Identification of Sordaria macrospora RED1, WAPL and SORORIN

RED1. Combined phylogeny-oriented structural homology searches using Foldseek, HMMsearch, and PSI-BLAST, together with multiple sequence alignments across fungal genomes, identified a single RED1 gene (SMAC4_01425) in Sordaria. The predicted Red1 protein (839-amino acids) contains a 244-residue C-terminal region (aa 583–769) analysed by DeepCoil2 to adopt a coiled-coil (CC) structure (S1A Fig). Multiple sequence alignments of the N-terminal residues from fungal orthologs (S1A Fig), combined with AlphaFold3 (AF3) structural modeling, revealed two conserved domains: (i) a 30-residue N-terminal motif likely mediating Hop1 interaction (ipTM: 0.71; pTM: 0.83; S1A, S1B and S1C Fig), and (ii) a 21-residue motif predicted to mediate Mek1 interaction (ipTM: 0.54; pTM: 0.62; S1A, S1B and S1D Fig). As previously reported in S. cerevisiae [43], AF3 modeling predicts that the Sordaria Red1 CC region mediates self-association (S1B and S1E Fig), consistent with its structural role in axis organization. In addition, AF3 modeling predicts interactions of Red1 with Rec114 and Zip4 consistent with functional roles in meiotic chromosome axis assembly [44,45] (S1B, S1F and S1G Fig). Together, our phylogenetic and structural modeling analyses strongly support that SMAC4_01425 encodes a bona fide RED1 homolog.

WAPL was identified by PSI-BLAST using the Schizosaccharomyces pombe Wapl protein sequence as a query. The Sordaria genome contains a single WAPL gene (SMAC4_07015). The predicted Wapl protein (852 amino acids) displays the domain organization shown in S2A Fig. It contains a conserved C-terminal domain (aa 380–726; S2A Fig), that defines the Wapl family and is predicted to be predominantly α-helical [13]. In contrast, the highly divergent N-terminal region (variable in both length and composition) is thought to mediate interaction with Pds5 [46]. AlphaFold2 (AF2) modeling (S2B and S2C Fig showing prediction confidence values) indicate that this N-terminal region is intrinsically disordered, as previously proposed for other Wapl orthologs. AF2 modeling of the C-terminal region (aa 380–852), combined with structural similarity searches against experimentally determined atomic structures in the Protein Data Bank, revealed that the Sordaria Wapl domain shares highest structural homology with the Ashbya gossypii Wapl domain (PDB: 3zik, Chain B; (S2D Fig). Structurally, the Sordaria Wapl domain adopts a fold of 25 consecutive α-helices, 15 of which align with the 21 α-helices present in A. gossypii ortholog (Z-score: 5.084). The Wapl domain can be subdivided into three regions, with the central region comprising six HEAT repeats [47].

SORORIN. Sordaria Sororin (SMAC4_08091) is a 701-amino acids protein, which, like all members of this family, is largely unstructured and poorly conserved except for the Sororin boxes [36]. Orthologs are found in human, insects, plants, and fungi except for S. cerevisiae which lacks a potential ortholog [36]. Sororin is characterized by a very short C-terminal domain (Sororin boxes), consisting of a polar linker (K/R-rich domain which varies in size between 10 and 20 amino acids) and a conserved motif predicted to form two α-helices and a β-strand [21]. Vertebrate Sororin and Wapl proteins interact with Pds5 through their YSR and FGF motifs [21,46]. The putative FGF and KEN box motifs, are not present in the Sordaria protein. In vertebrates, the KEN box mediates Sororin degradation during mitotic exit, ensuring timely cohesin removal. Absence of this motif in Sordaria suggests an alternative degradation mechanism, potentially involving ubiquitin ligases distinct from APC/CCdh1. See [36] for C-terminal domain alignment.

Role of Sororin, Wapl, and Red1 in vegetative growth, sexual development and ascospore viability

To assess the role of Sororin, Wapl, and Red1 during the vegetative phase, we compared the vegetative growth of the sororinΔ, waplΔ, and red1Δ null mutants with the wild-type growth rate. To do so, the four strains were inoculated on M2 minimal medium plates and incubated at 25 °C for four days until the plates were fully covered by the mycelium. Three plates were used for each strain and the mycelium growth was measured twice daily. All strains showed similar mycelial phenotypes. However, while the red1Δ mutant exhibited a growth rate indistinguishable from that of the wild type, the sororinΔ mutant showed only a slight decrease in growth rate (~10%, 2.00 cm/day compared to 2.26 cm/day for WT). In contrast, the waplΔ mutant displayed a pronounced and significant reduction in growth rate (~32%, 1.54 cm/day compared with the wild-type rate of 2.26 cm/day) (S3A Fig and S1 Data). These results indicate that Wapl, but not Red1, is required for wild-type vegetative growth, while the small reduction observed in sororinΔ does not support a functional requirement for Sororin in this process. We further observed that the sororinΔ waplΔ double mutant exhibits the same growth rate as the waplΔ single mutant, indicating that waplΔ is epistatic to sororinΔ with respect to vegetative growth phenotype (S3A Fig and S1 Data).

To analyse the transition from the vegetative to the sexual cycle, we monitored the protoperithecia (fruiting bodies) development over 5 days. No differences were observed in the timing or formation of protoperithecia between the null and wild-type strains, suggesting that Sororin, Wapl, and Red1 are not required for the initiation of the sexual cycle. However, clear defects were observed at later stages. Asci development and ascospore formation were assessed by dissecting perithecia after five days of growth. All three mutants exhibited impaired development compared with the wild-type strain, in both ascus differentiation and ascospore morphology (S3B Fig and S1 Data).

In the sororinΔ mutant, no asci containing eight wild-type-like ascospores were observed. Instead, 83.9% of the asci aborted before the sporulation step (blue arrow) as indicated by their numerous vacuoles (green arrow). The remaining 16.1% asci contained ascospores, but with abnormal shapes (pink asterisk).

In the waplΔ mutant, 70% of asci contained numerous vacuoles (green arrow) and did not form eight mature ascospores, while the remaining 30% of asci exhibited eight ascospores like the wild-type strain (red arrow).

In the red1Δ mutant, 41% of asci were either aborted or contained numerous vacuoles (blue and green arrows), while the remaining 59% contained eight wild-type like ascospores (red arrow).

Finally, we examined the ascospore germination of the three null mutants. For this, 80 ascospores were isolated from each strain and deposited on the germination medium. After 24 h at room temperature, the percentage of germinated ascospores was determined. When compared to wild type, the germination rate of the sororin∆ mutant was strongly reduced (~77%), while absence of Wapl resulted in a moderate reduction (~27%). In contrast, the red1Δ mutant showed a wild-type germination rate (S3C Fig and S1 Data).

Taken together, these results demonstrate that Sororin and Wapl are essential for successful sexual reproduction, affecting ascus development, ascospore morphology, and germination efficiency. By contrast, Red1 is dispensable for vegetative growth but still contributes to normal ascus and ascospore development.

Importantly, in all complemented strains, expressing ectopic GFP-tagged proteins in the corresponding null strains, growth rates, ascus development, ascospore morphology and germination rates were indistinguishable from wild type, demonstrating that the GFP-fusion proteins are functional and do fully complement the defects observed in sororinΔ, waplΔ and red1Δ mutants (S3A, S3B, S3C Figs and S1 Data).

Loading dynamics and localization of Red1, Rad21, Wapl and Sororin during wild-type meiotic prophase I

The localization of the four proteins was examined by single-cell imaging using fully functional GFP fusions: C-terminal GFP fusions for Rad21, Wapl and Sororin and an N-terminal GFP fusion for Red1 (hereafter referred to as GFP-Red1). All fusion proteins were expressed under the control of their native promoters.

To allow a reliable comparison with these newly identified proteins, we first confirmed the previously reported Sordaria localization patterns of Rec8-GFP, Hop1-GFP, and Spo76/Pds5-TdTomato in the wild-type strain [12,41,42] (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g001

As shown previously, Rec8, Spo76/Pds5, and Hop1 localize along the full length of the chromosome axes from early leptotene on and remain stably associated with the axes until the end of pachytene. Their signals provide robust markers for monitoring homolog recognition during zygotene and homolog synapsis during pachytene. Synapsis corresponds to the assembly of the synaptonemal complex, which aligns homologous chromosomes into close proximity (~100 nm), appearing as a single linear signal under fluorescence microscopy. At the end of pachytene, their signal progressively declines throughout the post-pachytene diffuse stage, and is no longer detectable by diplotene (Fig 1 and S4 Fig).

Consistent with its role as an axial-element protein, Red1 exhibits the same loading and localization dynamics as Hop1. GFP-Red1 is localized on the axis from very early leptotene on, and runs the full length of the chromosomes as smooth continuous lines until the end of pachytene (Fig 1, right).

In contrast, Wapl and Sororin are detected on chromatin at early leptotene, and their staining appears as short stretches on the axes at late leptotene. Both signals extend along the entire chromosome axis throughout pachytene (Fig 1).

Interestingly, the two kleisins Rec8 and Rad21 display distinct loading patterns at leptotene. Unlike Rec8, which is located on the chromosome axes at early leptotene, Rad21 colocalizes with the chromatin at leptotene but is only detectable on the axes from zygotene onward (Fig 1, left). This localization difference suggests that, in Sordaria, Rec8- and Rad21-containing cohesin complexes are sequentially loaded onto DNA during early prophase I. By pachytene, however, their staining patterns converge: both are detected along the full length of the axis and remain detectable as short stretches during the diffuse stage (Fig 1, left bottom). These observations indicate that, in Sordaria, both cohesin-complexes populations coexist during meiosis but exhibit distinct loading kinetics in early prophase I.

To address the relevance of Wapl, Red1 and Sororin together with Rad21, Rec8 and Spo76/Pds5 for chromosome-axis structure during meiotic prophase, we isolated and analyzed both single and double null alleles of the corresponding genes to test both their localization and roles in chromosome-axis morphology and synapsis.

Wapl controls the association of cohesin components and axis proteins with the meiotic chromosome axes and is required for synapsis and chromosome compaction

To determine the role of Wapl during prophase I, we analyzed the localization of Spo76/Pds5, Rec8, Rad21 and Sororin (Fig 2) as well as of the axis proteins Hop1 and Red1 (S5 Fig) in the wapl∆ mutant. Like in wild type, all six proteins localize along the homolog axes from leptotene throughout pachytene in the null mutants. However, the absence of Wapl results in the persistent association of these proteins with chromosome axes beyond pachytene, through the diffuse stage, diplotene, and up to anaphase I (Fig 2 and S5 Fig; S4 Fig for diplotene).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g002

This sustained persistent localization highlights the role of Sordaria Wapl as a releasing factor that promotes the timely dissociation of cohesin components (Spo76/Pds5, Rec8, Rad21, and Sororin) from meiotic axes, consistent with its conserved function in other organisms (review in [2]). Furthermore, our data reveal that Wapl also negatively regulates the retention of axis proteins Hop1 and Red1, revealing a previously uncharacterized function of Wapl in axis protein dynamics.

Interestingly, in the waplΔ mutant, Rad21 and Sororin are already detectable on chromosome axes from leptotene on, in contrast to their later axis localization in wild type. This premature association suggests that Wapl acts not only as a releasing factor at later stages, but also functions earlier during prophase I to prevent stabilization of Rad21 and Sororin on the axes (Fig 2).

We propose that Wapl regulates protein dynamics on meiotic chromosomes at two distinct stages: first, by limiting early axis association of Rad21 and Sororin during leptotene, and later, by facilitating their removal during the diffuse stage. While these activities may reflect a common underlying mechanism, our results do not exclude the possibility that Wapl interferes with initial loading via a separate pathway.

Despite proper localization of cohesin components and axis proteins (Red1 and Hop1) throughout prophase I, indicating that, axis assembly per se is not impaired in the waplΔ mutant, we observed that homologous chromosomes are paired but fail to achieve full synapsis.

To assess this defect, we classified the synaptic status of individual homolog pairs (bivalents) in each nucleus into three categories: (A) complete synapsis (arrow in S6A Fig), (B) partial synapsis (asterisk in S6A Fig), or (C) absence of synapsis (arrowhead in S6A Fig). (n = 34 nuclei, 238 bivalents; S6A Fig). Our analysis reveals that only 10,9% of the 238 bivalents display complete synapsis, whereas 89,1% exhibit synapsis defects: 45,8% show partial synapsis and 43,3% show a complete absence of synapsis (S1 Data, values are summarized in Table 1). Notably, we did not observe any nucleus in which all seven bivalents achieved full synapsis. These findings show that Wapl is essential for the establishment of complete homolog synapsis. This synapsis defect is further supported by the localization of the synaptonemal complex (SC) transverse filament protein Sme4/Zip1-GFP. In wild-type nuclei, Sme4/Zip1-GFP forms a continuous linear signal along the full length of each bivalent at pachytene and colocalizes with Spo76/Pds5-TdT (S6B Fig top). By contrast, in waplΔ mutants, Sme4/Zip1-GFP signal is frequently discontinuous (arrows in S6B Fig bottom), consistent with incomplete synapsis, and appears continuous only in fully synapsed bivalents.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Distribution of bivalent synapsis categories in waplΔ and hop1Δ waplΔ strains.

https://doi.org/10.1371/journal.pgen.1012001.t001

We further measured chromosome axis length using Spo76/Pds5-TdTomato as a marker in the waplΔ mutant. At pachytene, this marker allows accurate tracing of chromosome axes regardless of the synapsis status (S6A Fig for representative tracings). Axis length is significantly reduced in the waplΔ mutant compared to wild type (41.74 ± 5 μm, n = 34 vs. 66.12 ± 7 μm, n = 67; Welch’s t test p < 10^-4; S1 Data). Notably, similar axis lengths are observed in both synapsed and unsynapsed bivalents, indicating that chromosome compaction level is independent of synapsis extent.

Taken together, these findings demonstrate that Wapl regulates both the temporal association of cohesin and axis components as well as the higher-order organization of meiotic chromosomes.

Sororin protects Spo76/Pds5 and Rec8 from Wapl-mediated release during early prophase I but is not required for Hop1 or Red1 localization

Given that Sororin counteracts Wapl’s cohesin release activity through its association with Spo76/Pds5 in mouse cells [21], we investigated the impact of its absence on chromosome axis structure in Sordaria.

When compared to wild type, sororinΔ exhibits delayed axis localization of Spo76/Pds5 (Fig 3A) and Rec8 (Fig 3B): both proteins show weak chromatin-associated staining at leptotene, and become clearly associated with axes only from zygotene onwards (Fig 3A and 3B). These results suggest that Sororin is required for timely recruitment and/or stabilization of Spo76/Pds5 and Rec8 at leptotene, potentially via direct or indirect interactions, or by protecting them from premature Wapl-mediated release.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g003

In contrast, in the absence of Sororin, GFP-Red1 (Fig 3C) and Hop1-GFP (Fig 3D) are associated with the axes at leptotene like in wild type. Thus, recruitment of Red1 and Hop1 on the axes does not require Sororin nor the early presence of Spo76/Pds5 and Rec8. This finding differs from the observations made in S. cerevisiae, where Red1 loading partially depends on Rec8 [48,49]. However, it is consistent with data from mouse, where it has been shown that SYCP2 (Red1 homolog) localizes on axes in the absence of Rec8 [50,51]. Localization of GFP-Red1 along chromosome axes in the rec8∆ mutant (Fig 3E) strengthens our conclusions that, in Sordaria, Sororin is not required for Red1 loading on axes at leptotene.

Finally, analysis of double mutants shows that the absence of Wapl in the sororinΔ mutant restores Spo76/Pds5 and Rec8 axis association. Both proteins are again detected along the chromosome axes from early leptotene on in the sororinΔ waplΔ double mutant (Fig 3A and 3B), indicating that Sororin protects these components from Wapl-dependent release during early prophase I.

Consistent with Wapl’s role as a cohesin releasing factor, Spo76/Pds5 and Rec8 axis staining persists during the diffuse stage in the sororinΔ waplΔ double mutant, like in the waplΔ single mutant (Fig 3A and 3B), demonstrating that waplΔ is epistatic to sororinΔ with respect to Spo76/Pds5 and Rec8 localization.

Together, these results indicate that Sororin safeguards Spo76/Pds5 and Rec8 from premature Wapl-mediated release during early prophase I, whereas the localization of axis proteins Red1 and Hop1 is Sororin-independent, highlighting a specific protective role of Sororin in cohesin dynamics.

Hop1 protects Spo76/Pds5 from Wapl-mediated release during pachytene

Previous work has shown that Hop1 is required to maintain Spo76/Pds5 along chromosome axes at pachytene: although present along the axes during leptotene and zygotene, Spo76/Pds5 signal is lost in hop1Δ, specifically in regions that lack a synaptonemal complex ( [41] and Fig 4). This indicates that Hop1 is not required for Spo76/Pds5 loading on the chromosome axes at early stages but is essential for its local stabilization on chromosome axes during pachytene. We now show that Red1 displays a similar local destabilization pattern along homologs in hop1Δ during pachytene (S7A Fig), even though its initial loading onto chromosome axes is independent of Hop1 (S7B Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g004

To test whether this destabilization is mediated by Wapl, we examined Spo76/Pds5 localization in the hop1Δ waplΔ double mutant. In this background, Spo76/Pds5-TdT is visible along the entire length of the homologous chromosomes at pachytene (Fig 4). This result strongly suggests that Wapl is responsible for the partial loss of Spo76/Pds5 in the hop1Δ mutant and that Hop1 protects Spo76/Pds5 from Wapl-mediated removal at pachytene. Strikingly, GFP-Red1 localization is also rescued in the double mutant and is now located along the entire chromosomal axes (S7B Fig). The concomitant restoration of Red1 and Spo76/Pds5 in the hop1Δ waplΔ mutant suggests a functional interdependence between these proteins. This interdependence could reflect a scenario in which stabilization of Spo76/Pds5 indirectly promotes Red1 association.

We next assessed homolog synapsis in the hop1Δ waplΔ double mutant. The synapsis defects are similar to those observed in the single waplΔ mutant (Fig 4 and S7B Fig). Among the 175 bivalents analyzed (25 nuclei), only 11,5% achieve complete synapsis, whereas 88,5% exhibit synapsis defects: 23% show partial synapsis and 66% show a complete absence of synapsis (S1 Data, values are summarized in Table 1). Notably, the proportion of unsynapsed bivalents is significantly increased in the hop1Δ waplΔ double mutant compared to waplΔ (66% vs 43%, respectively, 2way Anova p < 10^-4), suggesting that Hop1 contributes to synapsis in the absence of Wapl, although whether this reflects a direct role in synapsis in this context or an indirect effect through cohesin dynamics remains to be determined.

Red1 is essential for Hop1 and Spo76/Pds5 axis localization, but is not required for axis association of Rad21 and Rec8

As both Sororin and Hop1 contribute to the axis localization of Spo76/Pds5 in a stage-specific manner, we next asked whether Red1, a known partner of Hop1 in budding yeast [49,52], is required for the localization of cohesin components on meiotic chromosome axes.

We first observed that Red1 is essential for the axis association of both Hop1 (Fig 5A) and Spo76/Pds5 (Fig 5B) during prophase I. In the red1Δ mutant, both proteins display diffuse nuclear or chromatin-associated signals and never form linear axis-associated structures (Fig 5A and 5B). Absence of these axis markers prevents the visualization and analysis of pairing and synapsis in the red1Δ mutant. We, therefore, staged nuclei as early, mid, and late prophase, based on their ascus sizes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g005

Absence of Hop1-GFP in the red1Δ mutant shows that, in Sordaria like in budding yeast (e.g., [53]), the localization of Hop1 on axes requires the presence of Red1 (Fig 5A), whereas Red1 can associate with axes independently of Hop1 (S7B Fig).

In contrast, in absence of Red1, Rec8-GFP and Rad21-GFP exhibit foci and short axis-like stretches but their staining does never extend along the full-length of the chromosomes (Fig 5C and 5D). Thus, both kleisins can associate with chromosome axes independently of Red1, and consequently of Hop1 and Spo76/Pds5. However, Rec8 and Rad21 localization is clearly impaired compared to wild type, where both proteins form continuous axis-associated signals. These results suggest that Red1, Hop1, and/or Spo76/Pds5 promote the stable axis association of Rec8 and Rad21 along chromosome axes during prophase I.

Red1 protects Spo76/Pds5 from Wapl release and ensures proper chromosome architecture

Given the role of Wapl in promoting the release of Spo76/Pds5, Rec8 and Rad21 from chromosomes in both sororinΔ and hop1Δ mutants (see above), we next asked whether the loss of Wapl could restore axis association of these proteins in a red1Δ mutant background. To address this, we examined their localization in the red1Δ waplΔ double mutant.

First, Hop1-GFP remains undetectable on chromosome axes in the red1Δ waplΔ double mutant (Fig 5A), showing that Hop1 axis association strictly requires Red1 and does not involve the Wapl activity. Surprisingly, although Spo76/Pds5 is completely absent from chromosome axes in the red1Δ single mutant, in the red1Δ waplΔ double mutant, Spo76/Pds5 becomes clearly detectable along the chromosome axes from early to late prophase I (Fig 5B). This finding strongly indicates that the loss of Spo76/Pds5 on axes in the absence of Red1 is independent of Hop1 and is strongly, if not exclusively, due to the Wapl release activity.

We next examined the localization of the kleisins Rec8-GFP and Rad21-GFP in the red1Δ waplΔ mutant. Unlike their predominantly punctate or discontinuous staining observed in the red1Δ mutant, both proteins now form continuous lines along chromosome axes in the double mutant (Fig 5C and 5D, respectively). Notably, Rad21-GFP is already present on axes at early prophase I, as in the waplΔ mutant (Fig 5D, Top panel), confirming that Wapl prevents its premature association with chromosome axes. Furthermore, Spo76/Pds5, Rec8 and Rad21 staining persists during late prophase in the red1Δ waplΔ mutant, similar to their staining pattern in the waplΔ mutant (Fig 5B, 5C and 5D, Bottom panel). These observations are in line with the Wapl releasing activity and show that waplΔ is epistatic to red1Δ with respect to Spo76/Pds5 localization.

However, despite the presence of Spo76/Pds5, Rec8, and Rad21 on axes in the red1Δ waplΔ mutant, chromosome organization remains severely defective. Throughout prophase I, chromatin appears entangled, with no discernible individual chromosomes, even at mid-prophase stages (defined by ascus size), when the 7 bivalents are clearly distinguishable in wild type and waplΔ (S8 Fig). These structural defects closely resemble those observed in the red1Δ single mutant (S8 Fig), indicating that red1Δ is epistatic to waplΔ regarding chromatin organization.

Thus, although cohesin complexes are retained on the axes in the red1Δ waplΔ mutant context, their presence alone is insufficient either to restore wild-type axis architecture or to drive the enhanced chromatin compaction characteristic of waplΔ. This discrepancy could reflect two non-exclusive scenarios: (i) cohesins are recruited but remain non-functional in the absence of Red1, or (ii) cohesins are functional, but Red1 plays an upstream architectural role indispensable for meiotic chromatin organization and progression.

Our data reveal, thus, a dual genetic relationship: while waplΔ restores axis association of Spo76/Pds5, Rec8, and Rad21 in the absence of Red1, red1Δ remains epistatic to waplΔ regarding the chromatin organization. This highlights the fact that the presence of cohesins on axes is not sufficient to ensure normal axis structures.

Taken together, these results indicate that in the absence of Red1, Wapl activity compromises the stable association of Rec8 and Rad21 with chromosome axes, either directly or indirectly through the loss of Spo76/Pds5. In the latter scenario, Spo76/Pds5 is a primary target of Wapl, and its stabilization on the axes promotes kleisin retention.

Rec8 and Rad21 stability relies on Spo76/Pds5, independently of direct Wapl targeting

To assess whether Wapl directly targets kleisins, we analyzed Rec8-GFP and Rad21-GFP localization in the spo76–1 mutant, which expresses an hypomorphic allele of the essential SPO76/PDS5 gene in Sordaria. In this mutant, Spo76–1-GFP localizes to chromatin and forms only discrete foci during prophase I but is never observed along the chromosome axes (S9 Fig). In the spo76–1 mutant, both Rec8 and Rad21 associate with chromosome axes and exhibit short axis-like stretches and foci during prophase I (Fig 6A and 6B, respectively). This discontinuous staining pattern contrasts with their continuous axis staining in wild type, indicating that Spo76/Pds5 is required for the stabilization of both kleisins on the chromosome axes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g006

Strikingly, deletion of WAPL, in spo76–1, failed to restore continuous axis staining of Rec8 or Rad21: both kleisins retained their discontinuous pattern observed in the spo76–1 single mutant (Fig 6A and 6B, respectively). These findings indicate that: (i) Wapl does not act directly on kleisins, (ii) Spo76/Pds5 protects kleisins from destabilization, and (iii) kleisins are targeted by a distinct pathway in the absence of Spo76/Pds5 from the axes.

Rec8 and Rad21 are destabilized by the SUMO-targeted ubiquitin ligase Slx8 in the absence of Spo76/Pds5 axis localization

Having established that the destabilization of both kleisins is independent of Wapl activity in the spo76–1 mutant, we next investigated whether this instability could result from the proteasome-mediated degradation.

In S. cerevisiae, the mitotic kleisin Scc1/Rad21 is targeted for degradation in the absence of Spo76/Pds5 via the ubiquitin-proteasome system. This process is mediated by the SUMO-targeted ubiquitin ligase (STUbL) complex Slx5-Slx8, which recognizes polySUMOylated cohesin subunits and promotes their ubiquitin-dependent proteolysis [26].

To assess whether this pathway contributes to kleisin destabilization during meiosis in the spo76–1 mutant, we identified the Sordaria homolog of SLX8 (see Methods), deleted the corresponding gene, and examined Rec8 and Rad21 localization along chromosome axes in the absence of Slx8.

Notably, deletion of SLX8 in the spo76–1 mutant, leads to a marked improvement in the axial staining of both kleisins, which appear sharper, more continuous, and consistently maintained during prophase I (Fig 6A and 6B).

This improvement of axis association indicates that, in Sordaria, Rad21 is targeted during meiosis via a Slx8-dependent SUMO-ubiquitin-proteasome pathway, as previously established for mitosis in budding yeast. Importantly, our results show that this regulatory mechanism also applies to the meiotic-specific kleisin Rec8, revealing that Rec8 is similarly susceptible to STUbL-mediated turnover when Spo76/Pds5 is absent from the chromosome axes.

Together, these findings demonstrate that both mitotic and meiotic kleisins are subject to SUMO-ubiquitin-mediated degradation, and that Spo76/Pds5 plays a central protective role, shielding these kleisins from proteasome degradation during meiotic prophase I.

However, deletion of SLX8 alone does not fully restore the wild-type Rad21 and Rec8 axis localization in the spo76–1 mutant, indicating that additional factors, which remain to be identified, contribute to kleisins destabilization.

Crosstalk between cohesin components and axis proteins

The meiotic chromosome axis is composed of the structural proteins Red1 and Hop1 together with cohesin complexes including SMC proteins, kleisins, and associated regulators. Our analysis shows that in Sordaria, like in budding yeast and mice [49,52,54], Hop1 recruitment to the axes strictly depends on the prior association of Red1 with axes, whereas Red1 axis loading is independent of Hop1.

Communication among cohesin subunits is well established and critical for both the assembly and regulation of the cohesin complex. In Sordaria, deletion of WAPL results in persistent retention of cohesin components, premature recruitment of Sororin and Rad21, enhanced chromosome compaction, and defective synapsis. These phenotypes parallel the Wapl functions reported in budding yeast, plants, Caenorhabditis elegans, and mammals, indicating that its canonical role as a cohesin releasing factor is conserved in Sordaria [13,37,55–59]. Interestingly, our results extend the Wapl function beyond its well-established role at late prophase I. We found that Wapl already acts at leptotene, preventing the early recruitment/localization of Sororin and Rad21 on axes. Thus, Wapl exerts a dual control: first, by delaying the association of these components at early prophase I, and later by promoting cohesin release after pachytene. Whether these two activities rely on the same mechanism, limiting stable binding to axes, or whether Wapl interferes with Sororin and Rad21 initial loading through a distinct pathway remains to be determined.

In mammals, Sororin counteracts Wapl and stabilizes cohesin complexes during meiosis [14,21,36]. Consistently, in Sordaria, Sororin is required to maintain Spo76/Pds5 and Rec8 on chromosome axes during early prophase I, protecting them from premature Wapl-mediated release. By contrast, Red1 and Hop1 localization is unaffected in the absence of Sororin, indicating that Wapl specifically targets cohesins and not these two axis proteins.

The interplay between axial elements and cohesins is a central question in meiosis. In S. cerevisiae, Rec8-containing cohesin complexes are required for proper Red1 loading, while Red1 and Hop1 contribute to axis integrity and recombination control [49,60,61]. In mouse, SYCP2/SYCP3 (Red1 homologs) can localize independently of Rec8 [50]. These studies highlight that dependencies between axis proteins and Rec8 may vary across species, although the molecular basis of this crosstalk remains unclear. Beyond Rec8, little is known about the interactions between other cohesin subunits and axis proteins.

Our results identify a new protective pathway in which Red1 and Hop1 stabilize Spo76/Pds5 against Wapl-mediated release. In the absence of Hop1, Spo76/Pds5 is specifically lost in the subset of chromosomal regions that lack synaptonemal complex [41], but this loss is reversed when Wapl is deleted, indicating that Hop1 normally shields Spo76/Pds5 from Wapl activity. Likewise, Red1 is essential for Spo76/Pds5 axis association throughout prophase I, and again Wapl removal restores Spo76/Pds5 localization in the red1Δ mutant. These findings demonstrate that Red1 and Hop1 do not simply serve as structural axis components but counteract Wapl-mediated release, revealing a new layer of regulation in axis-cohesin crosstalk.

Moreover, our results indicate that Hop1, Red1, together with Sororin safeguard Spo76/Pds5 in a stage-specific manner: Sororin during early prophase, Hop1 at pachytene, while Red1 provides continuous protection throughout prophase I. This coordinated protection ensures proper cohesin retention and axis integrity (Fig 7A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

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.

https://doi.org/10.1371/journal.pgen.1012001.g007

Finally, we observed that the persistent retention of cohesins at late pachytene, in the wapl∆ mutant, correlates with the persistent presence of the axis proteins Red1 and Hop1. Moreover, restoration of full-length Spo76/Pds5 localization in the hop1Δ waplΔ double mutant is accompanied by a restoration of Red1 on chromosome axes. Since Wapl acts specifically on cohesins, this parallel recovery implies that axis protein maintenance is a downstream consequence of cohesin stability. Conversely, in the red1Δ waplΔ double mutant, cohesins remain bound to chromosome axes, yet the axis structure is severely disorganized, emphasizing Red1’s central role in meiotic chromosome architecture. These results suggest that Red1 acts as a key structural organizer, essential for proper axis shaping. Alternatively, cohesins may persist in a non-functional state, impaired in loop extrusion or cohesion, indicating that, without Red1, they cannot maintain axis integrity. Clarifying the relative contributions of Red1 and cohesins will be crucial for understanding the mechanisms underlying axis assembly during meiosis.

Spo76/Pds5 is a central regulator of cohesin’s stability in Sordaria

Our data reveal that in Sordaria, Spo76/Pds5 functions as a central guardian of cohesin stability during meiotic prophase I. In the red1Δ mutant, Wapl efficiently removes Spo76/Pds5 from chromosome axes but is insufficient to fully displace Rec8 and Rad21, suggesting that Spo76/Pds5 is a major target of Wapl activity.

Strikingly, in the red1Δ waplΔ double mutant, Spo76/Pds5 localization is restored and both kleisins exhibit substantially improved stability along chromosome axes throughout pachytene and beyond. Moreover, in the spo76–1 hypomorphic mutant, kleisins are destabilized independently of Wapl (in spo76–1 or in spo76–1 waplΔ). These findings indicate that (i) Wapl does not act directly on Rec8 or Rad21 but exerts its activities indirectly via Spo76/Pds5 (ii) Spo76/Pds5 functions as a protector of kleisins, whose stability is otherwise compromised in its absence. Deletion of SLX8 in the spo76–1 mutant improves the axis association of both kleisins, demonstrating that, Spo76/Pds5 safeguards cohesins from Slx8-dependent SUMO-ubiquitin-proteasome-mediated degradation during meiotic prophase I in Sordaria.

Notably, we show here that, like its mitotic counterpart Rad21, the meiotic kleisin Rec8, is subject to STUbL-mediated turnover during meiosis.

The protective role of Spo76/Pds5 may involve multiple mechanisms, especially, it could mask post-translational modification sites on kleisins, such as SUMOylation sites, or recruit deSUMOylating enzymes, thereby preventing recognition by Slx8 and subsequent proteasomal degradation. The partial restoration of kleisin stability in spo76–1 slx8Δ suggests that additional, yet unidentified, factors contribute to kleisin destabilization, reflecting a multilayered regulatory network.

Comparisons with other model systems

Our findings in Sordaria highlight both conserved and unique features of cohesin regulation during meiotic prophase. The distinct timing of Rec8 and Rad21 loading, as well as Wapl-dependent control of cohesin retention, are broadly conserved across budding yeast, C. elegans, and mammals, indicating a common framework for coordinating axis association and higher-order chromosome organization [37,55,59]. In addition, in Sordaria, like in mammals, Sororin antagonizes Wapl by promoting cohesin stabilization [21,36]. In contrast, budding yeast and C. elegans lack a canonical Sororin ortholog; in these organisms, regulation of Wapl relies on alternative mechanisms, such as Smc3 acetylation and axis-associated proteins [21–23]. Furthermore, the recruitment of Red1 and Hop1 to chromosome axes in the absence of Sororin, Spo76/Pds5, or Rec8 parallels the behavior of mammalian SYCP2/SYCP3 (Red1 orthologs), but contrasts with budding yeast, where Sororin is absent and Red1 loading is partially Rec8-dependent [44,59,60].

At the same time, Sordaria exhibits several distinct features: (i) a stage-specific, coordinated protection of cohesins by Sororin, Hop1, and Red1, (ii) a central role for Spo76/Pds5 as a mediator and major target of Wapl activity, and (iii) Spo76/Pds5 dependent safeguarding of kleisins from SUMO-ubiquitin-dependent proteasome degradation during meiosis. These comparisons emphasize that while the core principles of Wapl regulation and cohesin dynamics are widely conserved, Sordaria provides an intermediate and informative model for dissecting both conserved and lineage-specific strategies of cohesin regulation and axis organization during meiosis.

Concluding remarks

In summary, Sororin, together with the axis proteins Red1 and Hop1, protects Spo76/Pds5 from the Wapl-mediated release. In turn, the presence of Spo76/Pds5 on chromosome axes safeguards kleisins from proteasomal degradation, at least in part via the Slx8-dependent STUbL pathway (Fig 7B). This multilayered protection is essential for maintaining cohesin integrity and axis architecture throughout meiotic prophase I.

Given that mutants lacking cohesin and/or axis proteins mostly exhibit also DNA double-strand breaks (DSBs) defects (review in [2]), it remains possible that some of the severe phenotypes observed, particularly in the red1Δ mutant, reflect altered DSB metabolism in addition to cohesin defects. Future studies will be essential to disentangle the contributions of recombination, axis assembly, and cohesin regulation, and to define how these pathways converge to ensure faithful axis structure during meiotic progression.

Homolog identification, alignments, phylogeny and models

Homologs were identified by iterative PSI-BLAST searches at NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi) or by ProtBLAST/PSI-BLAST and HHMER searches at MPI (https://toolkit.tuebingen.mpg.de) as well as by Foldseek searches (https://search.foldseek.com/search). Multiple sequence alignments (MSAs) generated by MAFFT 7.0 (http://mafft.cbrc.jp/alignment/server), with the auto mode and default parameters, were used as inputs at MPI PSI-BLAST or HHMER websites. Foldseek analyses were done with 3D models from the AlphaFold protein structure database ((https://alphafold.ebi.ac.uk/). Protein modeling was done with the AlphFold3 (AF3) server (https://alphafoldserver.com/). The predicted template modeling (pTM) score and the interface predicted template modeling (ipTM) score were provided by the AF3 website. Protein-protein interaction confidence was evaluated with PIZSA (https://cospi.iiserpune.ac.in/pizsa/). Atomic structures and pLDDT values were visualized in Chimera X (https://www.cgl.ucsf.edu/chimera/).

SLX8. Sordaria Slx8 (SMAC_06687) is a 418 amino acids protein that displays the two characteristics shared by the STUbL family proteins: one SUMO Interacting Motif (SIM) (VVDL) located at 309–312aa, and a RING-type E3 ubiquitin ligase domain located closer to its C-terminus (339–379aa). The RING domains were identified by SMART (http://smart.embl-heidelberg.de/) or Uniprot (https://www.uniprot.org/) and the putative SIMs were predicted using Jassa (http://www.jassa.fr) and GPS-SUMO (https://sumo.biocuckoo.cn).

Deletions of the RED1, SORORIN, WAPL and SLX8 genes

The null waplΔ, red1Δ, sororinΔ and slx8Δ mutants were obtained by replacement of their respective entire open-reading frame by the Escherichia coli hph gene, conferring hygromycin resistance. The 5’ and 3’ flanking sequences were amplified (S1 Data for primers). Using the Gibson assembly method (NEBuilder HiFi DNA Assembly mix; New England Biolab), both fragments were assembled with pBC-Hygro plasmid, previously cleaved by XbaI. This three-partner assembly yielded a deletion plasmid containing a complete deletion of the CDS for each gene. Each plasmid was linearized at the unique XbaI site to generate homologous recombination ends and used to transform a ku70Δ mutant strain, known to increase the frequency of homologous integration events. Transformants carrying a null allele were selected for their hygromycin resistance, and the presence of the deleted allele was confirmed through PCR analysis and DNA sequencing. Further crosses with a wild-type KU70 strain eliminated the ku70Δ allele.

Construction of fluorescent GFP-tagged fusion proteins

The GFP coding sequence (p-EGFP-1, Clontech) was fused immediately after the last C-terminal amino acid predicted from the WAPL and SORORIN open reading frames and directly after the initiating methionine codon of the RED1 ORF. For each gene, the GFP fusion coding sequence (CDS) is under the control of its respective promoter. For the GFP-Red1 and Sororin-GFP fusions, a helix-forming linker (LAEAAAKEAAAKEAAAKAAA; Arai et al., 2001) was inserted between GFP and the target protein. In the Wapl-GFP fusion, the linker contains multiple glycine residues (GSGGGGSGGGGG). All constructs were generated using the Gibson assembly method (NEBuilder HiFi DNA Assembly Mix; New England Biolabs). For the Wapl-GFP fusion, the WAPL 5’ flanking region and CDS (excluding the stop codon) were amplified (S1 Data for primers) and assembled with pJET-GSGGGGSGGGGG-GFP, previously digested with BamHI and XhoI. For the Sororin-GFP fusion, the same approach was used: the SORORIN 5’ flanking region and CDS (excluding the stop codon) were amplified and assembled with pBC-LAEAAAKEAAAKEAAAKAAA-GFP, which had been digested with SalI. In both cases, the restriction sites are located upstream of the GFP CDS. The GFP-LAEAAAKEAAAKEAAAKAAA-RED1 allele was generated through a two-step Gibson assembly into the pBC-GFP-LAEAAAKEAAAKEAAAKAAA plasmid (pBC-SK (+), which contains the GFP-helix-forming linker coding sequence inserted at the EcoRV site). First, the 5’ promoter region of RED1, amplified by PCR (S1 Data for primers), was inserted upstream of the GFP CDS. In the second step, the RED1 CDS (including the stop codon) and the 3’ flanking region containing the terminator sequence were inserted downstream, in-frame with the GFP-LAEAAAKEAAAKEAAAKAAA CDS (S1 Data for primers used to amplify this fragment).

After validation by sequencing, GFP-tagged versions of the three genes were ectopically integrated into the WT strain by co-transformation with a plasmid carrying the hygromycin resistance cassette. All resulting strains showed wild-type sporulation.

Complementation experiments with GFP-tagged proteins

To perform complementation experiments, the Wapl-GFP, GFP-Red1 and Sororin-GFP genes were introduced into the corresponding null mutant strains (waplΔ, red1Δ and sororin∆ strains, respectively) by genetic crosses. For each cross, the offspring were genotyped to select waplΔ Wapl-GFP, red1Δ GFP-Red1 and sororin∆ Sororin-GFP strains.

Assessment of vegetative and sexual phenotypes

To evaluate vegetative growth, strains were inoculated onto M2 minimal medium plates and incubated for four days at 25°C until the plates were fully covered by the mycelium. Three plates were used for each strain and the mycelium growth was monitored twice daily by measuring the growth front. The development of protoperithecia and mature perithecia was monitored during ten days. Asci development and ascospore formation were assessed by dissecting perithecia after five days of growth. Perithecia were cracked open using forceps, and ascus bouquets were mounted between slide and coverslip for microscopic observation. Ten ascus bouquets, corresponding to the production of 10 perithecia, were imaged per strain, and the numbers of normal asci (containing eight black ascospores) and abnormal asci were determined. To assess germination efficiency, 80 ascospores were isolated from perithecia and inoculated on germination medium (15g/l BactoAgar, 15g/l Bacto Peptone, 0,06 M Ammonium Acetate). Germination was scored after 24 h at room temperature.

Cytology

GFP, Tomato-red (TdT) and DAPI (0.5g/ml) signals were observed after fixation in 4% paraformaldehyde. Images were acquired using a Leica DM6 B widefield microscope equipped with a K5 monochrome camera and the THUNDER imager system (Leica Microsystems Inc. Wetzlar, Germany). Thunder computational clearing, which uses Leica Microsystems’ proprietary algorithms was used to enhance image contrast and resolution. The public domain software ImageJ (http://rsb.info.nih.gov/ij) was used for images processing. We used the line tracing tool of the Image J software to measure the axis lengths.