The biogenesis, regulation and functions of transitive siRNA in plants

Abstract

Small RNA (sRNA)-mediated RNA interference (RNAi) is a sequence-specific gene silencing mechanism that modulates gene expression in eukaryotes. As core molecules of RNAi, various sRNAs are encoded in the plant genome or derived from invading RNA molecules, and their biogenesis depends on distinct genetic pathways. Transitive small interfering RNAs (siRNAs), which are sRNAs produced from double-strand RNA (dsRNA) in a process that depends on RNA-dependent RNA polymerases (RDRs), can amplify and spread silencing signals to additional transcripts, thereby enabling a phenomenon termed "transitive RNAi". Members of this class of siRNAs function in various biological processes ranging from development to stress adaptation. In Arabidopsis thaliana, two RDRs participate in the generation of transitive siRNAs, acting cooperatively with various siRNA generation-related factors, such as the RNA-induced silencing complex (RISC) and aberrant RNAs. Transitive siRNAs are produced in diverse subcellular locations and structures under the control of various mechanisms, highlighting the intricacies of their biogenesis and functions. In this review, we discuss recent advances in understanding the molecular events of transitive siRNA biogenesis and its regulation, with a particular focus on factors involved in RDR recruitment. We aim to provide a comprehensive description of the generalized mechanism governing the biogenesis of transitive siRNAs. Additionally, we present an overview of the diverse biological functions of these siRNAs and raise some pressing questions in this area for further investigation.

Keywords: RDRs; siRNA; silencing; transitivity.

Figure 1

Phylogenetic distribution of RDR proteins in plants and nonplant eukaryotes (A) Identification of the putative RDRs in the proteome sequences from animals, fungi, plants, and algae. The solid and open circles indicate the presence and absence, respectively, of RDR in the corresponding species. (B) Phylogenetic tree of the six RDR proteins in Arabidopsis thaliana. (C) Phylogenetic tree of the RDR1, RDR2, and RDR6 subclades from representative plants. The solid dots indicate the six Arabidopsis RDR proteins. The homologous RDR proteins of the 10 species in (A) were retrieved by two-round BLASTP with DIAMOND using default parameters: (1) the protein sequences of RDR1, RDR2, and RDR6 of Arabidopsis were used as queries; and (2) sequences of all the homologous proteins identified in step 1 were used as queries. MAFFT, Fasttree, and MEGA were used for pairwise alignment, phylogenetic tree construction, and visualization, respectively. The 0.5 size bar represents a phylogenetic distance of 0.5 nucleotide substitutions per site measured on the graph.

Figure 2

The biogenesis pathways of RDR1- and RDR6-dependent transitive siRNAs (A) RDR1 targets viral RNA or a set of endogenous RNAs upon viral infection and then synthesizes dsRNAs, which are preferentially (larger blue oval) processed by DCL4 into 21-nt siRNAs. Only a small fraction of dsRNAs is processed by DCL2 to generate 22-nt siRNAs. These 21- and 22-nt siRNAs are typically loaded onto AGO1/AGO2 and less often onto AGO5, 7, and 10 to form a RISC. This assembly directs silencing of a broad spectrum of sequences homologous to the transitive siRNA through cleavage or translational repression, amplifying the antiviral signal. (B) PHAS , TAS, and TEs need primary miRNA to induce production of transitive siRNA. The miRNAs are loaded into AGO1/AGO7 to form a RISC, which mediates target RNA cleavage, followed by RDR6 recruitment. The dsRNAs synthesized by RDR6 are primarily cleaved into 21-nt siRNA by DCL4, with DCL2 playing a lesser role. The majority of RDR6-dependent transitive siRNAs are loaded into AGO1 to mediate silencing of cognate genes by transcript cleavage or translational repression, amplifying and spreading the silencing signal to additional transcripts. However, some siRNAs derived from TAS or TEs are associated with AGO4/AGO6 or other unidentified AGOs to direct DNA methylation. Particularly, during the reproduction stage, two types of PHAS genes are targeted by RDR6, and the resulting dsRNAs are processed into 21- and 24-nt reproductive phasiRNAs by DCL4 and DCL3/DCL5, respectively. The 21-nt reproductive phasiRNAs are loaded into AGO1 or AGO5 and could direct the cleavage of their cognate genes. The 24-nt reproductive phasiRNAs influence DNA methylation in cis, whereas their primary function and which AGOs they are loaded onto remain unclear. (C) Viral RNAs, transgene and endogenous coding RNAs, and NAT transcripts produce transitive siRNA without a defined sRNA trigger. Under certain conditions, such as being aberrant, these transcripts are converted into dsRNA, subsequently generating 21- or 22-nt siRNA by the action of DCL4 or DCL2, respectively. Because DCL4 outcompetes DCL2 for dsRNA substrate targeting, the siRNAs are predominantly 21 nt in length. These siRNAs are loaded into AGO1 to form a RISC, directing silencing of cognate RNAs through cleavage or translational repression. The ssRNA in blue indicates the original target RNA, gene1; RNA in blue and purple is the transcript of related target gene 2, and RNA in blue and orange is the transcript of related target gene 3. The complementary RNA produced by RDR1/RDR6 is in red.

Figure 3

The proposed mechanisms of RDR1/RDR6 recruitment (A) dsRNA produced by RDR1 and RDR6 may accumulate in a phospholipid-rich spherule upon viral infection. ALA1/ALA2 play an important role in the formation of spherules, into which RDR1/RDR6 are incorporated via an unknown manner. (B) SGS3 recognizes and binds to the 22-nt sRNA-AGO1 complex as well as to the miR390-AGO7 complex on the target RNA, and these associations promote the LLPS of SGS3. Subsequently, RDR6 is recruited into the LLPS condensates through the interaction between RDR6 and SGS3, forming an siRNA body where RDR6 synthesizes dsRNA. (C) Ribosome stalling on a transcript marks it as aberrant. If the decay process is deficient or the aberrant RNA outnumbers the decay capacity, stalled ribosomes on an aberrant RNA may facilitate the binding of SGS3 via an unclear mechanism and may subsequently recruit RDR6. In a few cases, aberrant RNA may still be detected in the absence of a RISC and stalled ribosomes and may recruit SGS3/RDR6 by an atypical mechanism.