody with in vivo activity [17,28]. Future work should expand strain coverage and titration series, map 1C1 and vaccine-elicited epitopes, integrate structural analysis of sE2-N2, and test combinations of glycan edits and presentation formats (e.g., nanoparticles or scaffolded E1E2) [37,38]. These steps will clarify breadth and durability and help align immunogen design with the neutralizing landscape of circulating HCV.

In summary, selective removal of the N2 glycan was associated with improved immunogenicity and a modest, genotype 1a–predominant neutralization signal under the conditions tested. This glycan editing appeared to preserve antigen integrity and was accompanied by strengthened humoral responses and measurable cellular immune readouts. Together with the activity observed for 1C1, these findings support rational glycan editing—beginning with N2—as a tractable approach to refine E2-based HCV vaccine candidates, while underscoring the need for structural confirmation and broader, more stringent efficacy testing across genotypes and experimental settings.

5. Conclusions

In conclusion, selective removal of the N2 glycan (N423) from the HCV E2 ectodomain maintained detectable expression/production features and ER localization in our assays and, under the conditions tested, was associated with stronger humoral and cellular immunity in mice. N2 vaccination increased anti-E2 antibody titers and showed greater inhibition of HCVcc at the tested dilutions, promoted IFN-γ-skewed cellular responses and CD8^+^ cytotoxic markers, enhanced in vitro CTL activity, and slowed the growth of E2-expressing tumors. In an infection setting using ICR^4R+^ mice, therapeutic sE2-N2 was associated with lower blood HCV RNA and reduced liver viral signals/measures compared with sE2. A monoclonal antibody (1C1) derived from the sE2-N2 regimen neutralized HCV in vitro and lowered viremia and hepatic signals after passive transfer. These findings support N2 glycan editing as a practical step for refining E2-based immunogens while motivating structural confirmation, broader strain coverage, and optimization of dosing and delivery in future studies.