Widespread remodeling of the transcriptome is a signature of cancer; however, little is known about the post-transcriptional regulatory factors including RNA-binding proteins (RBPs) that regulate mRNA stability, and the extent to which RBPs contribute to cancer-associated pathways. Here, by modeling the global change in gene expression based on the effect of sequence-specific RBPs on mRNA stability, we show that RBP-mediated stability programs are recurrently deregulated in cancerous tissues. Particularly, we uncovered several RBPs that contribute to the abnormal transcriptome of renal cell carcinoma (RCC), including PCBP2, ESRP2 and MBNL2. Modulation of these proteins in cancer cell lines alters the expression of pathways that are central to the disease, and highlights RBPs as driving master regulators of RCC transcriptome. This study presents a framework for the screening of RBP activities based on computational modeling of mRNA stability programs in cancer, and highlights the role of post-transcriptional gene dysregulation in RCC.
Data files for measuring the activity of RBP stability programs
- CAGEKID ... .txt.gz (gzipped text file, 8.6MB): The CAGEKID expression dataset that was used in this work (source). The values represent log2 fold-change between tumor and patient-matched normal tissue.
- TCGA ... .txt.gz (gzipped text file, 6.3MB): The TCGA ccRCC expression dataset that was used in this work (source). The values represent log10 fold-change between tumor and patient-matched normal tissue.
- StabilityCode ... .txt.gz (gzipped text file, 68.6KB): The RBP stability code that was used in this work (source). Each value indicates whether the corresponding RBP (columns) binds and regulates the stability of the corresponding transcript (rows).
Data files corresponding to RBP shRNA experiments
- RBP_shRNA ... .txt.gz (gzipped text file, 629.4 KB): The values represent log10 fold-change between the indicated RBP-specific shRNA infection and the average of two control shRNA infections from the same cell line. Genes with fewer than an average of 10 reads per sample are filtered out.
Scripts for measuring the activity of RBP stability programs
- glmnet.bootstrap.R (R script, 2.2KB): The script used in this work for inferring RBP activity from mRNA expression data.