Proliferation of legume nodule primordia is controlled by shoot-root signaling known as autoregulation of nodulation (AON). Mutants defective in AON show supernodulation and increased numbers of lateral roots. Here, we demonstrate that AON in soybean is controlled by the receptor-like protein kinase GmNARK ( Glycine max nodule autoregulation receptor kinase), similar to Arabidopsis CLAVATA1 (CLV1). Whereas CLV1 functions in a protein complex controlling stem cell proliferation by short-distance signaling in shoot apices, GmNARK expression in the leaf has a major role in long-distance communication with nodule and lateral root primordia.

The repression of Arabidopsis FLC expression by vernalization (extended cold) has become a model for understanding polycomb-associated epigenetic regulation in plants. Antisense and sense non-coding RNAs have been respectively implicated in initiation and maintenance of FLC repression by vernalization. We show that the promoter and first exon of the FLC gene are sufficient to initiate repression during vernalization; this initial repression of FLC does not require antisense transcription. Long-term maintenance of FLC repression requires additional regions of the gene body, including those encoding sense non-coding transcripts.

Polycomb Repressive Complex 2 (PRC2) is an essential regulator of gene expression that maintains genes in a repressed state by marking chromatin with trimethylated Histone H3 lysine 27 (H3K27me3). In Arabidopsis, loss of PRC2 function leads to pleiotropic effects on growth and development thought to be due to ectopic expression of seed and embryo-specific genes. While there is some understanding of the mechanisms by which specific genes are targeted by PRC2 in animal systems, it is still not clear how PRC2 is recruited to specific regions of plant genomes.

Summary: The FLC gene encodes a MADS box repressor of flowering that is the main cause of the late‐flowering phenotype of many Arabidopsis ecotypes. Expression of FLC is repressed by vernalization; maintenance of this repression is associated with the deposition of histone 3 K27 trimethylation (H3K27me3) at the FLC locus. However, whether this increased H3K27me3 is a consequence of reduced FLC transcription or the cause of transcriptional repression is not well defined. In this study we investigate the effect of changes in transcription rate on the abundance of H3K27me3 in the FLC gene body, a chromatin region that includes sequences required to maintain FLC repression following vernalization. We show that H3K27me3 is inversely correlated with transcription across the FLC gene body in a range of ecotypes and mutants with different flowering times. We demonstrate that the FLC gene body becomes marked with H3K27me3 in the absence of transcription. When transcription of the gene body is directed by an inducible promoter, H3K27me3 is removed following activation of transcription and H3K27me3 is added after transcription is decreased. The rate of addition of H3K27me3 to the FLC transgene following inactivation of transcription is similar to that observed in the FLC gene body following vernalization. Our data suggest that reduction of FLC transcription during vernalization leads to an increase of H3K27me3 levels in the FLC gene body that in turn maintains FLC repression.

The Arabidopsis mutant Atubp26 initiates autonomous endosperm at a frequency of ∼1% in the absence of fertilization and develops arrested seeds at a frequency of ∼65% when self-pollinated. These phenotypes are similar to those of the FERTILIZATION INDEPENDENT SEED (FIS) class mutants, mea, fis2, fie, and Atmsi1, which also show development of the central cell into endosperm in the absence of fertilization and arrest of the embryo following fertilization. Atubp26 results from a T-DNA insertion in the UBIQUITIN-SPECIFIC PROTEASE gene AtUBP26, which catalyzes deubiquitination of histone H2B and is required for heterochromatin silencing. The paternal copy of AtUBP26 is able to complement the loss of function of the maternal copy in postfertilization seed development. This contrasts to the fis class mutants where the paternal FIS copy does not rescue aborted seeds. As in the fis class mutants, the Polycomb group (PcG) complex target gene PHERES1 (PHE1) is expressed at higher levels in Atubp26 ovules than in wild type; there is a lower level of H3K27me3 at the PHE1 locus. The phenotypes suggest that AtUBP26 is required for normal seed development and the repression of PHE1.