Widespread roles for piRNAs and WAGO-class siRNAs in shaping the germline transcriptome of Caenorhabditis elegans

Abstract

Piwi-interacting RNAs (piRNAs) and small interfering RNAs (siRNAs) are distinct classes of small RNAs required for proper germline development. To identify the roles of piRNAs and siRNAs in regulating gene expression in Caenorhabditis elegans, we subjected small RNAs and mRNAs from the gonads of piRNA and siRNA defective mutants to high-throughput sequencing. We show that piRNAs and an abundant class of siRNAs known as WAGO-class 22G-RNAs are required for proper expression of spermatogenic and oogenic genes. WAGO-class 22G-RNAs are also broadly required for transposon silencing, whereas piRNAs are largely dispensable. piRNAs, however, have a critical role in controlling histone gene expression. In the absence of piRNAs, histone mRNAs are misrouted into the nuclear RNAi pathway involving the Argonaute HRDE-1, concurrent with a reduction in the expression of many histone mRNAs. We also show that high-level gene expression in the germline is correlated with high level 22G-RNA production. However, most highly expressed genes produce 22G-RNAs through a distinct pathway that presumably involves the Argonaute CSR-1. In contrast, genes targeted by the WAGO branch of the 22G-RNA pathway are typically poorly expressed and respond unpredictably to loss of 22G-RNAs. Our results point to broad roles for piRNAs and siRNAs in controlling gene expression in the C. elegans germline.

piRNAs prevent runaway amplification of siRNAs from ribosomal RNAs and histone mRNAs

“…While loss of piRNAs leads to a reduction in WAGO-class 22G-RNA levels from many mRNAs in the germline, core histones instead produce elevated levels (7,29). In this study, we find that ribosomal RNAs also hyperaccumulate 22G-RNAs in prg-1 mutants.…”

“…Without PRG-1 and polyA-binding proteins to help protect rRNAs and histones, 22G-RNAs may engage in a feedback loop that leads to runaway amplification by RNA-dependent RNA polymerases. This model may explain why hundreds of CSR-1 targets are also depleted of 22G-RNAs in prg-1 mutants (29). Perhaps a similar molecular sieve model could help to explain the highly abundant pachytene piRNAs found in mice, which also have expansive targeting capacity but lack a clear role in regulating gene expression (40).…”

“…Presumably the trapped RNAs are acted on by genelicensing factors, such as CSR-1, or gene-silencing factors, such as the WAGO Argonautes, or both. This would help to explain why essentially all germline mRNAs are targeted by piRNAs, while only a subset are silenced (23,29,39). The sheer abundance of rRNAs and histone mRNAs may overwhelm the cellular machinery that processes them.…”

“…We and others recently showed that histone transcripts are directed into an endogenous RNAi pathway in piwi/prg-1 mutants where they spawn high levels of 22G-RNAs, which is in some instances coincident with reduced histone mRNA levels (7,29). To determine if this phenomenon is unique to histones, we examined 22G-RNA production from all annotated protein coding genes using RNA-seq datasets from dissected distal gonads from wild type and prg-1 and mut-16 mutants (29). The majority of protein coding genes yielded 22G-RNA reads in wild type animals, and to a lesser extent in prg-1 and mut-16 mutants ( Fig.…”

Recovery from transgenerational RNA silencing is driven by gene-specific homeostasis

“…6e) and complete loss of EGO-1 results in sterility 21,22 . Furthermore, small RNAs made by these RdRPs do not always correlate with gene silencing 23 . Nevertheless, robust recovery of expression even after hundreds of generations of silencing suggests that silencing is actively established in every generation.…”