Rumors of Its Disassembly Have Been Greatly Exaggerated: The Secret Life of the Synaptonemal Complex at the Centromeres

The synaptonemal complex (SC) and the centromeres have long been thought to play their roles at quite different, nonoverlapping periods during meiosis. The SC is viewed as a hallmark structure that connects homologous chromosomes and is functionally critical for early prophase (zygotene and pachytene), while centromeres are known to assemble the kinetochore, act as a target for microtubules, and direct chromosome movement during anaphase I and II. However, several recent studies have suggested much earlier functions for centromeres (at least in some organisms) and detailed post-pachytene functions of the SC. First, studies in budding yeast, Drosophila, and higher plants provide evidence that centromeres play an important role in the initiation of synapsis in these systems [1–5]. Second, building on earlier work [6,7], recent studies have clearly demonstrated that in both yeast and flies the SC persists at the centromeres long after the end of pachytene— at least until late prophase [1,8,9]—and that these regions of centromeric pairing and synapsis play important roles in mediating segregation at anaphase I (see below). In this issue of PLoS Genetics, two studies of mammalian spermatocytes [10,11] demonstrate that while the centromeres are not involved in synaptic initiation in mammals, stretches of the SC at paired centromeres and at sites of crossing over do indeed persist beyond the breakdown of synapsis along the chromosome arms at the end of pachytene (see Figure 1). The SC is a proteinaceous railroadtrack–like structure that assembles in early meiotic prophase immediately following or concomitant with homolog pairing. This assembly process is often referred to as homolog synapsis. Although the SC is structurally complex, it may be thought of as consisting primarily of three components: lateral elements that run along the entire length of each homolog (and in mammals contain the SYCP3 protein); transverse filaments that, like the teeth of a zipper, serve to connect the two lateral elements of the SC (and include the mammalian SYCP1 protein); and a set of proteins located at the center of the SC (cleverly referred to as central element proteins). The question thus arises as to the function of these short regions of SC that survive the end of pachytene. Studies in yeast have clearly shown that the maintenance of centromeric pairing by short regions of the SC plays a role in mediating the segregation of both exchange and nonexchange chromosome segregation at meiosis I [8,9]. Similar data demonstrate that persistent pericentromeric pairings also play a critical role in mediating the segregation of non-exchange homologs in Drosophila [12,13]. In this issue of PLoS Genetics, the Pezza and Hunter groups show that the SC at mammalian centromeres also persists beyond the breakdown of synapsis along the chromosome arms at the end of pachytene in mouse spermatocytes [10,11]. Indeed, the centromeres are the last sites at which SC is observed. Both groups also demonstrate that the persistent pairing of centromeric regions is dependent on the transverse filament protein SYCP1. Moreover, Qiao et al. [11] characterize the structure of the SC at centromeres by structured illumination microscopy (SIM), revealing that the SC stretches at the centromere represent bona fide tripartite SC in which transverse filaments connect two lateral elements. Unlike SC breakdown along the chromosome arms following the end of pachytene, the much later dissolution of the SC at the centromeres may be a temporally complex process in which at least one SC component remains at the centromeres. Consistent with previous observations, both Bisig et al. [10] and Qiao et al. [11] show that while SYCP1 becomes undetectable prior to Nuclear Envelope Breakdown, SYCP3 remains associated with the centromeres until at least anaphase I in males. Qiao et al. also suggest that in some cases the persistence of SYCP3 at the centromere may provide a ‘‘bridge between the two homologous kinetochores’’ at diakinesis/metaphase I. Studies in other organisms support the view that such SYCP3 bridges may play important roles in facilitating segregation of achiasmate sex chromosomes [14]. However, centromeres do not appear to function as synaptic initiation sites in mammals. Indeed, both Bisig et al. and Qiao et al. demonstrate that centromeres are the last to pair, doing so only at zygotene–pachytene transition. Qiao et al. also use Rnf212 (Zip3 homolog) knock-out mice to explore whether or not centromeres can be ‘‘forced’’ to become synaptic initiation sites. In yeast zip3 mutants, synapsis occurs predominantly at the centromeres [2]. In Rnf212 spermatocytes, however, synapsis does not initiate at centromeres even though SC-associated centromere pairing at diplotene in the mutants was similar to that observed in wild-type. These data show clearly that,