Microtubule occupancy at kinetochores links checkpoint silencing with mitotic memory

18 The spindle assembly checkpoint (SAC) promotes faithful chromosome 19 segregation by delaying mitosis until all kinetochores attach to spindle 20 microtubules. Paradoxically, a p53-dependent memory mechanism - the “ mitotic 21 stopwatch” - blocks daughter cell proliferation after unusually prolonged mitoses. 22 Understanding how the SAC coordinates with the mitotic stopwatch is critical to 23 disentangle this conundrum. Here, we found that microtubule occupancy at 24 kinetochores is a cornerstone linking SAC silencing with mitotic memory. By 25 combining live-cell with super-resolution microscopy, FRAP , laser microsurgery 26 and molecular perturbations in Indian muntjac fibroblasts, we show that SAC 27 silencing at kinetochores is gradual, non-uniform and confined to highly-localized 28 microtubule attachments. Augmin promotes timely SAC silencing with maximal 29 microtubule occupancy at kinetochores, whereas MPS1/CDK1 inhibition 30 bypasses this condition. Partial microtubule occupancy delays SAC silencing, 31 increases segregation errors and blocks daughter cell proliferation. Thus, timely 32 SAC silencing with high microtubule occupancy avoids “bad memories” of mitosis 33 to allow daughter cell proliferation. 34 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 3

35 Faithful chromosome segregation during mitosis is essential for tissue 36 homeostasis. Errors in chromosome segregation may lead to aneuploidy 37 compromising cell viability, or drive chromosomal instability, a hallmark of human 38 cancers implicated in tumor evolution, metastasis, therapy resistance, and 39 reduced patient survival 1. To promote faithful chromosome segregation during 40 mitosis, the spindle assembly checkpoint (SAC) delays anaphase until all 41 kinetochores attach to mitotic spindle microtubules 2. At the root of the SAC 42 signaling cascade, a MAD1/MAD2 complex recruited to unattached kinetochores 43 initiates a catalytic process that activates free cytosolic MAD2 to bind CDC20, 44 BUBR1 and BUB3, producing a diffusible Mitotic Checkpoint Complex (MCC). 45 The MCC inhibits the Anaphase Promoting Complex/Cyclosome (APC/C), an 46 E3-ubiquitin ligase that otherwise targets Cyclin B1 and Securin for degradation 47 by the proteasome (reviewed in 3-5). This releases the cysteine-protease 48 Separase to cleave the Cohesin molecules that hold sister-chromatids together, 49 thereby triggering their initial separation in anaphase, while preventing SAC 50 reactivation during anaphase 6,7. 51 Upon kinetochore attachment to microtubules, SAC strength gradually 52 decreases as result of the removal of MAD1/MAD2, eventually halting the 53 catalytic cascade underlying MCC production 8,9. The removal of MAD1/MAD2 54 from kinetochores during SAC silencing is mediated by PP1 and PP2A 55 phosphatases, which are recruited by the outer kinetochore component KNL1 to 56 oppose the action of key mitotic kinases, such as CDK1 and MPS1 10-15. As 57 microtubules establish end-on attachments, MPS1 is outcompeted from NDC80 58 complexes at kinetochores 16,17, and the MAD1/MAD2 complex is removed by 59 dynein-mediated stripping as the fibrous corona disassembles18-22. 60 Pioneering correlative light and electron microscopy work in PtK1 cells 61 showed that their kinetochores bind more microtubules in anaphase than in 62 metaphase, indicating that SAC silencing during normal mitosis can occur at 63 ∼85% microtubule occupancy at kinetochores 23. Classic micromanipulation 64 experiments in grasshopper spermatocytes further indicated that weakly 65 attached kinetochores with less than 40% microtubule occupancy show greatly 66 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 4 diminished MAD1/MAD2 recruitment when compared with fully unattached 67 kinetochores, delaying but not preventing anaphase onset 24. Indeed, in this 68 system, high microtubule occupancy was found to be necessary for the reliable 69 and complete removal of MAD1/MAD2 from kinetochores 24. In contrast, two 70 recent studies in human HeLa cells overexpressing individual or combinations of 71 Hec1 phosphomimetic mutants to create intermediate average microtubule 72 attachment states at kinetochores estimated that 20–50% of the normal 73 metaphase microtubule occupancy is sufficient to trigger complete and uniform 74 removal of MAD1/MAD2 from kinetochores 25,26. However, this switch-like model 75 remains difficult to reconcile with the finding that the rate of MAD1/MAD2 loss 76 from kinetochores increases with microtubule occupancy, with SAC silencing 77 under conditions of low microtubule occupancy resulting in extensive mitotic 78 delays 24-26. A third study in human RPE1 cells used a microtubule 79 depolymerizing drug to reduce microtubule occupancy at kinetochores 27. Based 80 on direct microtubule counting by electron microscopy, this study showed that 81 ~70% microtubule occupancy at metaphase kinetochores is sufficient to silence 82 the SAC, with less than 5 min delay relative to controls, but at the price of 83 increasing the incidence of lagging chromosomes during anaphase 27. Thus, 84 although cells can silence the SAC with low microtubule occupancy, there may be 85 selective pressure for high microtubule occupancy at kinetochores before 86 initiating anaphase, ensuring proper chromosome segregation and avoiding 87 significant mitotic delays 27,28. This hypothesis gains weight in light of the recently 88 uncovered “mitotic stopwatch” complex formed by 53BP1 and the deubiquitinase 89 USP28 that stabilizes p53 in response to even moderate delays in mitosis, 90 leading to subsequent cell cycle arrest or death in G1, independently of DNA 91 damage 29-36. However, how SAC silencing is coordinated with the mitotic 92 stopwatch to control daughter cell proliferation remains unknown. 93 Here we set to investigate whether and how the SAC responds to 94 highly-localized microtubule attachments at kinetochores, while testing whether 95 the extent of microtubule occupancy underlying SAC silencing impacts daughter 96 cell proliferation. Overall, our data support a gradual, non-uniform (i.e. confined to 97 highly-localized microtubule attachments), SAC-silencing mechanism that is 98 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 5 assisted by Augmin to promote high microtubule occupancy at kinetochores for 99 timely and faithful chromosome segregation. Importantly, if cells ultimately 100 silence the SAC after experiencing significant mitotic delays due to low 101 microtubule occupancy at kinetochores, the resulting daughter cells are 102 bookmarked to prevent their proliferation and the potential propagation of 103 segregation errors. 104 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 6

105 106 MAD1 is gradually removed from Indian muntjac kinetochores in a 107 microtubule-dependent manner 108 According to the switch-like SAC-silencing model, kinetochores respond as a 109 single unit to silence the SAC after just a few microtubule attachments are 110 established 25,26. However, because human kinetochores cannot be resolved by 111 conventional light microscopy, this postulate has so far been impossible to test 112 experimentally. To overcome this limitation, we took advantage of the unique 113 cytological features of female Indian muntjac fibroblasts, which carry only six 114 chromosomes with naturally “super-resolved” kinetochores that arose by tandem 115 and centric fusions of smaller ancestral chromosomes 37-43. As a first step 116 towards investigating SAC silencing in this system, we determined the time 117 between MAD1 disappearance from kinetochores and anaphase onset by 118 tracking individual kinetochores in live, hTERT-immortalized, female Indian 119 muntjac fibroblasts stably expressing mScarlet-CENP-A (a constitutive 120 kinetochore marker) and Venus-MAD1 8. We found that MAD1 signal at 121 kinetochores became essentially undetectable ~20 min before anaphase onset 122 (Fig. S1A, B), in agreement with previous reports in other vertebrate cell lines 44,45 123 and consistent with the temporal framework of anaphase onset upon 124 laser-mediated ablation of the last unattached kinetochore 2. 125 To determine the kinetics of SAC silencing in response to microtubule 126 attachments at individual kinetochores in live Indian muntjac fibroblasts, we 127 combined stable expression of fluorescent Venus-MAD1 with carefully titrated 128 SiR-tubulin 41 under slight cell compression to improve visualization 42. As a 129 control, cytoplasmic Venus-MAD1 signal was simultaneously measured, and 130 photobleaching was found to be negligible throughout the course of the 131 experiment (Fig. S1C). MAD1 decayed gradually at individual kinetochores as 132 chromosomes bi-oriented and aligned at the equator during prometaphase 133 (Figure 1A, C), with a rate of decay that was well fit to a single exponential (Figure 134 1E and Fig.S2), in agreement with previous reports in Ptk2 cells 46. Noteworthy, 135 pole-proximal chromosomes aligned last and did so tangentially along the 136 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 7 curvature of the spindle edge. Moreover, their leading kinetochores were 137 enriched with MAD1 from the onset of congression, with the signal gradually 138 decreasing as chromosomes approached the equator (Figure 1A, A’, A’’). 139 Importantly, MAD1 typically decayed first on trailing kinetochores, likely reflecting 140 differences in the timing or efficiency in the formation of stable end-on 141 microtubule attachments compared with leading kinetochores. (Figure 1A, A’, A’’). 142 These results are fully consistent with congression models where pole-proximal, 143 mono-oriented chromosomes use plus-end-directed motors on the leading 144 kinetochore to initiate lateral gliding along pre-existing spindle microtubules 145 towards the equator 46-49, while at odds with models where stabilization of end-on 146 attachments and bi-orientation take place before pole-proximal chromosomes 147 initiate congression 50. 148 To test whether gradual MAD1 decay at kinetochores during mitosis depends 149 on microtubule occupancy, we monitored MAD1 levels in the presence of 1 µM 150 nocodazole, which depolymerized all spindle microtubules. As expected, under 151 these conditions, MAD1 localized on every kinetochore for over 20 min, without 152 any measurable decay (Figure 1B, C and Fig. S2). These data indicate that 153 gradual SAC silencing during mitosis depends on the establishment of 154 microtubule attachments at kinetochores. 155 156 MAD1 is rapidly removed from Indian muntjac kinetochores upon MPS1 or 157 CDK1 inactivation, regardless of microtubule occupancy 158 MPS1 and CDK1 kinases are critical to maintain SAC signaling by counteracting 159 the activities of PP1 and PP2A phosphatases required for mitotic exit 10-15. 160 Therefore, we next investigated the kinetics of MAD1 decay at kinetochores after 161 acute MPS1 or CDK1 inactivation with MPS1-IN-1 51 or RO 3306 52, respectively. 162 We found that MAD1 decayed more abruptly from kinetochores after MPS1 or 163 CDK1 inhibition, regardless of microtubule occupancy, reaching 50% of the initial 164 levels 3.8x or 1.5x faster than during normal mitosis, respectively (Figure 1B, C, E 165 and Fig. S2). These data indicate that inactivation of MPS1 or CDK1 can bypass 166 the requirement of a given microtubule occupancy at kinetochores to silence the 167 SAC. 168 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 8 MAD1 removal depends on kinetochore size 169 A prediction from switch-like models that favor independent microtubule binding 170 at kinetochores and where a given percentage of occupancy is key to uniformly 171 silence the SAC 25,26 is that small or large kinetochores would simultaneously 172 reach the same percentage of occupancy (and, consequently, have similar MAD1 173 half reduction times), despite binding a different total number of microtubules. 174 However, despite ample evidence of kinetochore size differences in humans 53-61, 175 this prediction is currently impossible to test experimentally in human cells for the 176 same reasons evoked before. In contrast, the large kinetochore from 177 chromosome 3+X in female Indian muntjac cells can be unequivocally 178 distinguished from those on the smaller chromosomes 1 and 2, allowing us to 179 track MAD1 decay as a function of kinetochore size 38,40,41. Surprisingly, we found 180 that MAD1 signal on the large kinetochores of chromosome 3+X requires ~38% 181 more time to become undetectable, when compared with smaller kinetochores on 182 chromosomes 1 and 2 (Figure 1D, E, and Fig. S2). These findings suggest that 183 switch-like models do not fully account for how the SAC is normally silenced. 184 185 MAD1 is immobile within kinetochores 186 In order to obtain a deeper understanding of the mechanisms underlying SAC 187 silencing we investigated MAD1 dynamics within kinetochores. Two possible 188 scenarios could be envisioned: 1) if MAD1 is mobile within kinetochores, few 189 microtubule attachments might be sufficient to “extinguish” all MAD1, for example 190 by Dynein-mediated stripping along microtubules 18,19 (Figure 2A); 2) if MAD1 is 191 immobile within kinetochores, Dynein-mediated stripping along microtubules 192 would be confined to the sites or domains where microtubules are attached 193 (Figure 2A). To distinguish between these two scenarios, we photobleached 194 Venus-MAD1 on a fraction (up to 50%) of the large kinetochore from 195 chromosome 3+X in the absence of microtubules. Clear predictions about MAD1 196 mobility within kinetochores could be made, depending on the FRAP profile. If 197 MAD1 is mobile within kinetochores, the photobleached fraction of the 198 kinetochore is expected to recover due to migration of fluorescent Venus-MAD1 199 molecules from the unbleached fraction, resulting in the convergence of signals 200 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 9 to an intermediate and uniform level along the entire kinetochore (Figure 2A’). In 201 contrast, if MAD1 is immobile within kinetochores, the photobleached fraction of 202 the kinetochore would be expected to only recover fluorescent Venus-MAD1 203 molecules that are exchanging with the cytoplasmic pool, while the unbleached 204 fraction would remain stable over time (Figure 2A’). We were successful in 205 partially photobleaching Venus-MAD1 in a fraction of the kinetochore, with 206 minimal/unintended photobleaching of Venus-MAD1 on the remaining fraction 207 (Figure 2B-B’’). Importantly, after photobleaching of the intended fraction of the 208 kinetochore, Venus-MAD1 fluorescence recovered up to ~50%, with a half 209 recovery time of 2-6 seconds, while remaining stable (or increasing slightly; see 210 ahead) over time in the unbleached fraction (Figure 2B-B’’, D, E). As a control, we 211 photobleached Venus-MAD1 on entire kinetochores, which resulted in identical 212 recovery parameters when compared to partial photobleaching, and in line with 213 previous measurements in other mammalian systems 44,45 (Figure 2C-C’’, D, E). 214 Because fluorescence recovery of the mobile fraction of Venus-MAD1 after 215 photobleaching entire kinetochores is exclusively due to exchange with the 216 cytoplasmic pool 44,45, we concluded that this pool also accounts for the observed 217 fluorescence recovery of Venus-MAD1 on partially photobleached kinetochores. 218 Taken together, these data indicate that MAD1 is essentially immobile within 219 kinetochores, suggesting that it can only be removed from spatially confined 220 regions upon the establishment of highly-localized microtubule attachments or by 221 a microtubule-independent SAC-silencing mechanism, such as the one involving 222 CDK1 or MPS1 inactivation. 223 224 MAD1 removal along individual kinetochores is non-uniform and confined 225 to highly-localized microtubule attachments 226 The conclusion from our FRAP experiments implies that, in a normal mitosis, 227 there may be low microtubule occupancy states in which MAD1 removal is limited 228 to regions where microtubule attachments are established. To test this prediction, 229 we monitored Venus-MAD1 decay along the entire kinetochore length as 230 microtubules attach during mitosis. Near-simultaneous tracking of individual 231 kinetochores, MAD1 and microtubules in living cells revealed a non-uniform 232 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 10 decay of MAD1 along kinetochores (Figure 3A, B). To further dissect the 233 relationship between non-uniform MAD1 decay and microtubule occupancy 234 along kinetochores, we used super-resolution CH-STED microscopy 62 and 3D 235 image rendering to inspect (fixed) late prometaphase cells with only partial MAD1 236 signal along kinetochores (Figure 4A, B and Fig. S3A, B). We found clear 237 examples where fractions of the kinetochore had end-on attached microtubules 238 with no detectable MAD1, while other fractions were devoid of microtubules and 239 clearly enriched for MAD1 (Figure 4A, B and Fig. S3A, B). A line scanning profile 240 along the entire length of several such individual kinetochores confirmed a 241 negative correlation between MAD1 signal and microtubule density in 15/21 242 (71.4%) of the cases (Figure 4C, D and Fig. S3A’, B’). 243 To improve the resolution of our analyses we divided the longitudinal axis of 244 partially attached kinetochores into multiple subsections and quantified MAD1 245 and microtubule intensity at each subsection (Figure 4E). We found a weak, yet 246 highly significant, negative correlation between MAD1 and microtubule signal 247 intensity along kinetochores (Figure 4F and Fig. S3C). Of note, the resulting 248 L-shape curve indicates that while high microtubule signal is a good predictor of 249 the absence of MAD1 at kinetochores, the opposite is not the case, suggesting 250 that there are regions at kinetochores without microtubules that have no 251 detectable MAD1, consistent with models evoking cooperative microtubule 252 binding at kinetochores 63,64. Overall, these data show that SAC silencing at 253 kinetochores is non-uniform and confined to highly-localized microtubule 254 attachments. 255 256 Local perturbation of kinetochore function results in a local SAC response 257 to microtubule attachments/detachments 258 To directly test how the SAC responds to localized microtubule 259 attachments/detachments, we used laser microsurgery to partially ablate 260 individual metaphase kinetochores (and, consequently, microtubules that were 261 attached to the ablated region) upon SAC silencing in live Indian muntjac 262 fibroblasts (Figure 5A). These cells stably expressed GFP-CENTRIN-1 to mark 263 spindle pole positions, 2xGFP-CENP-A to define the kinetochore target, and 264 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 11 mScarlet-MAD1 to monitor SAC status before and after surgery (Figure 5B). We 265 found that, within few minutes after partial kinetochore ablation, the unperturbed 266 sister kinetochore typically bent towards the attached spindle pole without any 267 detectable MAD1 (Figure 5A-C). However, ~5 min after bending, MAD1 was 268 often recruited to the tip of the bent unperturbed sister kinetochore, presumably 269 due to highly-localized microtubule detachments (Figure 5A-D). In other cases, 270 MAD1 was recruited to a more central region of the bent unperturbed sister 271 kinetochore (Fig. S4A, A’, B). Strikingly, in all cases MAD1 accumulation was 272 highly localized and distinct from the uniform accumulation of MAD1 on fully 273 unattached kinetochores resulting from nocodazole treatment (Figure 5B, D, E). 274 Therefore, local kinetochore perturbations result in a highly-localized SAC 275 response. 276 To assist us in the interpretation of the laser microsurgery data and determine 277 whether the SAC was responding to microtubule detachments in the bent 278 unperturbed sister kinetochore, we performed correlative super-resolution 279 CH-STED microscopy after fixation of cells that showed MAD1 recruitment upon 280 partial sister kinetochore laser ablation by live imaging (Figure 5F, F’ and Fig. 281 S4C, C’). We found that MAD1 accumulation, either at the tip or at a more central 282 region of the unperturbed sister kinetochore, negatively correlated with 283 microtubule density (Figure 5F’, F’’ and Fig. S4C’, C’’). We reasoned that MAD1 284 accumulation near the tip of the unperturbed kinetochore may result from highly 285 localized microtubule detachments due to lack of tension in the partially ablated 286 sister. In other cases, localized forces applied to both ends of the unperturbed 287 kinetochore may generate enough tension along its length to allow a few 288 microtubules to remain stably attached near the tip, resulting in more central 289 accumulation of MAD1. These results demonstrate that the SAC responds locally 290 to highly localized microtubule attachment/detachment events along 291 kinetochores. 292 293 Low microtubule occupancy at kinetochores after Augmin depletion 294 delays MAD1 removal and SAC silencing 295 296 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 12 Next, we sought to investigate the cellular consequences of low microtubule 297 occupancy at kinetochores. To do so we monitored SAC silencing after 298 RNAi-mediated depletion of HAUS6, a subunit of the Augmin complex required 299 for K-fiber maturation 43,65. We found that MAD1 decay from kinetochores was 300 slowed-down in HAUS6-depleted cells, as revealed by ~50% increase in the half 301 reduction time of MAD1 signal at kinetochores relative to mock-depleted controls 302 (Figure 6A-C, Figure S2). Moreover, even when HAUS6-depleted cells were 303 treated with the proteasome inhibitor MG132 for 1h to provide more time for 304 K-fiber maturation, many more (66% vs. 12%) cells had at least one 305 MAD1-positive kinetochore and overall higher MAD1 levels at kinetochores when 306 compared with control metaphase cells (but lower than prometaphase levels) 307 (Figure 6D, E). Closer inspection of mock- and HAUS6-depleted fixed cells with 308 CH-STED super-resolution microscopy revealed that, while kinetochores in 309 control mock-depleted metaphase cells treated with MG132 were abundantly 310 occupied with microtubules and without detectable MAD1 signal, in 311 HAUS6-depleted cells treated with MG132 MAD1 signal could still be detected 312 along the kinetochores, despite interspersed end-on microtubule attachments 313 (Figure 6F). However, unlike in control cells with clear partially attached 314 kinetochores that showed a negative correlation between MAD1 and microtubule 315 signal (r = -0.29, p<1e-10; Figure S5E), in HAUS6-depleted cells this correlation 316 was lost (r = -0.06, p=0.14; Figure S5D, E and Figure S6C), suggesting that the 317 resulting perturbations of microtubule occupancy were beyond our resolution 318 capacity (Fig. S5A-C and Fig. S6A-B’). Overall, these data are in line with 319 previous findings in human cells 25,26 indicating that low microtubule occupancy at 320 kinetochores delays MAD1 removal from kinetochores and consequently SAC 321 silencing. 322 323 Low microtubule occupancy delays SAC silencing, increases segregation 324 errors, and blocks daughter cell proliferation 325 To further characterize the cellular consequences of low microtubule occupancy 326 at kinetochores, we tracked HAUS6-depleted cells over several days by live-cell 327 microscopy and determined their fates (Figure 7A). Mitoses in HAUS6-depleted 328 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 13 cells were significantly delayed and error-prone, proportional to the extent of 329 HAUS6 depletion over time (Figure 7A’ and B). When inspecting the outcomes 330 resulting from mitotic delays in HAUS6-depleted cells more closely, we found 331 three regimens. The majority of cells with mitotic durations below 200 minutes 332 divided without segregation errors or spindle defects (Figure 7C). However, after 333 a delay of approx. 200 minutes, virtually all cells divided with mitotic defects 334 (Figure 7C), pointing towards persistently immature K-fibers at anaphase onset 335 27. Strikingly, most cells that were stuck in mitosis for longer than 400 minutes 336 eventually died in mitosis (Figure 7C). 337 To determine the fate of daughter cells from mothers that did divide despite 338 immature K-fibers, we defined a cohort of cells that underwent mitosis within 12 339 and 36 hours of HAUS6 depletion and followed their offspring for 60 hours 340 (Figure 7A). Strikingly, 74% of daughter cells arising from mitoses with durations 341 longer than approx. 90 minutes stopped proliferating, whereas only 32% of 342 daughter cells derived from shorter mitosis arrested (Figure 7D and S7F), in line 343 with previous findings in human cells, suggesting active surveillance by the 344 “ mitotic stopwatch” 29,31,32,34. To confirm that such a “memory” of mitotic duration 345 is conserved in Indian muntjac cells, we tracked daughter cell fates after 346 releasing cells from up to 6 hours of mitotic arrest mediated by the Eg5-inhibitor 347 monastrol (Fig. S7A). In agreement with our observations in HAUS6-depleted 348 cells, mother cells that experienced mitotic times longer than approx. 90 minutes 349 gave rise to daughters that showed increased cell cycle lengths or stopped 350 proliferating (>90 min mitosis - 61% of daughters arrested; <90 min mitosis - 30% 351 of daughters arrested; Fig. S7B-F). 352 Together, these observations suggest the presence of a conserved mitotic 353 stopwatch mechanism surveilling mitotic durations. Indeed, we were able to 354 detect a significant accumulation of USP28 bound to 53BP1 – two components of 355 the mitotic stopwatch complex – in Indian muntjac cells that experienced a 356 prolonged mitosis (note that available anti-p53 monoclonal antibodies do not 357 recognize the Indian muntjac protein under denatured conditions; Fig. S7G, H). 358 However, since prolonged mitotic durations under low microtubule occupancy 359 (i.e. HAUS6 depletion) or monastrol perturbation were tightly associated with 360 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 14 mitotic errors (Figures 7B, C and Fig. S7C, D) that may independently block 361 daughter cell proliferation 66-69, we sought to disentangle the two events. To this 362 end, the low chromosome number of Indian muntjac fibroblasts allowed us to 363 unequivocally exclude cells dividing with errors from our cell fate analysis (Figure 364 7E and Fig. S7E, E’, F). Remarkably, mother cells with increasing mitotic 365 durations - but dividing error-free - were still increasingly more likely to give rise to 366 daughter cells that arrest in G1 (Figure 7E and Fig. S7E, E’, F), suggesting active 367 “mitotic stopwatch” surveillance of mitotic duration. 368 369 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 15

370 While there is a large consensus that SAC silencing can occur with low 371 microtubule occupancy at kinetochores 23-27,46, whether and how SAC signaling 372 along vertebrate kinetochores responds to highly localized microtubule 373 attachments has remained a central open question. Current models suggest that 374 mammalian kinetochores respond as a single unit that globally halts SAC 375 signaling upon partial microtubule attachments 25,26,46. As so, low microtubule 376 occupancy would cause kinetochores to respond as a whole and predicts a 377 uniform loss of MAD1 along the kinetochores. Moreover, these switch-like 378 models imply the existence of a sensitive microtubule “counting” mechanism that 379 determines SAC silencing when a given occupancy at each individual 380 kinetochore is reached. However, in virtue of the sub-diffraction nature of most 381 vertebrate (humans included) kinetochores, this prediction had never been 382 directly tested and the essence of a putative microtubule “counting” mechanism 383 remains elusive. In this regard, Indian muntjac kinetochores offer a unique setting 384 to test these ideas. In stark contrast with the predictions from switch-like 385 SAC-silencing models, we found that the removal of the SAC protein MAD1 from 386 Indian muntjac kinetochores is gradual, non-uniform, correlates with local 387 microtubule density and kinetochore size, and depends on Augmin-mediated 388 maturation of K-fibers. 389 In line with previous work 46, MAD1 signal decay from Indian muntjac 390 kinetochores was well fit by a single exponential, indicating that SAC silencing in 391 this system faithfully replicates established kinetics in other vertebrate systems. 392 Switch-like models that favor independent microtubule binding at kinetochores 393 and where a given percentage of occupancy is key to uniformly silence the SAC 394 25,26, predict that the fraction of unattached sites is expected to decay 395 exponentially, with a timescale that does not depend on kinetochore size. 396 Interestingly, however, we found that MAD1 decay on large kinetochores was 397 38% slower compared to small kinetochores. While the underlying reasons for 398 these differences remain unclear, this finding suggests that switch-like models do 399 not fully account for how the SAC is normally silenced. Indeed, both live-cell 400 imaging and super-resolution CH-STED analysis in Indian muntjac cells revealed 401 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 16 a non-uniform distribution of MAD1 on partially attached kinetochores that 402 negatively correlated with microtubule occupancy. Moreover, laser microsurgery 403 experiments demonstrated that once silenced, the SAC remains sensitive to 404 highly-localized microtubule detachments from kinetochores, even when more 405 than 50% occupancy persists. Lastly, experimental perturbation of microtubule 406 occupancy at kinetochores by interfering with K-fiber maturation significantly 407 delayed MAD1 removal and SAC silencing. Therefore, although low microtubule 408 occupancy at kinetochores may eventually silence the SAC, there is an 409 unavoidable impact on the respective mitotic duration. This has important 410 implications in light of the recently-uncovered “mitotic stopwatch” mechanism, 411 where even mild mitotic delays elicit a subsequent 412 USP28-53BP1-p53-dependent cell cycle arrest in G1 29,31,32,35,70, and the 413 observation that SAC silencing with low microtubule occupancy increases the 414 incidence of lagging chromosomes during anaphase, potentially leading to 415 aneuploidy 27. As so, the mechanism underlying SAC silencing at individual 416 kinetochores is likely to have evolved to maximize microtubule occupancy within 417 a strict temporal window to avoid “bad memories” of mitosis, and we show here 418 that the Augmin complex, through its role in K-fiber maturation 43,65, is 419 instrumental in this process (Figure 8). 420 But how do partially attached kinetochores that are locally sensitive to 421 microtubule attachments ultimately remove all MAD1/MAD2? In line with a model 422 in which the number of unattached kinetochores ultimately determines the rate of 423 MCC production that proportionally inhibits the APC/C 8,9, partially-attached 424 kinetochores may produce less MCC than fully unattached kinetochores 24. 425 Consequently, the APC/C would become sufficiently active and allow for slower, 426 yet steady, Cyclin B1 degradation (and CDK1 inactivation), ultimately silencing 427 the SAC after an extensive mitotic delay 24. Whether kinetochores eventually 428 reach normal microtubule occupancy by the time the SAC is silenced after a 429 delay remains unclear (Figure 8), but it should be noted that even cells that are 430 unable to silence the SAC after extremely prolonged mitotic delays due to the 431 complete absence of microtubules degrade Cyclin B1 and eventually exit mitosis 432 due to residual APC/C activity 71,72. Our finding that CDK1 inhibition triggers SAC 433 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 17 silencing irrespective of microtubule occupancy further supports this 434 interpretation (Figure 8). 435 One may argue that the compound architecture of Indian muntjac 436 kinetochores does not reflect what is normally found in other vertebrate species, 437 including humans. However, we have no evidence that kinetochore architecture 438 in Indian muntjac reflects a significant deviation at the level of the core 439 microtubule-binding moduli, or other functional differences, including MAD1 440 decay during SAC silencing, when compared with other vertebrates 41,43,46,73. 441 Noteworthy, recent works have revealed that many human, mouse and chicken 442 centromeres/kinetochores also have a compound, often bipartite, architecture 443 59,61,74,75. In particular, the same human kinetochore could be found with one 444 sub-domain end-on attached, and the other laterally attached to microtubules via 445 the fibrous corona 74. This is consistent with our finding of non-uniform, highly 446 localized, MAD1 removal in response to local microtubule occupancy in Indian 447 muntjac kinetochores. Importantly, while in these cases MAD1 and local 448 microtubule density often negatively correlated along two apparently insulated 449 adjacent domains on the same large kinetochore of chromosome 3+X (simply 450 because these extreme cases are easier to detect), there were other cases 451 (particularly evident after Augmin perturbation) where this negative correlation 452 was observed along three or more sub-domains, without following a strict spatial 453 pattern (also evident in the distinct responses after laser microsurgery 454 experiments). Moreover, our finding that MAD1 is immobile within unattached 455 kinetochores indicates that it can only be removed upon the establishment of 456 highly-localized microtubule attachments (or eventually by a 457 microtubule-independent SAC-silencing mechanism, such as MPS1 or CDK1 458 inactivation). Overall, our data favors a model where anaphase onset is 459 determined by the gradual silencing of the SAC in response to highly localized 460 microtubule attachments along the entire kinetochore length, rather than a 461 uniform, switch-like mechanism. This model reconciles gradual SAC silencing 462 with increasing microtubule occupancy at kinetochores, without the need for a 463 dedicated microtubule “counting” system that triggers SAC inactivation. Most 464 relevant, our finding that microtubule occupancy at kinetochores is at the root 465 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 18 between timely SAC silencing and mitotic memory, identifies a critical 466 rate-limiting step with important implications for cell proliferation in mammals. 467 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 19

468 Cell lines and culture conditions 469 Indian muntjac cell lines were grown in Minimum Essential Media (MEM) 470 (Corning), supplemented with 10% FBS (GIBCO, Life Technologies). Indian 471 muntjac hTERT-immortalized fibroblasts were a gift from Jerry W. Shay 76. IM 472 fibroblasts stably expressing Venus-MAD1 were generated by lipofection of 473 pVenus-MAD1 (kind gift from Jakob Nilsson 77). RPE-1 hTERT-immortalized cells 474 were cultured in DMEM (Corning™, 15323531) supplemented with 10% FBS 475 (GIBCO, Life Technologies). All cells were kept at 37ºC in humidified conditions 476 with 5% CO 2. Transfection was performed by incubating the cells with 1:400 477 Lipofectamine2000 (Invitrogen) in Opti-MEM (GIBCO, Life Technologies) for 6h, 478 and stable and uniformly expressing cells were selected by FACS sorting (as 479 detailed in 42. For the production of Venus-MAD1 and mScarlet-CENP-A 480 co-expressing cell line, Venus-MAD1 cells were transduced with 481 pLVx-mScarlet-CENP-A lentiviral plasmid (generated in 43 ) . IM cells 482 co-expressing 2x-GFP-CENP-A, GFP-CENTRIN-1 and mScarlet-MAD1 were 483 produced by lentiviral transduction of pRRL-2x-EGFP-CENP-A (generated in 43 ), 484 pLVX-EGFP-C1-CENTRIN-1 (kind gift from Manuel Thery; Addgene plasmid # 485 73331) and pInducer-mScarlet-MAD1 (generated in this paper). Cells expressing 486 GFP-H2B were produced by lentiviral transduction of LV-GFP (kind gift from 487 Elaine Fuchs; Addgene plasmid # 25999). Lentiviral transduction was performed 488 as detailed in 42. Cells were incubated with lentivirus particles in the presence of 489 1:2000 Polybrene (Sigma-Aldrich; TR-1003) for 24h before supplied with fresh 490 complete media. Stable lines with sufficient fluorescence intensity were selected 491 by FACS sorting. 492 493 Plasmid design 494 pInducer-mScarlet-MAD1 was produced by cloning mScarlet (from 495 pRRL-mScarlet-CENP-A) and MAD1 (from pVenus-MAD1) into the pInducer-20 496 backbone under a Tet-inducible promotor. To construct the empty backbone 497 KASH-mCherry was removed by SalI + XhoI restriction digestion of pinducer 20 498 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 20 DN-KASH (kind gift from Daniel Conway; (Addgene plasmid # 125554)). 499 Restriction sites were amplified from the plasmid by PCR (fwd: 500 5’-TTAAAGGAACCAATTCAGGCTAGCACGCGTATATCTAGACCCAGCTTTCT-501 3’; rev: 5’-TAAGCGTAGTCTGGGACG-3’) and added to the emptied vector 502 through Gibson assembly. mScarlet and MAD1 were cloned into the NheI 503 linearized pInducer-20 backbone through Gibson assembly. 504 505 siRNA Experiments 506 Protein knockdown was performed as detailed in 42 . Briefly, IM fibroblasts seeded 507 at 70% confluency were starved with MEM supplemented with 5% FBS for 30 508 minutes before transfection. siRNA transfection was performed by adding 509 Lipofectamine RNAi Max (1:400; Invitrogen) and 50 nM of siRNA in serum 510 free-medium (Opti-MEM, Gibco) for 6h before replacing the solution with fresh 511 complete media. Cells were analyzed 72 hours after depletion. In long-term live 512 cell imaging experiments (see below), siRNAs were not replaced with fresh 513 medium and imaging started shortly after treatment. For the depletion of HAUS6, 514 the following sequence was used: 515 5’-GGUUGGUCCUAAGUUUAUU[dT][dT]-3’39. Cells mock transfected 516 (lipofectamine only) or transfected with luciferase-targeting siRNA were used as 517 controls. 518 519 Drug Treatments 520 For acute inhibition of Mps1 and CDK1 kinases, 4 µM MPS1-IN-1 (kind gift from 521 N.Gray 51) and 10 μM RO 3306 (ChemCruz; CAS 872573-93-8), respectively, 522 were added shortly after Nuclear Envelop Breakdown (NEB). To arrest cells in 523 G2, these were treated with 10 μ M RO 3306 for 14h. Microtubule 524 depolymerization was triggered using 1 or 3.3 µM of Nocodazole (Sigma-Aldrich; 525 CAS 31430-18-9) for 30 mins before the analysis. To increase mitotic duration 526 cells were arrested in prometaphase with 100 μL of Monastrol (Tocris Bioscience, 527 Cat. No.1305). Metaphase arrest was obtained using 5 µM MG132 528 (Sigma-Aldrich; CAS 133407-82-6). Fixed cell analysis using MG132 was 529 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 21 performed in the first 1.5 h after drug addition to avoid cohesion fatigue. 530 SiR-tubulin (Spirochrome; SC002) 78 was used to visualize microtubules, at 50 531 nM concentration incubated for 1 h prior to live-cell imaging. To induce p53 532 enrichment cells were treated with 5 μ M etoposide (Selleck Chemicals, S1225) or 533 3.3 µM Nocodazole for 16h. The same volume of DMSO (Sigma-Aldrich, D4540) 534 was used as a control. 535 536 Immunofluorescence 537 Indian muntjac fibroblasts were seeded on fibronectin coated coverslips (#1.5 538 thickness) 24h before the experiment. Depending on the microscopy routine, 539 different fixation protocols were used: for widefield imaging cells were incubated 540 with 4% paraformaldehyde (Electron Microscopy Sciences) in PBS for 10 min at 541 room temperature; for STED microscopy PFA 4% supplemented with 0.1%-0.2 % 542 Glutaraldehyde (Electron Microscopy Sciences) in Cytoskeleten Buffer (CB - 543 274 mM NaCl, 10 mM KCl, 2.2 mM Na 2HPO4, 0.8 mM KH 2PO4, 4 mM EGTA, 544 4 mM MgCl 2, 10 mM Pipes, 10 mM glucose, pH 6.1) for 10 min at room 545 temperature was used; for same cell correlative live confocal and 546 super-resolution STED microscopy pre-warmed 2x concentrated PFA + 547 Glutaraldehyde diluted in imaging media to a final concentration of 4% and 0.2%, 548 respectively, was added to the imaging chamber. After a 5 min incubation at the 549 microscope (37ºC), the fixation solution was replaced with PFA 4% 550 supplemented with 0.2 % Glutaraldehyde in CB for an additional 5 min at room 551 temperature. Autofluorescence was quenched by a 0.1% sodium borohydride 552 solution (Sigma-Aldrich) for 7min. Cells were permeabilized with CB-0.5%Triton 553 for 20-30 min and blocked with CB-0.05% Tween 20 with 10% FBS (blocking 554 buffer) for 1 hour at RT. Samples were incubated with primary antibodies diluted 555 in blocking buffer over-night at 4ºC. The following primary antibodies were used: 556 human anti-centromere antiserum (ACA, Fitzgerald; 90C-CS1058) 1:2000/1:200 557 (Widefield/ STED); anti-tyrosinated tubulin (Bio-Rad; MCA77G) 1:2000/1:100 558 (Widefield/ STED); anti α-tubulin (clone B-5-1-2, Sigma-Aldrich; T5168) 1:100 559 (correlative live-cell and fixed STED microscopy). Cells were washed with 560 PBS-0.05% Tween before incubation with the corresponding secondary 561 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 22 antibodies for 1h at RT – Alexa Fluor 568 and 647 (Thermo Fisher Scientific) 562 1:1000; or abberior STAR 580 (Abberior Instruments; ST580) and abberior 563 STAR-Red (Abberior Instruments; STRED) 1:200 for STED microscopy. DNA 564 was labeled with a quick incubation in the presence of 1 µg/mL 565 4’6’-Diamidino-2-phenylindole (DAPI) in PBS-0.05% Tween. The samples were 566 washed in PBS and mounted on glass slides with a mounting solution (20 mM 567 Tris pH8, 0.5 N-propyl gallate, 90% glycerol). 568 569 Monitoring kinetochore MAD1 levels in live recordings 570 Indian muntjac fibroblasts stably expressing Venus-MAD1 or co-expressing 571 Venus-MAD1 and mScarlet-CENP-A were plated on fibronectin coated glass 572 bottom 35 mm FluoroDish (World Precision Instruments; FD35-100), 36-48h 573 before imaging. Before imaging, normal culture media was replaced with 574 Leibovitz’s L15 medium (GIBCO, Life Technologies) supplemented with 10% 575 FBS and the cell-permeable microtubule dye, SiR-tubulin. Live-cell imaging was 576 performed on a temperature-controlled Nikon-Ti microscope equipped at the 577 camera port with a Yokogawa CSU-X1 spinning-disc head with Borealis and an 578 iXon+ DU-897 EM-CCD (Andor). Cells were imaged with an oil-immersion 60x 579 1.4 NA Plan-Apo DIC CFI objective (Nikon, lambda series), yielding a 176 580 nm/pixel sampling, or a 100x 1.4 NA Plan-Apo DIC CFI objective (Nikon, VC 581 series), yielding a 106 nm/pixel. Images were acquired using 488 nm, 561nm and 582 647 nm (Coherent) laser lines and acquisition was controlled by NIS Elements 583 AR software. To ensure that kinetochore could be tracked overtime with a 584 reduced number of optical sections, cells were imaged under modest cell 585 confinement. An image stack was acquired: every 6 seconds, 9 planes separated 586 by 0.5 µm (Figures 1, 6 and Fig. S1C); 1 min, 11 planes separated by 0.5 µm, Fig. 587 S1A, B); 10 seconds, 9 planes separated by 0.5 µm (Figure 3). Fluorescence 588 intensity of kinetochore Venus-MAD1 was manually tracked throughout mitosis 589 using the image analysis software Fiji 2.16.0 79. The recorded image stacks were 590 sum projected and a region was drawn encircling the kinetochore and three 591 different cytoplasmic background sites at each timepoint. MAD1 kinetochore 592 fluorescence intensity (S KT) was calculated by subtracting the averaged 593 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 23 cytoplasmatic signal (S BG), represented in the following equation: S KT, corrected = 594 SKT - S BG, averaged /g3400AKT/ABG S: raw integrated density; A: Area. The calculated 595 signal intensity was normalized to the followings: the clear last peak of the signal 596 decay curve in untreated/control, mock treated and siHAUS6 conditions; the first 597 frame after adding the inhibitor in MPS1i and CDK1i treatments; the average of 598 the first 10 timepoints in nocodazole treated cells. A single-phase exponential 599 decay function was fitted globally across curves polled from all kinetochores in 600 each experimental condition using the equation: Y = (Y0- Plateau) x (exp^(-Kx)) + 601 Plateau, in GraphPad Prism 9 (Boston, Massachusetts USA) . The half decay time 602 (t1/2) for each condition was derived from the fit according to t1/2 = ln(2)/K. To test 603 for differences in the rate of MAD1 decay (K, derived from the global fit) in 604 untreated vs treated cells or large vs small kinetochores between conditions an 605 extra sum-of-squares F-test was used. Quantification of the signal intensity ratio 606 between kinetochore and cytoplasm (Fig. S1B), was performed manually using 607 Fiji 2.16.0. Three regions were drawn encircling the kinetochore (KT), the 608 cytoplasm (cyto) and an area outside of the cell (background signal, BG) at define 609 time points (anaphase onset, 10, 20, 30 and 40 minutes before anaphase). The 610 kinetochore and cytoplasmic fluorescence intensities were calculated by 611 subtracting background signal, represented in the following equations: SKT, corrected 612 = SKT/AKT - S BG/ABG; Scyto, corrected = S cyto/Acyto - S BG/ABG; S: signal; A: Area. The 613 ratio of the corrected kinetochore over cytoplasmatic signals was then calculated. 614 To observe how Mad1 signal spatial distribution within the kinetochore changed 615 throughout time kymographs were produced from live recordings of IM cells 616 (Figure 3). Image stacks were sum projected using Fiji 2.16.0 and kymographs 617 were generated with a custom-written routine in MATLAB (Mathworks, Natick, 618 MA) that compensates for spindle rotation and translations previously described 619 in 64. 620 621 Cell confinement 622 For cell confinement, we adopt a cell confiner as previously described 39, 62 using 623 a custom-designed polydimethylsiloxane (PDMS, RTV615, GE) layout to fit a 624 35mm diameter fluorodish. A suction cup was custom-made with a 10:1 mixture 625 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 24 (w/w PDMS crosslinker A/ B) and baked on an 80 ºC hot plate for 30mins and left 626 to dry over-night before unmolding. A PDMS confinement slide was then 627 prepared with a 10:1 mixture (w/w PDMS crosslinker A/ B) molded into 628 micropillars with a height of 8 μ m (using a SU-8 wafer mold) polymerized onto a 629 10 mm round coverslip and baked for 95 °C for 15 min. The PDMS confinement 630 slide was attached to the PDMS suction cup and connected to a vacuum 631 generator (AF1-dual, Elveflow). 632 633 Quantification of MAD1 intra-kinetochore mobility via fluorescence recovery after 634 photobleaching (FRAP) 635 For FRAP experiments, IM cells stably expressing Venus-MAD1 were plated on 636 fibronectin coated 35mm FluoroDish (World Precision Instruments, FD35-100) 637 36-48 h before imaging. To elicit uniform and maximal kineotchore-MAD1 638 occupancy, cells were treated with nocodazole for 30 minutes before the 639 experiment. To ensure that the bleached kinetochore could be tracked overtime 640 with a reduced number of optical sections, an agar overlay was placed on top of 641 the cells decreasing the cell volume, as described in 63. Briefly, a 170 μ m thick 642 layer of 2% low-melting-point agarose in L15 medium supplemented with 10% 643 FBS was prepared in advance and incubated for 30 minutes with pre-warmed 644 L15 + 10% FBS containing nocodazole. Immediately before imaging, a small 645 piece of the agar layer was gently placed over the cells that were maintained in a 646 minimal amount of medium to prevent cell dehydration while ensuring 647 compression. FRAP experiments were conducted on a Leica Scanning Confocal 648 STELLARIS 8 FALCON (Leica Microsystems, Germany), equipped with an 649 incubation chamber and a power HyD S detector. Cells were imaged with a 650 water-immersion 86x 1.20 NA HC PL APO STED white objective (Leica 651 Microsystems, Germany), plus a 4.5x zoom, yielding an 86 nm/pixel sampling. 652 The 488-nm emission wavelength of a white Light Laser (WLL) was used to 653 bleach MAD1 signal from either part of or an entire kinetochore with 2 pulses of 654 0.3 seconds each. An image stack of 3 z-planes at a 0.5 μm interval was acquired 655 with the following routine: 3 pre-bleach frames, every 1.3 seconds; kinetochore 656 bleaching; post-bleach 5 frames, every 1.3 seconds for 5 frames, 10 frames, 657 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 25 every 2 seconds and then every 3 seconds for up to 55 seconds. The selected 658 representative images show sum-intensity projections of the image stacks. 659 Venus-MAD1 fluorescence at the kinetochore was manually tracked on the sum 660 projected image stacks using Fiji 2.16.0and a custom-designed script in Python 661 3.13. Four rectangular ROIS were defined corresponding to: (i) the bleached 662 region of the experiment kinetochore; (ii) the non-bleached region of the 663 experiment kinetochore; (iii) the unperturbed kinetochore from the pair (sister KT) 664 and (iv) a kinetochore from an unperturbed pair (control KT). Additionally, two 665 square ROIs with 0.86 μm sides were drawn over the cytoplasm to measure 666

fluorescence (BG). A custom-designed script was used to draw the 667 kinetochore ROIs with a constant width of 0.6 μm, and length defined by the 668 distance of the kinetochore extremities in the pre-bleach frame (manually marked 669 based on MAD1 signal using Fiji 2.16.0). In the case of the partially ablated 670 kinetochore the length of the non-bleached region was defined at the first frame 671 post-bleaching, and the length of the bleached region was defined as the 672 difference between the total kinetochore length and the length of the 673 non-bleached region. The long axis of the rectangle was defined based on the 674 vector described by the points marking the extremities of the non-bleached signal 675 and was adjusted at every time point to track the kinetochore motion. MAD1 676 kinetochore fluorescence intensity (SKT, corrected) was corrected for fluctuations in 677 the cytoplasmatic fluorescent signal as described in the following equation: S KT, 678 corrected = SKT / (a x time + b); SKT: kinetochore MAD1 raw integrated density; a, b: 679 constants provided by a linear fit of the cytoplasmatic MAD1 raw integrated 680 density. The corrected kinetochore signal was then normalized to the average of 681 the 3 pre-bleach values (SKT, normalized) and fit to a single exponential association 682 function: Y = Y0 + (Plateau – Y0) x (1-(exp^(-Kx))); Y0 value was constrained to 683 be the SKT, corrected at the first frame post-bleaching. Half recovery time (t 1/2) was 684 derived from the fit according to t1/2 = ln(2)/K. Fluorescence recovery percentage 685 (Recovery) was calculated according to Recovery = (Plateau – Y0) / (1 – Y0). 686 Fitting and downstream statistical analysis was performed using GraphPad Prism 687 9 (Boston, Massachusetts USA ) . Cells where bleaching efficiency was < 50%; 688 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 26 focal plane changes were > 1 μ m; or where the normalized MAD1 fluorescence 689 could not be fitted to a single exponential (if the 95 % confidence interval for any 690 of the parameters could not be determined) were excluded from the analysis. 691 Line intensity profiles showing Venus-MAD1 signal along the kinetochore were 692 measured from a ROI covering the entire kinetochore length and a width of 0.250 693 μm at selected times. 694 695 Quantification of kinetochore MAD1 signal distribution with Stimulated Emission 696 Depletion (STED) Microscopy 697 Indian muntjac fibroblasts stably expressing Venus-MAD1 were seeded on glass 698 coverslips and processed as described in the “Immunofluorescence” section. 3D 699 image stacks were acquired using an Abberior 'Expert Line' gated-STED 700 microscope, equipped with a Nikon Lambda Plan-Apo 1.4NA 60x objective lens. 701 CH-STED mode was implemented as described before 62. Images were acquired 702 using excitation wavelengths at 488 nm, 561 nm, and 640 nm and a single 703 doughnut-shape depletion beam at 775 nm, a 0.8 Airy unit pinhole and the STED 704 channel had a time-gate threshold of 500 ps. Pixel size was set to 30 nm in xy 705 and 200 nm in z and Image stacks covering near full spindle volume (~8 μm). The 706 selected representative images show sum-intensity projections of a fraction of 707 the Z-slices (5-10). In images that show merged channels, histograms were 708 cropped to allow visualization of all structures. Venus-MAD1 and tubulin signal 709 distribution was analyzed along individual kinetochores in sum projected 3D 710 stacks (5-10 slices) using Fiji 2.16.0. A line of 0.6 µm width was drawn along the 711 longitudinal axis of the kinetochore (defined by the ACA signal). Intensity line 712 profiles were extracted along the kinetochore for both MAD1 and tubulin 713 channels and background subtracted (average cytoplasmatic intensity value). 714 Corrected signal intensity was normalized to the maximum intensity value within 715 each kinetochore. To visualize and measure correlations between MAD1 and 716 tubulin signals, we calculated the intensity z-scores, standardized internally (for 717 each kinetochore). Each position along the kinetochore is represented as a point 718 in a 2D space where the z-scores are the x-y axes and an ensemble of 719 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 27 kinetochores was either overlaid in the same scatter plot or displayed and 720 calculated for isolated kinetochores. The product zMAD1.ztubulin at each data 721 point is the local contribution to an overall correlation. Averaging over all data 722 points yields the covariance normalized to the standard deviation product, known 723 as the Pearson correlation coefficient. The Matlab function corrcoef was used, 724 which calculates Pearson's correlations and p-values with a two-tailed Student 725 t-test for zero correlation. 3D reconstructions of the image stacks were obtained 726 with Imaris 9.3.1 (Oxford Instruments). 727 728 Laser microsurgery and correlative live-cell confocal and fixed-cell CH-STED 729 microscopy 730 Indian muntjac cells expressing 2x-GFP-CENP-A and GFP-CENTRIN-1 and 731 mScarlet-MAD1 were seeded on fibronectin coated 25 mm, no. 1.5, round 732 coverslips (CAT 10593054, fisher scientific) 48h prior to the experiment. To 733 induce mScarlet-MAD1 expression, cells were incubated with 1 ug/mL of 734 doxycycline for 24h. Before imaging the coverslips were mounted on a 35 mm 735 Attofluor™ Cell Chamber (CAT A7816, Invitrogen) and the media replaced with in 736 Leibovitz’s L15 medium (GIBCO, Life Technologies) supplemented with 10% 737 FBS. Cells were imaged using Nikon-Ti spinning-disk confocal microscope with a 738 100x 1.4 NA oil objective as detailed in the “Monitoring kinetochore MAD1 levels 739 in live recordings” section. Cells in metaphase were identified by having MAD1 740 negative kinetochores aligned at the equator and increased inter- kinetochore 741 and pole-pole distance (assessed by GFP-CENTRIN-1). After recording a 742 pre-surgery image stack, one of the large kinetochores (from the IM chromosome 743 3+X) was partially ablated with a 532 nm laser, controlled by a custom routine. A 744 detailed description of the microsurgery setup can be found in 80. 2-5 consecutive 745 pulses with a 0.35 μ m step and a 12 Hz repetition rate) were applied. The pulse 746 width was 10 ns and the pulse energy was 3.9–4.4 μ J. After surgery an image 747 stack (5 planes, 0.75 μm z-step) was recorded every minute until MAD1 signal 748 was detected at the kinetochore, for up to 30 min; after MAD1 detection the 749 acquisition routine was changed: 7 planes, 0.5 μm z-step every 20 seconds for 750 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 28 5 min. As a control for uniform MAD1 signal distribution, non-ablated cells were 751 imaged upon acute nocodazole treatment (5 to 60 min incubation). For 752 correlative live-cell confocal and fixed CH-STED microscopy analysis, once 753 MAD1 signal was detected at the kinetochore a fixation solution was added to the 754 imaging chamber. The location of the cell of interest was marked with a 755 laser-engraved pattern on the glass coverslip. The sample was processed for 756 staining as described in the “Immunofluorescence” section. After locating the 757 surgery cell with a 10x objective, a super-resolved image stack of the spindle was 758 acquired with CH-STED as described in the “ Quantification of kinetochore MAD1 759 signal distribution with Stimulated Emission Depletion (STED) Microscopy” 760 section (pixel size 35 nm in xy and 150 nm in z). Representative images show a 761 sum (live-cell data) or max-intensity (fixed-cell data) projection of all z-slices that 762 contain the kinetochore of interest. Histograms were adjusted per frame to allow 763 better visualization of the representative phenotypes. To measure MAD1 764 recruitment after laser microsurgery, kinetochores were manually tracked using 765 the GFP-tagged CENP-A as a reference in Fiji 2.16.0. kinetochore deformation 766 was defined as the first time point when an increase in the angle between the line 767 connecting the ends of the surgery kinetochore and the line connecting the ends 768 of the unperturbed kinetochore from the pair (sister kinetochore) was observed. 769 MAD1 recruitment was defined as the first time point when mScarlet-MAD1 770 signal was visible at the sister kinetochore. MAD1 signal distribution along the 771 sister kinetochore was measured at the time point when signal intensity appeared 772 highest, based on visual assessment. On the cells where correlative live-cell 773 spinning-disk confocal and CH-STED microscopy was performed, tubulin signal 774 distribution was measured in the fixed sample. Measurements were performed 775 on sum projected stacks containing the kinetochore pair of interest. A segmented 776 line of 0.8 μm width (for MAD1, live imaging) or 0.9 μm width (for tubulin, fixed cell 777 imaging) was drawn to span the full length of the kinetochore, as defined by the 778 CENP-A signal, starting from the end that retained its counterpart (or randomly in 779 nocodazole treated cells). Intensity line profiles were extracted along the 780 kinetochore for all channels and background subtracted (cytoplasmatic signal 781 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 29 near the kinetochore). The corrected intensity profile was normalized to the 782 maximum within each kinetochore, as exemplified for CENP-A: S CENP-A, norm = 783 (SCENP-A - S BG CENP-A ) / max (S CENP-A - S BG CENP-A ), S: mean pixel intensity. To 784 correct for intensity variations caused by kinetochore deformation or movement, 785 intensity profiles are shown as a ratio of MAD1/tubulin over CENP-A. Kinetochore 786 length was normalized to a relative scale ranging from 0 (start of the line profile) 787 to 1 (end of the line profile). The average MAD1/CENP-A ratio intensity profile for 788 each condition (Nocodazole vs Surgery), was calculated using the ‘Multiple 789 average curves’ function with linear interpolation on OriginPro 2022 (OriginLab 790 Corporation, Northampton, MA, USA). The coefficient of variation of the 791 MAD1/CENP-A ratio was calculated as the standard deviation of the 792 MAD1/CENP-A intensity values along each kinetochore, divided by their mean. 793 Cells that failed to recruit MAD1 within the mean recruitment time + 1 s.d. (~12 794 minutes) or exhibited only very faint MAD1 signal were excluded from the MAD1 795 distribution analysis. Statistical analysis and plots were done using GraphPad 796 Prism 9 (Boston, Massachusetts USA) . 797 798 Quantification of kinetochore MAD1 levels in prometaphase and metaphase 799 HAUS6 depleted cells 800 Control and HAUS6 depleted cells expressing Venus-Mad1 were treated with MG 801 132 for 1h and then fixed and stained for as detailed in the “Immunofluorescence 802 for wide-field” section. Samples were imaged on AxioImager Z1 equipped with a 803 CCD camera (ORCA-R2, Hamamatsu) operated by Zen software (Carl Zeiss, 804 Inc.). 3D image stacks covering the entire cell volume (0.23 μ m z-step) were 805 acquired using a 100× Plan-Apochromatic oil differential interference contrast 806

lens, 1.46 NA (Carl Zeiss Microimaging Inc.) and excitation 807 wavelengths at 488 nm, 561 nm, and 640 nm. The number of cells with MAD1 808 positive kinetochores was visually assessed for each condition. Kinetochore 809 Venus-MAD1 signal intensity was measured on a single focal plane using Fiji 810 2.16.0. Cytoplasmatic Venus-MAD1 signal was averaged from 3 regions outside 811 the spindle area. The ration between kinetochore and cytoplasm signal was 812 calculated according to the equation: S KT / S cytoplasm; S: mean pixel intensity. 813 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 30 Statistical analysis and plots were done using GraphPad Prism 9 ( Boston, 814 Massachusetts USA) . 815 816 Cell tracking and daughter fate analysis in live recordings 817 Indian muntjac fibroblasts stably expressing GFP-H2B were plated on fibronectin 818 coated µ-Slide 4 Well Glass Bottom (Ibidi GmbH, 80427) with Ham’s F-10 media 819 supplemented with 20% FBS 24h before imaging. Cells were recorded using the 820 Zeiss LSM 980 Airyscan confocal microscope built on an Axio Observer 7 821 inverted stand with a motorized piezo stage and active focus stabilization. A 822 controlled atmosphere chamber (CO2 and temperature) was used. Objective 823 lens was a dry 0.3NA 10x Plan Neofluar. Detection was performed with two 824 GaAsP detectors. Image acquisition was carried out with ZEN 3.11 software. 3D 825 image stacks (pixel size 830 nm in xy, 4 planes separated by 6.7 μm, pinhole Airy 826 unit of 1.23, scan speed of 7, no averaging) were recorded every 7 min using both 827 transmitted light and an excitation wavelength at 488 nm. In experiments where 828 mitotic delay was induced through monastrol treatment, cells were imaged in the 829 presence of the drug for 6h before washing out the drug. The fate of cell families 830 was tracked for 72h. In experiments where HAUS6 was depleted, cells were 831 treated with control or HAUS6 siRNAs shortly before imaging. Cell families were 832 then tracked for 96h. Cells were tracked manually using TrackMate 81. Single-cell 833 lineages and cell cycle durations (frame of anaphase mother cell to frame 834 anaphase daughter cell) were analysed by modifying previously published 82 835 Jupyter notebooks (7.0.8). We defined mitotic duration as the frame of mitotic 836 onset (condensing chromatin) until the first frame of anaphase (clearly separated 837 daughter chromatin). For each division event, we determined mitotic defects and 838 errors, and the fate of the daughters. For downstream analysis, filtering and 839 plotting, all data were integrated using Jupyter notebooks (Python 3.12) and the 840 pandas (2.2.2) 83 and Altair (5.4.1) 84 libraries. Fiji/ImageJ 79 macros and 841 analytical code for the daughter fate analysis is available at 842 https://github.com/TobiasKletter/MitoticStop. 843 844 Co-immunoprecipitation 845 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 31 Indian muntjac fibroblasts stably expressing GFP-H2B were grown on with Ham’s 846 F-10 media supplemented with 20% FBS. To obtain a mitotic arrested population 847 cells were treated for 16h with 3.3 μ M Nocodazole. 10 million mitotic cells were 848 collected by shake-off. The remaining interphase cells (NOC interphase) and 849 cells left untreated (asynchronous) were collected by cell scrapping. Cells were 850 pelleted at 300 g for 5 minutes, washed with cold PBS and cell pellets were snap 851 frozen in liquid nitrogen. Co-immunoprecipitation assay was optimized from the 852 protocol described in 29. Cell were resuspended in lysis buffer (20 mM Tris/HCl, 853 pH 7.5; 50 mM NaCl; 0.5% Triton X-100; 5 mM EGTA; 1 mM dithiothreitol; 2 mM 854 MgCl2; Pierce Protease and Phosphatase inhibitor mini tablets, EDTA-free 855 [Thermo Fisher Scientific; A32961]) and incubated for 45 minutes at 4ºC with 856 agitation. Cell lysates were centrifuged at 15 000 rpm for 5 minutes at 4ºC to 857 collect the supernatant. Protein concentration was quantified and a lysate 858 solution containing 1.5 mg of protein in 1 mL of lysis buffer was prepared for each 859 condition. Lysates were incubated with 3 ug of anti-53BP1 antibody (Novus, 860 NB100-304) or control rabbit IgG for 2h at 4ºC with agitation and subsequently 861 with 20 μ L of pre-washed (4x with lysis buffer) Pierce™ Protein A Magnetic Beads 862 (Thermo Fisher Scientific, 88845) for 1h at 4ºC with agitation. The non-binding 863 supernatant faction was collected and the beads were washed 4x with lysis 864 buffer. The bound fraction was eluted from the beads by resuspending in 2x 865 sample buffer (50 /i2 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 1% 866 β -mercaptoethanol, 12.5 /i2 mM EDTA, and 0.02% bromophenol blue) and 867 denatured at 95/i2 °C for 5/i2 min. Samples were analyzed through western blot. 868 869 Western blotting 870 Protein samples of 13-20 ug in sample buffer were loaded on a 4–15% 871 Mini-PROTEAN® TGX™ Precast Protein Gel (Bio-Rad, 4561084), mounted on a 872 Mini-PROTEAN vertical electrophoresis apparatus (Bio-Rad). NZYColour Protein 873 Marker II (NZYTech) was used as a ruler. Blotting was performed with a 874 Transfer-Blot Turbo transfer system (Bio-Rad) onto 0.2 µm Nitrocellulose 875 membranes. Membranes were blocked in blocking buffer (5% powder milk in 876 PBS Tween 0.1%) for 1h. Anti-53BP1 rabbit 1:1000 (Novus, NB100-304), 877 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 32 anti-USP28 rabbit (Abcam, ab126604), anti-p53 DO-1 mouse (Santa Cruz 878 Biotechnology, SC-126) and anti-GAPDH mouse 1:10 000 (Proteintech, 879 60004-1-Ig) primary antibodies were diluted in blocking buffer and incubated 1 h 880 (4/i2 °C) with agitation. Anti-rabbit HRP (Jackson ImmunoResearch, 111-035-003) 881 and anti-mouse light chain specific HRP (Jackson ImmunoResearch, 882 115-035-174) secondary antibodies were diluted 1:5000 in blocking buffer and 883 incubated 1 h (4 /i2 °C) with agitation. Membranes were developed with Clarity 884 Western ECL Blotting Substrate (Bio-Rad, 1705061) and detected in the 885 Chemidoc XRS system and Image Lab software (Bio-Rad). 886 887 888 889 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 33

890 We would like to thank all colleagues that generously provided reagents used in 891 this study, and Paula Sampaio and Maria Azevedo from i3S Advanced Light 892 Microscopy Facility for technical support. We are indebted to Franz Meitinger for 893 critical discussions and technical advice on the mitotic stopwatch. We also thank 894 CID Lab members for discussions and constructive feedback throughout the 895 course of this project, with special thanks to Daniela Andrade and Rui Martins for 896 their strong contribution during mitotic shake-offs. NO, JS-O and EW were 897 respectively supported by fellowships SFRH/BPD/118126/2016, 898 SFRH/BD/07730/2021 and 2022.13008.BD from Fundação para a Ciência e a 899 Tecnologia of Portugal. This project was funded by The European Union through 900 the European Research Council grant KAREVO (project 101140624) to HM. 901 902 Author contributions 903 Conceptualization: HM 904 Methodology: NO, JS-O, TK, AJP , EW 905 Investigation: NO, JS-O, TK 906 Visualization: NO, JS-O, TK, AJP 907 Funding acquisition: HM 908 Project administration: HM 909 Supervision: HM, AJP 910 Writing – original draft: NO, HM 911 Writing – review & editing: NO, JS-O, TK, AJP, EW, HM 912 913 Competing interests 914 Authors declare that they have no competing interests. 915 916 917 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 34 Figure legends 918 919 Figure 1. SAC protein MAD1 is gradually removed from the kinetochore in a 920 microtubule and mitotic kinases dependent manner. 921 (A) Representative live-cell recording images of Indian muntjac fibroblasts 922 expressing Venus-MAD1 (orange) and with SiR-tubulin labeled microtubules 923 (grey). Scale bar: 5µm. 924 Insets highlight an early congressing kinetochore pair (dashed square) and 925 dashed circles show polar kinetochores/chromosomes. Scale bar: 5µm. 926 (A’) Insets focusing on 1 half-spindle to highlight the congression of polar 927 chromosomes (dashed circles) from A. Scale bar: 5µm. 928 (A’’) Kymographs depicting Venus-MAD1 signal of the 2 polar chromosomes 929 shown in A’. The closest pole position is shown by the dashed line. Leading 930 (black asterisk) and trailing kinetochore (white asterisk) and indicated for each 931 pair. Scale bar: 5µm. 932 (B) Representative live-cell recording images of Indian muntjac fibroblasts 933 acutely treated with Nocodazole, MPS1 and CDK1 inhibitors. Venus-MAD1 934 (orange) and SiR-tubulin labeled microtubules (grey). Insets highlight a 935 kinetochore pair. Scale bar: 5µm. 936 (C) Normalized Venus-MAD1 fluorescence decay IM cells untreated (control) or 937 treated with Nocodazole, MPS1 and CDK1 inhibitors. Data represented as mean 938 ± 95% confidence interval (CI; Control n = 53 kinetochores, 7 cells, green; 939 Nocodazole n = 31 kinetochores, 8 cells, grey; MPS1i n = 63 kinetochores, 12 940 cells, yellow; CDK1i n = 18 kinetochores, 10 cells, blue; data pooled from at least 941 five independent experiments per condition). 942 (D) Normalized Venus-MAD1 fluorescence decay in large or small kinetochores 943 from untreated cells. Data are represented as mean ± 95% CI. (Large KTs n = 19 944 kinetochores, 7 cells, 5 independent experiments, dark green; small KTs n = 34 945 kinetochores, 5 cells, 4 independent experiments, light green). 946 (E) Summary table of the parameters extracted from a single exponential fitting 947 of Venus-MAD1 fluorescence decay across the condition shown in C and D. 948 P-values were calculated using extra sum-of-squares F-test comparing the rate 949 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 35 of MAD1 decay (K, derived from global fit) in untreated vs treated cells or large 950 vs small kinetochores. 951 952 Figure 2. MAD1 is immobile within kinetochores 953 (A) Schematic representation of Dynein-mediated MAD1/MAD2 complex 954 removal from kinetochores. 955 (A’) Predicted curves of MAD1 fluorescence recovery after partial kinetochore 956 bleaching assuming MAD1 to be mobile (left) or immobile (right) within the 957 kinetochore. Lines represent the fluorescence levels from the region of the 958 kinetochore that was bleached (light blue) or left unperturbed (dark blue). 959 (B), (C) Representative images of Indian muntjac cells pre- and post-bleach of 960 part (B) or an entire kinetochore (C). Insets: kinetochore pair that was perturbed. 961 Dashed line boxes indicate the ablated kinetochore region (partial bleach KT: 962 non-bleached region - dark blue; bleached region – blue; total bleach KT - 963 yellow). 964 (B’), (C’) Venus-Mad1 signal intensity profile along the kinetochore at pre-bleach 965 and at selected times post-bleach for kinetochores shown in B and C, 966 respectively. 967 (B’’), (C’’) Normalized Venus-Mad1 fluorescence recovery curves for cells shown 968 in B and C, respectively (Dark blue line non-bleached region; light blue bleached 969 region; yellow line totally bleached kinetochore; black and dark grey lines 970 unperturbed kinetochores from the bleach or a control kinetochore pair, 971 respectively; light grey cytoplasm; from fit curves are shown by the 972 semi-translucid line). 973 (D) Kinetochore Venus-MAD1 fluorescence recovery after partial or total 974 kinetochore bleach (Partial KT bleach n = 7; total KT bleach n = 8 from at least 975 two independent experiments). Dots show individual kinetochores. P = 0.2319, 976 Mann-Whitney test (two-tailed). 977 (E) Half recovery times of kinetochore Venus-MAD1 fluorescence after partial 978 or total kinetochore bleach. Dots show individual kinetochores. P = 0.5358, 979 Mann-Whitney test (two-tailed). 980 981 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 36 Figure 3. MAD1 removal along an individual kinetochore is not uniform 982 (A) Live-cell recording images of Indian muntjac fibroblasts stably co-expressing 983 Venus-MAD1 (orange) and mScarlet-CENP-A (blue) with SiR-tubulin labeled 984 microtubules (grey). Inset: kinetochore pair that shows non-uniform loss of 985 MAD1 signal. Dashed arrow indicates the kinetochore depicted in B. Scale bar: 986 5µm. 987 (B) Kymographs depicting Venus-MAD1 and mScarlet-CENP-A signal 988 distribution along the longitudinal axis of the kinetochore shown in A. Each 989 channel is shown on a greyscale and as a heatmap. Scale bar: 1µm 990 991 Figure 4. MAD1 removal along individual kinetochores correlates with local 992 microtubule density 993 (A) Representative CH-STED image of an Indian muntjac fibroblasts with 994 partially attached kinetochores. Microtubules (grey; STED), centromeres (ACA; 995 blue; STED) and Venus-MAD1 (orange; confocal). Insets: partially attached 996 kinetochore. Scale bar: 2µm. 997 (B) Three-dimensional reconstruction of the cell shown in A (different 998 perspective). Inset: partially attached kinetochore. The kinetochore pair of 999 interest is traced by a dashed line. Microtubules (grey), centromeres (ACA; blue) 1000 and Venus-MAD1 (orange). Scale bars, Left: 3µm; Right: 1µm. 1001 (C) Max normalized Venus-Mad1 (orange) and tubulin (microtubules; grey) 1002 signal intensity profile along the longitudinal axis of the kinetochore highlighted 1003 in A (arrowhead). 1004 (D) Pearson correlation coefficient between MAD1 and microtubules signal 1005 intensity along the kinetochore. Bars show individual partially attached 1006 kinetochores (n=21 kinetochores, 7 independent experiments). 1007 E) Representative distribution of kinetochore sub-regions showing positive, 1008 negative and no correlation between MAD1 (y-axis) and microtubules (x-axis) 1009 intensity z-scores. Dots represent distinct positions along the “kinetochore” 1010 shown below the graphs. NE and SW quadrants contribute positively, while NW 1011 and SE quadrants contribute negatively to the Pearson correlation coefficient (r). 1012 (F) Correlation map of MAD1 and microtubule intensity z-scores for the 1013 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 37 kinetochores shown in D. The underlying scatterplot contains data from multiple 1014 kinetochores, with each dot representing a sub-kinetochore region. For 1015 visualization, the scatterplot was converted into a regional density 1016 map. P-value was calculated with Student t-test for zero correlation (two-sided). 1017 1018 Figure 5. Local perturbation of kinetochore structure results in a localized 1019 SAC response that inversely correlates with microtubule occupancy 1020 (A) Schematic summary of the kinetochore response to partial laser 1021 microsurgery ablation. 1022 (B) Representative live-recording images of Indian muntjac cells expressing 1023 2x-GFP-CENP-A and GFP-CENTRIN-1 (both in blue) and mScarlet-MAD1 1024 (orange/ grey scale) upon partial kinetochore surgery or acute Nocodazole 1025 treatment. Insets: kinetochore pair that was perturbed. Arrows point to the 1026 ablated kinetochore region, pre- and post-surgery (solid and hollow arrow, 1027 respectively). Arrowheads show MAD1 recruitment. Scale bar: 2µm. 1028 (B’) Max normalized line intensity profile plot of MAD1/CENP-A signal ratio along 1029 the longitudinal axis of the kinetochores highlighted in A (surgery cell in orange; 1030 nocodazole cell in grey). 1031 (C) Dot plot showing the time from surgery to observable deformation (bending; 1032 n= 12 kinetochores) and MAD1 signal re-appearance (n = 14 kinetochores) at 1033 the sister kinetochore in cells where SAC recruitment was observed. Each dot 1034 represents an individual kinetochore from a different cell, recorded during 13 1035 independent experiments. P <0.0015, Wilcoxon matched-pairs signed rank test. 1036 (D) Max normalized MAD1/CENP-A ratio intensity profile along the longitudinal 1037 axis of kinetochores after partial surgery of their sisters (orange) or upon acute 1038 Nocodazole treatment (NOC; grey). Thin lines show individual kinetochores and 1039 thick lines show average curves (surgery n = 11 kinetochores/cells, 10 1040 independent experiments; Nocodazole n = 19 kinetochores/cells, 2 independent 1041 experiments). The graph depicts a cohort of cells that recruit MAD1 to the 1042 proximal tip of the kinetochore, relative to the surgery region (n = 11/16 MAD1 1043 recruiting kinetochores). 1044 (E) Coefficient of variation of MAD1/CENP-A signal ratio along kinetochores 1045 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 38 after partial surgery of their sisters. Dots represent individual kinetochores 1046 pooled from all cells where MAD1 recruitment was observed (surgery n = 16 1047 cells, 14 independent experiments; Nocodazole n = 19 cells, 2 independent 1048 experiments). P <0.0001, Mann-Whitney test. 1049 (F), (F’) Representative images from correlative live cell spinning disk (F) and 1050 fixed CH-STED microscopy (F’). Indian muntjac cells expressing 1051 2x-GFP-CENP-A and GFP-CENTRIN-1 (both in blue) and mScarlet-MAD1 1052 (orange/grey scale) were fixed after partial kinetochore surgery and stained to 1053 visualize microtubules (grey). Insets: kinetochore pair that was perturbed. 1054 Arrows point to the ablated kinetochore region, pre- and post-surgery (solid and 1055 hollow arrow, respectively). Arrowheads show MAD1 recruitment. Scale bar: 1056 2µm. 1057 (F’’) Max normalized line intensity profile plot of MAD1/CENP-A (live cell data 1058 just before fixation, orange) and microtubules/CENP-A (fixed cell data, grey) 1059 signal ratio along the longitudinal axis of the kinetochore highlighted in F and F’. 1060 1061 Figure 6. Partial microtubule occupancy at kinetochores after Augmin 1062 depletion delays MAD1 removal and SAC silencing 1063 (A) Representative live-cell recording images of Indian muntjac fibroblasts mock 1064 or HAUS6 siRNA treated. Venus-MAD1 (orange) and SiR-tubulin labeled 1065 microtubules (grey). Scale bar: 5µm.Insets highlight a kinetochore pair. Asterisks 1066 indicate the two kinetochores from the pair. Scale bar: 5µm. 1067 (B) Normalized Venus-MAD1 fluorescence decay from A. Data from at least five 1068 independent experiments are represented as mean ± 95% confidence interval 1069 (Mock n = 48 kinetochores, 10 cells, grey; siHAUS6 n = 47 kinetochores, 10 1070 cells, blue). 1071 (C) Summary table of the parameters extracted from a single exponential fitting 1072 of Venus-MAD1 fluorescence decay across the conditions shown in B. P-values 1073 were calculated using extra sum-of-squares F-test comparing the rate of MAD1 1074 decay (K, derived from global fit) in mock vs siHAUS6 treated cells. 1075 (D) Percentage of cells with at least one MAD1 positive kinetochore in mock 1076 (grey) or siHAUS6 (blue) treated cells after 1h of MG132 arrest. Bars show 1077 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 39 mean, error bars show s.d. from three independent experiments (n/g3410293 cells per 1078 condition). P <0.0001, prometaphase; P=0.0008, metaphase; unpaired t test 1079 (two-sided). 1080 (E) Dot plot showing the ratio between kinetochore and cytoplasmatic 1081 Venus-MAD1 signal intensity in mock (grey) or siHAUS6 (blue) treated cells 1082 fixed after 1h of MG132 arrest. Dots show individual kinetochores, bars show 1083 mean, error bars show s.d. from three independent experiments (n /g3410309 1084 kinetochores, /g341017 cells per condition). P<0.0001, Mann-Whitney test 1085 (two-sided). 1086 (F) Representative CH-STED image of an Indian muntjac control or HAUS6 1087 depleted cells treated with MG132 for 1h. Microtubules (grey; STED), 1088 centromeres (ACA; blue; STED) and Venus-MAD1 (orange; confocal). Inset: full 1089 (control) or partially (siHAUS6) attached kinetochore pair. Scale bar: 2µm. 1090 1091 Figure 7. Delayed SAC silencing due to partial microtubule occupancy at 1092 kinetochores activates the mitotic stopwatch to stop cell proliferation 1093 (A) Schematic summary of the experimental setup and downstream analysis. 1094 (A’) Mitotic duration of Indian muntjac cells expressing H2B-GFP with decreasing 1095 HAUS6 levels (time in the presence of control or HAUS6 siRNA; grey and blue, 1096 respectively). Data pooled from three independent experiments (siHAUS6 n = 1097 370, siCtrl n = 335). Circles denote mean in each time bin (bin size /i2 =/i2 12h), 1098 error bars show the s.e.m. 1099 (B) Bar graph showing the percentage of Indian muntjac cells that exited mitosis 1100 with a particular mitotic defect/ error at decreasing HAUS6 levels. Data pooled 1101 from three independent experiments (n = 469 cells). Bars show mean in each 1102 time bin (bin size/i2 =/i2 6h or 24h), error bars show the s.e.m. 1103 (C) Plot of the mitotic cell fate as a function of mitotic duration in Indian muntjac 1104 cells treated with siHAUS6 for 0 to 96h. Bars show individual cells (n = 469 cells 1105 pooled from three independent experiments). 1106 (D) Plot of daughter cells fate (Gen 2) as a function of mother cell mitotic 1107 duration (Gen 1). Only mother cells from the selected cohort (12-36h of siHAUS6 1108 treatment) are shown. Bars show individual daughter cells (n = 272 cells pooled 1109 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 40 from three independent experiments). 1110 (E) Bar graph showing the percentage of mother cells from the selected cohort 1111 that has 0, 1 or 2 daughter cells dividing at least once, as a function of mother 1112 cells time in mitosis. Only mother cells from the selected cohort (12-36h of 1113 siHAUS6 treatment) and that completed mitosis with no defects are shown. Data 1114 are binned ( 120 min n = 10 cells pooled from three independent 1116 experiments). 1117 (G) Analysis of 53BP1 immunoprecipitates from Indian muntjac fibroblast 1118 populations treated with nocodazole (3.3 µM) for 16 h. Cells were separated into 1119 interphase and mitotic cells via mechanical shake-off. As negative control, rabbit 1120 IgG were used. GAPDH as loading control. 1121 (H) Immunoblot on whole-cell lysates from human RPE1 hTERT cells and Indian 1122 muntjac fibroblasts (GFP-H2B), that were treated for 16h either with DMSO, 1123 etoposide (5 µM), or nocodazole (3.3 µM, mitotic shake-off) and probed for p53 1124 enrichment. GAPDH as loading control. 1125 1126 Figure 8 - Microtubule occupancy at kinetochores integrates with the 1127 mitotic stopwatch to ensure faithful daughter cell proliferation 1128 (Top) Augmin-mediated K-fiber maturation promotes maximal microtubule 1129 occupancy at kinetochores through gradual and highly localized MAD1/MAD2 1130 removal, resulting in timely SAC silencing. In this way, mitotic timing is optimal, 1131 fidelity is maintained, and daughter cells continue proliferating. 1132 (Bottom) When K-fiber maturation is impaired, for example through disruption of 1133 Augmin-mediated microtubule amplification, microtubule occupancy at 1134 kinetochores remains partial and SAC silencing is delayed. Eventually, the cell 1135 divides either after achieving full occupancy or while kinetochores are still only 1136 partially occupied. The resulting mitotic delay is registered by the mitotic 1137 stopwatch, which bookmarks daughter cells to halt proliferation. 1138 1139 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 41 Supplemental Figure legends 1140 1141 Figure S1. MAD1 is undetectable at kinetochores 20 minutes before 1142 anaphase 1143 (A) Representative live-cell recording images of Indian muntjac fibroblasts 1144 expressing Venus-MAD1 (orange, upper panels or grey, lower panels) and 1145 mScarlet-CENP-A with SiR-tubulin labeled microtubules (grey, upper panels). 1146 Scale bar: 5µm. 1147 (B) Dot plot showing the ratio between kinetochore and cytoplasmatic 1148 Venus-MAD1 signal intensity at selected time points. Dots show individual 1149 kinetochores, lines show mean ± s.d. (-40 min: 87 kinetochores, p< 0.0001; -30 1150 min: 123 kinetochores, p 0.9999; -10 1151 min: 126 kinetochores, p> 0.9444; Anaphase onset: 126 kinetochores; 15 cells 1152 from 7 independent experiments)., P value calculated with Kruskal-Wallis test 1153 with Dunn’s comparisons of mean fluorescence in each time point with 1154 anaphase onset. 1155 (C) Normalized signal intensity of cytoplasmatic Venus-MAD1 throughout live 1156 cell recording. Lines show selected cytoplasmatic regions (n= 27 spots, 9 cells, 6 1157 independent experiments). 1158 1159 Figure S2. MAD1 tracking on individual kinetochores throughout mitosis 1160 Single kinetochore tracks of Venus-MAD1 fluorescence decay shown in Figure 1 1161 and 6. Thin lines represent individual kinetochores (Control n = 53 kinetochores, 1162 7 cells, green; Control Large KT s n = 19 kinetochores, 7 cells, dark green; 1163 Control small KTs n = 34 kinetochores, 5 cells, light green; Nocodazole n = 31 1164 kinetochores, 8 cells, grey; MPS1i n = 63 kinetochores, 12 cells, yellow; CDK1i n 1165 = 18 kinetochores, 10 cells, blue; Mock n = 48 kinetochores, 11 cells, light grey; 1166 siHAUS6 n = 47 kinetochores, 10 cells, purple blue; data pooled from at least 1167 four independent experiments per condition). Thick lines show the single-phase 1168 exponential decay fit curve for all pooled data (except for Nocodazole: average 1169 cure curve). 1170 1171 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 42 Figure S3. Additional examples of partially attached kinetochores 1172 (A), (A) Representative CH-STED image of Indian muntjac fibroblasts with 1173 partially attached kinetochores. Microtubules (grey; STED), centromeres (ACA; 1174 blue; STED) and Venus-MAD1 (orange; confocal). Insets: partially attached 1175 kinetochores. Scale bar: 2µm. 1176 (A’), (B’) Max normalized Venus-Mad1 (orange) and tubulin (microtubules; grey) 1177 signal intensity profile along the longitudinal axis of the kinetochores highlighted 1178 in A and B, respectively (arrow head). 1179 (C) Correlation map of MAD1 and microtubule intensity z-scores for the all 1180 partially attached kinetochores analyzed in untreated cells. Each scatterplot 1181 contains data from a single kinetochore, with each dot representing a 1182 sub-kinetochore region. P-value was calculated with Student t-test for zero 1183 correlation (two-sided). 1184 1185 Figure S4. Additional examples of localized MAD1 recruitment upon 1186 kinetochore laser ablation 1187 (A) Representative live-recording images of Indian muntjac cells expressing 1188 2x-GFP-CENP-A and GFP-CENTRIN-1 (both in blue) and mScarlet-MAD1 1189 (orange/ grey scale) upon partial kinetochore surgery. Insets: kinetochore pair 1190 that was perturbed. Arrows point to the ablated kinetochore region, pre- and 1191 post-surgery (solid and hollow arrow, respectively). Arrowhead shows MAD1 1192 recruitment. Scale bar: 2µm. 1193 (A’) Max normalized line intensity profile plot of MAD1/CENP-A signal ratio along 1194 the longitudinal axis of the kinetochore highlighted in A. 1195 (B) Max normalized MAD1/CENP-A ratio intensity profile along the longitudinal 1196 axis of kinetochores after partial surgery of their sisters (orange) or upon acute 1197 Nocodazole treatment (NOC; grey). Thin lines show individual kinetochores and 1198 thick lines show average curves (surgery n = 5 kinetochores/cells, 5 independent 1199 experiments; Nocodazole n = 19 kinetochores/cells, 2 independent 1200 experiments). The graph depicts a cohort of cells that recruit MAD1 to a distal 1201 position relative to the surgery region (n = 5/16 MAD1 recruiting kinetochores). 1202 Data from Nocodazole treated cells is the same as shown in Figure 5. 1203 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 43 (C), (C’) Representative images from correlative live cell spinning disk (C) and 1204 fixed CH-STED microscopy (C’) in a cell that recruited MAD1 to distal position 1205 relative to the surgery region. Indian muntjac cells expressing 2x-GFP-CENP-A 1206 and GFP-CENTRIN-1 (both in blue) and mScarlet-MAD1 (orange/grey scale) 1207 were fixed after partial kinetochore surgery and stained to visualize microtubules 1208 (grey). Insets: kinetochore pair that was perturbed. Arrows point to the ablated 1209 kinetochore region, pre- and post-surgery (solid and hollow arrow, respectively). 1210 Arrowheads show MAD1 recruitment. Scale bar: 2µm. 1211 (C’’) Max normalized line intensity profile plot of MAD1/CENP-A (live cell data 1212 just before fixation, orange) and microtubules/CENP-A (fixed cell data, grey) 1213 signal ratio along the longitudinal axis of the kinetochore highlighted in C and C’. 1214 1215 Figure S5. Uniform perturbation of microtubule occupancy upon Augmin 1216 depletion 1217 (A) Representative CH-STED image of an Indian muntjac fibroblasts with 1218 partially attached kinetochores upon siHAUS6 treatment. Microtubules (grey; 1219 STED), centromeres (ACA; blue; STED) and Venus-MAD1 (orange; confocal). 1220 Insets: partially attached kinetochore. Scale bar: 2µm. 1221 (B) Three-dimensional reconstruction of the cell shown in A. Inset: partially 1222 attached kinetochore. The kinetochore pair of interest is traced by a dashed line. 1223 Microtubules (grey), centromeres (ACA; blue) and Venus-MAD1 (orange). Scale 1224 bars, Left: 3µm; Right: 1µm. 1225 (C) Max normalized Venus-Mad1 (orange) and tubulin (microtubules; grey) 1226 signal intensity profile along the longitudinal axis of the kinetochore highlighted 1227 in A (arrow head). 1228 (D) Pearson correlation coefficient between MAD1 and microtubules signal 1229 intensity along the kinetochore. Bars show individual partially attached 1230 kinetochores in siHAUS6 treated cells (n=13 kinetochores, 9 cells, 2 1231 independent experiments). 1232 (E) Correlation map of MAD1 and microtubule intensity z-scores for kinetochores 1233 in control (data from Figure 3) and in siHAUS6 treated cells (shown in D). The 1234 underlying scatterplot contains data from multiple kinetochores, with each dot 1235 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 44 representing a sub-kinetochore region. For visualization, the scatterplot was 1236 converted into a regional density map. P-value was calculated with Student 1237 t-test for zero correlation (two-sided). 1238 1239 Figure S6. Additional examples of partially attached kinetochores in 1240 Augmin depleted cells 1241 (A), (A) Representative CH-STED image of Indian muntjac fibroblasts siHAUS6 1242 treated with partially attached kinetochores. Microtubules (grey; STED), 1243 centromeres (ACA; blue; STED) and Venus-MAD1 (orange; confocal). Insets: 1244 partially attached kinetochores. Scale bar: 2µm. 1245 (A’), (B’) Max normalized Venus-Mad1 (orange) and tubulin (microtubules; grey) 1246 signal intensity profile along the longitudinal axis of the kinetochores highlighted 1247 in A and B, respectively (arrowhead). 1248 (C) Correlation map of MAD1 and microtubule intensity z-scores for the all 1249 partially attached kinetochores analyzed in siHAUS6 cells. Each scatterplot 1250 contains data from a single kinetochore, with each dot representing a 1251 sub-kinetochore region. P-value was calculated with Student t-test for zero 1252 correlation (two-sided). 1253 1254 Figure S7. Extended mitosis activates the mitotic stopwatch to stop cell 1255 proliferation in Indian muntjac cells 1256 (A) Schematic summary of the experimental setup and downstream analysis. 1257 (B) Daughter cell cycle duration as a function of mother mitotic duration in Indian 1258 muntjac cells expressing H2B-GFP upon monastrol-induced arrest and release. 1259 Bars show individual daughter cells (n = 305 cells pooled from two independent 1260 experiments). Note that only daughter cells that entered mitosis during imaging 1261 were included in the graph (see methods section). 1262 (C) Mitotic duration of Indian muntjac cells that enter mitosis within a given 1263 recording time bin. Data pooled from two independent experiments (n = 489). 1264 Circles denote mean in each time bin (bin size /i2 =/i2 6h), error bars show the 1265 s.e.m. 1266 (D) Bar graph showing the percentage of Indian muntjac cells that exited mitosis 1267 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 45 with no defects or a particular mitotic defect/ error after distinct mitotic durations 1268 (bin size /i2 =/i2 30 min). Only mother cells from the selected cohort (0-18h of 1269 imaging) are shown. Data pooled from two independent experiments (n = 142 1270 cells). Bars show mean in each time bin (bin size/i2 =/i2 6h or 24h), error bars show 1271 the s.e.m. 1272 (E) Daughter cells fate (Gen 2) as a function of mother cell mitotic duration (Gen 1273 1). Only mother cells from the selected cohort (0-18h of imaging) and that 1274 completed mitosis with no defects are shown are shown. Bars show individual 1275 cells (n = 123 cells pooled from two independent experiments). 1276 (E’) Percentages of cell families dividing only once (Gen 1 mothers only) versus 1277 a second (Gen 2 daughters) or a third time (Gen 3 daughters), after binning the 1278 cell families by mitotic duration of the Gen 1 mother. Circles show the two 1279 independent experiments, bars show the mean and errors show the s.e.m. 1280 (F) Summary of the percentage of daughter cells that arrest as a function of 1281 mother cell mitotic duration and cause of mitotic delay. 1282 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 46

1283 1284 1. Bakhoum, S.F., and Cantley, L.C. (2018). The Multifaceted Role of 1285 Chromosomal Instability in Cancer and Its Microenvironment. Cell 174, 1286 1347-1360. 10.1016/j.cell.2018.08.027. 1287 2. Rieder, C.L., Cole, R.W., Khodjakov, A., and Sluder, G. (1995). The 1288 checkpoint delaying anaphase in response to chromosome 1289 monoorientation is mediated by an inhibitory signal produced by 1290 unattached kinetochores. The Journal of cell biology 130, 941-948. 1291 3. Musacchio, A., and Salmon, E.D. (2007). The spindle-assembly 1292 checkpoint in space and time. Nature reviews 8, 379-393. nrm2163 [pii] 1293 10.1038/nrm2163. 1294 4. Lara-Gonzalez, P., Pines, J., and Desai, A. (2021). Spindle assembly 1295 checkpoint activation and silencing at kinetochores. Semin Cell Dev Biol 1296 117, 86-98. 10.1016/j.semcdb.2021.06.009. 1297 5. McAinsh, A.D., and Kops, G. (2023). Principles and dynamics of spindle 1298 assembly checkpoint signalling. Nature reviews 24, 543-559. 1299 10.1038/s41580-023-00593-z. 1300 6. Vazquez-Novelle, M.D., Sansregret, L., Dick, A.E., Smith, C.A., McAinsh, 1301 A.D., Gerlich, D.W., and Petronczki, M. (2014). Cdk1 inactivation 1302 terminates mitotic checkpoint surveillance and stabilizes kinetochore 1303 attachments in anaphase. Curr Biol 24 , 638-645. 1304 10.1016/j.cub.2014.01.034. 1305 7. Rattani, A., Vinod, P .K., Godwin, J., Tachibana-Konwalski, K., Wolna, M., 1306 Malumbres, M., Novak, B., and Nasmyth, K. (2014). Dependency of the 1307 spindle assembly checkpoint on Cdk1 renders the anaphase transition 1308 irreversible. Curr Biol 24, 630-637. 10.1016/j.cub.2014.01.033. 1309 8. Collin, P ., Nashchekina, O., Walker, R., and Pines, J. (2013). The spindle 1310 assembly checkpoint works like a rheostat rather than a toggle switch. 1311 Nature cell biology 15, 1378-1385. 10.1038/ncb2855 1312 ncb2855 [pii]. 1313 9. Dick, A.E., and Gerlich, D.W. (2013). Kinetic framework of spindle 1314 assembly checkpoint signalling. Nature cell biology 15, 1370-1377. 1315 10.1038/ncb2842 1316 ncb2842 [pii]. 1317 10. Rosenberg, J.S., Cross, F.R., and Funabiki, H. (2011). KNL1/Spc105 1318 recruits PP1 to silence the spindle assembly checkpoint. Curr Biol 21, 1319 942-947. 10.1016/j.cub.2011.04.011. 1320 11. Moura, M., Osswald, M., Leca, N., Barbosa, J., Pereira, A.J., Maiato, H., 1321 Sunkel, C.E., and Conde, C. (2017). Protein Phosphatase 1 inactivates 1322 Mps1 to ensure efficient Spindle Assembly Checkpoint silencing. Elife 6, 1323 e25366. 10.7554/eLife.25366. 1324 1 2. Grallert, A., Boke, E., Hagting, A., Hodgson, B., Connolly, Y ., Griffiths, 1325 J.R., Smith, D.L., Pines, J., and Hagan, I.M. (2015). A PP1-PP2A 1326 phosphatase relay controls mitotic progression. Nature 517, 94-98. 1327 10.1038/nature14019. 1328 13. Hayward, D., Alfonso-Pérez, T., Cundell, M.J., Hopkins, M., Holder, J., 1329 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 47 Bancroft, J., Hutter, L.H., Novak, B., Barr, F.A., and Gruneberg, U. (2019). 1330 CDK1-CCNB1 creates a spindle checkpoint-permissive state by enabling 1331 MPS1 kinetochore localization. The Journal of cell biology 218, 1332 1182-1199. 10.1083/jcb.201808014. 1333 14. Alfonso-Pérez, T., Hayward, D., Holder, J., Gruneberg, U., and Barr, F.A. 1334 (2019). MAD1-dependent recruitment of CDK1-CCNB1 to kinetochores 1335 promotes spindle checkpoint signaling. The Journal of cell biology 218, 1336 1108-1117. 10.1083/jcb.201808015. 1337 15. Allan, L.A., Camacho Reis, M., Ciossani, G., Huis In 't Veld, P.J., 1338 Wohlgemuth, S., Kops, G.J., Musacchio, A., and Saurin, A.T. (2020). 1339 Cyclin B1 scaffolds MAD1 at the kinetochore corona to activate the 1340 mitotic checkpoint. The EMBO journal 39, e103180. 1341 10.15252/embj.2019103180. 1342 16. Hiruma, Y ., Sacristan, C., Pachis, S.T., Adamopoulos, A., Kuijt, T., Ubbink, 1343 M., von Castelmur, E., Perrakis, A., and Kops, G.J. (2015). CELL 1344 DIVISION CYCLE. Competition between MPS1 and microtubules at 1345 kinetochores regulates spindle checkpoint signaling. Science (New York, 1346 N.Y 348, 1264-1267. 10.1126/science.aaa4055. 1347 17. Ji, Z., Gao, H., and Yu, H. (2015). CELL DIVISION CYCLE. Kinetochore 1348 attachment sensed by competitive Mps1 and microtubule binding to 1349 Ndc80C. Science (New York, N.Y 348, 1260-1264. 1350 10.1126/science.aaa4029. 1351 18. Howell, B.J., McEwen, B.F., Canman, J.C., Hoffman, D.B., Farrar, E.M., 1352 Rieder, C.L., and Salmon, E.D. (2001). Cytoplasmic dynein/dynactin 1353 drives kinetochore protein transport to the spindle poles and has a role in 1354 mitotic spindle checkpoint inactivation. The Journal of cell biology 155, 1355 1159-1172. 1356 19. Wojcik, E., Basto, R., Serr, M., Scaerou, F., Karess, R., and Hays, T . 1357 (2001). Kinetochore dynein: its dynamics and role in the transport of the 1358 Rough deal checkpoint protein. Nature cell biology 3, 1001-1007. 1359 20. Gassmann, R., Holland, A.J., Varma, D., Wan, X., Civril, F., Cleveland, 1360 D.W., Oegema, K., Salmon, E.D., and Desai, A. (2010). Removal of 1361 Spindly from microtubule-attached kinetochores controls spindle 1362 checkpoint silencing in human cells. Genes & development 24, 957-971. 1363 10.1101/gad.1886810. 1364 21. Ide, A.H., DeLuca, K.F., Wiggan, O., Markus, S.M., and DeLuca, J.G. 1365 (2023). The role of kinetochore dynein in checkpoint silencing is restricted 1366 to disassembly of the corona. Molecular biology of the cell 34 , ar76. 1367 10.1091/mbc.E23-04-0130. 1368 22. Barisic, M., Sohm, B., Mikolcevic, P ., Wandke, C., Rauch, V., Ringer, T., 1369 Hess, M., Bonn, G., and Geley, S. (2010). Spindly/CCDC99 is required for 1370 efficient chromosome congression and mitotic checkpoint regulation. 1371 M olecular biology of the cell 21, 1968-1981. 10.1091/mbc.e09-04-0356. 1372 23. McEwen, B.F., Heagle, A.B., Cassels, G.O., Buttle, K.F., and Rieder, C.L. 1373 (1997). Kinetochore fiber maturation in PtK1 cells and its implications for 1374 the mechanisms of chromosome congression and anaphase onset. J. 1375 Cell Biol. 137, 1567-1580. 10.1083/jcb.137.7.1567. 1376 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 48 24. Nicklas, R.B., Waters, J.C., Salmon, E.D., and Ward, S.C. (2001). 1377 Checkpoint signals in grasshopper meiosis are sensitive to microtubule 1378 attachment, but tension is still essential. Journal of cell science 114, 1379 4173-4183. 10.1242/jcs.114.23.4173. 1380 25. Kuhn, J., and Dumont, S. (2019). Mammalian kinetochores count 1381 attached microtubules in a sensitive and switch-like manner. The Journal 1382 of cell biology. 10.1083/jcb.201902105. 1383 26. Etemad, B., Vertesy, A., Kuijt, T.E.F., Sacristan, C., van Oudenaarden, A., 1384 and Kops, G. (2019). Spindle checkpoint silencing at kinetochores with 1385 submaximal microtubule occupancy. Journal of cell science 132. 1386 10.1242/jcs.231589. 1387 27. Dudka, D., Noatynska, A., Smith, C.A., Liaudet, N., McAinsh, A.D., and 1388 Meraldi, P . (2018). Complete microtubule-kinetochore occupancy favours 1389 the segregation of merotelic attachments. Nat Commun 9, 2042. 1390 10.1038/s41467-018-04427-x. 1391 28. Pereira, A.J., and Maiato, H. (2012). Maturation of the 1392 kinetochore-microtubule interface and the meaning of metaphase. 1393 Chromosome Res 20, 563-577. 10.1007/s10577-012-9298-8. 1394 29. Meitinger, F ., Belal, H., Davis, R.L., Martinez, M.B., Shiau, A.K., Oegema, 1395 K., and Desai, A. (2024). Control of cell proliferation by memories of 1396 mitosis. Science (New York, N.Y 383, 1441-1448. 1397 10.1126/science.add9528. 1398 30. Sparr, C., and Meitinger, F. (2025). Prolonged mitosis: A key indicator for 1399 detecting stressed and damaged cells. Current opinion in cell biology 92, 1400 102449. 10.1016/j.ceb.2024.102449. 1401 31. Meitinger, F., Anzola, J.V., Kaulich, M., Richardson, A., Stender, J.D., 1402 Benner, C., Glass, C.K., Dowdy, S.F., Desai, A., Shiau, A.K., and 1403 Oegema, K. (2016). 53BP1 and USP28 mediate p53 activation and G1 1404 arrest after centrosome loss or extended mitotic duration. The Journal of 1405 cell biology 214, 155-166. 10.1083/jcb.201604081. 1406 32. Lambrus, B.G., Daggubati, V., Uetake, Y ., Scott, P .M., Clutario, K.M., 1407 Sluder, G., and Holland, A.J. (2016). A USP28-53BP1-p53-p21 signaling 1408 axis arrests growth after centrosome loss or prolonged mitosis. The 1409 Journal of cell biology 214, 143-153. 10.1083/jcb.201604054. 1410 33. Fulcher, L.J., Batley, C., Sobajima, T., and Barr, F.A. (2025). Time as a 1411 danger signal promoting G1 arrest after mitosis. Trends in cell biology. 1412 10.1016/j.tcb.2025.06.001. 1413 34. Fong, C.S., Mazo, G., Das, T ., Goodman, J., Kim, M., O'Rourke, B.P., 1414 Izquierdo, D., and Tsou, M.F. (2016). 53BP1 and USP28 mediate 1415 p53-dependent cell cycle arrest in response to centrosome loss and 1416 prolonged mitosis. Elife 5. 10.7554/eLife.16270. 1417 35. Uetake, Y ., and Sluder, G. (2010). Prolonged prometaphase blocks 1418 d aughter cell proliferation despite normal completion of mitosis. Curr Biol 1419 20, 1666-1671. 10.1016/j.cub.2010.08.018. 1420 36. Uetake, Y ., and Sluder, G. (2018). Activation of the apoptotic pathway 1421 during prolonged prometaphase blocks daughter cell proliferation. 1422 Molecular biology of the cell 29, 2632-2643. 10.1091/mbc.E18-01-0026. 1423 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 49 37. Brinkley, B.R., Valdivia, M.M., Tousson, A., and Brenner, S.L. (1984). 1424 Compound kinetochores of the Indian muntjac. Evolution by linear fusion 1425 of unit kinetochores. Chromosoma 91, 1-11. 1426 38. Chi, J.X., Huang, L., Nie, W., Wang, J., Su, B., and Yang, F. (2005). 1427 Defining the orientation of the tandem fusions that occurred during the 1428 evolution of Indian muntjac chromosomes by BAC mapping. 1429 Chromosoma 114, 167-172. 10.1007/s00412-005-0004-x. 1430 39. Wurster, D.H., and Benirschke, K. (1970). Indian muntjac, Muntiacus 1431 muntjak: a deer with a low diploid chromosome number. Science 168, 1432 1364-1366. 10.1126/science.168.3937.1364. 1433 40. Comings, D.E., and Okada, T.A. (1971). Fine structure of kinetochore in 1434 Indian muntjac. Exp Cell Res 67, 97-110. 1435 41. Drpic, D., Almeida, A.C., Aguiar, P., Renda, F., Damas, J., Lewin, H.A., 1436 Larkin, D.M., Khodjakov, A., and Maiato, H. (2018). Chromosome 1437 Segregation Is Biased by Kinetochore Size. Curr Biol 28, 1344-1356 1438 e1345. 10.1016/j.cub.2018.03.023. 1439 42. Almeida, A.C., Drpic, D., Okada, N., Bravo, J., Madureira, M., and Maiato, 1440 H. (2020). Functional Dissection of Mitosis Using Immortalized 1441 Fibroblasts from the Indian Muntjac, a Placental Mammal with Only Three 1442 Chromosomes. Methods Mol. Biol. 2101, 247-266. 1443 10.1007/978-1-0716-0219-5_16. 1444 43. Almeida, A.C., Soares-de-Oliveira, J., Drpic, D., Cheeseman, L.P., 1445 Damas, J., Lewin, H.A., Larkin, D.M., Aguiar, P ., Pereira, A.J., and Maiato, 1446 H. (2022). Augmin-dependent microtubule self-organization drives 1447 kinetochore fiber maturation in mammals. Cell Rep 39, 110610. 1448 10.1016/j.celrep.2022.110610. 1449 44. Howell, B.J., Moree, B., Farrar, E.M., Stewart, S., Fang, G., and Salmon, 1450 E.D. (2004). Spindle checkpoint protein dynamics at kinetochores in living 1451 cells. Curr Biol 14, 953-964. 10.1016/j.cub.2004.05.053 1452 S0960982204003860 [pii]. 1453 45. Shah, J.V., Botvinick, E., Bonday, Z., Furnari, F., Berns, M., and 1454 Cleveland, D.W. (2004). Dynamics of centromere and kinetochore 1455 proteins; implications for checkpoint signaling and silencing. Curr Biol 14, 1456 942-952. 10.1016/j.cub.2004.05.046. 1457 46. Kuhn, J., and Dumont, S. (2017). Spindle assembly checkpoint 1458 satisfaction occurs via end-on but not lateral attachments under tension. 1459 The Journal of cell biology 216, 1533-1542. 10.1083/jcb.201611104. 1460 47. Kapoor, T.M., Lampson, M.A., Hergert, P ., Cameron, L., Cimini, D., 1461 Salmon, E.D., McEwen, B.F., and Khodjakov, A. (2006). Chromosomes 1462 can congress to the metaphase plate before biorientation. Science (New 1463 York, N.Y 311, 388-391. 311/5759/388 [pii] 1464 10.1126/science.1122142. 1465 48. Barisic, M., Aguiar, P., Geley, S., and Maiato, H. (2014). Kinetochore 1466 motors drive congression of peripheral polar chromosomes by 1467 overcoming random arm-ejection forces. Nature cell biology 16. 1468 10.1038/ncb3060. 1469 49. Itoh, G., Ikeda, M., Iemura, K., Amin, M.A., Kuriyama, S., Tanaka, M., 1470 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 50 Mizuno, N., Osakada, H., Haraguchi, T., and Tanaka, K. (2018). Lateral 1471 attachment of kinetochores to microtubules is enriched in prometaphase 1472 rosette and facilitates chromosome alignment and bi-orientation 1473 establishment. Scientific Reports 8, 3888. 10.1038/s41598-018-22164-5. 1474 50. Vukusic, K., and Tolic, I.M. (2025). CENP-E initiates chromosome 1475 congression by opposing Aurora kinases to promote end-on attachments. 1476 Nat Commun 16, 8537. 10.1038/s41467-025-64148-w. 1477 51. Kwiatkowski, N., Jelluma, N., Filippakopoulos, P ., Soundararajan, M., 1478 Manak, M.S., Kwon, M., Choi, H.G., Sim, T., Deveraux, Q.L., Rottmann, 1479 S., et al. (2010). Small-molecule kinase inhibitors provide insight into 1480 Mps1 cell cycle function. Nat Chem Biol 6, 359-368. 1481 10.1038/nchembio.345 1482 nchembio.345 [pii]. 1483 52. Vassilev, L.T., Tovar, C., Chen, S., Knezevic, D., Zhao, X., Sun, H., 1484 Heimbrook, D.C., and Chen, L. (2006). Selective small-molecule inhibitor 1485 reveals critical mitotic functions of human CDK1. Proceedings of the 1486 National Academy of Sciences of the United States of America 103, 1487 10660-10665. 10.1073/pnas.0600447103. 1488 53. Peretti, D., Maraschio, P ., Lambiase, S., Lo Curto, F., and Zuffardi, O. 1489 (1986). Indirect immunofluorescence of inactive centromeres as indicator 1490 of centromeric function. Human genetics 73, 12-16. 1491 54. Cherry, L.M., and Johnston, D.A. (1987). Size variation in kinetochores of 1492 human chromosomes. Human genetics 75, 155-158. 1493 55. Cherry, L.M., Faulkner, A.J., Grossberg, L.A., and Balczon, R. (1989). 1494 Kinetochore size variation in mammalian chromosomes: an image 1495 analysis study with evolutionary implications. Journal of cell science 92 ( 1496 Pt 2), 281-289. 1497 56. Sanchez, L., Martinez, P., and Goyanes, V. (1991). Analysis of 1498 centromere size in human chromosomes 1, 9, 15, and 16 by electron 1499 microscopy. Genome 34, 710-713. 1500 57. Tomkiel, J., Cooke, C.A., Saitoh, H., Bernat, R.L., and Earnshaw, W.C. 1501 (1994). CENP-C is required for maintaining proper kinetochore size and 1502 for a timely transition to anaphase. The Journal of cell biology 125, 1503 531-545. 1504 58. Nixon, F.M., Honnor, T.R., Clarke, N.I., Starling, G.P ., Beckett, A.J., 1505 Johansen, A.M., Brettschneider, J.A., Prior, I.A., and Royle, S.J. (2017). 1506 Microtubule organization within mitotic spindles revealed by serial block 1507 face scanning EM and image analysis. Journal of cell science. 1508 10.1242/jcs.203877. 1509 59. Logsdon, G.A., Rozanski, A.N., Ryabov, F., Potapova, T., Shepelev, V.A., 1510 Catacchio, C.R., Porubsky, D., Mao, Y ., Yoo, D., Rautiainen, M., et al. 1511 (2024). The variation and evolution of complete human centromeres. 1512 N ature. 10.1038/s41586-024-07278-3. 1513 60. Parl, F.F . (2024). Analysis of CENP-B Boxes as Anchor of Kinetochores in 1514 Centromeres of Human Chromosomes. Bioinform Biol Insights 18 , 1515 11779322241248913. 10.1177/11779322241248913. 1516 61. Kixmoeller, K., Tarasovetc, E.V., Mer, E., Chang, Y.W., and Black, B.E. 1517 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 51 (2025). Centromeric chromatin clearings demarcate the site of 1518 kinetochore formation. Cell. 10.1016/j.cell.2024.12.025. 1519 62. Pereira, A., Sousa, M., Almeida, A.C., Ferreira, L.T., Costa, A.R., 1520 Novais-Cruz, M., Ferras, C., Sousa, M.M., Sampaio, P., Belsley, M., and 1521 Maiato, H. (2019). Coherent-hybrid STED: high contrast sub-diffraction 1522 imaging using a bi-vortex depletion beam. Optics express 27, 8092-8111. 1523 10.1364/OE.27.008092. 1524 63. Zaytsev, A.V., Sundin, L.J., DeLuca, K.F., Grishchuk, E.L., and DeLuca, 1525 J.G. (2014). Accurate phosphoregulation of kinetochore-microtubule 1526 affinity requires unconstrained molecular interactions. The Journal of cell 1527 biology 206, 45-59. 10.1083/jcb.201312107 1528 jcb.201312107 [pii]. 1529 64. Volkov, V.A., Huis In 't Veld, P .J., Dogterom, M., and Musacchio, A. 1530 (2018). Multivalency of NDC80 in the outer kinetochore is essential to 1531 track shortening microtubules and generate forces. Elife 7. 1532 10.7554/eLife.36764. 1533 65. Stimac, V., Koprivec, I., Manenica, M., Simunic, J., and Tolic, I.M. (2022). 1534 Augmin prevents merotelic attachments by promoting proper 1535 arrangement of bridging and kinetochore fibers. Elife 11. 1536 10.7554/eLife.83287. 1537 66. Thompson, S.L., and Compton, D.A. (2010). Proliferation of aneuploid 1538 human cells is limited by a p53-dependent mechanism. The Journal of 1539 cell biology 188, 369-381. 10.1083/jcb.200905057. 1540 67. Santaguida, S., Richardson, A., Iyer, D.R., M'Saad, O., Zasadil, L., 1541 Knouse, K.A., Wong, Y .L., Rhind, N., Desai, A., and Amon, A. (2017). 1542 Chromosome Mis-segregation Generates Cell-Cycle-Arrested Cells with 1543 Complex Karyotypes that Are Eliminated by the Immune System. 1544 Developmental cell 41, 638-651 e635. 10.1016/j.devcel.2017.05.022. 1545 68. Hinchcliffe, E.H., Day, C.A., Karanjeet, K.B., Fadness, S., Langfald, A., 1546 Vaughan, K.T., and Dong, Z. (2016). Chromosome missegregation during 1547 anaphase triggers p53 cell cycle arrest through histone H3.3 1548 Ser31 phosphorylation. Nature cell biology 18, 668-675. 1549 10.1038/ncb3348. 1550 69. Giam, M., Wong, C.K., Low, J.S., Sinelli, M., Dreesen, O., and Rancati, G. 1551 (2020). P53 induces senescence in the unstable progeny of aneuploid 1552 cells. Cell Cycle 19, 3508-3520. 10.1080/15384101.2020.1850968. 1553 70. Phan, T .P ., Maryniak, A.L., Boatwright, C.A., Lee, J., Atkins, A., Tijhuis, A., 1554 Spierings, D.C., Bazzi, H., Foijer, F., Jordan, P .W., et al. (2021). 1555 Centrosome defects cause microcephaly by activating the 1556 53BP1-USP28-TP53 mitotic surveillance pathway. The EMBO journal 40, 1557 e106118. 10.15252/embj.2020106118. 1558 71. Brito, D.A., and Rieder, C.L. (2006). Mitotic checkpoint slippage in 1559 h umans occurs via cyclin B destruction in the presence of an active 1560 checkpoint. Curr Biol 16, 1194-1200. 10.1016/j.cub.2006.04.043. 1561 72. Gascoigne, K.E., and Taylor, S.S. (2008). Cancer cells display profound 1562 intra- and interline variation following prolonged exposure to antimitotic 1563 drugs. Cancer Cell 14, 111-122. 10.1016/j.ccr.2008.07.002. 1564 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 52 73. Renda, F., Magidson, V., Tikhonenko, I., Fisher, R., Miles, C., Mogilner, 1565 A., and Khodjakov, A. (2020). Effects of malleable kinetochore 1566 morphology on measurements of intrakinetochore tension. Open biology 1567 10, 200101. 10.1098/rsob.200101. 1568 74. Sacristan, C., Samejima, K., Ruiz, L.A., Deb, M., Lambers, M.L.A., 1569 Buckle, A., Brackley, C.A., Robertson, D., Hori, T., Webb, S., et al. (2024). 1570 Vertebrate centromeres in mitosis are functionally bipartite structures 1571 stabilized by cohesin. Cell. 10.1016/j.cell.2024.04.014. 1572 75. Altemose, N., Logsdon, G.A., Bzikadze, A.V., Sidhwani, P., Langley, S.A., 1573 Caldas, G.V., Hoyt, S.J., Uralsky, L., Ryabov, F.D., Shew, C.J., et al. 1574 (2022). Complete genomic and epigenetic maps of human centromeres. 1575 Science (New York, N.Y 376, eabl4178. doi:10.1126/science.abl4178. 1576 76. Zou, Y., Yi, X., Wright, W.E., and Shay, J.W. (2002). Human telomerase 1577 can immortalize Indian muntjac cells. Exp Cell Res 281, 63-76. 1578 77. Zhang, G., Kruse, T ., Guasch Boldu, C., Garvanska, D.H., Coscia, F., 1579 Mann, M., Barisic, M., and Nilsson, J. (2019). Efficient mitotic checkpoint 1580 signaling depends on integrated activities of Bub1 and the RZZ complex. 1581 The EMBO journal 38. 10.15252/embj.2018100977. 1582 78. Lukinavicius, G., Reymond, L., D'Este, E., Masharina, A., Gottfert, F ., Ta, 1583 H., Guther, A., Fournier, M., Rizzo, S., Waldmann, H., et al. (2014). 1584 Fluorogenic probes for live-cell imaging of the cytoskeleton. Nat Methods 1585 11, 731-733. 10.1038/nmeth.2972 1586 nmeth.2972 [pii]. 1587 79. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., 1588 Pietzsch, T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., et al. 1589 (2012). Fiji: an open-source platform for biological-image analysis. Nat 1590

9, 676-682. 10.1038/nmeth.2019. 1591 80. Pereira, A.J., Matos, I., Lince-Faria, M., and Maiato, H. (2009). Dissecting 1592 mitosis with laser microsurgery and RNAi in Drosophila cells. Methods 1593 Mol Biol 545, 145-164. 10.1007/978-1-60327-993-2_9. 1594 81. Tinevez, J.Y ., Perry, N., Schindelin, J., Hoopes, G.M., Reynolds, G.D., 1595 Laplantine, E., Bednarek, S.Y ., Shorte, S.L., and Eliceiri, K.W. (2017). 1596 TrackMate: An open and extensible platform for single-particle tracking. 1597

115, 80-90. 10.1016/j.ymeth.2016.09.016. 1598 82. Kletter, T., Munoz, O., Reusch, S., Biswas, A., Halavatyi, A., Neumann, 1599 B., Kuropka, B., Zaburdaev, V., and Reber, S. (2025). Cell state-specific 1600 cytoplasmic density controls spindle architecture and scaling. Nature cell 1601 biology 27, 959-971. 10.1038/s41556-025-01678-x. 1602 83. Jeff Reback, W.M., jbrockmendel, Joris Van den Bossche, Tom 1603 Augspurger, Phillip Cloud, gfyoung, Sinhrks, Simon Hawkins, Adam Klein, 1604 Matthew Roeschke, Jeff Tratner, T erji Petersen, Chang She, William Ayd, 1605 MomIsBestFriend, Marc Garcia, Jeremy Schendel, Andy Hayden, … 1606 Martin Winkel (2020). pandas-dev/pandas: Pandas 1.1.3 (v1.1.3). 1607 Z enodo. . 1608 84. VanderPlas, J., Granger, B., Heer, J., Moritz, D., Wongsuphasawat, K., 1609 Satyanarayan, A., Lees, E., Timofeev, I., Welsh, B., Sievert, S. (2018). 1610 Altair: Interactive Statistical Visualizations for Python. Journal of Open 1611 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 53 Source Software 3, 1057. https://doi.org/10.21105/joss.01057. 1612 1613 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint A D E 0 s 0 s 200 s120 s60 s 0 s 360 s240 s120 s 200 s 400 s 600 s SiR-Tubulin ControlNocodazoleMPS1iCDK1i Relative Venus-MAD1 levels Relative Venus-MAD1 levels Time (s) Time (s) Control Large KTs Small KTs 126 160 116 33 84 117 - 137 144 - 180 105 - 128 29 - 37 70 - 103 0.52 0.66 0.49 0.59 0.67 95 % CI R 2 MPS1i CDK1i 0 100 200 300 400 500 600 700 −0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0 100 200 300 400 500 600 700 −0.2 0.0 0.2 0.4 0.6 0.8 1.0 Control Nocodazole MPS1i CDK1i Control (Large KTs) Control (Small KTs)Venus-MAD1 Half reduction time (s) p-value <0.0001 <0.0001 0.0018 2x: 2x: 2x: 2x: A’ 0 s 336 s 498 s 570 s 690 s SiR-Tubulin Venus-MAD1 Venus-MAD1 B C 0 s 498 s336 s 570 s 1134 s 1320 s 1452 s240 s 690 s 864 s 1500 s 1674 s Pole 0 s 864 s 0 s 1674 s A’’ Pole * * ** ** 1 2 1 2 1 2 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint A C Total KT Bleach Figure 2 Time Total Partial 0 20 40 60 Fluorescence Recovery (%) 0.2319 Total Partial 0 2 4 6 8 Half recovery time (s) 0.5358 0.0 0.5 1.0 0 2 4 6 8 Normalized distance along KT Venus-MAD1 levels (KT/Cyto) Pre-Bleach Bleach 10 s 50 s Venus-MAD1 Partial KT Bleach 0 s-1 s 10 s 50 s Partial KT BleachRelative MAD1 levels Time t t Mobile Relative MAD1 levels Time Immobile Total KT Bleach 0.0 0.5 1.0 0 2 4 6 8 Normalized distance along KT Venus-MAD1 levels (KT/Cyto) Pre-Bleach Bleach 10 s 50 s B B’ C’ C’’ D E Relative Venus-MAD1 levels 0 10 20 30 40 50 60 0.0 0.5 1.0 Time (s) Recovery = 40% t1/2 = 5 s Bleach KT BGCtrl KT A’Mobile Immobile MAD1/MAD2 Dynein 0 s-1 s 10 s 50 s Venus-MAD1 B’’ Partial KT Bleach 0 10 20 30 40 50 60 0.0 0.5 1.0 Time (s) Recovery = 50% t1/2 = 7 s Relative Venus-MAD1 levels Non-Bleach Bleach Sis KT BGCtrl KT Bleach Bleach .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint Figure 3 A B 03:20 min 07:20 min 08:30 min00:00 min Cropped kinetochore Kinetochore kymograph MergeSiR-Tubulin t Venus-MAD1mScarlet-CENP-A Venus-MAD1 0 (s) 330 t mScarlet-CENP-A - + - + .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint Figure 4 Pearson correlation coefficient Individual kinetochores MergeA B C D 0 1 2 3 0.0 0.5 1.0 Distance along KT(µm) Relative signal intensity MAD1 MTs E Norm. intensity profile ZMTs Experimental data Strong positive correlation Strong negative correlationNo correlation r ≈ 1 r ≈ 0 r ≈ -1 Synthetic data F Microtubules ACA Venus-MAD1 Longitudinal KT line scanning Merge 3D rendering 3 3 -3-3 0 0 3 3 -3-3 0 0 ZMTs ZMTs ZMTs ZMAD1 ZMAD1 ZMAD1 ZMAD1 3 3 -3-3 0 0 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 S

 $*
 Q

F -1.0 -0.5 0.0 0.5 1.0 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint 00:00 min 08:00 min 15:20 min Fixed (CH-STED) A B’ Figure 5 B SurgeryNOC 0.0 0.5 1.0 0 1 2 3 4 6 8 Normalized distance along KT MAD1/CENP-A relative signal intensity D MAD1/CENP-A relative signal intensity SurgeryNOC 0.0 0.5 1.0 0 1 2 3 4 6 8 Normalized distance along KT NOC Surgery 0 50 100 150 Coefficient of variation (%) <0.0001E -01:00 min Live 12:00 min F 0.0 0.5 1.0 0.0 0.5 1.0 1.5 Normalized distance along KT Relative signal intensity MAD1/CENP-A MTs/CENP-A F’ After Surgery Nocodazole -01:40 min Pre-Surgery After SurgeryPre-Surgery F’’ C Partial laser ablation of the KT Longitudinal KT line scanning MAD1 (Ablated sister KT) (Ablated sister KT) KT Bending MAD1 Recruitment Time after surgery (min) 0.0015 0 2 4 6 8 10 12 14 16 18 3.6 2.8 ± 8.2 3.8 ± Mean SD ± 2xGFP-CENP-A; GFP-CENTRIN mScarlet-MAD1 2xGFP-CENP-A; GFP-CENTRIN mScarlet-MAD1 2xGFP-CENP-A; GFP-CENTRIN Microtubules After Surgery .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint A F Figure 6 D 0 20 40 60 80 100 % Cells w/ MAD1(+) KTs p<0.0001 p=0.0008 Prometaphase Metaphase siHaus6 Mock E Prometaphase Metaphase p<0.0001 p<0.0001 0.5 1.0 1.5 2.0 Venus-Mad1 levels (KT/cyto) siHAUS6 Mock B siHAUS6 Mock 0 100 200 300 400 500 600 700 Time (s) Relative V enus-MAD1 levels siHAUS6 Mock Merge +MG132 1h Control siHAUS6 Venus-MAD1ACAMicrotubules SiR-Tubulin 2x: 0 s 600 s 1200 s 1800 s Mock siHAUS6 2x: C Mock 142 222 128 - 158 186 - 272 0.53 0.29 Half reduction time (s) 95 % CI R 2 siHAUS6 p-value <0.0001 0 s 600 s 1200 s 1800 s Venus-MAD1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint Selected Gen 1 cohort (defect-free mitosis; n = 96 mother cells) siHAUS6 Tracking of cell fates Indian muntjac fibroblasts (GFP-H2B) 0 h 12 h 36 h 96 h Live-cell imaging Short mitosis Long mitosis A C A’ Gen 1 First mitosis Individual divisions 0 200 400 600 800 1,000 1,200 1,400 1,600 Mitotic duration (min) Selected Gen 1 cohort for daughter tracking Mitotic death Mitotic defect / errors No defect siHAUS6 All divisions < 30 120 < 30 - 60 60 - 90 90 - 120 Mother cell mitotic duration (min) ED 0 20 40 60 80 100 % of mother cells in cohort Selected Gen 1 cohort (n = 136 mother cells) N of daughters proliferating 0 1 2 0 100 200 300 400 Mother mitotic duration (min) Individual daughter cells Daughter arresting Daughter proliferating Figure 7 0 12 24 36 48 60 72 84 96 0 100 200 300 400 Mitotic duration (min) HAUS6 depletion time (h) siHAUS6 siCtrl B HAUS6 depletion time (h) 0 20 40 60 Percentage (%) Laggings Massive misseg. DNA bridge Misaligned Multip. division Cytok. failure Slippage <6 12-36 36-60 60-84 Mitotic defect /errors (n = 146 cells) .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint Kinetochore NEBD AnaphaseTime Time Microtubule MAD1/MAD2 SAC Strength APC/C Activity Threshold value Augmin-mediated K-fiber maturation Cyclin B1 / CDK1 Kinetochore NEBD Anaphase Microtubule MAD1/MAD2 SAC Strength APC/C Activity Threshold value Inefficient K-fiber maturation (e.g. Augmin depletion) Initial occupancy Partial occupancy Full occupancy Initial occupancy Partial occupancy Partial occupancy Daughters arrest Increased errors Daughters proliferate Accurate mitosis Cyclin B1 / CDK1 STOP Partial occupancy Full occupancy ? .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 28, 2026. ; https://doi.org/10.64898/2026.01.26.701783doi: bioRxiv preprint