Identification of 68 HLA-A24 and -A2-restricted cytotoxic T lymphocyte-inducing peptides derived from 10 common cancer-specific antigens frequently expressed in various solid cancers

Abstract

Targeting cancer antigens expressed in cancer cells is necessary to develop cancer-specific immunotherapy. We have performed immunohistochemical analysis of various solid cancer specimens, adding ROBO1, AFP, TGFBI, EphB4, CLDN1, and LAT1 to the previously studied glypican-3 (GPC3), HSP105α, FOXM1, and SPARC, and found that these 10 common cancer antigens are sufficient to cover most solid cancers. These antigens were frequently expressed in various solid cancers but shown to be rarely ex-pressed, with some exceptions, in non-cancerous normal organs adjacent to the cancer. In this study, we predicted 72 and 73 peptides that bind to HLA-A24 and -A2 in silico from the full-length amino acid sequences of these 10 common cancer antigens and immunized each HLA transgenic mouse with a cocktail of synthesized peptides together with the poly I:CLC three times weekly to analyze the antigen-specific immune response. As a result, 68 peptide sequences (30 and 38, respectively) were identified that had higher cytotoxic T lymphocyte (CTL) induction ability than GPC3 298-306 and GPC3 144-152 used in the clinical trials. Furthermore, experiments with cocktail peptide vaccines using mouse models expressing subcutaneous tumors of each antigen showed promising results in terms of safety and efficacy. These peptides identified in this study, derived from 10 common cancer antigens covering all solid cancers, are expected to be clinically applicable as cocktail peptide vaccines.

Keywords: Cancer-specific antigen; Common cancer antigen; Cytotoxic T lymphocyte (CTL); HLA-A2; HLA-A24; Peptide vaccine; mRNA vaccine.

Fig. 1

Expression frequency of 10 common cancer-specific antigens in various solid cancers. Representative examples of the expression patterns of 10 types of cancer-specific antigens common to each type of cancer are shown.

Fig. 2

Expression of 10 common cancer-specific antigens in non-cancerous normal organs adjacent to various cancers. Representative examples of the expression patterns of 10 common cancer-specific antigens in non-cancerous normal organs (pharynx, mammary gland, esophagus, lung, stomach, liver, gallbladder, pancreas, colon, kidney, ovary, uterus, and skin) adjacent to the cancer lesions of a total of 385 cases of surgically resected ad-vanced cancer are shown.

Fig. 3

Expression of HLA class I on cell membranes in various cancerous tissues and adjacent non-cancerous normal organs. The expression patterns of HLA class I in the cancerous areas and adjacent normal organs (pharynx, mammary gland, esophagus, lung, stomach, liver, gallbladder, pancreas, colon, kidney, ovary, uterus, and skin) of a total of 364 cases of surgically resected advanced cancer, head and neck cancer, breast cancer, esophageal cancer, lung cancer, gastric cancer, hepatocellular carcinoma, biliary tract cancer, pancreatic cancer, colorectal cancer, renal cancer, ovarian cancer, uterine cancer, melanoma, and pediatric cancer are shown.

Fig. 4

Screening method for epitope peptides derived from predicted common cancer antigens in HLA transgenic mice (Tgm). (A) The expression frequency of each antigen in cancerous and normal tissues of patients with various solid tumors was analyzed immunohistochemically, and DNA and RNA of antigen proteins commonly highly expressed in cancer tissues were extracted. Whole exon and RNA sequence analyses were performed using a next-generation sequencer, and peptide candidates predicted to bind to HLA were selected using an HLA-binding prediction algorithm. (B)Scheme of vaccinations and assays. For experiments using human HLA-transgenic mice (human HLA-Tg) experiments, peptides with wild-type sequences homologous to mice were selected and used for peptide vaccination with poly I:CLC as an adjuvant. After repeated vaccination with human HLA-Tg, spleen cells were collected, and immune responses to the peptides were evaluated using IFN -ELISPOT assay. (C)Representative pictures of ELISPOT assays with A02‐restricted GPC3 545-553, EphB4 47-55, EphB4 581-589, and FOXM1 377-385 peptides in vaccinated HLA‐A02 Tg mice. “No peptide”and“PMA + ionomycin”conditions were the negative and positive controls, respectively. (D) Representative images of ELISPOT assays using positive peptides for HLA A02 restricted-10 common cancer antigen peptides are shown, (E) and HLA A24 restricted-10 common cancer antigen peptides are shown. Several peptides were identified that showed higher immunoreactivity compared to the clinical trial peptides circled in red. The Figure was created with Biorender.com.

Fig. 5

The efficacy and safety of the cocktail vaccine were verified by ELISPOT assay and IHC staining. Peptides with high inducibility of CTLs were selected by screening, and a cocktail vaccine was made by combining these peptides. (A) Screening was performed using the same scheme as for the short peptide, and MCA205 tumor cells expressing the antigen were implanted on the second dose (Day 7). One week after the third vaccine, HLA-Tgmice were sacrificed, spleens were used for ELISPOT assay, and tumors and normal organs were analyzed by IHC staining. (B) The preceding 2 groups of cocktail vaccine experiments are shown. High immune response to tumor cells expressing antigens, and various peptides were confirmed by ELISPOT assay. (C) The transplanted MCA205-hEphB4 and MCA205-hCLDN1 were stained with EphB4 and CLDN1 to confirm the expression of each antigen positive cell. Each tumor was also stained with CD8 to assess lymphocyte infiltration within the tumor. (D) Normal organs of unvaccinated and vaccinated mice were stained with CD8 for safety evaluation. The Figure was created with Biorender.com.