Talk abstracts

Talk on Saturday 12:15-12:35pm submitted by Daniel Dominguez

High levels of intron-containing transcripts in aggressive cancers

Daniel Dominguez (Pharmacology, UNC Chapel Hill), Yi-Tsuan Tsai (Bioinformatics and Computational Biology, UNC Chapel Hill), Zefeng Wang (Chinese Academy of Sciences), Christopher Burge (Biology, MIT)

Abstract:
RNA processing defects can initiate transformation, promote tumor growth, enhance metastasis and confer drug resistance. We performed a comprehensive analysis of hundreds of kidney cancer transcriptomes and uncovered an unusual subset of tumors with abnormally high levels of most detected introns. We termed this subset of tumors, accounting for about 20% of kidney cancer cases, the IR-class (Intron Retention-class). We find that high intron retention levels are linked to a defect in the non-sense mediated decay pathway (NMD), as tumors with high intron retention also showed increased levels of NMD-associated exons (poison cassette exons) and other NMD-targeted transcripts. Analysis of ribosome profiling data revealed some intron-containing species associate with ribosomes, indicative of their presence in the cytoplasm and evasion of NMD. Thus, high levels of intron-containing transcripts are likely a result of defects in NMD rather than defects in splicing. Assessment of thousands of other tumor samples spanning a range of tumor types, revealed that this phenomenon is unique to kidney cancer. While no individual genetic alternation was enriched in the IR-class, IR-class tumors showed recurrent mutations in MTOR pathway members and altered expression of core energy metabolism pathways. Patients with IR-class tumors had a striking decrease in overall survival, living about half as long as patients with non-IR tumors. Analysis of a panel of kidney cancer cell lines revealed a similar pattern, wherein only certain cell lines exhibited high levels of intron retention and NMD-exons, providing tractable models for study of IR-class tumors. Our work links key cancer signaling pathways to NMD defects resulting in high levels of aberrant transcripts in patients with poor clinical outcomes.

Keywords: Intron Retention, Cancer, RNA processing