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.

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 and consistent 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.

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. We had previously observed that meiotic nuclei in early prophase were more likely to produce crossovers when DSBs were induced by excision of the Mos transposon in pch-2 mutants than in control animals but experimental caveats limited our ability to properly interpret this experiment (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).

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 late pachytene nuclei that received exogenous DSBs in mid-pachytene, and 24 hours after irradiation, when we monitor crossover formation in late pachytene nuclei that received exogenous DSBs in leptotene/zygotene, also known as the transition zone in C. elegans (Figure 2A). In order to detect crossovers in late pachytene, 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. In unirradiated control worms, 94% of late pachytene nuclei have 6 GFP::COSA-1 foci, one per chromosome pair (Figure 2B). In unirradiated pch-2 mutants, the percentage of nuclei have 6 GFP::COSA-1 foci drops to 82%, primarily because of the significant increase in meiotic nuclei with less than 6 GFP::COSA-1 foci (17%) (Figure 2B), consistent with our previous report that loss of PCH-2 leads to a decrease in crossover formation (Deshong et al., 2014). At 8 hours after irradiation, pch-2 mutants’ defect in crossover assurance becomes even more pronounced. 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 (Figures 2C and 2D). However, we did not observe a significant number of nuclei with greater than 6 COSA-1 foci in either backgrounds. This observation reinforces previous findings that exogenous DSBs introduced in mid-pachytene do not affect the number of crossovers in wildtype animals (Yokoo et al., 2012).

Next, we examined crossover formation 24 hours post-irradiation (Figures 2E and 2F). For this time point, we also detected a significant loss of crossover assurance in pch-2 mutants, compared to control worms. However, in contrast to the 8 hour time point, we also saw a significant increase in nuclei containing less than 6 COSA-1 foci in both control (13%) and pch-2 mutants (26%), compared to unirradiated worms (p values = 0.0008 and < 0.0001, respectively, Fischer’s Exact Test), indicating that even in control animals, extra DSBs in early meiotic prophase can disrupt crossover assurance. In addition, we also observed a significant increase in nuclei with more than 6 COSA-1 foci in pch-2 mutants (wildtype: 2%, pch-2: 13%), demonstrating 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 introduced in early meiotic prophase from becoming crossovers, supporting our recombination analysis (Figure 1).

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). When we treated both control and pch-2 mutant worms for 36 hours with auxin (Figure 3A), 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 (Figure 3D), we observed 12 DAPI stained bodies, or univalents, which are the 6 homolog pairs that have not formed chiasmata (Figure 3F). Thus, we can reliably remove SPO-11 from meiotic nuclei throughout the germline and prevent crossover formation.

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 four different timepoints in early meiotic prophase to determine if acute depletion of SPO-11 leads to any changes in crossover formation between control animals and pch-2 mutants (Figures 3A and D). In this experiment, we assayed crossover formation with two experimental approaches. Similar to our irradiation experiment (Figures 2E and 2F), we assayed crossover formation by analyzing the number of GFP::COSA-1 foci in late pachytene 24, 22, 20 and 18 hours after acutely depleting SPO-11 (Figure 3A). In addition to this approach, we analyzed the presence of bivalents in diakinesis (Figure 3D), which indicates the successful formation of chiasmata between homologous chromosomes. Since meiotic nuclei travel from late pachytene to diakinesis in approximately 12 hours (Deshong et al., 2014), the 36, 34, 32 and 30 hour timepoints to assay bivalent formation correspond to the 24, 22, 20 and 18 hour timepoints, respectively, in which GFP::COSA-1 were analyzed. We also performed a control experiment, treating worms with ethanol for 24 and 36 hour timepoints to ensure that this tagged version of SPO-11 was fully functional for crossover formation (Figures 3C and 3F).

At the 24 hour time point (Figures 3A and 3C), 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 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, when we assayed bivalent formation 12 hours later (see 36 hours auxin treatment in Figures 3D and 3F), we observed the same heterogeneity and 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.

We continued to monitor GFP::COSA-1 and bivalent formation at successive timepoints. At the 22 hour 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 (Figure 3C) and a reduction in the number of DAPI stained bodies (Figure 3F), suggesting that more DSBs are becoming crossovers. In our analysis of GFP::COSA-1 foci at this timepoint, there appeared to a bimodal distribution: the majority of nuclei in both genotypes fell into two clear classes, those with no COSA-1 foci (24% in control animals, 26% in pch-2 mutants) and those with 6 COSA-1 foci (42% in both control animals and pch-2 mutants) (Figure 3C). The number of crossovers continued to increase 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 (Figures 3C and 3F).

At the 22 hour and 34 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 (Figures 3C and 3F), 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 (Figure 3C), 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 (Figure 3F), indicating that more homolog pairs had successfully formed chiasmata in pch-2 mutants. No difference in GFP::COSA-1 foci or bivalent formation was observed at the 18 hour and 30 hours timepoints, respectively (Figures 3C and 3F). 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.

After DSB formation, a subset of DSBs are licensed to be repaired through a pro-crossover pathway and eventually winnowed to a stereotypical number of crossovers (Yokoo et al., 2012). In C. elegans, these crossover-eligible intermediates can be visualized by the loading of the pro-crossover factor MSH-5, a component of the meiosis-specific MutSγcomplex that stabilizes crossover-specific DNA repair intermediates called joint molecules (Janisiw et al., 2018; Kelly et al., 2000; Snowden et al., 2004). In the spatiotemporally organized meiotic nuclei of the germline, a functional GFP tagged version of MSH-5, GFP::MSH-5, begins to form a few foci in leptotene/zygotene (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 in a process called designation (Janisiw et al., 2018; Woglar and Villeneuve, 2018; Yokoo et al., 2012; see control in Figures 4A and 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).

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 (Figures 4A and 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 (Figures 4A and 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 (Figures 4A and 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, limiting the formation of crossover-eligible intermediates during early meiotic prophase, consistent with both our genetic analysis of recombination (Figure 1) and PCH-2’s 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 control different aspects of meiosis. A mutant version of the essential meiotic HORMAD, HIM-3^R93Y, binds its closure motif with reduced affinity in vitro and we have proposed that an analogous mutation in another essential meiotic HORMAD, HTP-3^H96Y, behaves similarly (Russo et al., 2023). These mutations suppress different meiotic defects in pch-2 mutants in vivo, indicating that 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, and the potential role for these early crossover-eligible intermediates, 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^R93Ymutants 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^R93Y single 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 and Villeneuve, 2018). These data also suggest that these mutations 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^H93Y double 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 (Figures 4C and 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, as visualized by GFP::COSA-1 foci (Figure 2B and 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 (Figures 4A and 4B), suggesting a complex relationship between having too many crossover-eligible intermediates and crossover assurance in C. elegans. To address this inconsistency, we generated wildtype animals and pch-2 mutants with both GFP::MSH-5 and a version of COSA-1 that has been endogenously tagged at the N-terminus with the epitope tag, OLLAS (Janisiw et al., 2018), a fusion of the E. coli OmpF protein and the mouse Langerin extracellular domain. 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 (Figures 4E and 4F; Janisiw et al., 2018; Woglar and 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, consistent with crossover designation (Figures 4E and 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::COSA-1 foci per row (Figures 4E and 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 (Figures 4E and 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 (Figures 4E and 4G), suggesting that the inability to reduce the number of crossover-eligible intermediates in pch-2 mutants, counterintuitively, prevents designation on some chromosomes and contributes to the observed loss of crossover assurance.

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 and 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 (Figures 5A and 5B). However, unlike what we have observed in other mutants that have defects in synapsis or changes in karyotype (Deshong et al., 2014; Patel et al., 2023), 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 (Figures 5A and 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.

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 (Figures 5C and 5D). These data argue against the possibility that PCH-2’s removal from the SC is simply in response to or coincident with crossover designation and instead, suggest that PCH-2’s removal from the SC somehow facilitates crossover designation and assurance. 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.

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 and 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 cell cycle kinase, CDK-1, on T452, producing a Polo box binding motif that recruits the polo-like kinase, 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, delaying meiotic progression as visualized by the extension of the transition zone (Figure 6A, Supplemental Figure 3).

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. These double mutants exhibited the same delay in meiotic progression as syp-1^T452A single mutants (Supplemental Figure 3). 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^T452A mutants, 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 and 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.

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:

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 (Bazan and Hillers, 2011; Wicks et al., 2001). 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 by backcrossing pch-2 mutants to worms of the CB4856 background at least eight times and verifying the presence of Hawaiian SNPs on all chromosomes tested in the recombination assay. These Hawaiianized pch-2 mutants were then mated with pch-2; bcIs39. Subsequent steps were performed as in the wild-type worms.

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 (Deshong et al., 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

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 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 Figure 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.

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 and meiotic progression, 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.

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.