Targeting Mutated Plus Germline Epitopes Confers Pre-clinical Efficacy of an Instantly Formulated Cancer Nano-Vaccine

Abstract

Personalized cancer vaccines hold promises for future cancer therapy. Targeting neoantigens is perceived as more beneficial compared to germline, non-mutated antigens. However, it is a practical challenge to identify and vaccinate patients with neoantigens. Here we asked whether two neoantigens are sufficient, and whether the addition of germline antigens would enhance the therapeutic efficacy. We developed and used a personalized cancer nano-vaccine platform based on virus-like particles loaded with toll-like receptor ligands. We generated three sets of multi-target vaccines (MTV) to immunize against the aggressive B16F10 murine melanoma: one set based on germline epitopes (GL-MTV) identified by immunopeptidomics, another set based on mutated epitopes (Mutated-MTV) predicted by whole exome sequencing and a last set combines both germline and mutated epitopes (Mix-MTV). Our results demonstrate that both germline and mutated epitopes induced protection but the best therapeutic effect was achieved with the combination of both. Our platform is based on Cu-free click chemistry used for peptide-VLP coupling, thus enabling bedside production of a personalized cancer vaccine, ready for clinical translation.

Keywords: germline; melanoma; mutated; neoantigen; personalized; vaccine; virus-like particles.

Figure 1

Algorithm for the generation of a personalized melanoma vaccine platform based on VLPs. First, tumor-specific epitopes need to be identified in a systematic way. Here, immunopeptidomics approach has been used to identify potential tumor-specific germline epitopes and whole exome sequencing to predict the mutated ones. In a second step, the identified and predicted epitopes should be prioritized and this was basically achieved by bioinformatics and stimulation of tumor-infiltrating cells using in vitro T-cell assay. Next, the selected epitopes are extended to ~13–14 amino acids long peptides using their flanking protein sequence for the goal of targeting CD8⁺ CTL. The extended peptides are then synthesized and coupled to CpG-loaded VLPs using Cu-free click chemistry. A mix-multi target VLP-based vaccine is proposed.

Figure 2

Bio-orthogonal Cu-free click chemistry; an efficient method for coupling antigens to VLPs to enhance their immunogenicity. (A) A sketch illustrating the coupling method using Bio-orthogonal Cu-free click chemistry (dibenzocyclooctyne NHS ester DBCO). Briefly, NHS ester reacts with Lys residues on VLPs and incorporates a cyclooctyne moiety which reacts with the azide labeled molecule forming a stable triazole linkage. (B) SDS-PAGE stained with coomassie blue showing the coupling efficiency of p33 to Qβ(CpGs)-VLP using SMPH cross-linker or DBCO cross-linker. Lane 1 protein marker, lane 2 Qβ-VLP monomer, lane 3 Qβ(CpGs)-VLP derivatized with SMPH cross-linker, lane 4 Qβ(CpGs)-p33 vaccine using SMPH cross-linker, lane 5 Qβ(CpGs)-VLP derivatized with DBCO cross-linker, lane 6 Qβ(CpGs)-p33 vaccine using DBCO cross-linker. Each extra band in lanes 4 and 6 indicates a peptide binds to a Qβ monomer. (C) Densitometric analysis of SDS-PAGE lanes 1, 4, and 6. (1) uncoupled Qβ-VLP, total 180 monomers, (2) Qβ(CpGs)-p33 vaccine using SMPH cross-linker, and (3) Qβ(CpGs)-p33 vaccine using DBCO cross-linker. Notice the percentage of the coupled peptides to Qβ-VLP monomers, total 180 monomers. (D) Percentage of CD8⁺ Tetramer⁺ cells (means ± SEM) in the spleen of vaccinated groups: group 1) Qβ(CpGs)-VLPs, group 2) Qβ(CpGs)-p33 vaccine using SMPH cross-linker, and group 3) Qβ(CpGs)-p33 vaccine using DBCO cross-linker. (E) Percentage of CD8⁺ IFN-γ⁺ secreting cells (means ± SEM) in the spleen of vaccinated groups. (F) CFSE in vivo lytic activity in the spleen of vaccinated groups, using the formula 100X(1–CFSE^Hi pulsed/CSFE^Low un-pulsed). Statistical analysis by Students t-test. (G) Representative FACS plots showing the percentage of CD8⁺ Tetramer⁺ cells in the spleen of vaccinated groups. (H) Representative FACS histogram of CFSE in vivo lytic activity in the spleen of vaccinated groups. (n = 3) mice per group, one representative of 3 similar experiments is shown. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

Figure 3

Identification and prediction of CD8⁺ T-cell epitopes of B16F10 melanoma cells by immunopeptidomics and whole exome sequencing. (A) Length distribution of peptides identified by immunopeptidomics. (B) 8 and 9 mers motifs identified by immunopeptidomics. (C) Whole exome sequencing results showing a heterozygous SNV (marked in red and orange) in the 13th exon of gene Kif18b at position Chr11:102,908,157 (ENSMUST00000021311: c.2367T>G, p.Lys739Asn), causing a K>N missense mutation. (D) Whole exome sequencing results showing a heterozygous SNV (marked in orange and green) in the 9th exon of gene Cpsf3l at position Chr4:155,886,970 (ENSMUST00000120794: c.1021G>A, p.Asp292Asn and ENSMUST00000030901: c.1087G>A; p.Asp314Asn), causing a D>N missense mutation.

Figure 4

Validation of germline and mutated epitopes identified by immunopeptidomics and predicted by whole exome sequencing. (A) Percentage of CD8⁺ IFN-γ⁺ secreting cells (means ± SEM) in duplicate, pre-gated on TILs. p33 peptide was used as a positive control and actin as a negative control. Statistical analysis by Student's t-test in comparison with Actin (–ve control). (B) Representative FACS plots showing the percentage of CD8⁺ IFN-γ⁺ secreting cells pre-gated on TILs and stimulated with Actin (–ve control), p33 (+ve control), ZFP518, Cpsf3l, PMEL17, and Kif18b peptides. One representative of 3 similar experiments is shown. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

Figure 5

Multi-target VLP-based vaccine using Cu-free click chemistry could significantly hinder the progression of the aggressive transplanted B16F10 tumor. (A) A sketch of the challenging therapeutic murine melanoma model based on injecting ~1 × 10⁶ B16F10 melanoma cell line into the flank of C57BL/6 RAG2^−/− mice. Twelve to thirteen days later the growing tumors are collected and processed for transplantation of ~2 mm³ into the flank of C57BL/6 WT mice. (B) Vaccination scheme of five groups: Qβ(CpGs)-VLPs, anti-CD25, germline-multi target vaccine (GL-MTV), Mutated-MTV, and Mix-MTV GL-MTV. (C) Representative percentage of CD4⁺ CD25^Hi in the periphery of the groups vaccinated with Qβ(CpGs)-VLPs, anti-CD25, and Mix-MTV+anti-CD25 on day 6 post-tumor transplantation. (D) Representative FACS plots showing the percentage of CD4⁺ CD25^Hi in periphery of the groups vaccinated with Qβ(CpGs)-VLPs, anti-CD25, and Mix-MTV+anti-CD25 on day 6 post-tumor transplantation. (E) Photographic images of s.c. B16F10 tumors on day 14 post-tumor transplantation. (F) Tumor volume mm³ (means ± SEM) measured on day 14 post-tumor transplantation for the groups vaccinated with Qβ(CpGs)-VLPs, anti-CD25, GL-MTV, Mutated-MTV and Mix-MTV, each dot represents a tumor. Statistical analysis by Student's t-test. (G) Individual tumor growth curves of s.c. B16F10 melanoma of the designated groups. (H) Combined tumor growth curves of the designated groups, arrows indicate start of treatment. Statistical analysis by AUC. The groups of GL-MTV, Mutated-MTV, and Mix-MTL were combined with anti-CD25mAb in all experiments. (n = 5) mice per group, one representative of 3 similar experiments is shown. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

Figure 6

Mix multi-target VLP-based vaccine increased CD8⁺ T-cell infiltration into B16F10 tumor and enhanced the survival. (A) Density of CD8⁺ T-cells (means ± SEM) in tumor, pre-gated on TILs in the groups vaccinated with Qβ(CpGs)-VLPs, anti-CD25, GL-MTV, Mutated-MTV, and Mix-MTV. The density measured by dividing the total number of CD8⁺ T-cells in each tumor by its volume. (B) Density of CD8⁺ IFN-γ⁺ secreting cells (means ± SEM) in tumor, pre-gated on TILs in the vaccinated groups. Statistical analysis by one-way ANOVA. (C) Representative FACS plots showing the total number of CD8⁺ T-cells in each tumor in the vaccinated groups. (D) Survival of mice in the vaccinated groups, mice were euthanized when the tumor reached 1,000 mm³. The arrows indicate vaccination time. Statistical analysis by log-rank test. The groups of GL-MTV, Mutated-MTV, and Mix-MTL were combined with anti-CD25mAb in all experiments. (n = 5) mice per group, one representative of 3 similar experiments is shown. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

Figure 7

Mix multi-target vaccine altered the myeloid composition of B16F10 tumor. (A) Percentage of Ly6G⁺ cells (means ± SEM) in TILs, pre-gated on CD45⁺ cells in the groups vaccinated with Qβ(CpGs)-VLPs, anti-CD25, GL-MTV, Mutated-MTV, and Mix-MTV. (B) Percentage of Ly6C⁺ cells (means ± SEM) in TILs, pre-gated on CD45⁺ cells in the vaccinated groups. Statistical analysis by one-way ANOVA. (C) Representative FACS plots showing the percentage of Ly6G⁺ and Ly6C⁺ cells in TILs in the vaccinated groups. (D) Correlation between the ratio of the percentage of Ly6G⁺ and Ly6C⁺ cells over the tumor volume mm³ in the vaccinated groups. Statistical analysis by linear regression. The groups of GL-MTV, Mutated-MTV, and Mix-MTL were combined with anti-CD25mAb in all experiments. (n = 5) mice per group, one representative of 3 similar experiments is shown. ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.