|M.Sc Student||Boaz Goldstein|
|Subject||Determining the Nature of RNAi and ADAR Interplay|
|Department||Department of Biology||Supervisor||Professor Lamm Ayelet|
|Full Thesis text|
This thesis attempts to unravel the interaction between RNA-editing and endogenous RNAi (RNA interference) in C. elegans. A-to-I RNA-editing is a process in which Adenines found in double stranded RNA (dsRNA) are converted to inosines. Inosine is read as guanosine by the transcription and translation machinery, resulting in effective A-to-G changes. RNAi is a process in which small RNAs (20-30nt) that are generated from transcripts regulate RNA expression. We focused on two populations: 26nt small interfering RNA (siRNA), which are cleaved from dsRNA by dicer (primary siRNAs), and 21-22nt siRNA which are polymerized by RNA-dependent-RNA-polymerase (RdRPs) (Secondary siRNAs). Both mechanisms, RNA editing and RNAi, are conserved and in contrast to mammalians, including human, they are not essential in C. elegans. Although RNA editing is an important process, not much is know about the mechanism and how it regulates dsRNA. There have been indications that RNA editing regulates siRNA in C. elegans, which lead us to the aim of this research to study the nature of this regulation by investigating siRNA that originates from editing sites.
To investigate small RNAs originating from editing sites, we developed a bioinformatics pipeline that identifies RNA editing sites in non-repetitive regions, and a visualization tool that assists analysis of sequences. We applied the pipeline to published and new mRNA sequences in order to produce a list of editing sites in C. elegans in non-repetitive regions. To reduce noise, we used the pipeline for phylogenetic profiling identify sequences from the same strain of C. elegans, since DNA matching mRNA sequences was not available. We discovered ~15,000 editing sites in non-repetitive regions, in various developmental stages. We focused on 77 genes, which have editing sites in their 3’UTR. Of these, 27 have editing site clusters. We found that mRNA expression levels do not change in ADAR mutants. Furthermore, editing sites are depleted of 26nt siRNA, but not 21-22nt siRNAs. 21-22nt siRNAs that originate from editing sites are polymerized over edited transcripts, and have matching T-to-C changes. We found no such changes in ADAR mutants, nor in random 3’UTR locations in wildtype, indicating that ADAR drives these changes. We observed increased expression of 21-22nt siRNAs in genes with 3’UTR RNA editing-clusters in ADAR mutants vs. wildtype, in embryo and L4 worms. We concluded that RNA editing inhibits generation of 26nt siRNA, but does not inhibit synthesis of 21-22nt siRNA, resulting in anti-sense nucleotide changes in 21-22 siRNA. Our results suggest that RNA editing occurs first and prevents dicer from generating 26nt primary siRNA, however RdRPs can still generate 22nt secondary siRNAs. The secondary siRNAs can be generated from RNA-edited transcripts. Therefore when RNA editing is knocked out it no longer interferes with siRNAs generation and siRNAs increase, maintaining any downregulation editing might have caused.
We also performed an analysis of Copy Number Variation data from cancer cells by omnibus univariate and multivariate analysis in order to test the ability of omnibus statistical tests to ascribe roles to genes in cancer driving pathways.