PgmNr C46: The Enhancer of zeste like protein Ezl1 is required for scnRNA selection and transcriptional repression of transposon-derived sequences in Paramecium tetraurelia.

Authors:
Andrea Frapporti 1 ; Olivier Arnaiz 2 ; Maoussi Lhuillier-Akakpo 1 ; Augustin de Vanssay 1 ; Caridad Miro Pina 1 ; Raphaël Margueron 3 ; Sandra Duharcourt 1


Institutes
1) Institut Jacques Monod, CNRS, UMR 7592, Sorbonne Paris Cité, Paris, France; 2) Institute of Integrative Biology of the Cell, UMR9198 CNRS CEA U Paris-Sud,91198 Gif-sur-Yvette, France; 3) Institut Curie, Paris Sciences et Lettres Research University, INSERM, U934, CNRS, UMR3215, Paris, France.


Abstract:

In the unicellular eukaryote Paramecium tetraurelia, differentiation of the somatic macronucleus (MAC) from the zygotic nucleus is characterized by massive and reproducible rearrangements of germline DNA, that include the elimination of transposable elements and 45,000 short, single-copy internal eliminated sequences (IESs). A specific class of small RNAs, the scnRNAs that are produced by the germline micronucleus (MIC) through a meiosis-specific RNAi pathway, is essential for DNA elimination of homologous sequences in the developing MAC. The mechanisms that allow specific recognition of germline-limited sequences within chromatin remain an open question. We identified recently a Paramecium homolog of mammalian Enhancer-of-zeste, Ezl1 that is required for the establishment of both H3K27me3 and H3K9me3 epigenetic modifications during development and for programmed genome rearrangements (Lhuillier-Akakpo, 2014). Here, we show that the Ezl1 protein is a bona-fide histone methyl-transferase, whose catalytic activity is necessary in vivo for histone H3 trimethylation on lysines 27 and 9 and for correct elimination of germline-limited DNA. We used RNA sequencing to measure steady state RNA levels from EZL1 knockdown (KD) cells at different time points during development. Analysis of small RNA datasets revealed that scnRNA biogenesis is not affected upon Ezl1 depletion as compared to control. Yet elimination of MAC-specific scnRNAs from the initial pool of scnRNAs, a process that normally leads to MIC-specific scnRNA enrichment during development, is abrogated in EZL1 KD. Ezl1 depletion also caused a sharp rise in transposon-derived transcripts. Altogether our data suggest that the histone methyltransferase Ezl1 is required for MIC-specific scnRNA selection and for transcriptional repression of transposon-derived sequences.