piRNAs of Caenorhabditis elegans broadly silence nonself sequences through functionally random targeting

Abstract

ABSTRACTSmall non-coding RNAs such as piRNAs serve as guides for Argonaute proteins, enabling sequence-specific, post-transcriptional regulation of gene expression. The piRNAs of Caenorhabditis elegans have been observed to bind targets with high mismatch tolerance, and appear to lack specific transposon targets, unlike piRNAs in Drosophila melanogaster and other organisms. These observations support a model in which C. elegans piRNAs provide a broad, indiscriminate net of silencing, acting in competition with siRNAs associated with the CSR-1 Argonaute that specifically protect self genes from silencing. However, the breadth of piRNA targeting has not been subject to in-depth quantitative analysis, nor has it been explained how piRNAs are distributed across sequence space to achieve complete coverage. Through a bioinformatic analysis of piRNA sequences, incorporating an original metric of piRNA-target distance, we demonstrate that C. elegans piRNAs are functionally random, in that their coverage of sequence space is comparable to that of random sequences. By possessing a sufficient number of distinct, essentially random piRNAs, C. elegans is able to target arbitrary nonself sequences with high probability. This result elucidates the mechanism by which newly transcribed mRNAs in C. elegans are classified as self or nonself, and has implications for piRNA evolution and biogenesis.