Genomics and regulatory functions of microRNAs and small silencing RNAs in Arabidopsis thaliana

Abstract

Small RNA-mediated gene silencing is a mechanism widely employed by eukaryotes to repress many loci including some involved in critical developmental transitions. In plants, endogenous small RNAs consist of two broad classes, the ~21-nucleotide (nt) microRNAs (miRNAs) and the diverse ~22-24-nt trans-acting and heterochromatic short interfering RNAs (siRNAs). In order to more extensively characterize the small RNA landscape in plants and to identify undiscovered miRNAs and siRNAs, we performed high-throughput sequencing of small RNAs. We generated a large dataset consisting of >340,000 unique sequences expressed in several representative plant organs and developmental stages including wild-type seedlings, flowers, leaves and siliques. Application of enhanced miRNA annotation criteria gleaned from the data revealed the existence of at least 38 apparently recently-evolved miRNAs that were much less abundant in plant tissues than the 26 conserved miRNA families, and had a greater diversity of predicted target genes. We characterized several of these miRNAs more closely. Our results supported a homeostatic auto-regulatory loop for DCL1 via the intron-embedded miR838, and elaborated on the prevailing model of DCLl-mediated miRNA biogenesis with the finding that at least two miRNAs (miR839 and miR822) are processed exclusively by DCL4. Several microRNA target sites were experimentally validated, including the miR823-directed cleavage of the DNA cytosine methylation factor CHROMOMETHYLASE3. We also identified a trans-acting siRNA-generating locus which we called TAS4, and confirmed that miR828 triggers phased siRNA production by specifying targeted cleavage of TAS4 transcripts.

(cont.) The evolving miRNAs described in this work may have lineage-specific roles, and their discovery potentiates future functional investigation of recently-emerged miRNAs and their evolution in Arabidopsis. Finally, we discovered thousands of endogenous candidate heterochromatic siRNAs of unknown function, the majority of which mapped to unannotated regions of the genome especially prone to generating siRNAs ("hotspots") or to repetitive or transposable elements. Our small RNA study suggests that a significant proportion of the genome is primed for the emergence of new miRNA families or for siRNA production, and expands the roles of these small RNAs in shaping regulatory circuits and transcriptome output.