Doubled-haploid induction generates extensive differential DNA methylation in Arabidopsis.

Doubled haploids (DHs) fix traits from hybrids in one generation. DH induction includes two changes in ploidy levels typically associated with variation in DNA methylation. However, DNA methylation patterns in DH plants and their biological significance are largely unknown. We generated three DH lines in Arabidopsis thaliana by crossing a haploid inducer with the accession Col-0, thus removing tissue culture and hybridization as a variable. DH induction produced thousands of differentially DNA methylated regions (DMRs), most of which were stochastic. Both haploidization and colchicine-induced genome duplication produced DMRs; the former mainly yielded DMRs at non-CG contexts, whereas the latter affected differential gene body methylation. Spontaneous genome doubling of haploid plants also induced DMRs in greater numbers than self-propagation. Our results provide the first evidence that haploid induction and genome doubling result in differential DNA methylation, offering a novel approach to induce epialleles.

Meiocyte size is a determining factor for unreduced gamete formation in Arabidopsis thaliana.

Polyploidy, the presence of more than two sets of chromosomes within a cell, is a widespread phenomenon in plants. The main route to polyploidy is considered through the production of unreduced gametes that are formed as a consequence of meiotic defects. Nevertheless, for reasons poorly understood, the frequency of unreduced gamete formation differs substantially among different plant species. The previously identified meiotic mutant jason (jas) in Arabidopsis thaliana forms about 60% diploid (2n) pollen. JAS is required to maintain an organelle band as a physical barrier between the two meiotic spindles, preventing previously separated chromosome groups from uniting into a single cell. In this study, we characterized the jas suppressor mutant telamon (tel) that restored the production of haploid pollen in the jas background. The tel mutant did not restore the organelle band, but enlarged the size of male jas tel meiocytes, suggesting that enlarged meiocytes can bypass the requirement of the organelle band. Consistently, enlarged meiocytes generated by a tetraploid jas mutant formed reduced gametes. The results reveal that meiocyte size impacts chromosome segregation in meiosis II, suggesting an alternative way to maintain the ploidy stability in meiosis during evolution.

The Bub1-TPR Domain Interacts Directly with Mad3 to Generate Robust Spindle Checkpoint Arrest.

The spindle checkpoint monitors kinetochore-microtubule interactions and generates a "wait anaphase" delay when any defects are apparent [1-3]. This provides time for cells to correct chromosome attachment errors and ensure high-fidelity chromosome segregation. Checkpoint signals are generated at unattached chromosomes during mitosis. To activate the checkpoint, Mps1Mph1 kinase phosphorylates the kinetochore component KNL1Spc105/Spc7 on conserved MELT motifs to recruit Bub3-Bub1 complexes [4-6] via a direct Bub3 interaction with phospho-MELT motifs [7, 8]. Mps1Mph1 then phosphorylates Bub1, which strengthens its interaction with Mad1-Mad2 complexes to produce a signaling platform [9, 10]. The Bub1-Mad1 platform is thought to recruit Mad3, Cdc20, and Mad2 to produce the mitotic checkpoint complex (MCC), which is the diffusible wait anaphase signal [9, 11, 12]. The MCC binds and inhibits the mitotic E3 ubiquitin ligase, known as Cdc20-anaphase promoting complex/cyclosome (APC/C), and stabilizes securin and cyclin to delay anaphase onset [13-17]. Here we demonstrate, in both budding and fission yeast, that kinetochores and KNL1Spc105/Spc7 can be bypassed; simply inducing heterodimers of Mps1Mph1 kinase and Bub1 is sufficient to trigger metaphase arrest that is dependent on Mad1, Mad2, and Mad3. We use this to dissect the domains of Bub1 necessary for arrest, highlighting the need for Bub1-CD1, which binds Mad1 [9], and Bub1's highly conserved N-terminal tetratricopeptide repeat (TPR) domain [18, 19]. We demonstrate that the Bub1 TPR domain is both necessary and sufficient to bind and recruit Mad3. We propose that this brings Mad3 into close proximity to Mad1-Mad2 and Mps1Mph1 kinase, enabling efficient generation of MCC complexes.