ATM signaling modulates cohesin behavior in meiotic prophase and proliferating cells

Cohesins are ancient and ubiquitous regulators of chromosome architecture and function, but their diverse roles and regulation remain poorly understood. During meiosis, chromosomes are reorganized as linear arrays of chromatin loops around a cohesin axis. This unique organization underlies homolog pairing, synapsis, double-stranded break induction, and recombination. We report that axis assembly in Caenorhabditiselegans is promoted by DNA-damage response (DDR) kinases that are activated at meiotic entry, even in the absence of DNA breaks. Downregulation of the cohesin-destabilizing factor WAPL-1 by ATM-1 promotes axis association of cohesins containing the meiotic kleisins COH-3 and COH-4. ECO-1 and PDS-5 also contribute to stabilizing axis-associated meiotic cohesins. Further, our data suggest that cohesin-enriched domains that promote DNA repair in mammalian cells also depend on WAPL inhibition by ATM. Thus, DDR and Wapl seem to play conserved roles in cohesin regulation in meiotic prophase and proliferating cells. The authors show that the assembly of the meiotic chromosome axis in worms depends on activation of the master DNA-damage response kinase ATM, which leads to destabilization of the cohesin-unloader WAPL. Similar ATM-dependent WAPL inhibition also occurs in cohesin-rich genomic regions upon DNA-damage induction.

Three-dimensional chromosome architecture is strongly influenced by cohesins. The core cohesin complex comprises four proteins: a heterodimer of two large ATPases, the structural maintenance of chromosomes (SMC) proteins Smc1 and Smc3; a HEAT repeat protein known as sister chromatid cohesion 3/stromal antigen (Scc3/SA/Stag); and an alpha-kleisin protein. Cohesins are related to condensins and the SMC5/6 complex, and homologous SMC complexes are found across all domains of life 1,2 .
Cohesin binding during DNA replication establishes sister chromatid cohesion, which is required for accurate chromosome segregation during mitosis and meiosis. Cohesins are also molecular motors that move along chromatin, forming loops that govern chromosome topology. They contribute to transcriptional regulation, chromosome condensation, and repair of DNA damage. The diverse chromosome architectures observed in different cell types are presumed to be modulated by cohesin subunits and additional regulatory factors, but how these factors collaborate to shape chromosomes in distinct contexts remains mysterious.
Remodeling of meiotic chromosomes to form an axis-loop structure is thought to be driven in part by expression of meiosis-specific cohesin subunits. All eukaryotes studied to date express one or more meiosis-specific kleisins, and some also have meiosis-specific SMC and/or Scc3/Stag isoforms 10,13,18 . However, it is largely unknown how the activities of meiotic cohesins differ from those in other cells, except that the Rec8 kleisin can be selectively protected from cleavage to keep sister chromatids together during the first meiotic division.
WAPL-1 reaccumulates in oocyte nuclei during later stages of meiotic prophase and contributes to removing COH-3/4 cohesin from the axes as chromosomes condense during diplotene and diakinesis 31 . However, this role is nonessential; wapl-1-null mutants are viable and fertile, with normal meiotic segregation, as indicated by a low production of male progeny, which arise through X chromosome nondisjunction and are diagnostic for meiotic errors 31,32 , in broods of self-fertilizing hermaphrodites. They do show hallmarks of reduced mitotic fidelity, including some embryonic and larval lethality and low-penetrance egg-laying and locomotion defects 32 .
WAPL-1 contains four consensus phosphorylation sites for CHK-2 (R-X-X-S/T). However, mutation of all four potential sites to nonphosphorylatable residues (wapl-1 4SA ) did not affect the localization of WAPL-1 in proliferating or meiotic nuclei (Extended Data Fig. 1a-d). Thus, we considered the possibility that regulation of WAPL-1 by CHK-2 might be indirect. We tested whether WAPL-1 downregulation requires the formation of meiotic DSBs, synapsis, and/or homologous pairing, three distinct meiotic events that depend on CHK-2 activity in early prophase 17,35,36 . Loss-of-function mutations 32 or auxin-induced depletion of SPO-11, which is essential for meiotic DSBs, did not alter WAPL-1 localization, indicating that WAPL-1 downregulation is independent of DSBs (Extended Data Fig. 1e-g). WAPL-1 localization was also unaffected when we depleted SYP-3, an essential component of the synaptonemal complex (SC), although chromosome synapsis was disrupted, confirming that depletion was effective (Extended Data Fig. 1e-g) 37,38 . Similarly, WAPL-1 localization was unaffected by co-depletion of PLK-1 and PLK-2, two orthologs of mammalian PLK1 that play partially overlapping roles in chromosome pairing and synapsis during early prophase 32,39,40 .
Previous studies in mammalian cells have shown that ATM promotes genome-wide enhancement of cohesin binding in response to ionizing radiation (IR)-induced DNA damage 41 . DDR signaling also governs localized cohesin binding at DNA-damage loci in budding yeast and mammalian cells [42][43][44][45] . Moreover, WAPL has been identified as a target of ATM and ATR in Arabidopsis 46 . We found that depletion of ATM-1 on chromatin. These include the 'loading' complex Scc2-Scc4, the acetyltransferase Eco1/Ctf7, the 'release factor' Wapl, and Pds5. Wapl destabilizes cohesin binding to chromosomes by a mechanism that is independent of kleisin cleavage. This activity is inhibited by Eco1-mediated acetylation of Smc3 but is promoted in some contexts by Pds5 (refs. [19][20][21][22][23][24][25][26][27]. Here, we investigate the mechanism of chromosome remodeling during the early meiotic (EM) prophase in C. elegans. This nematode expresses two types of meiotic kleisins: REC-8 and COH-3 and COH-4 (COH-3/4). COH-3/4 are closely related paralogs with overlapping roles and are thus regarded as a single type of kleisin 12 . REC-8 and COH-3/4 are essential for homolog pairing and synapsis but have distinct roles (Fig. 1a). REC-8 is expressed in premeiotic (PM) germ cells and is thus present during DNA replication 12 ; REC-8 cohesins mediate sister chromatid cohesion (SCC) that prevents inter-sister recombination and synapsis 28 . COH-3/4 are expressed only after replication; these complexes likely create chromatin loops that emanate from the axis and are more abundant than REC-8 cohesins 12,15,28,29 . No other meiosis-specific cohesin proteins have been identified in C. elegans. During most of meiotic prophase, cohesin complexes containing REC-8 and COH-3/4 localize along the length of chromosome axes. Following crossover designation at mid-pachytene, the two types of cohesins become enriched on reciprocal 'arms' of the bivalent to mediate two sequential rounds of segregation 12,30 .
Wapl/Rad61 is a widely conserved cohesin regulator that was identified in screens for radiation sensitivity in budding yeast and mitotic defects in Drosophila. Its best-known role is promoting the release of 'arm' cohesion during mitotic prophase. Here, we show that C. elegans WAPL-1 is downregulated by ATM-1 (ataxia telangiectasia mutated, ATM) at meiotic entry. This inhibition promotes or stabilizes COH-3/4 binding along chromosome axes. Surprisingly, we find that ATM-1 is activated at meiotic entry by the CHK-2 (Chk2) kinase, which, together with ECO-1 (Eco1) and PDS-5 (Pds5), preferentially protects cohesin complexes containing REC-8 from the effects of WAPL-1. Together, these findings reveal that constitutive activation of DDR kinases at meiotic entry reshapes the genome through cohesins to promote interhomolog interactions and meiotic recombination. Finally, we extend our observations to show that inhibition of WAPL by ATM promotes cohesin enrichment at sites of DNA damage in proliferating human cells.

CHK-2 leads to downregulation of WAPL-1 at meiotic entry
Immunolocalization of C. elegans WAPL-1 (Wapl) reveals diffuse nuclear localization in most tissues, including the germline. Its localization to chromatin drops abruptly at meiotic entry 31,32 . We were intrigued by this finding because depletion of WAPL from mammalian cells during interphase can lead to formation of 'vermicelli,' linear cohesin chromosome cores that resemble meiotic chromosome axes 33 . This reduction was more pronounced in dissected, immunostained gonads than in Fig. 1 | CHK-2 is required for downregulation of WAPL-1 at meiotic entry. a, Diagram of a C. elegans gonad containing proliferating germline stem cells and meiotic nuclei. The dashed line between red and blue nuclei indicates the boundary between PM and EM cells. Important meiotic events and the behaviors of REC-8 and COH-3/4 cohesins in each stage are shown. b, WAPL-1 immunostaining in the distal tip region of gonads, showing the persistence of WAPL-1 on meiotic chromosomes following CHK-2 depletion. Here and elsewhere, 'control AID ' indicates treatment of a negative control strain with auxin in parallel with the experimental strain(s). The control strain is isogenic except that it lacks any degron-tagged genes, that is, it contains the same TIR1 transgene as the experimental strain(s). Red and blue rectangles outline regions containing PM germline stem cell and EM nuclei, respectively, which are enlarged in c. Scale bar, 10 µM. Meiotic nuclei but not PM nuclei display chromosomes marked by the HORMA domain protein HTP-3 (magenta). In the merged images, WAPL-1 (W) is shown in green and HTP-3 (H) is shown in magenta. Scale bar, 2 µM. d, Quantification of WAPL-1 immunostaining in b. Lower and upper box ends represent the first and third quartiles, with the median indicated by the horizontal line within the box. All data points are shown, and the sample sizes are indicated. ****P < 0.0001 (two-sided Wilcoxon-Mann-Whitney test, adjusted by Bonferroni correction). a.u., arbitrary units. e, COH-3/4 and REC-8 immunolocalization (in green) in early pachytene nuclei, showing that WAPL-1 negatively regulates axial COH-3/4, but not REC-8, upon CHK-2 depletion. Scale bar, 2 µM. f,g, Quantification of the intensity of COH-3/4 and REC-8 immunostaining in auxin-treated animals of the indicated genotypes. e. Lower and upper box ends represent the first and third quartiles, with the median indicated by the horizontal line within the box. All data points are shown, and the sample sizes are indicated. ****P < 0.0001 (two-sided Wilcoxon-Mann-Whitney test, adjusted by Bonferroni correction). See 'Data presentation' for more details.

CHK-2 positively regulates ATM-1 activity
In the canonical DDR pathway in mammalian cells, CHK2 is an essential downstream transducer of ATM activity 49,50 . However, previous studies have found that C. elegans CHK-2 is dispensable for checkpoint activation in response to hydroxyurea and ionizing radiation in embryos and the adult germline, and is essential only for meiosis 35,36 . Like other meiosis-specific CHK2 orthologs, C. elegans CHK-2 lacks the amino-terminal SCD that mediates activation by ATM (Fig. 3a). Additionally, CHK-2-dependent phosphorylation of the nuclear envelope protein SUN-1 was observed in the absence of ATM-1 and ATL-1 (ref. 51 ). We further found that pairing center proteins HIM-8 and ZIM-1, ZIM-2, and ZIM-3, which are direct targets of CHK-2 (ref. 17 ), were phosphorylated when we depleted ATM-1, ATL-1, or both proteins (Fig. 3b,c), confirming that CHK-2 activity is independent of ATM-1 and ATL-1.
Since WAPL-1 persists on chromatin following depletion of either CHK-2 or ATM-1, we tested whether ATM-1 activity might depend on CHK-2. The intensity of immunostaining using an antibody against phosphorylated SQ/TQ (pS/TQ) 52 , which recognizes ATM/ATR substrates, was greatly reduced in meiotic nuclei following depletion of either CHK-2 or ATM-1 (Fig. 3d).
Formation of meiotic DSBs in C. elegans requires CHK-2 (refs. 35,53,54 ). We tested whether ATM-1 activity depends on DSBs by depleting the SPO-11 endonuclease, and found that this also reduced pS/TQ immunofluorescence, albeit less so than depletion of ATM-1 (Figs. 3d,e). This is consistent with our evidence that SPO-11 depletion does not affect WAPL-1 downregulation at meiotic entry (Extended Data Fig. 1a-c). Together, these results indicate that CHK-2 promotes basal levels of ATM-1 activity in the absence of meiotic DSBs.

A CHK-2 consensus site is essential for ATM-1 activity
Previous studies have shown that high concentrations of Chk2 can lead to self-activation in vitro and phosphorylation of H2AX and other S/T-Q sites in vivo 55,56 , suggesting that Chk2 may promote ATM/ATR activity under certain conditions. We aligned the amino acid sequences of ATM family proteins and found a conserved CHK-2 consensus phosphorylation motif (R-X-X-S/T) within their FAT domains, which are critical for ATM/ATR activation ( Fig. 4a) [57][58][59] . A mutation in this motif was identified in a person with leukemia with ATM deficiency 60 .
Mutation of the arginine and serine residues in this motif (atm-1 KA ) greatly reduced pS/T-Q immunofluorescence in meiotic nuclei (Fig. 4a-c and Extended Data Fig. 3), while the corresponding phosphomimetic mutation did not appear to alter ATM-1 activity (Fig. 4a,b). Additionally, phosphomimetic ATM-1 S1853D showed detectable, albeit reduced, activity, even when CHK-2 was depleted ( Fig. 4b-d). Depletion of either CHK-2 or SPO-11 in animals expressing ATM-1 S1853D resulted in similar levels of anti-pS/TQ immunofluorescence (Fig. 4e,f). Together, these observations suggest that break-independent phosphorylation of S1853 by CHK-2 promotes a basal level of ATM-1 activity, and DSBs further elevate its activity.
Together, these findings indicate that the persistence of WAPL-1 on meiotic chromosomes upon CHK-2 depletion may be a consequence of a failure to activate ATM-1. To further test this idea, we examined WAPL-1 immunostaining in atm-1 KA and atm-1 S1853D , alleles that showed defective and normal kinase activity, respectively, in early meiosis. CHK-2 showed normal activity in both cases (Fig. 4b-d). Animals expressing only the nonphosphorylatable atm-1 KA allele showed aberrant WAPL-1 staining, similar to that seen following depletion of CHK-2 or ATM-1 ( Fig. 5a-c). By contrast, the phosphomimetic atm-1 S1853D allele resulted in normal downregulation of WAPL-1, even when CHK-2 was depleted ( Fig. 5a-c). Moreover, nonphosphorylatable WAPL-1 2A persisted on EM chromosomes in atm-1 S1853D , indicating that ATM-1 S1853D regulates WAPL-1 through its N-terminal mini-SCD (Extended Data Fig. 5d,e). We tested whether atm-1 S1853D could bypass the requirement for CHK-2 in DSB induction, and found that RAD-51 foci were absent, indicating that CHK-2 promotes breaks independently of ATM activation (Extended Data Fig. 4a,b), which also validated the depletion efficacy of SPO-11.
PDS-5 (also known as EVL-14) is essential for gonad development in C. elegans, presumably owing to its mitotic functions 77 . The protein localizes to nuclei throughout the germline and is enriched along chromosome axes in meiotic nuclei (Extended Data Fig. 6a,b). Its abundance Article https://doi.org/10.1038/s41594-023-00929-5 on axes was not affected by loss of CHK-2 (Extended Data Fig. 6a-c), despite the reduction in COH-3/4 (above), supporting the idea that PDS-5 preferentially associates with REC-8 cohesin.
control AID (120)   The HORMA domain protein HTP-3 can be recruited to axes in the absence of either REC-8 or COH-3/4 cohesin, but not both 12 . Co-depletion of CHK-2 and PDS-5 resulted in loss of HTP-3 from axes ( Fig. 7g-i), corroborating the conclusion that both classes of cohesin were severely disrupted 12 . Localization of REC-8 and HTP-3 was restored by co-depletion of WAPL-1 (Fig. 7g-i). Together, these results reveal that downregulation of WAPL-1 and protection of REC-8 by PDS-5 are parallel pathways that contribute to cohesin stability and axis assembly at meiotic entry. Acetylation of cohesins by ECO-1 likely also contributes to axis formation, but this is only evident when WAPL-1 cannot be properly downregulated.

Cohesin enrichment at DNA damage foci is regulated by WAPL
Our identification of a conserved mini-SCD in Wapl homologs prompted us to explore whether WAPL downregulation is important in contexts other than meiosis. Studies in yeast and human cells have shown that the DDR kinases ATM and ATR mediate cohesin enrichment at DNA-damage loci, which promotes repair 41,44,86 . ATM has also been found to promote the enrichment of CTCF, a cohesin binding partner, at DNA-damage sites in human cells 87 . We thus investigated whether the establishment of cohesin-enriched domains at DNA-damage foci depends on ATM activity and/or WAPL downregulation.
We treated HeLa cells with DNA-damage-inducing agents, including the radiomimetic DNA-cleaving agent bleomycin and the topoisomerase II poison etoposide (ETO), both of which lead to DSBs that activate ATM. Following treatment with either agent, we observed foci that were positive for both γH2A.X (S139-phosphorylated histone H2A.X) (Fig. 8a) 88,89 and pS/TQ (Fig. 8b) 52 . The mitotic kleisin Rad21 was concentrated at many of these damage foci (Fig. 8a-d).
We next induced DNA damage after treating cells with specific inhibitors of ATM, ATR, and DNA-PK (Fig. 8e), three paralogous DNA-damage transducer kinases. Inhibition of ATM, but not ATR or DNA-PK, resulted in markedly reduced γH2A.X upon ETO treatment ( Fig. 8e and Extended Data Fig. 7a). Inhibition of ATM also largely eliminated damage-induced Rad21 foci (Fig. 8e,i).
To corroborate the specificity of the chemical inhibitors, we performed short interfering RNA (siRNA)-mediated knockdown of ATM and ATR (Fig. 8f). We confirmed knockdown of ATM using a commercial antibody (Extended Data Fig. 7c,d). Although we lacked a similar tool to monitor ATR abundance, we noted that nuclear volume was reduced by either ATR inhibition or ATR knockdown (Extended Data Fig. 7b,g). Consistent with our observations with small-molecule inhibitors, ATM knockdown resulted in a substantial decrease in nuclear γH2A.X intensity ( Fig. 8f and Extended Data Fig. 7f). Importantly, Rad21 was no longer enriched at damage foci marked by either γH2A.X or pS/TQ (Fig. 8f,j and Extended Data Fig. 7e). Together, these results indicate that the activity of ATM, but not ATR or DNA-PK, is required for cohesin enrichment at DNA-damage foci.
We next tested whether WAPL downregulation contributes to the assembly of these cohesin-enriched domains. We reasoned that if ATM promotes the formation of cohesin foci by phosphorylating WAPL, overexpression of nonphosphorylatable WAPL might impair focus formation. Human WAPL has two potential mini-SCDs (Fig. 8g). Importantly, neither overlaps with the FGF or YSR motifs, which mediate the interaction between WAPL and PDS5 (refs. 24,90 ). Overexpression of GFP-tagged wild-type WAPL or mutant proteins did not affect the appearance of γH2A.X following DNA damage (Fig. 8g-h), indicating that ATM signaling was not perturbed (Fig. 8h and Extended Data Fig. 7h). Overexpression of wild-type WAPL also did not significantly affect nuclear RAD21 foci following damage (Fig. 8h,k). However, overexpression of nonphosphorylatable GFP-WAPL (WAPL 5A ) blocked the formation of RAD21 foci very effectively, whereas GFP-WAPL 5D lacked this activity (Fig. 8h,k). To test whether the reduction of RAD21 foci in cells expressing GFP-WAPL 5A is due to WAPL-dependent cohesin release, we overexpressed a nonphosphorylatable WAPL protein lacking four amino acids ( 1116 MEDC 1119 ) that are critical for release activity ( Fig. 8g) 91,92 . Although this mutant protein (WAPL 5AΔ4 ) localized to nuclei, it had no effect on RAD21 foci following DNA damage (Fig. 8h,k). Interestingly, we also found that GFP-WAPL 5AΔ4 overexpression caused nucleus-wide clustering of cohesin, similar to the 'vermicelli' phenotype, although the cohesin threads appeared to be much thinner ( Fig. 8h) 33 , suggesting that this protein may act in a dominant negative fashion. Taken together, these results indicate the role of both ATM activity and WAPL mini-SCD in regulating local enrichment of cohesin at damage foci in mammalian cultured cells, similar to results for C. elegans meiotic chromosome axes.

Discussion
We have found that downregulation of WAPL by ATM promotes cohesin localization along meiotic chromosome axes in C. elegans and at DNA-repair foci in mammalian cells (Extended Data Fig. 8). A key function of meiotic axes is to regulate the outcome of repair of induced DSBs 93,94 , so it makes teleological sense that this assembly would be regulated by DDR signaling 95 . Our findings illuminate the role of cohesin regulators and how they are orchestrated by DDR during the unique cell cycle state of meiotic prophase. Expression of meiosis-specific cohesins is necessary but not sufficient for axis formation and function 96,97 . We find that CHK-2 promotes break-independent activation of ATM-1 at meiotic entry, which in turn promotes axis assembly through downregulation of WAPL-1, a key regulator of cohesin dynamics 21,22 . Previous studies have shown that WAPL reduces cohesin residence time on chromatin and modulates cohesin clustering and cohesin-dependent loop extrusion 33,98 . We find that downregulation of WAPL-1 is important Lower and upper box ends represent the first and third quartiles, with the median indicated by the horizontal line within the box. All data points are shown, and the sample sizes are indicated. ****P < 0.0001 (two-sided Wilcoxon-Mann-Whitney test, adjusted by Bonferroni correction). e, Immunostaining of ƔH2A.X and RAD21 in nuclei of HeLa cells treated with kinase inhibitors, followed by ETO to induce DNA damage, showing that ATM activity is required for cohesin concentration at DNA-damage foci. The chemical inhibitors, which affected specific kinases, were as follows: KU55933 (ATMi), VE-821 (ATRi), and NU7441 (DNA-PKi). The control group (CONi) was treated with DMSO. Scale bar, 10 µM. f, RAD21 immunostaining in nuclei of HeLa cells depleted of ATM and/or ATR prior to ETO-induced DNA damage, showing that ATM is required for cohesin concentration at DNA damage foci. siRNAs specific for each kinase were used for knockdown. A non-targeting siRNA pool (Methods) was used for control knockdowns (siCON). Scale bar, 10 µM. g, Domain architecture of human WAPL, indicating the positions of the YSR (tyrosine-serine-arginine) motif, FGF (phenylalanine-glycine-phenylalanine) motifs, the two SCD domains, the MEDC (methionine-glutamate-aspartate-cysteine) sequence, and the residues that were mutated in our transgenic constructs. h, Immunofluorescence of HeLa cell nuclei expressing GFP or GFP-WAPL, showing that overexpression of nonphosphorylatable WAPL inhibits cohesin concentration at DNA-damage foci. Scale bar, 10 µM. i,j,k, Quantification of the number of RAD21 foci in e, f, and h, as described in 'Image analysis' and 'Data presentation'. Lower and upper box ends represent the first and third quartiles, with the median indicated by the horizontal line within the box. All data points are shown, and the sample sizes are indicated. ****P < 0.0001 (two-sided Wilcoxon-Mann-Whitney test, adjusted by Bonferroni correction).
Our findings also reveal that phosphorylation by ATM inhibits WAPL. A role for ATM in meiotic axis morphogenesis has also been demonstrated in Arabidopsis 101 . ATM and ATR localize sequentially to chromosome axes during EM prophase in mice, consistent with roles in regulating cohesin along the axis 102 . Our results also reveal a key function for this regulatory pathway in the DDR in proliferating cells 103,104 .
We also find that the acetyltransferase ECO-1 contributes to the stability of axial cohesins, although this was apparent only when downregulation of WAPL was defective. This is consistent with observations from yeast, Drosophila, and Arabidopsis 64,105,106   which in turn antagonizes Wapl and promotes cohesion 68,107 . Analogous regulation of C. elegans COH-3/4 by CHK-2 has also been proposed 12 . Studies across diverse eukaryotes have established critical roles for Pds5 in regulating cohesin-dependent chromosomal eve nts 20,24,25,74,76,[108][109][110] . It plays a widely conserved role in meiosis and is often regarded as an axis component 75 , but its specificity seems to vary across species. In C. elegans, PDS-5 stabilizes REC-8 against WAPL-1, analogous to evidence from fission yeast 81 . However, loss of Pds5 in budding yeast or of PDS5 in Arabidopsis has little effect on the association of Rec8 (REC8) with meiotic chromosomes 82,83,111 . Nevertheless, budding yeast PDS5 mutants show SC formation between sister chromatids, rather than homologous chromosomes, a phenotype also seen in C. elegans and mouse spermatocytes lacking REC-8 or Rec8, respectively 28,82 . Thus, Pds5 may promote cohesive activity of Rec8 cohesin, even if it is dispensable for Rec8 localization to chromosome axes.
The functional interplay between Pds5 and Wapl in different organisms has been enigmatic. Some results have indicated that these factors act as a complex; in other cases, Pds5 antagonizes Wapl activity, as shown here. We found that WAPL-1 antagonizes cohesin localization even when PDS-5 is depleted (Fig. 7g,h), indicating that it can function independently of PDS-5. Most importantly, WAPL-1 depletion restores axial REC-8 cohesin upon PDS-5 depletion. Notably, the FGF and YSR motifs in vertebrate Wapl that mediate binding to Pds5 (refs. 24,90 ) are both absent from C. elegans WAPL-1 (ref. 71 ). C. elegans PDS-5 also has a relatively long, unstructured domain that may modulate its activities through mechanisms analogous to the function of Sororin in vertebrates and Dalmatian in Drosophila, both of which are cohesin-protecting factors that antagonize WAPL-dependent cohesin release 61 .
Taken together, the results of our work show that axis assembly is driven by the specialized roles of meiotic cohesins and their interactions with cohesin regulators, which in turn are controlled by specialized DDR signaling during meiotic prophase. Additionally, our study reveals a pathway that regulates cohesin activity to promote programmed induction and repair of DSBs in meiotic cells and repair of exogenous breaks in proliferating cells.

Online content
Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41594-023-00929-5.

Strain maintenance
All C. elegans strains were maintained on standard nematode growth medium (NGM) plates seeded with OP50 bacteria at 20 °C 112 . Young adult hermaphrodites (20-24 hours (h) post-L4) were used for immunofluorescence analysis. C. elegans is a laboratory model nematode that does not require ethical approval to study.

Strain construction
All new alleles used in this study were generated by CRISPR-Cas9-mediated genome editing. See Supplementary Table 1 for the list of strains used in this study. Briefly, Alt-R CRISPR-Cas9 crRNAs specific for target sites were mixed with a dpy-10-specific crRNA 113 at a molar ratio of 8:1. These were denatured and annealed to an equal quantity of tracrRNA (Integrated DNA Technologies) by being heated to 95 °C for 5 minutes (min), followed by 5 min at 25 °C. 1 µL of 100 µM hybridized tracrRNA/crRNA was combined with 2.5 µL of 40 µM S. pyogenes Cas9-NLS purified protein (QB3 MacroLab) and incubated at room temperature for 5 min. Next, 0.5 µL of 100 µM stock of an Ultramer DNA oligonucleotide (IDT) repair template containing 35-45 bp homology arms and the desired epitope/degron or mutation sequence was added to the mixture, for a total volume of 5 µL, and injected into the gonads of young adult hermaphrodites aged 24 h from the late L4 stage. Hermaphrodites were injected and maintained on individual plates at 20 °C for 3-4 days. Roller and Dumpy F1 progeny were singled, maintained at 20 °C for 3 days, and screened by PCR for the desired mutation or epitope tag. Candidate alleles were verified by Sanger sequencing. See Supplementary Table 2 for a complete list of crRNAs, repair templates, and genotyping primer sequences used in this study. Validations of key constructed strains are available in Supplementary Table 4.

Worm viability and fertility
To quantify brood sizes, male self-progeny, and embryonic viability, L4 hermaphrodites were plated individually and transferred to new plates daily for four consecutive days. Eggs were counted twice a day to minimize counting errors. Viable progeny and males were scored when they reached young adulthood.

Auxin-induced protein depletion in worms
Auxin-induced depletion of degron-tagged proteins was performed as previously described 114 . Hermaphrodites at the L4 stage were transferred to seeded plates containing 1 mM indole-3-acetic acid (IAA, auxin) and incubated for 24 h before analysis. For each experiment, strains being compared were treated in parallel using the same batch of auxin-containing plates. Degron and epitope tags were inserted into genes of interest by genome editing in a strain expressing TIR1 in the germline (see Supplementary Table 1 for detailed information), which was used as a control and treated in parallel with all other strains in each assay. The stability and function of degron-tagged proteins in the absence of auxin were validated by localization and/or phenotypic assays. The kinetics and efficacy of depletion were analyzed by immunolocalization, functional assays, and, where feasible, by western blots.

Plasmids
To express human WAPL in HeLa cells, sequences were inserted into the pcDNA3-acGFP vector, obtained from Addgene (cat. no. 128047). The WAPL coding sequence was divided into four ~1-kb fragments (sequences are available in Supplementary Table 3) and synthesized by Twist Bioscience. These fragments were inserted at the 3′ end of the GFP coding sequence using Gibson assembly 115 and verified by Sanger sequencing.

Transient transfection
For WAPL overexpression, HeLa cells were grown on coverslips in 6-well plates to 50% confluency. Then, 2.5 µg of purified plasmid DNA was mixed with 5 µL Lipofectamine 3000 (Thermo Fisher) and used for transfection according to the manufacturer's protocol. Cells were fixed for imaging 48 h after transfection. DNA-damage-inducing agents and/or kinase inhibitors were added 24 h before fixation.

Chemical treatments
DNA damage was induced by addition of 0.8 µM ETO or 0.4 µM bleomycin for 24 h. KU55933, VE-821, and NU7441 were added at 1 µM for 24 h. Chemicals were dissolved in DMSO (dimethylsulfoxide).

Immunofluorescence assays
Adult hermaphrodites were dissected on a clean coverslip in egg buffer (25 mM HEPES pH 7.4, 118 mM NaCl, 48 mM KCl, 2 mM EDTA, 0.5 mM EGTA) containing 0.01% tetramisole and 0.1% Tween-20. Samples were fixed for 2 min in egg buffer containing 1% formaldehyde and then transferred to a 1.5-mL tube containing PBS + 0.1% Tween-20 (PBST). After 5 min, the buffer was replaced with ice-cold methanol and incubated at −20 °C for an additional 10 min. Worms were washed twice with PBST, blocked with Roche blocking reagent diluted into PBST, and stained with primary antibodies diluted in blocking solution at 4 °C overnight. Samples were then washed with PBST and incubated with secondary antibodies that were diluted in blocking solution at room temperature for 1 h. Worms were washed twice with PBST and mounted in ProLong Diamond with DAPI (Invitrogen) before imaging.

Nature Structural & Molecular Biology
Article https://doi.org/10.1038/s41594-023-00929-5 For immunofluorescence of HeLa cells, coverslips in 6-well plates were washed with PBS and then fixed with 4% formaldehyde in PBS at room temperature for 10 min. After 3 washes with PBS, cells were permeabilized by addition of 0.5% Triton X-100 in PBS at room temperature for 5 min. They were rinsed with PBS and blocked with 5% BSA in PBS at room temperature for 1 h. Cells were then washed with PBS and incubated with primary antibodies diluted in 1% BSA at room temperature for 2 h. After another PBS wash, cells were incubated in secondary antibodies diluted in 1% BSA in PBS at room temperature for 1 h in the dark. Cells were then washed again with PBS and mounted in ProLong Diamond with DAPI before imaging.

Image analysis
To quantify the abundance of proteins in C. elegans germline nuclei, additive projections were generated from raw (undeconvolved) 3D data stacks after background subtraction using the rolling ball tool in ImageJ. Individual nuclei (regions of interest, ROIs) were manually segmented based on DAPI staining in ImageJ, and the integrated intensity within each ROI was calculated. For each condition, 80-200 nuclei from 3-4 representative gonads were quantified.
To quantify protein abundance in HeLa cell nuclei, individual nuclei (ROIs) were first segmented on the basis of DAPI fluorescence in an equatorial optical section from a 3D image stack using the 2D watershed tool (scikit-image library v0.18, Python 3.9). Protein abundance (integrated intensity) within this region was calculated from additive Z projections, similar to the approach we used to quantify proteins in C. elegans germline nuclei.
To quantify RAD21 enrichment at sites of DNA damage in HeLa cells, we developed an automated method to ensure consistency and minimize potential investigator bias. Following empirical optimization, the method was applied to each dataset using ImageJ macros. For experiments involving expression of GFP or GFP-WAPL, only GFP-positive cells were included; these were identified based on GFP fluorescence in equatorial sections using the Auto Threshold tool in ImageJ in 'Li' maximum entropy mode. Nuclear ROIs were segmented as described above. Peaks of immunofluorescence of DNA-damage markers (γH2A.X or pS/TQ) were segmented using the Auto Threshold tool in MaxEntropy mode. The resulting binary masks of damage-marker-enriched nuclear regions were used to segment and calculate the average intensity (integrated intensity/area) of RAD21 at DNA-damage regions from additive Z projections. The average background RAD21 intensity was calculated from the nuclear regions outside of these masks.

Data presentation
For data based on immunofluorescence in C. elegans germline nuclei, we show representative images of the distal regions of dissected gonads. All images are oriented with the distal tip on the left. They show the entire proliferative (PM) region and a similarly sized region containing nuclei in EM prophase. The boundary between PM and meiotic prophase is indicated by a dashed line. Figure labels indicate proteins that were depleted by auxin treatment. For immunofluorescence in HeLa cells, representative images of individual nuclei are shown with enlargements of fluorescent foci as insets.
For quantitative analysis of immunofluorescence, the integrated nuclear intensity or number of foci were measured as described above under 'Image analysis.' Tukey boxplots of data points from individual nuclei were generated using R. Boxes indicate the quartiles and median, and the median value is also indicated next to the box. The number of nuclei that were scored for each condition or group is shown in parentheses underneath the data points.

Statistical analysis
We used the Shapiro-Wilk test to determine whether our data for each condition showed a normal distribution. We used Student's t-test to compare data sets that were found to show a normal distribution (P > 0.05 by the Shapiro-Wilk test); otherwise, we used the Wilcoxon-Mann-Whitney test. P values were adjusted by Bonferroni correction when statistical analyses involved multiple tests on the same dataset. The number of asterisks indicates the calculated P values: **P < 0.01, ***P < 0.001, ****P < 0.0001. The exact values of the P values that are greater or equal to 0.01 are indicated on the plots.

Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability
Source data are provided with this paper.
HTP-3 marks chromosome axes in meiotic prophase nuclei. Scale bar, 2 µM. d,g, Quantification of WAPL-1 intensity in b and e, respectively. Lower and upper box ends represent the first and third quartiles, with the median indicated as the horizontal line within the box. All data points are shown with the sample sizes indicated below the boxes. ****P < 0.0001 (two-sided Wilcoxon-Mann-Whitney test, adjusted by Bonferroni correction). Data for plots in d and g are available as source data.