Translational inhibition and phase separation primes the epigenetic silencing of transposons

Abstract

19Transposons are mobile DNAs that can cause fatal mutations. To counteract these 20 genome invaders, the host genomes deploy small interfering (si) RNAs to initiate and 21 establish the epigenetic silencing. However, the regulatory mechanisms for the selective 22 recognition of transposons by the host genomes remain still elusive. Here we show that plant 23 transposon RNAs undergo frequent ribosome stalling caused by their inherently unfavourable 24 codon sequence usage. The ribosome stalling then causes the RNA truncation and the 25 localization to siRNA bodies, which are both critical prerequisites for the siRNA processing. 26 In addition, SGS3, the key protein in the siRNA biogenesis pathway, forms liquid droplets in 27 vitro through its prion-like domains implicating the role of liquid-liquid phase separation in 28 the formation of the siRNA bodies. Our study provides a novel insight into the regulatory 29 mechanisms for the recognition of invasive genetic elements which is essential for the 30 maintenance of genome integrity. 31 32 33 Keywords 34 transposon, codon optimality, ribosome stalling, easiRNA, RDR6-RdDM, liquid-liquid phase 35 separation, siRNA body, stress granule 36 37 38 45 of transposons 5 . While the canonical RdDM is mainly for reinforcing the silent state of 46 transposons, the recently proposed alternative RdDM pathway suppresses the active 47 transposons that are induced upon the epigenetic mutations and in several development 48 stages 3,6-10 . RNA-DEPENDENT RNA POLYMERASE 6 (RDR6)-SUPPRESSOR OF GENE 49 SILENCING 3 (SGS3) complex serves as a detector of transposon RNAs that selectively 50 processes them into 21 or 22-nt siRNAs (also referred to as epigenetically activated siRNAs, 51 easiRNAs) 11,12 . Transposon-derived siRNAs trigger both post-transcriptional repression by 52 cleaving the transposon RNAs (and then subsequently initiating the secondary siRNA 53 biogenesis) and establishment of the epigenetic silencing by recruiting DNA 54 methyltransferases to the target TE chromatin 5,7,8 . The biogenesis of 21 or 22-nt siRNAs in 55 plants is initiated by RDR6 which is an RNA-dependent RNA polymerase that forms double-56stranded RNAs by templating directly the target RNAs 13 . SGS3 is an RNA-binding protein 57 that interacts with RDR6 and is essential for its function 13 . The duplex RNAs are then sliced 58 to 21 or 22-nt siRNAs by DICER-LIKE 3 or 4 (DCL3/4) 8,14,15 . Since RDR6 templates are 59 directly derived from the RNAs of those mobile elements, RDR6 and the resulting easiRNAs 60 are deemed the first line of the host immune system against the invasive genetic elements. 61The RDR6-mediated easiRNA production pathway (RDR6-RdDM) is usually 62 prevented in the transcripts derived from genes by the RNA decay pathways [16][17][18] . This raises 63 4 an important question of how RDR6 specifically recognizes its TE targets and establish their 64 epigenetic silencing. Previous reports showed that the initial cleavage of target transcripts is a 65 critical prerequisite for RDR6 recog...

Nuclear and cytoplasmic RNA exosomes and PELOTA1 prevent miRNA-induced secondary siRNA production in Arabidopsis

“…We also show that the expression level of 21-nt miRNAs, not the accumulation of their RISC-mediated cleavage fragments, correlates with their ability to trigger secondary siRNAs in rrp4 mutants. We discuss these results in the light of our observation in the present work that mutation of PEL1 leads to miRNA-induced siRNA amplification similar, but not identical, to that observed in ski2, ski3 and rrp45b , and of recent evidence for the importance of ribosome stalling for siRNA production (53).…”

“…Ribosome profiling experiments demonstrate ribosome association with TAS precursors (73,74), and loss of the exportin-like protein SDE5 required for all tasiRNA production causes loss of TAS precursor transcripts from ribosome-associated fractions (75). Furthermore, detailed biochemical analyses reveal the importance of ribosome stalling in proximity to the 5’ miR390 site in TAS3 precursors (74), and ribosome stalling and collision at rare codons in transposable element mRNAs correlates with their production of 21-nt siRNAs (53) (observed in mutants defective in DNA methylation, and, therefore, often referred to as epigenetically activated siRNAs, or easiRNAs (25)). It appears, therefore, that stalled ribosomes, in particular in combination with RISC, act as a trigger of secondary siRNA production, perhaps by extending the time of RISC association with target mRNA by AGO interaction.…”