Disruption of the UPS leads to elevated pro-crossover protein levels in C. elegans

To systematically investigate the potential targets of the UPS in the germline of C. elegans, we adopted a data-independent acquisition (DIA) proteomics strategy. Germline-specific RNAi of uba-1, the sole E1-encoding gene in C. elegans, was employed to inhibit polyubiquitination initiation (Fig 1A and 1B) [45,46]. The resulting up-regulated proteins were then identified and analyzed compared with control RNAi. The proteomics data revealed a ~60% reduction in UBA-1 abundance, demonstrating the efficacy of our germline-specific RNAi (S1A Fig). Meanwhile, many known UPS targets, including the cyclins and WEE-1, were up-regulated (S1 Table), showing that the function of the UPS was indeed impaired. Biological Processes enrichment reveals that proteins functioning in meiotic cell cycle are enriched (S1B Fig). Among the enriched meiotic proteins are the evolutionarily conserved pro-crossover factors, such as COSA-1, ZHP-3, and CDK-2 (Figs 1C, 1D, and S1A; S1 Table), as well as BRC-1 and BRD-1 (Fig 1C and 1D), which were found to concentrate at crossover-designated sites, colocalizing with pro-crossover factors [47].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Knockdown of the UPS led to increased abundance of pro-crossover proteins.

(A) Schematic of the ubiquitin-proteasome system. E1 and RPN-6 were targeted by RNAi in this study. E1, ubiquitin-activating enzyme; E2, ubiquitin-conjugating enzyme; E3, ubiquitin ligase; CP, core particle; RP, regulatory particle. (B) Schematic of the germline of Caenorhabditis elegans and the workflow of DIA proteomics. (C) Volcano plot showing changes in protein level after germline-specific uba-1 RNAi compared with ctrl RNAi. Dots representing UBA-1, BRC-1, BRD-1, CDK-2, COSA-1 and ZHP-3 are highlighted. The volcano plot was executed with R package ggplot2. (D) Table listing the fold change, P value, and false discovery rate (FDR) for proteins highlighted in (C). (E) Western blot showing changes in protein level of COSA-1::3 × HA and ZHP-3::AID::3 × HA upon ctrl, rpn-6, and uba-1 RNAi treatments. (F) Immunofluorescence images of dissected gonad from cosa-1::3 × HA treated with ctrl, rpn-6, or uba-1 RNAi, stained for DNA (DAPI, blue), HIM-3 (magenta), and COSA-1 (green). (G) Quantification of immunofluorescence intensity of COSA-1 in (F). (H) Immunofluorescence images of dissected gonad from zhp-3::AID::3 × HA, treated with ctrl, rpn-6, or uba-1 RNAi, stained for DAPI (blue), HIM-3 (magenta), and ZHP-3 (green). (I) Quantification of immunofluorescence intensity of ZHP-3 in (H). (J) Immunofluorescence images of dissected gonad from ollas::cosa-1::3 × FLAG; spo-11, treated with ctrl, rpn-6, or uba-1 RNAi, stained for DAPI (blue), HIM-3 (magenta), and COSA-1 (green). (K) Quantification of immunofluorescence intensity of COSA-1 in (J). Signals from the distal tip of the gonad to the end of pachytene were quantified. n ≥ 3. The means ± SD are shown. P values are given by ordinary one-way ANOVA. All scale bars, 100 μm. The underlying data for Fig 1C can be found in S1 Table, and for Fig 1G, 1I, and 1K in S1 Data.

https://doi.org/10.1371/journal.pbio.3003868.g001

To validate the up-regulation of the pro-crossover factors, we performed immunoblotting on whole worm lysates treated with either control RNAi, uba-1 RNAi, or rpn-6.1 RNAi (hereafter referred to as rpn-6 RNAi). The latter is a subunit of 19S proteasome and has been used in a previous study to decrease proteasome function (Fig 1A) [48]. We found that pro-crossover proteins, such as COSA-1, MSH-5, and ZHP-3, were difficult to be detected in whole worm lysates, likely due to low abundance (Figs 1E, S1C, and S1D). Strikingly, following uba-1 and rpn-6 RNAi treatments, these proteins became readily detectable, such as MSH-5 (knock-in GFP::MSH-5 in S1C Fig), COSA-1 (detected by both knock-in HA-tagged in Fig 1E and transgenic GFP::COSA-1 in S1D Fig), ZHP-3 (knock-in HA-tagged in Fig 1E), and HIM-6 (knock-in eGFP::HIM-6 in S1C Fig), indicating a significant increase in their levels. Notably, the level of a GFP-tagged SC component SYP-3 remained unchanged (S1E Fig), suggesting that pro-crossover factors might be preferentially targeted over other meiotic proteins. Immunofluorescence analysis was then performed on dissected germline to further confirm the increase in protein abundance. Consistent with the immunoblotting results, a significant increase in COSA-1 and ZHP-3 was observed upon depletion of uba-1 and rpn-6 by RNAi (Fig 1F–1I). Notably, both transgenic overexpression lines (using promoters different from the endogenous ones) and endogenous knock-in alleles of cosa-1 exhibited identical trends in protein accumulation, strongly suggesting that the observed effects are mediated through post-translational regulation. This conclusion is further supported by RNA-seq data, which showed no significant alterations in the mRNA levels of known pro-crossover factors (S1F Fig and S2 Table).

We next set out to investigate the genetic requirement(s) for the up-regulation of pro-crossover factors. Proteolysis of pro-crossover factors has been proposed to facilitate crossover/non-crossover determination [33,42]. For instance, yeast Msh4 phosphorylation at crossover sites prevents its proteasomal degradation [42]. If the mechanism is conserved in C. elegans, the increase in pro-crossover protein levels following disruption of the UPS could be due to the non-degraded pro-crossover proteins from the non-crossover sites. To test this hypothesis, we reasoned that a DSB-deficient background, in which crossover designation cannot occur, would eliminate the need for the proposed proteolytic regulation of pro-crossover factors. We generated a spo-11 null mutant using CRISPR-Cas9 to block endogenously programmed DSB induction [49], as evidenced by the presence of 12 univalents in the diakinesis oocytes (S1G Fig). In uba-1 or rpn-6 RNAi treated spo-11 mutants, the increase in COSA-1 protein levels still occurred (Fig 1J and 1K), suggesting that UPS-mediated regulation of pro-crossover factors operates independently of SPO-11-dependent DSB formation.

Crossover designation is maintained despite depletion of UBA-1 or RPN-6

Our data demonstrate that the UPS regulates protein level of the pro-crossover factors. To further characterize the impacts of UPS knockdown beyond protein abundance, we carefully examined pro-crossover protein signals from meiotic entry to late pachytene, when CO designation happens (S2A Fig). Consistent with previous studies, knockdown of the UPS through rpn-6 or uba-1 RNAi led to the formation of polycomplexes containing the chromosomal axis component HIM-3 [32], along with shortened or even absent TZ region, which normally contains crescent-shaped nuclei (white arrows and yellow arrowheads in S2A Fig) [41]. In these knockdown conditions, elevated levels of COSA-1 and ZHP-3 were already detectable during early meiosis, primarily as diffuse signals in the nucleoplasm without forming distinct foci (S2A Fig, TZ and EP). Nevertheless, COSA-1 and ZHP-3 foci coalesced normally during mid-to-late pachytene, comparable to control (S2A Fig). Using an endogenous mCherry-tagged allele of cosa-1, cosa-1::mCherry [50], we observed a similar trend regarding COSA-1 dynamics, where the signal of COSA-1 is initially nucleoplasmic and later transitions to foci along the chromosomes. Notably, even though COSA-1 foci exhibited increased brightness and size upon rpn-6 or uba-1 RNAi, most likely due to elevated protein levels (Fig 2A), this enhancement was not associated with premature appearance of COSA-1 foci (Fig 2A, dotted white line denotes the LP region with COSA-1::mCherry foci), indicating that crossover designation timing is tightly controlled despite UPS impairment.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Knockdown of the UPS led to increased intensity and focus the number of pro-crossover protein.

(A) Fluorescence images of a portion of gonad from cosa-1::mCherry worms treated with ctrl, rpn-6, or uba-1 RNAi, extending from early pachytene through early diplotene. Note the overall increase in nucleoplasmic COSA-1::mCherry and COSA-1::mCherry focus intensity following rpn-6 and uba-1 knockdown. (B) Immunofluorescence images of diplotene nuclei from cosa-1::3 × HA worms treated with ctrl, rpn-6, or uba-1 RNAi, stained for DNA (blue), HTP-1 (magenta), and SYP-1 (green). (C) Fluorescence images of late pachytene nuclei from cosa-1::mCherry; GFP::msh-5 worms treated with ctrl, rpn-6, or uba-1 RNAi, showing COSA-1::mCherry (magenta) and GFP::MSH-5 (green). Individual nuclei were outlined by dotted circle. (D) Enlargements of two nuclei marked by the yellow star in (C). Arrowheads point to GFP::MSH-5 foci not overlapping with COSA-1::mCherry. (E) Immunofluorescence images of late pachytene nuclei from cosa-1::3 × HA worms treated with ctrl, rpn-6, or uba-1 RNAi, stained for DAPI (blue), HIM-3 (magenta), and COSA-1 (green). (F) Quantification of the percentage of late pachytene nuclei with the indicated number of COSA-1::3 × HA foci as shown in (E). P values are given by ordinary one-way ANOVA. Scale bars, 50 μm for (A), 5 μm for (B), (C), and (E), 1 μm for (D). The underlying data for Fig 2F can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003868.g002

We also carefully examined COSA-1 foci formation in the spo-11 mutant background. In agreement with previous studies, 1–2 COSA-1 foci were observed within the oocyte nuclei at late pachytene in spo-11 mutants (S2B and S2C Fig). Following rpn-6 RNAi, the number of foci remained largely unchanged, but their intensity was significantly increased, presumably due to elevated COSA-1 protein abundance (S2B and S2C Fig). Interestingly, uba-1 RNAi led to a marked increase in the number of COSA-1 foci, indicating a fundamental divergence in how ubiquitination and the proteasome pathway regulate COSA-1 localization in this background.

Crossover designation induces the asymmetric distribution of certain SC-related proteins, such as the axial element HTP-1 and the central element SYP-1, leading to chromosome remodeling [51]. We found that the knockdown of the UPS did not result in any significant changes in the protein levels of HTP-1 and SYP-1 (S3A–S3C Fig), but diplotene chromosome morphology appears altered, such as shortened or disorganized axes, as indicated by HTP-1 staining (Fig 2B). We speculate that this is related to the previously reported role of the UPS in regulating the dynamics of these proteins [32]. Nevertheless, remodeling still occurred following RPN-6 or UBA-1 depletion (Fig 2B), albeit with some deviations from wild type, further demonstrating that crossover designation remains intact.

COSA-1 and MSH-5 co-accumulate at crossover designation sites during late pachytene, exhibiting interdependent chromosomal localization [21]. To simultaneously monitor their dynamics, we generated a cosa-1::mCherry; GFP::msh-5 double-fluorescent strain [47]. Following UPS knockdown, while most COSA-1 foci colocalized with MSH-5, persistent MSH-5 foci lacking COSA-1 were consistently observed in late pachytene (Fig 2C and 2D, and arrowheads in Fig 2D). By quantification, we found the existence of both MSH-5-only and COSA-1-only foci, with the former being more frequent (S3D and S3E Fig). These persistent single-stained foci may represent either unprocessed recombination intermediates or non-degradable protein aggregates. These results suggest that the UPS may be involved in clearing pro-crossover proteins that have accumulated at non-crossover sites under physiological conditions. Given that MSH-5 accumulates at recombination intermediates earlier and forms more foci than COSA-1 [25,29], it presents more potential targets for UPS-mediated clearance, which could explain the higher frequency of MSH-5-only foci. Since the majority of COSA-1 foci co-localized with MSH-5 (S3D and S3E Fig), we next examined whether increased pro-crossover abundance is accompanied by an increase in crossover designation by quantifying focus number of COSA-1 per nucleus. Nuclear spreading immunofluorescence demonstrated significantly elevated number of COSA-1 focus (>6/nucleus; Fig 2E and 2F) after UPS depletion, indicating defective crossover interference.

E1-mediated ubiquitination is required for the dissociation of pro-crossover proteins from diakinesis chromosomes

Crossover designation does not necessarily mean successful crossover formation. Cross-species studies have shown that ubiquitination-dependent proteasomal degradation mediated by specific E3s actively promotes crossover formation [27,28,34,36–39,52]. Therefore, impairment of the UPS function would be supposed to disrupt crossover formation. In C. elegans, defects in crossover formation could be easily detected by examining the karyotypes of individual diakinesis-stage oocytes. In wild type, six DAPI-stained bodies are visible, representing six pairs of homologs linked by chiasmata, the cytological manifestations of crossover (Fig 3A). The appearance of individual chromosomes (univalents), or more than six DAPI-stained bodies, usually indicates a defect in crossover formation. To directly check the impact of UPS impairment on crossover formation, we examined bivalent formation in diakinesis. Univalent was never observed under rpn-6 or uba-1 knockdown conditions, implying efficient crossover formation (Fig 3B). Intriguingly, we observed that the pro-crossover factors COSA-1 and ZHP-3 remained associated with the chromosomes in diakinesis oocytes (Fig 3B), a pattern reminiscent of mutants defective in resolution of meiotic recombination intermediates [14].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Knockdown of E1 but not the proteasome led to persistent association of pro-crossover proteins with diakinesis chromosomes.

(A) Schematic drawing of meiotic events during oogenesis in the Caenorhabditis elegans germline. (B) Immunofluorescence images of diplotene (DP) and diakinesis (DK) nuclei from cosa-1::3 × HA or zhp-3::AID::3 × HA worms treated with ctrl, rpn-6, or uba-1 RNAi, stained for DAPI (blue), HIM-3 (magenta), and COSA-1 or ZHP-3 (green). (C) Schematic of the composition of the eukaryotic 26S proteasome. PBS-6 and RPN-6 are indicated. (D) Quantification of the percentage of diakinesis nuclei with chromosome-associated COSA-1::3 × HA foci for ctrl RNAi, rpn-6 RNAi, pbs-6 RNAi, rpn-6 + pbs-6 double RNAi, and uba-1 RNAi. (E) Immunofluorescence images of diakinesis (DK) nuclei from transgenic GFP::cosa-1 and zhp-3::GFP worms treated with ctrl, pbs-6, or uba-1 RNAi, stained for DAPI (blue), HIM-3 (magenta), and COSA-1 (green) or ZHP-3 (green). (F) Immunofluorescence images of DP and DK nuclei from cosa-1::3 × HA worms treated with ctrl RNAi, rpn-6 RNAi, and rpn-6 + pbs-6 double RNAi, stained for DAPI (blue), HIM-3 (magenta), and COSA-1 (green). All scale bars, 5 μm. The underlying data for Fig 3D can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003868.g003

During normal meiotic progression to diakinesis, pro-crossover proteins gradually dissociate from crossover sites, becoming undetectable in the majority of −3 and −2 and all −1 diakinesis oocytes [14] (Fig 3A and 3B). However, persistent chromosome-associated COSA-1 and ZHP-3 foci could be detected in −1 diakinesis oocytes following uba-1 knockdown (Fig 3B). In rpn-6 RNAi-treated oocytes, although large and bright chromosome-associated COSA-1 and ZHP-3 foci were readily detected in diplotene, a large proportion of these foci vanished as the oocyte progressed into −1 diakinesis (compare rpn-6 RNAi with uba-1 RNAi in Fig 3B and 3D). The remnant foci of COSA-1 and ZHP-3 also exhibited distinct localization pattern, where most foci were in the nucleoplasm of the diakinesis oocytes (Fig 3B, arrowheads), reflecting that COSA-1 and ZHP-3 complexes could be dissociated from chromosomes but were unable to be degraded. The persistent chromatin association of ZHP-3 and COSA-1 in uba-1 RNAi-treated oocyte was not due to a difference in protein levels, since both COSA-1 and ZHP-3 were similarly increased upon rpn-6 or uba-1 knockdown. Considering that RPN-6 belongs to the 19S regulatory particle lid complex, we then tested pbs-6 RNAi, which targets the 20S proteasome’s catalytic core (Fig 3C). Similar to rpn-6, pbs-6 knockdown led to upregulation of COSA-1, as revealed by western blot and IF (S4A–S4C Fig), but no chromosome-associated COSA-1 could be detected in −1 diakinesis oocytes either (Fig 3D).

We next used GFP-tagged transgenic cosa-1 and zhp-3 worms to further validate these findings. A similar pattern was observed, whereby the dissociation of COSA-1 and ZHP-3 from chromosomes was exclusively delayed upon uba-1 knockdown but not pbs-6 knockdown (Figs 3E and S4D). Considering the distinct effects of disrupting ubiquitination (uba-1) and proteasome (rpn-6, pbs-6) on the dissociation of the pro-crossover proteins, we performed a double knockdown by combining rpn-6 and pbs-6 in a single RNAi vector to test whether simultaneously disrupting both the regulatory particle and the catalytic core of the proteasome could block chromosomal dissociation of these factors. However, knocking down both rpn-6 and pbs-6 did not enhance the pro-crossover protein retention phenotype as compared with single RNAi; instead, most of the observed foci were either nucleoplasmic or chromosome-adjacent (Fig 3D and 3F).

Together, the above data indicate that ubiquitination, but probably not proteasomal degradation, is responsible for the dissociation of pro-crossover proteins from diakinesis chromosomes. This mechanism is reminiscent of how polyubiquitination directs the disassembly of stable complexes, such as the replisome (see the Discussion for further details).

The CDC-48^UFD-1/NPL-4 segregase is involved in the dissociation of pro-crossover proteins from meiotic chromosomes

The CDC-48^{UFD−1/NPL−4} segregase is best known for its function in the recognition and triaging of poly-ubiquitinated substrate(s) for proteasomal degradation or recycling [53,54] (schematic in Fig 4A). We therefore tested the hypothesis that a similar ubiquitin-segregase relay is involved in the process of crossover formation. To test this hypothesis, we first performed RNAi experiments targeting cdc-48 and npl-4. Similar to uba-1 RNAi treatment, cdc-48 or npl-4 depletion led to significantly persistent association of COSA-1 and ZHP-3 foci with diakinesis chromosomes, as foci could be detected even in −1 diakinesis oocytes (Figs 4B–4E and S5A–S5C). This is consistent with the idea that CDC-48 and NPL-4 function within the same pathway. Since npl-4 RNAi produced a stronger effect than cdc-48 RNAi (Figs 4C and S5A–S5C), we subsequently used npl-4 alone to represent the function of the underlying CDC-48^UFD-1/NPL-4 segregase complex. We also noticed an evident phenotypic difference between npl-4 RNAi and uba-1 RNAi when COSA-1::HA was monitored (20% versus 100%, Figs 3D and 4C). We reason that this likely reflects limitations in detection sensitivity, because when npl-4 RNAi was performed in transgenic GFP::cosa-1 and zhp-3::GFP worms, the penetrance of the delayed dissociation phenotype (70%–100%) (S5B and S5C Fig) approached that observed in uba-1 RNAi (100%) (S4D and S5B Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. The CDC-48^UFD-1/NPL-4 segregase is required for stage-specific dissociation of pro-crossover proteins from chromosomes.

(A) Schematic showing the fate of a polyubiquitinated substrate protein with the CDC-48^UFD-1/NPL-4 segregase. DUB, deubiquitinase. (B) Immunofluorescence images of diakinesis (−1, −2, and −3) nuclei from cosa-1::3 × HA worms treated with ctrl, cdc-48, or npl-4 RNAi, stained for DAPI (blue), HIM-3 (magenta), and COSA-1 (green). (C) Quantification of the percentage of diakinesis nuclei with COSA-1::3 × HA foci for (B). (D) Immunofluorescence images of diakinesis nuclei from zhp-3::AID::3 × HA worms treated with ctrl RNAi or npl-4 RNAi, stained for DAPI (blue), HIM-3 (magenta), and ZHP-3 (green). (E) Quantification of the percentage of diakinesis nuclei with ZHP-3::AID::3 × HA foci for (D) and (F). (F) Immunofluorescence images of −1 diakinesis nuclei from zhp-3::AID::3 × HA worms treated with ctrl RNAi, rpn-6 RNAi, or rpn-6 + npl-4 double RNAi, stained for DAPI (blue), HIM-3 (magenta), and ZHP-3 (green). All scale bars, 5 μm. The underlying data for Fig 4C and 4E can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003868.g004

Unlike uba-1 RNAi, npl-4 depletion did not result in an increase in COSA-1 or ZHP-3 (S5D–S5G Fig), even though both cdc-48^ufd-1/npl-4 and uba-1 knockdown resulted in delayed dissociation of pro-crossover factors at diakinesis. Therefore, in contrast to ubiquitination, CDC-48^UFD-1/NPL-4 specifically regulates the dissociation of pro-crossover factors from diakinesis chromosomes.

Based on these results, we propose a linear dissociation cascade with sequential steps: pro-crossover proteins are first polyubiquitinated, which leads to their recognition and displacement from chromosomes by the CDC-48^UFD-1/NPL-4 complex; this displacement is then followed by the final degradation by the proteasome. Next, in order to further test if this displacement step is dependent on CDC-48^UFD-1/NPL-4, we took advantages of the observation that knockdown of rpn-6 alone didn’t impair the dissociation of pro-crossover proteins. We would expect the phenotype to revert from the rpn-6 RNAi phenotype (non-chromosome-associated foci) to resemble the uba-1 or npl-4 RNAi phenotype (chromosome-associated foci). We performed double knockdown of rpn-6 and npl-4 and observed persistent chromosome-associated ZHP-3 foci in −1 diakinesis oocytes (Fig 4E and 4F), indicating npl-4 is epistatic to rpn-6 in the dissociation of pro-crossover factors. Collectively, our data establish that the CDC-48^UFD-1/NPL-4 segregase is required for efficient dissociation of pro-crossover complex from diakinesis chromosomes. However, we cannot formally exclude the possibility that another segregase is also involved.

UPS knockdown rescues the crossover defect in a hypomorphic allele of cosa-1

The persistent association of pro-crossover factors with meiotic chromosomes observed above could be explained by an alternative possibility that late recombination intermediates were not processed into mature crossovers after uba-1 or npl-4 knockdown, as observed for slx-4 mutants [14]. To further establish the role of the UPS in pro-crossover factor regulation and crossover formation, we introduced a crossover-deficient condition specifically induced by mutation in the pro-crossover factor COSA-1. P51 and D52 are two amino acids in COSA-1 that are mutated in the cosa-1-6A allele but not in cosa-1-4A, two alleles generated in a recent study from our lab [14]. Because cosa-1-6A exhibits more severe meiotic defects than cosa-1-4A, a double-alanine substitution mutant (P51A, D52A) was generated to assess the specific functional contribution of these residues, and this allele was named cosa-1^PD (Fig 5A). Like COSA-1-4A and COSA-1-6A [14], COSA-1^PD exhibited reduced binding affinity for other pro-crossover proteins such as MSH-5, while its interaction with CDK-2 was maintained in yeast two-hybrid assays (Fig 5B). Homozygous cosa-1^PD worms showed partial crossover defects, as evidenced by an average of 8.54 DAPI-stained bodies in diakinesis oocytes (Fig 5C and 5D). In contrast to other cosa-1 mutants [14], cosa-1^PD could be homozygous viable and fertile, with normal brood size, reduced progeny viability, and increased incidence of male offspring (Fig 5E–5G), making this allele ideal for probing UPS-dependent crossover regulation.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Knockdown of the UPS and NPL-4 rescued chromosomal accumulation of COSA-1 and bivalent formation in cosa-1^PD.

(A) Schematic of the primary structure of COSA-1. The cyan rectangle represents the N-terminus of COSA-1, which is predicted to contain an intrinsically disordered region (IDR) responsible for interactions with other pro-crossover factors. P51 and D52 are underlined and highlighted in green. (B) Yeast two-hybrid assay showing that COSA-1^PD can bind efficiently to CDK-2, but hardly binds to MSH-5. (C) Fluorescence images of diakinesis nuclei stained for DAPI, showing bivalent formation in ollas::cosa-1^PD::3 × FLAG worms treated with ctrl RNAi, 30% rpn-6 RNAi, 20% uba-1 RNAi, or npl-4 RNAi. (D) Quantification of the percentage of nuclei with the number of DAPI-stained bodies indicated in (C). (E–G) Quantification of brood size (E), progeny viability (F), and male progeny frequency (G) for ollas::cosa-1::3 × FLAG and ollas::cosa-1^PD::3 × FLAG, respectively. The means ± SD are shown. P values are given by t test. (H) Immunofluorescence images of late pachytene nuclei of the indicated genotypes stained for DAPI (blue), HIM-3 (magenta), and COSA-1 (green). (I) Quantification of the percentage of late pachytene nuclei with indicated number of COSA-1 or COSA-1^PD foci as shown in (H). Quantification for rpn-6 and uba-1 single RNAi treatment in the wild-type background was also included. P values are given by ordinary one-way ANOVA. All scale bars, 5 μm. The underlying data for Fig 5D–5G and 5I can be found in S1 Data.

https://doi.org/10.1371/journal.pbio.3003868.g005

Depletion of UBA-1 was thus performed in cosa-1^PD mutants. uba-1 partial RNAi was used to obtain well resolved and separated chromosomes for quantification. After 20% uba-1 RNAi treatment (diluted 1:4 with control RNAi), we observed partial rescue of bivalent formation defect in cosa-1^PD mutants, with an average of 6.47 DAPI-stained bodies compared with 8.54 in untreated controls (Figs 5C, 5D, and S6A). This argues against the idea that the persistent pro-crossover protein foci marked unresolved recombination intermediates. Instead, it strongly suggests that crossovers were efficiently generated upon uba-1 knockdown. Partial knockdown of rpn-6 (30%) also reduced the number of DAPI-stained bodies in cosa-1^PD, implying that the UPS negatively regulates crossover formation (Fig 5C and 5D).

In cosa-1^PD mutants, COSA-1 foci are barely detectable in late pachytene (Fig 5H), even though crossover designation still seems to happen as manifested by the formation of 6 MSH-5 foci at late pachytene (S6B and S6C Fig). We therefore investigated whether the uba-1 and rpn-6 RNAi-mediated rescue of bivalent formation involves restoration of COSA-1 localization and accumulation. Strikingly, knockdown of the UPS by either rpn-6 or uba-1 RNAi significantly recovered the number of COSA-1 foci (7.5 foci for uba-1 RNAi and 6.05 foci for rpn-6 RNAi) (Fig 5H and 5I). These findings strongly suggest that the UPS restricts crossover formation by controlling the stability and chromosomal retention of pro-crossover factors at the crossover designation sites.

We further used cosa-1^PD to test if the CDC-48^UFD-1/NPL-4 segregase also regulates crossover formation. Intriguingly, npl-4 RNAi was also able to partially rescue bivalent formation in cosa-1^PD mutants (Fig 5C and 5D), suggesting that it likewise participates in restricting crossover formation. However, unlike uba-1 and rpn-6, knockdown of npl-4 did not restore the accumulation of COSA-1 during late pachytene, indicating that the UPS may have a broader functional scope than the CDC-48^UFD-1/NPL-4 segregase (S6D Fig).

The UPS enforces crossover interference in C. elegans

Only one crossover per homolog pair is normally generated during meiosis in C. elegans, representing an extreme example of crossover interference [55]. The above data therefore indicates a role for the UPS in crossover interference. To directly test the involvement of the UPS in crossover interference, we measured the recombination rates on chromosome II using a recently established Insertion-Deletion-Polymorphisms method [56]. uba-1 RNAi increased both overall crossover frequency and the incidence of double crossovers and triple crossovers. Although the difference in the frequency of multiple crossovers between control RNAi and uba-1 partial RNAi does not reach statistical significance (Fisher’s exact test, one-tailed, P = 0.065), the presence of more than one crossover strongly supports the hypothesis that the ubiquitin system plays a critical role in restricting crossover designation and formation (Fig 6A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Knockdown of the UPS attenuated crossover interference.

(A) Crossover frequency and distribution on chromosome II of the indicated genotypes with different RNAi treatments were evaluated. Red-highlighted values denote the sum of the 5 intervals. n is the number of total progenies scored. The frequency of triple crossovers, double crossovers, single or zero crossover per chromosome is also indicated. (B) Crossover distribution of double crossover and triple crossover events detected. Blue indicates the Bristol N2-derived allele, and yellow indicates the Hawaiian-derived allele. The number indicates how many times double crossovers occurring at the same location were detected.

https://doi.org/10.1371/journal.pbio.3003868.g006

The recombination rates were next assessed in the cosa-1^PD background. Consistent with the presence of univalent at diakinesis and the HIM (High incidence of Male progeny) phenotype, cosa-1^PD exhibited a reduced recombination rate (Fig 6A, 0.426 versus 0.470). However, double crossovers were detected in cosa-1^PD, indicative of impaired crossover interference (Fig 6A). Partial uba-1 RNAi rescued the reduction in recombination rate of cosa-1^PD, consistent with the restrictive role of ubiquitination in crossover control (Fig 6A). Intriguingly, uba-1 RNAi in cosa-1^PD caused more severe crossover interference defects compared with equivalent uba-1 knockdown in the wild-type background, whereby the incidence of double and triple crossovers was further elevated (Fig 6A, compare the effect of 20% uba-1 RNAi treatment in N2 with cosa-1^PD). This is in agreement with our observations that a subset of nuclei in cosa-1^PD mutants contained more than 9 COSA-1 foci upon uba-1 RNAi, a number never reached in wild type with the same treatment (Fig 5I). Consistent with the formation of multiple crossovers, ring-like bivalents could be detected in both the wild-type and cosa-1^PD diakinesis oocyte following uba-1 RNAi (S7 Fig, yellow arrows). Notably, the frequency of multiple crossovers is significantly higher in the uba-1 partial RNAi group than control in the cosa-1^PD background (Fisher’s exact test, one-tailed, P = 0.0063), reinforcing the idea that the ubiquitin system restricts crossover formation. Given that COSA-1^PD exhibits severely impaired binding to other pro-crossover factors, such as MSH-5, these results suggest that the interaction between COSA-1 and MSH-5 may contribute to the enhancement of crossover interference regulation. A preliminary analysis of the available double and triple crossover events revealed no consistent spacing preference, which may indicate a fundamental breakdown of crossover spacing constraints following UBA-1 depletion (Fig 6B). This disruption of spatial patterning provides further evidence that the UPS ordinarily enforces crossover interference.

The UPS and the synaptonemal complexes (SC) differ in their mechanism underlying crossover interference

The synaptonemal complexes (SC) has long been implicated in crossover interference [57,58]. Since disruption of the UPS was previously shown to lead to defects in SC dynamics [32,33], the impacts of UPS knockdown on crossover designation and formation reported here might be simply attributed to the role of the UPS in synapsis. As the synaptonemal complex central region (SC-CR) proteins (e.g., SYP-1, -2, -3, -4, -5, or -6) are essential for crossover formation in C. elegans, it is not feasible to evaluate the function of the UPS in crossover patterning in mutants completely lacking any SYP protein. However, previous studies have shown that partial depletion of the SC central element, such as SYP-1, can attenuate crossover interference without compromising its crossover-promoting function of the SC [6,59–61]. Given that UPS knockdown in this study similarly impairs crossover interference, we compared partial depletion of SYP-1 with that of UBA-1 for their impacts on crossover formation in the cosa-1^PD background, attempting to uncover an SC-independent function of the UPS in crossover regulation.

We first evaluated syp-1 partial RNAi efficacy and found that it recapitulated what had been used to demonstrate the role of the SC in limiting crossovers: the TZ zone was extended, as would be expected if SC-CR function is disrupted (S8A and S8B Fig); a significant increase in the number of COSA-1 foci at late pachytene (average 7.42, S8C and S8D Fig) and six bivalents in diakinesis (S8E Fig) were observed, indicating defective crossover interference albeit successful formation of chiasmata for all six chromosome pairs [6]. In contrast to knockdown of uba-1 (Fig 5C, 5D, 5H, and 5I), partial depletion of SYP-1 could neither recover COSA-1 foci at late pachytene (S8F Fig), nor rescue bivalent formation of cosa-1^PD (S8G and S8H Fig). Instead, syp-1 partial depletion exhibited strong genetic interaction with cosa-1^PD in bivalent formation: the crossover formation defect was exaggerated as suggested by the increased incidence of univalents (S8G and S8H Fig). Together, our genetic evidence strongly suggests that the UPS mediates crossover control through mechanisms separable from its regulation of SC dynamics. However, we cannot exclude that synapsis defects may still partially contribute to the crossover patterning changes observed upon UPS disruption.

Meiotic proteins are direct targets of ubiquitination

The data above demonstrate that the UPS regulates the dynamics of pro-crossover proteins during meiosis prophase. Although the pro-crossover factor ZHP-3 itself is a putative E3 ligase [18,62], ubiquitination of pro-crossover proteins remains poorly characterized, likely due to their low endogenous abundance. To determine whether pro-crossover factors undergo direct ubiquitination, nuclear extracts were prepared following germline-specific rpn-6 RNAi, in order to enrich for potential polyubiquitinated proteins undergoing proteasomal degradation, and the ubiquitin-modified proteome was analysed by mass spectrometry. Consistent with previous findings that ubiquitin and proteasome localize along meiotic chromosome axes across species [32,33], SMC proteins, axis components, SC lateral components, and SC central components were found to undergo ubiquitination at multiple sites (Fig 7A and S3 Table). Intriguingly, pro-crossover proteins ZHP-3 and CDK-2 were found to possess reliable ubiquitin-modified residues (Fig 7A and 7B; S3 Table). Other meiotic proteins known to accumulate at the CO-designated sites during late pachytene, such as RMH-1 and BRD-1, and proteins localizing specifically on the short arm at late prophase, such as ZHP-1, ZHP-2, and PLK-2, are also subject to ubiquitination (Fig 7A and 7B; S3 Table). No ubiquitination site was identified for COSA-1, likely reflecting either the absence of ubiquitination for COSA-1 or its low abundance. Notably, the identified ubiquitinated proteins are not always accompanied with an increase in their abundance (Fig 7A), suggesting that these ubiquitination events may involve non-K48-linked poly-ubiquitin chains or mono-ubiquitination. While the functional consequences of these ubiquitination events require further investigation, these findings provide direct evidence that the crossover recombination repair machinery is broadly subject to ubiquitin-mediated regulation, and further support the hypothesis that the UPS directly regulates pro-crossover protein homeostasis through ubiquitination.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Pro-crossover proteins are subject to ubiquitination.

(A) Meiotic proteins closely related to crossover designation were identified as being subject to ubiquitination. Data from the DIA proteomics study are included when available. Fold change and P value are for the DIA proteomics based on three biological replicates. NA, not applicable. ND means the protein was not detected by the DIA MS. Ubiquitination sites were verified based on the extracted ion chromatogram. (B) Secondary mass spectrum showing evidence of ubiquitination modification(s) for ZHP-3 (two sites), CDK-2, and RMH-1. The peptide was broken into charged fragments. The fragments to the left are named as b ions, and the fragments to the right as y ions. The detected fragments are marked by blue b or red y followed by the number of the fragment’s amino acid. m/z for the detected fragment is shown on the plot, and the theoretical m/z is shown on the side. The detected value is almost equal to the theoretical value when the lysine (K) in the fragment is not ubiquitinated. The red-highlighted theoretical values are 114 Da (the molecular weight of Gly-Gly) less than the detected ones, indicative of ubiquitination. m/z, mass to charge ratio; + , single charged; ++, double charged. The underlying data for this figure can be found in S1 and S3 Tables.

https://doi.org/10.1371/journal.pbio.3003868.g007

C. elegans strains and maintenance

All strains in this study were maintained on NGM (nematode growth medium) plates seeded with OP50 bacteria. All experiments were performed at 20°C. All C. elegans strains were derived from a Bristol N2 background. S4 Table summarizes all strains and mutations used in this study. cosa-1::3 × HA, ollas::cosa-1^PD::3 × FLAG and spo-11 deletion were generated by CRISPR-Cas9 genome editing. Briefly, worms were injected with CRISPR/Cas9 injection mixture (10 μl volume), including 0.2 μl Cas9 (IDT, 10 μg/μl stock), 1 μg sgRNA, 10 ng/μl pCFJ90 and 10 ng/μl pCFJ104 injection marker plasmids, and a melted dsDNA donor cocktail (PCR donor with blunt ends or single stranded overhangs, 30 ng/μl) [68,69] or single-stranded templates (30 ng/μl) in the final injection mixture as the repair template. Gene editing-related crRNAs, repair templates, and primers for genotyping are shown in S5 Table. Progenies were genotyped by PCR to detect edits and subsequently validated by sequencing. The strains were outcrossed with N2 at least 3 times prior to analysis.

Analyses of brood size, progeny viability and incidence of males

L4 hermaphrodites were individually picked onto NGM plates (4 worms per plate, 3 plates total). The worms were transferred to fresh plates every 12 h, and the number of eggs were recorded for each plate until cessation of reproduction. The total egg output across all plates was averaged to determine the mean brood size per worm. The percentage of progeny viability was calculated as the total number of viable progenies divided by the total number of eggs. Unhatched embryos were counted 24 h after each worm transfer. The incidence of males was determined by dividing the number of males by the total number of hatched offspring.

DIA proteomics

Whole worm samples were ground into powder using liquid nitrogen. Approximately 20 mg of powder were resuspended in 200 μl lysis buffer (4% SDS, 150 mM Tris-HCl pH 8.0, 5 mM EDTA with protease inhibitors). The samples were boiled for 3 min and further ultrasonicated. Undissolved cellular debris were removed by centrifugation at 16,000g for 20 min. The supernatant was collected and quantified with a BCA Protein Assay Kit. DTT was added to a final concentration of 100 mM and boiled for 5 min. 200 µl UA buffer (8 M Urea, 150 mM Tris-HCl, pH 8.0) was then added to the samples and the samples were then transferred into 1.5 ml Nanosep tube (Pall Corporation, MWCO 10K) and centrifuged for 20 min at 14,000 g at 20°C to remove DTT. 100 µl 0.05 M iodoacetamide (IAA) in UA buffer was added to block reduced cysteine residues and the samples were incubated for 30 min in darkness and then centrifuged for 10 min at 12,000g. After 2× wash with 100 µl UA buffer and 100 µl 50 mM NH₄HCO₃, respectively, 40 µl Trypsin buffer (6 µg Trypsin in 40 µl 50 mM NH₄HCO₃) was added into the sample and incubated at 37°C for 20 h. The peptides were harvested by centrifugation, acidified by 0.1% TFA, desalted with the C18 stage tip, and subsequently dried by a refrigerated CentriVap concentrator. Finally, the tryptic digested peptides were re-dissolved with 0.1% FA. The concentrations of re-dissolved peptides were determined by Nanodrop One device. LC–MS/MS were performed on a Orbitrap Astral mass spectrometer coupled with Vanquish Neo UHPLC system (Thermo Fisher Scientific). Mass spectra were acquired in the DIA mode and the DIA MS data were analyzed using DIA-NN 1.8.1. The MS data were then searched against the Caenorhabditis elegans reference proteome database uniprotkb-Caenorhabditis elegans[6239]-27445-20241018.fasta. Volcano plot and scatter plots in Figs 1C and S1A were executed with R package ggplot2. Proteins that met the criteria of fold change greater than 1.5-fold and P value less than 0.05 were considered as significant differentially expressed proteins. GO enrichment analyses were carried out with the Fisher’s exact test, and FDR correction for multiple testing was also performed. Enriched GO was nominally statistically significant at the Fisher’s exact test P < 0.01. Sample preparation and MS analysis were conducted by Shanghai Bioprofile Technology Company (Shanghai, China).

Ubiquitin-modified proteome analysis

Isolated worm nuclei were used for the analysis. Following RNAi treatment, the adult worms were mixed with small shards of broken glass coverslips and shredded by vortexing in 15 ml conical tubes containing ice-cold nuclear isolation buffer (10 mM HEPES at pH 7.6, 1 mM EGTA, 1.5 mM MgCl₂, 10 mM KCl, 250 mM sucrose, 0.5 mM DTT, and protease inhibitor cocktail). Worm debris was removed by centrifugation at 30g for 5 min, and the supernatant containing nuclei was filtered first with a 100 µm mesh, followed by a second filtration with a 40 µm mesh. The nuclei were then collected by centrifugation at 2,500g for 10 min. Protein preparation steps from grounding to the addition of IAA to block reduced cysteine residues are the same as in the DIA proteomics protocol. After 30 min in darkness, cold acetone was added, and the samples were left in −20°C to precipitate protein overnight. The samples were then centrifuged for 10 min at 12,000g, and the pellets were washed with cold acetone, centrifuged again, and air-dried. 8 M UA buffer was added to re-dissolve the precipitated protein. Next, 50 mM NH₄HCO₃ was added to dilute UA to less than 1 M. Trypsin was then added into the samples, at the enzyme-to-protein ratio of 1:25 (W:W). After incubation at 37°C for 20 h, the samples were acidified by 0.1% TFA, desalted with Sep-PAK, and dried by a refrigerated CentriVap concentrator (Labconco). The trypsin-digested samples were reconstituted and the ubiquitin-modified peptides were enriched with PTMScan Ubiquitin Remnant Motif (K-ε-GG) Kit (Cell Signaling Technology, cat. no. 5562). After wash with IPA buffer, the modified peptides were eluted with 0.15% (v/v) TFA. The elution of the modified peptides was then desalted with a C18 Stage Tips column and dried by a refrigerated CentriVap concentrator (Labconco). Finally, the modified peptides were re-dissolved with 0.1% FA. LC-MS/MS were performed on a Orbitrap Astral mass spectrometer coupled with Vanquish Neo UHPLC system (Thermo Fisher Scientific). Mass spectra were acquired in the DIA mode, and the DIA MS data were analyzed using Spectronaut 19 (Biognosys AG). The MS data were then searched against the Caenorhabditis elegans reference proteome database uniprotkb-Caenorhabditis elegans[6239]-27445-20241018.fasta. Sample preparation and MS analysis were conducted by Shanghai Bioprofile Technology Company (Shanghai, China).

Immunoblot

Synchronized adult worms were washed into M9 buffer and lysed by adding 5×SDS sample buffer and heating to 95°C for 10 min with intermittent vortexing. Whole worm lysates were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore IPVH00010). Membranes were blocked in 5% milk in PBST for 1 h at room temperature and incubated with primary antibodies overnight at 4°C. Primary antibodies used are as follows: mouse anti-HA (1:1,000, BioLegend, 16B12), rabbit anti-GFP (1:2,000, ABclonal Technology, AE011), and mouse anti-ACTIN (1:3,000, Proteintech, #66009-1-Ig).

Immunofluorescence

Gonads from young adult hermaphrodites (48 h post L4 at 20°C) were dissected and stained as previously described [14]. Briefly, young adult hermaphrodites were dissected in 7 μl dissection buffer (25 mM HEPES pH 7.4, 118 mM NaCl, 48 mM KCl, 2 mM EDTA, 5 mM EGTA, 0.1% Tween-20) and briefly fixed in 1% formaldehyde. The dissected gonads were placed on a slide, covered with a coverslip, and flash-frozen in liquid nitrogen. The coverslip was quickly removed and the slide containing the gonads was then placed in −20°C methanol for 10 min. The slide was next washed three times in 1× PBST (0.1% Tween-20) and blocked with 1% BSA for 30 min at room temperature. The sample was then incubated with primary antibodies overnight at 4°C at the indicated dilutions as listed below.

Chromosome spreading of C. elegans germ cell nuclei was performed as previously described [14]. The gonads of adult worms, treated with RNAi for 48 h, were dissected in 5 μl low solution (0.2× PBS with 0.1% v/v Tween-20) on a 18 × 18 mm coverslip (ZEISS, thickness 0.170 ± 0.005 mm). Fifty microliters of spreading solution was then added and the dissected gonads were immediately distributed over the whole coverslip. Coverslips were put in 6 cm culture plates, left to dry overnight at room temperature, washed for 20 min in methanol at −20°C, and rehydrated by washing 3 times for 10 min in 1× PBST. After rehydrating, the samples were processed for immunofluorescence using antibodies at the concentrations listed below. Spreading solution (for one coverslip, 50 μl): 32 µl of fixative (4% w/v paraformaldehyde and 3.2 ± 3.6% w/v sucrose in water), 16 µl of Lipsol solution (1% v/v Lipsol in water) and 2 μl of sarcosyl solution (1% w/v of sarcosyl in water, Sigma-Aldrich L7414).

Primary antibodies used are as follows: mouse anti-FLAG (1:800, Sigma, F1804), GFP booster (1:400, Chromotek, gb2AF488), mouse anti-HA (1:500, BioLegend, 16B12), rabbit anti-HIM-3 (1:300, generated by ABclonal Technology) [14], guinea pig anti-HTP-1 (1:300, generated by ABclonal Technology) [14], and rabbit anti-SYP-1 (1:300, generated by ABclonal Technology) [14]. Secondary antibodies (Invitrogen Alexa 488 and Alexa 594) and DAPI (Sangon Biotech E607303) were used at 1:1,000 and 1:20 dilution, respectively. The samples were mounted in antifade mounting medium (Vectashield H-1000) and imaged using Olympus SpinSR10 with a 60× oil immersion and 1.42 NA objective.

Meiotic crossover recombination frequency assay

Meiotic crossover recombination frequencies were determined as previously described [14], using six insertion–deletion polymorphisms on Chromosome Ⅱ that differ between N2 Bristol and CB4856 Hawaiian. ollas::cosa-1^PD::3 × FLAG was crossed into a genetic background with Chr Ⅱ homozygous for Hawaiian DNA. Hawaiian males or homozygous cosa-1^PD mutant males with Chr Ⅱ homozygous Hawaiian were then crossed with hermaphrodites of the identical genotype in the N2 Bristol background to obtain mutant F1 strains heterozygous for Hawaiian. The F1 hermaphrodites were treated with either ctrl RNAi or 20% uba-1 RNAi for 24 h and then crossed with males of CB5584, a myo-2::GFP expressing strain, and allowed to lay eggs. After confirming the genotype of the F1s, the F2 offspring with high level of GFP signals in the pharyngeal muscles were analyzed for crossover recombination by PCR. Primers used are as follows: Chromosome II:

1. genetic position (gp) -15.4 B5: primer 1:5′-AACGACGCGATGCTATGGAT-3′, primer 2:5′-TGGAATTGAAACAGAACTCAGC-3′, N2: 1,100 bp, CB4856:824 bp;
2. gp -6.3 B7: primer 1:5′-ATTTGGGTGGGAACTGGAGG-3′, primer 2:5′-GCGTGCAGACATAAGATAGGG-3′, N2:634 bp, CB4856:425 bp;
3. gp 0.1 B10: primer 1:5′-ACCAGCAATAGGTCAAGGTCT-3′, primer 2:5′-CACGTCATTCGCCAGTCAAA-3′, N2: 819 bp, CB4856:484 bp;
4. gp 5.7 C3: primer 1:5′-ACATGGGAGCGACGGTTTTA-3′, primer 2:5′-CCCGACACCATAACACAACA-3′, N2:933 bp, CB4856:448 bp;
5. gp 17.5 C5: primer 1:5′-AGCCGTTACTCGCCATGAAA-3′, primer 2:5′-GCCAAACATCGGTCATCGGA-3′, N2:959 bp, CB4856:744 bp;
6. gp 23.1 C6: primer 1:5′-TTGTGTGCAAACACCGTCAC-3′, primer 2:5′-TCGGTCCGAAGGCAATCAAA-3′, N2:1220 bp, CB4856:811 bp.

RNAi, partial RNAi, and double RNAi

Feeding RNAi was used to achieve gene knockdown in C. elegans. The corresponding cDNAs were cloned into L4440 to construct the RNAi plasmid. The empty L4440 was used as control. The RNAi plasmid was transferred to E. coli HT115, which had tetracycline resistance. Single colonies were grown overnight in LB liquid medium containing ampicillin (Amp) and tetracycline (Tet) (both at 25 μg/ml final concentration) at 37°C and 225 rpm. The overnight bacterial cultures were transferred to fresh LB (+Amp, +Tet) at 1:50 (v/v) and incubated at 37°C until a value of optical density at 600 nm (OD₆₀₀) between 0.8 and 1 was reached. 3 mM IPTG was then added to induce dsRNA expression for 6–8 h at 20°C and 160 rpm. The induced bacterial cultures were then spread onto NGM plates supplemented with 2 mM IPTG, 25 μg/ml Amp, and 25 μg/ml Tet. L4-stage worms were transferred to the prepared RNAi plates and left at 20°C for 48 h, unless otherwise stated, before analysis.

For uba-1 partial RNAi, after obtaining bacterial cultures (empty L4440 control and uba-1 RNAi, respectively) with identical OD₆₀₀ values, IPTG was added to induce dsRNA expression. The IPTG-induced bacterial cultures were then mixed at predetermined ratios before being applied to IPTG plates to achieve titrated gene silencing. For 50% syp-1 and 50% uba-1 double RNAi, we mixed the IPTG-induced bacterial cultures at a 1:1 ratio before seeding. For rpn-6 and pbs-6 double RNAi, cDNAs of both rpn-6 and pbs-6 were cloned into the same L4440 plasmid. The remaining steps were identical to the standard RNAi procedures.

RNA-seq

Total RNA was extracted using Trizol (Thermo Fisher Scientific) and further treated with DNase to remove genomic DNA contamination. Isolation of mRNA was performed using the KAPA Stranded mRNA-Seq Kit (Roche Diagnostics GmbH), and the mRNA was then used for RNA-Seq library preparation with the KAPA Stranded mRNA-Seq Kit for Illumina (Roche Diagnostics GmbH). The library was then subjected to Illumina sequencing with paired-end 2 × 150 as the sequencing mode. After quality control, the clean reads were mapped to the C. elegans genome (assembly WBcel235) using the HISAT2 v2.1.0 (non-default parameters: --rna-strandness RF --dta) software. Gene expression levels were estimated using FPKM by StringTie v1.3.4d (non-default parameters: -e --rf). edgeR v3.24.2 was used to measure differential gene expression. The false discovery rate (FDR) control method was used to calculate the adjusted P values in multiple testing to evaluate the significance of the differences. Only gene with an adjusted P value <0.05 and |log2FC| ≥ 1 were used for analysis. Library preparation, sequencing, and data analysis were performed by Genefund Biotech (Shanghai, China).

Yeast two-hybrid analysis

The yeast two-hybrid assay was performed as previously described [70]. Full-length coding sequences for pro-crossover factors (COSA-1, MSH-5, and CDK-2) and COSA-1 mutants (COSA-1^PD) were cloned into plasmid pGADT7 or pGBKT7. Each pair of bait and prey plasmids was co-transformed into yeast strain AH109. The positive colonies were selected on a medium lacking tryptophan (-Trp) and leucine (-Leu) and were resuspended in 1 ml 1× PBS. Serial dilutions of the resuspension were plated on -His -Trp -Leu and -Trp -Leu solid media, and the plates were incubated at 25°C for 3 or 4 days to view the results.

Statistics

Prism 8 software (GraphPad) was used to generate graphs and perform statistical analyses. When it comes to the total immunofluorescence intensity of the germline, signals from the distal tip to the end of pachytene were quantified. P values were primarily calculated using an unpaired t test or ordinary one-way ANOVA. The significance of multiple crossover occurrence was assessed using a one-tailed Fisher’s exact test. P < 0.05 was considered significant. Sample sizes were not predetermined and experiments were not randomized. All the experiments for quantification were conducted at least three times.