Distinct roles of spindle checkpoint proteins in meiosis

SUMMARY Gametes are produced via meiosis, a specialized cell division associated with frequent errors that cause birth defects and infertility. Uniquely in meiosis I, homologous chromosomes segregate to opposite poles, usually requiring their linkage by chiasmata, the products of crossover recombination. 1 The spindle checkpoint delays cell-cycle progression until all chromosomes are properly attached to microtubules, 2 but the steps leading to the capture and alignment of chromosomes on the meiosis I spindle remain poorly understood. In budding yeast meiosis I, Mad2 and Mad3 BUBR1 are equally important for spindle checkpoint delay, but bio-rientation of homologs on the meiosis I spindle requires Mad2, but not Mad3 BUBR1 . 3,4 Here we reveal the distinct functions of Mad2 and Mad3 BUBR1 in meiosis I chromosome segregation. Mad2 promotes the pro-phase to metaphase I transition, while Mad3 BUBR1 associates with the TOGL1 domain of Stu1 CLASP , a conserved plus-end microtubule protein that is important for chromosome capture onto the spindle. Homologous chromosome pairs that are proﬁcient in crossover formation but fail to biorient rely on Mad3 BUBR1 - Stu1 CLASP to ensure their efﬁcient attachment to microtubules and segregation during meiosis I. Furthermore, we show that Mad3 BUBR1 -Stu1 CLASP are essential to rescue the segregation of mini-chromosomes lacking crossovers. Our ﬁndings deﬁne a new pathway ensuring microtubule-dependent chromosome capture and demonstrate that spindle checkpoint proteins safeguard the ﬁdelity of chromosome segregation both by actively promoting chromosome alignment and by delaying cell-cycle progression until this has occurred. RESULTS


In brief
Mukherjee et al. identify differential roles of Mad2 and Mad3 BUBR1 in meiosis, outside their canonical functions in delaying anaphase onset as part of the spindle checkpoint.While Mad2 promotes meiotic progression prior to metaphase I, Mad3 BUBR1 interacts with the microtubule regulator Stu1 CLASP to promote chromosome capture onto the spindle.The mitotic checkpoint complex (MCC) comprising Mad2-Mad3 BUBR1 -Bub3-Cdc20, an inhibitor of the anaphase promoting complex (APC Cdc20 ), is the canonical effector of the spindle checkpoint. 2 Accordingly, both Mad2 and Mad3 BUBR1 are required to impose a metaphase I delay in response to either unattached kinetochores or a lack of inter-homolog tension in meiosis. 3,4Furthermore, in an unperturbed meiosis, metaphase I and metaphase II are similarly shortened in the absence of either MAD2 5 or MAD3 or in the double mutant (Figures S1A-S1C).Although Mad3 BUBR1 was suggested to delay meiotic prophase, 6 the prophase marker, Zip1, a component of the synaptonemal complex, was present for a similar duration in wild-type, mad2D, mad3D, and mad2D mad3D cells (Figures S1D and  S1E).Despite their shared role, Mad2 and Mad3 BUBR1 also perform distinct functions in meiosis, 4,6 and some mutations affecting kinetochore or spindle function show differential synthetic interactions with mad2D and mad3D. 7Consistently, Mad2, but not Mad3 BUBR1 , is required for homolog biorientation in an unperturbed meiosis 4 (Figures S2A and S2B).Furthermore, we found that mad2D and mad3D have additive effects on the non-disjunction of homologs during meiosis I. Live imaging of meiotic cells with both chromosome V homologs labeled close to the centromere (CEN5-tdTomato) and carrying a spindle marker (GFP-TUB1) showed that homolog non-disjunction in meiosis I was only mildly elevated in mad3D ($5%), reached $10%-15% in mad2D, but was raised to $20% in mad2D mad3D (Figures 1A-1C).Together, these observations reveal distinct spindle checkpoint-independent functions of Mad2 and Mad3 BUBR1 in meiosis I (Figure 1D).

Mad2 ensures orderly M phase events after prophase exit
In C. elegans germ cells and cultured human cells, Mad1-Mad2, but not BubR1(Mad3), ensure a timely G2-mitosis transition. 8ad1-Mad2 curtails APC Cdc20 activity prior to mitosis independently of MCC formation to allow a threshold level of cyclin B to accumulate. 8Similarly, our live imaging of meiotic yeast cells carrying the kinetochore label Mtw1-tdTomato and the anaphase marker, Cdc14-GFP, showed that Mad1 and Mad2, but not Mad3 BUBR1 , promote the transition from prophase exit to metaphase I. Dispersed kinetochores in prophase re-cluster at prophase exit and split into two Mtw1-tdTomato clusters at metaphase I 9 (Figure 1E).Cdc14-GFP phosphatase is released from the nucleolus in anaphase I and II to reverse M phase phosphorylation 10,11 (Figure 1E).In wild-type and mad3D cells, Cdc14 release invariably followed the splitting of (legend continued on next page) the Mtw1-tdTomato cluster into two, reflecting the strict sequential order of metaphase I and anaphase I (Figures 1E  and 1F).Surprisingly, however, in mad1D and mad2D cells, Cdc14 release frequently occurred prior to the splitting of Mtw1-tdTomato (Figures 1E-1H).Consistent with their shorter metaphase I (Figure S1B), the time between prophase exit (kinetochore re-clustering) and anaphase I (Cdc14 release) was reduced in mad1D and mad2D cells (Figure 1G).However, mad1D and mad2D, but not mad3D, cells were delayed in reaching metaphase I after prophase exit (Figure 1I).Therefore, Mad1 and Mad2, but not Mad3 BubR1 , impose an order on M phase events and promote the timely transition between prophase and metaphase I. Premature Cdc14 activation leading to the untimely reversal of key phosphorylations and/or due to ectopic APC Cdc20 activity, as in C. elegans, 8 could cause defective homolog biorientation in mad2D cells.However, bulk Rec8 loss occurs at anaphase I onset in mad2D cells (Figure S1B), indicating that not all cell-cycle events downstream of APC Cdc20 are disrupted in mad2D cells.S1).Mad2 predominantly bound Mad1, Bub1, and Cdc20, while Mad3 BUBR1 was not enriched (Figure 2B), potentially indicating formation of a regulatory complex distinct from the MCC involved in promoting a timely prophase-metaphase I transition (Figure 1I), similar to that described in C. elegans. 8In contrast, Bub3 is the major Mad3 BUBR1 interactor, consistent with known direct binding events [12][13][14][15] (Figure 2A).Proteins of the outer kinetochore, including Ndc80 and Spc105 KNL1 -Kre28, to which Bub1 binds directly as part of its role in the spindle checkpoint, 16 were found only in Mad2 immunoprecipitates.In contrast, several distinct proteins, including tubulin subunits (Tub1, Tub2, and Tub3), three distinct phosphatases (Ptc3, Ptc7, and Rts1, a subunit of PP2A-B56), and the microtubuleregulator Stu1 CLASP , along with its binding partner Slk19 CENPF , were significantly enriched in the Mad3 BUBR1 purification.Direct comparison confirmed that Stu1 CLASP , Slk19 CENPF , and Ptc7, in addition to Bub3, were most significantly enriched in Mad3 BUBR1 over Mad2 purifications (Figure 2C).Since Mad3 BUBR1 promotes homolog segregation independently of the spindle checkpoint or Mad2 (Figures 1A-1C), relevant interactions should persist upon spindle checkpoint inactivation.Comparison of Mad3-FLAG interactors with or without Mad2 (to abrogate the checkpoint) in Cdc20-depleted cells (metaphase I arrest) showed that while association of Slk19 CENPF with Mad3 BUBR1 was greatly diminished in mad2D cells, the interactions with Stu1 CLASP and Ptc7 were maintained (Figure 2D).Stu1 CLASP , but not Slk19 CENPF , was also moderately enriched in Mad3 BUBR1 immunoprecipitates from prophase I-arrested cells (Figure S3B), suggesting that Stu1 interacts with Slk19 only at prophase exit, potentially in response to checkpoint activity.We conclude that Mad3 BubR1 associates with Stu1 CLASP independently of the spindle checkpoint and Mad2.
Mad3 BUBR1 interacts with Stu1 CLASP through its TOGL1 domain Stu1 CLASP is a member of the conserved CLASP family of microtubule regulators that suppress catastrophes and promote rescue of plus ends to direct chromosome capture and alignment in mitosis. 18We reasoned that the N-terminal TOGL1 domain (Figure 2E), which mediates Stu1 CLASP localization to kinetochores in mitosis, but not its binding to microtubules or viability, 17 2H).Therefore, Mad3 BUBR1 association with Stu1 CLASP requires its N-terminal TOGL1 domain.Slk19 CENP-F was also absent in Mad3-FLAG immunoprecipitates from stu1DTOGL1 prometaphase I cells, indicating that the Mad3-Slk19 CENP-F interaction is likely to be indirect, via Stu1 CLASP (Figure 2F).STU1 is essential for microtubule organization in mitosis and viability. 19Although virtually all wild-type and mad3D cells  underwent meiosis to produce four gametes, called a tetrad, over 75% of pCLB2-STU1 cells failed to produce spores, with the remaining cells predominantly producing dyads (two spores), while 80% of stu1DTOGL1 cells formed tetrads (Figure S3C).Therefore, although Stu1 is critical for sporulation, its N-terminal TOGL1 domain, which is required for Mad3 BUBR1 association, is not.Anaphase I spindles were observed in live wild-type, mad3D, and stu1DTOGL1 cells carrying GFP-TUB1 (to label microtubules), but not in pCLB2-STU1 cells (Figures S3D and S3E), indicating that Stu1 is required for bipolar spindle formation in meiosis I, similar to mitosis. 19Finally, purification of Stu1-GFP or Stu1D TOGL1-GFP revealed no major changes in interaction partners, including the retention of tubulin binding (Figures S3F-S3H), consistent with proficient meiotic spindle formation in stu1D TOGL1 cells (Figures S3D and S3E).We note that Mad3 BUBR1 was not recovered in either Stu1-GFP or Stu1DTOGL1-GFP immunoprecipitates (Figures S3F-S3H), indicating that only a minor fraction of cellular Stu1 CLASP interacts with Mad3 BUBR1 .Consistently, we found that Stu1 is around 8-to 10-fold more abundant than Mad3 in prophase and throughout the meiotic divisions in our recent whole-proteome dataset 20 (Figure S3I).We conclude that stu1DTOGL11 is a separation of functional allele that loses Stu1 CLASP interaction with Mad3 BUBR1 while retaining its ability to organize microtubules.
Stu1 CLASP kinetochore localization in meiosis does not require TOGL1 In mitosis, Stu1 is recruited to kinetochores via its TOGL1 domain, where it is particularly enriched when microtubules are not attached. 17However, Stu1DTOGL1-GFP localized similarly to wild-type Stu1-GFP in meiotic cells, being recruited to kinetochores at prophase exit before localizing also to the spindle in metaphase I and II, or the spindle midzone in anaphase I and II (Figures S4A and S4B).Quantification of the kinetochore localization at prophase I exit, when kinetochores cluster prior to spindle formation, revealed a small, but not statistically significant, decrease in Stu1DTOGL1-GFP compared to Stu1-GFP (Figures S4C and S4D).Stu1DTOGL1-GFP also localized to unattached kinetochores in meiotic cells treated with the microtubule-depolymerizing drug benomyl (Figure S4E), and its localization was independent of Mad3 (Figure S4F).It is unclear why the Stu1 TOGL1 domain is required for kinetochore localization in mitosis, 17 but not meiosis.We speculate that the CL domain of Stu1 CLASP (Figure 2E), which contributes to kinetochore localization in mitotic cells, 17 may be more important in meiosis, where kinetochores have a modified organization. 21d3 BUBR1 and the TOGL1 domain of Stu1 CLASP work together to promote meiosis I chromosome segregation We employed the stu1DTOGL1 separation-of-function allele to test the function of the Stu1 CLASP -Mad3 BUBR1 interaction.stu1DTOGL1 had no significant effect on the duration of prophase, metaphase I, or metaphase II, and mad2D and/or mad3D shortened metaphase I and II independently of stu1D TOGL1 (Figure S1).Therefore, the TOGL1 domain of Stu1 is not required for the canonical spindle checkpoint.Next, we tested whether Mad3 BUBR1 mediates meiotic chromosome segregation through Stu1 CLASP , in which case meiosis I nondisjunction rates in mad3D stu1DTOGL1 cells would be expected to be comparable to either single mutant.Homologs disjoined to the same pole in $5% of mad3D cells, as described above, while this was increased to $11% in stu1DTOGL1 cells, though this difference was not statistically significant (p = 0.37; Figure 3A).Homolog mis-segregation in mad3D stu1DTOGL1 cells was $13%, similar to stu1DTOGL1 alone (Figure 3A; p = 0.99).Therefore, although Stu1TOGL1 may have additional functions to Mad3 BUBR1 , the fact that mad3D does not increase homolog mis-segregation in stu1DTOGL1 cells indicates that Mad3 BUBR1 works in the same pathway as the TOGL1 domain of Stu1 CLASP .Since Mad3 BUBR1 function in meiosis I chromosome segregation is most evident in the absence of MAD2 (Figure 1B), we assessed the stu1DTOGL1 mutant in the mad2D background.This revealed that mad2D exacerbated the meiosis I chromosome segregation defect in mad3D and stu1DTOGL1 cells to a similar extent (Figure 3A).However, additional additive effects were not observed in the triple mad2D mad3D stu1DTOGL1 mutant.Therefore, Mad3 BUBR1 and the TOGL1 domain of Stu1 CLASP act in the same genetic pathway to promote meiosis I homolog segregation.In contrast, Mad2 acts in a distinct pathway (Figures 1B and 3A) and, unlike Mad3 BUBR1 , is important for homolog biorientation during meiosis I 4 (see also below) and for ordering M phase events (Figures 1E-1I).Taken together, our findings indicate that Mad3 BUBR1 -Stu1 CLASP rescues the  17 (F) List of proteins and their unique peptide counts as identified by one repeat of mass spectrometry after immunoprecipitation of Mad3-FLAG in the indicated strains.Note that all three strains have heterozygous pCLB2-STU1, with the other allele as indicated.Cells were harvested 60 min after release from prophase I. (G and H) Mad3 interaction with Stu1 is lost in stu1DTOGL1 cells.Volcano plots after mass spectrometry showing the relative enrichment of proteins immunoprecipitated with Mad3-FLAG in (G) wild-type versus stu1DTOGL1 and (H) wild-type versus pCLB2-STU1 prophase I-arrested cells.In (A)-(D), (G), and (H), the absence of a colored dot for a kinetochore protein, phosphatase, or tubulin in the volcano plot means that it was not detected in this experiment.See also Figures S3 and S4 and Table S1.segregation of homologs that fail to biorient, while Mad2 functions in an independent pathway.The Mad3 and Stu1-TOGL1 pathway rescues the segregation of chromosomes that lack crossovers Crossover recombination generates chiasmata that provide linkages between homologs, ensuring accurate meiosis I segregation.3 The occasional failure of a crossover can be tolerated, and so-called achiasmate or non-exchange chromosomes can undergo proper meiosis I segregation around 80% of the time, though the underlying mechanisms are not well understood.[22][23][24][25][26] Interestingly, Mad3 BUBR1 , unlike Mad2, is essential for the segregation of such an achiasmate chromosome pair.6 Current models posit that the synaptonemal complex, which zips homologous chromosomes together as they recombine, persists at centromeres to maintain homolog pairing beyond prophase to rescue the segregation of chromosomes that fail to cross over.27,28 However, Mad3 BUBR1 was found to act independently of centromere pairing and was instead proposed to mediate a prophase delay to allow achiasmate chromosome segregation.6 Because our live-cell imaging found no prophase delay in mad3D cells (Figure S1E), we instead hypothesized that Mad3 BUBR1 directs achiasmate chromosome segregation via engaging Stu1 CLASP .To test this idea, we introduced a pair of centromeric mini-chromosomes, one labeled with tdTomato (tetO-TetR-tdTomato), the other with GFP (lacO-GFP-LacI), into mad3D, stu1DTOGL1, and mad3D stu1DTOGL1 cells (Figure 3B).Since the mini-chromosomes are both small and divergent in sequence, they will not cross over and therefore represent a pair of achiasmate chromosomes.29 In wild-type cells, red and green signal segregated to the same pole at anaphase I in 25% of cells, while segregation of the achiasmate mini-chromosomes was essentially random ($50%) in mad3D cells, as expected 6 (Figure 3C).Crucially, stu1DTOGL1 cells, where the Mad3 BUBR1 -Stu1 CLASP interaction is abolished (Figures 2F-2H), also exhibit random segregation of achiasmate mini-chromosomes in meiosis I, as does the mad3D stu1DTOGL1 double mutant (Figure 3C).Therefore, the ability of Stu1 CLASP to bind Mad3 BUBR1 is critical for achiasmate chromosome segregation.
Mad3 BUBR1 enables chromosome-spindle interactions through Stu1 CLASP How might Stu1 CLASP -Mad3 BUBR1 contribute to the fidelity of chromosome segregation?1][32] Therefore, Mad3 BUBR1 may promote kinetochore capture or biorientation via Stu1 CLASP .To test this, we analyzed the position of CEN5-tdTomato foci relative to the metaphase I spindle prior to elongation at anaphase I (Figures 4A-4C).We considered instances where CEN5-tdTomato was located centrally on the spindle  axis as ''bioriented,'' while asymmetric CEN5-tdTomato foci on the spindle axis were scored as ''off center'' and foci that did not co-locate with the GFP-Tub1 signal were scored as ''off axis'' (Figure 4A).Consistent with a previous report, 4 mad2D showed defective biorientation, manifest as a significantly increased frequency of ''off-center'' CEN5-tdTomato foci (Figure 4B).However, compared to wild type, neither mad3D nor stu1DTOGL1 exhibited defective biorientation unless MAD2 was also deleted (Figure 4B).In contrast, the fraction of cells where CEN5-tdTomato was ''off axis,'' suggesting defective kinetochore capture by microtubules or stabilization of this attachment, was significantly increased over wild type in both mad3D and stu1DTOGL1 mutants whether or not Mad2 was present (Figure 4C).These data show that Mad2 is important for positioning chromosomes in the center of the spindle axis (bioriented), while Stu1 CLASP -Mad3 BUBR1 is important for chromosome association with the spindle.Therefore, Mad2 and Stu1 CLASP -Mad3 BUBR1 represent two distinct chromosome segregation pathways that respectively promote biorientation and chromosome-microtubule interactions, possibly through initial capture (Figure 4D).

Conclusions
The spindle checkpoint prevents catastrophic segregation in response to improper kinetochore-microtubule attachments.
Here we provide evidence that the components of this surveillance mechanism also contribute directly to the correction of improper or absent kinetochore-microtubule attachments.Incorporating these two key activities within the same proteins allows coordination of surveillance with segregation-promoting mechanisms (Figure 4D).Mad2 couples cell-cycle events as cells transition from prophase exit into mitosis, which could explain its role in promoting sister kinetochore biorientation, though the mechanism remains unclear.Mad3 BUBR1 facilitates chromosome alignment through Stu1 CLASP -dependent chromosome capture.We demonstrate that this pathway is critical to rescue chromosomes from lack of crossovers or a failure to biorient.Similar mechanisms may operate in mouse oocytes since Mad2 has a non-canonical function in curtailing APC activity at meiosis I exit 33 and BUBR1 is required for robust kinetochore microtubule attachments. 34Safeguarding mechanisms such as those we identify here are therefore likely to play key roles in protecting against errors that lead to aneuploidy in human meiosis.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:  2H, S3B, and S3F-S3H) of meiotic culture grown at OD 600 = 2.5 was harvested and washed once with sterile dH 2 O. Cells were pelleted and resuspended in 20% v/w 2 mM PMSF and snap frozen as small 'noodles' by releasing drops of cells into liquid nitrogen.These noodles were filled in metal canisters pre-cooled in liquid nitrogen and cells lysed by 5 rounds of 30/s speed for 3 min each in the twin bio-pulverizer Retsch MM400.Grindate was then emptied out of the canisters into a 50mL falcon tube and stored at À80 C. For immunoprecipitation, the cryogrindate was thawed and resuspended in 20% w/v H0.15M lysis buffer (25mM HEPES pH8, 2mM MgCl 2 , 0.1mM EDTA pH8.0, 0.5mM EGTA-KOH pH8.0, 15% glycerol, 0.1% NP-40, 150mM KCl) with phosphatase and protease inhibitors (CLAAPE, comprising 10 mg/mL each of chymostatin, leupeptin, antipain, pepstatin and E64, together with 2mM AEBSF, 0.8mM Na Orthovanadate, 0.2uM microcystin, 1x EDTA-free Roche protease inhibitor tablet, 2mM NEM, 4mM b-glycerophosphate, 2mM Na pyrophosphate, 10mM NaF).40U/ml of Benzonase (Novagen) was added to the lysate and incubated for 1h at 4 C with rotation to digest DNA.Samples were centrifuged for 10 min at 4000 rpm at 4 C and supernatant was collected in new pre-chilled falcon tubes.Protein concentration was determined by Bradford assay, and each lysate was adjusted to the same volume and protein concentration.50mL of each adjusted lysate was added to 10mL 4xLDS + 5% b-mercaptoethanol, boiled at 95 C for 5min and stored at À20 C as input.2mg ⍺-GFP (Roche) or 0.05 mg ⍺-FLAG (Sigma) previously conjugated to Protein G-dynabeads were added to each sample and incubated with rotation at 4 C for 2.5h.Dynabeads were concentrated using a pre-chilled magnet and the flow through was discarded.The beads were transferred into an eppendorf tube and washed once with buffer H0.15M with inhibitors and 2mM DTT, then three more times with buffer H0.15M with inhibitors.Beads were concentrated on the magnet, resuspended in 50mL 1xLDS + 5% b-mercaptoethanol and boiled at 70 C for 10min to elute.Samples were spun down at 13,200rpm for 5min before the eluate was transferred to a fresh eppendorf tube and stored at À20 C indefinitely for preparation for mass spectrometry.
In gel digestion of protein samples for mass spectrometry In-gel digestion was used to prepare samples for mass spectrometry in Figures 2A-2D and 2F.Yeast growth conditions and the immunoprecipitation protocol used was the same as above, with a few modifications.3L of SPO cultures at OD 600 = 2.5 were harvested and 500mL Protein G dynabeads previously conjugated to 50mL M2 ⍺-FLAG antibody were added to each extract which was made from approximately 15g of cryogrindate.Proteins were eluted from beads in 100mL 1xLDS + 5% b-mercaptoethanol, out of which 90mL was loaded on NuPAGE Novex 4-12% Bis-Tris Gel (Life Technologies) gels and run for 6min so that all proteins enter the gel.The gel was stained by incubating with agitation in Instant Blue (Abcam) and washed three times for 5min each with dH 2 O. Protein bands were cut from the gel and chopped into $1mm 3 pieces using a new clean scalpel, and the pieces were collected in an eppendorf tube.The pieces were submerged in 50mM ammonium bicarbonate (ABC) for 30min.ABC was discarded and 100% acetonitrile (ACN) was added until gel pieces were submerged and incubated for another 30min.$80ul 10mM DTT in 50mM ABC was added to the gel pieces and incubated for 30 min at 37 C. DTT solution was removed and gel pieces were resuspended in ACN for 5min and any excess liquid was removed.$80ul 55mM iodoacetamide dissolved in ABC was added to cover the pieces and incubated in the dark at RT for 20mins.The liquid was removed and gel pieces were incubated with 50mM ABC buffer for 5 min at 37 C, the ABC was removed, and then the gel pieces were incubated in ACN for 5 mins at 37 C.All liquid was removed and enough trypsin digestion mix (0.013 mg/ml trypsin, 10% ACN, 10mM ABC) was added to cover the gel pieces and left initially at 4 C and then at 37 C overnight for 12-15h in a moist chamber.0.1% or 10% TFA was added to the gel pieces in trypsin digestion mix to stop over-digestion of peptides and the solution was kept at room temperature for 15min to allow all peptides to diffuse out form the gel.1mL of sample was dropped on a pH paper to confirm that the solution has pH < 2.0.Filter-aided sample preparation (FASP) of protein samples FASP was used to prepare samples for mass spectrometry in Figures 2G, 2H, S3B, and S3F-S3H, as described, 48 with a few modifications.Proteins were eluted from beads by incubating in 30mL 0.1% Rapigest (Waters) dissolved in 50mM ABC at 37 C for 30min, removing the eluate and then repeating to obtain a further 30mL of eluate.Pooled eluates from the two elutions were stored at À20 C. On the day of trypsin digestion, 10% volume of 1M DTT was added to samples and boiled for 5 min at 95 C with agitation.Tubes were cooled to room temperature before adding 3x vol of 8M urea in 100mM Tris-HCl pH8.0 (UBB) to each sample.The whole sample was transferred onto Sartorius Stedim Biotech's Vivacon 500 MWCO 30 000 VN01H21 column and spun down at 10,000rpm for 10-15 min at room temperature to bind all peptides to the membrane.100mL of 55mM iodoacetamide dissolved in UBB was added, the tube shaken at 600rpm for 1 min at RT in a theromixer, and incubated in the dark for 30min before spinning the buffer through the column.The column was then washed once with 100mL UBB and twice with 100mL ABC.The column was completely dried before adding 60mL of trypsin digestion mix (0.013 mg/ml trypsin, 0.002% TFA, 50mM ABC) onto the column membrane.Columns were capped and sealed with parafilm before being shaken at 600rpm for 1 min at room temperature and then incubated at 37 C overnight for $15h in a moist chamber.Parafilm was removed and the columns were centrifuged to elute trypsin-digested peptides into new protein protein LoBind tubes containing 10mL of 10% TFA to stop the trypsin digestion.1mL of sample was dropped on a pH paper to confirm that the solution was pH < 2.0.

Mass spectrometry
Stage tips were prepared by inserting three Empire C18 disks (3M) inside a p200 pipette tip.20mL MeOH and 50mL 0.1% TFA was passed through the tip to calibrate the disks at the correct pH.All the liquid from the gel digestion or the in-column digestion was passed through the stage tip by microfuging for $10min.The tip was then washed again with 0.1% TFA and stored at À20 C.
Peptides were eluted in 40 mL of 80% acetonitrile in 0.1% TFA and concentrated down to 1 mL by vacuum centrifugation (Concentrator 5301, Eppendorf, UK).The peptide sample was then prepared for LC-MS/MS analysis by diluting it to 6 mL by 0.1% TFA.All LC-MS analyses were performed on an Orbitrap Fusion Lumos Tribrid Mass Spectrometer (Thermo Fisher Scientific, UK) both coupled on-line, to an Ultimate 3000 HPLC (Dionex, Thermo Fisher Scientific, UK).Peptides were separated on a 50 cm (2 mm particle size) EASY-Spray column (Thermo Scientific, UK), which was assembled on an EASY-Spray source (Thermo Scientific, UK) and operated constantly at 50 o C. Mobile phase A consisted of 0.1% formic acid in LC-MS grade water and mobile phase B consisted of 80% acetonitrile and 0.1% formic acid.Peptides were loaded onto the column at a flow rate of 0.3 mL min -1 and eluted at a flow rate of 0.25 mL min -1 according to the following gradient: 2 to 40% mobile phase B in 150 min and then to 95% in 11 min.Mobile phase B was retained at 95% for 5 min and returned back to 2% a minute after until the end of the run (190 min).Survey scans were recorded at 120,000 resolution (scan range 350-1500 m/z) with an ion target of 4.0e5, and injection time of 50ms.MS2 was performed in the ion trap at a rapid scan mode, with ion target of 2.0E4 and HCD fragmentation 49 with normalized collision energy of 27.The isolation window in the quadrupole was 1.4 Thomson.Only ions with charge between 2 and 6 were selected for MS2.Dynamic exclusion was set at 60s.

Analysis of mass spectrometry data
The MaxQuant software platform 39 version 1.6.1.0was used to process the raw files and search was conducted against the complete/reference proteome set of Saccharomyces cerevisiae SK1 strain (combined Saccharomyces Genome Database and in-house database -released in August 2019), using the Andromeda search engine. 50For the first search, peptide tolerance was set to 20 ppm while for the main search peptide tolerance was set to 4.5 p.m. Isotope mass tolerance was 2 ppm and maximum charge to 7. Digestion mode was set to specific with trypsin allowing maximum of two missed cleavages.Carbamidomethylation of cysteine was set as fixed modification and oxidation of methionine, was set as variable modification.Label-free quantitation analysis was performed by employing the MaxLFQ algorithm as described. 51Absolute protein quantification was performed as described. 52Peptide and protein identifications were filtered to 1% FDR.

Live cell imaging
To adhere cells, 5mL of ConA (5 mg/mL ConcanavalinA in 50mM CaCl 2 , 50mM MnCl 2 ) was spread at the bottom of chambers in 8-well glass-bottomed Ibidi dish (Thistle Scientific) using a plastic loop and incubated at 30 C for 15min.ConA was aspirated and the chambers washed three times with 500mL sterile dH 2 O and stored in the dark.To prepare cells for imaging, 10mL of meiotic cultures were started at OD 600 = 2.3 in SPO media.After 3h, 1mL of culture was spun down at 3000rpm for 1min.The pellet was resuspended in 300mL SPO media, added to the Ibidi dish and incubated for 20 min at 30 C. Wells were washed with 500mL SPO media twice before adding 400mL fresh SPO media.For the inducible Ndt80 block-release system, 200mL SPO was added while setting up the Ibidi dish, and another 200mL SPO with 2mM b-estradiol was added immediately before starting the time lapse imaging.For depolymerizing microtubules in Figure S6E, benomyl was pre-dissolved in SPO medium and was added at a final concentration of 50 mg/ml to the wells along with b-estradiol.Fluorescent microscopy was performed using Zeis Axioplan 2 microscope with 100x Plan ApoChromat NA 1.4 oil lens.Images were acquired through ORCA FLASH 4 CCD camera with auto-focus operated through Axiovision software and with 2x2 binning.GFP-Tub1 was imaged at 4% laser intensity for 80ms, CEN5-tdTomato was imaged at 4% intensity for 100ms, Mtw1-tdTomato was imaged at 10% intensity for 100ms or 150ms.Rec8-GFP was imaged at 5% intensity for 100 ms.Zip1-GFP was imaged at 1.5% intensity for 50 ms.Spc42-tdTomato was imaged at 5% intensity for 200 ms.Cdc14-GFP was imaged at 5% intensity for 100 ms.For all fluorescent channels, 9 z-slices of 0.7mm interval were captured.Brightfield was used for auto-focus and imaged only for the middle slice with 3V for 10ms.Chromosome segregation, metaphase I/II timing and Cdc14-GFP assays were imaged every 15min.Prophase timing (Zip1-GFP) was imaged every 8 min.Biorientation assays were imaged every 5min for 10h in total.

Image analysis
ImageJ software (National Institutes of Health) was used to max project the z-stacks and for visualising the images.To quantify signal intensity in Figure S6D, A circular region was drawn encompassing the region of GFP and tdTomato signal overlap, and the ratio of integrated density measurement of the GFP signal over the tdTomato signal was calculated.Final image assembly was conducted in Adobe Illustrator.

Imaging of GFP-labelled chromosomes in fixed cells
For the chromosome segregation assay reported in Figure S2, 150mL meiotic culture at OD 600 = 1.9 in SPO media was added to 15mL of 37% v/v formaldehyde in 1.5mL Eppendorf tubes and fixed for 8 min at room temperature.Tubes were then spun at 13,200rpm for 1min, supernatant removed and resuspended in 1mL 80% EtOH.Tubes were spun again for 30s, EtOH poured out, spun again for 15s and the remaining EtOH removed with a pipette.The pellet was resuspended in 20mL of 1 mg/ml DAPI and temporarily stored at 4 C for up to one week.3mL of cells were placed on a Superfrost microscope slide (Thermo Fisher Scientific), covered with a coverslip (VWR) and sealed with nail polish.The coverslip was pressed tightly against the slides and imaged on an Axioplan 2 microscope with 100x Plan ApoChromat NA 1.4 oil lens with 5% GFP, 10% tdTomato and 2% DAPI to visualize GFP and tdTomato dots and DNA.

Achiasmate minichromosome segregation assay
Diploid cells carried homozygous pURA3-GFP-LacI and heterozygous pURA3-TetR-tdTomato integrated into the genome, and AMp1963 and AMp1973 plasmids.The plasmids were maintained by selecting on synthetic complete glucose agar lacking both uracil and tryptophan (SD/-ura/-trp), before cells were inoculated consecutively in YPDA, YPA and SPO liquid media as described above for induction of meiosis and sporulation.Samples were collected 2h after inducing sporulation and every 30min thereafter until 4:30h.Cells were fixed and processed as described above ''Imaging of GFP-labelled chromosomes in fixed cells'', stored in 4 C and were counted on the same day.Only cells with a single GFP and single tdTomato focus at the binucleate stage were included in segregation scoring.Immunofluorescence 200mL of meiotic culture was centrifuged at 13,200rpm for 1min and resuspended in 500mL of 3.7% v/v formaldehyde in 0.1M KPi, pH6.4 (potassium phosphate buffer: 27.8mM K 2 HPO 4 and 72.2mM KH 2 PO 4 ) to fix overnight at 4 C. Cells were then spun down and washed three times with 1mL of 0.1M KPi buffer, and resuspended in 1mL of sorbitol-citrate (1.2M sorbitol, 0.1M KH 2 PO 4 , 36mM citric acid).Cells were spun down again and resuspended in digestion mix (200mL sorbitol-citrate, 20mL glusulase and 6mL 10 mg/ ml zymolase) and incubated at 30 C for 2h or until cells become phase-dark under light microscope.Once digested, cells were pelleted, washed with 1mL sorbitol-citrate and then resuspended in $50mL sorbitol-citrate.5mL of 0.1% polylysine was added to each well of multi-well slides (Thermo Fisher Scientific) for 5 min at room temperature before being rinsed with dH 2 O and air-dried.5mL of digested cells were added to each well and incubated for 10min, before aspiration and submerging in MeOH for 3min followed by acetone for 10s.5mL of Rat ⍺-tubulin (AbD Serotec) primary antibody diluted in 1:50 in PBS-BSA (1% w/v BSA, 0.04M K 2 HPO 4 , 0.01M KH 2 PO 4 , 0.15M NaCl, 0.1% w/v NaN 3 ) was added to each well and incubated in a moist chamber for 1h at room temperature.Primary antibody was aspirated and wells washed five times each with 5mL of PBS-BSA.5mL of Donkey anti-rat-FITC (Jackson ImmunoResearch) secondary antibody was added, and slides were incubated in a dark moist chamber for 1h, then each well was washed five times with 5mL of PBS-BSA.3mL of DAPI-mount (9mM p-phenylenediamine, 0.04M K 2 HPO 4 , 0.01M KH 2 PO 4 , 0.15M NaCl, 0.1% w/v NaN 3 , 50 ng/ml DAPI, 90% v/v glycerol) was added to each well, before the slide was covered with a glass coverslip and sealed with nail paint.Slides were stored at À20 C and visualized on a Zeiss Axioplan 2 microscope with 100x Plan ApoChromat NA 1.4 oil lens.

QUANTIFICATION AND STATISTICAL ANALYSIS
Statistical analysis and graphs were generated using Graphpad Prism 9 software (San Diego).Micrographs and graphs were assembled using Adobe Illustrator.Statistical details of all experiments are given in the figure legends.
Distinct, checkpoint-independent functions for Mad2 and Mad3 BUBR1 in meiosis I chromosome segregation

Figure 1 .
Figure 1.Mad2 and Mad3 BUBR1 have distinct, spindle checkpoint-independent functions in meiosis I (A-C) mad2D and mad3D show additive effects on homolog segregation in meiosis I. Cells carrying CEN5-tdTomato and GFP-TUB1 were induced to sporulate and live imaged.

Figure 2 .
Figure 2. Mad3 BUBR1 interacts with the TOGL1 domain of Stu1 CLASP (A-C) Immunoprecipitation and mass spectrometry of Mad2-FLAG and Mad3-FLAG during prometaphase/metaphase I. Volcano plots showing the relative enrichment of proteins immunoprecipitated with (A) Mad3-FLAG compared to no tag, (B) Mad2-FLAG compared to no-tag conditions, and (C) Mad3-FLAG versus Mad2-FLAG.Cells were harvested 75 min after release from a prophase I arrest (corresponding to prometaphase/metaphase I).(D) Mad3-FLAG interacts with Stu1 independently of the spindle checkpoint.Volcano plot showing the comparative enrichment of proteins identified by mass spectrometry in Mad3-FLAG immunoprecipitates from wild-type and mad2D cells harvested 6 h after inducing sporulation where progression beyond metaphase I was prevented by depletion of Cdc20 (pCLB2-CDC20).Rpl38 and Cft1 are likely contaminants.Results in (A)-(C) include data from three biological replicates, and (D) includes data from two biological repeats for each condition.Log 2 (fold change) between conditions is shown with corresponding p values.Dashed line indicates log 2 (fold change) = |2|.(E) Schematic of Stu1 protein with domains shown as identified by Funk et al.17 (F) List of proteins and their unique peptide counts as identified by one repeat of mass spectrometry after immunoprecipitation of Mad3-FLAG in the indicated strains.Note that all three strains have heterozygous pCLB2-STU1, with the other allele as indicated.Cells were harvested 60 min after release from prophase I. (G and H) Mad3 interaction with Stu1 is lost in stu1DTOGL1 cells.Volcano plots after mass spectrometry showing the relative enrichment of proteins immunoprecipitated with Mad3-FLAG in (G) wild-type versus stu1DTOGL1 and (H) wild-type versus pCLB2-STU1 prophase I-arrested cells.In (A)-(D), (G), and (H), the absence of a colored dot for a kinetochore protein, phosphatase, or tubulin in the volcano plot means that it was not detected in this experiment.See also FiguresS3 and S4and TableS1.

Figure 4 .
Figure 4. Mad3 BubR1 -Stu1 CLASP promote chromosome association with spindles during meiosis I (A-C) Live imaging reveals position and orientation of CEN5-tdTomato relative to the metaphase I spindle (GFP-Tub1).(A) Representative time series showing chromosome capture and alignment on the meiosis I spindle for the indicated genotypes and scenarios.Scale bar, 5 mm.(B and C) The percentage of cells where CEN5-tdTomato foci (which typically were observed as one or two foci) were located off the center of the spindle (B) or off the spindle axis (C) was scored in the last time point prior to anaphase I spindle elongation.Examples of correctly bioriented (gray box), off center (purple box), or off axis (yellow box) are shown in (A).Bar charts show mean of three biological replicates (n = 44-56, wild type; n = 52-62, mad3D; n = 36-52, stu1DTOGL; n = 41-49, mad3D stu1DTOGL1; n = 45-58, mad2D; n = 41-56, mad2D mad3D; n = 28-54, mad2D stu1DTOGL1; n = 25-58, mad2D mad3D stu1DTOGL1) with error bars representing standard deviation.ns, not significant; ****p < 0.0001, ***p % 0.001, one-way ANOVA (Tukey's multiple comparisons test).(D) Model for role of spindle checkpoint proteins in meiosis I chromosome segregation.Upon prophase exit, Mad1-Mad2 ensure Cdc14 phosphatase retention in the nucleolus to allow phosphorylation of key substrates important for progression to metaphase I.In prometaphase I, Mad3 engages Stu1 to facilitate chromosome association with microtubules.Also in prometaphase I, Mad2 ensures proper chromosome alignment in the center of the spindle through an unknown mechanism, potentially related to the earlier Mad2 function in preventing premature activation of Cdc14 phosphatase.Finally, in their canonical spindle checkpoint role, Mad2 and Mad3 assemble into the MCC to inhibit APC Cdc20 and delay anaphase I onset.

TABLE
d RESOURCE AVAILABILITY B Lead contact B Materials availability B Data and code availability d EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS B Yeast strains B Plasmids d METHOD DETAILS B Meiotic induction and growth of yeast cultures B Sporulation efficiency assay B Mass spectrometry methods