Multiple mechanisms limit meiotic crossovers.

Meiotic crossovers (COs) have two important roles, shuffling genetic information and ensuring proper chromosome segregation. Despite their importance and a large excess of precursors (i.e DNA double strand-breaks, DSBs), the number of meiotic COs is tightly regulated, typically one to three per chromosome pair. Nevertheless, the mechanisms that ensure DSBs repair mostly as non-crossovers, and the evolutionary forces that impose this constraint, are poorly understood. Following a specific genetic screen, we identified several proteins that antagonize crossover formation in Arabidopsis Thaliana. This includes the helicase FANCM and its two co-factors MHF1 and MHF2, the BLM-like helicase RECQ4, the TOPOISOMERASE3a (TOP3α) and the AAA-ATPase FIDGETIN-LIKE1. Strikingly, the concomitant disruption of several of these activities led to a nine-fold increase in CO frequency, without affecting chromosome segregation and fertility. This shows that several parallel pathways actively limit CO formation and supports the idea that crossover number is restricted not because of mechanical constraints but likely because of long-term costs of recombination. Furthermore, this demonstrates how manipulating a few genes holds great promise for increasing recombination frequency in plant breeding programs.

Raphael Mercier

INRA, Institut Jean Pierre Bourgin, Versailles

Turning rice meiosis into mitosis

Achieving apomixis, the asexual production of seeds identical to the maternal parent, in rice would facilitate the deployment of hybrid vigor in the small farmer field. Apomictic seed development is a two-step process combining the production of unreduced spores bearing the maternal chromosome complement, apomeiosis, followed by the triggering of a parthenogenetic embryo development.

As a first step towards apomixis, we reconstructed in rice the MiMe apomeiotic mutant first generated in Arabidopsis (D’Erfurth et al 2009, PLoS Biology). The MiMe mutant is created by combining pair1 and rec8 mutations, which lead to a mitotic-like first division instead of the normal first meiotic division, together with a mutation in the Omission of Second Division (OSD1) gene, preventing the second division from occurring. The triple homozygous mutant pair1 rec8 osd1 was morphologically normal and had a restored fertility compared to single or double rec8 and pair1 mutants. Male meiocyte observation showed the anticipated phenotype, the replacement of meiosis by a mitotic-like division resulting in unreduced diploid gametes. Selfed progeny plants of the MiMe mutant were all tetraploid and, as anticipated, heterozygous at parental polymorphic SNP markers, indicating that they derive from the fusion of diploid gametes genetically identical to the mother plant. The next step is now to trigger a parthenogenetic embryo development from the female gamete of MiMe.

Delphine Mieulet

CIRAD, UMR AGAP, Montpellier