Screen Novel Splicing Repressors Regulating SMN Exon7Splicing And Study The Associated Regulatory Mechanism

Spinal muscular atrophy (SMA) is an autosomal recessive neuromusculardisorder with high incidence, and it is the primary genetic cause of infant mortality.The genetic cause of SMA is a homozygous loss of SMN1. But its homologous geneSMN2with a C-to-T mutation on exon7generates truncated and unstable SMNprotein because of the low splicing efficiency of exon7. Increasing SMN2exon7splicing is considered to be a feasible and effective strategy to therapy SMA. Thus,investigating the mechanism of SMN2exon7splicing regulation is of importance.Using esiRNA silencing approach to identify SMN2splicing repressor amongheterogeneous nuclear ribonucleoproteins, this study has found the RNA-bindingprotein hnRNP U as a new RNA splicing factor that represses SMN2exon7splicing,together with the known SMN2repressor hnRNP A1and A2. However, unlike hnRNPA1and A2, hnRNP U repression of SMN2splicing is not associated with the C-to-Tmutation on exon7, neither with any of the four known ISS elements in theneighboring introns. We further show that the repression is not associated with analtered transcription rate of the SMN2gene either. These results indicate that hnRNPU may repress splicing of the alternative exon7of SMN2pre-mRNA through amechanism different from all the known ones.hnRNP U has both RNA and protein interaction domains, but has never beenreported to associated with alternative splicing before. We hypothesize that hnRNP Umay regulate SMN2splicing through an interaction with some other known splicing factor. This promoted us to screen the other splicing factors that repress SMN2splicing through about340putative human RNA binding protein genes including allthe known splicing factors. We obtained8other splicing repressors for SMN2splicing,strikingly,5of them are constitutive splicing factors for the3’ splice site selection,including SF1, U2AF65, PUF60, U2AF35and CHERP (a component of snRNP). Weshowed that hnRNP U interacts with U2AF65, U2AF35and SF1through anRNA-dependent manner, among which U2AF65is the strongest one. These resultsstrongly indicate that hnRNP U may interact with these constitutive splicing factors toaffect the reorganization and selection of the constitutive3’ splice site, and thereforeinhibits the selection of the alternative3’ splice site. Therefore, when this class ofproteins is down-regulated, the constitutive splice site becomes less competitive ininteracting with their common5’ splice site. As a result, the alternative splice sitebecomes relatively more competitive.In order to identify the RNA components through which hnRNP U interacts withU2AF65and other3’ss splicing factors, this study used CLIP-seq method to obtainthe genome-wide hnRNP U binding sites. It is shown that hnRNP U has no bindingsite in exon7and the adjacent intronic regions, consistent with our data showing thatthe recombinant hnRNP U does not bind these regions in vitro. However, we foundthat hnRNP U strongly binds to SmBP-box of U2snRNA, suggesting that hnRNP Uis a constitutive3’ splice site splicing factor. We propose that hnRNP U may berecruited to the3’ splice site through binding to U2snRNA, where it interacts withU2AF65and promotes the3’ splice site selection.CLIP-seq data show that hnRNP U binds to the GU-rich sequence motifs, and thebinding is strongly associated with some alternative splicing events, which indicatesthat hnRNP U may be a general splicing regulator. The transcriptome data have beenobtained for the hnRNP U-expressed and-silenced Hela cells, and we anticipate theglobal view of the involvement of this protein in regulating of the alternative splicingin Hela cells will be obtained after the data analysis is finished.