The use of self-adjuvanting nanofiber vaccines to elicit high-affinity B cell responses to peptide antigens without inflammation

Abstract

Balancing immunogenicity with inflammation is a central tenet of vaccine design, especially for subunit vaccines that utilize traditional pro-inflammatory adjuvants. Here we report that by using a nanoparticulate peptide-based vaccine, immunogenicity and local inflammation could be decoupled. Self-assembled β-sheet-rich peptide nanofibers, previously shown to elicit potent antibody responses in mice, were found to be non-cytotoxic in vitro and, remarkably, elicited no measurable inflammation in vivo-with none of the swelling at the injection site, accumulation of inflammatory cells or cytokines, or production of allergic IgE that were elicited by an alum-adjuvanted vaccine. Nanofibers were internalized by dendritic cells and macrophages at the injection site, and only dendritic cells that acquired the material increased their expression of the activation markers CD80 and CD86. Immunization with epitope-bearing nanofibers elicited antigen-specific differentiation of T cells into T follicular helper cells and B cells into germinal center cells, as well as high-titer, high-affinity IgG that cross-reacted with the native protein antigen and was neutralizing in an in vitro influenza hemagglutination inhibition assay. These responses were superior to those induced by alum and comparable to those induced by complete Freund's adjuvant. Thus, nanoparticulate assemblies may provide a new route to non-inflammatory immunotherapies and vaccines.

Keywords: Alum adjuvant; Nanoparticle vaccines; Non-reactogenic; OVA(323-339) peptide; Self assembly.

Figure 1. Structure and cytotoxicity of Q11-based materials compared to Imject Alum

(a) OVA-Q11 fibers and (b) Imject Alum observed by TEM. (c) J774.1 macrophages exposed to Alum (top) or Q11 (bottom) at 10 % of their standard adjuvant concentrations for 4 hr, with viable (green) and non-viable (red) cell staining. (d) Percentages of non-viable J774.1 macrophages after 4 hr exposure to Alum or Q11. Concentration is given as the fraction relative to the standard adjuvant concentrations used for mouse immunizations (40 mg/mL total Al/Mg salt for Imject Alum; 2 mM, or 7 mg/mL, for Q11 peptide).

Figure 2. OVAQ11 immunization does not generate detectable local inflammation

(a) Injection of mouse footpads with OVAQ11 did not cause inflammation, whereas pOVA-Alum caused significant swelling and acute inflammation and necrosis (white arrows in H & E-stained tissue sections collected at day 8). Scale bars are 200 μm. Schematics at the right illustrate the immunization conditions for the pOVA antigen assembled into nanofibers via the Q11 assembly domain or adsorbed on alum. N = 3 mice per group. (b) Cellular and cytokine responses were analyzed in lavage fluid 20 hr after intraperitoneal injection. (top) Phenotyping by flow cytometry indicated that OVAQ11 did not recruit inflammatory cells, similar to unadjuvanted pOVA or PBS, whereas pOVA-Alum recruited significantly higher numbers of inflammatory cells. N = 3 mice per group, one of two independent experiments is shown. (bottom) Bead-based immunoassay showed no increase in inflammatory cytokines after OVAQ11 immunization, whereas alum immunization induced 5 out of 6 cytokines in the panel. Abbreviations: Macrophages (mac), conventional dendritic cells (cDC), inflammatory DCs (iDC), plasmacytotoid DC (pDC), neutrophils (neut), inflammatory monocytes (iMono), eosinophils (eosi); monocyte chemotactic protein-1 (MCP-1), keratinocyte-derived chemokine (KC), granulocyte colony-stimulating factor (G-CSF), interleukin (IL). All error bars show mean ± 1 std. dev. *, p < 0.05 compared to OVAQ11 group (1-way ANOVA with Tukey post-hoc tests).

Figure 3. OVAQ11 is internalized by antigen presenting cells and induces targeted dendritic cell activation in vivo

(a) Gating strategy for CD11c⁺ F4/80⁻ MHCII⁺ dendritic cells (DC) and CD11c⁻F4/80⁺ macrophages (M) after i.p. immunization. (b) Uptake of pOVA-fluorescein or OVAQ11-fluorescein 20 hr after i.p. immunization. Both DCs and Macrophages acquired OVAQ11 in greater quantities (brighter fluorescein staining) and in a larger fraction of the cells than pOVA (pooled cells from n = 2 per group; representative of two independent experiments). (c) Internalization of OVAQ11 by DCs. Cells from the lavage fluid were stained by allophycocyanin-conjugated anti-fluorescein antibody (surface staining) or were pre-blocked with anti-fluorescein antibody, fixed, permeabilized, and finally stained by allophycocyanin-conjugated anti-fluorescein antibody (intracellular staining). Cells were double-positive for allophcocyanin and fluorescein only after permeabilization, indicating that all of the OVAQ11 was intracellular (pooled cells from n = 3 per group). (d) Activation status after OVAQ11 uptake. Only DCs that were positive for fluorescent-OVAQ11 had an increase in CD80 and CD86 expression; neither DCs from the same sample that were negative for OVAQ11 nor macrophages had increased expression of these markers (n = 3 per group).

Figure 4. OVAQ11 induces CD4+ T-cell differentiation into T follicular helper and cytokine-secreting cells

Using an OTII adoptive transfer model, (a) pOVA-specific Tfh development was measured by expression of the markers PD-1 and CXCR5 3.5 d after immunization with the indicated condition (n = 3; representative of two independent experiments). (b) IL-4 and IFN-γ production were quantified by ELISPOT assays in response to stimulation with pOVA. Cells were collected from draining lymph nodes 3.5 and 7 d after immunization in the OTII cell transfer system (left; n = 3 mice per time point, pooled to n = 6 for analysis; representative of two independent experiments), or at week 9 in the endogenous system (right; n = 3). Data represent mean + 1 SD. *, p < 0.05 compared to OVAQ11 group (Student’s T test or 2-way ANOVA with Tukey post-hoc tests).

Figure 5. OVAQ11 but not alum induces a germinal center B cell response that correlates with high-titer IgG

(a) Endogenous OVA-specific GC-B cells were detectable by day 14 post-immunization with OVAQ11, but not with pOVA-Alum. B cells labeled with biotin-pOVA were detected by double-positive staining with streptavidin-PE and streptavidin-APC. GC B cells within this population were identified by double-positive staining with GL-7 and FAS, quantified at the right as a percentage of the total pOVA-binding B cells. *, p < 0.05 (1-way ANOVA with Tukey post-hoc tests). Pooled data from two independent experiments. (b) IgG and IgM signals against pOVA were strongest after immunization with OVAQ11 and OVAQ11-Alum (left); OVA protein-Alum raised a strong response against the whole protein (upper right). Only alum-adjuvanted vaccines elicited a detectable IgE response at week 9 (lower right; *, p < 0.05 versus OVAQ11, 1-way ANOVA with Tukey post-hoc tests). Mice were immunized subcutaneously (s.c.) at week 0 and boosted with half-doses at weeks 4 and 8 (gray arrows). N = 5 mice per group. Error bars show mean + 1 SD. Representative of at least two independent experiments.

Figure 6. Antibodies induced by OVAQ11 are high affinity and neutralizing, similar to those raised by pOVA with CFA

(a) Schematic of the SPR experiment, measuring the mass bound to a pOVA-conjugated surface as first serum solution and then buffer were flowed over it. (b) Antibodies induced by OVAQ11 or OVAQ11-Alum bound in similar quantities as those induced by pOVA-CFA; R₀ is the quantity of bound antibody (measured in response units, RU) after 3 min of serum flow, and this value scaled linearly with the concentration of pOVA used to coat the surface. (c) Dissociation rates (k_off, s⁻¹) were not significantly different between OVAQ11, OVAQ11-Alum, and pOVA-CFA sera. Measurements of k_off were not possible for the pOVA-Alum group because of its low binding. N = 5 mice per group; representative of two independent experiments. Error bars show mean ± 1 SD. Significance tested by 1-way ANOVA with Tukey post-hoc tests. (d) To test the functionality of the antibodies, mice were primed and boosted according to the schedule shown, and haemagglutination inhibition (HAI) assays were performed by incubating the resultant sera with a strain of influenza A engineered to express pOVA in head of the hemagglutinin protein (pOVA-HA; [29] or a control (wildtype) strain. (e) HAI titer was defined as the highest dilution at which agglutination of chicken red blood cells by the influenza virus was inhibited by the sera. OVAQ11 and pOVA-CFA both induced neutralization of pOVA-flu in a majority of immunized mice; no response was observed to the control flu. *, p < 0.05 compared to a minimal protective titer of 32 (dashed line), by single sample t-test. Data is from one of two similar independent experiments.