Identification of key microRNAs involved in tumorigenesis and prognostic microRNAs in breast cancer.

Breast cancer is a commonly diagnosed cancer in women, and one of the leading causes of cancer-related death among female patients However, the key microRNAs involved in its tumorigenesis and microRNAs of prognostic values have not been fully understood. In the present study, we aimed to perform a systematic analysis of microRNA expression profiles to identify some key microRNAs associated with tumor initiation and prognosis. Using TCGA breast cancer datasets, we identified 110 differentially expressed microRNAs. The functional enrichment analysis of the upregulated microRNAs revealed signaling transduction pathways, such as Notch and Wnt signaling pathway, and metabolism-related pathways such as sugar and nucleotide sugar metabolism, and oxidative stress response. Moreover, multivariable Cox model based on three variables of hsa-mir-130a, hsa-mir-3677, and hsa-mir-1247 stratified patients into high-risk and low-risk groups, which showed significant prognostic difference. In addition, we also tested the performance of this model in patient cohorts of any specific breast cancer subtypes or different TNM stages. The high performance in risk prediction was also observed in all of breast cancer subtypes and TNM stages. We also observed that there were highly possible interactions between hsa-mir-130a and seven target genes. Among these target genes, VAV3 and ESR1 were predicted as the target genes of hsa-mir-130a, suggesting that hsa-mir-130a may function by regulating the expression of VAV3 and ESR1 in breast cancer. In conclusion, the stratification based on the multivariable Cox model showed high performance in risk prediction. The dysregulated microRNAs and prognostic microRNAs greatly improved our understanding of the microRNA-related molecular mechanism underlying breast cancer.