Ufd2p promotes efficient crossover formation by destabilizing Top2p during meiosis

Significance

Meiotic crossovers are essential for chromosome segregation and genetic diversity through the genetic exchange between homologous chromosomes. However, the molecular mechanisms governing crossover patterns remain incompletely understood. Here, we identify the E3 ubiquitin ligase Ufd2p as a key factor promoting efficient crossover formation. We show that Ufd2p regulates crossover patterns by mediating the ubiquitin-proteasomal turnover of Topoisomerase II, thereby modulating DNA negative supercoiling and influencing the strength of crossover interference. Strikingly, we show that the Ufd2p mammalian homologue UBE4B promotes crossover formation via a conserved mechanism involving Topoisomerase 2A and DNA supercoil dynamics. Our work identifies a previously uncharacterized factor required for crossover regulation and uncovers a ubiquitin-based topological control mechanism shaping meiotic crossover landscapes across eukaryotic species.

Abstract

Proper crossover (CO) formation in meiosis serves dual roles in ensuring accurate chromosome segregation and generating genetic diversity. However, the molecular mechanisms underlying CO number and distribution remain incompletely understood. Previous studies have implicated the ubiquitin–proteasome system in CO regulation, but specific regulators and mechanisms are poorly defined. Here, we identify the E3 ubiquitin ligase Ufd2p as a key regulator promoting efficient CO formation through a focused genetic screen in Saccharomyces cerevisiae. Deletion of UFD2 significantly reduces CO frequency by enhancing the strength of CO interference. Integrated multiomics analysis indicates that Ufd2p targets Topoisomerase II (Top2p) for ubiquitination and subsequent proteasomal degradation during meiosis. Deletion of UFD2 results in Top2p accumulation, which resolves DNA negative supercoils excessively and enhances CO interference in the nucleus, ultimately reducing CO numbers. We further show that the mammalian homolog of Ufd2p, UBE4B, plays a conserved role in promoting efficient CO formation by regulating TOP2A-dependent DNA negative supercoils dynamics. Notably, expression of mouse or human UBE4B in yeast restores CO formation and meiotic progression in UFD2 deletion cells, demonstrating functional conservation across species. Together, our work identifies Ufd2p as a previously uncharacterized regulator of CO formation and provides important insights into the conserved molecular mechanism, which operates through Top2p-mediated supercoils homeostasis.

Meiosis constitutes a fundamental cell division program in eukaryotes; it executes an evolutionarily conserved function in sexual reproduction by producing haploid gametes (1–3). Crossing-over (CO) is a meiosis-specific feature, which results in DNA segments that are reciprocally exchanged between paired homologous chromosomes (homologs) (4). Crossovers establish physical connections between homologs that are essential for their faithful segregation and the maintenance of genomic integrity (4–6). Simultaneously, COs reshuffle genetic material between homologs, thereby enhancing genetic diversity and promoting evolutionary adaptability (3, 4, 7).

Both the number and spatial distribution of COs are tightly regulated to maintain a balance between chromosomal integrity and genetic variability. Insufficient CO formation and aberrant CO distribution compromise adequate spindle-mediated tension between bivalents, causing chromosome missegregation and genomic instability (6, 8, 9). These defects contribute to meiotic failure and aneuploidy-associated conditions such as infertility, miscarriage, and congenital disorders, including trisomy 21 and Klinefelter syndrome (6, 8–14). Together, these earlier observations underscore the importance of precisely controlling CO number and positioning to balance genomic fidelity while enabling genetic diversity.

Meiotic recombination is initiated via programmed double-strand breaks (DSBs) mediated by topoisomerase-like Spo11p (15–19). This process occurs in the context of meiotic chromosomes and largely depends on the DNA loop-axis architecture (4, 5, 17, 20). Substantial experimental evidence indicates that meiotic chromosome length is positively correlated with both DSB and CO numbers, highlighting the role of higher-order chromosome structure in recombination control (6, 21). Three distinct mechanisms involved in this CO patterning are widely accepted. First, the “obligatory CO” ensures that each homolog pair receives at least one CO, thereby guaranteeing proper homolog disjunction (4, 22). Second, CO interference reduces the probability of nearby CO formation once a CO has occurred, resulting in nonrandom CO spacing along chromosomes (4, 20, 23, 24). Third, CO homeostasis stabilizes overall CO numbers despite fluctuations in DSB levels (24–26). Emerging evidence suggests that CO homeostasis may be mechanistically linked to CO interference (22, 24). Despite their biological importance, the molecular nature of the interference signal and the broader regulatory mechanisms controlling CO distribution remain largely undefined.

The ubiquitin–proteasome system has emerged as an important regulator of meiotic recombination, particularly in CO regulation. Ubiquitin localizes to meiotic chromosome axes and contributes to chromosome assembly, synapsis, DSB formation and repair, and CO maturation, all of which are tightly coupled to recombination outcomes (27–30). Dysregulation of ubiquitination has been shown to perturb meiotic progression and recombination outcomes (28). Several E3 ubiquitin ligases have been implicated in CO regulation. For example, HEI10, the mammalian ortholog of yeast Zip3p, stabilizes recombination intermediates to promote CO maturation (31, 32). RNF212B functions as an E3 ubiquitin ligase required for both CO designation and formation (33). The SCF^RMF1/2 complex mediates degradation of the meiosis-specific recombinase DMC1 to facilitate DSB repair and CO generation (34). Our previous work demonstrated that the Skp-Cullin-F-box (SCF) complex and Ufd4p cooperatively regulate chromosome axis length and CO number through a Pds5p-dependent ubiquitin-proteasome pathway (28). Despite these advances, how E3 ligases coordinate CO patterning remains incompletely understood.

In this study, we identify Ufd2p, an E3 ubiquitin ligase, as a key regulator of CO patterning through functional genetic screening in Saccharomyces cerevisiae. Deletion of UFD2 (ufd2Δ) leads to elevated CO interference accompanied by a reduction in CO number. Integrated multiomics analyses identify Top2p as the substrate of Ufd2p. Biochemical studies demonstrate that Ufd2p physically interacts with Top2p and promotes its ubiquitination-dependent proteasomal degradation. Loss of UFD2 results in Top2p accumulation, which reduces DNA negative supercoils on meiotic chromatin and enhances the strength of CO interference, thereby limiting CO formation. In mice, UBE4B, the mammalian homologue of Ufd2p, promotes efficient CO formation and operates through a conserved mechanism involving TOP2A degradation and regulation of DNA negative supercoil dynamics. Importantly, heterologous expression of mouse or human UBE4B in yeast rescues the meiotic progression defects, reduced CO numbers, and aberrant CO interference caused by UFD2 deletion. Together, our findings uncover a conserved regulatory pathway in which Ufd2p-mediated control of Top2p abundance modulates DNA topology to fine-tune CO interference and ensure efficient CO formation during meiosis.

Results

Ufd2p Is a Key E3 Ubiquitin Ligase Promoting Efficient CO Formation.

To identify E3 ubiquitin ligases involved in CO regulation, we first analyzed publicly available transcriptomic datasets of budding yeast meiosis to identify ligases with elevated expression during the meiotic recombination stage (Dataset S1) (35). This analysis yielded 39 candidate genes (Fig. 1A). Gene Ontology (GO) enrichment analysis revealed that 10 of these candidates were associated with homologous recombination, cell cycle regulation, DNA damage repair, and chromosome organization, processes closely linked to CO regulation (Fig. 1B).

Fig. 1.

Screening of E3 ubiquitin ligases required for CO formation in budding yeast. (A) Heatmap showing the expression dynamics of E3 ubiquitin ligase genes during yeast meiosis, based on RNA-seq data from Brar et al. (35). The data used to generate the heatmap are provided in Dataset S1. The upper schematic illustrates meiotic progression, and genes highly expressed during the meiotic recombination stage are highlighted in red. (B) Gene ontology (GO) analysis of the highly expressed E3 ubiquitin ligase genes identified in panel A during the meiotic recombination stage. (C) Representative images showing CO numbers in whole nuclei and on chromosome XV (Chr XV), analyzed by surface spreading and immunofluorescence. The lacO/LacI-GFP system marks the distal end of Chr XV (blue; red arrows), Zip3p (9×Myc-tagged) marks CO designation sites (green), and Zip1p labels the synaptonemal complex (SC; red). (D and E) Quantification of Zip3p focus numbers on whole nuclei (D) and on Chr XV (E) in WT and E3 ubiquitin ligase deletion mutants. From left to right, 95, 96, 46, 46, 37, 37, 57, 39, 51, 42, and 51 nuclei were analyzed. (F and G) Quantification of Msh4p foci in whole nuclei (F) and along Chr XV (G) in WT and ufd2Δ strains; 53 WT and 51 ufd2Δ nuclei were analyzed. (H) Schematic of the strain used for the spore-specific fluorescence-based assay to quantify CO frequency in WT and ufd2Δ strains, with Chr VIII labeled by CFP, GFP, and YFP at three distinct loci. (I) Representative images (Bottom) and corresponding schematic diagrams (Top) illustrating CO patterns across different genetic intervals. (J) Quantification of CO frequency in WT and ufd2Δ strains based on analysis of 1,925 WT and 2,113 ufd2Δ tetrads. (K) Analysis of meiotic division in WT and strains carrying deletions of E3 ubiquitin ligase. Data are from three independent experiments for each strain, with more than 200 cells examined at each time point. Error bars, SE. (L) Quantification of the spore formation in WT and strains carrying deletion of E3 ubiquitin ligase. More than 300 cells were analyzed for each strain at 24 hours in SPM. (M) Quantitative analysis of the spore viability in WT and strains carrying deletion of E3 ubiquitin ligase. More than 100 tetrads were analyzed for each strain. Error bar, SE in (D-G); SD in (H); 95% confidence interval in (J, L and M); ns, P >0.05; *, P <0.05; **, P <0.01; ***, P <0.001; Two-tailed Student’s t-test in (D-G); two proportion Z-test in (J, L and M).

To assess the functional relevance of these candidates, we generated knockout mutants for each of the 10 genes and quantified COs using Zip3p, a well-established marker of class I CO-designated sites that reliably reflects CO patterning (Fig. 1C) (24, 36, 37). Among all mutants examined, only deletion of UFD2 resulted in a significant reduction in Zip3p focus number in both pachytene nuclei and on chromosome XV (Chr XV) (Fig. 1 D and E), indicating a specific requirement for Ufd2p in efficient CO formation. This defect was independently validated using Msh4p, another core component of the class I CO marker, which similarly exhibited a marked decrease in focus numbers in ufd2Δ cells (Fig. 1 F and G) (38, 39). To further confirm Ufd2p’s regulatory function in CO formation, we employed a spore-specific fluorescence assay, an alternative well-established experimental approach for CO analysis (40, 41). By labeling defined chromosomal intervals with distinct fluorescent reporters and scoring recombinant spore products, we observed a significant reduction in CO frequency across both tested intervals in ufd2Δ strain relative to wild type (Fig. 1 H–J). Together, these complementary approaches establish Ufd2p as an essential regulator of efficient CO formation.

Given that COs are required to establish physical connections between homologous chromosomes and ensure accurate meiotic segregation, defects in CO formation frequently result in impaired meiotic progression, reduced sporulation efficiency, and decreased spore viability (4, 21, 42). We therefore examined meiotic outcomes in all 10 E3 ligase deletion mutants. Notably, only ufd2Δ cells exhibited pronounced defects in meiotic nuclear division, reduced sporulation efficiency, and diminished spore viability (SI Appendix, Fig. S1 A–E), phenotypes consistent with defective CO formation. Collectively, these findings identify Ufd2p as a key E3 ubiquitin ligase required for normal CO formation.

Ufd2p Localizes to Meiotic Recombination Sites.

To gain mechanistic insight into Ufd2p’s role during meiosis, we examined its temporal expression and subcellular localization throughout meiotic progression. RT-PCR and immunoblot analyses showed that Ufd2p is robustly expressed during meiotic prophase I (0 to 5 h in sporulation medium), coinciding with the period of DSB formation and CO designation (SI Appendix, Fig. S2 A–D). Immunofluorescence analysis revealed that Ufd2p forms punctate foci along meiotic chromosomes. Both the number of Ufd2p foci and their signal intensity increased progressively from leptotene to pachytene, as determined by Zip1p-based staging of synaptonemal complex assembly (Fig. 2 A and B and SI Appendix, Fig. S2 E–G). This dynamic pattern suggests that Ufd2p is recruited to chromosomes during the window of active recombination.

Fig. 2.

Ufd2p promotes COs formation by regulating the strength of CO interference. (A and B) Representative images (A) and quantitative analysis (B) of Ufd2p association with meiotic chromatin during prophase I. Meiotic stages are indicated as leptotene (Lep), zygotene (Zyg), and pachytene (Pac). A total of 43 leptotene, 46 zygotene, and 41 pachytene nuclei were analyzed. (C) Integrative Genomics Viewer (IGV) snapshot showing enrichment profiles of Ufd2p, Zip2p, Zip3p, Zip4p, and Msh4p at 4 h in SPM across a representative region of chromosome VII. (D) Quantification of Ufd2p enrichment at Zip2p, Zip3p, Zip4p, and Msh4p peak sites. (E) Classical coefficient of coincidence (CoC; Left) and best-fit simulated CoC (Right) analyses on Chr XV in WT and ufd2Δ strains, with L_CoC and L_BF values indicated. (F) Distributions of observed Zip3p foci (Left) and simulated CO numbers (Right) on Chr XV in WT and ufd2Δ strains; mean CO numbers are shown above each graph. (G) Modified CoC (MCoC) analysis of CO interference along Chr XV, with colored blocks indicating the number of intervals affected by interference and the average number of affected intervals and interference distance shown below. (H) Gamma distribution shape parameters describing distances between adjacent Zip3p foci on Chr XV. Independent experiments were performed, and histogram values represent averages. Error bars indicate SE in (B) and SD in (E, F, and H). ***P < 0.001; two-tailed Student’s t test.

Consistent with this interpretation, more than 80% of Ufd2p foci colocalized with the DSB repair proteins RPA, Rad51p, and Dmc1p (SI Appendix, Fig. S3 A–C), indicating that Ufd2p localizes to sites of meiotic DSB repair. To further define its genome-wide chromatin association, we performed chromatin immunoprecipitation sequencing (ChIP-seq) of Ufd2p during meiosis. This analysis revealed extensive overlap between Ufd2p-binding sites and peaks corresponding to ZMM proteins, including Zip2p, Zip3p, Zip4p, and Msh4p. Specifically, more than 70% of Ufd2p peaks overlapped with ZMM protein peaks, while approximately 80% of ZMM peaks coincided with Ufd2p enrichment (Fig. 2C and SI Appendix, Fig. S3 D–I). Quantitative enrichment analysis further demonstrated that Ufd2p is significantly enriched at ZMM proteins-marked CO sites relative to surrounding chromatin (Fig. 2D and SI Appendix, Fig. S3J). Together, these cytological and genomic data indicate that Ufd2p localizes to meiotic recombination sites and is positioned to directly regulate CO formation.

Ufd2p Attenuates CO Interference.

Multiple factors are known to influence CO number, including DSB formation and repair, chromosome axis length, and the strength of CO interference (21, 24, 43). To determine which of these factors underlies the reduced CO frequency observed in the ufd2Δ strain, we performed a systematic analysis of each potential contribution. We first examined whether UFD2 deletion affects meiotic DSB formation or repair. To this end, UFD2 deletion was introduced into a dmc1 mutant background, which accumulates unrepaired DSBs (44). Quantification of Rad51p foci revealed no significant difference in DSB levels between wild type (WT) and ufd2Δ strains (SI Appendix, Fig. S4 A and B). Kinetic analysis of Rad51p focus formation and resolution across meiotic progression showed only a minor delay in DSB repair in ufd2Δ, indicating that Ufd2p does not substantially influence DSB formation or repair dynamics (SI Appendix, Fig. S4 C and D).

A positive correlation between chromosome axis length and CO frequency has been identified in many organisms (3, 6, 21), and the cohesion-associated factor Pds5p has been reported to regulate CO numbers in a dose-dependent manner (21). We next assessed whether altered axis organization or and Pds5p abundance contributes to CO reduction in ufd2Δ. Immunostaining of the axial element protein Rec8p and the synaptonemal complex (SC) central element Zip1p revealed no significant changes in chromosome length upon UFD2 deletion (SI Appendix, Fig. S4 E–J). Consistently, immunoblot analyses showed that protein levels of Pds5p and Rec8p remained unchanged in ufd2Δ during meiotic prophase I (SI Appendix, Fig. S4 K–N). These results indicate that Ufd2p regulates COs formation independently of chromosome axis length and Pds5p abundance.

We next focused on CO interference, a key determinant of CO spacing and frequency. CO positions were mapped cytologically using Zip3p foci along Chr XV, and interference strength was quantified using four established and independent approaches: coefficient of coincidence (CoC), beam-film simulation, modified CoC (MCoC), and gamma distribution modeling (24, 26, 43). In ufd2Δ strain, the CoC curve exhibited a clear rightward shift, with the L_CoC increasing from 0.30 to 0.35 μm, indicating that the distance at which the following Zip3p-marked site is encountered in 50% of intervals has risen from 0.30 to 0.35μm. Concurrently, Zip3p focus numbers were significantly decreased in ufd2Δ strain (Fig. 2 E and F, Left panels and SI Appendix, Table S1).

To further characterize changes in CoC profiles and CO distribution, we applied best-fit simulations using the beam-film model (43, 45, 46). Experimental data from ufd2Δ strain were best matched by simulations in which the CO interference distance parameter (L_BF) was increased from 0.10 to 0.12, corresponding to an increase in effective interference distance from 0.30 to 0.35 μm (Fig. 2 E and F, Right panels and SI Appendix, Table S1). MCoC analysis independently confirmed enhanced CO interference in ufd2Δ, as reflected by increased inhibited interval numbers and a longer interference distance (L_MCoC) (Fig. 2G). Similarly, gamma distribution describes the frequency distribution of distances between adjacent COs, where larger shape parameters indicate stronger interference. Consistently, elevated shape parameters were observed in this modeling in ufd2Δ strain, further supporting enhanced CO interference strength resulting from Ufd2p depletion (Fig. 2H).

To generalize these findings, we extended the analysis to Chromosome III (Chr III), one of the smallest yeast chromosomes. As observed for Chr XV, all four analytical methods consistently demonstrated enhanced CO interference in ufd2Δ strain (SI Appendix, Fig. S5 A–E and Table S1). In addition, spore-specific fluorescence assays independently confirmed enhanced CO interference in ufd2Δ strain (SI Appendix, Fig. S5F). Together, these results establish that Ufd2p depletion globally enhances CO interference along meiotic chromosomes.

To distinguish whether altered CO patterning arises from changes in interference strength versus other recombination parameters, we performed best-fit simulations incorporating four key variables: interference distance, CO designation driving force, DSB number, and CO maturation efficiency. Among these parameters, only increased interference distance yielded strong concordance between simulated and experimental data in ufd2Δ strain (SI Appendix, Fig. S5F) (26, 43), supporting the conclusion that Ufd2p primarily attenuates CO interference. It was reported that CO interference is dispensable for the obligatory CO but required for maintaining the strength of CO homeostasis (22, 24, 26). To test whether Ufd2p affects these regulatory features, we tested the presence of an obligatory CO by quantifying the frequency of Chr III lacking any Zip3p foci and found no significant difference between WT and ufd2Δ strains, indicating that the obligatory CO remains preserved (SI Appendix, Fig. S5 G and H). We further assessed CO homeostasis in ufd2Δ strain by measuring CO output across strains with experimentally altered DSB levels, as described previously (24–26). Together with stronger CO interference, ufd2Δ strain exhibited intact CO homeostasis, as indicated by its ability to maintain CO numbers despite changes in DSB numbers (SI Appendix, Fig. S5I).

Collectively, these analyses demonstrate that Ufd2p promotes efficient CO formation primarily by restraining the strength of CO interference, without perturbing DSB formation, chromosome axis organization, obligatory CO assurance, or CO homeostasis.

Top2p Is a Meiosis-Specific Substrate of Ufd2p.

Ufd2p functions as an E3 ubiquitin ligase that extends polyubiquitin chains to target substrates for proteasomal degradation (47–49). We therefore hypothesized that Ufd2p promotes efficient CO formation by ubiquitinating and destabilizing a key meiotic regulator. To identify candidate substrates, we performed parallel quantitative proteomic and ubiquitinome analyses comparing WT and ufd2Δ cells collected at pachytene stage, when CO designation and interference are established (5 h in sporulation medium). Proteomic profiling identified 259 proteins with increased abundance in ufd2Δ cells, while ubiquitinome analyses revealed 3,217 ubiquitination sites with reduced modification across 1,539 proteins (Fig. 3 A and B). Intersection of these datasets yielded 118 proteins that simultaneously exhibited increased protein abundance and decreased ubiquitination upon UFD2 deletion (Fig. 3C). GO enrichment analysis of these candidates revealed significant overrepresentation of proteins involved in meiotic recombination and chromosome organization (Fig. 3D).

Fig. 3.

Ufd2p destabilizes Top2p during yeast meiosis. (A and B) Volcano plots showing differentially expressed proteins (A) and differentially ubiquitinated proteins (B) during meiotic prophase I (4 h in SPM) in ufd2Δ compared with WT strains. Three independent biological replicates were analyzed for each strain. Selection criteria were a fold change ≥ 1.5 and P < 0.05. Source data are provided in Dataset S3 (A) and Dataset S3 (B). (C) Venn diagram illustrating the overlap between upregulated proteins and downregulated ubiquitinated proteins in ufd2Δ cells during meiotic prophase I. (D) GO analysis of the overlapping protein set identified in (C), with dot size and adjacent numbers indicating the number of genes associated with each term. (E) Heatmap showing protein abundance at 4 h in SPM for meiotic recombination-related proteins identified in (D). (F and G) Coimmunoprecipitation analyses demonstrating the interaction between Ufd2p and Top2p using meiotic cells (4 h in SPM) from UFD2-9MYC and TOP2-3HA strains; mouse monoclonal anti-Myc and rabbit monoclonal anti-HA antibodies were used. (H) GST pull-down assay confirming a direct interaction between Ufd2p and Top2p. (I) Cycloheximide chase assay assessing Top2p stability following addition of 100 μM cycloheximide and 50 μM MG132 at pachytene (4 h in SPM); time points indicate hours after treatment. (J) Quantification of Top2p abundance from (I). (K) Representative images showing Top2p localization on meiotic chromatin during prophase I, with stages defined by Zip1p morphology. (L) Quantification of Top2p fluorescence intensity on chromatin; from leptotene to pachytene, 75, 58, and 37 WT nuclei and 34, 34, and 32 ufd2Δ nuclei were analyzed. (M) Immunoblot analysis of Top2p abundance in WT, ufd2Δ, top2-KR, and ufd2Δ/top2-KR strains. (N) Quantification of Top2p levels in (M). Three independent experiments were performed for (J) and (N). For plots in (J, L, and N), center lines indicate means. Error bars represent SD in (J and N) and SE in (L). *P < 0.05; **P < 0.01; ***P < 0.001; two-tailed Student’s t test.

Among these candidates, Topoisomerase II (Top2p) emerged as a particularly compelling target. Top2p showed one of the greatest reciprocal changes in abundance and ubiquitination in ufd2Δ cells and has been previously implicated in regulating CO interference through modulation of DNA supercoiling (Fig. 3 C–E) (24, 26). These observations suggested that Top2p may be a direct substrate of Ufd2p during meiosis. To test this hypothesis, we performed coimmunoprecipitation (Co-IP) assays, which revealed a robust physical interaction between Ufd2p and Top2p in meiotic extracts (Fig. 3 F and G). This interaction was further validated by GST pull-down assays using purified proteins, demonstrating a direct interaction between Ufd2p and Top2p in vitro (Fig. 3H). Together, these data identify Top2p as a candidate Ufd2p substrate during meiosis.

Ufd2p Promotes Proteasomal Degradation of Top2p During Meiosis.

To determine whether Ufd2p-mediated ubiquitination regulates Top2p stability, we performed cycloheximide (CHX) chase assays to monitor Top2p turnover during meiosis (50–52). In WT cells, Top2p protein levels declined progressively following CHX treatment, whereas inhibition of proteasomal activity with MG132 largely blocked this degradation (Fig. 3 I and J) (29). In ufd2Δ cells, and MG132 treatment provided no additional stabilization, indicating that Ufd2p promotes Top2p degradation through the ubiquitin–proteasome pathway (Fig. 3 I and J). To directly test whether ubiquitination of Top2p is necessary for its degradation, we mutated four ubiquitinated lysine residues of Top2p identified in our ubiquitinome analysis (K190, K477, K1007, and K1043) to arginine, generating a ubiquitination-deficient top2-KR allele. CHX chase assays revealed that Top2-KR was substantially stabilized compared to WT Top2p. Importantly, deletion of UFD2 did not further increase Top2-KR levels in the ufd2Δ top2-KR strain, indicating that Ufd2p and Top2p ubiquitination function in the same pathway to regulate Top2p stability (Fig. 3 B, I, and J).

We next examined the effects of Ufd2p deletion on Top2p abundance and chromatin association during meiotic prophase I. Immunoblotting and immunofluorescence analyses revealed significantly elevated Top2p levels in ufd2Δ cells relative to WT (Fig. 3 K–N and SI Appendix, Fig. S6 A–C). The top2-KR mutant phenocopied this increase, and ufd2Δ top2-KR cells showed no further elevation of Top2p abundance (Fig. 3 M and N). These biochemical and genetic analyses collectively establish that Ufd2p catalyzes Top2p ubiquitination, promoting its proteasomal degradation during meiosis.

Ufd2p Attenuates CO Interference by Promoting Top2p Degradation and Maintaining DNA Negative Supercoils.

DNA negative supercoiling has been shown to influence both CO number and the strength of CO interference during meiosis (26). Top2p is the primary enzyme responsible for resolving DNA negative supercoils, and Top2p activity directly affects CO patterning (24, 26). Based on our finding that Ufd2p promotes Top2p degradation, we hypothesized that Ufd2p might attenuate CO interference by maintaining appropriate levels of DNA negative supercoils through Top2p regulation. To test this hypothesis, we examined the genomic distribution of Ufd2p, Top2p, and DNA negative supercoils during meiosis. DNA negative supercoils were mapped using biotinylated trimethylpsoralen (bTMP), as well-established probe that preferentially intercalates into negatively supercoiled DNA and can be detected by immunofluorescence and ChIP-seq (53, 54). bTMP ChIP-seq analysis revealed that Ufd2p localized to genomic regions enriched for both Top2p and DNA negative supercoils, exhibiting highly similar genome-wide distribution patterns (Fig. 4A). Quantitative analysis further showed significant colocalization between Ufd2p occupancy and DNA negative supercoil enrichment among the genome (Fig. 4 B and C).

Fig. 4.

Ufd2p regulates CO interference through Top2p and DNA negative supercoils. (A) Representative IGV tracks showing enrichment of Ufd2p, Top2p, and DNA negative supercoils at 4 h in SPM across a region of chromosome III, with the reference genome displayed below (blue). (B and C) Quantification of Ufd2p enrichment at Top2p peaks (B) and DNA negative supercoil peaks (C). (D) Representative images of DNA negative supercoils in surface-spread meiotic nuclei detected by bTMP incorporation and NeutrAvidin DyLight staining. (E) Quantification of DNA negative supercoil signals from 0 to 5 h; for WT, 49, 31, 31, 36, 35, and 36 nuclei were analyzed, and for ufd2Δ strain, 39, 44, 39, 43, 44, and 44 nuclei were analyzed. (F) Immunoblot analysis of Top2p abundance in WT, ufd2Δ, and TOP2/top2Δ ufd2Δ strains. (G) Quantification of Top2p levels in (F) from three independent experiments. (H, I) Quantification of Zip3p foci in whole nuclei (H) and on Chr XV (I) in WT and indicated mutants; from left to right, 95, 96, 96, 199, and 178 nuclei were examined. (J) Distributions of observed Zip3p foci (Top) and simulated CO numbers (Bottom) on Chr XV, with average CO numbers shown above each plot. (K) CoC analysis (Top) and best-fit simulated CoC analysis (Bottom) on Chr XV, with L_CoC and L_BF values indicated. Three independent experiments were conducted per strain, with more than 200 bivalents analyzed per experiment; ten independent simulations per strain were performed, each using 5,000 bivalents (J and K). For plots, center lines indicate means. Error bars represent SE in (E, H, and I) and SD in (G, J, and K). Statistical significance: ns, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; two-tailed Student’s t test.

We next assessed global DNA negative supercoil levels during meiotic prophase I (0 to 5 h in SPM) using bTMP immunofluorescence (53, 55, 56). Compared with WT cells, ufd2Δ mutants exhibited a marked reduction in DNA negative supercoil levels, indicating that loss of Ufd2p leads to excessive resolution of negative supercoils on meiotic chromatin (Fig. 4 D and E). These findings support a model in which Ufd2p regulates the chromatin topology by targeting Top2p for degradation, thereby modulating the supercoiling environment to control CO formation and CO interference (26).

To directly test whether elevated Top2p levels underlie the CO defects observed in ufd2Δ cells, we genetically reduced Top2p dosage by deleting one TOP2 allele in the ufd2Δ background. Partial reduction of Top2p restored both CO numbers and CO interference strength to WT levels, effectively rescuing the meiotic defects caused by UFD2 deletion (Fig. 4 F–K and SI Appendix, Fig. S7 A and B and Table S1). Conversely, the ubiquitination-defective top2-KR mutant, which stabilizes Top2p, phenocopied the reduced CO number and enhanced CO interference observed in ufd2Δ cells (Figs. 3 K–N and 4 F–K and SI Appendix, Fig. S7 A and B). Importantly, UFD2 deletion did not further exacerbate these phenotypes in the top2-KR background, demonstrating that Ufd2p-mediated Top2p ubiquitination and degradation act within a single genetic pathway (Fig. 4 H–K and SI Appendix, Fig. S7 A and B and Table S1).

Collectively, these results establish that Ufd2p promotes efficient CO formation by attenuating CO interference through Top2p degradation, thereby maintaining a permissive level of DNA negative supercoils during meiotic prophase.

UBE4B Is Required for Efficient CO Formation in Mice.

Our findings in budding yeast demonstrate that Ufd2p attenuates CO interference through a Top2p-dependent mechanism. Given the strong evolutionary conservation of both Ufd2p and Top2p across eukaryotes (SI Appendix, Fig. S8 A–H), we next asked whether this regulatory pathway is conserved in mammals. Mammals encode two Ufd2p homologs, UBE4A and UBE4B. To determine their respective roles during meiosis, we generated germ cell-specific knockout mice by crossing Ube4a-flox or Ube4b-flox mice (both harboring loxP sites flanking exon 6) with Stra8-GFPCre mice, which induce recombination in type A1 spermatogonia (Fig. 5A). The resulting conditional knockout mice are referred to as Stra8-Ube4a^−/− and Stra8-Ube4b^−/−, respectively.

Fig. 5.

Ufd2p/UBE4B promotes efficient CO formation through a conserved mechanism. (A) Schematic illustrating the generation of Ube4a^F/F, Stra8-Ube4a^−/−, and Ube4b^F/F, and Stra8-Ube4b^−/− mice. Exon 6 of Ube4a and Ube4b were selected for deletion. Purple arrows indicate LoxP sites, and yellow arrows indicate Cas9/sgRNA cleavage sites. (B and C) Immunoblot analyses demonstrating reduced UBE4A (B) and UBE4B (C) protein levels in testes from Stra8-Ube4a^−/− and Stra8-Ube4b^−/−mice, respectively. GAPDH served as a loading control. (D) H&E-stained histological sections of testes (Top) and caudal epididymides (Bottom) from 8-wk-old Ube4b^F/F and Stra8-Ube4b^−/− mice. (E) Representative images of pachytene spermatocytes from Ube4b^F/Fand Stra8-Ube4b^−/−mice stained for SYCP3 (red) and MLH1 (green). (F) Quantification of MLH1 foci per cell in (E), based on analysis of 65 Ube4b^F/F and 38 Stra8-Ube4b^−/− spermatocytes. (G) Representative images of pachytene spermatocytes with chromosome lacking MLH1 foci. The chromosomes lacking MLH1 were outlined in red and shown at higher magnification in the Right panel. (H) Proportion of pachytene spermatocytes with chromosomes lacking MLH1 foci in Ube4b^F/Fand Stra8-Ube4b^−/−mice; 134 and 121 spermatocytes were analyzed, respectively. (I) Heatmap showing Top2a and Top2b expression during spermatogenesis, based on transcriptional data from the GametesOmics database (57). (J) Coimmunoprecipitation analysis of the interaction between mouse UBE4B and TOP2A. HEK293T cells were cotransfected with pCS2plus-6Myc-Top2a and pRK-Flag-Ube4b (or empty vector), harvested 36 h post-transfection, and subjected to immunoprecipitation with anti-c-Myc magnetic beads followed by immunoblotting with anti-FLAG and anti-MYC antibodies. (K) Immunoblot analysis of TOP2A protein levels in testis extracts from Ube4b^F/Fand Stra8-Ube4b^−/−. (L) Quantification of TOP2A abundance in (K) from three independent experiments. (M) Representative images of pachytene spermatocytes stained for SYCP3 (red) and TOP2A (green), with nuclei counterstained by DAPI (blue). (N) Quantification of TOP2A fluorescence intensity in (M), based on 37 Ube4b^F/Fand 36 Stra8-Ube4b^−/− spermatocytes. (O) Representative images of pachytene spermatocytes stained for SYCP3 (red) and bTMP probe (green) to assess DNA negative supercoils, with nuclei stained by DAPI (blue). (P) Quantification of DNA negative supercoil levels in (O); 32 spermatocytes per genotype were analyzed. (Q, R) Quantification of CO numbers in whole nuclei (Q) and on Chr XV (R) in WT, ufd2Δ, and ufd2Δ strains expressing human or mouse UBE4A or UBE4B. From left to right, 95, 96, 45, 40, 49, and 48 nuclei were analyzed in (Q), and 213, 95, 45, 40, 49, and 48 Chr XV were analyzed in (R). (S) Distributions of observed Zip3p foci (Top) and simulated CO numbers (Bottom) on Chr XV in WT, ufd2Δ, and ufd2Δ strains expressing human or mouse UBE4A or UBE4B; mean CO numbers are indicated above each graph. (T) Coefficient of coincidence (CoC; Top) and best-fit simulated CoC (Bottom) analyses on Chr XV for the same strains, with L_CoC and L_BF values indicated. Three independent experiments were performed per strain, with more than 200 bivalents analyzed per experiment; ten independent simulations per strain were conducted using 5,000 bivalents per simulation in (S and T). For all plots, center lines indicate means. Error bars represent SE in (F, N, and P–R), SD in (L, S, and T), and 95% CI in (H). Statistical significance: ns, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; two-tailed Student’s t test was used in (F, L, N, and P–R), and a two-proportion Z-test was used in (H).

Efficient deletion of Ube4a and Ube4b in these models was confirmed via immunoblotting of testis lysates (Fig. 5 B and C). While Stra8-Ube4a^−/− mice exhibited normal testis size and weight, Stra8-Ube4b^−/− mice displayed significantly reduced testis size and weight relative to Ube4b^F/F controls (SI Appendix, Fig. S9 A–D). Histological examination revealed severe disruption of seminiferous tubule architecture in Stra8-Ube4b^−/− testes, characterized by widespread germ cell degeneration and condensed nuclei (Fig. 5D). Periodic acid Schiff (PAS) and hematoxylin staining further showed defective spermatogenesis, including loss of postmeiotic cells, accumulation of pachytene spermatocytes with hypercondensed nuclei at stage IV, and metaphase spermatocytes with chromosome alignment defects at stage XII (Fig. 5D and SI Appendix, Fig. S9E). Consistent with these observations, no spermatozoa were detected in the cauda epididymis of Stra8-Ube4b^−/− mice, and total sperm counts were dramatically reduced compared to controls (Fig. 5D). In contrast, Stra8-Ube4a^−/− mice displayed normal seminiferous tubule organization and sperm output (Fig. 5D and SI Appendix, Fig. S9E). These results establish that UBE4B, but not UBE4A, is essential for male meiosis in mice.

To define the precise stage of disrupted meiosis, we quantified meiotic prophase I substages in Ube4b^F/F and Stra8-Ube4b^−/−testes. Stra8-Ube4b^−/− spermatocytes exhibited a marked accumulation at the pachytene stage, accompanied by a significant reduction in diplotene and diakinesis cells (SI Appendix, Fig. S9 E–J). Consistently, immunostaining for phosphor-histone H3 (H3Ser10p), a marker of metaphase spermatocytes, revealed a substantial decrease in metaphase I cells in Stra8-Ube4b^−/− testes (SI Appendix, Fig. S9K). Since the first spermatogenic wave is synchronized with pachytene cells appearing at postnatal day 16 (PD16) (58), we analyzed prepubertal testes and observed significant germ cell loss in Stra8-Ube4b^−/− mice after PD17 (SI Appendix, Fig. S9 L and M). TUNEL assays corroborated these results, revealing dead pachytene spermatocytes at stage IV (SI Appendix, Fig. S9 F, Upper panel) and metaphase spermatocytes at stage XII (SI Appendix, Fig. S9 F, Lower panel) in Stra8-Ube4b^−/− testes. The proportion of TUNEL-positive tubules was significantly higher in Stra8-Ube4b^−/− mice than in controls (SI Appendix, Fig. S9G). These findings suggest that Stra8-Ube4b^−/− spermatocytes are predominantly arrested and undergo cell death during the pachytene stage.

Given Ufd2p’s role in promoting efficient CO formation in yeast, we next assessed CO formation in Stra8-Ube4b^−/− spermatocytes using MLH1 immunostaining, a marker for COs (59). The number of MLH1 foci was significantly reduced in Stra8-Ube4b^−/− pachytene spermatocytes relative to controls (Fig. 5 E and F), while meiotic DSBs could be repaired in some Stra8-Ube4b^−/− pachytene spermatocytes as suggested by staining of DSB repair markers γH2AX and RPA2 (60, 61) (SI Appendix, Fig. S10 A–D). Importantly, the frequency of chromosomes lacking an MLH1 focus was not significantly increased (Fig. 5 G and H), indicating that the obligatory CO is preserved. These features parallel the yeast ufd2Δ phenotype and are consistent with enhanced CO interference.

Consistent with impaired CO maturation, Stra8-Ube4b^−/− spermatocytes also exhibited a significant reduction in HEI10 foci, a marker of designated and maturing CO sites (SI Appendix, Fig. S11 A and B). Furthermore, Stra8-Ube4b^−/− diplotene spermatocytes displayed complete homolog desynapsis (SI Appendix, Fig. S11C), indicating defective chiasma formation and stabilization. Collectively, these data demonstrate that UBE4B is required for efficient CO formation and maturation in mouse meiosis and functionally mirrors the conserved role of Ufd2p in regulating CO patterning.

Ufd2p/UBE4B Promotes Efficient CO Formation Via a Conserved Mechanism.

We next investigated whether UBE4B promotes CO formation in mammals through the same mechanistic pathway as Ufd2p in yeast, namely by regulating TOP2-mediated DNA negative supercoils. In mammals, two type II topoisomerases exist, TOP2A and TOP2B; however, transcriptomic analysis revealed that TOP2A is the predominant isoform expressed during mouse meiosis (Fig. 5I), indicating that it is the most likely functional counterpart of yeast Top2p. Coimmunoprecipitation assays revealed a physical interaction between UBE4B and TOP2A (Fig. 5J). Immunoblot analysis further revealed a significant increase in TOP2A protein levels in Stra8-Ube4b^−/−testes relative to control testes (Fig. 5 K and L). Immunofluorescence analysis localized this accumulation specifically to pachytene spermatocytes in Stra8-Ube4b^−/− mice (Fig. 5 M and N), indicating that UBE4B promotes TOP2A turnover during meiotic prophase I.

To determine the consequences of TOP2A accumulation on chromatin topology, we examined DNA negative supercoils using bTMP immunofluorescence in spermatocyte chromosome spreads. Stra8-Ube4b^−/− pachytene spermatocytes exhibited a pronounced reduction in negative supercoil signal compared to controls (Fig. 5 O and P), consistent with excessive TOP2A-mediated relaxation of DNA. These findings indicate that UBE4B regulates meiotic chromatin topology by limiting TOP2A abundance, thereby maintaining appropriate levels of DNA negative supercoils required for normal CO patterning.

To directly test the functional conservation of this pathway, we expressed mouse and human UBE4A or UBE4B in ufd2Δ yeast strain and assessed their ability to rescue meiotic defects. Expression of either mouse or human UBE4B fully restored Top2p protein abundance to WT levels, rescued meiotic progression, normalized CO numbers, and reinstated wild type CO interference patterns (Fig. 5 Q–T and SI Appendix, Fig. S12 A–J). In contrast, UBE4A expression resulted in only partial restoration of Top2p levels and yielded incomplete rescue of meiotic progression, CO number, and CO interference phenotypes (Fig. 5 Q–T and SI Appendix, Fig. S12 A–J). These findings establish that UBE4B, but not UBE4A, fulfills a substantial role in maintaining meiotic functions similar to those of Ufd2p.

The differential rescue efficiency between UBE4A and UBE4B correlates with their distinct sequence and structural similarity to yeast Ufd2p, as well as their relative capacity to promote Top2p degradation (SI Appendix, Figs. S8 E–H and S12 A–E). The complete rescue of yeast meiotic defects by UBE4B points to a functional specialization within the UBE4 protein family and highlights the evolutionary conservation of the Ufd2p–Top2p axis as a key regulator of CO interference.

Discussion

The number and spatial distribution of COs are critical for ensuring accurate homolog segregation and generating genetic diversity during meiosis. Although the ubiquitin–proteasome system has emerged as an important regulator of meiotic recombination, the specific ubiquitin ligases and molecular mechanisms that control CO patterning have remained incompletely understood. In this study, we identify the E3 ubiquitin ligase Ufd2p as a key regulator of efficient CO formation in budding yeast and demonstrate that it acts by targeting Top2p for ubiquitination and proteasomal degradation during meiotic prophase I. By controlling Top2p abundance, Ufd2p modulates DNA negative supercoil dynamics, thereby fine-tuning CO interference strength (Figs. 2 and 4 and SI Appendix, Figs. S12 and S13 and Table S1). In mammals, UBE4B, but not its paralog UBE4A, performs a conserved function in regulating CO formation by controlling TOP2A-dependent DNA negative supercoil dynamics (Fig. 5 and SI Appendix, Fig. S12). These findings define an evolutionarily conserved Ufd2p–Top2p regulatory axis as a fundamental mechanism governing CO interference.

A stress relief model has been proposed to elucidate and simulate CO interference (24, 43). According to this model, mechanical stress generated by chromosome-axis-constrained chromatin expansion accumulates along the axes, provoking local CO formation that intrinsically relieves stress and thereby establishes interference. This local change propagates outward along axes (“interference spreading”), inducing chromatin/axis compaction to suppress further COs (24, 43). The dynamics of DNA negative supercoils (through accumulation, relief, and redistribution across chromosomal domains) may serve as a physical substrate for this process and provide the physical basis for CO interference (24, 26, 43). Our results provide mechanistic support for this model by identifying Ufd2p as a regulator of DNA supercoil homeostasis via Top2p degradation. Loss of Ufd2p enhances Top2p-mediated unwinding of negative supercoils, thereby amplifying interference spreading and suppressing additional CO formation. By linking ubiquitin-dependent protein turnover to chromosome-wide topological regulation, our work provides molecular insight into how CO interference strength can be dynamically tuned during meiosis. DNA negative supercoils are intrinsic topological features of genomic DNA, arising from fundamental processes such as DNA replication and transcription (26, 53, 55), and their abundance and localization are dynamically modulated throughout the cell cycle. Our work establishes Ufd2p as a pivotal modulator of chromatin architecture, governing DNA negative supercoil dynamics via ubiquitin-dependent precision control over Top2p abundance and spatial distribution. This finding also provides possible insights into how cells establish and maintain appropriate topological states during critical nuclear processes.

Beyond its role in CO regulation, Ufd2p has been previously characterized as a ubiquitin chain elongation factor (E4) and ubiquitin signal converter that facilitates proteasomal degradation by remodeling ubiquitin linkages (48, 49). In somatic cells, Ufd2p and its homologs function in DNA damage responses and DSB repair (62, 63). Our findings reveal a distinct meiotic function for Ufd2p that operates downstream of DSB formation and repair initiation. Although Ufd2p depletion has little effect on meiotic DSB formation or repair efficiency (SI Appendix, Fig. S4 A–D), it profoundly affects recombination pathway and outcomes by modulating CO interference. This functional divergence likely reflects fundamental differences between somatic and meiotic DNA repair programs, as meiosis relies on homologous recombination while actively suppressing nonhomologous end joining (4, 64–66). Notably, a substantial fraction of Ufd2p does not localize to recombination sites (SI Appendix, Fig. S3), and multiple additional candidate substrates were identified in our proteomic analyses (Fig. 3 A–E), suggesting that Ufd2p may exert broader regulatory roles during meiosis beyond CO control.

Compared to yeast ufd2Δ cells, a more severe meiotic phenotype was observed in Stra8-Ube4b^−/− mice (Fig. 5 and SI Appendix, Figs. S1 A–C and S9). Consistent with previous studies, male mouse meiosis is generally more sensitive to perturbations in recombination factors (67, 68). This phenotypic divergence may be associated with the species-specific differences in baseline CO numbers, indicating the broader functional requirements of UBE4B during mammalian meiosis. Whole-genome analysis indicates that yeast undergoes approximately 90 COs along 16 homologs per meiosis, whereas mice have only about 22 COs along 20 homologs (31, 42). The higher CO number in yeast may provide greater redundancy, helping to ensure at least one obligatory CO per chromosome pair even when CO formation is impaired. Additionally, evidence of DSB repair defects in a subset of Stra8-Ube4b^−/− spermatocytes suggests that UBE4B may contribute to multiple meiotic processes beyond CO regulation (SI Appendix, Fig. S11).

Collectively, our findings define a conserved ubiquitin-based topological control mechanism in which the Ufd2p/UBE4B-Top2p/TOP2A axis regulates DNA negative supercoil dynamics to shape meiotic CO landscapes. By linking protein ubiquitination, DNA topology, and CO interference, this work advances our understanding of how physical chromosome properties are integrated with recombination control to ensure faithful meiosis across eukaryotic species.

Materials and Methods

Yeast Strains.

All S. cerevisiae strains used in this study were derived from the SK1 background, as described and listed in SI Appendix, Table S2. The plasmids used to construct the yeast strains are listed in SI Appendix, Table S3.

Meiotic Time Course, Sporulation, and Spore Viability analysis.

Yeast meiotic cultures were synchronized in SPS pregrowth medium (SPSII) as previously described (21, 42). Cells were then harvested and transferred to sporulation medium (SPM) to induce meiosis. At the indicated time points, aliquots were collected from meiotic cultures, fixed, and stained with DAPI. Nuclear divisions were monitored using an epifluorescence microscope (Nikon, TiU2). Sporulation efficiency was assessed microscopically at 24 h, and spore viability was determined by tetrad dissection on YPD plates.

Crossover Interference analysis.

Cross-over interference was evaluated using the classical coefficient of coincidence (CoC), beam-film simulation, modified CoC (MCoC), and gamma distribution analyses, as previously described (24, 26, 42, 43). Detailed protocols are provided in SI Appendix.

Mouse.

The Stra8-GFPCre mouse line was kindly gifted by Dr. Ming-Han Tong (Chinese Academy of Sciences Center for Excellence in Molecular Cell Science) (69). Detailed procedures for generating Strs8-Ube4a^−/− and Strs8-Ube4a^−/− mice were described in SI Appendix. All animal experiments were conducted in accordance with approved institutional animal care and use committee (IACUC) protocols of the Institute of Zoology, Chinese Academy of Sciences (Ethics No. 08-133), and Guangzhou Medical University (Ethics No. G2022-149).

Immunoblotting and Coimmunoprecipitation.

Tissues and transfected cells were lysed in ice-cold RIPA or TAP buffer, respectively, each supplemented with protease inhibitors. Lysates were centrifuged to collect supernatants. For coimmunoprecipitation, clarified lysates were incubated with specific antibodies and magnetic beads at 4 °C. Immunoprecipitants were then washed, resuspended in Laemmli buffer, and boiled. All protein samples were subsequently analyzed by SDS-PAGE followed by immunoblotting.

Quantification and Statistical Analysis.

Western blot signals and fluorescence intensities were quantified using ImageJ and analyzed with GraphPad Prism. The numbers of Zip3p, Rad51p, Ufd2p, Msh4p, MLH1, RPA, and HEI10, as well as the fluorescence intensities of Top2p, TOP2A, and bTMP, are presented as means ± SE. The abundances of Top2p, Ufd2p, and Pds5p measured by immunoblotting, CoC analyses, and the distributions of Zip3p numbers derived from both experimental data and simulations were reported as means ± SD. Sample sizes (n) for each experiment are specified in the corresponding figure legends. Data are presented as means ± SD, means ± SE, or 95% CI, as indicated in the figure legends. Statistical significance of differences between group means was assessed using a Student’s t test, while differences in frequencies were evaluated using a two-proportion Z-test. Statistical significance is denoted in the figures and figure legends as follows: ns, no significant; P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001.

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Dataset S02 (XLSX)

Dataset S03 (XLSX)

Dataset S04 (XLS)

Data, Materials, and Software Availability

The Ufd2p ChIP-seq data have been deposited in the NCBI Gene Expression Omnibus (GEO) under Accession No. GSE294728. Other ChIP-seq datasets were obtained from GEO or the Sequence Read Archive (SRA) under the following Accession Nos.: GSE103877 (Zip2, Zip3p, and Zip4p) (70), GSE169760 (Top2p) (71), and GSE201366 (DNA negative supercoils) (26). Proteomic and ubiquitinomic data have been deposited in China National Center for Bioinformation (CNCB) under Accession No.: OMIX013815. All other data are included in the manuscript and/or the supporting information.