A Rapid-Response Humoral Vaccine Platform Exploiting Pre-Existing Non-Cognate Populations of Anti-Vaccine or Anti-Viral CD4+ T Helper Cells to Confirm B Cell Activation

Abstract

The need for CD4+ T cell responses to arise de novo following vaccination can limit the speed of B cell responses. Populations of pre-existing vaccine-induced or anti-viral CD4+ T cells recognising distinct antigens could be exploited to overcome this limitation. We hypothesise that liposomal vaccine particles encapsulating epitopes that are recognised, after processing and B cell MHCII presentation, by pre-existing CD4+ T cells will exploit this pre-existing T cell help and result in improved antibody responses to distinct target antigens displayed on the particle surface. Liposomal vaccine particles were engineered to display the malaria circumsporozoite (CSP) antigen on their surface, with helper CD4+ epitopes from distinct vaccine or viral antigens contained within the particle core, ensuring the B cell response is raised but focused against CSP. In vivo vaccination studies were then conducted in C57Bl/6 mice as models of either vaccine-induced pre-existing CD4+ T cell immunity (using ovalbumin-OVA) or virus-induced pre-existing CD4+ T cell immunity (murine cytomegalovirus-MCMV). Following the establishment of pre-existing by vaccination (OVA in the adjuvant TiterMax® Gold) or infection with MCMV, mice were administered CSP-coated liposomal vaccines containing the relevant OVA or MCMV core CD4+ T cell epitopes. In mice with pre-existing anti-OVA CD4+ T cell immunity, these vaccine particles elicited rapid, high-titre, isotype-switched CSP-specific antibody responses-consistent with the involvement of anti-OVA T helper cells in confirming activation of anti-CSP B cells. Responses were further improved by entrapping TLR9 agonists, combining humoral vaccination signals 'one', 'two' and 'three' within one particle. Herpes viruses can establish chronic infection and elicit significant, persistent cellular immune responses. We then demonstrate that this principle can be extended to re-purpose pre-existing anti-MCMV immunity to enhance anti-CSP vaccine responses-the first description of a strategy to specifically exploit anti-cytomegalovirus immunity to augment vaccination against a target antigen.

Fig 1. Vaccine mechanism of action.

Pre-existing non-cognate CD4+ T cell help is exploited to enhance antibody responses to target antigens. B-cells with BCRs specific for the antigen on the particle surface are engaged by vaccine particles. Following internalisation, endosomal proteases generate peptide fragments from both the surface antigen and the helper antigen entrapped within the vaccine particle. These may be presented on MHC class II to cognate T cells (i.e. those that recognise the same antigen as the BCR—which may be rare for weak antigens) and non-cognate T cells (i.e. those that recognise the strong helper MHCII epitopes). Pre-existing CD4+ T cells that recognise the helper antigen (generated by pre-vaccination or infection) can provide costimulatory signals to the B-cell, resulting in confirmation of antibody production. These signals drive B-cell proliferation, differentiation, antibody production, somatic hypermutation, and isotype switching. Because the specificity of this antibody response was determined at the point of BCR-mediated antigen recognition, the antibody produced will be directed against the target antigen on the vaccine particle surface, and not against the core helper components.

Fig 2. Pre-existing CD4+ T cell immunity to epitopes contained within the core of liposomal vaccine particles enhances the antibody response to non-cognate target antigens on the particle surface.

Mice (n = 4) were pre-vaccinated with either PBS or OVA_323-339 in TMG at week 0 and week 2, prior to vaccination with CSP(OVA_323-339) liposomes at week 4. The effect of this pre-existing immunity to OVA_323-339 on anti-CSP antibody responses was measured over weeks 4 to 8 and compared using two-way ANOVA with Bonferroni’s post-test. Asterisks indicate the statistical significance of the difference between OVA_323-339 + TMG (black circles, solid black line) and PBS + TMG (grey squares, solid grey line) pre-vaccinated groups. This difference was found reproducibly in multiple experiments and a typical result is shown (b). Anti-OVA_323-339 responses were also measured in both PBS and OVA_323-339 pre-vaccinated mice administered CSP(OVA_323-339) liposomes (c). The effect of liposomal particle dose was examined by vaccinating mice (n = 4) with five-fold dilutions of CSP(OVA_323-339) vaccine (the highest concentration containing 50 μg of CSP and 10 μg of OVA_323-339). Anti-CSP responses were compared using a one-way ANOVA with Tukey’s post-test for multiple comparisons (d). Pre-vaccination with PBS or OVA_323-339 in TMG was then performed either intramuscularly or subcutaneously. Subsequently, CSP(OVA_323-339) liposomal vaccines were then administered either subcutaneously (e) or intramuscularly (f) and mean anti-CSP responses were compared over time using two-way ANOVA with Bonferroni’s post test. Asterisks indicate the difference between groups shown in blue and grey groups at the indicated time points.

Fig 3. Pre-existing CD4+ T cell immunity to non-cognate core epitopes drives the rapid production of isotype-switched, higher avidity antibodies against target antigens.

CSP-specific IgG1 (a), IgG2b (b), IgG2c (c) and IgG3 (d) responses to vaccination with CSP(OVA_323-339) were quantified in mice (n = 4) pre-vaccinated with either PBS + TMG or OVA_323-339 + TMG using two-way ANOVA with Bonferroni’s post-test. (e) The anti-OVA_323-339 IgG1/IgG2c response pattern at the time of liposomal vaccination was compared with the anti-CSP IgG1/IgG2c response four weeks after CSP(OVA_323-339) vaccination in PBS + TMG or OVA_323-339 + TMG pre-vaccinated mice (n = 4) using unpaired, two-tailed t tests. If Welch’s correction is used for unequal variances, the significance of difference between the anti-OVA_323-339 IgG1 and IgG2c responses reduces to * (p < 0.05). (f) The concentration of the chaotropic agent sodium thiocyanate required to decrease the anti-CSP ELISA signal by 50% was determined in serum from mice (n = 4) pre-vaccinated with either PBS or OVA_323-339 in TMG four weeks after vaccination with CSP(OVA_323-339) liposomes and compared with unpaired, two-tailed t tests.

Fig 4. Particle-encapsulated TLR9 agonists further enhance immunogenicity.

Liposomal particles were produced to contain CpG DNA TLR9 agonists in addition to OVA_323-339 peptides. Mice (n = 4) were vaccinated with either PBS or OVA_323-339 in TMG and then administered CSP(OVA_323-339) or CSP(OVA_323-339+CpG) liposomes. The total anti-CSP antibody response was compared between groups (a). The isotype of the anti-CSP response was quantified for IgG1 (b), IgG2b (c), and IgG2c (d). Mean responses in groups pre-vaccinated with OVA_323-339 and then vaccinated with liposomes with or without CpG were compared (solid and dashed black lines) using two-way ANOVA with Bonferroni’s post test. Asterisks indicate the difference between groups shown in blue and grey groups at the indicated time points.

Fig 5. Non-cognate anti-viral CD4+ T cell populations in mice chronically infected with murine cytomegalovirus can enhance antibody production against target antigens.

Mice (n = 4) were infected with MCMV. IFNγ and IL-4 responses to the MCMV m09 peptide were measured by ELISPOT assay in splenocytes from mice infected with MCMV for 8 weeks, or non-MCMV-infected matched controls and compared using unpaired, two-tailed t tests (a). 8 weeks post-infection, mice—including uninfected matched controls—were vaccinated with CSP-coated liposomes containing the m09 peptide. Each vaccine dose contained 0.5 μg of CSP and 0.1 μg of m09 peptide. The total anti-CSP response and CSP-specific IgG1 (c), IgG2b (d), and IgG2c (e) were quantified. Responses were compared using two-way ANOVA with Bonferroni’s post-test. The IgG1/IgG2c profile of the anti-MCMV response was also assessed by ELISA (n = 2) and is presented alongside anti-CSP IgG1 and anti-CSP IgG2c levels from vaccinated infected and uninfected mice (n = 4) with levels compared using unpaired, two-tailed t tests (f).