Screening and identification of HLA-A2-restricted neoepitopes for immunotherapy in endocrine therapy-resistant breast cancer

Abstract

Endocrine therapy has shown significant clinical efficacy in estrogen receptor alpha (ERα)-positive breast cancer management, but the emergence of therapy-resistant mutations significantly undermines treatment outcomes, frequently leading to disease progression and metastasis. Among these resistance mechanisms, mutations in the ESR1 gene are particularly prevalent, detectable in 76% of endocrine therapy-resistant tumor specimens. The identification of immunogenic neoepitopes derived from mutant ESR1 offers a promising therapeutic avenue for patients with endocrine therapy-resistant breast cancer. In this study, we systematically investigated the mutational landscape of ESR1 across various cancer types, with particular emphasis on mutation frequency and spectrum analysis. Our findings revealed that non-synonymous ESR1 mutations predominantly occurred in breast cancer, clustering at four distinct hotspot sites: K303, E380, Y537 and D538. We further characterized the mutation prevalence at these hotspots across different breast cancer subtypes. Through comprehensive screening, we identified eight human leukocyte antigen (HLA)-A*0201 restricted immunogenic neoepitopes derived from ESR1 hotspot mutations. These neoepitopes demonstrated the capacity to elicit specific cytotoxic T lymphocytes (CTLs) responses both in vitro and in vivo. The induced CTLs exhibited specific recognition and cytotoxic activity against both T2A2 cells loaded with mutant neoepitopes and HLA-A*0201-positive breast cancer cells transfected with minigene encoding mutant neoepitopes. Notably, adoptive transfer of T cells primed with a peptide pool containing these eight neoepitopes significantly suppressed tumor growth and enhanced CD8⁺ T cells infiltration within tumor tissue. These findings suggest that the identified neoepitopes represent promising candidates for the development of tumor shared neoantigen vaccines.

Keywords: ESR1; HLA-A2; Hotspot mutations; Immunotherapy; Neoepitopes.

Fig. 1

Mutation of ESR1 in different tumor types. A. ESR1 mutation in different tumor tissues. B. ESR1 mutation rate in penis. C. ESR1 mutation rate in liver. D. ESR1 mutation rate in pancreas. E. ESR1 mutation rate in prostate. F. ESR1 mutation rate in oesophagus. G. ESR1 mutation rate in endometrium. H. ESR1 mutation rate in skin. I. ESR1 mutation rate in breast. J. ESR1 mutation type rate in breast. K. ESR1 hot spot mutation in breast ductal epithelial hyperplasia. L. ESR1 hot spot mutation in breast ductal carcinoma. M. ESR1 hot spot mutation in ER-PR positive breast carcinoma. N. ESR1 hot spot mutation in breast lobular carcinoma. O. ESR1 hot spot mutation in ER-positive breast carcinoma. P. ESR1 hot spot mutation in breast carcinoma. Q. ESR1 hot spot mutation in breast ductolobular carcinoma.

Fig. 2

Immunoreactivity of mutant peptides targeting T2A2 cells in vitro. CTLs were induced with autologous MUT peptide-pulsed DCs from peripheral blood lymphocytes of HLA-A2⁺ healthy donors. A-B. CTLs were collected and co-cultured with T2A2 cells loaded with MUT/WT peptide and then were detected for IFN-γ production and FasL expression. C. CTLs were co-cultured with T2A2 cells that were pulsed with MUT peptide or corresponding WT peptide at an effector/target (E/T) ratio of 12.5:1, 25:1, and 50:1, respectively. LDH cytotoxicity killing assay was used to detect the immunogenicity of mutated epitope peptides. The T2A2 cells loaded with WT peptide group was defined as the negative control group. Data were represented as means ± SD. Statistical significance was determined by unpaired Student’s t-test, *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 3

CTLs obtained by MUT peptide pool induction in HLA-A2.1/K^b transgenic mice specifically recognize mutant epitopes on MDA-MB-231 cells. Splenocytes and lymph node cells from HLA-A2.1/K^b transgenic mice immunized with either a peptide pool (Peptide pool group) or normal saline (Vehicle group) in combination with CpG ODN 1826 (30 µg per mouse) were restimulated in vitro with MUT peptides for 5 days. Then the splenocytes and lymph node cells were co-cultured with different target cells to detect the CTL response in (A-B) intracellular cytokine assay and (C) cytotoxicity assay. Then transfected cells MDA-MB-231-WT and MDA-MB-231-MUT were severed as stimulator cells and target cells. The MDA-MB-231-WT group was defined as the negative control group. Data were represented as means ± SD. Statistical significance was determined by unpaired Student’s t-test, *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4

CTLs obtained by MUT peptide pool induction in HLA-A2.1/K^b transgenic mice are able to specifically distinguish antigenic epitopes loaded on T2A2 cells. Immunogenicity of the MUT peptide in HLA-A2.1/K^b transgenic mice. Splenocytes and lymph node cells from HLA-A2.1/K^b transgenic mice immunized with either a peptide pool (Peptide pool group) or normal saline (Vehicle group) in combination with CpG ODN 1826 (30 µg per mouse) were restimulated in vitro with MUT peptides for 5 days. Then CTLs were co-cultured with different target cells to detect the CTL response in (A-B) intracellular cytokine assay and (C) cytotoxicity assay. The T2A2 cells loaded with WT peptide pool and T2A2 cells loaded with MUT peptide pool were severed as stimulator cells and target cells. The T2A2 cells loaded with WT peptide pool group was defined as the negative control group. Data were represented as means ± SD. Statistical significance was determined by unpaired Student’s t-test, **P < 0.01, ***P < 0.001.

Fig. 5

CTLs obtained from MUT peptide pool induction in HLA-A2.1/K^b transgenic mice specifically recognize single mutant epitopes loaded on T2A2 cells. Splenocytes and lymph node cells from HLA-A2.1/K^b transgenic mice immunized with either a peptide pool (Peptide pool group) or normal saline (Vehicle group) in combination with CpG ODN 1826 (30 µg per mouse) were restimulated in vitro with MUT peptides for 5 days. Then CTLs were co-cultured with different target cells to detect the CTL response with cytotoxicity assay. The T2A2 cells loaded with single WT peptide and T2A2 cells loaded with single MUT peptide were severed as stimulator cells and target cells. The T2A2 cells loaded with single WT peptide group was defined as the negative control group. Data were represented as means ± SD. Statistical significance was determined by unpaired Student’s t-test, *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 6

Antitumor activity of mutant epitope peptide-specific T cells in NOD/SCID mice bearing MDA-MB-231-MUT tumor cells. MDA-MB-231-MUT cells (5 × 10⁶ cells/mouse) were implanted subcutaneously in the flanks of immunocompromised NOD/SCID mice. Tumor-bearing mice received intravenous administration of the following: (1) normal saline, (2) non-induced CTLs (T cell group), (3) peptide-pool-induced CTLs (Induced T cell group). A. The scheme of tumor model. B. Tumor volume. C. Tumor weight. D. Tumor photos. E. The secretion of IFN-γ in mouse serum samples. F-H. The percentage of infiltrating CD8⁺ T cells in tumor tissues, spleen and lymph node was detected by flow cytometry. I. Infiltration of CD8⁺ T cells in tumor tissues was detected by immunofluorescence. Scale bar = 100 μm. Data were represented as means ± SD. Statistical significance was determined by unpaired Student’s t-test, *P < 0.05, **P < 0.01, ***P < 0.001.