MicroRNAs (miRNAs or miRs) are ~22 nt noncoding RNAs which target complementary gene transcripts for translational repression or mRNA cleavage. Having been implicated in the initiation and progression of human cancers, miRNAs regulate processes such as cell growth, differentiation, and apoptosis. A productive miRNA:mRNA interaction can occur with as little as six consecutive nucleotides, through pairing between the 5′-seed of the miRNA (located in nucleotides 2–7) and sequences which are largely localized in the 3′-untranslated regions (UTRs) of mRNA targets; consequently, a given miRNA can potentially impact hundreds of genes within and across diverse signaling pathways.
MiRNAs, along with gene copy number alterations and methylation of gene promoter regions, globally influence gene expression, which ultimately determines cellular behavior. The Cancer Genome Atlas (TCGA) project is a large-scale collaborative effort which seeks to comprehensively catalogue the molecular aberrations in various cancers. While the recent initial report from TCGA on 489 high-grade serous ovarian adenocarcinomas (487 of which had corresponding miRNA data) presented a broad molecular picture of the disease, much in terms of miRNAs remains to be elucidated. Here, we comprehensively survey the miRNAs within the TCGA ovarian dataset, making use of the various molecular profile data types, miRNA and mRNA in particular, that have all been generated for the same set of tumors. Previous miRNA expression profiling studies of ovarian cancer have defined differentially expressed miRNAs in cancer relative to the corresponding normal control though in this present study, such a large dataset (n = 487 patients) allows us to more fully explore the diversity of miRNAs within a single ovarian cancer subtype.
Here, we present a number of findings on miRNAs in ovarian cancer, from various integration-based analyses. These analyses revolve around the basic question of whether the general rules of miRNA behavior, as we currently understand them, can be supported by corroborating patterns within human cancers. Our study serves to reinforce our current notions of basic miRNA biology, to demonstrate how current in silico miRNA-gene targeting predictions may be refined through integrative analysis, and to demonstrate the rich resource of TCGA in identifying miRNA candidates for functional targeting in cancer. Our study also provides second-level data mining results for molecular biologists to more deeply explore specific miRNA-associated pathways in ovarian cancer.
This article was published in PLoS ONE and is free to access online.