Single-cell analysis identifies LYPD6B as a tumor-intrinsic candidate associated with immunotherapy non-response in breast cancer

Abstract

Immune checkpoint blockade (ICB) induces durable clinical responses in a subset of breast cancer patients, but many show limited benefit from anti-PD-1/PD-L1 therapy. Characterizing differences in the tumor microenvironment (TME) and tumor-intrinsic immune evasion programs between responders (R) and non-responders is critical to enhance immunotherapy efficacy and rational combination strategies. Three public scRNA-seq datasets of PD-(L)1-treated breast cancer (51 patients; 327,022 cells) were integrated. Patients were categorized as R or NR by clinical outcomes. Paired pre-/post-treatment samples were analyzed for TME composition, immune activation, IFN-response programs and T cell dynamics. NR-associated DEGs were screened to identify tumor-intrinsic non-response drivers, with validations including CRISPR knockout, drug-repurposing, molecular assays, immunohistochemistry and pan-cancer analyses. Relative to NR, R had increased CD8+ T cell infiltration, higher interferon-response activity in immune cells and myeloid/B-cell remodeling. Pseudotime analysis showed T cells progressed from activation to cytotoxic differentiation then exhaustion in R, matching activation-associated negative feedback in effective ICB. Tumor-intrinsic screening identified LYPD6B, a membrane gene upregulated in NR cancer cells, whose high expression suppressed antigen processing/presentation and IFNα/β signaling. Functional assays confirmed high LYPD6B expression in breast cancer cell lines; its ablation impaired proliferation and clonogenic growth and induced apoptosis. Drug-repurposing analyses found venetoclax binds LYPD6B and recapitulates its anti-proliferative effects. IHC and pan-cancer analyses verified LYPD6B’s tumor-cell localization and its association with immune infiltration and checkpoint expression. We identify LYPD6B as a tumor cell-intrinsic candidate linked to immune-evasive transcriptional programs and PD-(L)1 non-response in breast cancer, making it a potential target for combination immunotherapy.