Abstract
Breast cancer (BC) is the leading cause of cancer death in women. Bidirectional regulation between BC and gut microbiota (GM) is established, but GM's mechanistic role in BC pathogenesis remains unclear. Public BC/control samples and GM genome-wide association study data underwent Mendelian randomization to identify GM-BC associations and GMRGs. DEGs between BC and controls were analyzed. Candidate genes were derived from intersecting DEGs and GMRGs. Machine learning identified biomarkers, validated by expression analysis. GSEA, immune infiltration, drug screening with molecular docking, and scRNA-seq were performed. Intersecting 3455 DEGs with GMRGs yielded eight candidates; MCM6 and NR3C1 were validated as biomarkers, enriched in DNA replication pathways. Immune infiltration showed 13 differential immune cells, with macrophages notably influencing biomarkers. Etoposide exhibited strong binding to biomarkers via docking. scRNA-seq identified epithelial cells as key, with stage-dependent biomarker expression. This study redefines BC as a microbiome-regulated network, identifying the MCM6/NR3C1 biomarker pair for early diagnosis and microbiome-targeted interventions.