Sumoylation Of TARBP2 Regulates The Efficiency Of Small RNA Interference

Small non-coding RNA induced gene silencing is one of the basic but complicated molecular mechanisms for regulation of gene expression. However, the underlying mechanisms how the efficiency of mi RNA/si RNA-induced gene silencing is fine-tuned represents a critical issue unresolved in the field of non-coding RNA research. In mammalian, mature mi RNAs are generated by a two-step processing pathway. Firstly, pri-mi RNAs are cleaved by the microprocessor complex Drosha-DGCR8 to produce pre-mi RNAs in cell nucleus. Secondly, pre-mi RNAs are exported to cytoplasm by Ran GTP-Exportin 5, and then are recognized and cleaved by Dicer-TARBP2 complex into the duplex mi RNAs, which are sequentially embedded into Ago2, a core component of RISC(RNA-induced silencing complex), and trigger gene silencing. There is increasing evidence for that the RISC assembly contains two key steps, including the duplex mi RNA loading into RISC and the mi RNA duplex unwinding for gene silencing. In the latter, one strand of mi RNA duplex is selected by Ago2 as the guide strand, which is incorporated into Ago2 to form the functional center of the RISC, while another strand is degraded. But the mechanisms in the first step are poorly defined. Post-translational modification(PTM) is one of the most critical regulation levels in life. In small RNA pathway, abnormality of modifications of RNA binding proteins can cause dysregulation of mi RNAs. A number of research has indicated that dysregulation of mi RNAs often leads to human diseases including cancer. Whether PTMs of key proteins in RNAi pathway play a role in tumorgenesis is still unclear.In the first part, we provided evidences that TARBP2 was SUMOylated at K52 in vitro and in vivo. External signals such as growth signal stimuli firstly activated phosphorylation of TARBP2 mediated by the MAPK/Erk signaling pathway, and subsequently upregulated TARBP2 SUMOylation, which stabilized TARBP2 itself by inhibition of the K48-linked ubiquitylation of TARBP2, thereby preventing the degradation through the ubiquitin proteasome pathway. We also found that SUMOylation level of TARBP2 could be regulated by external stimuli, such as H2O2 treatment and hypoxia.In the second part, we found although SUMOylation of TARBP2 appeared to have no significant influence on the overall production of mi RNAs, it played a role in the regulation of small RNA-induced gene silencing. Mechanically, SUMOylated TARBP2 with Dicer significantly recruited more Ago2 to build RLC(RISC-loading complex) through the direct interaction of the SUMO1 conjugated to TARBP2 with the SIMs of Ago2. Simultaneously, SUMOylation of TARBP2 promoted more precursor mi RNAs/si RNAs loading into the RLC. These led to Ago2, which is an active component of RISC, was more stabilized and mi RNAs/si RNAs bound by TARBP2-Dicer were efficiently transferred to Ago2. Thus, the guide strands were highly loaded into Ago2 to form the functional center of RISC, which remarkably improved the efficiency of RISC to silence specific m RNAs by degradation or translational inhibition. Meanwhile, we found SUMOylation of TARBP2 was required for its roles in suppression of tumor growth and tumor cell migration. Taken together, this study demonstrates that TARBP2 can be SUMOylated in vivo and in vitro at K52 for the first time. Our data suggests that SUMOylation of TARBP2 can increase its binding with Ago2 and pre-mi RNAs, which can regulate the efficiency of small RNA-induced gene silencing. We propose a detailed model about post-translational modification regulating efficiency of RNA interference, which can help us understanding small RNA-induced gene silencing. This study also provides a possible molecular etiology for tumorgenesis.