Fusion of Complement Fragment C3d Enhances Germinal Center Responses to HIV-1 Envelope Glycoproteins

Abstract

Eliciting sustained germinal center (GC) responses is critical for the development of an effective HIV-1 vaccine, yet HIV-1 envelope glycoprotein (Env) immunogens often fail to elicit GC responses required for the maturation of cognate B cells that secrete broadly neutralizing antibodies (bNAbs). Effective antigen recognition is important for initial B cell priming, activation, and GC engagement. Since complement opsonization contributes to antigen recognition, we investigated whether C3d fusion could enhance the GC response of the stabilized HIV-1 Env immunogen based on a consensus sequence (ConM Env). Our results demonstrate that ConM Env-C3d induced potent HIV-specific B cell activation in vitro compared to ConM Env alone. We also observed that the C3d fusion enhanced antigen presentation by human tonsil-derived follicular dendritic cells (FDCs) to HIV-specific B cell lines. Moreover, mouse immunization studies combining ConM Env-C3d with the AddaS03 adjuvant revealed significantly enhanced early GC formation and prolonged antigen display and retention on FDCs for up to 56 days, highlighting improved antigen persistence within GCs. These immunological enhancements, including a more focused early antibody response, correlated with improved virus neutralization. Additionally, we observed sex-based differences in immune responses, with female mice showing stronger antibody responses and enhanced antigen retention compared to males. These findings suggest that C3d fusion can enhance GC engagement and improve the immunogenicity of HIV-1 vaccines.

Keywords: B cell activation; FDCs; HIV-1 Env; antigen retention; complement C3d; germinal center response; neutralizing antibodies.

Figure 1:. Conformation and antibody binding of ConM Env and ConM Env-hC3d.

(A) AlphaFold3 prediction of the ConM Env (purple)-hC3d (orange) fusion construct. The trimer base epitope is indicated with an arrow. The flexible glycine-serine (GS) linker (yellow) may allow for rotation of the hC3d subunits, potentially obscuring immunodominant trimer base epitopes. (B) Superposition of ConM Env-hC3d onto a hC3D-complement receptor 2 (CR2) structure (PDB: 3OED). The predicted angle of the hC3d domain allows for CR2 binding. (C)ELISA binding assay of complement receptor 2 (CR2)/CD21 with ConM Env, ConM Env-hC3d, recombinant hC3d, and the SARS-CoV-2 Wuhan Spike as a negative control. (D) ELISA binding curves for antibodies PGT145, PGT121, PGT151, PGDM1400, VRC01, 2G12, RM19R against the ConM Env (left) and ConM Env-hC3d (right) constructs, with SARS-CoV-2-binding mAb COVA1-16 as negative control.

Figure 2:. Activation of B cells expressing bNAb PGT121 by ConM Env and ConM Env-hC3d.

Ramos B cell activation of PGT121 IgG B cells as measured by calcium (Ca²⁺) flux assay using equimolar amounts of ConM Env, ConM Env-hC3d, or ConM Env supplemented with recombinant hC3d (ConM Env + hC3d) at concentrations of 10 µg, 1 µg, 0.1 µg, 0.01 µg, and 0.001 µg (Env-equivalent mass). ConM Env and recombinant hC3d were co-administered at a molar ratio matching the stoichiometry of ConM Env-hC3d. A baseline without antigen was established between 0–30 s, after which antigen was added to the B cells (30–50 s).

Figure 3:. FDC-mediated presentation of ConM Env-C3d to HIV-1 Env-specific B cells.

(A) Percentage of live PDPN⁺ CD21⁺ CD35⁺ FDCs positive for binding ConM Env-AF647 or ConM Env-hC3d-AF647, presented either as the antigen alone, within an immune complex (IC; 2G12 mAb), or within a complement-opsonized immune complex (CO-IC; active human serum + 2G12 mAb). (B) Schematic of antigen extraction by PGT121 B cells from human FDCs. Fluorescently labeled antigen, either alone or within an immune complex (IC), binds to FDCs via C3d/C3b–CD21 interactions or antibody Fc–FcγR engagement. After incubation, unbound antigen is removed by washing. HIV-specific PGT121 Ramos B cells are then added, followed by a second wash. Antigen extraction is assessed by flow cytometry, measuring fluorescent signal on HIV-1 Env-specific B cells. Schematic was created with BioRender.com. (C) Percentage of live CD19⁺ GFP⁺ PGT121-expressing Ramos B cells positive for binding ConM Env-AF647 or ConM Env-hC3d-AF647, presented by FDCs as described in the schematic of Figure 5B. Antigen was presented either alone, within an immune complex (IC; 2G12 mAb), or as a complement-opsonized IC (CO-IC; active human serum + 2G12 mAb). A SARS-CoV-2 spike-specific Ramos B cell line (COVA1-16) was included as a negative control to assess antigen specificity in the antigen-CO-IC condition. Bar plots depict the mean ± SD. Each dot represents one donor; paired comparisons were performed using two-tailed paired t-tests. Significance is indicated as: ns, not significant (p > 0.05); *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 4:. Humoral responses following immunization with ConM Env or ConM Env-mC3d, with or without AddaS03 adjuvant.

(A) Immunization and sampling schedule. Mice received 10 µg ConM Env or ConM Env-mC3d (equimolar) with or without AddaS03. Serum, lymph nodes (LNs), and spleens were collected at indicated time points. (B) ConM Env-specific IgG levels measured by multiplex binding assay at week 4 (left), week 8 (middle), and week 12 (right), plotted as area under the curve (AUC). Data are stratified by sex with male and female mice shown in blue and maroon, respectively. (C) ConM Env-specific IgG responses in female mice at week 12, measured by multiplex binding assay and presented as AUC. (D) Differential IgG binding ratio of ConM Env to V1V2 switch variant (ConM Env / ConM V1V2 switch), calculated from AUC values at week 8 and 12. Ratios >2.5 (shaded area) indicate a significant V1V2-focused response. (E) Differential IgG binding ratio of ConM Env to base-mutant (PAS50) variant (ConM Env / ConM PAS50) at week 8 and 12. Ratios >2.5 (shaded area) indicate a significant base-focused response. (H) Neutralization titers (ID50) against ConS pseudovirus at week 8 and (I) at the final bleed (week 12). Data are stratified by sex, with male (blue) and female (maroon) mice shown separately. Data are shown as scatter plots or box-and-whisker plots (min to max), with medians indicated. Each dot represents one mouse. Statistical comparisons were performed using two-tailed Mann–Whitney U tests. Significance: ns (p > 0.05); * (p < 0.05); ** (p < 0.01); *** (p < 0.001); **** (p < 0.0001).

Figure 5:. Antigen capture, retention, and GC B cell responses in draining lymph nodes following ConM Env and ConM Env-mC3d immunization.

(A) Immunization schedule of mice for analyzing antigen localization in draining lymph nodes (LNs) after immunization with ConM Env-AF647 and ConM Env-mC3d-AF647, each supplemented with AddaS03 adjuvant. Six mice per group were used, split evenly between males and females (3 each). Left side depicts the timeline for evaluating antigen presence at 24 hours and 7 days post-injection. On the right, a similar setup is used, with the addition of VRC01, an anti-HIV-1 Env-specific monoclonal antibody (mAb), injected intraperitoneally (i.p.) one day prior to subcutaneous administration of AF647-labeled antigen to assess formation and localization of antigen-immune complexes. (B) Representative percentage of live CD19⁺ CD95⁺ GL7⁺ GC B cells in lymph nodes from mice treated as described in Figure 5A, assessed at 24 h and 7 days post-immunization with ConM Env-AF647, ConM Env-mC3d-AF647, or ConM Env-AF647 with prior i.p. administration of VRC01 (ConM Env + VRC01). (C) Percentage of antigen-AF647⁺ FDCs (CD45⁻,B220⁻,CD31⁻,PDPN⁺,CD21/35^hi,Madcam1⁺) at 24 hours and 7 days post-immunization in mice treated with ConM Env-AF647, ConM Env-mC3d-AF647, and ConM Env-AF647 + VRC01. Sex differentiation is shown with females in maroon and males in light blue. (D) Immunization schedule of mice for assessing long-term antigen retention and effect of GCs in draining lymph nodes. Left: mice were immunized with ConM Env-AF647 or ConM Env-mC3d-AF647. Right: mice were immunized with ConM Env-AF647 or ConM Env-mC3d-AF647 and boosted on day 28 with their AF488-labeled counterparts. All immunizations included AddaS03 adjuvant. Mice were culled at day 56 (n=6 per group, 3 male and 3 female). (E) Percentage of total antigen-AF647⁺ FDCs (prime-specific) for mice treated as described in Figure 5D, assessed at 56 days post-immunization. Sex differentiation is shown with females in maroon and males in light blue. (F) Percentage of total antigen-AF488⁺ FDCs (boost-specific) for mice treated as described in Figure 5D, assessed at 56 days post-immunization. Sex differentiation is shown with females in maroon and males in light blue. (G) Percentage of ConM Env-specific CD19⁺ B cells assessed at 56 days post-immunization, for mice treated as described in Figure 5D. (H) Percentage of CD19⁺ CD95⁺ GL7⁺ GC B cells for mice treated as described in Figure 5D. (I) Median fluorescence intensity (MFI) of CR2/CD21 expression in female and male mice immunized as in Figure 5A, measured on all CR2⁺ cells and on FDCs (CD21^hi Madcam1⁺) from lymph nodes at 24 h and 7 days post-immunization (pooled). (J) Median fluorescence intensity (MFI) of CR2/CD21 expression in female and male mice immunized as in Figure 5D, measured on all CR2⁺ cells and on FDCs (CD21^hi Madcam1⁺) from lymph nodes at 56 days post-immunization. Data are shown as mean ± SEM, median with scatter, or box-and-whisker plots (min to max), as indicated. Each dot represents one mouse. Statistical comparisons were performed using two-tailed Mann–Whitney U tests, except in panels I and J (unpaired t-tests). Significance: ns (p > 0.05); * (p < 0.05); ** (p < 0.01); *** (p < 0.001).