Population-Associated Molecular Variation in Histologically Normal Breast Tissue Is Associated with Distinct Baseline Transcriptional States

Abstract

Population-associated molecular variation in breast tissue may contribute to differences in tissue biology and disease susceptibility. Still, the extent to which such variation is shaped by underlying tissue state remains unclear. We performed a pilot RNA-seq and lipidomic analysis of histologically normal breast tissue from African American (AA) and Caucasian White (CW) individuals. Unsupervised transcriptomic analysis identified two baseline tissue states, G1 and G2, representing the dominant axis of molecular variation and associated with epithelial-enriched and vascular-enriched tissue contexts, respectively. Across the full cohort, AA and CW samples showed minimal transcriptomic differences. However, within G1, 191 genes were differentially expressed between AA and CW samples, with coordinated enrichment of extracellular matrix organization and proliferative/cytoskeletal processes in AA samples; these patterns were consistent across enrichment methods and sensitivity analyses. No comparable population-associated transcriptional signal was detected in G2. Lipidomic profiles showed limited separation and no robust population-associated differences after correction for multiple testing. Together, these pilot findings suggest that population-associated molecular variation in histologically normal breast tissue may be state-dependent, becoming detectable within a specific baseline transcriptional context rather than uniformly across the cohort. These results further underscore baseline tissue state as a major source of variation in small, heterogeneous bulk-tissue cohorts and provide a framework for future cell-resolved studies of tissue variation and disease susceptibility.