Aurora B kinase erases monopolar microtubule-kinetochore arrays at the meiosis I-II transition

iScience Article Aurora B kinase erases monopolar microtubulekinetochore arrays at the meiosis I-II transition Sergio VillaConsuegra, Vı ´ctor A. Tallada, Juan Jimenez [email protected] (V.A.T.) [email protected] (J.J.) Highlights Aurora B kinase re-localizes from mid-spindle to kinetochores at anaphase I exit Aurora B spindle midzone releases requires importin aImp1 function Kinetochore re-located Aurora B erases monopolar arrays at meiosis interkinesis Delaying Aurora B relocation leads to chromosome segregation errors in meiosis II Villa-Consuegra et al., iScience 26, 108339 November 17, 2023 ª2023 The Author(s). https://doi.org/10.1016/ j.isci.2023.108339 ll OPEN ACCESS iScience Article Aurora B kinase erases monopolar microtubule-kinetochore arrays at the meiosis I-II transition Sergio Villa-Consuegra, 1 Vı ´ctor A. Tallada, 1, *and Juan Jimenez 1,2, * SUMMARY During meiosis, faithful chromosome segregation requires monopolar spindle microtubule-kinetochore arrays in MI to segregate homologous chromosomes, but bipolar in MII to segregate sister chromatids. Using fission yeasts, we found that the universal Aurora B kinase localizes to kinetochores in metaphase I and in the mid-spindle during anaphase I, as in mitosis; but in the absence of an intervening S phase, the importin aImp1 propitiates its release from the spindle midzone to re-localize at kinetochores during meiotic interkinesis. We show that ‘‘error-correction’’ activity of kinetochore re-localized Aurora B becomes essential to erase monopolar arrangements from anaphase I, a prerequisite to satisfy the spindle assembly checkpoint (SAC) and to generate proper bipolar arrays at the onset of MII. This microtubulekinetochore resetting activity of Aurora B at the MI-MII transition is required to prevent chromosome missegregation in meiosis II, a type of error often associated with birth defects and infertility in humans. INTRODUCTION In eukaryotic cells, the microtubule-based mitotic spindle generates forces to align the condensed chromosomes at the metaphase plate and then pull the sister chromatids in opposite directions to segregate them into two daughter cells during anaphase. The spindle is composed of two types of microtubules. Dynamic kinetochore microtubules (KtMTs) capture and pull chromosomes, whereas interpolar microtubules (ipMTs) connect and separate the two spindle poles. 1–3 The kinetochore is the key structure assembled at the centromere that mediates the interactions of chromosomes with the KtMTs of the spindle. This large proteinaceous structure is composed of two submodules: the inner and outer kinetochore. While the inner kinetochore persists with centromeres throughout the cell cycle, the outer kinetochore attaches microtubules to the inner kinetochore and assembles only to condensed chromosomes during mitosis. 4 In fission yeast, however, in contrast to those of metazoans, most of the kinetochore components tend to be constitutive throughout the mitotic cell cycle, and only few outer factors associate at mitosis. 5–7 Kinetochores can initially bind to microtubules in any configuration, but in mitosis, precise chromosome segregation requires that the outer kinetochores of sister chromatids ultimately bind to KtMTs from opposite spindle poles (amphitelic arrangement). These bipolar junctions are carefully regulated by the error-correction mechanism driven by Aurora B kinase (Ark1) 8,9 and the spindle assembly checkpoint (SAC). 10,11 Incorrect arrangements are destabilized and only correct bioriented arrays are stabilized, ensuring accurate chromatids segregation. 12 The spindle segregates the sister chromatids during anaphase, and once segregated, the chromosomes decondense, and the KtMTkinetochore complexes disassemble to initiate a new cell cycle. In S. pombe, centromeres are clustered at the spindle pole bodies (SPBs) in the interphase of mitotic cycles. 13 Meiosis is a conserved event of sexual reproduction in eukaryotic organisms. Unlike the mitotic cycle, in which M-phase and S-phase alternate, meiosis consists of two rounds of chromosome segregation after a single round of DNA replication, leading to the generation of haploid gametes from diploid germ cells. In meiosis I (MI), homologous chromosomes segregate toward opposite poles (reductional segregation), whereas in meiosis II (MII), sister chromatids separate from each other (equational segregation). 14 These different types of chromosome segregation depend on different arrays of spindle attachment to chromosomes. While the bipolar MI segregation of homologous chromosomes requires sister kinetochores of each chromosome to be mono-oriented to capture microtubules from the same pole (syntelic arrangement), 15 in MII, biorientation occurs similarly to mitosis in that sister kinetochores are attached to microtubules from opposite spindle poles (amphitelic arrangement). 16 These distinct MI and MII arrangements of KtMTs-kinetochores are critical to produce haploid cells by two consecutive meiotic divisions. However, the events and mechanisms regulating the dynamics of these different KtMT-Kinetochore arrays at the time interval from the exit of anaphase I to the onset of premetaphase II (the MI-MII transition) are not well understood. In this study, by using imp1Dmutants that keep Ark1 sequestered at the midzone of hyperstable MI spindles, 17 we demonstrate that the relocation of 1 Centro Andaluz de Biologı ´a del Desarrollo, Universidad Pablo de Olavide/Consejo Superior de Investigaciones Cientı ´ficas, Carretera de Utrera Km1, 41013 Seville, Spain 2 Lead contact *Correspondence: [email protected] (V.A.T.), [email protected] (J.J.) https://doi.org/10.1016/j.isci.2023.108339 iScience 26, 108339, November 17, 2023 ª2023 The Author(s). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1 ll OPEN ACCESS Figure 1. Effects of delayed MI spindle disassembly on sister chromatid segregation (A) Duration (minutes) of meiosis I (MI), interkinesis (IK) and meiosis II (MII) in h90 wild-type (wt; n = 30) versus h90 imp1D(n = 30) zygotic cells measured from the spindle (mCherry-atb2) and SPB (Sid2-Tomato) dynamics. imp1Dzygotic cells show a negative IK time is due to the MI spindle degradation delay that leads to a temporal coexistence of both MI and MII spindles in the cell. On the right, dot plot comparison of the time (min) between the start of spindle nucleation in meiosis I and meiosis II (SI-SII time) in h90 wild-type (wt; n = 30) versus h90 imp1D(n = 30) zygotic cells. p value was calculated using Student’s ttest (p value ns). (B) Schematic set up of spindle dynamics in h90 wild-type (wt) vs. h90 imp1Dzygotic cells at meiosis I and meiosis II from Figure 1A data. ‘‘IK’’ represents the interkinesis period whereas ‘‘overlap’’ represents the time of coexistence of MI and MII spindle. (C) Time-lapse microscopy images of h90 wt and h90 imp1Dzygotic cells (dotted white lines) expressing mCherry-Atb2 (microtubule marker), Sid2-Tomato (SPB) and Ndc80-GFP (kinetochores) throughout MI and MII. Normal genome segregation (wt control) and uneven segregation of sister chromatids during MII (white ll OPEN ACCESS 2iScience 26, 108339, November 17, 2023 iScienc e Article Aurora B from the spindle midzone to kinetochores is required to reset KtMT-kinetocore arrangements at the MI-MII transition, a key step to ensure accurate chromosome segregation during the two consecutive meiotic divisions in the absence of an intervening S phase. RESULTS Delayed meiosis I spindle disassembly is associated with chromosome segregation defects in meiosis II By studying spindle dynamics in live fission yeast cells, we previously demonstrated that in mitosis, the importin aImp1 is essential for triggering spindle disassembly in the spindle midzone at the end of anaphase. 18 During meiosis, we also determined that zygotes depleted for Imp1 delayed spindle disassembly in meiosis I, indicating that, as in mitosis, this importin ais also essential for spindle dissolution in meiosis I, but not in meiosis II, where spindle disassembly is triggered by the virtual Nuclear Envelope Breakdown (vNEBD). 17 Surprisingly, we observed that in this imp1Dmutant background, MII spindles assembled even in the presence of persistent MI spindles, thus producing zygotic cells with coexisting MI and MII spindles. We observed that this property correlated with chromosome segregation defects in meiosis II, 17 suggesting that spindle coexistence may lead to MII chromosome missegregation. To better understand the mechanism leading to MI-MII spindle cohabitation in zygotic cells and its relationship with chromosome missegregation events, we examined the dynamics of spindles, SPBs and chromosomes in individual living zygotes (see STAR Methods). As shown in Figure 1A, despite the extensive delay in MI spindle dissolution caused by the imp1Dmutation, the MII spindle initiated assembly as in wt cells, with no significant delay after MI initiation. This observation suggests that spindle dynamics in MI and MII divisions are independent events programmed at the onset of meiosis, 19 resulting in frequent coexistence of MI and MII spindles when the timing of MI spindle dissolution overlaps with MII spindle assembly in imp1Dzygotic cells, without affecting the total duration of meiosis II (see Figure 1B). More importantly, we confirmed that the coexistence of MI and MII spindles may cause chromosome segregation errors in MII (Figures 1C–1F and S1F), as previously suggested. 17 Occasionally, tags used to mark kinetochore proteins may interfere with microtubule-kinetochore interactions. To discard tagging effects for this observation, we examined chromosome segregation related to spindle dynamics during meiosis using three independent kinetochore markers, the outer kinetochore protein Ndc80-GFP, 20 the p-lacO-lacI-GFP construct, 21 and the Mis6-GFP inner centromere marker. 22 As described above, persistent MI spindles in imp1Dzygotes resulted in frequent lagging chromosomes and missegregation in meiosis II irrespective of the marker used (Figures 1C–1E and S1 and Video S1), indicating that proper disassembly of the MI spindle is required for the accurate segregation of sister chromatids in MII. Strikingly, we observed that in imp1Dzygotic cells with coexisting MI and MII spindles, the non-segregating chromosomes in the second meiosis remained attached to the SPB that is still bound to MI spindle, referred here as the old SPB (Figure 1F). In Bub1-depleted budding yeasts, failure to correct initial meiotic attachments causes most chromosomes to travel to the old SPB. 23 Similarly, our results suggest that in these Imp1-depleted cells, the KtMTs-kinetochore arrays that segregated the homologous chromosomes likely persist into metaphase II. Altered dynamics of spindle assembly checkpoint factors in Imp1-depleted zygotes Such is the importance of proper assembly of spindle microtubules to sister kinetochores for accurate chromosome segregation, that eukaryotic cells have evolved a genetic pathway known as the spindle assembly checkpoint (SAC). 24 The SAC monitors the attachment of chromosomes to the spindle during metaphase, delaying the onset of anaphase until all chromosomes are properly attached and under tension. 25 This delay allows cells enough time to correct improper attachments, ensuring accurate chromosome segregation during mitosis 26 and in Figure 1. Continued arrows) with chromosome retention at MII onset (blue arrows) in imp1Dbackgrounds are shown. Numbers on top represent time in minutes. On the right, enlarged images highlight the mentioned stages of meiosis progression (metaphase II (84 min) and anaphase II (96 min)). Scale bar = 5mm. (D) Time-lapse microscopy images of h90 wt and h90 imp1Dzygotic cells (dotted white lines) expressing mCherry-Atb2 (microtubule marker), Sid2-Tomato (SPB) and Mis6-GFP (kinetochores) throughout MI and MII. Normal genome segregation (wt control) and uneven segregation of sister chromatids during MII with chromosome retention at MII onset (white arrows) in imp1Dbackgrounds are shown. Numbers on top represent time in minutes. On the right, enlarged image highlighting the referred stage of meiosis progression (metaphase II; 72 min). Scale bar = 5mm. (E) Time-lapse microscopy images of h90 wt and h90 imp1Dzygotic cells (dotted white lines) expressing mCherry-Atb2 (microtubule marker), Sid2-Tomato (SPB) and CENII-lacO lacI-GFP (centromere II) throughout MII. Normal genome segregation (wt control) and uneven segregation of sister chromatids during MII (white arrows) with chromosome retention at MII onset (blue arrows) in imp1Dbackgrounds are shown. Remanent MI spindle during MII is indicated (asterisk). Numbers at the top represent time in minutes. On the right, enlarged image highlights the mentioned stage of meiosis progression (anaphase II, 24 min). Scale bar = 5mm. (F) Time-lapse microscopy images of h90 wt and h90 imp1Dzygotic cells (dotted white lines) expressing mCherry-Atb2 (microtubule marker), Sid2-Tomato (SPB) and CENII-lacO lacI-GFP (centromere II) throughout MII. MI spindle persistence in imp1Dzygotic cells leads to the retention of chromosomes near the old SPBs (blue arrows), resulting in the appearance of lagging chromosomes (white arrows). Scale bar = 5mm. On the center, percentage of MII nuclei (n = 88) with centromere II (CENII) retained near the old SPB (the one attached to the MI spindle; dark gray), the new SPB (the one not attached to the MI spindle; black) or with no retention (wild-type-like; light gray). Segregation pattern of CENII was assessed in time-lapse fluorescence microscopy of h90 imp1Dzygotic cells with a fluorescent tagging at CENII (lacO insertions and expression of LacI-GFP). Retention phenotype was defined as meiosis II nuclei in which CENII is localized near one of the SPBs at prometaphase II for more than 5 min. On the right, percentage of meiosis II nuclei with CENII segregation defects in meiosis II of h90 wt (n = 136) and h90 imp1D(n = 189) zygotic cells. Lagging chromosomes (black) were defined as meiosis II nuclei in which at least one CENII dot does not colocalize with SPBs during anaphase II. Chromosome missegregation (gray) was defined as meiosis II nuclei in which a CENII signal is observed on a single SPB during anaphase II. ll OPEN ACCESS iScience 26, 108339, November 17, 2023 3 iScienc e Article ll OPEN ACCESS 4iScience 26, 108339, November 17, 2023 iScienc e Article both meiotic divisions. 27,28 Thus, the duration of metaphase-to-anaphase transition (referred here as metaphase duration) can be used as a readout of SAC activation. 19,29 To investigate whether persistent MI spindles found in Imp1-depleted zygotes could led to MII chromosome missegregation by interfering proper KtMT-kinetochore attachments, we analyzed SAC activation. To this end, we first determined the metaphase duration at MI and MII in imp1Das compared to wild-type (wt), and in different genetic backgrounds to either inactivate SAC (mad2D) or activate it specifically in meiosis I (rec12D) or meiosis II (clr4D) respectively. In rec12Dmutants, chromosomes bind to the spindle incorrectly due to the lack of chiasmata and therefore SAC delays the onset of anaphase in MI. 29 As shown in Figures 1C and 1D, both wt and imp1Dstrains segregate homologous chromosomes accurately in meiosis I. Both strains show also similar timing for anaphase I initiation and the rec12Dmutation delays anaphase onset in both strains in a similar fashion (see in Figure 2A, left panel). Depletion of Mad2 (mad2Dbackgrounds), which abolishes SAC function, 10 rendered also similar metaphase duration in imp1Dand in wild-type cells. These results indicate that SAC signaling is fully functional during MI in imp1Das in wt zygotic cells. Using the clr4Dmutation, in which SAC specifically delays anaphase initiation in meiosis II, 19,30 we forced SAC activation at this stage in wt and imp1Dbackgrounds. As expected, chromosome and spindle dynamics show that the clr4D-mediated SAC activation delays anaphase II onset in the wt background cells. However, this delay was poor in the imp1Dclr4Ddouble mutant indicating that the SAC, despite evident KT missegregation (see in Figure 1), is not fully active during MII in zygotes depleted for Imp1. In fact, metaphase duration in imp1Dclr4Dzygotes is similar to that observed in imp1Dmad2Dzygotes (Figure 2A, right panel), suggesting that the SAC is not properly signaling the errors leading to chromosome missegregations observed in imp1Dcells. Among all known components of SAC complex, Mad2 is at the bottom and Aurora B kinase (Ark1) at the top of the kinetochore localization hierarchy. 31 To obtain clues about the molecular bases of the SAC malfunction in Imp1-depleted cells, we next investigated the localization of these two key elements along the meiotic cycle. Mad2 is one of the best-characterized SAC factors. 25 Binding of this protein is stimulated by kinetochores that fail to attach to the spindle or to generate tension, in order to delay anaphase onset by directly inhibiting the APC/C Cdc20 activity. 32 Mad2-GFP localization was examined in living zygotes and its amount quantified by GFP fluorescence intensities during meiotic progression. In wild-type control cells, Mad2GFP localizes at kinetochores in metaphase I until late anaphase I and delocalizes at the MI-MII transition (interkinesis). This SAC factor is recruited again in pro-metaphase II to the kinetochores until sister chromatids successfully segregate in anaphase II. However, in imp1Dzygotes, an important fraction of Mad2-GFP remains at kinetochores during the MI-MII transition (Figures 2B and 2C). The persistence of Mad2GFP at the kinetochores, together with the previous observation that chromosomes remain attached to the old SPB at the MII anaphase onset (see Figure 1F), suggest that syntelic KtMT-kinetochores arrangements required for homologs segregation are not resolved during the MI-MII meiotic window in this mutant. Spindle midzone-kinetochore re-localization of the Aurora B-chromosomal passenger complex complex at the meiosis Imeiosis II transition In mitosis, Aurora B plays a well-established role in correcting erroneous KtMT-kinetochore attachments during the SAC response. 33–35 Loss of Aurora B in budding yeast (Ipl1) or fission yeast (Ark1) meiosis leads to massive chromosome missegregation due to the failure to correct initial erroneous attachments. 9,36–38 Aurora B forms a stable complex with INCENP (the Ark1-Pic1 complex in S. pombe), and together with borealin (Nbl1) and survivin (Bir1), constitutes the conserved chromosomal passenger complex (CPC). The CPC localizes to various regions at different times during mitosis, where the enzymatic subunit, Aurora B kinase, regulates key mitotic events. 39 In S. pombe mitosis, Aurora B-CPC does not associate with the SPB/kinetochore complex in interphase. However, following mitotic commitment, Ark1 concentrates particularly on prometaphase-metaphase kinetochores. During anaphase, Ark1 distributes along the spindle, and it gets restricted to the midzone as the spindle extends. 40 To assess Aurora B-CPC localization dynamics in wt and imp1Dmeiosis, we visualized Ark1-GFP but also other three GFP-tagged CPC components (Pic1, Nbl1, Bir1). All four proteins followed identical localization dynamics in the progression of meiosis Figure 2. Mad2 and Aurora B/Ark1 kinase analysis during the meiotic cycle (A) Duration of metaphase I (MTI; left) and duration of metaphase II (MTII; right) in minutes as considered in the above cartoons, corresponding to the phase I (initial spindle growth) and phase II (constant length) of spindle nucleation dynamics prior to spindle elongation at anaphase. The times were measured from zygotic cells analyzing the spindle (mCherry-atb2) and SPB (Sid2-Tomato) dynamics, in the control and mutant backgrounds indicated at the bottom. p values were calculated using non-parametric Kruskal-Wallis one-way ANOVA. Relevant significant (****p < 0.0001) and non-significant (ns) differences are indicated. (B) Time-lapse microscopy images of h90 wt and h90 imp1Dzygotic cells (dotted white lines) expressing mCherry-Atb2 (microtubule marker), Sid2-Tomato (SPBs) and Mad2-GFP throughout MI and MII. Normal Mad2 dynamics (white arrows) in wt control (upper panel) and abnormal persistence of Mad2 dots at interkinesis (IK; yellow arrows) in imp1Dbackground with MI spindle disassembly delay (overlapping). Numbers on top represent time in minutes. Scale bar = 5mm. (C) Mad2-GFP fluorescence intensity (AU) dynamics throughout meiosis I (PI: prophase I, MTI: metaphase I and AI: anaphase I) and meiosis II (PII: prophase II, MTII: metaphase II and AII: anaphase II) of h90 wt (blue; n = 6) and h90 imp1D(red; n = 6) zygotic cells. Average values (bold lines) and Standard Deviation (colored area) are indicated. Different meiotic stages are indicated within black dashed lines. (D) Fluorescence images of the h90 wt and h90 imp1Dcells (dashed white lines) expressing mCherry-Atb2 (microtubule marker), Sid2-Tomato (SPBs), Ndc80mTurquiose2 (Centromeres) and Pic1-GFP (CPC) during anaphase I and metaphase II (MTII) (from Figure S4). Pic1-GFP re-localizes to both daughter nuclei in MII, but under persistent MI spindles, Pic1-GFP remains trapped at the spindle midzone (white arrows) decreasing its protein amount in the kinetochores of MII (yellow arrows). Scale bar = 5mm. ll OPEN ACCESS iScience 26, 108339, November 17, 2023 5 iScienc e Article Figure 3. Kinesin-6 klp9 deletion (klp9D) restores Aurora B-CPC spindle midzone to kinetochore re-localization at interkinesis (A) Time-lapse fluorescence images of a representative h90 wt cell (left panel) and h90 imp1Dcell (right panel) expressing mCherry-Atb2 (microtubule), Sid2Tomato (SPBs), Pic1-GFP (CPC), and klp9-mTurquoise2. Remanent MI spindle during MII is indicated by asterisks. Normal midzone disassembly (wt control) ll OPEN ACCESS 6iScience 26, 108339, November 17, 2023 iScienc e Article (Figure S2). To avoid any potential tagging-related interference with Ark1 enzymatic functions, we followed Pic1-GFP/INCENP constructs to analyze Aurora B/Ark1 localization thought the Ark1-Pic1-GFP complex. 41,42 By analyzing the Ark1-Pic1-GFP complex, we observed that during MI, Aurora B-CPC accumulates at kinetochores from prophase to the onset of anaphase I, when this kinase redistributes to the spindle midzone during spindle elongation, similar to its localization during mitosis (Figures 2D upper panels and S2). 40,43–45 Midzone-associated Aurora B-CPC is then released upon spindle disassembly at the exit of meiosis I. Importantly, in the absence of an intervening S phase, released Ark1/CPC from the MI spindle midzone directly targets the kinetochores during the MI-MII transition (Figures 2D upper panels and S2). As compared to wild-type, imp1Dzygotes delay spindle midzone disassembly at MI, which often remains assembled as the zygote enters MII (Figure 1B). 17 Remarkably, in these imp1Dzygotes, Aurora B-CPC remains trapped at the persistent midzone during the MI-MII transition and enters MII with poor or undetectable Aurora B-CPC signal at kinetochores (Figures 3A, S2,S3A, and S3B). Since kinetochore localization of Aurora B kinase is required for error-correction activity at KtMT-kinetochore arrays and full SAC signaling in mitosis, 11,46 we hypothesize that sequestered Aurora B at the spindle midzone during the MI-MII transition hinders KtMT-kinetochore corrections and proper SAC action at the MII onset in imp1Dcells (Figure 2A), resulting in lagging and chromatids missegregation during anaphase II (Figures 1C–1F and S1). Ectopic release of the midzone-trapped Aurora B-chromosomal passenger complex restores its kinetochore re-localization in Imp1-depleted zygotes It has been described that Aurora B association with the spindle midzone in mitosis is dependent on Kinesin-6, 47,48 the Klp9 protein in fission yeast. 49,50 Accordingly, Pic1-GFP co-localizes with Klp9-GFP in the spindle midzone during anaphase I in S. pombe wild-type zygotes, and both proteins remain associated to the midzone until disassembly, at which point Aurora B-Pic1-GFP is found to relocate to MII centromeres (Figures 3A left and S2). Accordingly, in Imp1-depleted zygotes both proteins remain attached to the persistent MI spindle midzone, a small or no signal of Klp9 nor Pic1 being detected at kinetochores during the MI-MII transition (Figures 3A right and S2). Importantly, Klp9 depletion (klp9Dmutant) reduces CPC accumulation in the midzone (even though the spindle I remains assembled), significantly restoring its kinetochore localization (Figures 3B and 3C) and a wt-like dynamics of the Aurora B complex at the MI-MII transition in imp1Dcells (Figures 3B and 3D). We identified up to three abnormal phenotypical classes of Ark1-Pic1 localization defects with different frequencies in imp1Dmutant zygotes (cartoon in Figures 3D and S3), most of them being efficiently suppressed by Klp9 depletion (Figure 3D). Ase1 protein stabilizes interdigital spindle midzone. Lack of this protein lead to weaker and fragile spindles that randomly break down prematurely while elongating. 51–53 Accordingly, ase1Dsupresses hyperelongated spindles at mitosis in imp1Dcells. 18 Interestingly, imp1D ase1Dcells that break up the spindle before the interkinesis free up the Aurora B-CPC complex from the spindle midzone and, as described for Klp9 depletion, imp1Dase1Dzigotes significantly restores wt-like localization dynamics of Aurora B-Pic1-GFP at the MI-MII transition (Figures S6B and S6C). Thus, release of midzone-trapped Aurora B-CPC in imp1D,either directly througth Klp9 depletion or indirectly by premature spindle disolution in the ase1Dmutant background, re-establishes its kinetochore localization during interkinesis. Released Aurora B-chromosomal passenger complex restores normal chromosome MII segregation and spindle assembly checkpoint function in Imp1-depleted zygotes To determine whether midzone-sequestered Aurora B in persistent MI spindles leads to MII chromatid segregation errors, we studied spindle and chromosome dynamics in imp1Dzygotic cells depleted for the Klp9 Kinesin-6 along with respective controls. While the depletion of Klp9 have not a significant impact on the delay of MI spindle disassembly compared to imp1Dzygotes (Figure S5A), lack of Klp9 suppresses chromosome segregation errors at meiosis II in imp1Dzygotes (Figure 4A). MI and MII coexistent spindles can also be observed in imp1Dklp9D zygotic cells. Thus, segregation errors in meiosis II are unlikely due to abnormal interpolar spindle persistence itself but to Aurora B-CPC sequestration far away from the MII kinetochores. Similar results were observed by using Ase1 depletion to release AuroraB-Pic1-GFP from the midzone (Figure 4B). Quantitative analysis of chromosome retention in older SPBs during the MI-MII transition and segregation errors at the beginning of anaphase II in imp1Dzygotes reinforce this observation (Figures 4C, 4D, S5B, S6B, S6 and S6C). Furthermore, Ark1 release mediated by klp9 deletion significantly restores SAC function in metaphase II as well (Figure 4E). Thus, midzone-associated Aurora B-CPC release permits the zygote to erase monopolar arrays at kinetochores during the MI-MII transition, underlying a key step required to assembly sister KT-KtMT bioriented arrays and proper SAC function at the MII onset. Figure 3. Continued with Klp9-mTurq2 and Pic1-GFP foci disappearance (white arrows) and abnormal Klp9-mTurq2 and Pic1-GFP persistence at MII onset (yellow arrows) in imp1D backgrounds are shown. Numbers at the top represent time in minutes. Scale bar = 5 mm. (B) Time-lapse fluorescence images of a representative h90 klp9Dcell (top panel) and h90 klp9Dimp1Dcell (bottom panel) expressing mCherry-Atb2 (microtubule), Sid2-Tomato (SPBs), and Pic1-GFP (CPC). Pic1 localizes to the entire anaphase I spindle (yellow arrows). The lack of Klp9 allows its relocalization at prometaphase II nuclei (white arrows) even in the absence of imp1. Numbers at the top represent time in minutes. Scale bar = 5 mm. (C) Fluorescence intensity quantification (AU) of Aurora B-CPC (Pic1-GFP) at metaphase II kinetochores in h90 wt (n = 17), h90 imp1D(n = 15), h90 klp9D(n = 15), and h90 klp9Dimp1D(n = 15) zygotic cells. p values were calculated using an ordinary one-way ANOVA. Statistically significant (*p < 0.05; ****p < 0.0001) and non-significant (ns) differences are indicated (n R15). (D) Percentage of h90 wt (n = 20), h90 imp1D(n = 27), h90 klp9D(n = 16), and h90 klp9Dimp1D(n = 15) zygotic cells with the different types of Aurora B-CPC relocalization phenotypes that we appreciate (classified into Type I-IV drawn in Cartoons on the right; see also Figure S3). ll OPEN ACCESS iScience 26, 108339, November 17, 2023 7 iScienc e Article Figure 4. Deletion of Kinesin-6 klp9 or Ase1 restores sister chromosome segregation in imp1Dmeiosis II (A) Time-lapse fluorescence images of a representative h90 klp9Dimp1Dcell expressing mCherry-Atb2 (microtubule), Sid2-Tomato (SPBs), and Ndc80-GFP (kinetochores). Remnant MI spindle during MII is indicated by asterisks. Normal segregation of sister kinetochores during MII with no chromosome retention at MII onset (white arrows) in klp9Dimp1Dbackgrounds is shown. Numbers on top represent time in minutes. Scale bar = 5 mm. (B) Time-lapse fluorescence images of a representative h90 ase1Dimp1Dcell expressing mCherry-Atb2 (microtubule), Sid2-Tomato (SPBs), and Ndc80-GFP (kinetochores). Remanent MI spindle during MII is indicated (asterisks). Normal genome segregation of sister chromatids during MII with no chromosome retention at MII onset (white arrows) in ase1Dimp1Dbackgrounds is shown. Numbers at the top represent time in minutes. Scale bar = 5 mm. (C) Quantification of meiosis II nuclei with CENII segregation defects in meiosis II. Segregation pattern of centromere II was assessed in time-lapse fluorescence microscopy of h90 wt (n = 136), h90 imp1D(n = 189), h90 klp9D(n = 116), h90 klp9Dimp1D(n = 146), h90 ase1D(n = 146) and h90 ase1Dimp1D(n = 139) cells as before. Lagging chromosomes (black) and Chromosome missegregation (gray) was defined as previously described (see upper cartoon). p values were calculated using Chi-square test. Relevant significant (*p < 0.05) and non-significant (ns) differences are indicated. ll OPEN ACCESS 8iScience 26, 108339, November 17, 2023 iScienc e Article STAR+METHODS KEY RESOURCES TABLE RESOURCE AVAILABILITY Lead contact Further information and requests for resources, strains and reagents should be directed to and will be fulfilled by the lead contact, Juan Jimenez ([email protected]). Materials availability Plasmids and strains generated are available upon request to the lead contact. Data and code availability All data reported in this paper will be shared by the lead contact upon request. This paper does not report original code. Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request. EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS S. pombe strains and culture Genotypes of strains used or built in this study can be found in Table S1. The growth and genetics of S. pombe were conducted using established methods. 66 In homothallic S. pombe h 90 strains, isogenic daughter cells from germinating spores switch their mating type undergoing frequent conjugation-sporulating cycles at sporulation conditions. Taking advantage of this characteristic, h 90 strains were used to examine spindle dynamics and chromosome segregation during meiotic divisions in zygotes with different mutant backgrounds. Deletions and strains expressing proteins tagged with GFP, mCherry, mTomato, or mTurquoise2 were generated and confirmed by PCR, following previous protocols. 67 Ark1-GFP tagging was checked by comparing the fluorescent signal with that of the KGY7900 strain (Ark1-GFP h+) from the K. Gould lab. 68 Double or multiple mutants were produced through mating, tetrad dissection, and selection based on drug resistance, auxotrophy markers, or the presence of fluorescent markers under the microscope. In general, yeast cells were grown in YES media. When selecting an auxotrophic mutant, minimal media without the supplement was used. Generally, the growth temperature for S. pombe was 30C. Live-cell imaging was performed on h 90 strains that were grown in YES media at 25C for 18–24 hours. Meiosis was induced by transferring REAGENT or RESOURCE SOURCE IDENTIFIER Chemicals, peptides, and recombinant proteins 1-NM-PP1 Santa Cruz Biotechnology SC-203214 Glycine max lectin Sigma L1395 Experimental models: Organisms/strains A list of S. pombe strains used in this study can be found in Table S1. N/A N/A Oligonucleotides A list of the oligonucleotides used in this study can be found in Table S2. N/A N/A Recombinant DNA pFA6a-GFP(S65T)-kanMX6 J. Ba ¨hler pFA6a-GFP(S65T)-kanMX6 pFA6akanMX6 J. Ba ¨hler pFA6akanMX6 pFA6a-mTurq2-natMX6 M. Sato pFA6a-mTurq2-natMX6 Software and algorithms ImageJ Schneider et al. 2012 65 https://imagej.net/software/fiji/ Graphpad Prism 8.0. Graphpad Software https://www.graphpad.com BioRender BioRender Software https://www.biorender.com Other m-Slide 8 well Ibidi Cat#: 80826 FCS2 chamber Bioptechs Cat#: 060319-2-03 ll OPEN ACCESS iScience 26, 108339, November 17, 2023 15 iScienc e Article cells to sporulation plates (SPA) and incubating them at 28C for 10-12 hours until enough cells had mated. For all imaging assays, cells were cultured in SPA liquid media with supplements. Aurora B kinase inhibition assays were performed by adding 10 mM of the ATP analog 1-NMPP1 to the imaging media. METHOD DETAILS Aurora B kinase inhibition assays To evaluate Aurora B kinase inhibition, we utilized the ark1-as3 analog-sensitive allele, 36 which enables specific and rapid deactivation of Aurora kinase via the ATP analog 1-NM-PP1. To selectively inhibit Aurora during meiosis II, we mounted cells in a m-Slide eight-well chamber (Ibidi, 80826) previously coated with 100 mL of 1 mg/mL soybean lectin (Sigma-Aldrich, L1395) and added 10 mM of 1-NM-PP1 (Santa Cruz Biotechnology, SC-203214) to the imaging media after 10 minutes of time-lapse imaging. We focused our observations on cells that had initiated meiosis II (n between 170 and 376) to ensure that we captured only those cells that had undergone accurate chromosome segregation in meiosis I. To ensure that we had enough MII divisions, we selected those fields through the visualization of microtubules (mCherry-Atb2) where there was at least one cell at metaphase I-anaphase I. This ensured that when adding the inhibitor, this cell was on the verge of beginning meiosis II. Live-cell microscopy For live-cell imaging, cells were mounted in a Bioptechs FCS2 chamber using SPA liquid media (1% glucose, 7.3 mM KH2PO4, vitamins and 45 mg/L adenine, histidine, leucine, lysine, and uracil). To compare fluorescent signals between two or more strains, they were stuck in next, but separate drops of lectin so that the capturing settings and possible bleaching are identical between the control and experimental strains. The images were obtained using a spinning-disk confocal microscope (IX-81, Olympus; CoolSNAP HQ2 camera, Plan Apo-chromat 100x, 1.4 NA objective, Roper Scientific) and Metamorph software, with the temperature maintained at 25C. Time-lapse images were captured at intervals of 2 or 5 minutes, with 20 slices with a Z step of 0.35 mm, every time point over a period of 5-6 hours. For 3-color fluorescence analysis (refer to Figures 2D, 3A, and S4), images were acquired using a Zeiss Observer 7 inverted microscope equipped with Zeiss Plan-Apochromat 63X/1.40 Oil DIC and Alpha Plan-Apochromat 100x/1.46 Oil DIC lenses, coupled with a spinning Disk Confocal Yokogawa CSU-W1 head with excitation lasers and filters from 3i (Intelligent Imaging Innovations). Device control and image capture were performed using SlideBook6 software. Image analysis and measurement of spindle and chromosome dynamics The analysis of image stacks was conducted using ImageJ software. Maximum intensity projections were generated for subsequent analysis. To evaluate chromosome segregation at meiosis II, only cells with accurate chromosome segregation during meiosis I were selected for analysis. The duration of prometaphase and metaphase was determined by measuring spindle polymerization dynamics (mCherry-Atb2). Specifically, we measured the duration of phase I, during which the spindle elongates as a short spindle, and phase II, during which the spindle remains relatively constant in length. Anaphase onset was identified as the beginning of phase III, during which the spindle further elongates and eventually disappears. QUANTIFICATION AND STATISTICAL ANALYSIS Quantification of fluorescent signals The fluorescence intensity was measured by defining a region of interest around the targeted area and subtracting fluorescence within an equivalent background area. To analyze fluorescence intensity dynamics, a region of interest was traced and the average intensity per pixel was calculated while subtracting the background over time. To maintain consistency, the fluorescence images of both the wild-type and mutant samples were acquired simultaneously. Statistical analysis Graphs and statistical analyses were generated using Microsoft Excel and Prism 8.0 (GraphPad Software). The mean and standard deviation (SD) are represented in the graphs, and the number of cells scored from at least three independent experiments is indicated as n. Two-group comparisons were performed using unpaired Student’s t-test, while multiple-group comparisons were analyzed using non-parametric Kruskal-Wallis one-way ANOVA and ordinary one-way ANOVA. Chi-square tests were utilized to compare the frequencies of phenotypes between conditions in Figures 4C, 5B, and 5C. Statistical significance was considered for p-values lower than 0.05 (*). Further details of statistical analysis are given in the figure legends. ll OPEN ACCESS 16 iScience 26, 108339, November 17, 2023 iScienc e Article