Jeremy E Wilusz - US grants

DESCRIPTION (provided by applicant): Most of the eukaryotic genome is transcribed, yielding a complex repertoire of transcripts that includes tens of thousands of individual noncoding RNAs with little or no predicted protein-coding capacity. Among these are well-studied small RNAs, such as microRNAs, as well as many other classes of small and long transcripts whose functions and mechanisms of biogenesis are less clear - but likely no less important. The MALAT1 locus is over-expressed in many human cancers and produces an abundant long nuclear-retained noncoding RNA. Despite being an RNA polymerase II transcript, we previously showed that the 3' end of MALAT1 is not produced by canonical cleavage/polyadenylation but instead by recognition and cleavage of a tRNA-like structure by RNase P. This results in the generation of a second noncoding RNA from the MALAT1 locus known as mascRNA that is tRNA-like and exported to the cytoplasm. mascRNA is significantly more evolutionarily conserved than the long MALAT1 transcript; however, the function of mascRNA and its role in cancer initiation/progression have not been explored. In Specific Aim 1, I will use a newly developed expression plasmid that recapitulates MALAT1 3' end processing to efficiently overexpress mascRNA in tissue culture cells. Changes in gene expression and cellular phenotype induced by modulating the expression of mascRNA will be identified, allowing paradigms for how tRNA-like small RNAs function in mammalian cells to be revealed. In Specific Aim 2, I will characterize the molecular mechanisms by which the 3' end of the long MALAT1 transcript is stabilized despite the absence of a canonical poly(A) tail. These experiments will reveal new insights into how long transcripts not subjected to cleavage/polyadenylation are made resistant to degradation and function in gene expression. As there are very likely other noncoding RNAs besides MALAT1 that are processed at their 3' ends via non-canonical mechanisms, next-generation sequencing technology will be used in Specific Aim 3 to specifically identify the 3' ends of long poly(A) minus RNAs. Nearly all previous studies characterizing the transcriptome have used a poly(A) selection step to enrich for messenger RNAs and deplete abundant housekeeping RNAs, such as ribosomal RNAs. However, this step also removes all long RNAs that lack poly(A) tails and, therefore, most transcripts subjected to non-canonical 3' end processing mechanisms. By using a novel library construction method, the mature 3' ends of these previously hidden RNAs will be revealed and characterized, providing insights into unexpected regulatory mechanisms that may control RNA stability, localization, or translation efficiency. In the short term, this career development awar will allow me to greatly expand my research into new, previously unexplored areas during the K99 phase. The excellent training environment in the Sharp lab and MIT will greatly facilitate not only the mentored research but also endow me with all the necessary skills to transition to an independent academic faculty position. In the long term, I am confident that these experiments will provide a foundation on which my own independent research program can grow and flourish. In summary, by identifying the functional role of tRNA-like small RNAs as well as characterizing the mechanisms that generate and stabilize non-canonical 3' ends of long RNAs, these innovative studies will reveal key new insights into the regulation, functions, and processing of noncoding RNAs that are relevant in human cancer.

PROJECT SUMMARY Most of the eukaryotic genome is transcribed, yielding a complex repertoire of protein-coding mRNAs and noncoding RNAs. Most long RNA polymerase II transcripts are thought to have a 5' cap and a 3' poly(A) tail, which protect the transcript from degradation as well as recruit the translation machinery. However, our recent work has revealed a number of abundant transcripts that are generated by non-canonical mechanisms and either lack a poly(A) tail (e.g. MALAT1) or have covalently linked ends (e.g. circular RNAs). MALAT1, which is commonly mis-regulated in many human cancers, ends in a triple helical structure that supports both RNA stability and translation. Likewise, thousands of protein-coding pre-mRNAs are non-canonically spliced to produce circular RNAs that are resistant to degradation by exonucleases, and some of these circular RNAs exceed the abundance of their associated linear mRNA by a factor of 10. Because these non-polyadenylated RNAs and circular RNAs are structurally distinct from canonical mRNAs, they are subjected to different biogenesis and post-transcriptional control mechanisms as well as likely bound by unique factors. However, little is currently known about how the fates of these non-canonical RNAs are controlled. We thus propose two complementary projects to address these gaps in knowledge. First, we will characterize how non- polyadenylated linear mRNAs, such as transcripts ending in a triple helix or viral-derived sequences, are stabilized and efficiently translated. We propose that just as poly(A) binding protein (PABP) binds the poly(A) tail and stem-loop binding protein (SLBP) binds the histone stem-loop to help recruit the translation machinery, there are likely unique factors that directly bind these less characterized 3' ends to regulate their expression. These novel mechanisms would thus allow these specific RNAs to be regulated by unique signaling cascades or only expressed in particular tissues. Using RNAi screening and binding assays, the key trans-acting factors and the mechanisms by which they recruit the ribosome to these RNAs will be identified. We additionally will identify other sequences that can functionally replace a poly(A) tail, thereby revealing new paradigms for how mRNA 3' ends are generated and regulated. Second, we will determine how circular RNA expression is controlled by cellular cues to impact cellular differentiation events. Most circular RNAs are expressed in a tissue-specific manner, yet the underlying mechanisms by which their expression levels are regulated are unknown. By profiling changes in circular RNA expression as cells differentiate, a set of transcripts that are dramatically altered will be identified. The mechanisms by which these circular RNAs are regulated and function will subsequently be characterized, revealing new insights into how these unexpected outputs of protein-coding genes control cell identity. In total, these innovative studies will reveal important insights into how non-canonical transcripts are post-transcriptionally regulated and function to impact both normal and diseased states.

PROJECT SUMMARY/ABSTRACT For a protein-coding gene to perform its cellular function, it must first generate RNA transcripts that are expressed at the appropriate level and properly processed. This is no small feat when one considers that RNA polymerase II can prematurely terminate and that nascent transcripts can be acted upon by a variety of RNA processing machines, including ones that yield mature transcripts lacking a canonical 5' cap or 3' poly(A) tail. A major focus of our laboratory has thus been to identify and characterize novel ?non-canonical? processing pathways that can act on nascent RNAs. Here, we propose to build upon our recent work to study two such mechanisms that are widely employed across eukaryotic genomes. First, we will mechanistically dissect how the Integrator (Int) complex catalyzes premature transcription termination at hundreds of protein-coding genes. Integrator was long known to be critical for the biogenesis of small nuclear RNAs (snRNAs), but we recently showed that Integrator also binds to many protein-coding loci and attenuates production of their full- length mRNAs, in some cases by more than 100-fold. This is because the IntS11 RNA endonuclease directly cleaves nascent mRNAs, triggering degradation of the transcripts and premature transcription termination. Nevertheless, it remains poorly understood how Integrator is assembled, regulated, and recruited to protein- coding genes as most of the other subunits in the complex have no known function and lack obvious paralogs or known protein domains. Our preliminary data indicate that non-catalytic Integrator subunits have distinct roles at snRNA vs. protein-coding gene loci, and thus we will characterize in detail how these subunits are recruited and function. Crosslinking mass spectrometry will further be used to define physical interfaces between Integrator subunits, thereby revealing novel insights into how Integrator is globally assembled and controlled. Second, we will investigate why many protein-coding genes generate circular RNAs with covalently linked ends. Some of these non-canonical transcripts are greater than 10-fold more abundant than their associated linear mRNAs. This suggests the main function of these genes may be to produce circular RNAs, but the physiological functions of almost all mature circular RNAs remain unknown. We thus will use high-throughput screening coupled to detailed biochemical studies to identify critical functions for circular RNAs and their underlying molecular mechanisms. We further will systematically identify factors that modulate circular RNA levels, especially post-transcriptionally, as very little is currently known about how the fate and decay rates of these transcripts are controlled. Characterization of these key regulatory mechanisms will not only provide important insights into endogenous circular RNAs, but may also ultimately enable circular RNAs to become novel long-lasting therapeutic modalities. In total, these innovative studies will reveal new, fundamental insights into how the Integrator complex and circular RNAs are regulated and control protein- coding gene outputs to impact normal and diseased states.