Introduction

Meiosis is a specialized type of cell division that reduces chromosome number by half, resulting in the production of four genetically diverse haploid gametes, so that fertilization during sexual reproduction restores diploidy. This process occurs in two stages: meiosis I, in which homologous chromosomes are partitioned, and meiosis II, in which sister chromatid are segregated. The regulation of meiosis is crucial for ensuring that both genetic recombination and chromosome segregation occur accurately.

During prophase I, homologous chromosomes pair and this pairing is stabilized through a process called synapsis, in which a protein structure called the synaptonemal complex (SC) holds homologs together. This close association between homologs during synapsis facilitates crossover formation, the process in which double strand breaks (DSBs) are deliberately introduced into the genome, repaired by meiosis-specific mechanisms that exchange DNA between homologous chromosomes and generate the chiasmata, or linkage, that promotes accurate meiotic chromosome segregation.

Therefore, disruptions in pairing, synapsis, or recombination prevent the formation of chiasmata and can lead to meiotic errors such as nondisjunction, resulting in aneuploid gametes. Errors in human meiosis are associated with birth defects, such as Down and Turner syndromes, infertility and miscarriages, underscoring the importance of understanding meiosis to human health (Hassold & Hunt, 2001).

In addition to the fundamental role that recombination plays in ensuring accurate chromosome segregation, crossover recombination accomplishes another important function: it generates new haplotypes for natural selection to act upon to drive evolution. Thus, to assure a random assortment of alleles on a population level, the distribution of crossovers may be as tightly regulated as their number. The significance of controlling both crossover number and distribution is clearly illustrated by the existence of mechanisms such as crossover assurance, in which every pair of homologous chromosomes gets at least one crossover; crossover homeostasis, in which the number of crossovers remains relatively invariant even if the number of recombination precursors change; and crossover interference, in which the presence of a crossover inhibits the formation of a crossover nearby (Gray & Cohen, 2016). DSBs typically vastly outnumber crossovers in most organisms and are introduced gradually throughout early prophase. Therefore, to accomplish this precise level of control, meiotic crossover recombination and the decision about which DSBs become crossover-eligible intermediates, and eventually crossovers, is implemented progressively throughout meiotic prophase (Cole et al., 2012; Joshi et al., 2015; Morgan et al., 2021; Yokoo et al., 2012).

PCH-2, also known as TRIP13 in mammals, is an evolutionarily ancient AAA-ATPase that plays a significant role in regulating meiosis across different organisms, including M. musculus (mice), S. cerevisiae (budding yeast), D. melanogaster (fruit flies), and C. elegans (worms) (Bhalla, 2023). PCH- 2 and its orthologs structurally remodel a family of proteins with HORMA domains (HORMADs) to control their function (Gu et al., 2022). HORMADs participate in a variety of signaling events and can exist in at least three structurally distinct conformations: a “closed” conformation, which they adopt when they bind a short peptide sequence in their own protein sequence or another protein (also called a closure motif) and their C-terminus wraps around this motif to stabilize the interaction; an “open” conformation when unbound, in which their C-terminus is discretely tucked against the HORMA domain; and an “extended” conformation, which is an intermediate between the two (Gu et al., 2022). PCH-2 and its orthologs convert the closed version of HORMADs to the open or extended versions, playing an important role in recycling HORMADs during signaling. During meiosis, closed versions of meiotic HORMADs assemble on chromosomes to form meiotic chromosome axes, which are essential for pairing, synapsis and recombination between homologous chromosomes (Kim et al., 2014). The remodeling of meiotic HORMADs by PCH-2 to an “open” or “extended” conformation is thought to reduce the levels of HORMADs on chromosomes (Börner et al., 2008; Cuacos et al., 2021; Lambing et al., 2015; Wojtasz et al., 2009) and/or increase their dynamic association and dissociation (Russo et al., 2023), modulating and coordinating homolog pairing, synapsis and recombination during prophase.

The conserved role of the PCH-2/HORMAD module in meiosis has been difficult to characterize, in part, we have argued, because of the evolutionary innovation that is an inherent aspect of sexual reproduction (Bhalla, 2023). Moreover, in some systems, such as budding yeast and plants (Herruzo et al., 2021; Yang et al., 2020), PCH-2 orthologs not only remodel meiotic HORMADs on meiotic chromosomes but also perform this function in the cytoplasm to make meiotic HORMADs available for their role(s) in meiotic nuclei. This dual role can complicate functional analyses, particularly in pch-2 null mutants. However, all meiotic systems exhibit defects in the number and distribution of crossovers when PCH-2 function is abrogated by mutation (Deshong et al., 2014; Joshi et al., 2009; Joyce & McKim, 2009; Lambing et al., 2015; Roig et al., 2010; Zanders & Alani, 2009). Unfortunately, the lack of a clear pattern when analyzing these defects in recombination in pch-2 mutants has contributed to an inability to develop a unified, integrated model of PCH-2 function in the field.

In C. elegans, meiotic HORMADs localize to meiotic chromosomes independently of PCH-2 (Deshong et al., 2014). Here, we exploit this system to show that PCH-2 is required to control the number and distribution of crossovers by antagonizing the formation of crossovers. This antagonism produces different consequences depending on the stage of meiotic prophase. In early meiotic prophase, PCH- 2 inhibits DSBs from becoming crossovers, ensuring that they are more widely distributed than sites of initial DSB formation and/or homolog interactions. Later in meiotic prophase, PCH-2 is responsible for winnowing the numbers of crossover-eligible intermediates on synapsed chromosomes, contributing to the reinforcement of crossover-eligible intermediates, designation of crossovers and ultimately, crossover assurance. Genetic analysis demonstrates that PCH-2’s regulation of crossover-eligible intermediates is through one of three essential meiotic HORMADs, HIM-3. Finally, we link PCH-2’s effect on early DSBs in early meiotic prophase to cell cycle stage, demonstrating that both limit early DSBs from becoming crossovers. We propose that PCH-2’s remodeling of HIM-3 on meiotic chromosomes destabilizes crossover-eligible intermediates throughout meiotic prophase, contributing to the progressive implementation of meiotic recombination to control the number and distribution of crossovers and accomplish crossover control.

Results

PCH-2 controls the number and distribution of crossovers in similar patterns on multiple chromosomes

We had previously shown that loss of PCH-2 reduced the frequency of double crossovers and genetic length of both an autosome (chromosome III) and the X chromosome, albeit not uniformly among genetic intervals (Deshong et al., 2014). To determine whether a more obvious pattern could be observed, we expanded our analysis by monitoring recombination genetically in both wildtype and pch-2 mutant animals using five single nucleotide polymorphisms (SNPs) that spanned 95% of Chromosomes, I, III, IV and the X chromosome (Figure 1). We excluded Chromosome II from our analysis because of the potential difficulty combining Hawaiian SNPs with the pch-2 mutation, which is linked to Chromosome II, and Chromosome V because of our use of the bcIs39 transgene to identify cross progeny, which may disrupt recombination on that chromosome.

Figure 1.

In wildtype animals, we observed multiple double crossovers, ranging from 1-13 depending on the chromosome. In a majority of these double crossovers (69%), one crossover was at the end of the chromosome where pairing and synapsis initiate, also called the Pairing Center (PC) (MacQueen et al., 2005), suggesting some relationship between where chromosomes might make initial contacts and the likelihood of double crossovers. On all chromosomes analyzed, we observed no double crossovers in pch-2 mutants and this difference was statistically significant for Chromosomes I, III and X. Moreover, there was a striking shift of crossovers to the PC end of all four chromosomes tested.

This shift in the distribution of crossovers to the PC ends of chromosomes was generally accompanied by a reduction in crossovers in the center of chromosomes. In C. elegans, the center of chromosomes are where double strand breaks are less numerous (Nadarajan et al., 2021; Yu et al., 2016) and where genes are more abundant. In the case of the X chromosome and chromosome III, recombination at the non-PC end was also reduced and appeared to more closely resemble the physical map at this end of these chromosomes. Thus, PCH-2 ensures a wider distribution of crossovers across chromosomes, away from regions that are more likely to undergo early homolog interactions (PC ends) and with more double strand breaks (both the PC and non-PC ends of chromosomes), and towards the center of chromosomes, where double strand breaks are less abundant.

PCH-2 prevents exogenous DSBs early in meiotic prophase from becoming crossovers

We previously showed that PCH-2 promotes crossover formation and crossover assurance through its regulation of the meiotic HORMAD, HIM-3, in C. elegans (Russo et al., 2023). Moreover, this promoting role seems linked to its localization to the synaptonemal complex (Patel et al., 2023).

However, the observed shift in the distribution of crossovers in pch-2 mutants suggests that PCH-2 may also play a role in inhibiting crossovers and that this role may be occurring in early meiotic prophase, when chromosomes are undergoing initial homolog interactions. Previous data from our lab had suggested this possibility, but experimental caveats limited our ability to properly interpret these experiments (Deshong et al., 2014).

To explicitly test this possibility, we took advantage of both the spatiotemporal organization of meiotic nuclei in the C. elegans germline, and the observation that nuclei travel in an assembly line process at a stereotypical pace to late pachytene (Jaramillo-Lambert et al., 2007), where we can cytologically assess crossover formation by staining for the essential crossover factor, COSA-1 (Yokoo et al., 2012). We performed irradiation experiments to introduce exogenous DSBs at different timepoints during meiotic prophase and analyzed PCH-2’s role in crossover formation (Figure 2A).

Figure 2.

In this experiment, we tested the effects of ionizing radiation at two distinct time points: 8 hours after irradiation, when we can monitor crossover formation in meiotic nuclei that received exogenous DSBs in mid-pachytene, and 24 hours after irradiation, when we monitor crossover formation in meiotic nuclei that received exogenous DSBs in leptotene/zygotene, also known as the transition zone in C. elegans. In order to detect crossovers, control and pch-2 mutants germlines were stained for GFP::COSA-1, which allowed us to visualize crossovers, and DAPI, which allowed us to visualize DNA. At 8 hours after irradiation, we observed a significant increase in nuclei containing less than 6 COSA-1 nuclei in pch-2 mutants (26%), when compared to control (9%) germlines (Figure 2B, 2C). We did not observe a significant number of nuclei with greater than 6 COSA-1 foci in either backgrounds. This observation is consistent with our and others’ previous findings: exogenous DSBs in mid-pachytene do not affect the number of crossovers in wildtype animals (Yokoo et al., 2012) and loss of PCH-2 leads to a decrease in crossover formation (Deshong et al., 2014).

Next, we examined crossover formation 24 hours post-irradiation (Figure 2D, 2E). For this time point, we also saw a significant increase in nuclei containing less than 6 COSA-1 foci (control: 13%, pch-2: 26%), similar to the 8 hour time point. However, in contrast to the 8 hour time point, we also observed a significant increase in nuclei with more than 6 COSA-1 foci in pch-2 mutants (wildtype: 2%, pch-2: 13%), indicating that loss of PCH-2 leads to increased crossover formation when nuclei in the transition zone get more DSBs. In other words, PCH-2 inhibits exogenous DSBs in early meiotic prophase from becoming crossovers, consistent with our analysis of recombination on multiple chromosomes.

PCH-2 prevents SPO-11-induced DSBs from becoming crossovers in early meiotic prophase

Our previous experiment indicated that PCH-2 is preventing exogenous DSBs in early meiotic prophase from becoming crossovers (Figure 2). To test whether this was also the case for programmed meiotic DSB formation, we used the auxin-inducible degradation (AID) system to remove the enzyme that is responsible for programmed meiotic DSBs, SPO-11 (spo-11::AID::3XFLAG) (Dernburg et al., 1998; Keeney et al., 1997). Since meiotic nuclei travel through the germline at about one cell per row per hour (Jaramillo-Lambert et al., 2007), we took advantage of this precise spatiotemporal resolution to test five different timepoints to determine if acute depletion of SPO-11 leads to any changes in crossover formation between control animals and pch-2 mutants (Figure 3A). We assayed crossover formation by analyzing both the number of GFP::COSA-1 foci in late pachytene and the presence of bivalents in diakinesis, which indicates the successful formation of chiasmata between homologous chromosomes. Meiotic nuclei travel from late pachytene to diakinesis in approximately 12 hours (Deshong et al., 2014).

Figure 3.

When we treated both control and pch-2 mutant worms for 36 hours with auxin, we observed a complete loss of GFP::COSA-1 foci (Figure 3C). When these animals were treated with auxin for 48 hours, to allow nuclei without SPO-11 an additional 12 hours to reach diakinesis, we observed 12 DAPI stained bodies, or univalents, which are the 6 homolog pairs that have not formed chiasmata (Figure 3E). Thus, we can reliably remove SPO-11 from meiotic nuclei throughout the germline and prevent crossover formation.

Having established that we can reliably remove SPO-11 with the AID system, we performed the same experiment at additional timepoints that remove SPO-11 at different stages of early meiotic prophase, monitoring crossover formation both at late pachytene and 12 hours later in diakinesis: 24 hours to monitor GFP::COSA-1 foci and 36 hours to monitor bivalent formation; 22 (GFP::COSA-1 foci) and 34 (bivalent formation) hours; 20 (GFP::COSA-1 foci) and 32 hours (bivalent formation); and 18 (GFP::COSA-1 foci) and 30 hours (bivalent formation). We also performed a control experiment, treating worms with ethanol for the 24 and 36 hour timepoints to ensure that this tagged version of SPO-11 was fully functional for crossover formation (Figure 3C, 3D).

At the 24 hour time point, when meiotic nuclei in the transition zone do not receive SPO-11-induced DSBs, we observe a range of GFP::COSA-1 foci in late pachytene in both control animals and pch-2 mutants with a majority of nuclei having 0 or 1 focus (control and pch-2 averages: 1.8 GFP::COSA-1 foci), indicating a loss of crossover formation, albeit with some heterogeneity. This was confirmed when we monitored bivalent formation at the 36 hour timepoint in both strains. This heterogeneity likely reflects that it takes 1-4 hours to significantly affect DSB formation through acute degradation of SPO-11 at this stage of meiosis in C. elegans (Hicks et al., 2022). In addition, there was no statistically significant difference in the number of GFP::COSA-1 foci between control animals and pch-2 mutants. However, we did observe a slight, statistically significant difference in the number of DAPI stained bodies between control animals (average number of DAPI stained bodies: 10.6) and pch-2 mutants (average number of DAPI stained bodies: 9.6), suggesting that even at this very early time point, we can detect a role for PCH-2 in preventing early SPO-11-induced DSBs from becoming crossovers (Figure 3D, 3E).

We continued to monitor GFP::COSA-1 and bivalent formation at successive timepoints. At the 22 hours and 34 hour timepoints, where SPO-11 depletion begins later in the transition zone, we saw an increase in the number of GFP::COSA-1 foci and a reduction in the number of DAPI stained bodies, suggesting that more DSBs are becoming crossovers. This trend continued at the 20 hour and 32 hour timepoints, eventually reaching wildtype numbers for GFP::COSA-1 and DAPI stained bodies by the 18 hour and 30 hour timepoints, respectively.

At the 22 hour and 36 hour timepoints, we did not detect any statistically significant difference in the number of GFP::COSA-1 foci or DAPI stained bodies between control animals and pch-2 mutants, even when we analyzed the number of nuclei that had greater than 6 GFP::COSA-1 foci. In contrast, while the number of GFP::COSA-1 foci was not statistically significantly different between control animals and pch-2 mutants at the 20 hour time point, the number of nuclei with greater than 6 GFP::COSA-1 foci was significantly higher in pch-2 mutants (control: 3%, pch-2: 10%, p value = 0.009, Fischer’s Exact Test), similar to our analysis of GFP::COSA-1 24 hours after irradiation. Even more strikingly, the number of DAPI stained bodies was significantly lower in pch-2 animals than control animals at the corresponding 32 hour timepoint, indicating that more homolog pairs had successfully formed chiasmata in pch-2 mutants. The inconsistency between the average number of GFP::COSA-1 foci and DAPI stained bodies at this time point can be explained by our ability to more reliably detect pch-2’s loss of crossover assurance when quantifying GFP::COSA-1 foci than when analyzing bivalent formation (Deshong et al., 2014; Russo et al., 2023). No difference in GFP::COSA- 1 foci or bivalent formation was observed at the 18 hour and 30 hours timepoints, respectively. These data indicate that during a narrow time frame of early meiotic prophase, likely during leptotene/zygotene given the delay in limiting DSB formation with auxin-induced degradation of SPO- 11, PCH-2 prevents SPO-11-induced DSBs from becoming crossovers and chiasmata.

PCH-2 is required for timely loading and removal of MSH-5 on meiotic chromosomes through its regulation of HIM-3

After DSB formation, a subset of DSBs are repaired through a pro-crossover pathway and eventually winnowed to a stereotypical number of crossovers. In C. elegans, this can be visualized by the loading of the pro-crossover factor MSH-5, a component of the meiosis-specific MutSγ complex that stabilizes crossover-specific DNA repair intermediates called joint molecules (Janisiw et al., 2018; Kelly et al., 2000; Snowden et al., 2004). GFP::MSH-5 begins to form a few foci in the transition zone, becoming more numerous in early pachytene before decreasing in number in mid pachytene to ultimately colocalize with COSA-1 marked sites in late pachytene (Janisiw et al., 2018; Woglar & Villeneuve, 2018; see control in Figure 4A, 4B). The mechanism through which MSH-5 functions with the SC and other meiosis factors to ultimately form crossovers has not been established but may involve phosphorylation by cyclin-dependent kinases (Haversat et al., 2022; Zhang et al., 2021).

Figure 4.

Given that we have shown that PCH-2 prevents early DSBs from becoming crossovers (Figures 2 and 3), we tested whether we could cytologically detect this inhibition at the level of MSH-5 behavior in control animals and pch-2 mutants. We generated wildtype and pch-2 mutants with GFP::MSH-5 and quantified the average number of GFP::MSH-5 foci per row of nuclei from the transition zone to late pachytene (Figure 4A, 4B). We observed a statistically significant increase in the average number of GFP::MSH-5 foci per row in the transition zone in pch-2 mutants (Figure 4A, 4B, p value < 0.0001, Student’s t-test), indicating that PCH-2 typically limits MSH-5 loading at this early stage of meiotic prophase. The average number of GFP::MSH-5 foci per row were similar in number in early pachytene in both backgrounds. However, we also observed a substantial increase in the average number of GFP::MSH-5 foci per row in pch-2 mutants in the mid and late pachytene regions (Figure 4A, 4B), indicating that PCH-2 also promotes the turnover or maturation of GFP::MSH-5 foci at these later stages of meiotic prophase. Thus, PCH-2 prevents the loading of the crossover-promoting factor, MSH-5, during early meiotic prophase, consistent with both our genetic analysis of recombination (Figure 1) and its role in preventing early DSBs from becoming crossovers (Figures 2 and 3).

However, PCH-2 also appears to limit the number of crossover-eligible intermediates in mid and late pachytene, an unexpected observation given that the most severe recombination defect in pch-2 mutants is the loss of crossover assurance (Deshong et al., 2014).

We previously showed that PCH-2 genetically interacts with meiotic HORMADs to effectively control different aspects of meiosis: PCH-2 regulates pairing and synapsis through its regulation of HTP-3 and crossover recombination through its regulation of HIM-3 (Russo et al., 2023). To determine which meiotic HORMAD PCH-2 might be regulating to affect GFP::MSH-5’s loading and removal on meiotic chromosomes, we constructed gfp::msh-5;htp-3^H96Y, gfp::msh-5;pch-2;htp-3^H96Y, gfp::msh-5;him-3^R93Y and gfp::msh-5;pch-2;him-3^R93Y mutants and quantified GFP::MSH-5 foci throughout the germline.

Both htp-3^H96Y and him-3^R93Y mutants showed a drastic increase in the average number of GFP::MSH- 5 per row of nuclei throughout the germline (Figures 4C and 4D, Supplemental Figure 1). For example, compared to control germlines that peaked at approximately an average of 20 MSH-5 foci per row in early pachytene and then decreased in mid-pachytene, him-3^R93Y germlines had closer to an average of 30 GFP::MSH-5 foci per row in early pachytene. Moreover, the average number of GFP::MSH-5 foci per row increased further in mid and late pachytene, achieving peaks closer to an average of 35 foci (Figure 4D). These very high numbers of GFP::MSH-5 were striking but do not produce an increased number of crossovers in htp-3^H96Y or him-3^R93Ysingle mutants (Russo et al., 2023), raising the possibility that not all of them reflect functional, crossover-specific intermediates, similar to what has been observed in the absence of synapsis (Woglar & Villeneuve, 2018). These data also suggest that these mutations, which disrupt the ability of meiotic HORMADs to adopt the “closed” conformation (Russo et al., 2023), affect the behavior of GFP::MSH-5 on meiotic chromosomes.

Quantification of GFP::MSH-5 foci in pch-2;htp-3^H93Ydouble mutants was similar to htp-3^H93Y single mutants, suggesting that PCH-2’s effect on the behavior of GFP::MSH-5 foci was not through its regulation of HTP-3 (Supplemental Figure 1). In stark contrast to pch-2;htp-3^H93Ydouble mutants, the pch-2;him-3^R93Y double mutant showed a dramatic decrease in the overall average of MSH-5 foci per row throughout the germline (Figure 4C, 4D). These averages fell between the averages in control and pch-2 mutant animals, particularly in the transition zone and late pachytene (Figure 4D), consistent with our previous report that this double mutant has fewer defects in crossover formation than either single mutant (Russo et al., 2023). These data indicate that PCH-2’s effect on promoting the removal of GFP::MSH-5 in both the transition zone and mid to late pachytene is through its regulation of HIM- 3.

pch-2 single mutants exhibit a loss of crossover assurance (Deshong et al., 2014). However, our quantification of GFP::MSH-5 in pch-2 single mutants in mid to late pachytene showed an increased number of crossover-eligible intermediates (Figure 4A, 4B), suggesting a complex relationship between having too many crossover-eligible intermediates and crossover assurance in C. elegans. To address this inconsistency, we generated strains with both OLLAS::COSA-1 (Janisiw et al., 2018) and GFP::MSH-5 in wildtype animals and pch-2 mutants. Consistent with our and others’ previous findings, we observe a gradual reduction in the average number of GFP::MSH-5 foci per row of nuclei in control animals as these nuclei approach the end of pachytene, ultimately converging with the average number of OLLAS::COSA-1 foci per row (Figure 4E, 4F) (Janisiw et al., 2018; Woglar & Villeneuve, 2018). In addition, we consistently observed co-localization of GFP::MSH-5 and OLLAS::COSA-1 foci in control animals, with a few MSH-5 foci persisting without COSA-1 (Figure 4E, 4F).

By contrast, we detected higher average numbers of GFP::MSH-5 foci per row across all late pachytene nuclei in pch-2 mutants and this average never converged upon the average number of OLLAS-1::COSA-1 foci per row (Figure 4E, 4F). To test whether there was a correlation between the number of GFP::MSH-5 foci and the loss of crossover assurance in pch-2 mutants, we determined the average number of GFP::MSH-5 foci in meiotic nuclei with 6 OLLAS::COSA-1 foci and those with less than 6 OLLAS::COSA-1 foci in both control and pch-2 mutant animals (Figure 4E, 4G, Supplemental Figure 2). Indeed, we observed that pch-2 mutant nuclei with less than 6 OLLAS::COSA-1 foci had significantly higher numbers of GFP::MSH-5 foci, compared to both control and pch-2 mutant nuclei with 6 OLLAS::COSA-1 foci (Figure 4E, 4G), suggesting that the inability to reduce the number of crossover-eligible intermediates in pch-2 mutants, counterintuitively, contributes to the observed loss of crossover assurance.

PCH-2 is removed from the synaptonemal complex when crossovers are designated

We have previously shown that PCH-2 localization to the SC is extended when there are partial defects in synapsis or changes in karyotype, producing an increase in crossovers and a loss of crossover interference (Deshong et al. 2014; Patel et al., 2023). These data led us to hypothesize that PCH-2’s presence on chromosomes promotes crossover formation. However, our quantification of GFP::MSH-5 indicates that PCH-2 is required to limit the number of crossover-eligible intermediates, in direct contrast to our proposed hypothesis (Figure 4). Therefore, we decided to revisit what role PCH-2 localization to the SC might play in regulating crossover formation. To this end, we localized PCH-2 in dsb-2 mutants, in which DSB formation is substantially reduced and fewer crossovers are formed in C. elegans (Rosu et al., 2013). We performed this experiment because, in budding yeast, similar mutants that reduce DSB formation rely on Pch2 to successfully complete meiosis (Joshi et al., 2009; Zanders & Alani, 2009).

Because the recombination defect in dsb-2 mutants worsens with age (Rosu et al., 2013), we looked at PCH-2 localization in both young (24 hours post L4 larval stage) and older animals (48 hours post L4). We observed a unique and striking localization pattern in dsb-2 mutants that we had not observed before. When stained for PCH-2 and GFP::COSA-1, nuclei retain PCH-2 onto chromosomes far into late pachytene, past the normal region when PCH-2 typically is removed from chromosomes (Figure 5A, 5B). However, unlike what we have observed in other mutants that have defects in synapsis or changes in karyotype, this retention of PCH-2 is not uniform among all late pachytene nuclei in dsb-2 mutants. Instead, most nuclei lose PCH-2 localization in late pachytene while some retain it. We tested whether there was a relationship between the retention of PCH-2 and the number of GFP::COSA-1 foci and found that 96% of nuclei that lose PCH-2 have at least 1 GFP::COSA-1 focus (Figure 5A, 5B). In older animals (see 48 hours post-L4), this pattern was even more clear: 77% of nuclei without PCH-2 had 1 or more GFP::COSA-1 focus and 93% of meiotic nuclei without a GFP::COSA-1 focus retained PCH-2.

Figure 5.

These data indicate that PCH-2’s removal from the SC is in response to, coincident with or facilitates crossover designation. To distinguish between these possibilities, we analyzed GFP::COSA-1 foci in dsb-2::AID (Zhang et al., 2018) and dsb-2::AID;pch-2 worms, reasoning that if PCH-2’s removal is in response to or coincident with designation, we should not detect any differences in the number of GFP::COSA-1 foci in dsb-2::AID and dsb-2::AID;pch-2 worms. We performed these experiments with the auxin-inducible degradation system because of variability in the dsb-2 mutant background that affected reproducibility of our experiments.

Upon auxin treatment of dsb-2::AID worms, we observed a statistically significant (p value < 0.0001, Mann-Whitney U test) decrease in the number of GFP::COSA-1 foci, in comparison with ethanol- treated worms, verifying that we can reliably knock down DSB-2 with the AID system (Figure 5D).

When we performed the same experiment in dsb-2::AID;pch-2 worms, the average number of GFP::COSA-1 foci further decreased, indicating that fewer crossovers are designated when DSBs are reduced and PCH-2 is absent (Figure 5C, 5D). These data suggest that PCH-2’s presence on the SC prevents crossover designation. Given the correlation between elevated GFP::MSH-5 foci and the loss of crossover assurance we observe in pch-2 mutants, we propose that PCH-2 is retained on meiotic chromosomes to ensure that extra crossover-eligible intermediates are removed and crossover designation is delayed until crossover assurance can be guaranteed in C. elegans.

PCH-2 and high CHK-2 activity control the fate of early double strand breaks

In C. elegans, meiotic cell cycle entry and progression depends on the activity of CHK-2, the meiosis- specific ortholog of the DNA-damage kinase Chk2/CHEK2 (Baudrimont et al., 2022; Castellano-Pozo et al., 2020; Kim et al., 2015; MacQueen & Villeneuve, 2001; Zhang et al., 2023). Meiotic nuclei in leptotene/zygotene are characterized by high CHK-2 activity, which drops to intermediate activity in mid-pachytene (Kim et al., 2015; Zhang et al., 2023). In late pachytene, CHK-2 activity is inactivated by the recruitment of polo-like kinase, PLK-2, to the SC, which enables crossover designation (Zhang et al., 2023). Thus, CHK-2 activity also has consequences for the progression of meiotic recombination and DNA repair.

Given that early DSBs are prevented from becoming crossovers by PCH-2 in early meiotic prophase, we wanted to test if high CHK-2 activity in leptotene/zygotene also contributed to the fate of these early DSBs. To evaluate this possibility, we used syp-1^T452A mutants (Sato-Carlton et al., 2018). The SC component, SYP-1, is phosphorylated by CDK-1 on T452, producing a Polo box binding motif that recruits PLK-2 to the SC, contributing to the inactivation of CHK-2 when chromosomes are synapsed (Brandt et al., 2020; Zhang et al., 2023). Therefore, in syp-1^T452A mutants, CHK-2 activity remains high throughout most of meiotic prophase in the C. elegans germline (Figure 6A).

Figure 6.

We quantified the total number of GFP::COSA-1 foci in late pachytene nuclei in syp-1^T452A mutants and observed a significant decrease in the average number of GFP::COSA-1 foci when compared to control animals (Figures 6B and 6C), consistent with previous findings that syp-1^T452A mutants delay designation and have fewer crossovers (Zhang et al., 2023). These data are also consistent with the possibility that some DSBs fail to become crossovers when CHK-2 activity remains high. If this hypothesis is correct, we predict that the combination of high CHK-2 activity and loss of pch-2 should produce more crossovers. We generated pch-2;syp-1^T452A double mutants to test this hypothesis.

When we quantified GFP::COSA-1 in pch-2;syp-1^T452A double mutants, we detected a significant increase in the average number of GFP::COSA-1 foci compared to syp-1^T452Asingle mutants (Figure 6C), in strong support of our hypothesis.

To further verify that the pch-2 mutation suppresses the crossover defect in syp-1^T452Amutants, we also monitored bivalent formation in diakinesis nuclei. syp-1^T452A single mutants exhibit an average of 7.12 DAPI staining bodies in diakinesis, higher than both control animals and pch-2 single mutants (Figure 6D). By contrast, pch-2;syp-1^T452A mutants had an average of 6.05 DAPI staining bodies (Figure 6D, 6E), directly supporting our COSA-1 analysis and indicating that both PCH-2 function and high CHK-2 activity collaborate to control the fate of DSBs and prevent some of them from becoming crossovers in early meiotic prophase.

Discussion

We have shown that PCH-2 antagonizes crossover formation throughout meiotic prophase and that this regulation occurs through one of the three essential meiotic HORMADs in C. elegans, HIM-3 (Figure 4). We propose that PCH-2 remodels HIM-3 on meiotic chromosomes to destabilize crossover-eligible intermediates, thus limiting which DSBs will become crossovers (Figure 7). However, this antagonism has different consequences depending on when during meiotic prophase it occurs, underscoring the importance of temporal regulation of these events. During leptotene/zygotene, when CHK-2 activity is high (Kim et al., 2015; Zhang et al., 2023) and PCH-2 is present as foci on chromosomes (Deshong et al., 2014), PCH-2 prevents crossover formation at some sites of initial DSB formation and early homolog interactions (Figure 7B). In this way, PCH-2 promotes a wider distribution of crossovers across the genome (Figure 1). In pachytene, when CHK-2 activity has decreased (Kim et al., 2015; Zhang et al., 2023) and PCH-2 is localized to the SC (Deshong et al., 2014), PCH-2 winnows crossover-eligible intermediates, ensuring their reinforcement, designation and crossover assurance (Figure 7B). When there are defects in recombination, such as partial synapsis (Deshong et al., 2014), changes in karyotype (Patel et al., 2023) or too few DSBs (Figure 5), PCH-2 persists on the SC to prevent designation, guarantee crossover assurance and some degree of homeostasis, independent of an additional feedback mechanism that increases DSB formation (Patel et al., 2023). That persistence of PCH-2 also disrupts crossover interference in two of these situations (Deshong et al., 2014; Patel et al., 2023) strongly suggests that crossover interference is mechanistically linked to assurance and homeostasis.

Figure 7.

him-3^R93Y;pch-2 double mutants have stronger crossover assurance than either single mutant but less than wildtype animals (Russo et al., 2023), a phenotype which can now be explained by the behavior of GFP::MSH-5 foci in these double mutants (Figure 4C, 4D). We have shown that HIM-3^R93Y mutant protein can adopt the closed conformation and loads on meiotic chromosomes similar to wildtype HIM- 3 (Russo et al. 2023). In vitro analysis shows that HIM-3^R93Y binds its closure motif with reduced affinity, likely affecting its ability to adopt the closed conformation in vivo (Figure 7A and Russo et al. 2023). The proposed role of PCH-2 in meiosis is to remodel meiotic HORMADs from the closed conformation to the extended one (Figure 7A), regulating their association with chromosomes (Bhalla, 2023). However, since meiotic HORMADs are not visibly depleted from chromosomes during meiotic progression in C. elegans (Couteau et al., 2004; Couteau & Zetka, 2005; Goodyer et al., 2008; Martinez-Perez & Villeneuve, 2005), this genetic interaction supports a role for PCH-2 in temporarily reducing the occupancy of meiotic HORMADs on meiotic chromosomes, destabilizing interactions with partner proteins that modulate the progression and fidelity of meiotic recombination (Russo et al., 2023). Thus, PCH-2’s remodeling of HIM-3 would disrupt protein-protein interactions that underlie crossover-eligible intermediates, destabilizing them and reducing their number on chromosomes, contributing to crossover control (Figure 7B). Since meiotic HORMADs are essential for meiotic chromosome axis structure and function, our data therefore support a role for the meiotic axis in crossover control, as suggested by previous reports (Chu et al., 2024; Girard et al., 2023; Lambing et al., 2020; Nabeshima et al., 2004). However, the disassembly of crossover-eligible intermediates would also release pro-crossover factors present at these sites to concentrate at other, more stable sites on the SC, contributing to reinforcement and designation during synapsis (Girard et al., 2023; Yokoo et al., 2012). Moreover, these results raise the possibility that this behavior of PCH-2 and/or meiotic HORMADs might be regulated by post-translational modifications associated with crossover control, such as ubiquitination, SUMOylation and phosphorylation (Gray & Cohen, 2016).

Given that mutation of PCH-2 produces changes in the number and distribution of crossovers across multiple model systems, we argue that controlling crossover distribution and number is the conserved role of PCH-2 in meiosis (Bhalla, 2023). We have previously proposed that the conserved role of PCH- 2 is to coordinate meiotic recombination with synapsis (Bhalla, 2023). With these results, we further refine this model. Early in meiotic prophase, PCH-2 remodels meiotic HORMADs to prevent some DSBs from becoming crossover-eligible intermediates, widening the recombination landscape beyond early homolog interactions and/or sites that tend to be more favorable for DSB formation, also known as ‘hot spots.” For example, this may explain the localization of TRIP13, the PCH-2 ortholog in mice, to telomeres (Chotiner et al., 2024), sites that experience early homolog interactions due to the organization of meiotic chromosomes in the bouquet formation (Scherthan et al., 1996). This antagonism may also expand the regions of the genome that initiate synapsis in organisms that use DSBs to accomplish this event. This possibility is supported by the observation that loss of TRIP13 in mammals produce meiotic chromosomes that exhibit partial asynapsis (Roig et al., 2010), particularly near regions that may act as barriers to SC polymerization (Brown et al., 2005; Roig et al., 2010).

Limiting which DSBs becomes crossover-eligible intermediates in early meiotic prophase also ensures that meiotic recombination overlaps with synapsis, either completely, as in C. elegans (Yokoo et al., 2012) or partially, as in budding yeast, plants and mice (Capilla-Perez et al., 2021; Cole et al., 2012; Joshi et al., 2015; Morgan et al., 2021). Once synapsis is complete, PCH-2 continues to remodel meiotic HORMADs on chromosomes to control the gradual implementation of crossover number and distribution, reinforcing the important role that synapsis plays in mediating crossover control (Durand et al., 2022; Libuda et al., 2013).

Unexpectedly, the inability to reduce the number of crossover-eligible intermediates in pch-2 mutants, as visualized by GFP::MSH-5 foci, does not produce extra crossovers but a loss of crossover- assurance in C. elegans (Figure 4), a somewhat counterintuitive result. One interpretation of these data is that crossover-eligible intermediates may be more numerous but absent from some chromosomes in pch-2 mutants, explaining the loss of crossover assurance. Since the absence of crossover intermediates in C. elegans is accompanied by premature desynapsis of individual chromosomes (Machovina et al., 2016; Pattabiraman et al., 2017) and chromosomes in pch-2 mutants delay desynapsis (Deshong et al., 2014), we do not favor this interpretation. Instead, we propose that having too many crossover-eligible intermediates can be as deleterious to crossover assurance as having too few (Figure 7B). This possibility is further supported by the loss of crossover assurance we detect in irradiated wildtype worms, which is exacerbated in pch-2 mutants (Figure 2).

This phenomenon, where crossover-eligible intermediates need to be winnowed to some threshold number to ensure crossover assurance, may also explain the loss of crossover assurance also observed in Trip13 deficient mice (Roig et al., 2010) and on small chromosomes in budding yeast (Chakraborty et al., 2017). Alternatively, the counterintuitive relationship between the number of crossover-eligible precursors and crossover assurance in pch-2 mutants we observe might reflect an additional layer of regulation during crossover formation specific to C. elegans. Since C. elegans chromosomes are holocentric, crossovers play an additional role organizing chromosomes for the ordered release of sister chromatid cohesion during meiosis I (Martinez-Perez et al., 2008; Nabeshima et al., 2005) and extra crossovers can be deleterious to accurate chromosome segregation (Hollis et al., 2020). By contrast, in Arabidopsis, a system that appears to be able to tolerate an extraordinarily high number of crossovers with little to no effect on chromosome segregation (Durand et al., 2022), PCH2’s inability to localize to the SC produces an increase in crossover formation (Figure 7), as visualized by both MLH1 foci and the formation of chiasmata (Yang et al., 2022). Once again, an overarching theme that becomes apparent in our model is that PCH-2 may play a common role in different systems, with dramatic variations in phenotypic consequences given species-specific requirements and constraints.

We were not surprised to see high numbers of double crossovers on almost every chromosome in our genetic analysis of recombination in wildtype worms, given our previous analysis (Deshong et al., 2014). However, we were surprised to see that the majority of them were found near Pairing Centers and sites of synapsis initiation, suggesting a relationship between early homolog interactions and the formation of double crossovers. When we revisited our previous data, we observed similar patterns. In addition, we do not detect these double crossovers cytologically in C. elegans, even when using the OLLAS::COSA-1 reporter, which has been reported to identify double crossovers in spermatogenesis not visualized by GFP::COSA-1 (Cahoon et al., 2023). Crossovers that are cytologically marked by COSA-1 are known as Class I crossovers, which depend on pro-crossover factors such as MSH-5 and ZHP-3, and exhibit crossover control (Gray & Cohen, 2016). These data raise the intriguing possibility that these double crossovers are the product of the alternate, Class II, pathway of crossover formation, which relies on a different suite of proteins, does not respond to crossover control and contributes to varying degrees in different model systems (Gray & Cohen, 2016; Youds et al., 2010). Thus, important corollaries of our model may be that PCH-2 specifically coordinates recombination with synapsis to promote Class I crossovers, limit Class II crossovers and that Class II crossovers are more likely to form early in meiosis, prior to synapsis. Therefore, variations in the contribution of the Class II crossover pathways to crossover recombination and the degree of cross-talk between Class I and Class II pathways among model systems might reflect the degree to which crossover formation overlaps with synapsis (Bhalla, 2023; Gray & Cohen, 2016; Yokoo et al., 2012). Furthermore, these corollaries are entirely consistent with PCH-2’s absence from the genome of fission yeast (Wu & Burgess, 2007), a system in which chromosomes do not undergo synapsis, crossovers do not exhibit interference and all crossovers are dependent on the Class II pathway (Hollingsworth & Brill, 2004).

Our results also have some additional important implications about the regulation of DSB formation across the genome in C. elegans. We explain the shift in the recombination landscape away from the central regions of chromosomes and toward PC ends in pch-2 mutants (Figure 1) as a result of early DSBs becoming crossovers at the expense of later DSBs, suggesting that when DSBs happen in meiotic prophase affects where they happen in the genome. Specifically, we propose that chromosome arms, which are gene poor, receive DSBs early during (or even throughout) DSB formation and the center of chromosomes, which are gene rich, receive DSBs later. A similar regulation of the timing of DSB formation has been demonstrated in budding yeast, where the DSB landscape across the whole genome expands when time in prophase is increased (López Ruiz et al., 2024) and small, highly recombinogenic, chromosomes, get more DSBs later in meiotic prophase (Murakami et al., 2020; Subramanian et al., 2019). However, this temporal regulation has not been reported previously in C. elegans and suggests that this phenomenon is more widely conserved. This expansion of the DSB landscape in C. elegans to include the center regions of chromosomes later in meiosis may be a deliberate attempt for recombination to create new haplotypes for evolution to act on, despite the relative paucity of DSBs and the observation that they can result in chromosome missegregation (Altendorfer et al., 2020).

Finally, our work raises some important questions about the functional role(s) of DSBs in meiosis, aside from their contributions to crossover formation. Hicks and colleagues were the first to report that early DSBs do not contribute to crossover formation in C. elegans (Hicks et al., 2022). Here we show that these early DSBs are prevented from becoming crossovers by both PCH-2 activity and cell cycle stage, specifically in leptotene/zygotene, when homologs are initiating pairing and synapsis. In budding yeast, similar, early-occurring DSBs have been characterized as “scout DSBs” because of their preference for repair from sister chromatids, versus homologous chromosomes, and their proposed role in contributing to homolog pairing (Borde & de Massy, 2015; Joshi et al., 2015). The homolog bias that these “scout DSBs” do display seems dependent on budding yeast PCH2 (Joshi et al., 2015) but interpreting this experiment is complicated by the fact that Pch2 is also required to make the budding yeast meiotic HORMAD, Hop1, available for its loading onto meiotic chromosomes (Herruzo et al., 2021).

In contrast to budding yeast, C. elegans does not rely on DSBs to promote homolog pairing and initiate synapsis (Dernburg et al., 1998); in worms, cis-acting sites called Pairing Centers are essential for homolog pairing and synapsis (MacQueen et al., 2005). There is certainly some support in the literature for DSBs playing a role in supporting pairing and synapsis in C. elegans (Guo et al., 2022; Mlynarczyk-Evans et al., 2013; Roelens et al., 2015). However, if these early DSBs were contributing to pairing and synapsis, we would expect to see a genetic interaction between htp-3^H96Y and pch-2 mutations in the installation of GFP::MSH-5 in the transition zone; we previously reported that htp- 3^H96Y suppresses the acceleration of pairing and synapsis of pch-2 mutants, particularly when Pairing Center function is compromised (Russo et al., 2023). Instead, we favor the possibility that these early DSBs are generated to amplify the signaling of DNA damage kinases to prime them for their role in recombination, a role that has also been proposed for “scout” DSBs (Joshi et al., 2015). That ATM-1, a conserved DNA damage kinase that is downstream of CHK-2 in C. elegans, relies on DSBs for full activity, supports this proposed, conserved role (Yu et al., 2023).

Our work provides an important framework to finally understand the role of PCH-2 in controlling the number and distribution of crossovers, a role that we argue is its conserved role. While specific details may vary across systems, we propose that PCH-2 remodels meiotic HORMADs throughout meiotic prophase to destabilize crossover-eligible precursors, coordinating meiotic recombination with synapsis, contributing to the progressive implementation of meiotic recombination and guaranteeing crossover assurance, interference and homeostasis.

Methods and Materials

C.elegans Genetics and Genome Engineering

The C. elegans Bristol N2 was used as the wild-type strain. All strains were maintained at 20°C under standard conditions unless stated. Mutant combinations were generated by crossing. The following mutants and rearrangements were used:

LGII: pch-2(tm1458), meIs8 ([pie-1p::GFP::cosa-1 + unc-119(+)], dsb-2(me96), dsb-2(ie58[dsb- 2::AID::3xFLAG])

LGIII: cosa-1(ddr12[OLLAS::cosa-1]); htp-3(vc75)

LGIV: him-3(blt9), spo-11(ie59[spo-11::AID::3xFLAG]), msh-5[ddr22(GFP::msh-5)], ieSi38 [sun- 1p::TIR1::mRuby::sun-1 3’UTR + Cbr-unc-119(+)]

LGV: syp-1(icm85[T452A]), nT1[qIs51], bcIs39 (Plin-15::ced-1::GFP)

Genetic analysis of Recombination

The wildtype Hawaiian CB4856 strain (HI) and the Bristol N2 strain were used to assay recombination between single nucleotide polymorphisms (SNPs) on Chromosomes I, III, IV and X (Wicks et al., 2001; Bazan et al., 2011). The SNPs, primers, enzymes used for restriction digests and expected fragment sizes are included in Supplemental Table 1. To measure wild-type recombination, N2 males containing bcIs39 were crossed to Hawaiian CB4856 worms. Cross-progeny hermaphrodites were identified by the presence of bcIs39 and contained one N2 and one CB4856 chromosome.

These were assayed for recombination by crossing with CB4856 males containing myo-2::mCherry. Cross-progeny hermaphrodites from the resulting cross were isolated as L4s, and then cultured individually in 96- well plates in liquid S-media complete supplemented with HB101. Four days after initial culturing, starved populations were lysed and used for PCR and restriction digest to detect N2 and CB4856 SNP alleles. For recombination in pch-2 mutants, strains homozygous for the CB4856 background of the relevant SNPs were created, then mated with pch-2; bcIs39. Subsequent steps were performed as in the wild-type worms.

Immunostaining

DAPI staining and immunostaining was performed as in (Russo et al. 2023), 20 to 24 hours post L4 unless otherwise noted. For analyzing bivalents, the same protocol was implemented with the exception that hermaphrodites were dissected and DAPI stained 48 hours post late L4 stage, unless otherwise noted.

The following primary antibodies were used at the indicated dilutions: alpaca anti-GFP Booster (ChromoTek, gb2AF488) was used at 1:1000; rat anti-OLLAS (Invitrogen, PIMA516125) was used at 1:1000; and rabbit anti PCH-2 (Bhalla 2014) was used at 1:500. The following secondary antibodies were used at the indicated dilutions: anti-rabbit Cy3 (Jackson Labs) was used at 1:500 and anti-rat Cy5 (Jackson Labs) was used at 1:500

Irradiation Experiments

Control and pch-2 mutant L4’s were aged 12-14 hours before being exposed to 1,000 rad (10 Gy) of X-ray radiation using a Precision MultiRad 160 X-irradiator (Precision X-Ray Inc.). Germlines were then fixed and stained, 8 hours and 24 hours post irradiation.

Auxin-induced Degradation Experiments

Auxin treatment was performed by transferring young adult worms (aged 12-14 hours post-L4) to bacteria-seeded plates containing auxin or 99% ethanol at specific time points, except for experiments in which L4s were transferred directly to bacteria-seeded plates containing auxin or ethanol (the 36 hour and 48 hour time points in Figures 3C and E and the experiments in Figures 5C and D). The natural auxin indole3-acetic acid (IAA) was purchased from Alfa Aesar (#A10556). A 400 mM stock solution in ethanol was prepared and was stored at 4°C for up to one month. Auxin was diluted to 100mM, and 100ul was spread onto NGM plates. Plates were allowed to dry before seeding with fresh OP50 culture. Plates were left at 20°C for 2-3 days in the dark to allow for bacterial lawn growth.

Imaging and Quantification

All images were acquired using a DeltaVision Personal DV system (Applied Precision) equipped with a 100X N.A. 1.40 oil-immersion objective (Olympus), resulting in an effective XY pixel spacing of 0.064 or 0.040 µm. Three-dimensional image stacks were collected at 0.2-µm Z spacing and processed by constrained, iterative deconvolution. Image scaling, analysis and maximum-intensity projections were performed using functions in the softWoRx software package.

For analysis of GFP::MSH-5 foci, sum projections were generated using ImageJ for each image of the germline. ImageJ plugins Cell Counter, ROI Manager, and Find Maxima were used to identify and quantify GFP::MSH-5 foci by row from transition zone to the end of pachytene. The threshold value was set depending on background conditions to ensure minimal signals were identified. Foci were only quantified if they co-localized with DAPI staining. For all genotypes, three germlines per genotype were analyzed and representative germlines are shown.

Graphing and Statistical Analysis

Data was analyzed using Python 3.8 and Prism for statistical significance. All datasets were tested for normality using the Shapiro-Wilk test. For Figures 1, 2 and 5B, Fisher’s exact test was used to determine significance. For Figures 3, 4, 5D and 6, Mann-Whitney U test was used to determine significance.

Author contributions

B.P., M.G., and N.B. designed the experiments. B.P., M.G., A.H., E.L., and V.O. performed the experiments. B.P. and N.B. analyzed the data. B.P. and N.B. wrote the initial draft of the manuscript and B.P., M.G., A.H. and N.B. revised the manuscript. N.B. acquired funding.

Acknowledgements

We would like to thank Josh Arribere, Pete Carlton, Abby Dernburg, Nicola Silva and Anne Villeneuve for valuable strains and reagents. We would also like to thank the members of the Bhalla lab for careful review of the manuscript. This work was supported by the NIH (grant numbers R35GM141835 [N.B.], R25GM051765 [V.O.] and T34GM140956 (A.H. and V.O.). Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440).