The Molecular Mechanism of SNORD116 Action and Possible SNORD116 Substitution Strategies

The region on chromosome 15q11.13 that is not expressed in subjects with PWS contains a cluster of 27 SNORD116 copies. Microdeletions encompassing only SNORD116 result in a Prader-Willi like phenotype, suggesting a central role for SNORD116 in PWS. SNORDs (C/D box snoRNAs) are small non-coding RNAs that have C and D boxes as characteristic sequence elements. The C and D box form a characteristic RNA structure, the ‘RNA kink’ that is stabilized by the 15.5 (NHP2L1) protein, which triggers the assembly of a snoRNPs that contains two structural proteins (NOP58 and NOP58) and fibrillarin, an enzyme that performs 2’-O-methylation of target RNAs. These target RNAs bind to SNORD-sequences between the C and D boxes, termed antisense boxes. However, so far, no RNA-targets for SNORD116 have been determined and bioinformatic analysis could not predict RNAs interacting with SNORD116 that could be validated. We previously found that a fraction of SNORD116 molecules is processed into a shorter RNAs, due to cleavage near the C’ box. Importantly, this C’ box has the sequence RTGAGTGA in all SNORD116 copies which deviates from the human consensus (RTGATGA). The additional G will likely inhibit the formation of an RNA kink-structure with the D’ box and prevent subsequent stabilization through the 15.5kD protein. Using a mixed oligonucleotide, we could knock down all SNORD116 copies and using RNAseq, we determined target genes. We found that SNORD116 upregulates microexons and regulates the stability of numerous pre-mRNAs. We postulate that SNORD116 binds to target RNAs as a non-canonical snoRNAs lacking fibrillarin and changes splice site selection and stability. This will be tested in three Specific Aims: 1. Determine the targets of individual copies of SNORD116 We will knock down individual SNORDs in SH-SY5Y cells to test whether individual SNORDs have distinct targets. Using the same approach, we will knock down SNORD116 in mouse neuoblastoma cell lines to determine genes that are co-regulated in mouse and human, which will allow to identify direct targets containing compensatory mutations. These targets will be tested in the SNORD116 knockout mouse (4). 2. Validate RNA:RNA interaction using mutational analysis Using known HuR and PTB1 binding sites in c-fos, YAP1, SERPINE1, FAT1 and TGFB1, we will test the binding of SNORD116 to the predicted motifs using compensatory mutations in factor competition in transfection assays. 3. Test the effect of an oligo GRAAAGCUGAACAAAAUGAGGUGA on SNORD116 targets. Using transfection assays, we will test whether the oligo corresponding to the SNORD116 consensus site has a similar influence on gene expression as SNORD116 and could be a therapeutic approach. We will first use HeLa cells that do not express SNORD116 and later neurons derived from iPS cells.