Expression of tumor antigens within an oncolytic virus enhances the anti-tumor T cell response

Abstract

Although patients benefit from immune checkpoint inhibition (ICI) therapy in a broad variety of tumors, resistance may arise from immune suppressive tumor microenvironments (TME), which is particularly true of hepatocellular carcinoma (HCC). Since oncolytic viruses (OV) can generate a highly immune-infiltrated, inflammatory TME, OVs could potentially restore ICI responsiveness via recruitment, priming, and activation of anti-tumor T cells. Here we find that on the contrary, an oncolytic vesicular stomatitis virus, expressing interferon-ß (VSV-IFNß), antagonizes the effect of anti-PD-L1 therapy in a partially anti-PD-L1-responsive model of HCC. Cytometry by Time of Flight shows that VSV-IFNß expands dominant anti-viral effector CD8 T cells with concomitant relative disappearance of anti-tumor T cell populations, which are the target of anti-PD-L1. However, by expressing a range of HCC tumor antigens within VSV, combination OV and anti-PD-L1 therapeutic benefit could be restored. Our data provide a cautionary message for the use of highly immunogenic viruses as tumor-specific immune-therapeutics by showing that dominant anti-viral T cell responses can inhibit sub-dominant anti-tumor T cell responses. However, through encoding tumor antigens within the virus, oncolytic virotherapy can generate anti-tumor T cell populations upon which immune checkpoint blockade can effectively work.

Fig. 1. The hMet + S45Y ß-Catenin sleeping beauty transposon system induces liver tumors which are infiltrated by PD-1⁺ CD8 T cells.

Following hydrodynamic injection of hMet + S45Y ß-Catenin plasmids, animals were euthanized at 4 and 7 weeks for pathologic studies. A Livers (animals A1 and A4) demonstrated appreciable neoplasia with a well-differentiated morphology and minimal immune infiltration by week 4. This progressed by week 7 (animals C1 and C4) with neoplastic cells overtaking the majority of the tumor, with large pseudo-cysts, fat deposits, and signs of extramedullary hematopoiesis (more prominent in C4). B Livers from a control mouse and from a mouse 7 weeks after hMet + S45Y ß-Catenin hydrodynamic tail-vein injection. C, D Following hydrodynamic injection of hMet + S45Y ß-Catenin, animals were euthanized at day 22 and day 29 and analyzed for CD8+ (p = 0.0162) (C) and CD4+ (p > 0.1) (D) T cells by flow cytometry. D Similar to (C), tumor-infiltrating CD4 T cells were evaluated. E, F Proportions of CD8+ (Tim3- PD1- Naïve vs. Day 22 and 29, p < 0.0001; Tim3 + PD1+ Naïve vs. Day 22 and 29, p < 0.0001) (E) and CD4+ (p > 0.1 except for Tim3- PD1- Naïve vs. Day 22, p = 0.0260) (F) T cells expressing Tim3 and/or PD1⁺ in livers of tumor bearing or tumor naïve mice at days 22 and day 29. G Proportions of PD-L1⁺ CD86⁺ dendritic cells (DCs) in mice 22 and 29 days after hMet + S45Y ß-Catenin injection compared to tumor naïve mice. (PD-L1 + CD86+ Naïve vs. Day 22, p = 0.0054; PD-L1 + CD86- Naïve vs. Day 29, p = 0.0028; PD-L1 + CD86- Day 22 vs. Day 29, p = 0.0002). H Hematoxylin and eosin staining combined with PD-L1 immunohistochemistry of the livers of FVB mice at weeks 4, 5, 6, and 7 after hMet + S45Y ß-Catenin hydrodynamic tail-vein injection. 10x magnification. Significance for (C, D) determined by ordinary one-way ANOVA, data are presented as mean values +/− SEM. Significance for (E–G) was determined by 2-way ANOVA with multiple comparisons using Tukey’s multiple comparisons test, data are presented as mean values +/− SEM.

Fig. 2. Responsiveness of SB-HCC to CD8-mediated immune checkpoint blockade is abolished by VSV virotherapy.

Animals in all arms underwent hydrodynamic injection of hMet + S45Y ß-Catenin at day 0. N = 67 animals (A–C). A–C Control IgG or anti-PD-L1 ICI therapy was initiated at day 21 for 6 doses (days 21, 23, 25, 28, 30, 31) in non-depleted mice or in mice depleted of CD8 (A), CD4 (B) or NK cells (C) as shown (8 total doses). Survival with time is shown. D Mice hydrodynamically injected with hMet + S45Y ß-Catenin at day 0 were treated with PBS or VSV-mIFNß on days 21, 23, and 25 (3 doses, 10⁸ pfu/dose), followed by control IgG isotype or anti-PD-L1 on days 28, 30, 32, 35, 37, 39 (6 doses, 200 µg per dose). Survival with time is shown. E Mice hydrodynamically injected with hMet + S45Y ß-Catenin at day 0 were treated with control IgG isotype or anti-PD-L1 on days 21, 23, 25, 28, 30, 33 (6 doses, 200 µg per dose) and with PBS or VSV-mIFNß on days 40, 42 and 44 (3 doses, 10⁸ pfu/dose). Survival with time is shown. Significance was determined through survival curve comparison testing using a log-rank Mantel-Cox test.

Fig. 3. IT treatment with VSV-mIFNß generates a dominant anti-viral CD8 T cell response which replaces the anti-tumor T cell response.

A Treatment regimen. Following hydrodynamic injection of hMet + S45Y ß-Catenin on day 0, animals were treated with anti-PD-L1 on day 21 for 6 doses concurrently with VSV-mIFNß for 3 doses. All animals were euthanized on day 38. Livers were processed into single-cell suspensions and analyzed by the Mayo CyTOF core facility. N = 16, 4 animals per group, groups were pooled for Rphenograph tSNE analysis, 22 groups. Scatterplot and mean heat map shown. B Following initial analysis (A), 9 distinct immune populations were identified and pooled data was re-analyzed using FlowSOM tSNE analysis. ‘Exhausted CD8 T cells’ were identified by expression of CD8, Tim-3, LAG-3, CD39, and PD-1 expression, ‘B cells’ by CD38 and CD19; ‘CD4 T cells’ by CD4; ‘Memory CD4 T Cells’ by CD4, CCR7, and CD62L; ‘Memory CD8 T Cells’ by CD8, CCR7, and CD62L; ‘NK Cells’ by NK1.1; ‘NKT Cells’ by NK1.1 and CD8; and ‘Anti-viral CD8 T Cells’ by CD8, GranzB, and PD-L1. C Pooled analysis of lymphocytes within livers of tumor-free mice, first 10,000 events. D Pooled analysis of lymphocytes within livers of SB-HCC-bearing mice, first 10,000 events. E Pooled analysis of lymphocytes within livers of SB-HCC bearing mice treated with anti-PD-L1, first 10,000 events. F Pooled analysis of lymphocytes within livers of SB-HCC-bearing mice treated with anti-PD-L1 and VSV-mIFNß, first 10,000 events.

Fig. 4. Expression of a HCC_TAA within VSV overcomes the loss of anti-TAA CD8 + T cells by VSV virotherapy.

A, B Following hydrodynamic injection of HCC-OVA (day 0), mice were treated with anti-PD-L1 (day 6 for 6 doses). On day 23 livers (A) and spleens (B) were processed to single-cell suspensions and analyzed by flow cytometry for SIINFEKL tetramer positive CD8 + T cells. (n = 12, 4 per group, representative flow data shown). C–E Following hydrodynamic injection of HCC-OVA (day 0), mice were treated with anti-PD-L1 (day 6 for 6 doses) followed by VSV-GFP or VSV-IFNß (3 doses, days 18,20,22). On day 23 livers (Naïve vs. HCC-OVA + isotype p = 0.0004, HCC-OVA + isotype vs. HCC-OVA + αPDL1 p = 0.0037; HCC-OVA + αPDL1 vs. HCC-OVA + VSV-GFP, HCC-OVA + αPDL1 + VSV-GFP, HCC-OVA + VSV-mIFNβ, and HCC-OVA + αPDL1 + VSV- mIFNβ p < 0.0001) (D) and spleens (Naïve vs. HCC-OVA + αPDL1 p = 0.0111; HCC-OVA + αPDL1 vs. HCC-OVA + VSV-GFP p = 0.0345, HCC-OVA + VSV-mIFNβ p = 0.0091, and HCC-OVA + αPDL1 + VSV- mIFNβ p = 0.0226) (E) were analyzed by flow cytometry for SIINFEKL tetramer positive CD8 + T cells. (n = 28, 4 animals per group, data are presented as mean values +/− SEM). F–H Following hydrodynamic injection of HCC-OVA (day 0), mice were treated with anti-PD-L1 (day 23, 6 doses) followed by VSV expressing ovalbumin (VSV-OVA) for 3 doses (day 90, 92, 94). On day 98 livers and spleens were analyzed by flow cytometry for SIINFEKL tetramer positive CD8 + T cells. (n = 12, 4 per group, data are presented as mean values +/− SEM. Representative flow data is shown as well as the pooled ratio of SIINFEKL-tetramer⁺ to VSV-N⁺ CD8 T cells). Tumor-Free + Isotype + VSV-OVA vs. HCC + αPDL1 + VSV-OVA p = 0.0032. HCC + Isotype + VSV-OVA vs. HCC + αPDL1 + VSV-OVA p = 0.0074. G Tumor-Free + Isotype + VSV-OVA vs. HCC + αPDL1 + VSV-OVA p = 0.0059. HCC + Isotype + VSV-OVA vs. HCC + αPDL1 + VSV-OVA p = 0.0162. H Significance was determined through ordinary one-way ANOVA with Tukey’s multiple comparison tests, flow gating strategies shown in Supplementary Fig. 4.

Fig. 5. Putative HCC_TAA selected for high-level expression and MHC binding affinity.

A SB-HCC tumor-bearing mice were euthanized on day 18 with livers harvested for RNA extraction. RNA samples were then subjected to RNA-seq analysis, identifying 10 highest relative gene expression compared to non-tumor bearing livers. B, C The 10 most-expressed genes identified and full-length sequences were filtered through NET MHC 2.0 binding affinity algorithm to identify octamer or nonamer peptides whose binding affinity for H2Kb (B) or H2Kd (C) was below a threshold of 500 nM, and whose corresponding wild-type peptides had a binding affinity to the same molecules above 500 nM. These corresponded to Lcn2, Lect2, and SMAGP. D, E C57Bl/6 mice (n = 12, 3 per group, data are presented as mean values +/− SEM) were injected intravenously with 10⁷ plaque-forming units (pfu) of VSV-mIFNß, VSV-Lcn2, VSV-Lect2, or VSV-SMAGP. 10 days later 10⁶ splenocytes were co-cultured with a 1:1:1 mixture of SB-HCC explants 1, 2 and 3 at an effector:target (E:T) ratio of 10:1. Supernatants were assayed 48 h later for IL-17 (D) or IFNγ (VSV-mIFNβ vs VSV-Lcn2 p = 0.0040) (E). Significance was determined through ordinary one-way ANOVA with Tukey’s multiple comparison tests. Statistical significance was set with * indicating a p value less than 0.05, ** <0.01, *** <0.001, and **** <0.0001.

Fig. 6. VSV expressing the putative HCC_TAA Lcn2 improves upon therapy with VSV-IFNß in combination with ICI.

A Following hydrodynamic injection of hMet + S45Y ß-Catenin (day 0), mice was treated with anti-PD-L1 (200 µg/injection; days 5, 7, 9, 12, 14, 16) followed by 10⁷pfu of VSV-IFNß, VSV-IFNß-Lcn2 or with 3 × 10⁶ pfu each of (VSV-IFNß-Lect2 + VSV-IFNß-Lcn2 + VSV-IFNß-Smagp) (day 21, 22, 23). N = 32. B, C In a repeat of the protocol of (A), spleens were harvested 10 days following the last dose of virus and 10⁶ splenocytes were co-cultured with SB-HCC 1,2,3 tumor targets at an effector:target ratio of 10:1 with supernatant assayed 48 h later for (B) IL-17 and (C) IFNγ. N = 15, data are presented as mean values +/− SEM, VSV-mIFNβ + αPDL1 vs. αPDL1 p = 0.0018, αPDL1 vs. VSV-Lcn2, -Lect2, -SMAGP + αPDL1 p = 0.0006. Significance for (A) was determined through survival curve comparison testing using a log-rank Mantel-Cox test and (B, C) was determined through ordinary one-way ANOVA with Tukey’s multiple comparison tests.

Fig. 7. VSV-SB-HCC1,2,3 cures mice with SB-HCC mediated by CD8 T cells and associated with Th1 and Th17 responses.

A cDNA from three murine Sleeping Beauty HCC tumor-bearing livers, (SB HCC 1,2&3 Library), was pooled, cloned and amplified by PCR. The Library was then cloned into the VSV backbone plasmid between the Xho1 and Nhe1 sites. Figure created with BioRender.com under CC BY-NC-ND. B Composition of the VSV-derived ASMEL and SB HCC 1,2&3 Libraries was validated using PCR from the equivalent of 10⁸ pfu of the VSV-ASMEL or VSV-SB HCC 1,2&3 with gene specific primers to the HCC-specific Lcn2, Lect2, and Smagp genes or to the melanoma specific TYRP1 and GP100 genes. All p values noted are <0.0001. C Following hydrodynamic injection of hMet + S45Y ß-Catenin (day 0), animals were treated on days 21, 23, 25, 28, 30, 32 with anti-PD-L1 (100 µg/injection) with, or without, 10⁷ pfu of VSV-SB-HCC1,2,3 or 3 × 10⁶ pfu each of (VSV-IFNß-Lect2 + VSV-IFNß-Lcn2 + VSV-IFNß-Smagp) on days 38,40,42. (N = 32, 8 mice per group). D, E 10⁶ splenocytes were co-cultured with a 1:1;1 mixture of live SB-HCC 1,2,3 explant cells as targets at an effector:target ratio of 10:1 with supernatant assayed 48 h later for (D) IL-17 and (E) IFNγ. (N = 12, 3 mice per group, for (D) all p values noted are <0.0001, for (E) αPDL1 + VSV-Lcn2, -Lect2, -SMAGP vs. VSV-SB-HCC1,2,3 + αPDL1 p = 0.0230, VSV-SB-HCC1,2,3 + αPDL1 vs. VSV-SB-HCC1,2,3 + IgG p = 0.0437) (F). Following hydrodynamic injection of hMet + S45Y ß-Catenin (day 0), animals were treated on days 21, 23, 25, 28, 30, 32 with control isotype IgG or with anti-PD-L1 (200 µg/injection) with, or without, PBS or 10⁷pfu ASMEL (VSV-melanoma cDNA library) or 10⁷ pfu VSV-SB-HCC1,2,3 on days 38,40,42. Depleting antibodies were given concurrently with either anti-PD-L1 or IgG isotype control for 6 doses. (N = 65, 8 mice per group except VSV-SB-HCC1,2,3 + Control IgG, n = 9). Significance comparisons for (Fig. S8) are included in the Supplement, (B) was determined using 2-way ANOVA with multiple comparisons using Tukey’s multiple comparisons test, (C, F) were determined through survival curve comparison testing using a log-rank Mantel-Cox test, and (D, E) were determined through ordinary one-way ANOVA with Tukey’s multiple comparison tests.