Transcriptional and Post-Transcriptional Regulation of Retroelements

Retroelements make up nearly 40% of the human genome. Their ability to mobilize and cause insertional mutagenesis makes them potentially harmful. Dysregulated retroelements are implicated in cancer and neurological disease as well as in aging. Therefore, there are strict regulatory mechanisms in place to restrain retroelements. Transcriptional silencing of retroelements is mainly achieved by heterochromatinization and DNA methylation. Our lab has demonstrated that p53 can also transcriptionally repress retroelements, however, the mechanism by which p53 mediates repression was unknown. Due to the repetitive nature and large copy numbers of endogenous retroelements, it is challenging to study how individual retroelements are repressed by p53. Hence, as an alternative approach, we used a p53 biosensor with unique sequence and a visual readout (GFP) to study the mechanism of repression by p53. Using this in vivo p53 biosensor in Drosophila, we show that stimulus-dependent transactivation and constitutive transrepression is mediated by distinct p53 isoforms. Canonical transactivation functions do not adequately explain tumor suppression by p53. Since p53 is mutated in nearly all human cancers and retroelements are derepressed in many p53-driven cancers, it is possible that p53 suppresses tumors by restraining retroelements. Therefore, understanding the mechanism of retroelement repression by p53 can possibly provide further insights into its tumor suppressive mechanism. In addition to actual genomic integration events, retrotransposition intermediates like RNA, dsRNA, RNA:DNA hybrids and reverse transcribed cDNA are also potentially oncogenic due to induction of inflammatory response pathways and there exists multiple post-transcriptional regulatory pathways that target these retrotransposition intermediates for degradation. To test whether retroelement RNAs, specifically those derived from LINE-1, are oncogenic, we developed a CRISPR-based programmable RNase platform to eliminate LINE-1 RNAs. While this platform was effective against reporters, it did not perform well against endogenous LINE-1 RNAs. In the process, however, we made a surprising observation that LINE-1 RNAs were enriched in the nucleus. A comprehensive empirical and bioinformatic analysis revealed that this nuclear localization pattern was highly generalizable and widely conserved across elements and species. Our observations raise the possibility that nuclear retention of retroelement transcripts is an additional post-transcriptional regulatory mechanism that limits retrotransposition.