Search for chromosomal instability aiding variants reveal naturally occurring kinetochore gene variants that perturb chromosome segregation

Summary Chromosomal instability (CIN) is a hallmark of cancers, and CIN-promoting mutations are not fully understood. Here, we report 141 chromosomal instability aiding variant (CIVa) candidates by assessing the prevalence of loss-of-function (LoF) variants in 135 chromosome segregation genes from over 150,000 humans. Unexpectedly, we observe both heterozygous and homozygous CIVa in Astrin and SKA3, two evolutionarily conserved kinetochore and microtubule-associated proteins essential for chromosome segregation. To stratify harmful versus harmless variants, we combine live-cell microscopy and controlled protein expression. We find the naturally occurring Astrin p.Q1012∗ variant is harmful as it fails to localize normally and induces chromosome misalignment and missegregation, in a dominant negative manner. In contrast, the Astrin p.L7Qfs∗21 variant generates a shorter isoform that localizes and functions normally, and the SKA3 p.Q70Kfs∗7 variant allows wild-type SKA complex localisation and function, revealing distinct resilience mechanisms that render these variants harmless. Thus, we present a scalable framework to predict and stratify naturally occurring CIVa, and provide insight into resilience mechanisms that compensate for naturally occurring CIVa.

6][27][28][29] However, an increase in parental relatedness can increase the chances of biallelic variants leading to a full loss of protein.So, a list of chromosomal instability-aiding variants (CIVa) tolerated in healthy humans in the heterozygous but not homozygous form is critical for precision medicine.Also, information on highly prevalent CIVa candidates in healthy populations are valuable tools to probe how a predicted loss of function (LoF) variant in a chromosome segregation gene is tolerated in cells or specific tissues, expanding our knowledge of chromosome segregation mechanisms and in turn, CIN preventing pathways.
The Genes and Health (GH) database encompasses data from 100,000 individuals of Pakistani and Bangladeshi origin with a high incidence of consanguineal marriages.Genetic databases such as the Genome Aggregation database gnomAD (a coalition of several international population-specific and disease-specific databases 30 ) and somatic mutation databases such as COSMIC (database of cancer patients organized by tissue types 31 ) are powerful datasets for predicting harmful LoF variants in kinetochore genes.However, no scalable framework to identify and rank the impact of harmful CIVa has been proposed.

OPEN ACCESS
microtubule-binding proteins that form large complexes, the Astrin, SKA, and Ndc80 complexes.By combining live-cell imaging and variant protein expression, we demonstrate the adverse impact of Astrin p.Q1012* variant that causes defective localization and chromosome segregation.Unexpectedly, two LoF variants, SKA3 p.Q70Kfs7 and Astrin p.L7Qfs*21, are tolerated in cells.SKA3 p.Q70Kfs7 may not be incorporated into the larger SKA complex whereas Astrin p.L7Qfs*21 generates a short isoform that is functional during mitosis.Thus, using live-cell microscopy and variant isoform study, we present a scalable framework to stratify harmful CIVa, and reveal resilience mechanisms that compensate for LoF kinetochore protein variants.

Unexpected LoF variants of essential kinetochore genes in healthy humans
To identify CIVa candidates, we conducted the first bioinformatics screen for predicted LoF variants in 135 chromosome segregation genes using the GH database 32 (heterozygous and homozygous variants, from multiple databases, are collated in Table S1).To screen for LoF variant candidates, we analyzed the prevalence of premature stop-codons, due to a missense mutation or nucleotide loss/gain associated frameshift, in at least two different individuals (germline variants and allelic prevalence listed in Table S1).4][35] To test whether two LoF variants in the SPAG5/Astrin gene, p.(Q1012*) and p.(L7Qfs*21) are specific to the population surveyed in the GH database, we screened wider population databases, COSMIC (cancer database) and gnomAD (population and disease-specific database) (Figure 1A).The Astrin p.(L7Qfs*21) variant is not found in the COSMIC database, but found in gnomAD as both heterozygous and homozygous forms (Table S1; Figure 1A), 36 and the variant is exclusive to South Asians (with an allele frequency of 2%) in the 1000 Genomes Project Phase 3. 37 In contrast, a variant of SKA3, SKA3Q70Kfs*7 is found in healthy humans, 32 and also reported in multiple cancers including hematopoietic, lymphoid, thyroid, pancreatic, esophageal and lung 31 (Figure S1A).In summary, some of the kinetochore gene variants in the GH database can be found in other databases.We present the first queryable resource of CIVa candidates, including their allele frequencies, in 135 human chromosome segregation genes collated across different databases here: https://github.com/Draviam-lab/CIVa.
To test whether Astrin variants are selectively depleted in cancers, we probed the incidence of somatic mutations in Astrin and Ndc80 (an interactor of Astrin) 35,38 by comparing multiple tumor tissues for variants in five gene categories: (a) MCPH genes, (b) MVA genes, (c) the Astrin-SKAP complex, (d) Astrin-SKAP interactors, and (e) TP53 and BRCA1 (tumor suppressor genes, as positive controls).As expected, mutations in TP53 were found in all tissue types, including gastrointestinal, placenta, and pleura (Figure S1B).TP53 mutation was above 50% in most tumor types, showing its high prevalence (Figure S1B).In contrast, mutations in BRCA1, MCPH genes and most of the Astrin-SKAP interacting partners, including Ndc80 complex, 35,38 are not present in all tissue types suggesting tissue-specificity (Figure S1B).In addition, the frequencies of mutations in the Astrin-SKAP and Ndc80 complexes are much lower compared to MCPH or BRCA1 genes (Figure S1B, see box).Thus, somatic mutations in the Astrin-SKAP, NDC80 complex, and MVA genes are not high, highlighting the uniqueness of the two LoF gene variants uncovered in the CIVa database: Astrin p.(Q1012*) and p.(L7Qfs*21).
The rare variant astrin p.Q1012* shows impaired kinetochore localization We set out to develop an easily scalable framework for quantifying the impact of CIVa candidates by probing their localization and function in spindle checkpoint proficient HeLa transformed epithelial cell line.The Astrin p.(Q1012*) variant is predicted to eliminate Astrin's C-terminal tail that is required for the protein's recruitment to kinetochores. 35To assess Astrin p.(Q1012*) variant localization, we transiently expressed Yellow Fluorescent Protein (YFP)-tagged Astrin wild-type (WT) or p.Q1012* in HeLa, both in the presence and absence of endogenous Astrin (Figure S2A).We depleted endogenous Astrin using siRNA and expressed siRNA-resistant YFP-tagged Astrin WT or p.Q1012* variant protein (Figure S2B).Immunostaining studies of metaphase-arrested mitotic cells, with equatorially centered spindles, 39 showed that unlike YFP-Astrin WT protein, the p.Q1012* variant fails to localize at kinetochores identified by anti-centromere (CREST) anti-sera (Figures 1B and  1C).Qualitative analysis of Astrin's enrichment at kinetochores (crescent-like signals) showed that while YFP-Astrin WT localizes on the mitotic spindle and kinetochores (as crescents), Astrin p.Q1012* localizes exclusively on the mitotic spindle but is not enriched on kinetochores (Figures 1B, 1C, and S2C).While the mislocalization of Astrin p.Q1012* is most striking in Astrin siRNA treated cells, the kinetochore localization defect can be observed in cells with no siRNA treatment (see in the following section).We compared p.Q1012* localization against two previously reported Astrin C-terminal mutants with different extents of kinetochore localization defects 35 (Figures S2D-S2F).Quantitative analysis of Astrin crescents at the outer-kinetochores of immunostained cells showed that the kinetochore enrichment of Astrin p.Q1012* is impaired similarly to the Astrin D70 deletion mutant and more severely compared to the Astrin 4A mutant (Figures S2E and S2F).Together these findings demonstrate Astrin p.Q1012* as a variant exhibiting a severe loss of kinetochore localization.
To probe whether mature kinetochore attachments can form in variant expressing cells, we used the end-on attachment marker, Astrin-SKAP complex. 7Live-cells coexpressing mKate2-Astrin showed normal Astrin localization in cell expressing SKA3-WT or Q70Kfs*7 tagged at the C or N-termini, demonstrating mature end-on attachments and confirming the lack of variant localization at kinetochores (Figure 2A).Immunostaining using antibodies against 156-177 a.a of SKA3 (Figure S3A) showed that in cells expressing GFP-SKA3 p.Q70Kfs*7, endogenous SKA3 localizes normally at kinetochores marked with CREST antisera (Figure 2B), which suggest normal endogenous SKA complex formation.In summary, the SKA3 Q70Kfs*7 variant leads to a truncated SKA3 that does not localize at kinetochores and does not interfere with chromosome congression or Astrin recruitment, indicating normal chromosome-microtubule attachments in cells expressing SKA3 p.(Q70Kfs*7).
Another high-frequency SKA3 p.(R27*) variant was found in COSMIC and gnomAD (6805 heterozygous; 0 homozygous; gnomAD) but not in the GH database.Homozygous variants of SKA3 p.(R27*) do not exist despite the prevalence of the heterozygous form, relative to other variants in the same gene (Figure S4A).Similar to SKA3 p.(Q70Kfs*7), SKA3 p.R27* expression did not affect chromosome congression but resulted in the expression of a short fragment that is expected to include the dimerization domain of SKA3 (Figures S3, S4B, and S4C).We conclude SKA3 p.(Q70Kfs*7) and p.(R27*) may be generally harmless in the presence of full-length SKA3, explaining their monoallelic prevalence across ancestries.Thus, our 3-step framework to predict CIVa candidates, mimic their expression, and analyze their impact using single-cell studies can help stratify harmless versus harmful variants in chromosome segregation genes (Figure 2C).
Screening for alternative translation start sites in Astrin p.(L7Qfs*21) using Kozak consensus sequences ((gcc)gccRccAUGG) predicted two sites, N-454 and N-823, which precede the region encoding coiled coil stretches in Astrin.To mimic these start sites, we generated two Astrin N-terminal deletion mutants, D151 and D274 tagged with YFP at their N-termini (Figure S5B).Comparing their corresponding protein sizes using immunoblotting showed that the two N-terminally tagged Astrin deletion mutants, D151 and D274, migrated at 144 kDa and 131 kDa, respectively (Figure 3D, higher than endogenous Astrin as expected).Importantly, YFP-Astrin D151 migrated similarly to Astrin p.L7*-GFP, in anti-Astrin and anti-GFP antibody-stained immunoblots, indicating that Astrin p.(L7Qfs*21) can promote the expression of an N-terminally truncated protein starting from 152 a.a. of Astrin.
While Astrin D151 mutant localizes normally at the kinetochore, D274 mutant exhibits reduced kinetochore localization, compared to mKate2-Astrin that was either coexpressed in live-cells (Figures 3E and 3F) or separately expressed as YFP-Astrin WT control in fixed-cell studies (Figures S6A-S6C).However, in the absence of endogenous Astrin, D274 mutant localizes normally at the kinetochore similar to Astrin (C) Cartoon of a three-step scalable framework used to predict, mimic and stratify CIVa on the basis of their impact on kinetochore protein localization and function assessed using microscopy assays.Bioinformatic predictions on LoF variants in kinetochore genes and their allelic prevalence across COSMIC, gnomAD, and ELGH databases are collated into the CIVa database.CIVa in kinetochore genes are mimicked in single cells using different protein expression tools and their impact is quantitatively assessed to stratify harmful and harmless variants in monoallelic and biallelic forms.S6E, 2D, and 2E) and it allows the recruitment of SKAP (a member of the Astrin-SKAP complex 38 ) that requires Astrin for its localization at kinetochores. 17Thus in the absence of full-length protein, shorter isoforms of Astrin can localize at the kinetochore, and recruit SKAP normally.
During interphase, both Astrin D274 and Astrin D151 are normally excluded from the nucleus and present at growing microtubule ends marked by EB1 an interactor of Astrin complex, 17,40 suggesting normal interphase localization and function (Figure 3G).We conclude that Astrin p.(L7Qfs*21) variant expresses a shorter isoform lacking the first 151 a.a of Astrin, which localizes normally at microtubule-ends, the mitotic spindle and kinetochores.Thus, the Astrin p.(L7Qfs*21) variant reveals alternative Kozak usage as a resilience mechanism allowing normal chromosome-microtubule attachments.

Astrin pQ1012* abrogates endogenous Astrin-SKAP localization and microtubule-mediated pulling of chromosomes
Of the variants we analyzed, Astrin p.(Q1012*) is unique with impaired kinetochore localization (Figure 1B) even in the presence of endogenous Astrin.As the Astrin-SKAP complex is a dimer, 38 we hypothesized that Astrin p.Q1012* may disrupt the localization of the endogenous Astrin-SKAP complex.To test this, we transiently expressed the Astrin WT and the p.Q1012* variant in HeLa cells and analyzed the kinetochore localization of endogenous Astrin and SKAP (Figure S7A).Immunostaining showed that Astrin localizes as a crescent at an average of 68% of kinetochores in Astrin p.Q1012* expressing cells compared to 96% of kinetochores in Astrin WT expressing cells (Figures S7B  and S7C).Similarly, endogenous SKAP localizes at 86% of kinetochores in Astrin p.Q1012* expressing cells compared to 97% of kinetochores in WT expressing cells (Figures S7D and S7E).These findings indicate that the expression of Astrin p.Q1012* can disrupt the kinetochore localization of the endogenous Astrin-SKAP complex in a dominant-negative manner.
Astrin's C-terminal tail, lost in Astrin p.Q1012*, serves two important roles: it delivers PP1 phosphatase to the kinetochore which stabilizes end-on kinetochore-microtubule attachments and it enables microtubule-mediated pulling which ensures maximum enrichment of Astrin-SKAP at kinetochores. 19,35,42So, we investigated the extent to which microtubule-mediated pulling and Astrin p.Q1012* localization is reduced following Astrin p.Q1012* expression.For this, we measured inter-centromeric distances by co-expressing CENPB-DsRed, a centromere marker, with YFP-tagged Astrin WT or p.Q1012* (Figures S8A and S8B).The range of inter-centromere distances was reduced in cells expressing YFP-Astrin p.Q1012* compared to YFP-Astrin WT (Figures S8B-S8D, Videos S1 and S2).This difference in inter-centromere distances was more striking following the normalization of inter-centromere distances of each pair to its unstretched state (marked T 0 ), indicating sustained reduction in pulling forces.Tracking the fate of sister kinetochores for 5 min showed that although centromeric stretching can be observed in Astrin p.Q1012* expressing cells, the maximum inter-centromeric distances are reduced compared to Astrin WT expressing cells (Figures S8E and S8F).Thus, stable microtubule-mediated pulling of kinetochores is reduced following Astrin pQ1012* variant expression.
Quantifying the dynamic loss of kinetochore-bound Astrin p.Q1012* relative to WT in metaphase kinetochores through time required automated analysis.We developed a computational workflow using segmentation and particle tracking tools (see STAR Methods) to apply CENPB-DsRed intensities as a mask, to measure YFP-Astrin p.Q1012* and WT protein intensities in dynamically stretching sister kinetochores (Figure S9A).While Astrin WT signals at kinetochore were on average 1.1-to 1.3-fold higher than signal intensities in the cytoplasm, the variant was 0.9-fold relative to cytoplasmic signal intensities, indicating a reduction of Astrin p.Q1012* at the kinetochore (Figure S9B).Tracking changes through time showed a steady reduction in p.Q1012* associated kinetochore intensities (Figure S9C).Thus, the severe reduction of Astrin p.Q1012* variant levels at kinetochores correlates well with the sustained reduction in microtubule-mediated pulling across sister kinetochores.

Astrin p.Q1012* expressing cells display prolonged mitosis
The kinetochore localization defect of Astrin p.Q1012* is similar to that observed in a C-terminal deletion mutant, Astrin D70, which impedes chromosome segregation. 35So we investigated the fate of mitotic cells expressing YFP-tagged Astrin p.Q1012* by generating a tetracyclineinducible HeLa FRT/TO TM YFP-Astrin Q1012* cell line and acquired time-lapse images every 6 min for 10 h in the presence of SiR-DNA (a DNA Lagging chromatids in anaphase tracker) following a brief exposure to tetracycline 43 (Figure 4A).On average 70% of Astrin p.Q1012* expressing cells completed mitosis compared to 95% of Astrin WT expressing cells (Figure S10A).Astrin p.Q1012* expressing cells that completed mitosis displayed anaphase onset (AO) delay.The time from nuclear envelope break down (NEBD) to AO was 1.5-fold longer in YFP-Astrin p.Q1012* expressing cells compared to YFP-Astrin WT expressing cells (96 min for p.Q1012* versus 60 min for WT), showing significant delay in AO (Figures 4B and  4C).Both the delay in AO and increased incidence of mitotic failure show that the Astrin p.Q1012* variant promotes chromosome missegregation.
To investigate the cause for the prolonged mitosis in Astrin p.Q1012* expressing cells, we analyzed chromosome congression using timelapse movies.Only an average of 56% of Astrin p.Q1012* expressing cells congressed their chromosomes compared to 90% of Astrin WT expressing cells (Figures 4B and 4D).Additionally, cells expressing Astrin p.Q1012* are 2-fold slower in chromosome congression compared to those expressing Astrin WT; (t50 of NEBD to metaphase: 48 min for p.Q1012* versus 24 min for WT; Figures 4B and 4D).Importantly, only an average of 30% of Astrin p.Q1012* expressing cells, maintain chromosome congression compared to 76% of Astrin WT expressing cells (Figure S10B).Thus, both the establishment and maintenance of congressed chromosomes are disrupted in YFP-Astrin Q1012* expressing cells, confirming defects in maintaining stable chromosome-microtubule attachment which can cause chromosome missegregation.

Astrin p.Q1012* expression promotes chromosome missegregation
To investigate whether the prolonged mitosis, reduced microtubule pulling and chromosome congression defects induced by Astrin p.Q1012* have an impact on chromosome segregation accuracy, we analyzed the presence of lagging chromosomes in anaphase cells.Time-lapse movies showed that an average of 70% of Astrin p.Q1012* expressing cells presented lagging chromosomes during anaphase compared to 15% of Astrin WT expressing cells (Figures 4B and 4E).Moreover, immunostaining studies showed that an average of 46% of Astrin p.Q1012* expressing anaphase cells display lagging chromatids compared to 9% in Astrin WT expressing cells (Figures 4F-4H).Thus, Astrin p.Q1012* variant significantly increases the incidence of missegregating chromosomes and lagging chromatids during anaphase.We conclude that the expression of the naturally occurring variant Astrin p.(Q1012*) despite the presence of Astrin full-length protein (as in heterozygous/monoallelic form) is likely to interfere with chromosomal stability in humans.Unlike previously reported Astrin variants (p.[(G1064E*3)]; [(K409Pfs*19)]) presenting clinical features (microcephaly) due to compromised centrosomal localization 20 here, we present the first assessment of Astrin variants' loss of kinetochore localization and its impact on chromosome segregation.

DISCUSSION
We present the first comprehensive survey of CIVa in chromosome segregation genes, and their allelic prevalence, exploiting genome sequencing efforts across multi-ethnic populations.Using a 3-step scalable framework, we predict and stratify CIVa candidates in chromosome segregation genes (Figure S11).We identify a rare LoF variant in the microtubule-associated outer-kinetochore protein Astrin p.Q1012*; this variant is harmful as it impairs the localization and function of endogenous protein in a dominant negative manner (Figure 4B), showcasing it as a CIN aiding variant in heterozygous (monoallelic) form.Second, we report a high-frequency Astrin p.L7Qfs*21 variant which expresses a shorter Astrin that localizes and functions normally at microtubule-ends and kinetochores, revealing alternate Kozak usage as a resilience mechanism to cope with harmful LoF variants in homozygous and heterozygous forms.Third, we report SKA3 p.Q70Kfs*7 which does not affect the function of endogenous full-length SKA3 but may be harmful in homozygous form.Thus, the CIVa database and the framework to stratify the impact of CIVa candidates can help shed light on the origins of CIN in a variety of pathologies.
We exploit the HeLa epithelial cell line that has a robust spindle checkpoint allowing a rapid quantitative assessment of subtle congression defects that promote CIN. 44,45Of the three Astrin variants we explore p.Q1012* is unique in disrupting chromosome alignment, reducing microtubule-mediated pulling and increasing chromosome missegregation (Figure S11).Unlike previous studies that show the role of Astrin C-terminal tail in preventing chromosome missegregation using conditions lacking endogenous Astrin, 35,42 the current study reveals the dominant negative impact of Astrin p.Q1012* variant in the presence of endogenous full length Astrin, a condition that closely mimics the naturally occuring monoallelic variant.Whether the p.Q1012* variant affects other cells or tissue types will be informative to explore using nontransformed cells from different tissues.
The SKA3 variant p.(Q70Kfs*7) introduces a premature termination leading to a truncated protein predicted to disrupt the multimerization domain within the SKA complex which can weaken microtubule attachments and disrupt chromosome alignment. 34,46,47In cells expressing the SKA3 variant p.(Q70Kfs*7), we find normal alignment of chromosomes and proper localization of endogenous SKA3 at kinetochores, suggesting normal assembly of the endogenous SKA complex.The large GFP tag could interfere with SKA3 variant Q70Kfs*7 function.Nevertheless, recent genomics and proteomics analysis of UK biobank samples show that of the 691 gene-level signals from protein truncating variants, 99.4% were associated with decreased protein levels. 48We propose inability to disrupt the larger SKA complex as an explanation for the high prevalence and tolerance of the SKA3 p.(Q70Kfs*7) variant in heterozygous form.
1][32] In addition, we report a homozygous Astrin start-loss variant (heterozygous/homozygous individuals: 223/12 GH, 1761/25 gnomAD and 21/0 GenomeAsia 100K and UK10K databases), 30,32,37 indicating its presence in multiple ancestries with a higher incidence among Europeans. 30Heterozygous Astrin start-loss and p.(L7Qfs*21) variants are also listed on TOPmed databases. 49,50How are these nonsense variants tolerated ?Our study of p.(L7Qfs*21) localization and analysis of alternate Kozak usage reveals a mitotically functional short isoform of Astrin exposing new pathways in cells that can compensate for the Astrin start-loss and p.L7Qfs*21 variants.Our single-cell studies show that the short isoform localizes better in the absence of full length endogenous Astrin, which is in alignment with the biallelic forms of the variant across different ancestries.The loss of Astrin's N-terminus may be non-pathogenic in humans, revealing resilience pathways that N-terminal premature stop codon variants of coiled-coil proteins may use by relying on shorter isoform expression.Thus, our single-cell studies to assess CIVa protein localization and their mitotic function provide a scalable framework, that can take advantage of Artificial Intelligence guided large-scale image analysis, 51 to exploit genetic variant prevalence across ancestries and to stratify CIVa relevant to a variety of CIN syndromes.

Limitations of the study
Here, we explore the impact of CIVa candidates using the HeLa cell line, a transformed basal carcinoma line expressing HPV oncogenes; 52 it will be insightful to consider these studies in nontransformed cell cultures without HPV gene expression. 41The framework developed here uncovered the first CIVa in the Astrin gene; this needs to be expanded to stratify CIVa candidates seen in cancers (e.g.,CENPC 31 or Histone genes 53 ) with varying extent of CIN.

Live-cell imaging
For live-cell imaging studies, cells were seeded onto 4-well cover glass chambered dishes (Lab-Tek; 1064716) and transferred to Leibovitz's L15 medium (Invitrogen;11415064) for imaging.For low-resolution live-cell imaging, HeLa FRT/TO YFP-Astrin cells were synchronised using a double thymidine block.100 nM sirDNA (Tebu-bio; SC007) was added 10 hours before image acquisition to stain for DNA.3Z-planes, 0.6 mm apart, were acquired using a 40X/0.95UPlanSApoair objective on an Applied Precision DeltaVision Core microscope equipped with a Cascade2 camera under EM mode.Imaging was performed at 37 C using a full-stage incubation chamber set upto allow normal mitosis progression and microtubule dynamics.
For high-resolution live-cell imaging, cells were transfected with plasmid vectors 24 hours before an hour-long 10 mM MG132 treatment to arrest mitotic cells in metaphase.3Z-planes, 0.6 mm apart, were acquired using a 100X/1.40UPlanSApooil immersion objective on an Applied Precision DeltaVision Core microscope equipped with a Cascade2 camera under EM mode.For live-cell CFP imaging, Applied Precision DeltaVision Elite microscope equipped with an EDGE sCMOS_5.5 camera with a 60X oil-immersion objective was used.Imaging was performed at 37 C using a full-stage incubation chamber set up to allow normal mitosis progression and microtubule dynamics.SoftWorxä distance measurement tool was used to find inter-centromeric distances.Additional analysis was conducted on Microsoft Excel and graphs were plotted using GraphPad Prism 9ä.

Kinetochore particle tracker
The Kinetochore-Particle-Tracker was developed in Python 3, using python's image processing library scikit-image in Anaconda Environment and Jupyter Notebook.Data analysis was done in RStudio with the package ggplot.Figure panels were generated using matplotlib, ggplot and jupyter-notebook.Initial image pre-processing was done in ImageJ.To measure the kinetochore intensities in 3D images of time-lapse movies, the CENPB-dsRed signal was first detected to identify the location of kinetochores by applying an edge detector filter and a suitable threshold.Small particles were removed, and the holes were filled by performing morphological operations.Next, by burning the CENPB signal mask on the YFP-Astrin channel image, we extracted the mean particle intensities of YFP-Astrin.The cytoplasmic intensity was measured by creating a binary mask to segment the cell from the background and identifying a ring-shaped region as a proxy for the cytoplasm.The source code is available for download at Github: https://github.com/Draviam-lab/Kinetochore-Particle-Tracker.

CIVa database
CIVa database was developed in GitHub Pages using JavaScript, HTML and CSS.This database can be queried on the Gene Symbol or the Uniprot ID.The Git Page is available at: (https://draviam-lab.github.io/CIVa/).The source code is available at Github (https://github.com/Draviam-lab/CIVa).

QUANTIFICATION AND STATISTICAL ANALYSIS
All experiments were repeated multiple times, as indicated in figure legends.Data were pooled and, if required, analyzed further in Microsoft Excel, and plotted in GraphPad Prism (v9.0;GraphPad Software, La Jolla, CA). Figure legends specify the n, errors, and the statistical test used.Data distributions were tested for normality using the D'Agostino-Pearson omnibus normality test and statistical differences among conditions were calculated using One-way ANOVA with DUNNET correction, Two-way ANOVA with Sidak correction, Chi-square test, Mann-Whitney U test (non-parametric) or paired t-test (parametric) in GraphPad Prism (v9.0;GraphPad Software, La Jolla, CA).Differences were considered significant if the p-value was <0.05 (*), <0.01(**), <0.001(***), or < 0.0001 (****), as indicated in each figure legend.

Figure 1 .
Figure 1.CIVa study reveals LoF kinetochore gene variants in healthy humans (A) Chromosomal instability variant (CIVa) analysis of SPAG5 gene (encoding Astrin) comparing its variant prevalence in COSMIC, gnomAD, and ELGH databases.Images of lollipop graphs show the positions and number of occurrences of different types of variations in the SPAG5 gene, including potential CIVa sites.A black dot indicates a truncating mutation, a green dot indicates a missense mutation and a purple dot indicates other mutation types.The red font indicates key LoF variants.Numbers in brackets indicate the number of homozygous occurrences.(B) Representative immunofluorescence images of Astrin wild type and p.Q1012* expressing cells treated as in Figures 1 and S2A and probed for GFP and CREST.DNA was stained with DAPI.Scale bars: 5 mm in uncropped images and 1 mm in insets.(C).Violin plot showing differences in YFP-tagged Astrin or pQ1012* intensities at the kinetochores.Ratio of kinetochore and cytoplasmic intensities are shown.Colors represent independent experiments.Mann-Whitney U test was performed for statistical significance.''****'' represents p < 0.0001.(D) Representative deconvolved images of live mitotic cells coexpressing CENPB-DsRed (centromere marker) and N-terminal GFP-tagged SKA3 wild-type or p.Q70Kfs*7 as indicated (n = 24 wild-type and 15 p.Q70Kfs*7 mitotic cells).(E) Violin plot showing differences in GFP tagged SKA3 wild-type or p.Q70Kfs*7 intensities on individual kinetochores, marked using centromeric marker CENPB-DsRed in metaphase arrested cells as indicated in D (n values of cells as indicated).Ratio of kinetochore and cytoplasmic intensities are shown.Circles represent individual kinetochores and colors represent independent experiments.Mann-Whitney U test was performed for statistical significance.''****'' represents p < 0.0001.The black bars represent mean values.Scale bars: 5 microns in uncropped images and 1 micron in insets.

Figure 2 .
Figure2.High-frequency SKA3 p.Q70Kfs*7 does not disrupt end-on chromosome-microtubule attachment (A) Representative deconvolved images of live mitotic cells coexpressing mKate2-Astrin (end-on attachment marker) along with N-terminally tagged GFP-SKA3 wild-type or p.Q70Kfs*7 or C-terminally tagged SKA3-CFP wild-type or p.Q70Kfs*7 as indicated (n = 15 wild-type and 12 p.Q70Kfs*7 mitotic cells).Samples collated from at least 3 independent experimental repeats.Scale bars: 5 mm in uncropped images and 1 mm in insets.(B) Representative immunofluorescence images of GFP tagged SKA3 wild-type or p.Q70Kfs*7 expressing cells treated with MG132 and probed using anti-GFP and anti-SKA3 antibodies and CREST antisera.Anti-SKA3 antibody recognizes 156-177 a.a which follows the premature stop codon in SKA3 p.Q70Kfs*7.(C) Cartoon of a three-step scalable framework used to predict, mimic and stratify CIVa on the basis of their impact on kinetochore protein localization and function assessed using microscopy assays.Bioinformatic predictions on LoF variants in kinetochore genes and their allelic prevalence across COSMIC, gnomAD, and ELGH databases are collated into the CIVa database.CIVa in kinetochore genes are mimicked in single cells using different protein expression tools and their impact is quantitatively assessed to stratify harmful and harmless variants in monoallelic and biallelic forms.

Figure 3 .%
Figure 3. Short Astrin isoform localizes on microtubules and at kinetochores (A) Representative immunofluorescence images of Astrin wild type and Astrin p.7* expressing cells treated with MG132 and immunostained using anti-GFP antibody and CREST antisera (as a kinetochore marker).DNA was stained with DAPI.Scale bars: 5 mm in uncropped images and 1 mm in insets.(B) Violin plot showing YFP-Astrin intensity at the kinetochores.The solid line is the median and the dotted lines are quartiles.Circles represent individual kinetochores and colors represent different sets.Mann-Whitney U test was performed for statistical significance.''***'' represents p < 0.001.(C) Boxplot showing Astrin localization at the kinetochores scored as high, medium, low and no crescents (as in Figure S2A).Symbols represent independent experiments.A chi-square test was performed for statistical significance.''****'' represents p < 0.0001.(D) Immunoblots of HeLa cell lysates expressing Astrin-GFP (wild type and p.7*) and YFP-Astrin (wild type, D151 and D274) and probed for GFP and Astrin.Yellow and green asterisks refer to bands corresponding to GFP fusion protein and endogenous Astrin, respectively.Purple asterisk refer to non-specific bands in anti-GFP immunoblot.(E) Representative deconvolved images of live-cells co-expressing YFP-tagged Astrin wild type and either D151 or D274 Astrin mutant following MG132 treatment.Scale bars: 5 mm in uncropped images and 1 mm in insets.(F) Violin plot showing the ratio of D151 or D274 Astrin mutant and Astrin wild-type intensities at kinetochores.The solid line is the median and the dotted lines are quartiles.Circles represent individual kinetochores.Mann-Whitney U test was performed for statistical significance.''****'' represents p < 0.0001.(G)Representative immunofluorescence images of YFP-tagged Astrin wild type, D151 or D274 expressing cells immunostained using antibodies against GFP (for YFP tag) and the growing microtubule-end marker, EB1 that associates with Astrin-SKAP complex.40,41Scale bars: 5 mm in uncropped images and 5 mm in insets.

Figure 4 .
Figure 4. Astrin p.Q1012* prolongs mitosis and increases the incidence of lagging chromosomes (A) Experimental regimen showing methodology for double thymidine-based cell cycle synchronization and controlled Tet inducible expression and imaging of Astrin wild-type or variant expressing cells.(B) Representative time-lapse images of Astrin wild type and p.Q1012* cells treated as in A. Arrows mark chromosomes that fail to remain congressed and anaphase lagging chromatids.NEBD is nuclear envelope breakdown and AO is anaphase onset.Scale bars: 15 mm.(C) Cumulative frequency graph showing the time taken from NEBD to AO. T50 indicates the time taken by 50% of cells to complete mitosis.(D) Cumulative frequency graph showing the time taken from nuclear envelope break down (NEBD) to the formation of the metaphase plate.T50 indicates the time taken by 50% of cells to congress chromosomes.(E) Violin plot showing the percentage of cells with lagging chromosomes at anaphase.The solid line represents the median and the dotted lines represent the quartiles.Each dot represents an independent set.Paired t-test was performed to find statistical significance.''*'' represents p < 0.05.(C-E) Data represent four independent sets.(F) Experimental regimen (G).Representative immunofluorescence images of Astrin wild type and p.Q1012* expressing anaphase cells with lagging chromosomes.Cells were treated as in F and probed for GFP and CREST.DNA was stained with DAPI.Scale bars: 5 mm in uncropped images and 1 mm in insets.(H) Violin plot showing the percentage of anaphase cells with lagging chromosomes.The solid line represents the median and the dotted lines represent quartiles.Each dot represents an independent experiment.One-way ANOVA with DUNNET correction was performed to find statistical significance.''*'' and ns represent p < 0.05 and ''not significant'', respectively.