CENP-V is required for proper chromosome segregation through interaction with spindle microtubules in mouse oocytes

ARTICLE CENP-V is required for proper chromosome segregation through interaction with spindle microtubules in mouse oocytes Dalileh Nabi1,5, Hauke Drechsler 2, Johannes Pschirer1, Franz Korn1, Nadine Schuler1, Stefan Diez 2,3,4, Rolf Jessberger 1,7✉& Mariola Chacón 1,6,7✉ Proper chromosome segregation is essential to avoid aneuploidy, yet this process fails with increasing age in mammalian oocytes. Here we report a role for the scarcely described protein CENP-V in oocyte spindle formation and chromosome segregation. We show that depending on the oocyte maturation state, CENP-V localizes to centromeres, to microtubule organizing centers, and to spindle microtubules. We find that Cenp-V−/−oocytes feature severe deficiencies, including metaphase I arrest, strongly reduced polar body extrusion, increased numbers of mis-aligned chromosomes and aneuploidy, multipolar spindles, unfocused spindle poles and loss of kinetochore spindle fibres. We also show that CENP-V protein binds, diffuses along, and bundles microtubules in vitro. The spindle assembly checkpoint arrests about half of metaphase I Cenp-V−/−oocytes from young adults only. This finding suggests checkpoint weakening in ageing oocytes, which mature despite carrying mis-aligned chromosomes. Thus, CENP-V is a microtubule bundling protein crucial to faithful oocyte meiosis, and Cenp-V−/−oocytes reveal age-dependent weakening of the spindle assembly checkpoint. https://doi.org/10.1038/s41467-021-26826-3 OPEN 1Institute of Physiological Chemistry, Medical Faculty Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany. 2B CUBE-Center for Molecular Bioengineering, Technische Universität Dresden, Dresden, Germany. 3Cluster of Excellence Physics of Life, Technische Universität Dresden, Dresden, Germany. 4Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany. 5 Present address: Department of Neuropediatrics Charité-Universitätsmedizin Berlin, Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin, Germany. 6 Present address: CABIMER, Centro Andaluz de Biología Molecular & Medicina Regenerativa, Sevilla, Spain. 7 These authors contributed equally: Rolf Jessberger, Mariola Chacón. ✉email: [email protected];[email protected] NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications 1 1234567890():,; Mammalian oogenesis is burdened with comparably high rates of failures that may result in chromosome missegregation and aneuploidy (for reviews see1–9). These errors happen during oocyte meiosis, which includes a very prolonged period of arrest, and increase with advancing age. Several mechanisms were proposed to contribute to this problem. One prominent mechanism is a gradual deterioration of cohesin (for reviews see6,7,10–12), which ought to maintain sister chromatid cohesion throughout oocyte lifespan. Cohesin is loaded onto oocyte chromosomes at the entry into meiosis with no or very little replenishment later. The increasing loss of sister chromatid cohesion with consequently loss of chiasmata in meiosis I and weakening of sister centromere cohesion up to anaphase II is considered a major factor in age-dependent oocyte chromosome mis-segregation. Several other mechanisms likely also contributes to oocyte ageing and problems associated with it (for reviews see7,13–19). Telomere damage also increases with progressing age and may cause chromosome end fusions and thus aberrant chromosome segregation. Likewise, erroneous attachment of microtubules to the kinetochores may lead to mis-segregation if not corrected during a spindle assembly checkpoint (SAC) mediated arrest. It appears as if the control mechanisms provided by the SAC also weaken with advancing age although this has not yet been clearly demonstrated. Thus, for persistent maintenance of oocyte chromosome integrity and proper segregation a well-balanced system of monitoring, error correction, and strict control pathways are required. This, however, is not fully sufficient as it fails in the long term. In this communication we describe an additional factor that plays a central role in the control of oocyte meiosis, CENP-V. Human CENP-V is a 275 amino acid protein described so far in only very few publications, initially by William Earnshaw in 2008. This group identified it as a kinetochore component in proteome analysis of human mitotic chromosomes20. The 29.7 kDa protein CENP-V (pI 9.8) is highly conserved among vertebrates. In HeLa cells, it localizes to the mitotic kinetochores from prometaphase to metaphase and moves to the spindle mid-zone in anaphase. Cysteines within CENP-V may coordinate Zn++ which possibly mediates DNA binding. When overexpressed in HeLa cells, CENP- V causes hypercondensation of pericentromeric heterochromatin. Knock-down of CENP-V in HeLa cells leads to decompaction of chromosomes, decondensation of the heterochromatin. Metaphase alignment of chromosomes in CENP-V depleted HeLa cells is impaired and lagging chromosomes appear in anaphase20.Outof the context of mitosis, CENP-V appears to be involved in the regulation of microtubule structures as it promotes cell migration by targeting Src family kinases to the microtubule network in the leading edge of migrating cells21. CENP-V also supports tubulin acetylation at the basal bodies and axonemes of primary cilia22 –a posttranslational tubulin modification that is thought to enhance the resilience of microtubules to mechanical stress23. In this work, we investigated the role of CENP-V in oocytes as their particularly delicate meiotic and spindle features may require functional CENP-V during maturation. We describe CENP-V as an important regulator of oocyte microtubulekinetochore associations and chromosome segregation, whose absence causes age-dependent failure of the SAC and aneuploidy. Results CENP-V changes its localization during meiosis in mouse oocytes. CENP-V has been described as a component of the kinetochore from prometaphase to metaphase during mitosis20. To test whether this localisation is conserved in meiotic prophase I we stained mouse embryonic oocyte spreads with an anti- CENP-V antibody (Fig. 1a) raised against the full-length mouse protein and validated on CENP-V deficient cells (see next section). Unless otherwise specified, all experiments described below were done using young adults, i.e. 6 weeks to 3 months old mice. In wild-type (wt) mice we observed CENP-V to appear in pachytene yielding a diffuse signal. At the end of meiotic prophase in diakinesis CENP-V was found in the vicinity of centromeres and the surrounding heterochromatin as observed by co-staining with the centromere marker ACA (Fig. 1a). This pattern is similar to that seen in spermatocyte prophase I, where CENP-V localized to the centromere region and external to the centromere protein CENP-A from the pachytene stage onwards, similar to what has been described for mitotic HeLa cells20 (Sup. Figure 1). However, CENP-V was excluded from the spermatocyte X/Y chromosomes, the sex body chromatin. The distribution of CENP-V during meiotic resumption was analyzed in mouse oocytes at different stages of maturation. To allow fixed and live-cell imaging, oocytes expressing CENP-V-GFP and H2B-RFP after injection of Cenp-V-Gfp and H2B-Rfp mRNA at the arrested germinal vesicle (GV) stage were used (Fig. 1b). At the stage of germinal vesicle break down (GVBD) the CENP-V-GFP signal covered the entire chromosomes. Since Honda et al. suggested an association of CENP-V with microtubules21, some GVBD oocytes were fixed and probed with an anti γ-tubulin antibody as a marker for microtubule organising centres (MTOCs). Metaphase I and II spindles were stained with SiR Tubulin (Fig. 1b and supplementary mov. S1). During meiotic resumption we observed three different pools of CENP-V: (i) at the chromosomes, (ii) at the MTOCs (Fig. 1b, upper row), and (iii) at the meiotic spindles. At metaphase I CENP-V is mainly enriched at the spindle and this was maintained throughout metaphase II (Fig. 1b, middle and lower rows). To get an unbiased quantification of the localization of CENP-V to the spindle, we measured the Pearson’scoefficient, using an imaging algorithm to measure the colocalization along z-stacks between CENP-V-GFP and SiR Tubulin. Notably, in the spindle pole Pearson’scoefficient was close to 1, indicating an almost complete colocalization. On the contrary, colocalization of CENP-V and SiR Tubulin decreased towards the spindle center, suggesting that CENP-V does not show a homogenous distribution in the spindle but rather accumulates at the pole (Fig. 1c). The same pattern was observed by staining fixed oocytes with an anti-CENP-V antibody (Fig. 2a, Sup. Figure 2). This suggests that during female meiosis CENP-V expands its localization from around the centromere region in prophase to along metaphase I chromosomes, to the MTOCs, and to microtubules during metaphase I and PB extrusion (PBE). Deletion of Cenp-V arrests mouse oocytes at metaphase I. Would the lack of CENP-V affect oocytes maturation, meiotic progression? CENP-V homozygous null mice (Cenp-V−/− )were generated by Cre recombinase-mediated deletion of the floxed exon 3 of Cenp-V,confirmed by DNA sequencing. We first assessed the loss of CENP-V in Cenp-V−/−mice by immunofluorescence (IF) and western blot (WB) using commercial and in-house generated anti- CENP-V antibodies. The absence of CENP-V signals from chromosome spreads and from WB of Cenp-V−/−oocytes or ovary confirmed loss of the protein, and these experiments also validated the antibodies (Fig. 2a, Sup. Figure 3a, b). Cenp-V−/−mice were viable and females but not males were subfertile. Subfertility was observed when Cenp-V−/−females were bred with wt, heterozygous, or homozygous CENP-V deficient males (Sup. Figure 4a). Because we observed subfertility only in Cenp-V−/−females we focused on theroleofCENP-Vinoocytemeiosis. Cenp-V+/+and Cenp-V−/−oocytes were collected, maintained, and synchronized in prophase with intact germinal vesicles (GVs) by incubation in milrinone-containing medium. Two hours after removal of milrinone Cenp-V+/+and Cenp-V−/−oocytes ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 2NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications underwent germinal vesicle breakdown (GVBD), i.e. meiotic resumption. In average, Cenp-V−/−oocytes required significantly longer to undergo germinal vesicle breakdown (GVBD) (Fig. 2b; Cenp-V+/+: 83.7% ± 16.18 at 2.5 h; Cenp-V−/−: 54% ± 15.33 at 2.5 h). Twenty hours after the GV stage, 74.3% ± 6.24 of Cenp-V +/+but only 41.6% ± 7.95 of Cenp-V−/−oocytes have extruded the first polar body (PB). This process was also significantly slower in Cenp-V−/−oocytes than in Cenp-V+/+oocytes (Fig. 2c). To test whether this reduction in PB extrusion was due to a specific role of CENP-V during meiosis and not due to an unspecific effect brought about by the absence of CENP-V from all tissues in the Cenp-V−/−mouse strain we performed similar experiments using NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 ARTICLE NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications 3 aCenp-Vfl/flconditional deletion strain crossed with GDF9-iCre or SPO11-Cre. The oocyte-specific GDF9-iCre expresses Cre shortly after birth, i.e. in oocytes that have just reached dictyate arrest24. Spo11-Cre is expressed shortly after entry into meiosis25. The decrease in PBE in Cenp-Vfl/fl,Gdf9-iCre +and Cenp-Vfl/fl, Spo11-Cre +mice was very similar to that seen in Cenp-V−/− oocytes confirming the specific role of CENP-V during oocyte meiosis (Fig. 2d, e). Similar results were also obtained using Vasa- Cre26 which is expressed in oocytes at around early zygotene (not shown). Together this data suggests that CENP-V is not required in prophase I but is required after dictyate arrest. To further check possible deleterious effects of CENP-V deficiency on oogenesis we quantified the number of GV cells per female and performed histological analysis of ovaries from Cenp-V+/+and Cenp-V−/−mice. No differences were observed in GV cell numbers, and follicular development appeared to be normal as evaluated by hematoxylin/eosin staining (Fig. 2f, left two columns, Sup. Figure 3b). To assess a possible role of CENP- V during ageing we quantified the GV cell numbers and performed life imaging experiments in Cenp-V−/−females older than 12 months. We found a reduction of the GV cell number due to ageing but not due to the absence of CENP-V (Fig. 2f). Surprisingly, the reduction in PBE found in young Cenp-V−/− oocytes vanished in oocytes from old females (Fig. 2g). With the control steady at app. 78 %, 44,7 ± 14,62 % of the young the Cenp- V−/−oocytes underwent PBE compared to 71,7 ± 6,2 of old Cenp- V−/−oocytes. Assuming that the errors caused by the absence of CENP-V remain the same in young and old oocytes, this hints at the weakening of the key control mechanism, the spindle assembly checkpoint (SAC). Together, these results suggest that CENP-V plays an important role for faithful oocyte maturation and chromosome segregation. Indeed, we observed an increased frequency of MII oocytes harboring abnormal numbers of pairs of sister chromatids, i.e. aneuploidy (Fig. 2h, i). To assess aneuploidy in meiosis II oocytes we counted chromosomes, i.e. pairs of sister chromatids, on MII metaphase chromosome spreads. Missegregation leads to loss of chromosomes or increased numbers and thus any deviation from 20 pairs of sister chromatids is considered an aneuploidy. This analysis revealed almost 4-fold increased rates of aneuploidy in absence of CENP-V. Sister chromatid cohesin itself was not lost since we did not detect single, unpaired chromatids. The limited number of oocytes present in aged mice precluded an extensive analysis of aneuploidy in those, but when mice of 9, 16, and 21 months of age were analyzed (together 15 oocytes for littermate controls and 10 oocytes for Cenp-V−/−oocytes), we observed 46 % of cells for the control and 60 % of the Cenp-V−/−oocytes showing aneuploidy. Generally, aneuploidy increases with age, for example through cohesion decay. While due to low numbers the increase in aged Cenp-V−/−oocytes is only a trend, the data indicates a higher frequency of aneuploidy in aged CENP-V deficient oocytes which, however, contrary to young oocytes still proceed to the PB stage, consistent with the above data. Chromosome alignment is erroneous and delayed in Cenp-V−/− oocytes. A failure to perform PBE could be triggered by chromosome mis-alignment which can lead to SAC activation and thus to arrest of oocytes before PBE27. Since CENP-V accumulated at the centromeric region and at the meiotic spindle, both key structures for proper chromosome alignment and kinetochore capture28,we decided to monitor chromosome alignment in the absence of CENP-V. Oocytes proficient and deficient for CENP-V were analysed by time-lapse fluorescence imaging of microtubules (β-TUBULIN- GFP) and chromosomes (H2B-RFP) after injection of the respective mRNAs (Fig. 3and supplementary mov. S2 and S3). The H2B-RFP signal was followed until the spindle reached the cortex prior to PBE (Fig. 3a). Ten hours after GV, 75% of the control cells showed full chromosome alignment at the metaphase plate compared to 40% of Cenp-V−/−cells (Fig. 3b). These values were very similar to the respective frequencies of PBE (Fig. 3c) and to the values of PBE in non-injected cells (Figs. 2,3c), suggesting that chromosome misalignment contributes to the decrease in PBE and that injection per se had no effect. To test whether the lack of CENP-V and no other cause reduced PBE and chromosome mis-alignment we also injected Cenp-V-Gfp mRNA into Cenp-V−/−cells. Representative frames of the time course are shown in Fig. 3a, which includes an example of a Cenp-V−/−oocyte overexpressing β-TUBULIN-GFP/ H2B-RFP that stayed arrested at MI and a Cenp-V−/− oocyte overexpressing CENP-V-GFP that successfully extruded its PB (supplementary mov. S3 compare to supplementary mov S4). Cenp-V−/−oocytes expressing CENP-V-GFP restored both, chromosome alignment and PBE to control values (Fig. 3b, c), confirming that the lack of CENP-V causes the chromosome mis-alignment that leads to the reduction of PBE. For clarity of the results we do not represent the percentage of PBE and chromosome alignment of the Cenp-V+/+oocytes expressing CENP-V-GFP as they behave like control oocytes (Fig. 1b, supplementary mov. S1). Inhibition of Mps1 restores progression to metaphase II. The reduction in PBE in the absence of CENP-V correlates with chromosome misalignment, which is the likely cause of oocyte arrest at MI. Interestingly, we observed a decrease in PBE only in oocytes from young females, but not in oocytes from females 12 months or older. We speculated that the SAC increasingly fails to prevent anaphase in ageing oocytes carrying chromosome and/ or spindle aberrations. An age-dependent decline in SAC stringency has been previously suggested2,5,29,30. We hence tested if the SAC is responsible for the arrest of young Cenp-V−/−oocytes before PBE. Thus, we inhibited the SAC kinase Mps1 by adding reversine to the culture medium of Cenp-V−/−and Cenp-V+/+ Fig. 1 CENP-V localises to the centromere regions during meiotic prophase and to MTOCs, chromosomes, and spindle during oocyte maturation. aCENP-V distribution during meiotic prophase in Cenp-V+/+embryonic chromosome spreads by immunofluorescent staining. Chromosome axes (AEs/ LEs) were stained with anti-SYCP3 (blue); CENP-V was probed with anti-CENP-V (green); centromeres and surrounded regions by anti-ACA. CENP-V signal started to appear at pachytene and concentrates at the centromere regions (yellow arrows) at diakinesis; n=19 cells from 2 independent experiments. Scale bar =5μm. bCENP-V distribution during mouse oocyte maturation in whole cells. Live imaging performed with Cenp-V+/+oocytes injected by mRNA of H2B-Rfp (blue) and Cenp-V-Gfp (green). Maximum projections from representative oocytes at germinal vesicle break down (GVBD), metaphase I, and metaphase II stages are shown. Scale bar =10 μm. Some GVBD cells were fixed and immunostained with anti γ-Tubulin as an MTOC marker. Spindle at metaphase I and II were visualised by SiR Tubulin dye. Polar body is indicated as PB. Note that at the early stages of maturation CENP-V distributes along the entire chromosomes and colocalizes with MTOCs (yellow arrow heads). At metaphase I and II CENP-V appears mainly in the spindle and enriches at the spindle poles. See also supplementary movie 1; n=12 cells from 3 independent experiments. cQuantification of colocalization of CENP- V with the spindle pole by determining the respective Pearson’s correlation coefficient; n=10 cells from 3 independent experiments; the color code highlights the frequency of dots present in a certain region of the scatter plot (from blue to yellow and white with increasing frequencies). ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 4NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications oocytes at the GV stage31 (Fig. 4). PBE in reversine-treated control oocytes was accelerated and PBE in Cenp-V−/−cells was fully restored to control values. This suggests that the SAC arrests Cenp-V−/−oocytes of young females at metaphase I. Analysis of chromosome mis-alignment in reversine-treated oocytes from young adults revealed a significant increase. This indicates that upon SAC inhibition many more oocytes carrying misaligned chromosomes proceed to PBE and MII (Fig. 4b). Notably, unlike young Cenp-V−/−oocytes, aged Cenp-V−/−oocytes proceeded to PBE like controls (Fig. 2g). In the aged Cenp-V−/−oocytes a strongly increased frequency of cells carrying misaligned chromosomes was observed (Fig. 4b). The absence of MI arrest in aged oocytes despite this presence of mis-aligned chromosomes confirms that the SAC is not as active anymore in oocytes from NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 ARTICLE NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications 5 old females. Thus, CENP-V deficiency reveals age-dependent weakening of the SAC in oocytes. CENP-V maintains spindle integrity in vivo and binds, diffuses along, and bundles microtubules in vitro. It has been suggested that the absence of CENP-V disturbs microtubule dynamics21. A failure in microtubule dynamics could lead to problems in chromosome capture and alignment. Thus, we asked whether the lack of CENP-V affects oocyte spindle integrity. Cenp-V−/− oocytes were fixed and stained for microtubules at different time points; at 7 h after GV to study the recently formed MI spindle Fig. 2 Cenp-V−/−oocytes are impaired in MI progression. a Expression of CENP-V protein in Cenp-V+/+and Cenp-V−/−oocytes. CENP-V is probed by anti CENP-V antibody (green), DNA is stained by DAPI (blue).Scale bar =5μm. bCumulative percentage of GVBD cells over time in Cenp-V+/+and Cenp-V−/−oocytes. cCumulative percentage of first polar body (PB) extrusion over time in Cenp-V+/+and Cenp-V−/−oocytes. dCumulative percentage of first polar body extrusion over time in Cenp-Vfl/flGdf9cre- and Cenp-Vfl/flGdf9cre +oocytes. eCumulative percentage of first polar body extrusion over time in Cenp-Vfl/flSpo11cre- and Cenp-Vfl/flSpo11cre +oocytes. fNumber of oocytes in young (< 6 months) and old (>12 months) Cenp-V+/+and Cenp-V−/− females as an average number of GV cells per female, and corresponding SD. gPercentage of first polar body (PB) extrusion after 20 h culture of oocytes from young (< 6 months) and old (>12 months) Cenp-V+/+and Cenp-V−/−females, shown as an average percentage per female. hPercentage of MII oocytes showing aneuploidy. iExamples of MII oocyte chromosome spreads of control and Cenp-V−/−oocytes. The numbers of chromosomes are indicated; size bar =50 μm. Statistic: (b–h); data were expressed as mean ± SD and statistical differences were tested by a paired two tailed T-test (***p< 0.001, **p < 0.01, *p< 0.05). The exact pvalues are available in the source data file.T-test results between Cenp-V−/−(c), Cenp-Vfl/flGdf9cre (d) Cenp-Vfl/flSpo11cre (e) lines were not significant, i.e. they all showed the same phenotype. (b); statistical differences refer to the average of the data points between 1 and 5 h after GV. (c,d,e,f); statistical differences refer to the average of all data points. (b,c,f,g)n=303 oocytes from 6 independent experiments; (d,e)n=134 oocytes from 3 independent experiments and (i)n=20 cells from 3 independent experiments. Fig. 3 Chromosome alignment is reduced and delayed in Cenp-V−/−oocytes. Polar body extrusion and chromosome alignment return to control values upon Cenp-V-Gfp mRNA injection. (a) Selected time frames from a time-lapse experiment showing progression through meiotic prophase until metaphase II in Cenp-V+/+and Cenp-V−/−oocytes expressing β-TUBULIN-GFP (magenta), H2B-RFP (cyan, large image in grey) and/or CENP-V-GFP (green) after mRNA injection. At 09:30 h in metaphase I, the spindle is assembled and chromosomes are aligned in the equatorial plate (white dash line) in Cenp-V+/+ compare to Cenp-V−/−oocyte. Note that Cenp-V−/−oocytes show normal chromosome alignment and extrude the PB only when expressing CENP-V-GFP. Oocytes were imaged after 4 h release from GV every 30 m. Maximum intensity projection is shown. See also supplementary Movies S2,S3,S4). Scale bar =10 µm. bPercentage of cells that show chromosome alignment overtime of the experiment shown in (a). in Cenp-V+/+and Cenp-V−/−oocytes. Quantification finishes before the first polar body extrusion. Note that at 10 h, 75 % of control oocytes have reached full chromosome alignment compared to the 35% of Cenp-V−/−oocytes, and only the injection of Cenp-V-Gfp (green square line) mRNA restores the values to the control. cCumulative percentage of first polar body extrusion overtime of the experiment shown in (a). Note that Cenp-V−/−oocytes properly extrude the PB only when expressing CENP-V- GFP (green square line) and not when expressing β-TUBULIN-GFP/H2B-RFP. See also supplementary movie S1. Statistic: (b,c) data were expressed as mean ± SD and statistical differences were tested by a two way Anova test (***p< 0.001, **p<0.01,*p< 0.05). The exact pvalues are available in the source data file. Statistical differences refer to the average of all data points. Anova test between Cenp-V+/+and Cenp-V+/+injected was not significant. Anova test between Cenp-V−/−and Cenp-V−/−β-Tubulin-Gfp/H2B-Rfp injected was not significant. n=249 cells from 5 different experiments. ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 6NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications and between 10-16 h to study the late MI spindle after arrest and after the spindle has reached the cortex. At late metaphase I, between 10-16 h of culture, 46% of Cenp-V−/−oocytes but virtually no control oocytes showed aberrant spindles suggesting that the lack of CENP-V indeed disrupts spindle integrity (examples are shown in Fig. 4c and Sup. Figure 5a). In those cells that still show a defined spindle shape, the spindle pole width was measured. We observed that lack of CENP-V leads to loss of the canonical barrel-shaped spindles as the poles are less defined, diffusely widened, in Cenp-V−/−oocytes (Sup. Figure 5b). As previous work suggests that CENP-V can bind microtubules on its own22, we hypothesised that CENP-V-microtubule interactions might directly stabilise the acentrosomal meiotic spindle. To test this hypothesis, we expressed recombinant Fig. 4 Inhibition of MPS I or ageing allow Cenp-V−/−oocytes to proceed to anaphase I. a Cumulative percentage of first polar body extrusion (PBE) over time in Cenp-V+/+and Cenp-V−/−oocytes treated with reversine. Note that the addition of reversine restores PBE to control values. The addition of reversine accelerates and increases PBE in both Cenp-V+/+and Cenp-V−/−oocytes. bPercentage of cells that harbour misaligned chromosomes at MII in young oocytes treated with reversine and in old cells. Reversine increases the misalignment also in control cells compared to non-treated cells. 4a: statistical differences refer to the average of all data points. Statistic: (a,b) data were expressed as mean ± SD and statistical differences were tested by a two way Anova test (***p< 0.001, **p< 0.01, *p< 0.05). The exact pvalues are available in the source data file. Statistical differences refer to the average of all data points. Cenp-V+/+and Cenp-V−/−oocytes treated with reversine show not a significant difference. (a)n=112 oocytes from 3 independent experiments and (b)n=10 cells per condition from 4 independent experiments. cExample of Cenp-V+/+and Cenp-V−/−oocytes treated with reversine in young females and one example of Cenp-V−/−oocyte >12 months for comparison. Note that the absence of CENP-V in young females arrests half of the cells in MI and only after the addition of reversine all oocytes proceed to anaphase and extrude the PB. Microtubules are stained by anti-ß-tubulin (magenta), centromeres are stained with anti ACA (green) and DNA is stained by Hoechst 33258 (cyan). Enlarged images in grey shown the single channel of the chromosomes and spindle. Scale bar =10 µm. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 ARTICLE NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications 7 His 6 -CENP-V-eGFP in insect cells and purified the fusion protein by immobilised metal affinity and size exclusion chromatography (Supp. Figure 6). Biophysical properties of CENP-V were studied in vitro. We first quantified the microtubule-binding capacity of CENP-V-eGFP, by titrating increasing amounts of protein onto surface-bound, taxol-stabilised ATTO 647N-labelled microtubules in 80 mM PIPES, pH 6.8, 1 mM MgCl 2 , 1 mM EGTA (BRB80) and visualising the interaction by total internal reflection fluorescence (TIRF) microscopy (Fig. 5a, Sup. Figure 7A, B and Sup. Mov. S5). The binding of CENP-V-eGFP to the microtubule lattice was quantified by calculating the average GFP-intensity per micrometer for at least 100 microtubules per condition for three independent technical replicates. The resulting binding saturation curve was best fitted with a cooperative binding (Hill) model, ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 8NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications suggesting that binding of CENP-V to the microtubule lattice is cooperative (Fig. 5b). Binding of CENP-V-eGFP occurred with a K D of 53.2 ± 4.5 nM Fig. 5b), while single microtubule-associated particles were already evident at 1 nM (Fig. 5a–c). To characterise the lattice behaviour of single CENP-V-eGFP particles, we next recorded three-minute time-lapse movies at 100 milliseconds time resolution, visualising the movement of 1 nM CENP-V-eGFP in BRB80 and at 25 °C on surface-bound, taxol-stabilised ATTO 647N-labelled microtubules by TIRF microscopy. In general, CENP-V-eGFP particles stayed attached to the microtubule lattice, with only a few associations or dissociation events to be observed within up to fifteen minutes (Fig. 5c, supplementary Mov. S5, supplementary Fig. 8A, B). Lattice-bound CENP-V particles can be forced to dissociate from the microtubule by an excess of free tubulin - but not by BSA - indicating that the CENP-V-tubulin interaction is specific (Sup.Fig.8C).Particlesshowedcomplexmotilityonthe microtubule lattice that can be described either by i) continuous, rather “fast”diffusion (Fig. 5c I, II), ii) very slow diffusion (Fig. 5c III, IV) with interdispersed short but fast diffusing episodes (Fig. 5c, white arrowheads)oriii)amixturethereof, where CENP-V particles switch between modes. Furthermore, motile particles eventually became stationary for extended periods on the microtubule lattice (Fig. 5cIV,blue arrowhead to end of kymograph, switch motile to stationary and Fig. 5cV, beginning of kymograph to blue arrowhead, switch stationary to motile) or remained associated with the microtubule end (Sup Fig. 8A). At microtubule intersection, CENP-V-eGFP particles occasionally switched microtubule tracks and explored both microtubules (Sup Fig. 8D). A mean squared displacement analysis including all CENP-V-eGFP particles that showed some episodes of movement (excluding stationary episodes at microtubule ends or microtubule intersections) during the duration of the time-lapse movie, produced a rather low apparent diffusion coefficient of D=0.0192 ± 0.0006 µm s−1(n=136) (Fig. 5d). As expected from the complex CENP-V motility on the microtubule lattice (Fig. 5C), however, the diffusion coefficient of single CENP-V-eGFP particles are distributed over at least one order of magnitude with maximum values of D=0.15 µm s−1(supplementary Fig. 8E). Interestingly, motile CENP-V-eGFP particles also showed a rather broad intensity distribution with four local maxima at 131 ± 4, 259 ± 4, 432 ± 11 and 760 ± 79 arbitrary units (a.u., n=249) (Fig. 5e). A comparison with the average intensity loss of 118 ± 8 a.u. (n=50 bleaching steps from 28 particles) per GFP during bleaching events of these particles suggests that - in line with the observed cooperative binding behaviour - microtubule bound CENP-V can form multimers on the microtubule lattice (Fig. 5f, supplementary Fig. 9). Hence, we hypothesised that the broad distribution of diffusion coefficients might be a direct cause of the inhomogeneous particle size distribution. In this case, we would expect larger multimeric particles to interact stronger with the microtubule lattice causing slower diffusion rates compared to smaller particles. Subsampling of the data by grouping the particles in 125 arbitrary unit bins, however, showed that the distribution of diffusion coefficients does not change significantly with increasing particle size (Supplementary Fig. 8F). This suggests that the broad distribution of diffusion coefficients observed for single CENP-V particles is caused by peptide properties other than multimerisation. The cooperative binding behaviour of CENP-V and its capacity to multimerise (on the microtubule lattice) is reminiscent of other microtubule-associated proteins like PRC1/Ase1 that structurally reinforce the mitotic spindle by bundling interpolar microtubules in antiparallel overlaps of the spindle midzone32–36. We hence asked, whether also CENP-V has the capacity to drive microtubule bundle formation. To test this, we incubated taxol-stabilized ATTO 647N-labelled microtubules with 50 nM CENP-V-eGFP or non-GFP-tagged CENP-V for 15 m at room temperature, immobilized them on the surface of glass flow-cambers, and flushed out non-bound CENP-V prior to imaging. Indeed, both 50 nM CENP-V and CENP-V-eGFP induced the formation of large microtubule bundles that were absent from the buffer-only control (Fig. 5g). In summary, we conclude that CENP-V is a bona fidae microtubule-associated protein that binds cooperatively to the microtubule. It remains attached to the microtubule for a prolonged period exploring it by diffusion and it forms multimers that are able to cross-link and bundle microtubules. It therefore shows all characteristics of a structural MAP like PRC1/Ase1. This could explain the loss of spindle integrity we observed in CENP-V deficient oocytes in vivo. Thus we measured microtubule bundling in the meiotic spindle (Sup. Fig. 10). To remove unstable microtubules, MI ocytes were cold-treated before fixation. Microtubules were visualised by anti-Tubulin-GFP antibody. The spindle was divided in five equal regions and the thickness of the most and least prominent microtubule in each region was measured and defined as a maximum and minimum level of bundling37. We observed a reduction in both, maximal and minimal microtubule thickness, i.e. bundling, in Cenp-V−/− oocytes (Sup. Fig. 10a). The maximum level of bundling in Cenp- V−/−oocyte spindles does not even reach the minimum grade of bundling in controls. The overall level of spindle fluorescence was also reduced in Cenp-V−/−cells compared to the control oocytes Fig. 5 Multimers of CENP-V bind with high affinity to, diffuse along and bundle microtubules. a Binding of recombinant CENP-V-eGFP (top panel row)to taxol-stabilised ATTO 647N-labelled microtubules (bottom panel row) at the indicated CENP-V concentrations. Please refer to Sup. Figure 7A for pictures of the full concentration range and supplementary. Mov. S5. Bar equals 10 µm. bSaturation binding curve of CENP-V-eGFP showing the concentrationdependent fluorescence intensity of microtubule-associated CENP-V-eGFP in arbitrary units (arb.units) per micrometer. The dissociation constant (K d ) and the Hill-coefficient (h) are given. Error bars indicate the standard error of the mean of three independent experiments. Blue line is a Hill-, dashed grey line a Michaelis-Menten-model fit. (c) Representative kymographs (I-V) showing the behaviour of 1 nM CENP-V-eGFP on the lattice of taxol-stabilised ATTO 647N-labelled microtubules in BRB80 at 25 °C. Blue arrowheads indicate transitions between static and motile periods. White arrowheads indicate fast shift events during slow diffusion episodes. Scheme to the right indicates the extend of the respective microtubule lattice (red) and the microtubule ends (asterix). (d) Mean square displacement (MSD) analysis of motile CENP-V-eGFP particles. Data is fitted with a linear regression (black line); blue error bars are the standard error of the mean. (e) Histogram showing the initial fluorescence intensity of motile CENP-V-eGFP particles –i.e., the mean intensity during the first (in average75) movie frames. Blue lines indicate Gaussian fits to the multiple peaks of the histogram adding up to the cumulative Gaussian fit shown as agreen line.Dashed lines indicate the local maxima of the respective fits. The respective peak intensity values in arbitrary units (a.u.) ± standard error of the mean are given above. (f) Plot showing the intensity loss during single bleaching events (n=50) of motile CENP-V-eGFP particles (n=25). Solid and dashed lines indicate median and mean (98.9 a.u. and 118.3 a.u. ± 8.2 s.e.m. respectively). (g) Surface immobilised axol-stabilised ATTO 647N-labelled microtubules after pre-incubation with either buffer, 50 nM CENP-V or CENP-V-eGFP (top panel row). The corresponding eGFP-channels are shown in the bottom panel row. Scale bar =20 µm. Representative images of three independent experiments are shown. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 ARTICLE NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications 9 30. Touati, S. A. et al. Mouse oocytes depend on BubR1 for proper chromosome segregation but not for prophase I arrest. Nat. Commun. 6, 6946 (2015). 31. Santaguida, S., Tighe, A., D’Alise, A. M., Taylor, S. S. & Musacchio, A. Dissecting the role of MPS1 in chromosome biorientation and the spindle checkpoint through the small molecule inhibitor reversine. J. Cell Biol. 190, 73–87 (2010). 32. Mollinari, C. et al. PRC1 is a microtubule binding and bundling protein essential to maintain the mitotic spindle midzone. J. Cell Biol. 157, 1175–1186 (2002). 33. Kapitein, L. C. et al. Microtubule-driven multimerization recruits ase1p onto overlapping microtubules. Curr. Biol. 18, 1713–1717 (2008). 34. Bieling, P., Telley, I. A. & Surrey, T. A minimal midzone protein module controls formation and length of antiparallel microtubule overlaps. Cell 142, 420–432 (2010). 35. Subramanian, R. et al. Insights into antiparallel microtubule crosslinking by PRC1, a conserved nonmotor microtubule binding protein. Cell 142, 433–443 (2010). 36. Lansky, Z. et al. Diffusible crosslinkers generate directed forces in microtubule networks. Cell 160, 1159–1168 (2015). 37. Mullen, T. J. & Wignall, S. M. Interplay between microtubule bundling and sorting factors ensures acentriolar spindle stability during C. elegans oocyte meiosis. PLoS Genet 13, e1006986 (2017). 38. Greaney, J., Wei, Z. & Homer, H. Immunofluorescence staining of K-Fibers in mouse oocytes using cold fixation. Methods Mol. Biol. 77-87, 2018 (1818). 39. Kouznetsova, A., Kitajima, T. S., Brismar, H. & Hoog, C. Post-metaphase correction of aberrant kinetochore-microtubule attachments in mammalian eggs. EMBO Rep. 20, e47905 (2019). 40. Schuh, M. & Ellenberg, J. Self-organization of MTOCs replaces centrosome function during acentrosomal spindle assembly in live mouse oocytes. Cell 130, 484–498 (2007). 41. Clift, D. & Schuh, M. A three-step MTOC fragmentation mechanism facilitates bipolar spindle assembly in mouse oocytes. Nat. Commun. 6, 7217 (2015). 42. Polak, B., Risteski, P., Lesjak, S. & Tolic, I. M. PRC1-labeled microtubule bundles and kinetochore pairs show one-to-one association in metaphase. EMBO Rep. 18, 217–230 (2017). 43. Kajtez, J. et al. Overlap microtubules link sister k-fibres and balance the forces on bi-oriented kinetochores. Nat. Commun. 7, 10298 (2016). 44. Tolic, I. M. Mitotic spindle: kinetochore fibers hold on tight to interpolar bundles. Eur. Biophys. J. 47, 191–203 (2018). 45. Ruhnow, F., Zwicker, D. & Diez, S. Tracking single particles and elongated filaments with nanometer precision. Biophys. J. 100, 2820–2828 (2011). 46. Vallot, A. et al. Tension-induced error correction and not kinetochore attachment status activates the SAC in an Aurora-B/C-dependent manner in oocytes. Curr. Biol. 28, 130–139 e3 (2018). 47. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012). Acknowledgements This work was supported by a grant from the Deutsche Forschungsgemeinschaft to RJ (DFG, JE150/25-1), by a grant from the European Union to RJ (H2020, GermAge), H.D. is supported by the Deutsche Forschungsgemeinschaft (DFG, project number 417890911); M.C. was also supported by the VI Plan Propio Program of Universidad de Sevilla. We thank the following facilities for their excellent support: Core Facility Cellular Imaging, Medical Faculty Carl Gustav Carus, Technische Universität Dresden; and the Protein Expression, Purification and Expression facility, the Light Microscopy facility, the Transgenic Core facility of the Max Planck Institute of Molecular Biology and Genetics, Dresden. Author contributions DN, HD, NS, SD, RJ, MC contributed to writing the manuscript; DN, HD, JP, FK, NS, MC contributed to generating the data included in this manuscript; all authors contributed to the design and/or supervision of the work. Funding Open Access funding enabled and organized by Projekt DEAL. Competing interests The authors declare no competing interests. Additional information Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41467-021-26826-3. Correspondence and requests for materials should be addressed to Rolf Jessberger or Mariola Chacón. Peer review information Nature Communications thanks P. Todd Stukenberg and the other anonymous reviewer(s) for their contribution to the peer review this work. Reprints and permission information is available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. © The Author(s) 2021 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-26826-3 16 NATURE COMMUNICATIONS | (2021) 12:6547 | https://doi.org/10.1038/s41467-021-26826-3 | www.nature.com/naturecommunications