The number of meiotic crossovers (COs) is tightly regulated within a narrow range, despite a large excess of molecular precursors. The factors that limit COs remain largely unknown. Here, using a genetic screen in Arabidopsis thaliana, we identified the highly conserved FANCM helicase, which is required for genome stability in humans and yeasts, as a major factor limiting meiotic CO formation. The fancm mutant has a threefold-increased CO frequency as compared to the wild type. These extra COs arise not from the pathway that accounts for most of the COs in wild type, but from an alternate, normally minor pathway. Thus, FANCM is a key factor imposing an upper limit on the number of meiotic COs, and its manipulation holds much promise for plant breeding.

SCP3 is a meiosis-specific structural protein appearing at axial elements and lateral elements of the synaptonemal complex. We have analysed the behaviour of SCP3 and the cohesin subunit Rad21 in mouse spermatocytes by means of a squashing technique. Our results demonstrate that both proteins colocalize and are partially released from chromosome arms during late prophase I stages, although they persist at the interchromatid domain of metaphase I bivalents. Thus, Rad21 cannot be considered a `mitotic'-specific variant, but coexists with Rec8. During late prophase I SCP3 and Rad21 accumulate at centromeres, and together with the chromosomal passenger proteins INCENP and aurora-B kinase, show a complex `double cornet'-like distribution at the inner domain of metaphase I centromeres beneath the associated sister kinetochores. We have observed that Rad21 and SCP3 are displaced from centromeres during telophase I when sister kinetochores separate, and are not present at metaphase II centromeres. Thus, we hypothesise that Rad21, and the superimposed SCP3 and SCP2, are involved in the monopolar attachment of sister kinetochores during meiosis I, and are not responsible for the maintenance of sister-chromatid centromere cohesion during meiosis II as previously suggested.

These authors contributed equally to this work. SUMMARYThe movement of chromosomes during meiosis involves location of their telomeres at the inner surface of the nuclear envelope. Sad1/UNC-84 (SUN) domain proteins are inner nuclear envelope proteins that are part of complexes linking cytoskeletal elements with the nucleoskeleton, connecting telomeres to the force-generating mechanism in the cytoplasm. These proteins play a conserved role in chromosome dynamics in eukaryotes. Homologues of SUN domain proteins have been identified in several plant species. In Arabidopsis thaliana, two proteins that interact with each other, named AtSUN1 and AtSUN2, have been identified. Immunolocalization using antibodies against AtSUN1 and AtSUN2 proteins revealed that they were associated with the nuclear envelope during meiotic prophase I. Analysis of the double mutant Atsun1-1 Atsun2-2 has revealed severe meiotic defects, namely a delay in the progression of meiosis, absence of full synapsis, the presence of unresolved interlock-like structures, and a reduction in the mean cell chiasma frequency. We propose that in Arabidopsis thaliana, overlapping functions of SUN1 and SUN2 ensure normal meiotic recombination and synapsis.

. * For correspondence (fax +34 91 394 4844; e-mail pradillo@bio.ucm.es). † E. Sá nchez-Morá n and J.L. Santos acted as joint senior authors. SUMMARYThe eukaryotic recombinases RAD51 and DMC1 are essential for DNA strand-exchange between homologous chromosomes during meiosis. RAD51 is also expressed during mitosis, and mediates homologous recombination (HR) between sister chromatids. It has been suggested that DMC1 might be involved in the switch from intersister chromatid recombination in somatic cells to interhomolog meiotic recombination. At meiosis, the Arabidopsis Atrad51 null mutant fails to synapse and has extensive chromosome fragmentation. The Atdmc1 null mutant is also asynaptic, but in this case chromosome fragmentation is absent. Thus in plants, AtDMC1 appears to be indispensable for interhomolog homologous recombination, whereas AtRAD51 seems to be more involved in intersister recombination. In this work, we have studied a new AtRAD51 knock-down mutant, Atrad51-2, which expresses only a small quantity of RAD51 protein. Atrad51-2 mutant plants are sterile and hypersensitive to DNA double-strand break induction, but their vegetative development is apparently normal. The meiotic phenotype of the mutant consists of partial synapsis, an elevated frequency of univalents, a low incidence of chromosome fragmentation and multivalent chromosome associations. Surprisingly, nonhomologous chromosomes are involved in 51% of bivalents. The depletion of AtDMC1 in the Atrad51-2 background results in the loss of bivalents and in an increase of chromosome fragmentation. Our results suggest that a critical level of AtRAD51 is required to ensure the fidelity of HR during interchromosomal exchanges. Assuming the existence of asymmetrical DNA strand invasion during the initial steps of recombination, we have developed a working model in which the initial step of strand invasion is mediated by AtDMC1, with AtRAD51 required to check the fidelity of this process.

The temporal and functional relationships between DNA events of meiotic recombination and synaptonemal complex formation are a matter of discussion within the meiotic field. To analyse this subject in grasshoppers, organisms that have been considered as models for meiotic studies for many years, we have studied the localization of phosphorylated histone H2AX (c-H2AX), which marks the sites of double-strand breaks (DSBs), in combination with localization of cohesin SMC3 and recombinase Rad51. We show that the loss of c-H2AX staining is spatially and temporally linked to synapsis, and that in grasshoppers the initiation of recombination, produced as a consequence of DSB formation, precedes synapsis. This result supports the idea that grasshoppers display a pairing pathway that is not present in other insects such as Drosophila melanogaster, but is similar to those reported in yeast, mouse and Arabidopsis. In addition, we have observed the presence of c-H2AX in the X chromosome from zygotene to late pachytene, indicating that the function of H2AX phosphorylation during grasshopper spermatogenesis is not restricted to the formation of c-H2AX foci at DNA DSBs.