| Description |
Adenosine deaminases that act on RNA (ADARs) catalyze adenosine-to-inosine (A-to-I) conversion within double-stranded RNA (dsRNA). Since inosine prefers to base-pair with cytidine, it is read by cellular machinery like the ribosome as guanosine. Thus, A-to-I RNA editing can alter the translation of edited codons in cellular mRNAs. However, genome-wide A-to-I editing studies have demonstrated that editing in coding regions is exceedingly rare in most organisms. Instead, ADAR editing is abundant in noncoding sequences associated with protein-coding genes, particularly in introns and untranslated regions (UTRs). By extension, dsRNA structures must also be prevalent in such noncoding regions. These observations raise questions as to the purpose of RNA editing and dsRNA structure in gene-associated sequences. In this dissertation, I explore the physiological functions of ADAR enzymes and their noncoding dsRNA substrates in the nematode Caenorhadbitis elegans. First, I describe in Chapter 2 how ADARs prevent processing and silencing of cellular dsRNAs by the antiviral RNA interference (RNAi) machinery. Using RNAseq, I defined ADAR-edited dsRNAs, or editing-enriched regions (EERs), expressed during four stages of C. elegans development. I found that, in adr-1;adr-2 mutants, EERs gave rise to abundant ~23 nucleotide (nt) small interfering RNAs (siRNAs), and were involved in silencing their associated genes by an RNAi-dependent mechanism. Additionally, disruption of the 26G endogenous siRNA (endo-siRNA) pathway in adr-1;adr-2 mutant backgrounds caused a synthetic phenotype that was rescued by deleting factors involved in antiviral RNAi. These results suggest that ADARs limit RNAi activity against cellular dsRNAs, presenting a striking functional parallel to mammalian ADAR1, which prevents aberrant innate immune signaling by the antiviral dsRNA sensor MDA5. Though the work in Chapter 2 suggests that gene-associated dsRNAs can effect transcriptional silencing, in Chapter 3, I detail analyses suggesting that dsRNA-associated genes in fact exhibit higher-than-expected expression. I used three computational methods to define genome-wide loci in C. elegans encoding dsRNA structures, observing their enrichment on autosome distal arms. Despite that genes in distal arm regions are overall less highly expressed and less likely to be essential than genes in autosome centers, dsRNAs in these regions were enriched within essential and highly expressed genes. These analyses could not explicitly determine if gene expression patterns correlated with dsRNA formation or another property common to these loci. However, they suggest that dsRNA structures may function as important gene regulatory elements. In Chapter 4, I propose additional experiments to test contributions of dsRNA structure to gene expression regulation. |
