Fig 1.
Localization of Rec8, Rad21, Spo76/Pds5, Sororin, Wapl, Hop1 and Red1 during wild-type meiotic prophase I.
Single-cell fluorescence imaging of GFP- or TdTomato-tagged proteins shows that all seven proteins localize to chromosome axes from zygotene to late pachytene. Meiotic stages were assigned based on chromatin morphology, synapsis status and ascus size. At leptotene, Rec8, Spo76/Pds5, Hop1 and Red1 appear as continuous lines, whereas Rad21, Wapl and Sororin display diffuse chromatin-associated signals. At pachytene, all seven proteins exhibit a continuous localisation along the entire length of the chromosomes. Their signals progressively decline after pachytene and become restricted to short segments during the post-pachytene diffuse stage. For each strain and for each prophase stage, we analysed a minimum of 20 nuclei n ≥ 20. Scale bars: 1 µm.
Fig 2.
Localization patterns of Spo76/Pds5, Rec8, Rad21, and Sororin in absence of Wapl.
Localization patterns of the four cohesin components in wild type (WT) and wapl∆ from leptotene to the diffuse stage. Note that: (i) In WT, Rad21 and Sororin are not detected on chromosome axes at leptotene and appear only from zygotene on, whereas in wapl∆ they are already associated with the axes at leptotene. (ii) At pachytene, Spo76/Pds5, Rec8, Rad21, and Sororin form single continuous lines along each of the seven bivalents because being synapsed at ~100 nm, homologous chromosome axes are indistinguishable in fluorescence microscopy. In contrast, in wapl∆, the two homologous axes are visible because of synapsis defects (see details and defects in text). (iii) As clear from the magnification bars (identical in WT and mutant), chromosome axes are shorter in wapl∆ compared to WT (details in text). (iv) At the diffuse stage, all four cohesin components persist along chromosome axes in wapl∆, whereas in WT they are restricted to short segments. n ≥ 20 nuclei for each strain and for each prophase stage. Scale bars: 1 µm.
Table 1.
Distribution of bivalent synapsis categories in waplΔ and hop1Δ waplΔ strains.
Fig 3.
Sororin prevents premature Wapl-mediated release of Spo76/Pds5 and Rec8 but is dispensable for Red1 and Hop1 localization.
(A-B) Localisation of Spo76/Pds5-TdT and Rec8-GFP in sororinΔ single and sororinΔ waplΔ double mutants compared to their localization in WT and wapl∆. In sororinΔ, Spo76/Pds5-TdT and Rec8-GFP show only weak chromatin-associated staining at leptotene, and become only visible on axes from zygotene. In contrast, in the sororinΔ waplΔ double mutant, both proteins are located along chromosome axes from early leptotene on, and persist along axes throughout the diffuse stage, similarly to their location in the waplΔ single mutant. (C-D) In the absence of Sororin, GFP-Red1 and Hop1-GFP are associated with chromosome axes at leptotene, as in WT. (E) GFP-Red1 in rec8Δ. Like in WT, Red1 associates with chromosome axes at leptotene in the absence of Rec8. n ≥ 20 nuclei for each strain and for each prophase stage. Scale bars: 1 µm.
Fig 4.
Localization of Spo76/Pds5 in hop1Δ single and hop1Δ waplΔ double mutants compared with its localization in WT and wapl∆.
In hop1Δ, although Spo76/Pds5-TdT initially associates with chromosome axes at leptotene, its signal is specifically lost from certain regions at pachytene, thus forming 13-14 short segments [41]. In contrast, in the hop1Δ waplΔ double mutant, Spo76/Pds5-TdT is now visible along the entire chromosome axes at pachytene and persists throughout the diffuse stage like in the waplΔ single mutant. n ≥ 20 nuclei for each strain and for each prophase stage. Scale bars: 1 µm.
Fig 5.
Hop1, Spo76/Pds5, Rec8 and Rad21 localizations in red1∆ and red1∆ wapl∆.
(A-B) Hop1-GFP and Spo76/Pds5-TdT localization in red1Δ single and red1Δ waplΔ double mutants compared to their localization in WT and wapl∆ single mutant. Because reliable axis markers are lacking and chromatin morphology is severely disrupted in red1Δ, nuclei were staged as early, mid and late prophase based on ascus size. WT nuclei of comparable ascus size were used as reference, with early corresponding to leptotene, mid-prophase to pachytene, and late prophase to the diffuse stage. In red1Δ both proteins fail to associate with axes appearing only as diffuse chromatin-associated signal throughout prophase. In the red1Δ waplΔ double mutant, Hop1-GFP (A) remains absent on axes while Spo76/Pds5 (B) is located along chromosome axes from early to late prophase. (C-D) Rec8-GFP and Rad21-GFP axis localization in red1Δ single and red1Δ waplΔ double mutants compared to their localization in WT and wapl∆ single mutant. (C) In the single red1Δ mutant, Rec8 still associates with axes throughout prophase I but is observed only as short stretches or foci, contrasting with its continuous axis staining seen in WT. In contrast, in the red1Δ waplΔ double mutant, Rec8 forms now continuous lines along axes from early to late prophase, like in the waplΔ single mutant. (D) At early prophase, Rad21 is chromatin-associated in red1Δ like in WT. At mid prophase, Rad21 still associates with axes in red1Δ but, unlike in WT, its staining is never observed along the full chromosome axis length. Absence of Wapl in the red1Δ background restores premature and continuous Rad21 axis association from early prophase on, with staining persisting through late prophase as in waplΔ. n ≥ 20 nuclei for each strain and for each prophase stage. Scale bars: 1 µm.
Fig 6.
Slx8, but not Wapl, destabilizes Rec8 and Rad21 in the spo76-1 mutant.
(A-B) Rec8-GFP (A) and Rad21-GFP (B) in spo76-1 single, spo76-1 waplΔ, and spo76-1 slx8Δ double mutants at mid-prophase. In spo76-1, like in red1Δ (above), chromatin is diffuse and axis markers do not allow reliable staging. Therefore, nuclei were staged based on ascus size, using WT nuclei of comparable size as reference. In the spo76-1 mutant, Rec8 and Rad21 still associate with chromosome axes but appear as short stretches or discrete foci. Absence of Wapl in the spo76-1 background does not restore continuous localization of the two kleisins, indicating that Wapl does not act directly on Rec8 or Rad21. Deletion of SLX8 in the spo76-1 mutant markedly improves Rec8 and Rad21 axis staining, resulting in sharper and more continuous signals along chromosome axes (n ≥ 20 nuclei). These results indicate that, in the absence of Spo76/Pds5 on the axes, both kleisins are destabilized via a Slx8-dependent SUMO-ubiquitin-proteasome pathway. Scale bars: 1 µm.
Fig 7.
Model for the regulation of cohesin association with meiotic chromosome axes.
(A) Red1 and Hop1 axis proteins protect the association of cohesin components, Spo76/Pds5 and kleisins, with chromosome axes from Wapl-mediated release. In the absence of Wapl, Spo76/Pds5 and kleisins remain associated with the chromosome axes at the diffuse stage, demonstrating that Wapl ensures the timely cohesin release by the end of pachytene. Red1, Hop1, and Sororin act in a stage-specific manner to protect cohesins: Sororin during leptotene, Hop1 during pachytene, and Red1 throughout prophase I. (B) Spo76/Pds5 is a central Wapl target and is stabilized by the combined action of Red1, Hop1, and Sororin. Consistent with this model, these axis proteins become dispensable (for Spo76/Pds5 localization) when Wapl is absent, in agreement with the epistatic interaction of waplΔ over hop1Δ, red1Δ, and sororinΔ regarding Spo76/Pds5 axis localization. In turn, Spo76/Pds5 on chromosome axes protects kleisins from ubiquitination and proteasomal degradation via the STUbL Slx8 pathway, likely by masking post-translational modification sites.