Adenoviral-vectored neoantigen vaccine augments hyperexpanded CD8+ T cell control of tumor challenge in mice

Abstract

Background: Neoantigens are promising immunogens for cancer vaccines and are often delivered as adjuvanted peptide vaccines. Adenoviral (Ad) vectors have been shown to induce strong CD8⁺ T cell responses as vaccines against SARS-CoV-2, Ebola, and Zika, but their utility as neoantigen delivery vectors remains largely unexplored. In this study, we examine how an Ad-vectored neoantigen vaccine would impact tumor immunity compared with a peptide neoantigen vaccine.

Methods: We generated Ad serotype 26 (Ad26) vaccine candidates encoding B16-F10-ovalbumin (OVA) and MC38-specific neoantigens. Ad26 vaccines were compared with adjuvanted peptide delivery as prophylactic vaccines in B16-F10-OVA and MC38 challenge models. Immune responses induced by the best Ad26 vaccine (Ad26.VP22.7Epi) were compared with peptide vaccination systemically and within the tumor. Following vaccination with Ad26.VP22.7Epi, peptide, or sham, tumor-infiltrating CD45⁺ cells were analyzed using single-cell RNA sequencing (scRNA-seq) and T cell receptor sequencing (TCR-seq) to identify vaccine-induced differences in the tumor microenvironment.

Results: Single-shot Ad26 vaccines induced greater neoantigen-specific interferon-γ CD8⁺ T cell immune responses than two-shot adjuvanted peptide vaccines in mice, and Ad26.VP22.7Epi also provided superior protective efficacy compared with the peptide vaccine following tumor challenge. Ad26.VP22.7Epi induced a robust immunodominant CD8⁺ T cell response against the Adpgk neoantigen, while the peptide vaccine-induced lower responses against both Adpgk and Reps1 neoantigens. scRNA-seq analysis of CD45⁺ tumor-infiltrating cells demonstrated that both Ad26.VP22.7Epi and peptide vaccine-induced similar numbers of infiltrating CD8⁺ T cells. However, Ad26.VP22.7Epi induced CD8⁺ T cells showed more upregulation of T cell maturation, activation, and Th1 pathways compared with peptide vaccine induced CD8⁺ T cells, suggesting improved functional T cell quality. TCR-seq of these tumor-infiltrating lymphocytes also demonstrated that Ad26.VP22.7Epi generated larger T cell hyperexpanded clones compared with the peptide vaccine.

Conclusions: These results suggest that the Ad26.VP22.7Epi vaccine led to improved tumor control compared with the peptide vaccine due to increased T cell hyperexpansion and functional activation. Our data suggest that future cancer vaccine development strategies should focus on inducing functional hyperexpanded CD8⁺ T cell responses and not only maximizing tumor infiltrating CD8⁺ T cell numbers.

Keywords: T cell; Vaccine.

Figure 1. B16-F10-OVA neoantigen vaccine trial: (A) Timeline of vaccination and challenge for prophylactic and therapeutic challenge. (B) C57BL/6 mice were immunized with 10⁹ viral particles of Ad26.OVA or 50 μg of OVA protein with Poly I:C and anti-CD40 ab. Mice were bled 27 days later and whole blood intracellular cytokine staining was used to determine immunogenicity. OVA targeting cellular immune response was measured by IFNγ secreting CD8⁺ T cells isolated from whole blood. Individual mice are denoted by single dots. (C) C57BL/6 mice were challenged with 10⁵ B16-F10-OVA cells implanted in the right flank. Mean tumor growth of B16-F10-OVA tumors. Bars display mean±the SEM. (D) Survival curve of B16-F10-OVA challenged mice. N=9 for sham. N=10 for OVA protein and Ad26.OVA respectively. (E) C57BL/6 mice were challenged with 10⁵ B16-F10-OVA cells. Four days later the mice were immunized with 10⁹ viral particles of Ad26.OVA, 50 μg of OVA protein with Poly I:C and anti-CD40 ab, or sham. (F) Survival curve of B16-F10-OVA challenged mice vaccinated therapeutically. Ad26, adenovirus serotype 26; ICS, intracellular cytokine staining; IFN, interferon; OVA, ovalbumin; Poly I:C, polyinosinic acid:polycytidylic acid; WB, whole blood.

Figure 2. Generation and Immunogenicity of MC38 Ad26 cancer vaccines: (A) Immunogen design for four different Ad26 MC38 vaccine candidates. (B) In vitro mRNA expression of each vaccine candidate in A549 cells. N=3 per group. (C) C57BL/6 mice were immunized intramuscularly with 10⁹ vp of each Ad26 vaccine. Four weeks later splenocytes were harvested to measure the immune response against the 7Epi immunogen as a pool of seven peptides. Immune measured by IFN-g secreting CD8⁺ T cells isolated from the spleen. Representative flow plots shown. N=5 per group. Individual mice are denoted by single dots. Black bars represent median. Black dotted line shows limit of detection for the assay. Ad26, adenovirus serotype 26; IFN, interferon; mRNA, messenger RNA.

Figure 3. MC38 vaccine protection: (A) Vaccination schedule for Ad26 and peptide vaccines. (B,C) C57BL/6 mice were immunized intramuscularly with 10⁹ viral particles of Ad26 or 50 μg of each peptide with Poly I:C and anti-CD40 ab. Mice were bled 1 day prior to challenge and whole blood intracellular cytokine staining was used to determine immunogenicity. Whole blood intracellular cytokine staining (WB-ICS) showing prechallenge immunogenicity as measured by IFNγ⁺ and CD107a⁺ secreting CD8⁺ T cells after stimulation with 7Epi immunogen pool. Mann-Whitney test was performed *=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001. (D) C57BL/6 mice were challenged with 3×10⁵ MC38 tumor cells implanted in the right flank. After day 12 postchallenge, tumors were measured 2–3 times per week. Dots represent measurement time points. Bars display the mean±the SEM. Two-way ANOVA test used. Significance values represent significance compared with the sham group. Sham group in figures (D,E) consists of naïve, adjuvant, and Ad26.Luciferase immunized mice. Black * represent significance comparing Ad26.VP22.7Epi to sham and white * represent significance comparing peptide to sham. (E) Survival curve of MC38 challenged mice. Mantel-Cox test conducted to determine significance ****=p<0.0001. (F) Spearman correlation of mouse cytokines measured in (B,C) to tumor growth. Blue represents negative correlation and red represents positive correlation. FDR correction used. *=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001. Tumor growth studies were the result of multiple independent experiments with the following total N: Sham=89, peptide+adjuvant=19, Ad26.7Epi=28, Ad26.7Epi.VP22=26, Ad26.VP22.7Epi=30, Ad26.Shark=28. WB-ICS was performed on a subset of mice. Ad26, adenovirus serotype 26; ANOVA, analysis of variance; FDR, false discovery rate; IFN, interferon; Poly I:C, polyinosinic acid:polycytidylic acid; Treg, regulatory T cell.

Figure 4. Immune recall response and deconvolution: (A) Study timeline. (B) C57BL/6 mice were challenged with 3×10⁵ MC38 cells implanted in the right flank. Tumors were measured 2–3 times a week beginning at day 11. Dots represent measurement time points. Bars display the mean±the SEM. Significance measured by two-way ANOVA test. Significance values represent significance compared with peptide+adjuvant group. (C) At day 0, one cohort of mice was sacrificed and splenocytes were extracted. Cells were stimulated with each peptide of the 7Epi immunogen individually. IFNγ secreting CD8⁺ T cells were used to measure immune response. (D,E) At day 14 following tumor challenge, mice were sacrificed. Splenocytes and tumor infiltrating lymphocytes were harvested and measured for immune response against 7Epi individual peptides. (C–E) One-way ANOVA test used for each peptide. (F) Spearman correlations of immune responses and cell populations with tumor growth. Blue indicates negative correlation and red indicates a positive correlation. *=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001. Ad26, adenovirus serotype 26; ANOVA, analysis of variance; IFN, interferon.

Figure 5. scRNA-seq analysis of tumor-infiltrating lymphocytes: (A) Timeline of study. (B) CD45⁺ positive TILs from all mice were integrated and UMAP was generated. Cell populations denoted with * were deemed as “rare cell populations” and combined in pie-chart analysis. (C) Gene heatmap of each cell sequenced. Each vertical slice represents a single cell. Y-axis contains different genes that are used as markers for each cell label. X-axis contains what each cell type is labeled by ImmGen labeling as: Yellow represents good agreement between expected gene expression for cell type and observed gene expression of the cell, purple means low agreement. Gene heatmap shows excellent agreement of cell labels. (D) Percentage of CD8⁺ T cells found in the tumor of each mouse. *=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001. Black lines represent medians. (E–G) UMAP overlaying canonical markers for (E) T cells, (F) CD8⁺ T cells, and (G) monocytes. Ad26, adenovirus serotype 26; DC, dendritic cell; ILC, innate lymphoid cell; NK, natural killer; scRNA, single-cell RNA; Tgd, gamma delta T cells; TILs, tumor-infiltrating lymphocytes; UMAP, uniform manifold approximation and projection.

Figure 6. Differential gene expression pathway analysis of tumor infiltrating CD8 T cells: (A) Graphical representation of differentially expressed genes in CD8⁺ T cells between Ad26.VP22.7Epi vaccinated mice and peptide-vaccinated mice. Red genes are statistically significantly upregulated in Ad26.VP22.7Epi CD8⁺ T cells and blue genes are upregulated in peptide CD8⁺ T cells. Gray genes are not significant. Ingenuity Pathway Analysis of (B) select immune pathways or (C) larger multipathway categories for Ad26.VP22.7Epi compared with sham and peptide. Orange represents upregulation in Ad26.VP22.7Epi CD8⁺ T cells compared with other group and blue represents down regulation of the pathway. (D) C57BL/6 mice were immunized prior to challenge (following figure 3A). Two days and 1 day prior to challenge mice were administered an anti-CD8 depletion antibody or an isotype control. Mice were challenged with 3×10⁵ MC38 cells implanted in the right flank. Tumors were measured five times a week beginning at day 7. Dots represent measurement time points. Bars display the mean±the SEM. Ad26, adenovirus serotype 26; IL, interleukin; PD-1, programmed cell death protein-1; PD-L1, programmed death-ligand 1.

Figure 7. TCR-seq analysis of TILs: (A) UMAP of cells with complete TCR sequences integrated across different vaccine conditions. (B) Comparison of key T cell populations between vaccine conditions. *=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001. Black lines represent medians. (C) UMAP overlay of top TCR clonotype for each vaccine condition. Overlayed onto integrated UMAP from (A). (D) Graphical representation of top 10 VDJ gene usage clonotypes per vaccine platform (dark colors) overlayed on the total number T cells both normalized per animal. (E) Relative abundance of each clonotype group within each animal. (F) Pearson correlation of tumor growth to T cell subpopulations. Blue represents negative correlation and red represents positive correlation. *P<0.05. Ad26, adenovirus serotype 26; TCR-seq, T cell receptor sequencing; Tfh, T follicular helper cell; TIL, tumor-infiltrating lymphocytes; Treg, regulatory T cell; UMAP, uniform manifold approximation and projection; VDJ, variability-diversity-joining.