Quantifying Downstream Regulatory Output as a Way to Understand the Biogenesis Pathways of Endogenous siRNAs in C. elegans

In the germline of the model nematode Caenorhabditis elegans ( C. elegans ), a network of abundant endogenous small RNAs act to regulate gene expression during gamete formation. This network is complex, and includes 4 biochemically distinct classes of small RNAs that act as cofactors for at least 12 Argonaute proteins. The biogenesis pathways that generate this network of regulatory molecules remain poorly understood. One particularly abundant class of small RNA, the 22G RNAs, are synthesized de novo by RNA‐dependent RNA polymerases (RdRPs) and target more than a thousand transcripts in the C. elegans germline. 22G RNAs bind to either the Argonaute CSR‐1, which licenses gene expression, or to a member of the WAGO clade of Argonautes, which silence gene expression through either transcriptional or post‐transcriptional mechanisms. Biogenesis of at least 65% of WAGO‐bound 22G RNAs depends on an upstream “trigger” event initiated by a small RNA of the piRNA class. These 22G RNAs, derived as secondary RNA species from the piRNA system, silence their target genes primarily at the transcriptional level. Strikingly, production of these 22GRNAs, and the gene silencing they carry out, has the ability to be passed to progeny even in the absence of the initial piRNA trigger. We are interested in asking whether additional, piRNA‐independent pathways for WAGO‐class 22G RNAs exist in the worm. There is evidence that some 22G RNAs may not be dependent on piRNAs for their biogenesis, as their levels are unaffected by the loss of the piRNA‐interacting Argonaute prg‐1 . However, because transgenerational inheritance of 22G RNAs after loss of prg‐1 could explain this observation, it is difficult to use traditional genetic approaches to definitively ask whether there exists a piRNA‐independent biogenesis pathway for 22G RNAs. To evaluate this possibility, we are using RNA‐seq on an Illumina platform to compare levels of pre‐mRNA and mRNA in genetic backgrounds that have or lack 22G RNAs. This comparison will allow for the determination of whether a particular transcript is regulated by a post‐transcriptional or transcriptional mechanism. Because piRNA‐triggered 22G RNAs regulate their targets largely through a transcriptional mechanism, we predict that transcripts regulated exclusively at the post‐transcriptional level represent a set of piRNA‐independent 22G RNA targets, and that this set will overlap with the several hundred transcripts whose 22Gs remain expressed in the prg‐1 mutant background. This project requires a novel method to sequence low abundance pre‐mRNA and mRNA at low cost. In order to enrich both pre‐mRNAs and mRNAs while depleting rRNA and tRNA, we are using enzymatic degradation of polymerase I and III transcripts to enrich for capped transcripts. Based on RT‐qPCR of rRNA and selected mRNAs, this method appears to deplete rRNA relative to mRNA by 10–20 fold. We are currently performing RNA‐seq experiments to further validate this method. This method of enrichment also has the benefit of being directly translatable to all eukaryotic systems as it does not require the species‐specific primers that subtractive hybridization approaches do. In addition, we expect that this method will decrease the 3′ bias typically found with poly‐A selection. Ultimately, this project will contribute to our understanding of the mechanisms that generate a network of biochemically similar but functionally distinct small noncoding RNAs with critical roles in the development of germ cells.