Targeting neoantigens conserved across organs and species overcomes tumor immune escape

Abstract

Neoantigen-targeted immunotherapies hold promise for cancer treatment, but current personalized approaches are time-consuming and costly. Here, we identify neoantigens encoded by Ptprs and Igf2r that are shared across murine mismatch repair-deficient colorectal and breast tumors and unexpectedly conserved in human colorectal, endometrial, gastric, and prostate cancers. These neoantigens elicit spontaneous, organ-spanning CD8+ T cell-mediated memory responses that are enhanced by immune checkpoint blockade. Vaccination with mRNA/lipid nanoparticles encoding these conserved neoantigens suppresses tumor growth across prophylactic and therapeutic models, including checkpoint-resistant orthotopic tumors. Tumor rejection is accompanied by antigen spreading, abscopal effects, and infiltration by clonally diverse T cells, dendritic cells, and MHC I/II+ macrophages producing CXCL9/10, CCL5/8, and TNF. Tumor cells also show activation of innate and adaptive pathways, including MHC and ISGs overexpression. Our results uncover a conserved anti-tumor immune mechanism and support the development of off-the-shelf neoantigen vaccines across tissues and species.

Keywords: Abscopal effect; Conserved immune responses; Immune escape; Shared neoantigens; Vaccination.

Extended Figure 1). Protein structure properties and spatial gene expression of the candidate neoantigens and colocalization with immune genes.

A Prediction of the protein structures according to the presence of FS del mutation of interest using Alphafold. B SpatialRNAseq, expression of the genes encoding candidate MSH2 KO neoantigens in CT26 MSH2 KO and WT tumors and colocalization with Cd8 expression. C SpatialRNAseq, expression of the genes encoding candidate CT26 neoantigens in a CT26 WT tumor. D SpatialRNAseq, expression of the immune genes in a CT26 MSH2 KO tumor and a CT26 WT tumor.

Extended Figure 1). Protein structure properties and spatial gene expression of the candidate neoantigens and colocalization with immune genes.

A Prediction of the protein structures according to the presence of FS del mutation of interest using Alphafold. B SpatialRNAseq, expression of the genes encoding candidate MSH2 KO neoantigens in CT26 MSH2 KO and WT tumors and colocalization with Cd8 expression. C SpatialRNAseq, expression of the genes encoding candidate CT26 neoantigens in a CT26 WT tumor. D SpatialRNAseq, expression of the immune genes in a CT26 MSH2 KO tumor and a CT26 WT tumor.

Extended Figure 2). Characterization of mRNA vaccine and shared peptides immunogenicity.

A Gel electrophoresis of linearized plasmid compared to non-digested plasmid (ND). B Bioanalyzer electrophoresis mRNA. C B16F10 antigen presentation assay following stimulation with MSH2 KO RNA-Lipoplexes formulations. Flow cytometry gating strategy to quantify antigen presenting (H2-Kb/SIINFEKL+) B16F10 cells. D Tumor volume or weight of CT26 MSH2 KO or 4T1 MSH2 KO orthotopic tumors according to the presence of anti-PD-1 (100 ug twice a week 7 days after tumor inoculation) and mRNA vaccine encoding shared MMRd neoantigens (5 μg mRNA-LNPs at 5 days and 10 days after tumor inoculation). N=5 to 6, Mann-Whitney U Test, pvalue<0.05. Data are represented as mean ± SEM. E 4T1 MSH2 KO tumor growth following vaccination with mRNA/LNPs vaccine encoding neoantigens not present in 4T1 tumors. F Infiltration of 4T1 and CT26 MSH2 KO tumors by immune cells on mice with prophylactic vaccination (flow cytometry). N=3 to 8, Mann-Whitney U Test, pvalue<0.05. Data are represented as mean ± SEM. G Spheroid killing and Elispot assays were performed with blood cells from vaccinated mice and stimulation with WT or mutated peptides shared by MSH2 KO tumors and predicted to be immunogenic identified in Table 1. N=4, Mann-Whitney U Test, pvalue<0.05. 10,000 CT26 MSH2 KO cells (GFP+) per spheroid, cocultured with 100,000 blood cells from vaccinated mice and stimulated with IL-15 and peptides. Death is measured with propidium iodide. N=4, Mann-Whitney U Test, pvalue<0.05. H Elispot and CT26 WT spheroid killing assays were performed with blood cells from vaccinated mice and stimulation with WT or mutated peptides shared by CT26 tumors regardless MSH2 status and predicted to be immunogenic identified in Table 2. N=4, Mann-Whitney U Test, pvalue<0.05. I ScRNAseq expression of genes encoding candidate MSH2 KO neoantigens identified in table 1 by WES and MHCflurry. J Neoantigens from Ptprs P622LFs and Igf2r D1317TFs mutations shared in our human Lynch and MSI-H tumor cohort, with predicted affinity, best allele binding, processing score, presentation score and percentile by MHCflurry. Igf2r D1317TFs mutation is expressed in the CRC tumor of one of our Lynch patients (HLA-type: A*23:01 A*36:01 B*57:01 B*53:01 C*07:01 C*04:01). K Elispot was performed with blood from this Lynch CRC patient and stimulation with WT or mutated peptides shared by MSH2 KO tumors and predicted to be immunogenic. N=4, Mann-Whitney U Test, pvalue<0.05. L TCRs shared by several primary tumors on the same patient. M VAF of mutated Ptprs and Igf2r before and after vaccination. N CT26 MSH2 KO cells were transduced to express GFP. Tumor volume was measured after 3 weeks according to therapeutic vaccination targeting MMRd neoantigens. Flow cytometry was performed for GFP-tetramer-H2-Kd staining in the spleen of these mice. N=4, Mann-Whitney U Test, pvalue<0.05. Data are represented as mean ± SEM. O C11 tumor engraftment rate at baseline, 1 week or 6 weeks after a first CT26 MSH2 KO rejection. N=5.

Extended Figure 3). Tumor volume and immune infiltration of CT26 and 4T1 MMRd and MMRp tumors.

A/B/CTumor volume and engraftment in vivo over 28 days (200,000 cells/tumor). N= 6 to 27, Mann-Whitney U Test, pvalue<0.05. A Tumor volume of CT26 MSH2 KO tumors according to the presence of a CT26 MSH2 KO tumor on the other flank. B Tumor volume of CT26 MSH2 KO tumors according to the presence of a 4T1 MSH2 KO tumor on the other flank. Tumor volume of 4T1 MSH2 KO tumors according to the presence of a CT26 MSH2 KO tumor on the other flank. C Tumor volume of CT26 WT tumors according to the presence of a CT26 MSH2 KO tumor on the other flank. D Infiltration of CT26 and 4T1 MSH2KO tumors by immune cells on mice with a CT26 MSH2 KO tumor an the left flank and a 4T1 MSH2 KO tumor on the right flank (flow cytometry) N= 7 to 8, Linear regression, pvalue<0.05. E scRNAseq and TCRseq clustering on immune cells infiltrating tumors expressing TCRs in WT or MSH2 KO tumors. The dataset contains only α/β T-cell receptors. Filtering to exclude all cells that don’t have at least one full pair of receptor sequences and multichain cells. Each dot represents a clonotype, e.g. cells with identical receptor configurations. The size of each dot refers to the number of cells with the same receptor configurations. Overexpressed genes in the more abundant clonotype (clonotype 19). F Spleen composition from mice that rejected or not a CT26 MSH2 KO tumor. G Infiltration of CT26 WT tumors by immune cells from mice that rejected or not a MSH2 KO CT26 tumor. H Infiltration of MSH2 KO spheroids by immune cells from mice that rejected or not a MSH2 KO tumor. I Infiltration of MSH2 KO CT26 spheroids by immune cells from mice that rejected or not a MSH2 KO tumor. J/K/L Tumor volume in vivo over 28 days (200,000 cells/tumor). Unilateral or bilateral injections. N= 4 to 27, Mann-Whitney U Test, pvalue<0.05. Data are represented as mean ± SEM. J Tumor volume of CT26 MMRd tumors according to the presence of a CT26 MMRd tumor on the other flank. K Tumor volume of CT26 WT tumors according to the presence of a CT26 MMRd tumor on the other flank and ICB (anti-PD-1 + anti-CTLA4 + anti-LAG3 + anti-TREM2 twice a week after day 7, 100 ug each). Data are represented as mean ± SEM. L Tumor volume of 4T1 MMRd tumors according to the presence of a CT26 MMRd tumor on the other flank and ICB (anti-PD-1 + anti-CTLA4 + anti-LAG3 + anti-TREM2 twice a week after day 7, 100 ug each) (8). Data are represented as mean ± SEM.

Extended Figure 3). Tumor volume and immune infiltration of CT26 and 4T1 MMRd and MMRp tumors.

A/B/CTumor volume and engraftment in vivo over 28 days (200,000 cells/tumor). N= 6 to 27, Mann-Whitney U Test, pvalue<0.05. A Tumor volume of CT26 MSH2 KO tumors according to the presence of a CT26 MSH2 KO tumor on the other flank. B Tumor volume of CT26 MSH2 KO tumors according to the presence of a 4T1 MSH2 KO tumor on the other flank. Tumor volume of 4T1 MSH2 KO tumors according to the presence of a CT26 MSH2 KO tumor on the other flank. C Tumor volume of CT26 WT tumors according to the presence of a CT26 MSH2 KO tumor on the other flank. D Infiltration of CT26 and 4T1 MSH2KO tumors by immune cells on mice with a CT26 MSH2 KO tumor an the left flank and a 4T1 MSH2 KO tumor on the right flank (flow cytometry) N= 7 to 8, Linear regression, pvalue<0.05. E scRNAseq and TCRseq clustering on immune cells infiltrating tumors expressing TCRs in WT or MSH2 KO tumors. The dataset contains only α/β T-cell receptors. Filtering to exclude all cells that don’t have at least one full pair of receptor sequences and multichain cells. Each dot represents a clonotype, e.g. cells with identical receptor configurations. The size of each dot refers to the number of cells with the same receptor configurations. Overexpressed genes in the more abundant clonotype (clonotype 19). F Spleen composition from mice that rejected or not a CT26 MSH2 KO tumor. G Infiltration of CT26 WT tumors by immune cells from mice that rejected or not a MSH2 KO CT26 tumor. H Infiltration of MSH2 KO spheroids by immune cells from mice that rejected or not a MSH2 KO tumor. I Infiltration of MSH2 KO CT26 spheroids by immune cells from mice that rejected or not a MSH2 KO tumor. J/K/L Tumor volume in vivo over 28 days (200,000 cells/tumor). Unilateral or bilateral injections. N= 4 to 27, Mann-Whitney U Test, pvalue<0.05. Data are represented as mean ± SEM. J Tumor volume of CT26 MMRd tumors according to the presence of a CT26 MMRd tumor on the other flank. K Tumor volume of CT26 WT tumors according to the presence of a CT26 MMRd tumor on the other flank and ICB (anti-PD-1 + anti-CTLA4 + anti-LAG3 + anti-TREM2 twice a week after day 7, 100 ug each). Data are represented as mean ± SEM. L Tumor volume of 4T1 MMRd tumors according to the presence of a CT26 MMRd tumor on the other flank and ICB (anti-PD-1 + anti-CTLA4 + anti-LAG3 + anti-TREM2 twice a week after day 7, 100 ug each) (8). Data are represented as mean ± SEM.

Extended Figure 4). Single cell analysis of immune cells in CT26 MSH2 KO tumor with or without vaccination and anti-PD-1 therapy.

ScRNASeq of immune cells in CT26 MSH2 KO tumor with or without vaccination and anti-PD-1 therapy. More significantly upregulated genes in each immune cluster (Wilcoxon test). N=3 tumors per condition.

Extended Figure 5). TCRseq analysis paired with scRNAseq analysis on CT26 MSH2 KO tumor according to vaccination and anti-PD-1 therapy.

A/B scRNAseq clustering on immune cells infiltrating tumors expressing TCRs. The dataset contains only α/β T-cell receptors. Filtering to exclude all cells that don’t have at least one full pair of receptor sequences and multichain cells. C/D Clonal expansion, number of cells or percentage for each clonotype. E Clonal expansion and diversity. F/G Clonotype abundance in each cluster and in each condition. H Most abundant V-genes and exact combinations of VDJ genes. I Spectratype and length distribution of CDR3 regions. J Convergent evolution. K Clonotype modularity. L UMAP of the most abundant clones. Analysis was performed using scirpy and scanpy.

Extended Figure 6). Single cell analysis of tumor cells in CT26 MSH2 KO tumor with or without vaccination and anti-PD-1 therapy.

ScRNASeq of tumor cells in CT26 MSH2 KO tumor with or without vaccination and anti-PD-1 therapy. More significantly upregulated genes in each condition (Wilcoxon test), and gene enrichment analysis by metascape. N=3 tumors per condition. A Upregulated genes and pathways in untreated tumor cells. B Upregulated genes and pathways in tumor cells following PD-1 blockade. C Upregulated genes and pathways in tumor cells following vaccination. D Upregulated genes and pathways in tumor cells following PD-1 blockade + vaccination.

Figure 1). Mutations and candidate neoantigens are shared between MMRd murine and human tumors in multiple organs.

Using whole exome sequencing, strelka, mutec2 and varcode, we measured the number of indel and snv mutations modifying the coding regions of genes in the CT26, C11 and 4T1 WT and MMRd cell lines ex vivo before tumor inoculation and after 28 days of tumor growth in vivo. 4 tumors were pooled together for each condition before WES. A Comparison of the number of indel mutations modifying the coding regions compared to normal tissue (spleen). B Comparison of the number of snv mutations modifying the coding regions compared to normal tissue (spleen). C Number of indel and snv mutations modifying the coding regions shared between each cell line and the CT26 MMRd cell line. D ScRNAseq of tumor cells was generated, following the exclusion of immune cells, in both CT26 WT and MMRd tumors. n=3 mice per group. ScRNAseq expression of genes encoding candidate MMRd neoantigens identified in table 1 by WES and MHCflurry. Wilcoxon test, p<0.05. E ScRNAseq expression of genes encoding candidate CT26 neoantigens identified in Table 2 by WES and MHCflurry. F SpatialRNAseq, expression of one of these genes encoding candidate MMRd neoantigens (Ptprs) in a CT26 MMRd tumor. Colocalization with Cd8+Prf1+ T cells was measured using scanpy and squidpy. N=3. G Sequences of the frameshift deletion mutations in Ptprs and Igf2r genes shared in both murine and human MSI-H tumors, with predicted affinity, best allele binding, processing score, presentation score and percentile by MHCflurry. H Percentage of MSI-H and MSS patients sharing Ptprs P622LFs mutation based on reanalyzed cbioportal data (29) with n = 623 MSI-H patients, including n = 318 colorectal cancer (COAD), n = 43 prostate cancer (PRAD) and n = 181 endometrial cancer (UEC). I Percentage of MSI-H and MSS patients sharing Igf2r D1317TFs mutation based on analysis of cbioportal TCGA data with n = 78 colorectal cancer (COAD), n = 69 stomach cancer (STAD) and n = 132 endometrial cancer (UEC). J Graphical summary of the identification of mutations and candidate neoantigens that are shared between MMRd murine and human tumors in multiple organs.

Figure 2). Anti-tumor immune memory is conserved across organs and targets shared neoantigens through CD8+ T cells.

A Tumor engraftment rate of CT26, C11 and 4T1 WT and MMRd tumors (200,000 cells/tumor) and correlation with the number of mutations (Pearson correlation). N = 10 to 165. B Tumor engraftment rate of CT26, C11 and 4T1 WT and MMRd tumors (200,000 cells/tumor) following a first CT26 MMRd rejection and correlation with the number of mutations shared with the rejected CT26 MMRd tumor (Pearson correlation). N = 10. C Spleen composition from mice that rejected or not a first CT26 MMRd tumor (flow cytometry). CT26 WT tumor immune infiltration from mice that rejected or not a first CT26 MMRd tumor (flow cytometry). N = 4. Data are represented as mean ± SEM. D Anti-CD4 or anti-CD8 antibodies were used to deplete these populations before tumor rechallenge (100ug per mice for 4 days after a first CT26 MMRd tumor rejection and then once a week). N = 4 to 10. Data are represented as mean ± SEM.

Figure 3). Targeting neoantigens shared by human and murine tumors across multiple organs delays tumor growth and induces an abscopal effect.

A/B Tumor volume of CT26 and 4T1 WT and MMRd tumors (200,000 cells/tumor) in vivo. N = 5 to 10, Mann-Whitney U Test, pvalue<0.05. Data are represented as mean ± SEM. A Tumor volume of CT26 MMRd, 4T1 MMRd, CT26 WT and 4T1 WT tumors according to the presence of mRNA vaccine encoding shared MMRd neoantigens (5 μg mRNA-LNPs 10 days and 5 days before tumor inoculation). B Tumor volume of CT26 MMRd, 4T1 MMRd, CT26 WT and 4T1 WT tumors according to the presence of anti-PD-1 (100 ug twice a week 7 days after tumor inoculation) and mRNA vaccine encoding shared MMRd neoantigens (5 μg mRNA-LNPs at 5 days and 10 days after tumor inoculation). PBS was used as a negative control. C Percentage of CT26 and 4T1 MMRd tumor complete elimination after ICB and therapeutic vaccination combinations (%). Vaccination efficacy in orthotopic settings for both models was also confirmed in Extended Figure 2. D Elispot assay was performed with blood cells from vaccinated mice and stimulation with WT or mutated peptides shared by MMRd tumors and predicted to be immunogenic identified in Table 1. N=4, Mann-Whitney U Test, pvalue<0.05. E Intracellular cytokine staining of CD8+ T cells following stimulation of blood T cells from one MMRd CRC patient with the Igf2r indel mutation with fs-neoantigen peptides (patient data and sequences are in Extended Figure 2). DMSO as negative control, CEFT as positive control. N=2. Data are represented as mean ± SEM. F CT26 WT and CT26 MMRd tumors were inoculated on opposite flanks and mice were treated with anti-PD-1 (100 ug twice a week 7 days after tumor inoculation) and mRNA vaccine encoding shared MMRd neoantigens (5 μg mRNA-LNPs at 5 days and 10 days after tumor inoculation). Tumor volume of CT26 WT tumors. N= 5 to 10, Mann-Whitney U Test, p value<0.05. Data are represented as mean ± SEM. G Elispot assay was performed with blood cells from vaccinated mice with bilateral CT26 WT/MMRd tumors and stimulation with control WT or mutated peptides shared by CT26 tumors regardless MSH2 status and predicted to be immunogenic identified in Table 2. N=4, Mann-Whitney U Test, pvalue<0.05.

Figure 4). Targeting conserved neoantigens promotes the recruitment of MHC+ macrophages, stem-like and proliferating CD4+ and CD8+ T cells, and innate responses.

A ScRNAseq UMAP of immune cells in CT26 MMRd tumors with or without anti-PD-1 therapy or vaccination encoding shared MMRd neoantigens (5 μg mRNA-LNPs at 5 days and 10 days after tumor inoculation). UMAP of immune clusters subsequent to Leiden clustering by scanpy. B Percentage of cells in each immune cluster. C Gene expression within each immune cluster. D ScTCRseq UMAP of immune cells in CT26 MMRd tumors with or without anti-PD-1 therapy or vaccination encoding shared MMRd neoantigens. UMAP of immune clusters subsequent to Leiden clustering by scanpy and clonal expansion of T cell clones. Identification of T cell clones, number of cells in each clonotype for each tumor type and gene expression within each main T cell clone (clone of interest : blue vs other clones : orange). E Alphafold prediction of the binding of the neoantigens shared across organs and species (KPSAPLKTL for Ptprs, APQPSISTV for Igf2r) with murine MHC I (H2Dd) and the TCRs from the most expended T cell clones in the tumors. F ScRNAseq of tumor and myeloid cells in CT26 MMRd tumors with or without anti-PD-1 therapy or vaccination encoding shared MMRd neoantigens (5 μg mRNA-LNPs at 5 days and 10 days after tumor inoculation). Expression of innate immune genes in tumor and myeloid cells.