Coordination between polycomb repressive complex 2 (PRC2) and miRNA-mediated gene silencing

Gene expression is controlled by an intricate network of regulatory proteins and RNAs. Misregulation of gene regulatory networks can cause aberrant gene expression changes that can lead to disease or developmental abnormalities. Some of the major gene regulatory networks are formed by transcription factors and non-coding RNAs including miRNAs. Understanding subnetworks of genetic interactions formed by genes involved in key disease pathways can provide novel therapeutic targets. My primary research goal was to understand the regulatory crosstalk between a transcriptional regulator, polycomb repressive complex 2 (PRC2) and miRNAs. Crosstalk between these two gene regulators has not been systematically addressed. We pursued this crosstalk in Glioblastoma, one of the deadliest forms of brain cancer. We utilized several genome-wide approaches to elucidate direct and indirect genetic interactions involving PRC2, miRNAs and their respective protein-coding gene targets. We found that PRC2 and miRNAs coordinately silence a set of genes by forming a feed-forward regulatory network. In this network, PRC2 directly represses its protein-coding targets and also indirectly represses them by activating miRNAs. A significant fraction of genes repressed by PRC2 through this feed-forward network included interferon stimulated genes (ISG). In addition to repressing ISGs, we found that PRC2 also activated an independent set of ISGs indirectly by repressing miRNAs that target them. RNA-seq is a widely used technology to determine the transcript levels of multiple genes in parallel. Additionally, RNA-seq technology can also be extended to determine the binding of RNA binding proteins and transcriptional dynamics. Conclusions from RNA-seq experiments rely heavily on the quality of sequencing reads and their accurate mapping to the reference genome. Through analysis of datasets that we produced and those that are published by other labs, we found that several of these datasets were contaminated with artefactual reads produced as a result of mispriming during reverse transcription. We further show that failure to remove these artefactual reads can lead to misinterpretation of data. To address this issue, we provide an alternative approach to prepare RNA-seq libraries to avoid mispriming entirely and also provide a computational tool to filter out misprimed reads from sequencing datasets