Function of RNaseIII-like proteins in plants. Implication in the epigenome. – Epi-RNase

The biochemical study of RNaseIII-like proteins will be performed on soluble recombinant proteins expressed in and purified from bacteria. The obtention of functional recombinant proteins that clive RNA in vitro is a key point of the biochemical part of this project. Swap experiments, which will allow the structure-function relationship analysis, require to add restriction sites between the domains of these proteins and the verification that these new sites do not perturbate protein activities. The genetic study of RNaseIII-like proteins involves the identification of defective mutants as well as trasngenic plants overexpressing each of these proteins.The verification that mutants actually lack expression is thought considering the difficulty to detect RTL1 and RTL3.The verification that transgenic plants actually overexpress these proteins involves adding tags and checking that these tags do not affect protein activities. the inventory of miRNA and siRNA will be performed by hightrought-put sequencing and bioinformatic analysis. Epigenetic modifications will be scored specifically at genomic sites corresponding to small RNA changes using methylation-sensitive enzymes, or globally on the entire genome after bisulfite treatment. Histone modifications will be analyzed after chromatin purifications, followed by specific or global analyses.

Hightrought-put sequencing of plants overexpressing RTL1 followed by bioinformatic analysis revealed a specific effect of RTL1 on siRNA at the whole-genome level. RTL1 overexpression inhibits all forms of post-transcriptional (trans)gene silencing (S-PTGS and IR-PTGS). The mutagenesis of the RNaseIII domain and the deletion of the double-stranded RNA binding domain each compromise RTL1 activity. The expression of the RTL1 gene is induced in response to virus infection, but most viruses inhibit RTL1 activity. Some loci that specifically produce siRNAs in plants overexpressing RTL1 have been identified. The production of these siRNAs requires DCL2, DCL3 or DCL4, indicating that RTL1 acts upstream the DCLs. An His-RTL1 protein was purified from bacteria. In vitro RTL1 clives the RNA targets identified in vivo. This activity requires the RNaseIII domain. Several clivage sites have been mapped by primer extension in vivo and in vitro. RTL1 likely cuts blunt at a 4-5 nucleotide consensus sequence within stem-loop structures. RTL1 is part of a large protein complex (700 kDa). Several potential partners have been identified by mass spectrometry followed by sequencing.
Hightrought-put sequencing of plants overexpressing RTL2 followed by bioinformatic analysis revealed a limited effect of RTL2 at a few loci. RTL2-dependant siRNAs are generally 24-nt long and correlate with DNA methylation. In vitro, RTL1 and RTL2 cut different RNA targets. Domain swaps between RTL1 and RTL2 have been generated to analyse the specificity of these proteins towards their targets.

The study of the specificity of RTL1 and RTL2 and of their respective substrates will be continued. In planta analyses will be started to localize the proteins within the cell. The rôle of RTL1 in plant antiviral défense wil be studied, as well as the mechanisms by which viruses inhibit RTL1 activity.

Nahid Shamandi1, Matthias Zytnicki2, Cyril Charbonnel3,, Gersende Lepère1, Allison C. Mallory1, Julio Sáez-Vásquez3, and Hervé Vaucheret1 Plants encode a general siRNA suppressor that is induced and suppressed by viruses (submitted)

RNA silencing is a conserved eukaryotic gene regulatory mechanism, integral for taming endogenous (repetitive elements and transposons) or exogenous (viruses and bacteria) invasive nucleic acids to minimize their impact on genome integrity and function. RNA silencing also is essential for controlling the expression of protein coding genes throughout development or in response to environmental stimuli. RNA silencing is guided by small RNAs, which mostly are produced by a class of RNaseIII enzymes called Dicer. Recently, a growing number of Dicer-independent small RNAs have been found, resulting from the action of RNA-dependent RNA polymerases, RNaseH-like Argonaute proteins or exoribonucleases. Small RNAs also are subjected to regulatory mechanisms. For example, long RNAs that sequestrate small RNAs, or nucleases that specifically degrade small RNAs have recently been identified. However, the list of enzymes that regulate the fate of small RNAs or small RNA precursors is far from being complete.
In this project, we will examine the role of uncharacterized RNaseIII enzymes in the regulation of small RNA in plants. These enzymes exhibit RNaseIII and double-stranded RNA binding (DRB) domains but lack RNA helicase and PAZ domains that Dicer enzymes have. Therefore, they resemble bacterial or fungal RNaseIII enzymes and thus are referred to as RTL (for Rnase Three Like). Genetic, molecular and bioinformatic results obtained by Partner 1 and Partner 2 indicate that Arabidopsis RTL1 and RTL2 enzymes limit and stimulate small RNA accumulation, respectively, while biochemical results obtained by Partner 3 indicate that RTL2 exhibits RNA cleavage activity. Partner 1 also has observed that modulating the expression of RTL enzymes results in heritable epigenetic modifications. Specifically, our data suggest that RTL1 is an endogenous suppressor of 21-nt and 24-nt siRNA functions. Exogenous and endogenous suppressors of 21-nt siRNA function have already been reported; however, suppressors of 24-nt siRNA function have never been reported before. Because 24-nt siRNA travel at distance to establish epigenetic marks in cells where they are not produced, and because RTL1 is mostly expressed in roots, which are important sensors in plants, it is of particular interest to decipher the role of RTL1 in suppressing 24-nt function. Our unpublished data also suggest that RTL2 over-expression increase the production of siRNA at specific loci, and that RTL2 over-expression have the potential to generate novel epigenetic marks. Because RTL2-induced epigenetic marks are inherited after elimination of the RTL2 trigger, specific induction of RTL2 under particular conditions may play an important role in the adaptive response of plants to stress. Overall, our data suggest that RTL1 and RTL2 suppress and enhance epigenetic marks, respectively, thus contributing to the epigenome.
This project aims at understanding the function and biological role of RTL enzymes by a combination of genetics, bioinformatics and biochemistry approaches. We will characterize the biochemical function of RTL1 and RTL2 in vitro and in vivo to define the specificity of their RNaseIII and DRB domains, and identify their RNA substrates and protein partners. We will also attempt to characterize RTL3, a gene that resembles a fusion between RTL1 and RTL2 but which does not appear to be expressed under laboratory conditions, to determine if it could also contribute to the plant epigenome. We will explore the physiological consequences of RTL deregulation by analyzing the transcriptome and methylome of plants that have inherited phenotypic modifications after elimination of the RTL trigger, and we will map the most stable heritable traits and eventually clone the corresponding loci.