Search

Long noncoding RNAs (lncRNAs) are rapidly evolving and thus typically poorly conserved in their sequences. How these sequence differences affect the characteristics and potential functions of lncRNAs with shared synteny remains unclear. Here we show that the syntenically conserved lncRNA Firre displays distinct expression and localization patterns in human and mouse. Single molecule RNA FISH reveals that in a range of cell lines, mouse Firre (mFirre) is predominantly nuclear, while human FIRRE (hFIRRE) is distributed between the cytoplasm and nucleus. This localization pattern is maintained in human/mouse hybrid cells expressing both human and mouse Firre, implying that the localization of the lncRNA is species autonomous. We find that the majority of hFIRRE transcripts in the cytoplasm are comprised of isoforms that are enriched in RRD repeats. We furthermore determine that in various tissues, mFirre is more highly expressed than its human counterpart. Our data illustrate that the rapid evolution of syntenic lncRNAs can lead to variations in lncRNA localization and abundance, which in turn may result in disparate lncRNA functions even in closely related species.

Many of the biological functions performed by RNA are mediated by RNA-binding proteins (RBPs), and understanding the molecular basis of these interactions is fundamental to molecular biology. Here, we present MPRNA-immunoprecipitation (MPRNA-IP), an adaptation of the previously developed massively parallel RNA assay (MPRNA), and a new avenue for in vivo high-throughput dissection of RNA-protein interactions. By using custom pools of tens of thousands of RNA sequences containing systematically designed truncations and mutations, we are able to identify RNA domains, sequences, and secondary structures necessary and sufficient for protein binding in a single experiment. We show that this approach is successful for multiple RNAs of interest including NORAD, MS2, and human telomerase RNA, and we describe statistical models for identifying RNA domains and parsing the structural contributions of RNA in these interactions. By blending modern and classical approaches, MPRNA-IP provides a novel high-throughput way to elucidate RNA-based mechanisms behind RNA-protein interactions.

BackgroundSeveral long noncoding RNAs (lncRNAs) have been shown to function as central components of molecular machines that play fundamental roles in biology. While the number of annotated lncRNAs in mammalian genomes has greatly expanded, their functions remain largely uncharacterized. This is compounded by the fact that identifying lncRNA loci that have robust and reproducible phenotypes when mutated has been a challenge.ResultsWe previously generated a cohort of 20 lncRNA loci knockout mice. Here, we extend our initial study and provide a more detailed analysis of the highly conserved lncRNA locus, Taurine Upregulated Gene 1 (Tug1). We report that Tug1 knockout male mice are sterile with complete penetrance due to a low sperm count and abnormal sperm morphology. Having identified a lncRNA loci with a robust phenotype, we wanted to determine which, if any, potential elements contained in the Tug1 genomic region (DNA, RNA, protein, or the act of transcription) have activity. Using engineered mouse models and cell-based assays, we provide evidence that the Tug1 locus harbors three distinct regulatory activities – two noncoding and one coding: (i) a cis DNA repressor that regulates many neighboring genes, (ii) a lncRNA that can regulate genes by a trans-based function, and finally (iii) Tug1 encodes an evolutionary conserved peptide that when overexpressed impacts mitochondrial membrane potential.ConclusionsOur results reveal an essential role for the Tug1 locus in male fertility and uncover three distinct regulatory activities in the Tug1 locus, thus highlighting the complexity present at lncRNA loci.

The study of survival outliers of glioblastoma (GBM) can have important implications on gliomagenesis as well as in the identification of ways to alter clinical course on this almost uniformly lethal cancer type. However, current studied epigenetic and genetic signatures of the GBM outliers have failed to identify unifying criteria to characterize this unique group of patients. In this study, we profiled the global DNA methylation pattern of mainly IDH1 wild type survival outliers of glioblastoma and performed comprehensive enrichment analyses with genomic and epigenomic signatures. We found that the genome of long-term survivors in glioblastoma is differentially methylated relative to short-term survivor patients depending on CpG density: hypermethylation near CpG islands (CGIs) and hypomethylation far from CGIs. Interestingly, these two patterns are associated with distinct oncogenic aspects in gliomagenesis. The hypomethylation pattern at the region distant from CGI is associated with lower rates of de novo mutations while the hypermethylation at CGIs correlates with transcriptional downregulation of genes involved in cancer progression pathways. These results extend our understanding of DNA methylation of survival outliers in glioblastoma in a genome-wide level, and provide insight on the potential impact of DNA hypomethylation in cancer genome.