Virus-like vaccines against HIV/SIV synergize with a subdominant antigen T cell vaccine

Abstract

Background: In non-human primates (NHPs) and humans, partial protection from HIV/SIV infection or suppression of replication is achievable by Env-binding antibodies and Gag-specific CD8+ T-cells targeting protective epitopes. Unfortunately, such T-cell responses are frequently dominated by responses to non-protective, variable epitopes. In this study we attempt to combine three independent approaches, each developed to prevent immunodominance of non-protective epitopes. These approaches were (1) vaccines consisting exclusively of putatively protective p24 Gag highly conserved elements (CEs), (2) vaccines using solely subdominant antigens which were acutely protective in a recent NHP trial, and (3) virus-encoded virus-like particle vaccines (virus-like vaccines/VLVs) using heterologous Env and Gag sequences to enable selection of broadly cross-reactive responses and to avoid immunodominance of non-conserved sequences in prime-boost regimens as previously observed.

Methods: We vaccinated outbred CD1 mice with HIV-1 clade B Gag/Env encoded in an adenoviral prime and SIVmac239 Gag/Env in an MVA boost. We combined this completely heterologous immunization regimen and the homologous SIVmac239 Gag/Env immunization regimen with an additional prime encoding SIV CEs and accessory antigens Rev, Vif and Vpr (Ad-Ii-SIVCErvv). T-cell responses were analyzed by intracellular cytokine staining of splenocytes and antibody responses by trimer-specific ELISA, avidity and isotype-specific ELISA.

Results: Env dominance could be avoided successfully in the completely heterologous prime-boost regimen, but Env immunodominance reappeared when Ad-Ii-SIVCErvv was added to the prime. This regimen did however still induce more cross-reactive Gag-specific CD8+ T-cells and Env-specific antibodies. Including Ad-Ii-SIVCErvv in the homologous prime-boost not only elicited accessory antigen-specific CD8+ memory T-cells, but also significantly increased the ratio of Gag- to Env-specific CD8+ T-cells. The CD4+ T-cell response shifted away from structural antigens previously associated with infection-enhancement.

Conclusion: The homologous Gag/Env prime-boost with Ad-Ii-SIVCErvv prime combined acutely protective CD8+ T-cell responses to subdominant antigens and Env-binding antibodies with chronically protective Gag-specific CD8+ T-cells in outbred mice. This vaccine regimen should be tested in an NHP efficacy trial.

Keywords: Adenoviral vectors; Antibodies; Heterologous viral vectored prime-boost immunization; Human immunodeficiency virus; Subdominant antigen vaccine; T-cells; Virus-like particles; Virus-like vaccines.

Fig. 1

Design of adenoviral vectors and characterization in vitro and in vivo. a Schematic representation of the adenoviral vectors used in the study: Ad-HIVB encodes HIV-1 consensus clade B (HIVconB) Gag, P2A preceded by a glycine-serine-glycine linker (GSG; not noted in the figure) and HIV-1 JR-FL (clade B) Env; Ad-SIV encodes SIVmac239 Gag and Env separated by GSG and P2A; Ad-Ii-SIVCErvv encodes mouse Ii aa 1–75 (Ii) fused to SIV CEs and SIVmac239 Rev, Vif and Vpr. b Vero cells were infected with 50 IFU/cell Ad-HIVB and were stained after 2 days with the monoclonal antibodies VRC01, PGT145 and PGT151 targeting HIV-1 Env (shown in dark grey). APC-labelled anti-human IgG Fc served as a secondary antibody and the cells were analyzed by flow cytometry. Uninfected stained cells served as a control (light grey). c Vero cells were infected with 50 IFU/cell Ad-HIVB and after 48 h VLPs were purified from the cell culture supernatant. The VLP samples were analyzed by SDS-PAGE followed by Western Blot, which was stained with an anti-P2A antibody. Cells infected with an Ad5 vector not encoding P2A (Ad-noP2A) served as a negative control. d CD1 mice were vaccinated with Ad-HIVB (n = 5) and the induced T-cell responses were analyzed 18 days later by intracellular cytokine staining of stimulated splenocytes. The peptide pools used for stimulation are noted on the X-axis. The total numbers of IFNγ+ CD8+ and CD4+ T-cells per spleen were measured by flow cytometry. Horizontal lines indicate the geometric mean and significant differences are marked by asterisks with *(p ≤ 0.05). e CD1 mice were immunized with Ad-Ii-SIVCErvv (n = 10). After 14 days CD8+ and CD4+ T-cell responses to the vaccine antigens were analyzed as in d

Fig. 2

CD8+ and CD4+ T-cell responses elicited by heterologous HIV-prime SIV-boost regimen. We vaccinated CD1 mice with a, c only the priming vectors Ad-HIVB or Ad-HIVB together Ad-Ii-SIVCErvv or b, d boosted them with MVA-SIV approx. 9–10 weeks after the prime (n = 10 for each group). 10 days after the boost we measured the numbers of IFNγ+ CD8+ (a, b) and CD4+ (c, d) T-cells per spleen by stimulating splenocytes with the noted peptide pools and analyzing them by intracellular cytokine staining followed by flow cytometry. Horizontal lines mark the geometric mean and significant differences are indicated by asterisks with *(p ≤ 0.05), **(p ≤ 0.01) and ***(p ≤ 0.001)

Fig. 3

Heterologous HIV-prime SIV-boost regimen induced broadly reactive Env-specific antibody responses. Serum samples were taken 10 days after the boost and analyzed for antibody binding to lyzed pseudoviruses carrying a HIV-1 clade B (HIV B), b SIVmac239 and c HIV-1 clade C (HIV C) Env, by ELISA. Horizontal lines indicate the geometric mean and significant differences are marked by asterisks with *(p ≤ 0.05), **(p ≤ 0.01) and ***(p ≤ 0.001)

Fig. 4

CD8+ and CD4+ T-cell responses induced by homologous SIV-prime SIV-boost regimen. CD1 mice were immunized with the priming vectors Ad-SIV (n = 9) or Ad-SIV and Ad-Ii-SIVCErvv (n = 10) and boosted with MVA-SIV 8 weeks after the prime. 10 days later CD8+ and CD4+ T-cell responses were measured by stimulating splenocytes with the noted SIVmac239 peptide pools and analyzing them by intracellular cytokine staining followed by flow cytometry. a, c show the total numbers of IFNγ+ CD8+ (a) and CD4+ (b) T-cells per spleen and b, d the percentage of IFNγ+ TNFα+ CD8+/CD4+ T-cells of IFNγ+ CD8+/CD4+ T-cells for samples with more than 10⁴ IFNγ+ CD8+ T-cells. Horizontal lines mark the geometric mean and significant differences are indicated by asterisks with *(p ≤ 0.05), **(p ≤ 0.01) and ***(p ≤ 0.001)

Fig. 5

Addition of Ad-Ii-SIVCErvv to homologous SIV-prime SIV-boost regimen did not impact Env-specific antibody responses. Serum samples were taken at the time of the intracellular cytokine staining and analyzed for antibody binding to lyzed pseudoviruses carrying SIVmac239 Env by ELISA. a shows the antibody responses as titers and b as the AUC. In c the avidity index is depicted and in d the antibody responses detected with anti-IgG isotype-specific antibodies as the ratio to the not isotype-specific antibody (pan-IgG). Horizontal lines indicate the geometric mean, except in c where it depicts the mean because of a single mouse not having a measurable avidity index