MST1 and CPNE1 Emerge as Candidate Targets in Breast Cancer: Integration of Mendelian Randomization, Multi-omics Validation, and Functional Interrogation

Abstract

Background Breast cancer remains a leading cause of female mortality worldwide, with therapeutic benefit limited by tumor heterogeneity and drug resistance. Identification of novel therapeutic targets through integration of genetic causality inference and functional validation is urgently needed. Methods Plasma protein quantitative trait loci (pQTL) from the Fenland cohort (~10,700 individuals) and the FinnGen R10 SomaScan subset (n = 828) were integrated with breast cancer genome-wide association study data from the Breast Cancer Association Consortium (122,977 cases and 105,974 controls). Causal protein-disease relationships were inferred using summary-data-based Mendelian randomization (SMR), with colocalization and HEIDI tests. Multi-level validation was performed using TCGA-BRCA transcriptomic data. Functional validation included CCK-8, EdU, wound healing, Transwell invasion, and flow cytometry apoptosis analysis evaluated CPNE1 knocdown and/or overexpression of the MST1 gene which encodes macrophage-simulating protein (MSP). Results SMR identified 23 proteins in Fenland and 10 in FinnGen at FDR < 0.05. MST1/MSP and CPNE1 were supported in both datasets with PP.H4 ≥ 0.80 and non-significant HEIDI tests. Genetically predicted circulating MSP was positively associated with breast cancer risk, whereas circulating CPNE1 showed an inverse association. TCGA-BRCA showed higher CPNE1 mRNA in tumors (P = 2.57×10⁻25) and lower MST1 mRNA (P = 3.04×10⁻23). CPNE1 expression was highest in triple-negative breast cancer and correlated positively with clinical stage (ρ = 0.211), whereas MST1 expression was lowest in triple-negative breast cancer and correlated inversely with stage (ρ = −0.164). In T47D and MDA-MB-231 cells, CPNE1 knockdown and MST1 overexpression each reduced proliferation, migration, and invasion and increased apoptosis; the combined group showed greater changes than either single intervention. Conclusion This study prioritizes the MST1 gene, which encodes MSP, and CPNE1 as candidate proteins for further investigation in breast cancer. However, the circulating-protein associations, tumor-mRNA patterns, and cell-autonomous perturbations represent distinct and directionally discordant biological contexts. The findings therefore support context-dependent candidate roles and justify mechanistic, in vivo, and formal interaction studies, but do not yet establish therapeutic efficacy or synergy.