Many plant cells can be reprogrammed into a pluripotent state that allows ectopic organ development. Inducing totipotent states to stimulate somatic embryo (SE) development is, however, challenging due to insufficient understanding of molecular barriers that prevent somatic cell dedifferentiation. Here we show that Polycomb repressive complex 2 (PRC2)-activity imposes a barrier to hormone-mediated transcriptional reprogramming towards somatic embryogenesis in vegetative tissue of Arabidopsis thaliana. We identify factors that enable SE development in PRC2-depleted shoot and root tissue and demonstrate that the establishment of embryogenic potential is marked by ectopic co-activation of crucial developmental regulators that specify shoot, root and embryo identity. Using inducible activation of PRC2 in PRC2-depleted cells, we demonstrate that transient reduction of PRC2 activity is sufficient for SE formation. We suggest that modulation of PRC2 activity in plant vegetative tissue combined with targeted activation of developmental pathways will open possibilities for novel approaches to cell reprogramming.

Background Stable gene repression is essential for normal growth and development. Polycomb repressive complexes 1 and 2 (PRC1&2) are involved in this process by establishing monoubiquitination of histone 2A (H2Aub1) and subsequent trimethylation of lysine 27 of histone 3 (H3K27me3). Previous work proposed that H2Aub1 removal by the ubiquitin-specific proteases 12 and 13 (UBP12 and UBP13) is part of the repressive PRC1&2 system, but its functional role remains elusive. Results We show that UBP12 and UBP13 work together with PRC1, PRC2, and EMF1 to repress genes involved in stimulus response. We find that PRC1-mediated H2Aub1 is associated with gene responsiveness, and its repressive function requires PRC2 recruitment. We further show that the requirement of PRC1 for PRC2 recruitment depends on the initial expression status of genes. Lastly, we demonstrate that removal of H2Aub1 by UBP12/13 prevents loss of H3K27me3, consistent with our finding that the H3K27me3 demethylase REF6 is positively associated with H2Aub1. Conclusions Our data allow us to propose a model in which deposition of H2Aub1 permits genes to switch between repression and activation by H3K27me3 deposition and removal. Removal of H2Aub1 by UBP12/13 is required to achieve stable PRC2-mediated repression.

Several pathways control time to flowering in Arabidopsis thaliana through transcriptional and posttranscriptional gene regulation. In recent years, mRNA processing has gained interest as a critical regulator of flowering time control in plants. However, the molecular mechanisms linking RNA splicing to flowering time are not well understood. In a screen for Arabidopsis early flowering mutants we identified an allele of BRR2a. BRR2 proteins are components of the spliceosome and highly conserved in eukaryotes. Arabidopsis BRR2a is ubiquitously expressed in all analyzed tissues and involved in the processing of flowering time gene transcripts, most notably FLC. A missense mutation of threonine 895 in BRR2a caused defects in FLC splicing and greatly reduced FLC transcript levels. Reduced FLC expression increased transcription of FT and SOC1 leading to early flowering in both short and long days. Genome-wide experiments established that only a small set of introns was not correctly spliced in the brr2a mutant. Compared to control introns, retained introns were often shorter and GC-poor, had low H3K4me1 and CG methylation levels, and were often derived from genes with a high-H3K27me3-low-H3K36me3 signature. We propose that BRR2a is specifically needed for efficient splicing of a subset of introns characterized by a combination of factors including intron size, sequence and chromatin, and that FLC is most sensitive to splicing defects.

SUMMARYThe study and eventual manipulation of leaf development in plants requires a thorough understanding of the genetic basis of leaf organogenesis. Forward genetic screens have identified hundreds of Arabidopsis mutants with altered leaf development, but the genome has not yet been saturated. To identify genes required for leaf development we are screening the Arabidopsis Salk Unimutant collection. We have identified 608 lines that exhibit a leaf phenotype with full penetrance and almost constant expressivity and 98 additional lines with segregating mutant phenotypes. To allow indexing and integration with other mutants, the mutant phenotypes were described using a custom leaf phenotype ontology. We found that the indexed mutation is present in the annotated locus for 78% of the 553 mutants genotyped, and that in half of these the annotated T-DNA is responsible for the phenotype. To quickly map non-annotated T-DNA insertions, we developed a reliable, cost-effective and easy method based on whole-genome sequencing. To enable comprehensive access to our data, we implemented a public web application named PhenoLeaf (http://genetics.umh.es/phenoleaf) that allows researchers to query the results of our screen, including text and visual phenotype information. We demonstrated how this new resource can facilitate gene function discovery by identifying and characterizing At1g77600, which we found to be required for proximal-distal cell cycle-driven leaf growth, and At3g62870, which encodes a ribosomal protein needed for cell proliferation and chloroplast function. This collection provides a valuable tool for the study of leaf development, characterization of biomass feedstocks and examination of other traits in this fundamental photosynthetic organ.