UCLA

Mature microRNAs (miRNAs) are typically 22-25 nucleotide-long transcripts that broadly coordinate cellular stress responses by modifying the activity of complementary messenger RNAs (mRNAs). In particular, the miRNA hsa-miR-34 is well characterized as a tumour suppressor- its expression modifies the proliferation, metastasis, and apoptotic responses of cancer cells, and it is both upregulated by the key tumour suppressor p53 after ionizing radiation (IR) and downregulating factors that downregulate p53, resulting in a radiation-responsive feedback loop. De novo miRNA production starts by transcribing longer, capped and polyadenylated transcripts before cleavage in the nucleus, export to the cytoplasm, and a second round of processing resulting in the mature-length transcript with a 5’ phosphate, which is required for both the activity and stability of mature miRNAs- consequently, mature miR-34 derived from de novo transcription and processing is typically observed approximately 6h after IR. However, prior research from our lab has identified a stable population of 5’ unphosphorylated miR-34a, which is rapidly phosphorylated in response to IR, resulting in miR-34 activity prior to and independent of de novo transcription. Since the 5’ phosphate is important for the stability of mature miRNAs, this work aims to investigate mechanisms responsible for stabilizing and coordinating this unphosphorylated pool of miR-34. Mass spectrometry of probable interacting partners of unphosphorylated miR-34 identify nuclear paraspeckles as a likely reservoir for unphosphorylated miR-34. Paraspeckles are nuclear subcompartments comprised of several key proteins crosslinked to the long noncoding RNA (lncRNA) NEAT1_2, creating a phase separation that prevents mixing of internal compartments with the rest of the nucleoplasm. We hypothesize that this phase separation maintains the unphosphorylated pool by preventing its interaction with cellular factors responsible for degrading RNAs. This result would align well with prior studies of paraspeckle function, which have been implicated in DNA repair responses and processing of some miRNAs To study the role of phase separation on the stability and coordination of unphosphorylated miR-34a, we generated Cas9-mediated deletions of the triple helix domain of NEAT1_2, which is required for phase separation. Accordingly, ablation of paraspeckle formation by this method decreased the activity of miR-34 shortly after irradiation without affecting unrelated miRNA species or de novo miR-34 transcription or processing. Similarly, cells with intact paraspeckles demonstrate punctate nuclear localization of miR-34a but not other miRNAs, which is lost on deletion of nuclear paraspeckles. Interestingly, contrary to prior reports that IR induces the rapid phosphorylation and cytoplasmic relocalization of unphosphorylated miR-34, IR increased the number of nuclear miR-34 foci observed in both wildtype and paraspeckle-knockdown cells, with no change in nuclear localization of other miRNAs. Reverse transcriptase quantitative polymerase chain reaction of the processing intermediates of miR-34 in irradiated cells implies that these nuclear foci are likely newly-transcribed, p53-dependent pri-miR-34. These results demonstrate that nuclear paraspeckles interact with unphosphorylated miR-34 mimics and are required for the IR-responsive activity and nuclear localization of the unphosphorylated pool. These results did not extend to other miRNA species shown to lack unphosphorylated pools, implying a distinct mechanism for regulating miRNA activity and localization. Future work investigating miRNA activity within and around the paraspeckle may reveal novel unphosphorylated miRNAs and mechanisms