Abstract
Experimental breakthroughs have provided unprecedented insights into the genes involved in cancer. The identification of such cancer driver genes is a major step in gaining a fuller understanding of oncogenesis and provides novel lists of potential therapeutic targets. A key area that requires additional study is the posttranscriptional control mechanisms at work in cancer driver genes. This is important not only for basic insights into the biology of cancer, but also to advance new therapeutic modalities that target RNA—an emerging field with great promise toward the treatment of various cancers. In the current study we performed an in silico analysis on the transcripts associated with 800 cancer driver genes (10,390 unique transcripts) that identified 179,190 secondary structural motifs with evidence of evolutionarily ordered structures with unusual thermodynamic stability. Narrowing to one transcript per gene, 35,426 predicted structures were subjected to phylogenetic comparisons of sequence and structural conservation. This identified 7,001 RNA secondary structures embedded in transcripts with evidence of covariation between paired sites, supporting structure models and suggesting functional significance. A select set of seven structures were tested in vitro for their ability to regulate gene expression; all were found to have significant effects. These results indicate potentially widespread roles for RNA structure in posttranscriptional control of human cancer driver genes.
Highlights
Identification of cancer driver genes is an ongoing process [1, 2]
Reducing these structures to one transcript per gene resulted in a total of 35,426 predicted structures with at least one nucleotide exhibiting an average z-score of < -1 after the initial scan (S6 Table). Of these cancer driver -1 (CD-1) structures, about 4% were found in 50UTR sequences, 52% in coding sequences, and 44% in 300UTR sequences
We have predicted structured regions in cancer-related mRNAs and have determined whether these regions are evolutionarily significant through covariation
Summary
Identification of cancer driver genes is an ongoing process [1, 2]. The ability to separate genes whose mutations are not directly responsible for the progress of neoplasticity (passengers) compared to genes whose mutations stimulate neoplasticity and malignancy (drivers) is important for future cancer therapeutic targeting. New computational methods are uncovering previously unappreciated oncogenes and tumor suppressors. One such recently developed method considered the nucleotide context in which mutational events occur to distinguish. Mutpanning alone identified 460 genes; some that were previously known and others that were previously unappreciated potential cancer driver genes. Another strategy is to combine several computational methods to identify both known and novel driver genes. About one-quarter of these had not been previously recognized in the Cancer Gene Census [4] Combined, these methods point to 800 genes (228 in common, 232 unique to Mutpanning alone, 340 unique to the compendium) that have potential to drive cancer development and progression
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
