Mannose-binding lectin and complement mediate follicular localization and enhanced immunogenicity of diverse protein nanoparticle immunogens

Abstract

Nanoparticle (NP) vaccine formulations promote immune responses through multiple mechanisms. We recently reported that mannose-binding lectin (MBL) triggers trafficking of glycosylated HIV Env-immunogen NPs to lymph node follicles. Here, we investigate effects of MBL and complement on NP forms of HIV and other viral antigens. MBL recognition of oligomannose on gp120 nanoparticles significantly increases antigen accumulation in lymph nodes and antigen-specific germinal center (GC) responses. MBL and complement also mediate follicular trafficking and enhance GC responses to influenza, HBV, and HPV particulate antigens. Using model protein nanoparticles bearing titrated levels of glycosylation, we determine that mannose patches at a minimal density of 2.1 × 10^-3 mannose patches/nm² are required to trigger follicular targeting, which increases with increasing glycan density up to at least ∼8.2 × 10^-3 patches/nm². Thus, innate immune recognition of glycans has a significant impact on humoral immunity, and these findings provide a framework for engineering glycan recognition to optimize vaccine efficacy.

Keywords: lymph node trafficking; mannose-binding lectin; nanoparticle; vaccine.

Figure 1.. eOD-GT8 60mer nanoparticles accumulate in lymph nodes and localize to the FDC network of follicles in an MBL-dependent manner

(A) C57Bl/6 or MBL KO mice (n = 4–5/group) were immunized with 2 μg eOD equivalent AlexaFluor 647-labeled eOD-GT8 60mer and saponin adjuvant. Three or seven days post immunization, draining inguinal lymph nodes were excised for whole-tissue fluorescence imaging. Error bars indicate SEM; *, p <0.05 by Mann-Whitney test. (B) C57Bl/6 or MBL KO mice (n = 5/group) were immunized with 2 μg eOD equivalent AlexaFluor 647-labeled eOD-GT8 60mer or eOD monomer and saponin adjuvant. One or three days post immunization, lymph nodes were snap-frozen and cryosectioned into six sections each for confocal imaging. Blue, CD35; red, eOD-GT8; scale bars denote 100 μm. (C) C57Bl/6 or MBL KO mice (n = 4/group) were immunized with 2 μg AF647-labeled eOD or 2 mg eOD equivalent AF647-labeled eOD-GT8 60mer and saponin adjuvant. Draining lymph nodes were harvested on day 7 and cleared for confocal imaging. Shown are the percentage of antigen-positive follicles among all follicles within each individual draining lymph node. Error bars indicate SEM; ****, p <0.0001; ns, not significant by Mann-Whitney test.

Figure 2.. MBL KO mice have reduced germinal center and antibody responses to eOD-60mer immunization

C57Bl/6 or MBL KO mice (n = 10/group) were immunized with 2 μg eOD-equivalent eOD-GT8 60mer and saponin adjuvant or PBS control. (A–G) Flow cytometry analysis of GC B cell responses in draining inguinal lymph nodes 12 days post immunization. Shown are representative flow cytometry plots gating for antigen-specific B cells (A), absolute counts of B220⁺eOD⁺CD4⁻ antigen-specific B cells (B), representative plots of GC B cells (C) and antigen-specific GC B cells (D), absolute counts of total B220⁺GL7⁺CD4⁻CD38^low GC B cells (E), absolute counts of total B220⁺GL7⁺eOD⁺CD4⁻CD38^low antigen-specific GC B cells (F), and representative histograms of eOD signal MFI among all cells (gray) and antigen-specific GC B cells (blue) and the antigen-specific GC B cell eOD signal MFI from each sample (G). Error bars indicate SEM; **, p <0.01; ns, not significant by Mann-Whitney test. (H) Serum eOD-specific IgG titers over time in mice immunized with eOD-GT8 60mer. Error bars indicate SEM; *, p <0.05; **, p <0.01; ***, p <0.0001 relative to WT by one-way ANOVA followed by Tukey post hoc test.

Figure 3.. High-mannose glycans are required for MBL-driven follicular accumulation of eOD-GT8 nanoparticles

(A) BLI analysis of eOD-GT8 60mer glycan variants binding to immobilized recombinant murine MBL2 as a function of eOD particle concentration. (B–D) C57Bl/6 mice (n = 5/group) were immunized with 2 μg eOD equivalent eOD-GT8 60mer glycan variants and saponin adjuvant. Shown are average intensity Z projections through 360 μm of cleared draining lymph nodes harvested on day 7 (B, blue, CD35; red, eOD-GT8 60mer; scale bars denote 500 μm), and analyses of normalized total eOD-GT8 60mer signal per Z plane of cleared lymph nodes (C) and percent eOD-60mer signal found within follicles (D). Error bars indicate SEM; points represent average values between paired draining lymph nodes from one animal; *, p <0.05; ***, p <0.001; ****, p <0.0001, ns = not significant by one-way ANOVA followed by Tukey post hoc test.

Figure 4.. HA-8mer follicular accumulation and immunogenicity are dependent on MBL

(A and B) C57Bl/6 or MBL KO mice (n = 5/group) were immunized with 5 μg AlexaFluor 647-labeled influenza HA-8mer particles and saponin adjuvant. Draining lymph nodes were harvested on day 7 and cleared for confocal imaging. Shown are total HA-8mer signal per Z plane of cleared tissues (A) and percent HA-8mer signal found within follicles (B). Error bars indicate SEM; points represent average values between paired draining lymph nodes from one animal; **, p <0.01, ns = not significant by Mann-Whitney test. (C) C57Bl/6 or MBL KO mice (n = 5/group) were immunized with 5 μg influenza HA-8mer particles and saponin adjuvant. Shown are serum hemagglutinin-specific IgG titers 5 and 6 weeks post immunization. Error bars indicate SEM; **, p <0.01 by Mann-Whitney test. (D and F) Absolute counts of total GC B cells (D, *, p <0.05 by Mann-Whitney test) and absolute counts of antigen-specific GC B cells at day 12 (E, **, p <0.01 by Mann-Whitney test) and average MFI of antigen specificity stain among antigen-specific GC B cells (F, p = 0.39 by Mann-Whitney test).

Figure 5.. HPV16 L1 and HBsAg nanoparticles exhibit complement-dependent follicular accumulation and immunogenicity

(A and B) BLI binding curves of unmodified and PNGase F-treated HPV16 L1 (A) or unmodified HBsAg (B) to immobilized recombinant murine MBL2 as functions of antigen concentration. (C) C57Bl/6 mice or MBL KO mice (n = 5/group) were immunized with 0.1 μg AlexaFluor 647-labeled HPV16 L1 and saponin adjuvant. Seven days later, lymph nodes were harvested, cleared, and imaged by confocal microscopy. Shown are average intensity Z projections through 360 μm of tissue; shown is staining for CD35 (blue) and antigen (red), scale bars denote 500 μm. (D) Serum HPV16 L1-specific IgG titers over time in mice (n = 5/group) immunized with 0.1 μg HPV16 L1 and saponin adjuvant. Error bars indicate SEM, p = 0.92 compared with WT one-way ANOVA. (E) Absolute counts of germinal center B cells (B220⁺GL7⁺CD4⁻CD38^low) and antigen-specific germinal center B cells (B220⁺GL7⁺HPV16 L1⁺CD4⁻CD38^low) from WT and MBL KO mice (n = 5/group) at day 12 following immunization with 0.1 μg HPV16 L1 and saponin adjuvant. Error bars indicate SEM; *, p < 0.05 by Mann-Whitney test. (F) C57BL/6 mice or MBL KO mice (n = 5/group) were immunized with 5 μg AlexaFluor 647-labeled HBsAg and saponin adjuvant. Seven days later, lymph nodes were harvested, cleared, and imaged by confocal microscopy. Shown are average intensity Z projections through 360 μm of tissue; shown is staining for CD35 (blue) and antigen (red), scale bars denote 500 μm. (G) Serum HBsAg-specific IgG titers over time in mice immunized with 5 μg HBsAg and saponin adjuvant. Error bars indicate SEM; *, p <0.05 compared with WT by one-way ANOVA followed by Tukey post hoc test. (H) Absolute counts of germinal center B cells (B220⁺GL7⁺CD4⁻CD38^low) and antigen-specific germinal center B cells (B220⁺GL7⁺HBsAg⁺CD4⁻CD38^low) from WT and C3 KO mice 12 days after immunization with 5 μg HBsAg and saponin adjuvant. Error bars indicate SEM; *, p <0.05 by Mann-Whitney test.

Figure 6.. Differentially glycosylated nanoparticles accumulate in follicles in a mannose density-dependent manner

(A) Design models of glycosylated I53–50 nanoparticles with either 240 glycans (left) or 120 glycans (right) displayed on the particle. The left particle is assembled with 20 glycosylated I53–50A trimeric subunits (protein in gray and glycans in green) and 12 non-glycosylated I53–50B pentameric subunits (orange) to display 240 glycans on the particle. The right particle is assembled with 10 glycosylated and 10 non-glycosylated I53–50A trimeric subunits, and 12 non-glycosylated I53–50B pentameric subunits to display 120 glycans. The two-component nature of I53–50 particles (i.e., each particle is composed of 20 trimers and 12 pentamers) enabled titration of glycan densities on the particle through varying the molar ratio of non-glycosylated to glycosylated I53–50A trimeric subunits; glycosylation of I53–50A trimers was either native or high-mannose for each particle formulation. (B) Mean glycan distances calculated from the particle structure for NPs with titrated levels of total glycans. (C) BLI analysis of serially glycosylated I53–50 nanoparticles binding to immobilized recombinant murine MBL2 as a function of I53–50 nanoparticle concentration. (D) The apparent dissociation constant (K_D) of immobilized murine MBL2 binding to each I53–50 high-mannose glycoform was determined by BLI analysis using a global 1:1 binding model applied to the three highest I53–50 concentrations. (E and F) C57Bl/6 mice (n = 5/group) were immunized with 5 μg I53–50 high-mannose glycan variants and saponin adjuvant. Shown are average intensity Z projections through 360 μm of cleared draining lymph nodes harvested on days 3 and 7 (E, blue, CD35; red, I53–50; scale bars denote 500 μm), and quantification of the percent I53–50 signal found within follicles (F). Error bars indicate SEM; points represent average values between paired draining lymph nodes from one animal; *, p <0.05; ****, p <0.0001, ns = not significant by one-way ANOVA followed by Tukey post hoc test. (G) Absolute counts of germinal center B cells and antigen-specific germinal center B cells from WT and MBL KO mice 12 days after immunization with 5 μg I53–50 and saponin adjuvant. Error bars indicate SEM; *, p <0.05; ns = not significant by one-way ANOVA followed by Tukey post hoc test.