Homologous recombination, sister chromatid cohesion, and chromosome condensation in mammalian meiosis

In the life cycle of sexually reproducing eukaryotes, haploid and diploid generations of cells alternate. Two types of cell division occur in such a life cycle: mitosis and meiosis. They are compared in chapter 1 . Haploid and diploid cells can multiply by mitoses. A mother cell undergoing mitosis produces two daughter cells each with the same chromosome complement and combination of alleles as the mother cell. Only diploid cells can undergo meiosis. A cell undergoing meiosis produces four haploid cells with different combinations of alleles. Meiosis thus contributes to genetic diversity and is responsible for the transition from the diploid to the haploid phase in the life cycle. Meiosis has probably evolved from mitosis by modifications of the cell cycle, chromosome behavior and recombinational repair.The cell cycle has been modified because in meiosis two nuclear divisions (meiosis I and II) follow a single round of DNA-replication (S-phase), whereas in mitosis, one S-phase is followed by one division.Chromosome behavior has been modified in several respects. In meiotic prophase (prophase I), homologous chromosomes pair and associate by a zipper like structure, the synaptonemal complex (SC). SCs consist of two axial elements (AEs), which each support the two sister chromatids of one homolog. During the course of prophase I, numerous transverse filaments connect the axial elements of the two homologs along their entire length, a process called synapsis. Within the context of SCs, homologous recombination (HR) takes place between the paired chromosomes. At meiosis I, homologous chromosomes disjoin, which brings about the reduction in ploidy level (from diploid to haploid); at meiosis II, the sister chromatids separate, like in mitosis. Proper chromosome segregation in the two successive meiotic divisons is ensured by a modification in the regulation of sister chromatid cohesion. In mitosis, cohesion is released at once all along the chromosomes at the metaphase to anaphase transition. In meiosis, in contrast, cohesion is lost in two steps: chromosome arm cohesion is lost at meiosis I, and centromeric cohesion at meiosis II.Meiotic HR has probably evolved from mitotic homologous recombinational repair (HRR), and this has also been accompanied by several modifications: (1) in contrast to mitotic HRR, meiotic HR is actively induced by the cell; (2) meiotic HR occurs at a 100- to 1000-fold higher frequency than mitotic HRR; and (3) meiotic HR prefers a non-sister chromatid of the homologous chromosome as template, whereas mitotic HRR prefers the sister chromatid. In this thesis I focus on these modifications that the mitotic cell cycle must have undergone to yield a reliable meiotic cycle in which chromosomes are properly duplicated and segregated, and retain their integrity.In chapter 2 , we describe the immunolocalization of two proteins that are involved in the early steps of meiotic recombination, Rad50 and Mre11, in spermatocytes of mouse and rat. We found a similar localization of the two proteins in spermatocytes, which we expected, because Rad50 and Mre11 make part of the same protein complex. In early prophase I (pre-leptotene until early zygotene) Rad50 and Mre11 were present throughout the nucleus. In slightly later stages (mid and late zygotene), the two proteins concentrated in distinct domains, around segments of AEs that were not yet connected by transverse filaments. In the next stage (pachytene), both proteins disappeared from the nucleus except from the pair of sex chromosomes (XY-bivalent), where they remained until the end of diplotene. Unlike other proteins involved in meiotic recombination, Rad50 and Mre11 did not associate visibly with the AEs. We propose that Mre11 and Rad50, together with other proteins, prepare chromatin throughout the early meiotic prophase nucleus for the initiation of meiotic recombination. Possibly, only a small fraction of the Rad50 and Mre-containing (pre)recombination complexes associates transiently with AEs, where further steps in meiotic recombination can take place.In mitotic cells, Mre11 and Rad50 are involved in two major pathways of double-strand DNA-break repair (DSB repair), namely nonhomologous endjoining (NHEJ) and homologous recombinational repair (HRR). NHEJ connects DNA-ends irrespective of their sequence; it is an error-prone DSB repair mechanism, which predominates in mitotic cells. In contrast, mitotic HRR and meiotic HR repair DSBs accurately. In meiosis, DSBs should be repaired by HR, not by NHEJ. We wondered what determines the choice between the two repair pathways. Therefore we analysed the interactions engaged by Mre11 in mammalian somatic and meiotic cells ( chapter 3 ). We found a physical interaction between Mre11 and Ku70, a protein that is essential for NHEJ, but not for HR or HRR. Mre11 depended on Ku70 for the formation of subnuclear foci that are assembled upon irradiation of mitotic cells, and that are supposed to represent DNA repair complexes. Nevertheless, Ku70 and Mre11 were differentially expressed during meiosis. In the mouse testis, Mre11 and Ku70 colocalised in nuclei of somatic cells and in the XY bivalent of pachytene spermatocytes. However, in early meiotic prophase, when meiotic recombination is initiated, Mre11 was abundant (like Rad50, see Chapter 2 ), while Ku70 was not detectable. We propose that Ku70 acts as a switch between the two DSB repair pathways. If present, Ku70 would destine DSBs for NHEJ by binding to DNA-ends and attracting other factors for NHEJ, including Mre11. If absent, the DNA-ends and Mre11 can participate in the meiotic HR pathway.In chapters 4 to 6, we focus on meiotic adaptations of sister chromatid cohesion. Cohesion is ensured by a four-protein complex, cohesin. In chapter 4 we analyse the localization of two components of the cohesin complex, Smc1 and Smc3, relative to the AEs of SCs by immunofluorescence. For the recogition of AEs, we used antibodies against two AE-components, Scp2 and Scp3. Smc1 and Smc3 localized in a beaded structure along the AEs. Furthermore, we found that Scp2 and Scp3 co-immuneprecipitated with Smc3 from testis extracts. We also showed interaction between Smc1 and Scp3, using immunoaffinity chromatography. Together, these data suggest interactions between cohesin components (cohesins) and AE-components in vivo.In chapter 5, we describe the identification of a meiosis-specific variant of Smc1, Smc1b. Smc1bis highly homologous to Smc1 (further designated as Smc1a) except for a unique C-terminal, basic, DNA binding motif. Smc1bis specifically expressed in the testis and co-immuneprecipitates with Smc3 from testis nuclear extracts. Immunolocalization of Smc1bin sections of rat testis revealed that Smc1bis localized along the AEs throughout prophase I. In spread spermatocyte nuclei, we found that Smc1b, like Scp3, is present along AEs from leptotene until diakinesis, when AEs disassemble. Smc1bremained present in the centromeric region until metaphase II, together with Scp3, and disappeared at the onset of anaphase II. This localization pattern of Smc1bis consistent with a role of the protein in maintaining sister chromatid cohesion between centromeres until anaphase II. Possibly, a meiosis-specific isoform of Smc1awas needed in mammals for the modified regulation of cohesion in meiosis, and/or Smc1bwas required for the assembly of protein complexes involved in meiotic HR.In chapter 6 , we describe the localization of another meiosis-specific cohesin, Rec8, relative to known AE-components. Rec8 appeared shortly before premeiotic S-phase in the nucleus and formed AE-like structures (Rec8-AEs) in the absence of Smc1b, Smc3, Scp2 and Scp3. In the subsequent stage of prophase I (leptotene) the Smcs and Scps appeared, and localized along the Rec8 AEs. Initially, they formed dots along Rec8-AEs, which later extended and fused until they lined Rec-AEs along their length. Rec8 persisted along the AEs throughout prophase I. In metaphase I, Smc1b, Smc3, Scp2 and Scp3 disappeared from the chromosome arms and accumulated around the centromeres, where they stayed until anaphase II. In striking contrast, Rec8 persisted along the chromosome arms until anaphase I and near the centromeres until anaphase II. We conclude that the first steps in AE assembly do not require Smc1b, Smc3, Scp2 and Scp3. We propose that Rec8 provides a basis for AE-formation. Furthermore, Smc1b, Smc3, Scp2 and Scp3 cannot provide cohesion during the earliest stages of meiotic prophase, nor arm cohesion during metaphase I. We propose that Rec8 then provides cohesion. Furthermore, we found evidence for interaction of Rec8 with proteins involved in meiotic HR: Rad50 and Rad51 and/or Dmc1 co-immunoprecipitate with Rec8. Possibly, Rec8 provides also a basis for assembly of meiotic recombination complexes. We hypothesize that the replacement of mitotic cohesin Scc1 by meiotic cohesin Rec8 was necessary to allow the assemly of AEs. The AEs in turn were required both for the altered regulation of cohesion and for blocking the sister chromatid as template for repair of meiotic DSBs by HR.From chapters 4-6 it has become clear that cohesins form the basis of AEs in meiosis. AEs are single axial structures that are shared by the two sister chromatids. At the beginning of diakinesis, this shared chromatid axis disappears, and at the end of diakinesis, two axes appear, which each support one individual chromatid (I call these "single-chromatid axes"). I wandered if cohesin plays a role in the establishment of the single-chromatid axis at the end of diakinesis. I therefore wrote an overview of chromosomal axial structures in mitosis and meiosis ( chapter 7 ). I included in my overview the literature from the pre-immunocytochemistry era, when axes were visualized by silver staining. Silver staining reveals axial structures in mitosis and meiosis. In mitosis, a shared sister chromatid axis, (one axis supporting two sister chromatids) is not normally seen, whereas many publications have appeared about the mitotic single-chromatid axis. Some protein components which make part of the mitotic single-chromatid axis have been identified, as well as the DNA sequences (the scaffold attachment regions or SARs) by which chromatin is attached to this axis. with respect to meiosis, in contrast, many publications have appeared about the silver-stainable shared chromatid axis (which should correspond to the AE), while little is known about the meiotic single-chromatid axis. Although no hard conclusions could be drawn about the role of cohesin in the establishment of the single-chromatid axis in mitosis or meiosis, many ideas for useful and new experiments came up, which can fill the gap in today's knowledge of chromosomal axial structures.