Abstract
BACKGROUND: Fine particulate matter (PM(2.5)) has been associated with breast cancer specific mortality, particularly for women with Stage I cancer. We examined the biological pathways that are perturbed by PM(2.5) exposures by analyzing gene expression measurements from breast tissue specimens. METHODS: The Nurses' Health Studies (NHS and NHSII) are prospective cohorts with archival breast tissue specimens from breast cancer cases. Global gene expression data were ascertained with the Affymetrix Glue Human Transcriptome Array 3.0. PM(2.5) was estimated using spatio-temporal models linked to participants' home addresses. All analyses were performed separately in tumor (n = 591) and adjacent-normal (n = 497) samples, and stratified by estrogen receptor (ER) status and stage. We used multivariable linear regression, gene-set enrichment analyses (GSEA), and the least squares kernel machine (LSKM) to assess whether 3-year cumulative average pre-diagnosis PM(2.5) exposure was associated with breast-tissue gene expression pathways among predominately Stage I and II women (90.7%) and postmenopausal (81.2%) women. Replication samples (tumor, n = 245; adjacent-normal, n = 165) were measured on Affymetrix Human Transcriptome Array (HTA 2.0). RESULTS: Overall, no pathways in the tumor area were significantly associated with PM(2.5) exposure. Among 272 adjacent-normal samples from Stage I ER-positive women, PM(2.5) was associated with perturbations in the oxidative phosphorylation, protein secretion, and mTORC1 signaling pathways (GSEA and LSKM p-values <0.05); however, results were not replicated in a small set of replication samples (n = 80). CONCLUSIONS: PM(2.5) was generally not associated with breast tissue gene expression though was suggested to perturb oxidative phosphorylation and regulation of proteins and cellular signaling in adjacent-normal breast tissue. More research is needed on the biological role of PM(2.5) that influences breast tumor progression.