Genetic and epigenetic mechanisms underlying the regulation of flowering time

Developmental transitions and responses to the environment have a tight epigenetic control. Especially the switch to flowering is important for plants because it allows (sexual) reproduction, and should not occur unless the conditions are right. External and internal signals are relayed via FT/TFL1 genes; genes with a great impact on flowering time. The PRC1/PRC2 system plays an important role in the repression of flowering, indicated by the aberrant flowering phenotype of its mutants. Via deposition of histone modifications H2Aub1 and H3K27me3 it keeps genes stably silenced, so that transitions and responses do not happen as a result of random fluctuations in the internal or external environment. In the first manuscript we focussed on PRC2-component MSI1 in Arabidopsis. MSI1 is the nucleosome-binding core component of PRC2 and other chromatin-modifying complexes. We found that MSI1 is also a core part of a histone de-acetylase complex together with histone de-acetylase HDA19. We further found that this complex represses the ABA-mediated salt stress response by de-acetylating the ABA receptors. The second manuscript focused on flowering time in the invasive species Ambrosia artemisiifolia. During the last couple of centuries it invaded Europe, where it currently thrives mostly in the south-east. Earlier flowering populations have since been found in the north, suggesting local adaptation. We studied this early-flowering trait and found that it is inherited dominantly, and that it is maladaptive under long vegetation periods. We also identified the FT/TFL1 genes in this species, and found that a combination of expression changes in the FT-like floral activator and TFL1-like floral repressor likely underlies the altered flowering time. In the third manuscript we tried to shed light on repressive H2A de-ubiquitination in Arabidopsis. Previously, it has been observed that loss of UBP12/13-mediated H2A deubiquitination causes loss of H3K27me3 and re-activation of some PRC2 targets. We show now that this holds true on a genome-wide level, and that the genes targeted by UBP12/13 are those affected in H3K27me3 maintenance and expression. We also showed that H2Aub1 not only recruits PRC2, but likely also recruits H3K27 demethylase REF6. We show that H2Aub1 therefore puts genes in a state responsive to stimuli, and that stable repression requires its removal.