肿瘤长非编码RNA表达模式与黑人和白人女性的乳腺癌预后

Background : Black women are more likely than White women to develop aggressive estrogen receptor (ER) negative breast tumors and have a poor prognosis. The biological mechanisms underlying these disparities remain largely unknown. Long noncoding RNAs (lncRNAs) are key regulators of gene expression, and lncRNA dysregulations can contribute to breast cancer carcinogenesis and progression. Studies, however, have focused on White women and have been limited to characterize genome-wide lncRNA expression patterns and their functional roles. Importantly, the prognostic relevance of dysregulated lncRNAs remains unclear. Motivated by the biological importance and research gap, we completed lncRNA expression profiling in breast tissues in a large cohort of Black and White women. Methods : We characterized genome-wide lncRNA expression patterns in breast tumors from 873 Black and 319 White women using a capture-based RNA sequencing platform. Data were normalized, and differential expression analysis by subgroups was conducted using DESeq2. Cox regression and Cox LASSO models were applied to examine associations between lncRNA expression and survival outcomes. Differentially expressed lncRNAs (DElncRNAs) were defined as log2 fold change ≥1.0 and FDR <0.05. Results : We found that lncRNAs were frequently dysregulated in breast cancers and appeared to have tumor subtype-specific expression patterns. Some of these DElncRNAs, such as MRPS30-DT and GATA3-AS1 were highly expressed in ER+ tumors, while LINC02487, LINC00511 and AFAP1-AS1 were highly expressed in ER- tumors. We also identified lncRNAs that were differentially expressed between Black and White women in tumors overall and by ER status. For example, overexpression of these lncRNAs, such as LINC00470, SNX10-AS1, LINC01139, GACAT2, has been shown to promote tumor growth and progression, and many others are novel. These ER- or race-associated DElncRs-correlated protein-coding genes are enriched in several cancer-related pathways, such as estrogen-response, K-RAS, keratinization, and multiple GPCR-related signaling. We further identified lncRNAs that are significantly associated with breast cancer specific survival, such as GATA3-AS1, MNX1-AS1, MAPT-IT1, with known associations with cancer progression and prognosis, while many others remain elucidated. Using Cox LASSO modeling, we developed a lncRNA signature, with its composite score, associated with breast cancer specific survival after adjusting for covariates. Conclusions : These results indicate that there are unique lncRNA expression patterns by ER subtype and between racial groups, which may contribute to aggressive tumor biology and cancer prognosis, and that can help inform the development of targeted strategies for prevention and therapeutics. Future study is needed to fully understand the mechanisms underlying biological functions and their clinical implications.