Integrated transcriptomic meta-analysis provides novel insights into breast cancer molecular subtypes

Abstract

AbstractBreast cancer is the most common cancer in women worldwide. There are four major breast cancer subtypes (luminal A, luminal B, HER2-enriched and triple-negative/TNBC). TNBC is the most aggressive form with the worst prognosis. However, the differences among the subtypes have not been completely established at the molecular level, thereby limiting therapeutic and diagnostic options for TNBC. We performed a meta-analysis of microarray and RNA-sequencing data to identify candidate genes with an expression-based association in each molecular subtype of breast cancer. The protein interaction network of the candidate genes was analyzed to discover functionally significant gene clusters and hub genes. Potential therapeutic candidates for TNBC were explored through gene-miRNA interactions. We identified 316, 347, 382, and 442 candidate genes in luminal A, luminal B, HER2 and TNBC subtypes, respectively. A total of 135 (26 up- and 109 down-regulated) candidate genes were shared by all four subtypes. Functional analysis of the candidate genes indicated up-regulation of ‘cell cycle’ and ‘p53 signaling’ pathways and down-regulation of multiple signaling pathways. COL10A1 was found to be highly up-regulated in all subtypes. It may be a good target for research towards multiple types of applications, including therapeutics. KIF4A, a commonly up-regulated X-chromosome gene was significantly associated with the survival of breast cancer patients. Protein interaction network and centrality analysis revealed that low-moderately differentially regulated genes play an important role in functional cascades across proteins in breast cancer subtypes and may be potential candidates for therapeutics. Targeting FN1 (fibronectin 1), the key up-regulated hub gene by miR-1271 5p may be an important molecular event to be targeted for potential therapeutic application in TNBC.