Lymph node targeted multi-epitope subunit vaccine promotes effective immunity to EBV in HLA-expressing mice

Abstract

The recent emergence of a causal link between Epstein-Barr virus (EBV) and multiple sclerosis has generated considerable interest in the development of an effective vaccine against EBV. Here we describe a vaccine formulation based on a lymph node targeting Amphiphile vaccine adjuvant, Amphiphile-CpG, admixed with EBV gp350 glycoprotein and an engineered EBV polyepitope protein that includes 20 CD8⁺ T cell epitopes from EBV latent and lytic antigens. Potent gp350-specific IgG responses are induced in mice with titers >100,000 in Amphiphile-CpG vaccinated mice. Immunization including Amphiphile-CpG also induces high frequencies of polyfunctional gp350-specific CD4⁺ T cells and EBV-specific CD8⁺ T cells that are 2-fold greater than soluble CpG and are maintained for >7 months post immunization. This combination of broad humoral and cellular immunity against multiple viral determinants is likely to provide better protection against primary infection and control of latently infected B cells leading to protection against the development of EBV-associated diseases.

Fig. 1. Design of a lymph node-targeted subunit vaccine for EBV.

a Schematic representation of EBVpoly protein, showing the “beads on a string” structure. EBVpoly is a polyprotein containing 20 CD8⁺ T cell epitopes from eight EBV antigens. The peptide sequence, HLA type and source antigen for each of the epitopes are listed in b with the proteasome liberation sequence highlighted in red. c Schematic representation of the adjuvant, AMP-CpG. d Vaccine design and mechanism of vaccination including delivery to the lymph nodes following subcutaneous injection. EBVpoly is an engineered multi-epitope protein immunogen consisting of multiple HLA-restricted T cell target epitopes; gp350 is a predominant viral target for neutralizing antibody responses; AMP-CpG is a lymph node-targeted TLR-9 agonist. AMP-CpG binds to endogenous albumin at the injection site and albumin chaperones AMP-CpG into the lymph nodes in concert with passive transport of EBVpoly and gp350. Lymph node APCs internalize and process EBVpoly and the individual epitopes are presented on HLA Class I to CD8⁺ T cells. Created with BioRender.com.

Fig. 2. EBVpoly stimulation of human PBMCs expands EBV-specific CD8⁺ T cells.

a PBMCs from healthy EBV seropositive patients were stimulated with EBVpoly for 1 hour, then expanded for 14 days in the presence of IL-2. Expanded PBMCS were re-stimulated with the individual antigen peptides from EBVpoly in an ICS assay. Shown are frequencies of IFNγ, IL-2, TNFα, and/or CD107⁺ CD8⁺ T cells. White bars indicate no peptide stimulation. Colored bars represent responses to each annotated peptide designated based on the first three amino acid residues of the cognate epitope. b Shown are representative flow cytometry dot plots of IFNγ vs TNFα for each donor demonstrating responses to the designated peptide re-stimulation. Data are representative of one experiment. Source data are provided as a Source Data file.

Fig. 3. AMP-CpG enhances delivery of EBVpoly to the lymph node alongside comprehensive immune activation.

C57Bl/6J mice were immunized with 8 µg EBVpoly-AF594 and 10 µg gp350-AF647 admixed with 1.2 nmol soluble or AMP-CpG (n = 6 mice per group, 2 lymph nodes per mouse). a–d Quantification of total radiant efficiency in inguinal and axillary lymph nodes analyzed ex vivo by IVIS at a 24 and b 48 h post primer dose. Mock treatments represent fluorescence-negative controls collected from mice receiving equivalent amounts of AMP-CpG and unlabeled antigens. c, d IVIS images of five representative lymph nodes in a and b. e, f Quantification of cytokine concentrations in lymph nodes from a and b by Luminex. Depicted are mock-subtracted average Z-scores of protein analyte concentrations in lymph nodes. Cytokines are clustered into functional groups: (1) growth factors, (2) Th1/inflammatory cytokines, (3) Th2/regulatory cytokines, (4) chemokines, and (5) inflammasome-associated cytokines. Mock-treated animals were administered vehicle alone. Data are representative of one experiment. Values depicted are mean ± standard deviation. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, e, f ANOVA, Dunnett multiple comparisons test, FDR 2-stage, step-up Benjamini, Krieger, and Yukatiele analysis, Q = 5%. Source data are provided as a Source Data file.

Fig. 4. Vaccination with AMP-CpG induces robust polyfunctional EBV-specific T cell responses in splenocytes.

a–h HLA-B*35:01 transgenic mice (n = 6 per vaccine group and n = 4 per control group) were immunized at weeks 0, 3 and 6 with 40 μg EBVpoly and 10 μg gp350 proteins admixed with 1.2 nmol soluble or AMP-CpG and T cell responses were analyzed at week 7. Control groups were immunized with soluble CpG or AMP-CpG alone. Splenocytes were stimulated with a–d EBV CD8⁺ T cell peptides or e–h gp350 OLPs in an ICS assay. a, e Shown are frequencies of IFNγ, IL-2, TNFα, and IFNγ TNFα double positive CD8⁺ or CD4⁺ T cells, with b, f corresponding representative dot plots. c, g Pie chart representations of the functional T cell profiles. Pies represent the capacity of T cells to secrete any (1, 2, or 3) of the three cytokines IFNγ, TNFα, and IL-2. d, h Frequencies of cytokine⁺ d CD8⁺ T cells and h CD4⁺ T cells in splenocytes to indicated HLA transgenic mice. Data are representative of one experiment. Values depicted are mean ± standard deviation. *p < 0.05; **p < 0.01 by two-sided Mann-Whitney test applied to cytokine⁺ T cell frequencies. If statistics are not indicated, then the comparison was not significant. The exact p-values are as follows. a Soluble vaccine to soluble CpG-0.0095, AMP vaccine to soluble vaccine-0.0087, and AMP vaccine to AMP CpG-0.0095. d Soluble vaccine to soluble CpG for B*35:01-0.0095, AMP vaccine to soluble vaccine for A*24:01-0.0043, AMP vaccine to AMP-CpG for A*24:01-0.0190, AMP vaccine to AMP-CpG for B*35:01-0.0095, and AMP vaccine to soluble vaccine for B*35:01-0.0087. e Soluble vaccine to soluble CpG-0.0095, AMP vaccine to soluble vaccine-0.0087, and AMP vaccine to AMP CpG-0.0095. h Soluble vaccine to soluble CpG for A*24:01-0.0095, soluble vaccine to soluble CpG for B*35:01-0.0095, AMP vaccine to soluble vaccine for A*02:01-0.0152, AMP vaccine to soluble vaccine for B*35:01-0.0087, AMP vaccine to AMP-CpG for A*24:01-0.0095, AMP vaccine to AMP-CpG for B*35:01-0.0095, AMP vaccine to AMP-CpG for A*02:01-0.0095, and AMP vaccine to AMP-CpG for B*08:01-0.0095. Source data are provided as a Source Data file.

Fig. 5. AMP-CpG induces strong serum IgG and neutralizing antibody responses targeting gp350.

HLA-B*35:01 transgenic mice (n = 6 per vaccine group and n = 4 per control group) were immunized at weeks 0, 3, and 6 with 40 μg EBVpoly and 10 μg gp350 proteins admixed with 1.2 nmol soluble or AMP-CpG. Control groups were immunized with soluble CpG or AMP-CpG alone. Humoral responses to gp350 were assessed in splenocytes and serum from immunized mice at week 7 by ASC ELISPOT, ELISA or neutralization assay. Shown are a ex vivo ASC ELISPOT measured frequency of gp350-specific ASCs per 3 × 10⁵ splenocytes, b expanded ASC ELISPOT measured frequency of gp350-specific ASCs per 3 × 10⁵ splenocytes, c longitudinal serum IgG titers, d Ig subtype titers at week 7 in pooled serum samples, and e EBV-neutralizing antibody titers in EBV-seropositive and EBV-seronegative donors alongside longitudinal EBV neutralization titers in pooled serum samples (n = 6, serum samples pooled per vaccine group). Values depicted are mean ± standard deviation. Not detected (n.d.) values are shown on the baseline. Dotted line indicates the detected average level of EBV NT₅₀ among EBV seropositive subjects. Data are representative of one experiment. *p < 0.05; **p < 0.01 by two-sided Mann-Whitney test. If statistics are not indicated, then the comparison was not significant. The exact p-values are as follows. a Soluble vaccine to soluble CpG-0.0095 and AMP vaccine to AMP-CpG-0.0095. b Soluble vaccine to soluble CpG-0.0095, AMP vaccine to AMP CpG-0.0095 and AMP vaccine to soluble vaccine-0.030. c AMP vaccine to soluble vaccine at 4 weeks-0.0260, AMP vaccine to soluble vaccine at 7 weeks-0.0087. Source data are provided as a Source Data file.

Fig. 6. Vaccination with AMP-CpG maintains EBV-specific CD8⁺ and CD4⁺ T cells responses for >7 months.

a–f HLA-B*35:01 transgenic mice (n = 9 per vaccine group and n = 5 per control group) were immunized at weeks 0, 3, and 6 with 40 μg EBVpoly and 10 μg gp350 proteins admixed with 1.2 nmol AMP-CpG and T cell responses were analyzed at long-term timepoints. The control group was immunized with AMP-CpG alone. Splenocytes were stimulated with a–c EBV CD8⁺ T cell peptides or e, f gp350 OLPs in an ICS assay. Shown are frequencies of total IFNγ, IL-2, and TNFα a CD8⁺ or e CD4⁺ T cells over time; polyfunctional cytokine⁺ b CD8⁺ or f CD4⁺ T cell frequencies at different timepoints; and post-boost week 31 c CD8⁺ or f CD4⁺ T cell response to a recall vaccination at week 30, compared to week 29 pre-boost responses. Data are representative of one experiment. Values depicted are mean ± standard deviation. Black arrows indicate immunization days. *p < 0.05; **p < 0.01; and ***p < 0.001 by two-sided Mann–Whitney test applied to cytokine⁺ T cell frequencies. If statistics are not indicated, then the comparison was not significant. The exact p-values are as follows. a AMP vaccine to AMP-CpG at week 4, 7, 21, and 29-0.0095. b AMP vaccine to AMP-CpG at weeks 4, 7, 21, and 29-0.0095. c Post recall to pre-recall-0.0016. d AMP vaccine to AMP-CpG at weeks 4, 7, 21, and 29-0.0095. e AMP vaccine to AMP-CpG at weeks 4, 7, 21, and 29-0.0095. f Post recall to pre-recall-0.0004. Source data are provided as a Source Data file.

Fig. 7. Vaccination with AMP-CpG induces durable gp350-specific IgG responses.

HLA-B*35:01 transgenic mice (n = 9 per vaccine group and n = 5 per control group) were immunized at weeks 0, 3, and 6 with 40 μg EBVpoly and 10 μg gp350 proteins admixed with 1.2 nmol AMP-CpG. The control group was immunized with AMP-CpG alone. Humoral responses to gp350 were assessed in splenocytes and serum from immunized mice at week 7 by ASC ELISPOT or ELISA assay. Shown are a ex vivo ASC ELISPOT measured frequency of gp350-specific ASCs per 3 × 10⁵ splenocytes, b week 31 ex vivo ASC recall response to a recall vaccination at week 30, c longitudinal serum IgG titers, d week 31 IgG titers to a recall vaccination at week 30, e longitudinal Ig subtype titers in pooled serum samples, and f week 31 recall EBV neutralization titers to a recall vaccination at week 30, compared to 29 pre-boost responses using pooled serum samples. Data are representative of one experiment. Values depicted are mean ± standard deviation. Black arrows indicate immunization days. Dotted lines in c, e indicate the assay LOD. *p < 0.05; **p < 0.01; ***p < 0.001 by two-sided Mann–Whitney test. The exact p-values are as follows. a AMP vaccine to AMP-CpG at week 4, 7, 21, and 29-0.0095. b Post recall to pre-recall-0.0004. c AMP vaccine to AMP-CpG at week 3, 4, 6, 7, 14, 21, 29-0.0095. d Post recall to pre-recall-0.0004. Source data are provided as a Source Data file.

Fig. 8. Adoptive immunotherapy with EBVpoly-stimulated T cells with or without serum antibodies from vaccinated mice confers protection against EBV-associated B cell lymphoma in vivo.

a Study schema. Three groups of 6–8 week old NRG mice (n = 6/group) were treated with PBS (100 µL i.v.), EBVpoly-stimulated T cells (2 × 10⁷ cells/mouse i.v.) in combination with serum from vaccinated mice containing gp350-specific antibodies (100 µL/mouse i.p.) or EBVpoly-stimulated T cells alone (2 × 10⁷/mouse i.v.) on day −1. On day 0, these mice were engrafted with EBV-transformed lymphoblastoid cells (LCLs) (4 × 10⁶ cells/mouse s.c.). Seven days after LCL engraftment, mice were treated again as outlined for day −1 and monitored for the outgrowth of EBV lymphomas. b shows tumor volume measured using Vernier callipers. Each data point shows mean ± SEM of tumor volume in mice treated with PBS (Group 1), EBVpoly-stimulated T cells in combination with serum antibodies (Group 2) or EBVpoly-stimulated T cells alone (Group 3). Dotted line indicates maximum ethical limit of tumor volume (500 mm³). c shows frequencies of EBV-specific CD8⁺ T cells in splenocytes and in blood of animals from groups 2 and 3 on day of killing. d, e frequencies of human B cells in spleen and blood at the completion of follow up and animal killing. Data are representative of one experiment with n = 6 per group. Values depicted are mean ± s.e.m. not significant (ns), *p < 0.05, ***p < 0.001, and ****p < 0.0001 by two-sided Mann–Whitney test. The exact p-values are as follows. d PBS to T cells + Serum-0.0022 and PBS to T cells-0.0022. e PBS to T cells + Serum-0.0022 and PBS to T cells-0.0095. Source data are provided as a Source Data file. Created with BioRender.com.