In silico Prediction of Promiscuous HLA Class I and II T-Cell Epitopes from MED12 for Therapeutic Vaccine Development against Breast Phyllodes Tumours

Abstract

Breast phyllodes tumours are uncommon fibroepithelial neoplasms that may exhibit aggressive local behaviour and have limited systemic treatment options. Recurrent activating mutations in mediator complex subunit 12 (MED12), an X-linked gene, occur frequently in fibroepithelial tumours, including phyllodes tumours. This study used a computational approach to identify T-cell peptide candidates derived from the MED12 protein for possible therapeutic vaccine development. MED12-derived peptides were screened for susceptibility to proteasomal cleavage, transporter associated with antigen processing (TAP) transport, HLA class I and II binding, antigenicity, and T-cell immunogenicity. For HLA class I, RPRPRAYYL (1731–1739) showed the broadest predicted binding profile, whereas LYHTHLRPR (1725–1733) had the highest predicted immunogenicity score. For HLA class II, LHLFKTPQL (1173–1187) showed the broadest predicted binding profile, and VDPYRPVRL (1852–1866) had the highest antigenicity score. The workflow assessed whether the selected peptide candidates could support cytotoxic and helper T-cell responses across multiple HLA restrictions. These findings indicate that MED12-derived peptides may provide preliminary candidates for peptide-based immunotherapy research in breast phyllodes tumours. However, the results are based solely on in silico predictions and do not establish therapeutic efficacy. Experimental HLA-binding assays, T-cell activation studies, mutation-specificity assessment, toxicity and allergenicity testing, and relevant preclinical validation are required before the selected peptides can be considered for vaccine development.