PCH'ing Together an Understanding of Crossover Control

Meiosis is a specialized form of cell division involving one round of chromosome replication followed by two rounds of segregation, thereby producing daughter cells with half the genomic equivalent of the progenitor. In most organisms, doublestrand breaks (DSBs) are introduced into the genome following premeiotic S-phase. These breaks are repaired almost exclusively from the homologous chromosome via repair pathways that yield either a crossover or non-crossover recombination product [1,2]. Of particular importance are the crossovers, which tether homologous chromosomes and ensure accurate segregation at the first meiotic division (MI) [3]. Chromosomes that fail to cross over have significantly higher rates of non-disjunction at MI, which produces aneuploid gametes, causing miscarriages and birth defects in humans. It should be no surprise then that most eukaryotes possess a sophisticated mechanism to control meiotic recombination. Consider the situation in an individual mouse meiocyte. More than 200 DSBs are made; however, only a subset of these precursors are repaired as crossovers, while the rest are repaired as noncrossovers (Figure 1) [4]. Thus, central to the ‘‘crossover control’’ mechanism is a decision to direct a given DSB to either a crossover or non-crossover fate [5]. This process ensures that each homolog pair receives at least one crossover (often referred to as the obligate crossover), and also regulates the spatial distribution of crossovers along chromosomes such that, if a chromosome receives two or more crossovers, they tend to occur further apart than expected by chance (referred to as crossover interference) (Figure 1B) [6]. Furthermore, it has been shown that when the number of DSBs is reduced, crossovers tend to be maintained at the expense of non-crossovers (a phenomenon called crossover homeostasis) [7,8]. It has been proposed that the obligate crossover, crossover interference, and crossover homeostasis are all manifestations of a single or closely related set of molecular processes, but this hypothesis remains to be rigorously tested [7,9,10]. Almost a century after the first observation of crossover control [11], we still know very little about the underlying mechanism(s). Most of the proteins that have been shown to influence crossover control in budding yeast appear to function downstream of the crossover/noncrossover decision. One such class of proteins, commonly referred to as ZMMs (Zip1/2/3/4, Msh4/5, Mer3), is specifically required for the repair of DSBs into crossovers that exhibit interference [2]. Deletion of any of the ZMM genes causes accumulation of intermediates in the crossover pathway and subsequent prophase arrest [12]. Two articles in this issue of PLoS Genetics identify a role for the Pachytene Checkpoint gene, PCH2, in crossover control in Saccharomyces cerevisiae [13,14]. PCH2, which encodes a putative AAA+-ATPase, was initially identified in yeast as a checkpoint factor due to suppression of a zip1D arrest in a pch2D mutant. This and other observations led to the hypothesis that Pch2 helps monitor chromosome synapsis during meiotic prophase [15,16]. However, studies in yeast, flies, and mice revealed that Pch2 is not just a checkpoint factor, but that it is also required for chromosome axis organization and DSB repair [17,18,19]. Interestingly, PCH2 is widely conserved in organisms that construct a synaptonemal complex and exhibit crossover interference, but is absent from organisms such as Schizosaccharomyces pombe that do not exhibit these features [15]. This observation suggested that Pch2 might also function in crossover control. Recent analysis in yeast demonstrated a small reduction in crossover numbers in pch2 mutants at the HIS4LEU2 recombination hotspot [19], but data available at the time did not allow evaluation of crossing over genome-wide and also did not address whether crossover control was normal. In studies published in this issue of PLoS Genetics, the Alani and Börner groups [13,14] have examined these issues in detail. When crossover frequencies were measured across several genetic intervals on chromosomes III, VII, and VIII, Zanders et al. [13] and Joshi et al. [14] observed either no difference or an increase (depending on the interval) in pch2D strains. Importantly, these analyses demonstrated that the crossovers formed in pch2D mutants show reduced interference. Recent studies have indicated that decreased crossover interference is associated with a concomitant decrease in crossover homeostasis [8]. To investigate this relationship, both Zanders et al. and Joshi et al. measured spore viability of pch2D strains carrying various hypomorphic alleles of the topoisomerase-like protein, Spo11. These hypomorphic alleles decrease the number of DSBs [20]. If crossover homeostasis and crossover interference are separate manifestations of a common crossover control mechanism, then an interference-defective mutant would also be expected to show defects in homeostasis, and thus a decrease in DSBs in such a mutant should result in fewer crossovers that are randomly distributed throughout the genome. Such a scenario would in turn be expected to result in an increase in the frequency of chromosome pairs without a crossover, causing reduced spore viability because of MI non-