Bioinformatics Seminar
Time:
Venue: Na
30 August 2016
NaSystems analysis identifies micro-RNAs that influence melanoma phenotype switching
Joseph CursonsWEHI Bioinformatics
Micro-RNAs (miRs) are a class of non-coding RNA that target mRNA transcripts through complementary binding to modulate mRNA stability and/or translation. Most miRs target multiple mRNAs and can thus exert post-transcriptional control in a co-ordinated manner. This allows miRs to regulate complex phenotypic changes such as epithelial-mesenchymal transition (EMT) ; with important clinical implications through resistance to chemotherapeutic agents and immuno-modulation.
Here we present a novel systems analysis ; which we use to examine miR and mRNA abundance over 57 primary melanoma cell lines – the Ludwig Melbourne melanoma (LM‑MEL) panel. Pearson’s correlation and mutual information were applied to identify strong statistical associations that suggested miR-mediated downregulation of mRNA transcript abundance across different cell lines. High confidence interactions were extracted from the sequence-based miR-target prediction databases ; TargetScan and DIANA‑microT ; as an independent ; mechanistic interaction set. These data were used to filter our list of statistical associations providing a list of predicted ; putative miR‑mRNA relationships in melanoma. Matrigel invasiveness assays were also performed across 24 of the cell lines ; allowing us to identify relationships that may be implicated in melanoma phenotype switching.
While a number of these miR-mRNA relationships have been observed and validated in human cell lines ; our analysis also identified a large number of novel putative interactions. We performed enrichment analyses to identify miRs with a large number of ‘active’ relationships ; and functional annotations for the target lists that suggested melanoma phenotype switching. Our analysis suggests that hsa-miR-29b-3p (miR-29b) has a role in regulating melanoma cell invasiveness through the co-ordinated targeting of several mRNA transcripts. We validated this prediction to show that miR-29b reduces mRNA transcript abundance for LAMC1 ; PPIC and LASP1 ; and protein abundance for LAMC1 and PPIC. Finally ; we use antago-miRs and siRNAs against these targets to show that modulation of miR-29b ; LAMC1 and PPIC influences melanoma cell invasiveness within a 3D tumour sphere assay.
;;;