β-tubulin phosphorylation by Chk1 is required for normal spindle formation during cell division

Abstract

The mitotic spindle is a microtubule-based apparatus that is responsible for accurate segregation of chromosomes into two daughter cells. In this study, we show that the DNA damage kinase Chk1 is required for optimal density and efficient nucleation of spindle microtubules during unperturbed mitosis in vertebrate cells. Chk1 phosphorylates β-tubulin at the identified conserved site threonine-285 (T285) in vitro, and at mitotic centrosomes in prometaphase and metaphase. Impaired β-tubulin-T285 phosphorylation correlates with improper spindles, delayed anaphase onset, erroneous chromosome alignment and segregation, unequal daughter cell-size and reduced cell proliferation. The ATR-interacting protein ATRIP promotes localization of ATR kinase and the mediator protein TopBP1 to mitotic centrosomes; furthermore, interaction of ATRIP with ATR and TopBP1 is required for Chk1 activation and β-tubulin-T285 phosphorylation. These results identify a signaling pathway that promotes spindle maturation and function in human cells, through Chk1-mediated β-tubulin-T285 phosphorylation.

Chk1, the DNA damage signaling kinase, has also a role in mitosis, where it phosphorylates beta-tubulin. This post-translational modification stabilizes the microtubules of the mitotic spindle to ensure faithful chromosome segregation.

Introduction

During mitosis, the cell must accurately divide its replicated chromosomes into two daughter cells, a task performed by a microtubule-based machine called the mitotic spindle^1–3. Despite its identification almost 150 years ago, understanding how the mitotic spindle assembles and performs its functions remains a challenge^4,5. Errors in spindle formation can lead to chromosome missegregation and aneuploidy, spindle misorientation or improper daughter cell-size, which are associated with genetic disorders, human birth defects, or carcinogenesis^6–10. As a result, investigating the mechanisms that regulate spindle formation can help us better understand the causes of serious human diseases.

The mitotic spindle consists primarily of filaments called microtubules (MTs)¹¹. MTs are composed of αβ-tubulin heterodimers that associate head-to-tail to form protofilaments and typically 13 protofilaments are connected side-by-side to form the hollow MT cylinder^12,13. In animal cells, the bipolar spindle consists of three different MT subclasses¹⁴: kinetochore-MTs connect sister chromatids to the spindle poles to mediate chromosome alignment at the metaphase plate and sister chromatid segregation during anaphase. Overlap (interpolar) MTs on the other hand, emanate from the centrosome and extend toward the center of the spindle where they form antiparallel bundles with overlap MTs from the opposite pole, whereas astral MTs extend from the poles outwards to anchor the spindle within the cytoplasm¹⁴. By signaling to the cell cortex, non-kinetochore MTs regulate spindle positioning in metaphase and early anaphase, define the location of the cleavage plane and deliver proteins required for dividing the cell into two in cytokinesis^1,2,14.

At the early stages of mitosis, MT bundles emanating from the centrosome grow and shrink, thereby exploring the cellular space, until their plus-ends become attached to the paired sister kinetochores to form kinetochore-fibers^15–17. This search and capture mechanism leads to the assembly of the bipolar spindle and the establishment of bipolar kinetochore-MT attachments^15,18,19. Then, forces exerted to kinetochores by kinetochore-fibers and on chromosome arms by plus end-directed kinesin motors promote chromosome congression and alignment on the metaphase plate^20–23. In anaphase, changes in kinetochore-fiber dynamics promote their shortening leading to chromosome segregation^24–26. Lastly, cytokinesis physically divides the cell to produce two genetically identical daughter cells²⁷.

To avoid chromosome segregation errors, the mitotic spindle has a built-in checkpoint mechanism, the mitotic spindle checkpoint, which delays anaphase onset until all sister kinetochores are bipolarly attached to kinetochore-MTs^28,29. Inefficient kinetochore-MT interactions can delay anaphase due to persistent spindle checkpoint activation thus increasing the time spent in mitosis and reducing cell proliferation^17,30. Elevated MT assembly rates on the other hand can cause transient spindle geometry defects and promote chromosome missegregation, perhaps by overwhelming the capacity of the cellular error correction machinery³¹.

The position and orientation of the mitotic spindle during metaphase and early anaphase also define the site of cytokinesis cleavage furrow formation²⁷. In animal cells, the central spindle and astral MTs specify the position of the cytokinesis furrow by regulating the formation of the actomyosin contractile ring to remodel the plasma membrane^32–34. In most vertebrate cells, the cytokinesis furrow is positioned in the middle of the cell, resulting in equal-sized daughter cells³⁵. In response to uncentered spindles in metaphase or early anaphase in symmetrically dividing animal cells, cortical dynein becomes displaced from the cell cortex near the proximal spindle pole, causing the mitotic spindle to move towards the cell-center by pulling forces exerted on astral MTs of the distal pole^36–40. For this purpose, Polo-like kinase 1 (Plk1) kinase activity at the proximal spindle pole inhibits the interaction of the dynein-dynactin complex, which mediates spindle movement and centering, with receptor proteins at the cell cortex in a distance-dependent manner^39,41. Furthermore, Ran-GTP gradients promote the correct localization of dynein-receptor proteins near the spindle poles to promote proper spindle positioning^35,39,42,43. Because daughter cell-size affects the relative amount of cytoplasmic factors and organelles, daughter cell-size control can play a critical role in cell physiology and could protect against carcinogenesis or developmental defects^8,10,35.

The centrosome is the primary MT-organizing center in animal cells. It is comprised of two centrioles surrounded by an amorphous pericentriolar material containing hundreds of proteins, including those that help to organize and nucleate, i.e., initiate MT formation from individual tubulin dimers, MTs^44,45. A major component of the pericentriolar material is γ-tubulin which, together with gamma tubulin complex proteins (GCPs) 2 and 3, forms the γ-tubulin small complex (γ-TuSC)⁴⁶. Several γ-TuSCs form the large γ-tubulin ring complex (γ-TuRC), which includes GCPs 4, 5 and 6⁴⁶. The γ-TuRC provides a template onto which tubulin heterodimers are added thus promoting MT polymerization^47–50. Chromatin-induced MT formation^51–56 or MT nucleation from pre-existing MTs within the spindle also contribute to spindle assembly^57–60. Once nucleated, MT growth and shrinkage are regulated by microtubule-associated proteins (MAPs), such as MT polymerases, which promote MT growth by favoring a straight tubulin oligomers conformation within the MT lattice, and MT depolymerases, which promote MT catastrophe by favoring the intrinsic curved conformation of the tubulin dimer^61–65. The tubulin dimer is also subject to posttranslational modifications, such as acetylation, detyrosination/tyrosination, polyglutamylation, polyglycylation or phosphorylation (reviewed in refs. ^66–69; however, in most cases, the biological roles of these modifications are incompletely understood.

The conserved ataxia-telangiectasia mutated- and Rad3-related (ATR) and Checkpoint kinase 1 (Chk1) kinases are master regulators of the DNA damage response in mammalian cells^70–72. When DNA replication is inhibited, for example due to ultraviolet radiation-induced DNA lesions, inhibition of DNA polymerases or deoxyribonucleoside triphosphate (dNTP) depletion, DNA polymerases become uncoupled from the replicative helicase; as a result, tracts of single-strand DNA are generated and coated with the heterotrimeric Replication Protein A (RPA)⁷³. ATR is recruited to such tracts through association of its partner protein ATR-interacting protein (ATRIP) with RPA-coated single-strand DNA^74–77 and is activated by interaction with DNA Topoisomerase II-binding protein 1 (TopBP1)^78,79. Catalytically active ATR phosphorylates Chk1 at multiple sites including serines 317 and 345 within the C-terminal regulatory domain, and these phosphorylations are essential for Chk1 activation^80–83. In turn, active Chk1 dissociates from chromatin and phosphorylates several protein substrates to delay cell cycle progression and implement the DNA damage response^70,71. In the absence of DNA damage, Chk1 phosphorylates the Cell Division Cycle 25B (Cdc25B) phosphatase at centrosomes and/or inside the nucleus in interphase cells, to prevent premature entry to mitosis^84–86. Chk1 and ATR also exhibit non-canonical roles during cell division⁸⁷. More specifically, Chk1 is required for efficient chromosome alignment and segregation during unperturbed mitosis; furthermore, Chk1 phosphorylates Aurora B kinase at serine 331 (S331) at prometaphase kinetochores to fully promote Aurora B catalytic activity and spindle checkpoint signaling^88–90. ATR is activated at centromeric R-loops in prometaphase and is required for accurate chromosome segregation and complete Aurora B activation, presumably through Chk1-mediated Aurora B phosphorylation^91,92. Also, ATR, ATRIP and TopBP1 localize to mitotic centrosomes^93–95; however, the biological importance of this localization is unclear.

In the present study, we show that, in unperturbed mitosis, ATR, ATRIP and TopBP1 functionally interact to activate Chk1 at mitotic centrosomes. In turn, Chk1 phosphorylates β-tubulin at the identified conserved site threonine 285 (T285) in prometaphase and metaphase to promote optimal density and efficient nucleation of spindle MTs in vertebrate cells. This is a function of Chk1 in promoting mitotic spindles that is separate from its previously established roles in regulating mitotic entry or phosphorylating Aurora B-S331, and is independent of DNA damage. Improper mitotic spindles in the absence of β-tubulin-T285 phosphorylation correlate with erroneous chromosome alignment and segregation, delayed anaphase onset, generation of unequal-sized daughter cells and reduced cell proliferation. These results identify a mechanism that promotes normal spindle formation and function, by rewiring DNA damage signaling proteins at centrosomes to impose β-tubulin-T285 phosphorylation to promote efficient nucleation of spindle MTs.

Results

Chk1 is required for optimal density of spindle MTs in several vertebrate cell lines

Chk1 is required for mitotic cell division, we therefore investigated a potential role for Chk1 in the mitotic spindle. Human colon carcinoma BE cells depleted of Chk1 by siRNA (siChk1) or treated with the selective Chk1-inhibitor UCN-01 (Chk1i) exhibited reduced density of spindle MTs in prometaphase compared with controls (Fig. 1a, b), and this was not due to reduced total levels of α- or β-tubulin (Supplementary Fig. 1a-c). Expression of Chk1 fused to GFP (GFP:Chk1^R) resistant to degradation by a second Chk1-2 siRNA (siChk1-2) by introduction of appropriate silent point mutations, but not expression of GFP-only, rescued MT density after depletion of the endogenous Chk1 compared with GFP controls, showing the effect was Chk1-specific (Fig. 1c–e and Supplementary Fig. 1d). Treatment of Chinese hamster ovary cells stably expressing WT Aurora B (CHO^WT)⁹⁰, or treatment of human embruonic kidney (HEK 293 T; HEK) cells with UCN-01 (Chk1i) or with a second small molecule Chk1 inhibitor (SB218078, Chk1i-2) also diminished the density of spindle MTs in prometaphase compared with controls (Fig. 1f, g, and Supplementary Fig. 1e, f). Furthermore, avian b-lymphoma DT40 cells in which Chk1 was ablated by gene targeting (Chk1-/-)⁹⁶ exhibited reduced spindle MTs compared with Chk1 + /+ controls (Supplementary Fig. 1g, h). These results show that, although mitotic spindles form in Chk1-deficient cells, they are less robust compared with controls, suggesting that Chk1 is required for spindle maturation in various vertebrate cell lines.

Fig. 1. Chk1 inhibition reduces the density of spindle MTs.

a, b Prometaphase spindles and relative α-tubulin fluorescence intensity in control BE cells, cells transfected with Chk1 siRNA (siChk1) or with a combination of Chk1 and Cdc25B siRNAs (siChk1+siCdc25B), or treated with 300 nM UCN-01 (Chk1i) for 5 h. ***P = 2.92523E-19 (control vs siChk1), 1.51241E-24 (control vs Chk1i) by ANOVA and Student’s t test. c–e Prometaphase spindles and relative α-tubulin fluorescence intensity in BE cells expressing siRNA-resistant GFP:Chk1^R or GFP-only, in the absence (control) or presence of a second Chk1 siRNA (siChk1-2). ***P = 1.25652E-10 (GFP control vs GFP+siChk1-2) by ANOVA and Student’s t test. f, g Prometaphase spindles and relative α-tubulin fluorescence intensity in CHO cells expressing wild-type (WT) or mutant S331E Aurora B after induction with Tetracycline, in the absence (control) or presence of Chk1i. ***P = 1.2116E-20 (CHO^WT control vs CHO^WT +Chk1i), 3.07985E-20 (CHO^S331E control vs CHO^S331E +Chk1i) by ANOVA and Student’s t test. h Protocol for induction of DNA damage in mitotic cells. i,j Prometaphase spindles and relative α-tubulin fluorescence intensity in BE cells after induction of DNA damage by treatment with 10 μM etoposide. ***P = 2.04716E-22 (control vs Chk1i) by ANOVA and Student’s t test. k Relative γ-H2AX fluorescence intensity in control or Chk1i cells after etoposide-treatment. l Protocol for MT re-growth after cold treatment followed by fluorescence microscopy. m–o Mitotic spindles and relative α-tubulin fluorescence intensity in cells recovering from ice-cold medium in the absence (control) or presence of Chk1i. Values in control 10 min were set to 1. ***P = 1.8163E-31 (control vs Chk1i 2 min), 2.58819E-46 (control vs Chk1i 4 min), 2.67264E-27 (control vs Chk1i 6 min), 2.81557E-39 (control vs Chk1i 8 min), 2.75807E-39 (control vs Chk1i 10 min) by ANOVA and Student’s t test. Numbers next to each bar indicate n. Error bars show the SD from the mean. Bars, 5 μm.

Because Chk1 activity can delay onset of mitosis, we investigated whether improper spindles in Chk1-deficient cells were caused by premature mitotic entry^84,86,97. Simultaneous depletion of Chk1 and Cdc25B by siRNA (siChk1+siCdc25B) in BE cells to delay mitotic entry⁸⁵, did not rescue spindle MT density compared with cells depleted of Chk1-only (Fig. 1a, b and Supplementary Fig. 1i). As a positive control for the double Chk1/Cdc25B siRNA-treatment, we examined spindle rotation in metaphase cells after treatment with the proteasome inhibitor MG132 for 1 h⁸⁵. Control cells exhibited a spindle angle (θ) of approximately 3 degrees corresponding to spindle poles (marked by pericentrin) located on the same image plane (slice) and to mitotic spindles parallel to the apical cell surface (Supplementary Fig. 1j–l). Chk1-depletion on the other hand, increased spindle rotation to an average of 17 degrees, resulting in spindle poles located to different image slices; however, simultaneous depletion of Chk1 and Cdc25B restored spindle rotation compared with controls (Supplementary Fig. 1l), showing that double Chk1/Cdc25B depletion works in our hands⁸⁵. These results suggest that reduced spindle MT density in Chk1-deficient cells was not caused by premature entry to mitosis.

Aurora B catalytic activity is required for optimal spindle formation⁵⁵. In CHO cells, expression of phosphomimetic Aurora B-S331E in which Aurora B-S331 is mutated to glutamic acid (CHO^S331E) did not rescue spindle MT density after Chk1 inhibition compared with controls (Fig. 1f, g), indicating that reduced spindle MT density in Chk1-deficient cells was not caused by impaired Aurora B-S331 phosphorylation^90,98.

To investigate whether interphase defects can affect MT dynamics in Chk1-deficient cells⁹⁹, mitotic cells were isolated by treating BE cells with nocodazole followed by shake-off (Supplementary Fig. 1m). Mitotic cells treated with Chk1i for 1 h exhibited reduced spindle MT density and this was not due to reduced total levels of α- or β-tubulin compared with controls, showing that Chk1 promotes spindle MTs in a mitosis specific manner (Supplementary Fig. 1n–q).

MT dynamics can also be affected by DNA damage¹⁰⁰. To investigate whether regulation of spindle MT density by Chk1 is linked to DNA damage, cells were arrested in prometaphase by nocodazole and treated with etoposide for 2 h in the continuous presence of nocodazole to induce DNA damage in mitosis (Fig. 1h). After release from nocodazole and etoposide, Chk1-inhibition by Chk1i reduced spindle MTs compared with controls in the presence of similar levels of DNA damage as determined by γ-H2AX staining (Fig. 1i–k). Furthermore, asynchronous cells treated with Chk1i for 1 h without detectable γ-H2AX staining exhibited reduced spindle MTs compared with controls (Supplementary Fig. 2a, b). These results show that Chk1 promotes spindle MT density both in the presence or absence of damaged DNA, suggesting that Chk1 regulates spindle MT formation independently of DNA damage.

Cdk1 regulates MT dynamics¹⁰¹ and Chk1 can sustain Cdk1 activity in mitosis by preventing Cdk1 inhibitory phosphorylations¹⁰². Importantly, expression of Cdk1:GFP-AF, harboring nonphosphorylatable mutations of T14 to alanine and Y15 to phenylalanine⁹⁸, or wild-type (WT) Cdk1:GFP did not rescue spindle MT intensity after Chk1-inhibition compared with controls, indicating that Chk1 regulates spindle MT density independently of Chk1 effects on mitotic Cdk1 activity (Supplementary Fig. 2c–e).

In conclusion, these results suggest that Chk1 is required for proper mitotic spindle density in vertebrate cells independently of DNA damage or previously established Chk1 functions in mitosis, such as regulation of mitotic entry, phosphorylation of Aurora B-S331, or maintenance of mitotic Cdk1 activity.

Chk1 promotes spindle MT nucleation

To investigate a potential role for Chk1 in spindle MT nucleation, BE cells were treated with ice-cold medium for 1.5 h to completely depolymerize spindle MTs and, after recovery of cells in warm medium, MT re-growth in prometaphase in the absence or presence of Chk1-inhibitor was examined by fluorescence microscopy (Fig. 1l). Chk1-deficient cells exhibited reduced MT re-growth after cold treatment compared with controls (Fig. 1m–o), suggesting that Chk1 promotes efficient spindle MT nucleation.

Active Chk1 localizes to mitotic centrosomes

In control cells, Chk1 fused to GFP (GFP:Chk1) or phosphorylated (active) Chk1-serine 345 (pS345) localized to centrosomes (marked by γ-tubulin) in cells in late S-/ G2-phase (exhibiting uncondensed chromatin with duplicated centrosomes that were close together), and also in prophase, prometaphase and metaphase, whereas localization of active Chk1 at centrosomes was reduced in anaphase cells (Fig. 2a–c and Supplementary Fig. 2f). Chk1 was also detected at prometaphase kinetochores (marked by Hec1 staining) when the Chk1 signal was relatively overexposed, in agreement with previous findings (Supplementary Fig. 2g)⁸⁸. These results suggest a role for Chk1 at mitotic centrosomes from prophase to metaphase.

Fig. 2. Expression of nonphosphorylatable T285A GFP:β1-tubulin reduces spindle MT density.

a Localization of Chk1 fused to GFP (GFP:Chk1) in prometaphase. b,c Localization of phosphorylated Chk1-S345 (pS345) and relative pS345 fluorescence intensity at centrosomes in BE cells. Mean ± SD from n cells. Values in prometaphase cells were set to 1. Numbers next to each bar indicate n. ***, P = 2.83529E-17 (prometaphase vs anaphase), 6,60308E-07 (metaphase vs anaphase) by ANOVA and Student’s t test. d Co-immunoprecipitation from mitosis-enriched cell extracts. Precipitated β- tubulin (β-tub) or Chk1 were detected by Western blotting. Ab, antibody; NS, nonspecific. Molecular weights were inferred by aligning the autoradiography film with the corresponding nitrocellulose membrane containing the protein samples and a colorimetric molecular weight marker. e Chk1 kinase assay using recombinant Chk1 and purified fragments of human α- or β1-tubulin fused to GST, or GST-only as substrates. Autoradiography (³²P) and western blotting (WB) of the same gel. Asterisks mark the expected size of the corresponding full-length proteins. f Sequence alignment of β1/β- tubulin from various organisms. The identified Chk1 phosphorylation sites are in red and T285 is boxed and indicated by an arrowhead. Amino acid numbers refer to the human protein. g–j Prometaphase spindles and relative GFP or α-tubulin fluorescence intensity in HEK cells expressing siRNA-resistant wild-type (WT), T274A, S278A, T285A, T285E or T290A GFP:β1-tubulin, in the presence of β1-tubulin siRNA (si β1-tubulin) i ***, P = 6.58944E-10 (GFP:β1-tubulin^R WT+si β1-tubulin vs GFP:β1-tubulin^R T285A+si β1-tubulin), 8.44181E-10 (GFP:β1-tubulin^R T285A+si β1-tubulin vs GFP:β1-tubulin^R T285E+si β1-tubulin) by ANOVA and Student’s t test. j ***, P = 2.96969E-13 (GFP:β1-tubulin^R WT+si β1-tubulin vs GFP:β1-tubulin^R T285A+si β1-tubulin), 5.97209E-11 (GFP:β1-tubulin^R T285A+si β1-tubulin vs GFP:β1-tubulin^R T285E+si β1-tubulin) by ANOVA and Student’s t test. k–m Localization of phosphorylated β1-tubulin-T285 (pT285) and pT285 fluorescence intensity at centrosomes in BE cells in the absence (control) or presence of Chk1 siRNA (siChk1) or si β1-tubulin (si β1-tub), or after treatment of cells with 300 nM UCN-01 (Chk1i) for 5 h. Mean ± SD from n cells. Values in WT or control cells were set to 1. Numbers next to each bar indicate n. ***P = 4.29586E-39 (control vs Chk1i), 1.9655E-26 (control vs siChk1), 2.46118E-39 (control vs si β1-tub) by ANOVA and Student’s t test. Bars, 5 μm. Inset scale bars, 1 μm.

Chk1 phosphorylates β-tubulin in vitro

Endogenous Chk1 co-precipitated with soluble β-tubulin from cell extracts, after enrichment of cells in prometaphase by release from a nocodazole block (Fig. 2d). To investigate whether α/β-tubulin is a potential substrate of Chk1, purified fragments of human α1- or β1-tubulin fused to GST were incubated with recombinant Chk1 in kinase reactions in vitro. Chk1 phosphorylated GST:β1 tubulin (amino acids 254-444) relatively strongly, compared with GST:β1 tubulin (1-253), GST:α1 tubulin (1-200) and (201-344), or GST-only (Fig. 2e). To map the Chk1 phosphorylation sites on β1-tubulin, phosphorylated GST:β1-tubulin (254-444) was analyzed by liquid chromatography-mass spectrometry and four novel Chk1 phosphorylation sites, namely threonine 274 (Τ274), serine 278 (S278), threonine 285 and threonine 290 (T290) were identified (Supplementary Fig. 2h). These sites are conserved in higher eukaryotes and the budding yeast (Fig. 2f) and are inside a protein sequence that is 95-100% identical in all human β-tubulin isoforms, but is only 30% identical between α- and β-tubulin (Supplementary Fig. 2i).

Expresion of nonphosphorylatable β1-tubulin-T285A impairs spindle MT density

To investigate the importance of β-tubulin phosphorylation by Chk1, human wild-type (WT) or nonphosphorylatable mutant β1-tubulin proteins, in which the individual Chk1 phosphorylation sites were changed to alanine, fused to GFP were constructed and made resistant to degradation by introduction of silent point mutations inside the β1-tubulin siRNA target sequence (GFP:β1-tubulin^R). After depletion of the endogenous protein, expression of T285A or GFP-only, but not T274A, S278A or T290A, GFP:β1-tubulin^R reduced the density of spindle MTs in prometaphase, as determined by GFP- or α-tubulin staining, compared with cells expressing the wild-type (WT) or a phosphomimetic T285 to glutamic acid (T285E) mutant transgene (Fig. 2g–j; Supplementary Fig. 2j–n, and Supplementary Fig. 3a, b). These results suggest that phosphorylation of β-tubulin-T285, but not T274, S278 or T290, is required for optimal spindle MTs. T285 conforms to the minimum Chk1 consensus sequence R-x-x-S/T (Fig. 2f)¹⁰³ and is inside the β-tubulin M-loop (microtubule loop) connecting beta strand S3 and helix H9, which is the central element for the formation of lateral interactions between protofilaments and an essential part of the taxol-binding pocket (Supplementary Fig. 3c)^13,104–106.

Chk1 phosphorylates β-tubulin-T285 at centrosomes in prometaphase

To investigate β-tubulin-T285 phosphorylation in mitosis, an antiphospho-β1-tubulin-T285 antiserum was raised against the human protein sequence. To verify T285-phosphorylation and demonstrate the specificity of the antiserum, bacterially expressed WT or nonphosphorylatable mutant T285A GST:β1-tubulin (254-444) were used as substrates in Chk1 in vitro kinase assays and T285 phosphorylation was detected by western blotting. WT GST:β1-tubulin (254-444) was not pre-phosphorylated in the absence of Chk1; furthermore, mutation of T285 to alanine or inhibition of Chk1 activity by UCN-01 (Chk1i) impaired T285 phosphorylation compared with the WT protein+Chk1 (Supplementary Fig. 3d, e). These results show that Chk1 phosphorylates β1-tubulin-T285 in vitro.

Phosphorylated β-tubulin-T285 (pT285) localized to centrosomes in control cells in prometaphase (Fig. 2k) by confocal microscopy; furthermore, this localization was unaffected when cells were extracted in the presence of the protein phosphatase inhibitor microcystin^107,108, suggesting that pT285 is not sensitive to phosphatase activity (Supplementary Fig. 3f). Depletion of Chk1 or β1-tubulin by siRNA or inhibition of Chk1 catalytic activity by Chk1i diminished phospho-T285 staining compared with control cells (Fig. 2k–m). Inhibition of Cdk1 by treatment of mitotic cells with the Cdk-inhibitor roscovitine (Cdki) in the presence of the proteasome inhibitor MG132 to block progression beyond metaphase, did not reduce T285 phosphorylation at centrosomes, indicating that T285 is not phosphorylated by Cdk1 (Supplementary Fig. 3g–i). Phospho-T285 staining was impaired after incubation of the anti-pT285 antiserum (Ab) with the phosphorylated peptide phospho-T285 compared with the unphosphorylated peptide T285 synthetic peptides (Supplementary Fig. 3j, k); furthermore, in cells depleted of the endogenous β1-tubulin, expression of nonphosphorylatable mutant T285A GFP:β1-tubulin^R impaired T285 phosphorylation at centrosomes compared with controls expressing the WT protein, verifying that the anti-pT285 antiserum is specific for the phosphorylated β-tubulin-T285 by immunofluorescence (Supplementary Fig. 3l, m). Also, in control cells recovering from ice-cold medium (Fig. 1l), phosphorylated β-tubulin-T285 localized to and around the centrosomes (marked by pericentrin) and colocalized with α-tubulin at spindle poles; furthermore, T285 phosphorylation was diminished in Chk1-deficient cells (Fig. 3a–c), suggesting that Chk1 phoshorylates β-tubulin at spindle poles from relatively early on during spindle formation. Chk1 colocalized with phospho-T285 at spindle poles in prometaphase cells (Supplementary Fig. 4a). Also, T285 phosphorylation at centrosomes was relatively high in prometaphase and metaphase, but was reduced in anaphase, telophase-late cytokinesis, or interphase cells (Fig. 3d–f). These results show that Chk1 phosphorylates β-tubulin-T285 at centrosomes in prometaphase and metaphase.

Fig. 3. Expression of phosphomimetic T285E GFP:β1-tubulin rescues spindle MT density in Chk1-deficient cells.

a–c Localization of phosphorylated β1-tubulin T285 (pT285) and relative pT285 fluorescence intensity in BE cells recovering from ice-cold medium in the absence (control) or presence of 300 nM UCN-01 (Chk1i). Mean ± SD from n cells. Values in control were set to 1. ***P = 7.89586E-18 (control vs Chk1i) by ANOVA and Student’s t test. d–f Localization and relative pT285 fluorescence intensity in BE cells at different stages of mitosis or in interphase. Mean ± SD from n cells. Values in prometaphase were set to 1. Arrows indicate centrosomes. Insets show high magnifications of centrosomes or spindle poles. Inset scale bars, 1 μm. **P = 0.002193698 (prometaphase vs metaphase); ***P = 4.4776E-32 (prometaphase vs anaphase), 4.17967E-44 (prometaphase vs telophase-late cytokinesis), 1.21694E-69 (prometaphase vs interphase), 3.03195E-23 (metaphase vs anaphase), 4.71907E-13 (anaphase vs telophase-late cytokinesis) by ANOVA and Student’s t test. g–i Prometaphase spindles and relative GFP fluorescence intensity. HEK cells expressing siRNA-resistant wild-type (WT), T285A or T285E GFP:β1-tubulin in the presence of β1-tubulin siRNA (si β1-tubulin), were untreated or treated with 300 nM UCN-01 (Chk1i) for 5 h. Mean ± SD from n cells. Values in WT were set to 1. ***P = 2.08845E-20 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R WT + Chk1i + si β1-tubulin), 2.09372E-19 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin), 4.6577E-24 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin), 1.63587E-21 (GFP:β1-tubulin^R WT + Chk1i + si β1-tubulin vs GFP:β1-tubulin^R T285E + Chk1i + si β1-tubulin) by ANOVA and Student’s t test. j–m Mitotic spindles and relative α-tubulin fluorescence intensity in cells recovering from ice-cold medium. Values in WT 15 min were set to 1. Numbers next to each bar indicate n. Error bars show the SD from the mean. ***P = 1.85494E-09 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 5 min), 1.33617E-05 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 5 min), 1.73179E-09 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 10 min), 6.51642E-08 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 10 min), 6.90146E-18 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 15 min), 1.38161E-09 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 15 min), 4.16017E-12 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 20 min), 2.21822E-08 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 20 min) by ANOVA and Student’s t test. Bars, 5 μm. n Chk1 kinase and MT sedimentation assay. Recombinant Chk1 was incubated in kinase reactions with 1 μM purified porcine brain tubulin as substrate, in the absence or presence of exogenous ATP. After tubulin polymerization and centrifugation, supernatants (S) and pellets (P) were analysed for phosphorylated β-tubulin-T285 (pT285) or total α-tubulin by western blotting.

Phosphorylation of β-tubulin-T285 by Chk1 promotes spindle MT density

In HEK cells depleted of the endogenous β1-tubulin, expression of the phosphomimetic mutant T285E, but not WT, GFP:β1-tubulin^R rescued the density of spindle MTs after Chk1-inhibition compared with Chk1-proficient WT controls (Fig. 3g–i and Supplementary Fig. 4b). In contrast, expression of nonphosphorylatable mutant T285A GFP:β1-tubulin^R diminished spindle MT density in the absence or presence of Chk1i (Fig. 3g–i and Supplementary Fig. 4b). These results suggest that Chk1-mediated T285 phosphorylation is required for optimal spindle MT density.

β-tubulin-T285 phosphorylation promotes MT nucleation

To investigate the importance of T285 phosphorylation for nucleation of spindle MTs, cells depleted of the endogenous protein and expressing WT, T285A or T285E GFP:β1-tubulin^R transgenes were analysed for MT re-growth during recovery from a prolonged treatment with ice-cold medium, essentialy as in Fig. 1l. We found that cells expressing the nonphosphorylatable T285A GFP:β1-tubulin exhibited reduced spindle MTs at several time-points post-recovery compared with those expressing the WT or the phosphomimetic T285E protein (Fig. 3j–m). To further investigate the effect of Chk1 phosphorylation on MT nucleation, porcine brain tubulin was phosphorylated by Chk1 in the presence of ATP in vitro, kinase reactions were briefly returned on ice to depolymerize potentially formed MT polymers¹⁰⁹, and phosphorylated tubulin (p-tubulin) was then allowed to polymerize under conditions that favor partial tubulin-polymerization to MTs, i.e., in the presence of 1 mM GTP and without glycerol (Supplementary Fig. 4c)^110,111. After polymerization, samples were centrifuged to separate the MT pellet (P) from the unpolymerized tubulin in the supernatant (S) and the incorporation of phospho-T285 tubulin in the MT pellet and supernatant was analyzed by western blotting (Supplementary Fig. 4c). Phosphorylated β-tubulin-T285 was detected almost exclusively in the MT pellet fraction and was diminished in the abscence of Chk1 or in the absence of exogenous ATP in the Chk1 kinase reaction, indicating that phosphorylated β-tubulin-T285 is preferentially incorporated in polymerized MTs (Fig. 3n). Residual T285 phosphorylation in the absence of exogenous ATP could be due to ATP co-purifying with recombinant Chk1. Also, in mitotic cells isolated by nocodazole and shake-off, phosphorylated β-tubulin-T285 was predominantly found in the polymerized tubulin pellet (P) fraction, by using a MT partitioning assay that examines tubulin partitioning between the polymer and soluble forms (Supplementary Fig. 4d, e). These results suggest that β-tubulin-T285 phosphorylation by Chk1 promotes MT nucleation.

Chk1 binds and bundles microtubules in vitro

Chk1 was mainly detected in the MT pellet fraction after MT polymerization in vitro (Fig. 4a). Also, Chk1 was found in the polymerized and soluble tubulin fractions by MT partitioning assay in mitotic cells (Supplementary Fig. 4f), showing that Chk1 associates with polymerized MTs. Both the N-terminal kinase domain and the C-terminal regulatory region of Chk1, but not GST-only, associated with β-tubulin by GST-pull downs in cell extracts; however, this association was relatively reduced compared with the full-length GST:Chk1, indicating that sequences in both the N-terminal and C-terminal regions contribute to Chk1-interaction with tubulin (Supplementary Fig. 4g, h). Chk1 also bundled MTs in vitro and, for 1 μM tubulin, approximately 0.25 μM Chk1 was required to observe robust bundling (Supplementary Fig. 4i, j). We conclude that Chk1 directly binds and bundles MTs in vitro.

Fig. 4. Expression of nonphosphorylatable T285A GFP:β1-tubulin reduces cold-stable MT polymers, increases chromatin misalignment and missegregation, and delays onset of anaphase.

a Chk1 kinase and MT co-sedimentation assay. Recombinant Chk1 was incubated in kinase reactions with 1 μM purified porcine brain tubulin as substrate, in the absence or presence of exogenous ATP. After tubulin polymerization and centrifugation, supernatants (S) and pellets (P) were analysed for precipitated Chk1 or α-tubulin by western blotting. b Cold-stable MT polymers. HEK cells expressing siRNA-resistant wild-type (WT), T285A or T285E GFP:β1:tubulin in combination with β-tubulin siRNA (si β1-tubulin) were treated with 10 μg/ml MG132 for 1 h and incubated in ice-cold medium for 10 min. c, d Relative GFP and α-tubulin intensity in prometaphase cells from (b). Mean ± SD from n cells. Values in WT were set to 1. c ***P = 1.5534E-07 (GFP:β1-tubulin^R WT + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285A + si β1-tubulin + MG132), 2.13631E-07 (GFP:β1-tubulin^R T285A + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285E + si β1-tubulin + MG132) by ANOVA and Student’s t test. d ***P = 5.94232E-16 (GFP:β1-tubulin^R WT + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285A + si β1-tubulin + MG132), 5.26763E-12 (GFP:β1-tubulin^R T285A + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285E + si β1-tubulin + MG132) by ANOVA and Student’s t test. e Chromosome misalignment. HEK cells were transfected as in (b) and treated with 10 μg/ml MG132 for 1 h. f Frequency of metaphases with misaligned chromatin in cells from (e). Error bars show the SD from the mean from four independent experiments (n = 92, 94, 82). ***, P = 0.000245955 (GFP:β1-tubulin^R WT + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285A + si β1-tubulin + MG132), 0.000411049 (GFP:β1-tubulin^R T285A + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285E + si β1-tubulin + MG132) by ANOVA and Student’s t test. g Chromatin missegregation. Arrows show misaligned or missegregated chromatin. h Frequency of anaphases with missegregated chromatin in cells from (g). Error bars show the SD from the mean from three independent experiments (n = 64, 62, 65). ***P = 0.000314545 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin), 0.000506684 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin) by ANOVA and Student’s t test. i, j HEK 293 T cells were transfected with siRNA-resistant WT, T285A or T285E GFP:β1:tubulin in combination with si β1:tubulin and monitored by phase-contrast live-cell microscopy. Time is from cell rounding. Asterisks mark the beginning of anaphase. In (j), cells were treated with 2 μM AZ3146 (Mps1i) immediately before filming. k Time from cell rounding to anaphase, calculated from time-lapse movies as in (i, j). Mean ± SD from n cells. Numbers next to each bar indicate n. ***P = 3.9613E-07 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin), 1.4536E-05 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin), 3.74997E-16 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285A + Mps1i + si β1-tubulin) by ANOVA and Student’s t test. l, m Localization of Zwilch. BE cells transfected as in (b) were treated with 10 μg/ml MG132 for 1 h. Inset scale bars, 1 μm. n Relative Zwilch kinetochore intensity in prometaphase cells from (l, m). Mean ± SD from n kinetochores. Numbers next to each bar indicate number of kinetochores (n). Values in T285A were set to 1. ***P = 4.0823E-213 (GFP:β1-tubulin^R WT + si β1-tubulin + MG132 vs GFP:β1-tubulin^R T285A + si β1-tubulin + MG132) by ANOVA and Student’s t test. Bars, 5 μm.

β-tubulin-T285 phosphorylation is required for stable kinetochore-MT attachment, optimal chromatin alignment and segregation

To investigate the importance of T285 phosphorylation for kinetochore-MT attachment, HEK cells depleted of the endogenous β-tubulin and expressing WT or mutant transgenes were treated with the proteasome inhibitor MG132, incubated in ice-cold-medium for 10 min to depolymerize non-kinetochore MTs and, after fixation, stable kinetochore MT fibers were visualized by confocal microscopy (Supplementary Fig. 4k)¹¹². Cells expressing nonphosphorylatable T285A GFP:β1-tubulin^R exhibited reduced cold-stable MT polymers compared with WT controls or cells expressing the phosphomimetic T285E protein, suggesting that T285 phosphorylation is required for stable kinetochore-microtubule attachments (Fig. 4b-d). Unstable kinetochore-MT interactions can lead to misaligned chromosomes and chromosome segregation errors¹⁷. Cells expressing T285A GFP:β1-tubulin^R and treated with MG132 exhibited increased chromosome misalignement compared with those expressing the WT or T285E proteins, by confocal microscopy analysis of fixed samples (Fig. 4e, f and Supplementary Fig. 4l). Furthermore, expression of T285A GFP:β1-tubulin^R associated with increased chromosome mis-segregation compared with WT or T285E proteins (Fig. 4g, h). These results show that β-tubulin-T285 phosphorylation is required for optimal kinetochore-MT attachment, chromatin alignment and segregation in human cells.

Phosphorylated β-tubulin-T285 is required for optimal mitotic progression

Unstable kinetochore-MT attachments can lead to delayed anaphase onset. Live-cell imaging with the use of phase contrast¹¹³ showed that, after depletion of the endogenous protein, cells expressing nonphosphorylatable T285A GFP:β1-tubulin^R exhibited delayed progression from cell rounding to anaphase compared with cells expressing the WT or phosphomimetic T285E transgenes (Fig. 4i, k and Supplementary Movies 1–3) thus suggesting that T285 phosphorylation is required for timely onset of anaphase. Mitotic progression, chromosome alignment and segregation were also examined after staining cells with the DNA dye Biotracker, by live-cell imaging. We found that cells expressing T285A GFP:β1-tubulin^R exhibited increased chromatin misalignment in metaphase, increased chromosome misegregation in anaphase and delayed progression from chromatin condensation to anaphase compared with WT controls (Supplementary Fig. 4m–q and Supplementary Movies 4–7), in agreement with our findings from fixed cells. We also found that delayed entry to anaphase in T285A cells correlated with increased localization of the spindle checkpoint protein Zwilch, a component of the Rod-Zwilch-ZW10 (RZZ) protein complex that is essential for the mitotic spindle checkpoint delay^114,115, to metaphase-like kinetochores compared with WT controls (Fig. 4l–n). Furthermore, treatment of cells with the Mps1 kinase inhibitor AZ3146 (Mps1i) to inhibit spindle checkpoint signaling^116,117, impaired the anaphase delay in T285A cells and these cells entered anaphase with similar kinetics with Mps1i-treated WT or T285E cells (Fig. 4j, k and Supplementary Movies 8–10). These results suggest that impaired β-tubulin-T285 phosphorylation can lead to delayed anaphase onset due to spindle checkpoint activation.

β-tubulin-T285 phosphorylation is required for equal-sized cell division

Defects in spindle MTs can lead to spindle mispositioning and generation of unequal-sized daughter cells^39,40. To investigate daughter cell-size, we measured the ratio (R) of the areas of the two daughter cells (R = L/S; L - large; S - small) in late cytokinesis. In control HEK cells, R had an average value of 1.08 ± 0.11 corresponding to equivalently-sized daughter cells; however, the R-value was increased to 1.33 ± 0.24 in cells treated with Chk1i (Fig. 5a, b). Similarly, Chk1-depleted BE cells exhibited increased R value (1.33 ± 0.20) compared with controls (1.09 ± 0.08; Supplementary Fig. 5a, b), showing that Chk1-deficient cells exhibit unequal-sized cell divisions. Also, Chk1-deficient cells in metaphase after treatment with MG132 exhibited increased ratio (r) of the distances of the two spindle poles from the cell cortex (r = l/s; l - long; s - short), indicating uncentered metaphase spindles, compared with controls exhibiting r ≈ 1 (Supplementary Fig. 5c–f). Uncentered spindles in Chk1-deficient cells can be caused by impaired displacement of cortical dynein from the proximal spindle pole^39,40. By using a human cell line stably expressing dynein heavy chain (DHC) fused to GFP (DHC:GFP), we found that Chk1-deficient cells with uncentered metaphase spindles exhibited asymmetric localization of dynein heavy chain (DHC) to the cell cortex similar with controls, suggesting that Chk1 is not required for dynein displacement (Supplementary Fig. 5g, h). Instead, HEK cells expressing the nonphosphorylatable T285A GFP:β1-tubulin^R exhibited increased R-value compared with cells expressing the WT or the phosphomimetic T285E protein; furthermore, restoring spindle MT density by expression of T285E, but not WT or T285A, GFP:β1-tubulin rescued equal daughter cell-size after Chk1-inhibition compared with WT controls (Fig. 5d–f). These results suggest that Chk1-mediated T285 phosphorylation promotes centered spindle positioning and equal-sized cell division by promoting optimal spindle MT density in human cells.

Fig. 5. Depletion of ATR, ATRIP or TopBP1 impairs β-tubulin-T285 phosphorylation and reduces the density of spindle MTs.

a, b Daughter cell-size. Cell images and dot plot of the ratios (R) of the areas of the two daughter cells (R = L/S; L, large; S, small) in cytokinesis. HEK cells were untreated (control) or treated with 300 nM UCN-01 (Chk1i) for 5 h. ***P = 4.80166E-09 (control vs Chk1i) by ANOVA and Student’s t test. c–f Cells were transfected with siRNA-resistant wild-type (WT), T285A or T285E GFP:β1-tubulin in combination with β-tubulin siRNA (si β1-tubulin) and treated with Chk1i for 5 h. Points represent the ratio R per daughter cell-pair. ***P = 0.00035909 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R WT + Chk1i + si β1-tubulin), 0.000746204 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin), 0.001286149 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin), 0.000917042 (GFP:β1-tubulin^R WT + Chk1i + si β1-tubulin vs GFP:β1-tubulin^R T285E + Chk1i + si β1-tubulin) by ANOVA and Student’s t test. g, h Localization and relative intensity of phosphorylated Chk1-S345 (pS345) at centrosomes in prometaphase. BE cells were untransfected (control) or transfected with ATR (siATR), Chk1 (siChk1), TopBP1 (siTopBP1) or ATRIP (siATRIP) siRNAs. ***P = 2.36497E-37 (control vs siATR), 3.76358E-34 (control vs siTopBP1), 2.26112E-43 (control vs siATRIP) by ANOVA and Student’s t test. i–k Localization and relative intensity of phosphorylated β-tubulin-T285 (pT285) at prometaphase centrosomes. Arrows indicate centrosomes. ***P = 2.18032E-36 (control vs siATR), 5.66377E-40 (control vs siTopBP1), 4.17971E-45 (control vs siATRIP) by ANOVA and Student’s t test. l, m Mitotic spindles and relative α-tubulin fluorescence intensity. Mean ± SD from n cells. Numbers next to each bar indicate n. Values in control were set to 1. ***P = 1.5607E-29 (control vs siATR), 1.01391E-24 (control vs siTopBP1), 4.76862E-30 (control vs siATRIP) by ANOVA and Student’s t test. Bars, 5 μm.

ATR, ATRIP and TopBP1 are required for Chk1 activation and β-tubulin-T285 phosphorylation at mitotic centrosomes

ATR, ATRIP and TopBP1 activate Chk1 in the DNA damage response; we therefore investigated a potential role for these proteins in centrosomal Chk1 activation and spindle formation. We found that cells depleted of ATR, Chk1, ATRIP or TopBP1 by siRNA exhibited diminished localization of phosphorylated (active) Chk1-serine 345 to prometaphase centrosomes compared with controls (Fig. 5g, h and Supplementary Fig. 5i–k). Depletion of ATR, ATRIP or TopBP1 also diminished β-tubulin-T285 phosphorylation at centrosomes and reduced the density of spindle MTs compared with controls (Fig. 5i–m). Expression of phosphomimetic T285E, but not WT or nonphosphorylatable T285A, GFP:β1:tubulin^R rescued spindle MT density after ATR-inhibition by VE821 (ATRi) compared with WT controls (Fig. 6a–d). Also, mitotic cells isolated by nocodazole and shake-off (Supplementary Fig. 1m) and treated with ATRi for 1 h exhibited reduced spindle MT density compared with controls, further supporting that ATR promotes spindle MTs in a mitosis specific manner (Supplementary Fig. 5l, m). These results show that ATR, ATRIP and TopBP1 are required for centrosomal Chk1 activation, β-tubulin-T285 phosphorylation, and optimal spindle MT density.

Fig. 6. Expression of phosphomimetic T285E GFP:β1-tubulin rescues spindle MT density in ATR-deficient cells.

a–d Mitotic spindles and relative GFP or α-tubulin fluorescence intensity. HEK cells expressing siRNA-resistant wild-type (WT), T285A or T285E GFP:β1:tubulin in combination with β1-tubulin siRNA (si β1-tubulin) were treated with 10 μM VE821 (ATRi) for 5 h. Mean ± SD from n cells. Values in WT were set to 1. c ***P = 1.06692E-13 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R WT + ATRi + si β1-tubulin), 1.80427E-12 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + ATRi + si β1-tubulin) by ANOVA and Student’s t test. d ***P = 2.78153E-20 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R WT + ATRi + si β1-tubulin), 9.47272E-14 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + ATRi + si β1-tubulin) by ANOVA and Student’s t test. e–j Localization and relative intensity of ATR, TopBP1 or ATRIP at centrosomes in prometaphase. BE cells were untransfected (control), or transfected with ATRIP siRNA (siATRIP) or TopBP1 siRNA (siTopBP1). Mean ± SD from n cells. Values in control were set to 1. Insets show high magnifications of kinetochores. Inset scale bars, 1 μm. f ***P = 1.05572E-20 (control vs siATRIP) by ANOVA and Student’s t test. h ***P = 2.99768E-21 (control vs siATRIP) by ANOVA and Student’s t test. k–n Localization and relative intensity of ATR or TopBP1 at prometaphase centrosomes in HEK cells expressing siRNA-resistant WT GFP:ATRIP, GFP:ATRIP-del or GFP:ATRIP-top, in the presence of ATRIP siRNA (siATRIP). Mean ± SD from n cells. Numbers next to each bar indicate n. Values in WT were set to 1. m ***P = 1.34988E-28 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-del + siATRIP) by ANOVA and Student’s t test. n ***P = 2.96671E-27 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-top + siATRIP) by ANOVA and Student’s t test. Arrows indicate centrosomes. Bars, 5 μm.

ATRIP promotes localization of ATR and TopBP1 to mitotic centrosomes

Depletion of ATRIP by siRNA impaired localization of ATR and TopBP1 to centrosomes in prometaphase, and this was not due to reduced total amounts of ATR or TopBP1 proteins by western blotting (Fig. 6e–h and Supplementary Fig. 5n, o). In contrast, depletion of ATR or TopBP1 did not interfere with the localization of ATRIP to mitotic centrosomes compared with controls (Fig. 6i, j and Supplementary Fig. 5p). These results show that ATRIP promotes localization of ATR and TopBP1 to centrosomes in prometaphase.

Interaction of ATRIP with ATR and TopBP1 is required for Chk1 activation, β-tubulin-T285 phosphorylation and optimal spindle MT density

To further investigate the molecular interactions between ATR, ATRIP and TopBP1 at mitotic centrosomes, truncated ATRIP (Δ658-690; ATRIP-del) lacking the required amino acids for binding to ATR⁷⁶, or mutant ATRIP LLSS 332 to AAAA (ATRIP-top) that interacts with ATR but cannot bind to TopBP1⁷⁹ fused to GFP were generated and these proteins were resistant to degradation by siATRIP, which targets a sequence inside the 3’ untranslated region of human ATRIP mRNA (Supplementary Fig. 5q, r). After depletion of the endogenous protein, expression of ATRIP-del, but not ATRIP-top, impaired localization of ATR to centrosomes in prometaphase compared with controls expressing WT GFP:ATRIP (Fig. 6k, m). Furthermore, expression of ATRIP-top, but not ATRIP-del, diminished localization of TopBP1 to centrosomes compared with controls (Fig. 6l, n). Remarkably, expression of ATRIP-del or ATRIP-top diminished Chk1-serine 345 and β-tubulin-T285 phosphorylations at centrosomes, and reduced the density of spindle MTs compared with WT controls (Fig. 7a–f). These results suggest that interaction of ATRIP with ATR and TopBP1 is required for Chk1 activation and β-tubulin-T285 phosphorylation at mitotic centrosomes, and for optimal spindle maturation.

Fig. 7. Expression of GFP:ATRIP-del or GFP:ATRIP-top impairs Chk1-S345 phosphorylation at centrosomes and reduces spindle MT density.

a–d Localization and relative intensity of phosphorylated Chk1-S345 (pS345) or phospho-β-tubulin-T285 (pT285) at centrosomes in prometaphase. HEK cells were expressing siRNA-resistant wild-type (WT) GFP:ATRIP, GFP:ATRIP-del or GFP:ATRIP-top in the presence of ATRIP siRNA (siATRIP). b ***P = 1.54248E-08 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-del + siATRIP), 1.14594E-09 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-top + siATRIP) by ANOVA and Student’s t test. d ***P = 2.25859E-34 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-del + siATRIP), 1.7048E-33 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-top + siATRIP) by ANOVA and Student’s t test. e, f Mitotic spindles and relative α-tubulin fluorescence intensity in cells treated as in (a). Mean ± SD from n cells. Numbers next to each bar indicate n. Values in WT were set to 1. Arrows indicate centrosomes. Bars, 5 μm. ***P = 1.7304E-20 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-del + siATRIP), 1.87099E-25 (GFP:ATRIP WT + siATRIP vs GFP:ATRIP-top + siATRIP) by ANOVA and Student’s t test. g Protocol for cell proliferation assays. h Bar graphs showing proliferation of HEK cells expressing siRNA-resistant WT, T285A or T285E GFP:β1-tubulin in combination with β1-tubulin siRNA (si β1-tubulin). Error bars show the SD from the mean from three independent experiments (n = 3). **P = 0.00675329 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 2 days), 0.009172511 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 2 days), ***P = 0.00075432 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 3 days), 0.000479961 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 3 days), **P = 0.007566468 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin 4 days), 0.003343471 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin 4 days) by ANOVA and Student’s t test. i Phase contrast images of cells from (h). Bars, 200 μm.

Phosphorylated β-tubulin-T285 is required for cell proliferation

To investigate the importance of β-tubulin-T285 phosphorylation for tumor growth, we measured cancer cell proliferation on tissue culture dishes and colony formation on soft agar, which reflects anchorage independent cell growth in tumors¹¹⁸, at different times post-plating. After depletion of the endogenous protein, expression of nonphosphorylatable T285A GFP:β1-tubulin^R reduced cell proliferation on plastic dishes and colony formation on soft agar compared with cells expressing the WT or the phosphomimetic T285E transgenes (Fig. 7g–i and Fig. 8a–c), showing that T285 phosphorylation is required for efficient cancer cell proliferation.

Fig. 8. Expression of nonphosphorylatable T285A GFP:β1-tubulin reduces cell proliferation.

a Protocol for colony formation assays. b Colony formation assay. HEK cells were transfected with siRNA-resistant wild-type (WT), T285A or T285E GFP:β1-tubulin in combination with β1-tubulin siRNA (si β1-tubulin). The bar graph shows the number of colonies formed on day 13 after plating. Error bars show the SD from the mean from five independent experiments (n = 3940, 2874, 3888). ***P = 2.21936E-05 (GFP:β1-tubulin^R WT + si β1-tubulin vs GFP:β1-tubulin^R T285A + si β1-tubulin), 2.92961E-05 (GFP:β1-tubulin^R T285A + si β1-tubulin vs GFP:β1-tubulin^R T285E + si β1-tubulin) by ANOVA and Student’s t test. c Images of cell colonies from (b). Bars, 5 mm. d Model for the role of Chk1 in mitotic spindle formation and function. p, phosphorylation.

Discussion

On the basis of these findings, we propose the following model (Fig. 8d): In unperturbed mitosis, the DNA damage protein ATRIP promotes recruitment of ATR and TopBP1 to mitotic centrosomes. Interaction of ATRIP with ATR and TopBP1 is required for phosphorylation and activation of Chk1 kinase by ATR; in turn, active Chk1 phosphorylates β-tubulin-T285 in prometaphase and metaphase to promote efficient spindle MT nucleation and optimal spindle MT density, which is essential for timely anaphase onset, proper chromosome alignment and segregation, equal daughter cell-size and efficient cell proliferation.

These findings describe a mechanism that contributes to the generation of proper mitotic spindles in vertebrate cells and propose that, in the absence of damaged DNA, mitotic centrosomes act as signaling hubs where DNA damage response proteins are rewired to promote spindle MT nucleation, indicating a cross-talk of the DNA damage response and cell division machineries to ensure genome integrity⁸⁷. Because Chk1-inhibition impairs spindle microtubule density in several vertebrate cell lines tested, we propose that this mechanism is conserved in several species. Our findings are in agreement with previous studies showing localization of ATR, ATRIP and TopBP1 to mitotic centrosomes and propose a biological role for this localization^93–95. By assigning mitotic functions to ATR and Chk1 at centrosomes, the cell can perhaps efficiently coordinate mitotic entry, through Chk1-mediated regulation of Cdc25B phosphatase activity, with mitotic spindle formation^84,85. It is also perhaps noteworthy that active Chk1 is phosphorylated at the DNA damage site S345 at mitotic centrosomes, but not at prometaphase kinetochores as shown in previous studies, indicating that populations of Chk1 in separate mitotic compartments are wired differently^88,119.

The mechanism of ATR/ ATRIP recruitment to mitotic centrosomes in the absence of single stranded DNA requires further investigation. One possibility is that ATRIP directly binds to centrosomal proteins: for example, interaction of ATRIP with centrosomal proteins such as NIMA-Related Kinase 1 (Nek1) or Coiled-Coil Domain Containing 28B (CCDC28B) in different contexts by co-immunoprecipitation assay or yeast two-hybrid screen has been reported^120,121. Alternatively, an unidentified ATRIP-interacting protein could mediate ATRIP-binding to centrosomal components¹²². It may also be interesting to examine whether Claspin (a key mediator protein in the DNA damage response) or ETAA1 (an ATR activator at centromeric R-loops in mitosis) contribute to ATR and Chk1 activation at centrosomes in prometaphase^91,92.

Our results identify the conserved T285 as a phosphorylation site on β-tubulin that promotes optimal density of spindle MTs. Several post-translational modifications on α-/β-tubulin have been previously reported; however, in most cases, their biological significance remains unclear^66–69. The majority of these modifications accumulate on stable MTs over time and occur at the α-/ β-tubulin C-terminal tails, which are accessible on the MT polymer surface and can potentially interact with MAPs. Outside of the C-terminal tail region, β-tubulin can also be phosphorylated at Ser172 by Cdk1 and this phosphorylation prevents tubulin polymerization into growing MTs¹¹¹. Our results report a post-translation tubulin modification that promotes spindle MT density to the best of our knowledge, contributing to our understanding of how mitotic spindle formation is regulated. Furthermore, our results suggest that β-tubulin-T285 phosphorylation has crucial consequences for cell physiology, as it is required for correct chromosome alignment and segregation, timely mitotic progression, equal-sized cell division and efficient cell proliferation. Because errors in these processes are associated with human genetic disorders or birth defects^6–10, our results propose functions for ATR and Chk1 that could protect against human diseases and cancer, through their role in spindle maturation.

How does β-tubulin-T285 phosphorylation promote spindle MT density? T285 is inside the β-tubulin M-loop that is essential for the formation of lateral interactions between protofilaments^13,104–106. Therefore, one possibility is that phosphorylation of this site changes the electrostatic interactions between β-tubulin side-chains at the interdimer interphase to facilitate assembly of tubulin protofilaments into MT polymers^13,106. Perhaps consistently, we found that phosphorylated β-tubulin-T285 has a relatively high propensity to incorporate into polymerized MTs in vitro. However, tubulin phosphorylation by Chk1 did not increase MT nucleation in vitro (Fig. 3n) and this could be due to phosphorylation of a relatively small amount of β-tubulin by Chk1 under our experimental conditions. Alternatively, T285 phosphorylation may facilitate the interaction of β-tubulin with structural MAPs and/or microtubule-inner proteins (MAPs that bind to tubulin from the luminal side), which could stabilize the inter-protofilaments interactions to promote MT formation^65,123. T285 is conserved as a serine residue inside the α-tubulin M-loop, raising the possibility that the α-tubulin site is also phosphorylated in mitosis. Because the α-tubulin site does not match the minimum Chk1 consensus sequence R/K-x-x-S/T, perhaps a different kinase phosphorylates this residue inside α-tubulin¹⁰³.

Phosphorylated β-tubulin-T285 is visualized around mitotic centrosomes, where nucleation of spindle MTs occurs, but not inside the spindle by confocal microscopy. Although dephosphorylation of T285 inside the mitotic spindle cannot be formally excluded, inhibition of phosphatases by treatment with microcystin did not promote pT285 spindle MT localization. One possibility is that, because T285 is not located on the MT polymer surface, the phospho-specific antibody cannot access this residue after tubulin incorporation to MTs. Perhaps consistently, incorporation of phosphomimetic T285E β-tubulin into the mitotic spindle is similar to the WT protein. Also, phosphorylation of β-tubulin-T285 around centrosomes is higher in prometaphase and metaphase compared with cells later in mitosis or in interphase, and this coincides with a sharp increase in centrosomal γ-tubulin levels^124,125. We propose that β-tubulin-T285 phosphorylation is required for efficient MT nucleation from the centrosomes in the early stages of mitosis.

We also show that β-tubulin-T285 phosphorylation is required for cancer cell proliferation. Microtubule-targeting agents that bind and disrupt MTs, such as taxanes and nocodazole-analogs, are successfully used in cancer-treatment as an effective way to delay or stop cancer cells from exiting mitosis; however, they exhibit important limitations, such as cytotoxicity to non-tumorigenic cells and cancer cell resistance to chemotherapy^126,127. It will be interesting to examine whether impaired β-tubulin-T285 phosphorylation by pharmacological inhibition of Chk1 or ATR kinases synergizes with MT poisons to inhibit cancer cell proliferation, and whether this potential synergistic function could be exploited therapeutically. In summary, this paper describes a mechanism that is required for spindle maturation and function in vertebrate cells.

Methods

Primary antibodies

Mouse monoclonal antibodies against beta-tubulin (G8, sc-55529; used for western blotting in BE cells), Chk1 (G-4, sc-8408; used for western blotting), ATR (C-1, sc-515173), ATRIP (F-7, sc-365383), TopBP1 (B-7, sc-271043), Hec1 (C-11, sc-515550), GST (B-14, sc-138) and Myc (9E10, sc-40) were from Santa Cruz Biotechnology. Rabbit polyclonal antibodies against Chk1 (FL-476, sc-7898; used for immunoprecipitations), CENP-B (H-65, sc-22788), GFP (sc-8334; used in western blotting in ATRIP:GFP transfected cells; Supplementary Fig. 5q, r) and Cdc25B (H85, sc-5619) were also from Santa Cruz Biotechnology. Mouse monoclonal antibodies against actin (AC-40), alpha-tubulin (DM1A) and gamma-tubulin (GTU-88) were from Sigma-Aldrich. Rabbit monoclonal antibody against phospho-Chk1 Ser345 (133D3, mab#2348 T) and rabbit polyclonal antibody against ATRIP (2737S; used for western blotting in ATRIP:GFP transfected cells; Supplementary Fig. 5q, r) were from Cell Signaling Technology. Mouse monoclonal (ab28144) and rabbit polyclonal (ab4448) antibodies against pericentrin were from Abcam. Rabbit polyclonal antibody against β-tubulin (NB600-936SS; used for western blotting in tubulin:GFP transfected HEK cells; Supplementary Fig. 2j, l) was from Novus Biologicals. Rabbit polyclonal antibody against Turbo-GFP (PA5-22688; used in western blotting in tubulin:GFP transfected cells; Supplementary Fig. 2k, n) was from Thermo Fisher Scientific. Anti-pT285 polyclonal antiserum was generated in rabbits by immunization against the phosphorylated peptide phospho-T285 (280-290aa: QYRAL[pThr]VPELT-Cys) of human β1 tubulin (ProteoGenix). Rabbit polyclonal antibody against human Zwilch was from A. Musacchio (Max Planck Institute of Molecular Physiology, Dortmund, Germany).

Secondary antibodies

For immunofluorescence, goat anti-rabbit IgG Alexa Fluor 633-conjugated (A21071) antibody was from Invitrogen, Thermo Fisher Scientific. Goat anti-rabbit IgG FITC (fluorescein)-conjugated (111-096-047), goat anti-mouse IgG FITC (fluorescein)-conjugated (115-096-072), goat anti-rabbit IgG Rhodamine (TRITC)-conjugated (111-025-046) and sheep anti-mouse IgG Rhodamine (TRITC)-conjugated (515-025-072) antibodies were from Jackson ImmunoResearch. For Western blotting, horse anti-mouse IgG HRP-linked (7076) and goat anti-rabbit IgG HRP-linked (7074) antibodies were from Cell Signaling.

Recombinant proteins

Recombinant N-terminal GST tagged, human full length active Chk1 protein was from Millipore (Cat# 14-346).

Plasmids and cloning

Plasmid pEGFP-N1 coding for GFP under cytomegalovirus promoter was obtained from Takara Bio and plasmid GFP/hTUBB1 encoding human β1-tubulin (TUBB1) fused to TurboGFP into the pCMV6-AC-GFP vector was a gift from John H. Miller (Victoria University of Wellington, New Zealand)¹²⁸. Plasmid EGFP/hATRIP encoding human ATRIP gene fused to EGFP inside the pcDNA3.1zeo vector was a gift from Minoru Takata (Kyoto University, Japan)¹²⁹. Plasmids wild-type pCdc2-GFP and pCdc2-GFP-AF coding Cdk1 (Cdc2) in which Thr14 and Tyr15 were mutated to alanine and phenylalanine, respectively, fused to GFP in pEGFP-N1 vector were from R. Muschel (Oxford Institute for Radiation Oncology, Oxford, England, UK)¹³⁰. Plasmid pEGFP/Chk1 encoding human Chk1 fused to EGFP into pEGFP-N1 vector (Takara Bio Inc.) was from Addgene (22888). Sequencing of this plasmid showed that Chk1 exhibited mutation of aspartic acid-130 to alanine (D130A), and this mutation was reversed to obtain the WT pEGFP/Chk1 plasmid, as previously described¹³¹. Plasmids pGEX4T1/Chk1(1-476), pGEX4T1/Chk1(1-292) and pGEX4T1/Chk1(273-476) coding for protein sequences of avian Chk1 fused to GST were generated by amplifying the respective sequences by PCR from a pcDNA3.1zeo/avChk1 plasmid⁹⁶ and inserting them into the pGEX4T1 vector (GE Healthcare) as EcoRI-XhoI fragments.

Mutagenesis

Point mutations were generated by using the Q5 site-directed mutagenesis kit (New England Biolabs). To generate the pGFP/hTUBB^R plasmid encoding an siRNA-resistant form of human β1-tubulin fused to TurboGFP, the GFP/hTUBB plasmid was used to introduce T786A and C789A point mutations giving resistance to the β1-tubulin siRNA. To generate the pGFP/hTUBB^R(T285A) plasmid, the pGFP/hTUBB^R plasmid was used to introduce an A853G point mutation. To generate the pGFP/hTUBB^R(T285E) plasmid, the pGFP/hTUBB^R plasmid was used to introduce A853G and C854A point mutations. For pGFP/hTUBB^R(T274A) or pGFP/hTUBB^R(T290A) plasmids the pGFP/hTUBB^R plasmid was used to introduce, respectively, A820G, or A868G point mutations. For pGFP/hTUBB^R(S278A), the pGFP/hTUBB^R plasmid was used to introduce the A832G and G833C point mutations. To generate the ATRIP-top mutant, the EGFP/hATRIP plasmid was used and sequence 994-CTCCTGAGTAGT-1005 of human ATRIP was changed to 994-GCCGCGGCTGCT-1005. To generate ATRIP-del, the EGFP/hATRIP plasmid was used and the sequence 1972-2070 of human ATRIP was deleted. To generate the siRNA-resistant form of GFP:Chk1, the WT pEGFP/Chk1 plasmid was used to introduce A1044C, T1047C, and T1050A point mutations giving resistance to the Chk1-2 siRNA.

siRNA sequences

Human Chk1 (sc-29269; a pool of three individual siRNAs: 5΄-GCGUGCCGUAGACUGUCCA-3’, 5’-AACUGAAGAAGCAGUCGCAGU-3΄, 5΄-AACCAGAUGCUCAGAGAUUCU-3΄), ATR (sc-29763; a pool of three individual siRNAs: 5’-GAAGCAACAUUUAGUGAAA-3’, 5’-GUAGACUAAUGGAAUUCAA-3’, 5’-GCGUACAUGUAGAUUUCAA-3’), Cdc25B (sc-37552; a pool of three individual siRNAs: 5’- CCUUCAAGGAUGAGCUAAA-3’, 5’- CCCAGUCUGUUGAGUUAGU-3’, 5’- CACUGAGCAAGUUGAGAAA-3’), ATRIP (sc-44800, targeting a sequence inside the 3’ untranslated region of human ATRIP mRNA; 5’-CUCCUUUCCUUACCACAUC-3’) and TopBP1 (sc-41068; 5’-UGAUGGGCGGACGAGUAUA-3’) siRNAs were from Santa Cruz Biotechnology. Human Chk1-2 siRNA (5′-CCACAUGUCCUGAUCAUAU-3′) was from Thermo Fisher Scientific and β1-tubulin siRNA (Cat#4390824, siRNA ID: s285; 5’-CCACGUCUCCAUUUCUUUA-3’) was from Ambion. Only the sense sequences of the siRNA duplexes are shown.

Cell lines

Human colon carcinoma BE cells (diploid cells that contain an oncogenic Kras-G13D mutation as well as the BRAF-G463V oncogenic mutation), were a gift from Simon Wilkinson and Christopher Marshall (Institute of Cancer Research, London, UK)¹⁰⁷. Chinese hamster ovary cells expressing wild-type (CHO^WT) or mutant (T991G, C992A and T993G changing S331 to glutamic acid) 6xMyc-tagged human Aurora B cDNA (CHO^S331E) cloned into the pcDNA/FRT/TO vector and introduced into CHO cells stably expressing the tetracycline repressor (T-REx; Invitrogen), were previously described^90,98. Wild-type avian B-lymphoma DT40 cells (Chk1 +/+) or DT40 cells in which both chk1 alleles were disrupted by gene targeting (Chk1 -/-) were as described⁹⁶. A human cervical carcinoma HeLa cell line stably expressing dynein heavy chain fused to GFP (DHC:GFP) was a gift from Dr Iain Cheeseman (Whitehead Institute for Biomedical Research, Cambridge Massachusetts, USA)⁴⁰.

Cell culture and treatments

Human colon carcinoma BE cells, human embryonic kidney (HEK 293 T; HEK) cells, cervical carcinoma HeLa and chinese hamster ovary (CHO) cell lines were grown in DMEM (Biosera) containing 10% FBS (Biosera), 100 U/ml penicillin, and 100 μg/ml streptomycin at 37 °C in 5% CO₂. Avian B-lymphoma DT40 cells were cultured in DMEM containing 10% FBS, 1% chicken serum (Biosera), 10⁻⁵M β-mercaptoethanol, 100 U/ml penicillin, and 100 μg/ml streptomycin at 39.5 °C in 5% CO₂. Cells were treated with 300 nM UCN-01 (Chk1i; Sigma, Cat# 539644), 2.5 μΜ SB218078 (Chk1i-2; Tocris Bioscience, Cat# 2560), 10 μg/ml MG132 (Calbiochem), 2 μM AZ3146 (Mps1i; Axon MedChem, Cat # HY-14710), 10 μM VE821 (ATRi; Sigma, Cat# SML1415), 200 ng/ml nocodazole (Sigma-Aldrich), 20 μM roscovitine (Cdki, Sigma-Aldrich) or 10 μM etoposide (Sigma-Aldrich) as appropriate. Kinase inhibitors were added to the medium 5 h before fixation for analysis by indirect immunofluorescence microscopy, or as indicated. For time-lapse imaging, inhibitors were added to the medium immediately before filming. To induce expression of Aurora-B transgenes, CHO^WT or CHO^S331E cells were treated with 17 ng/ml or 50 ng/ml Tetracycline (Sigma), respectively, for 16 h prior to analysis or further treatment with drugs^90,98. Addition of Tetracycline stimulated accumulation of 6xMyc-Aurora-B^WT or 6xMyc-Aurora-B^S331E at approximate levels 10-fold higher than the endogenous protein and this level of expression was shown to disrupt endogenous Aurora-B functions while maintaining correct localization of 6xMyc-Aurora-B to centromeres^90,98. siRNA duplexes were transfected into BE or HEK cells 24 h before analysis using Lipofectamine 2000 (Invitrogen), unless otherwise stated. For expression of exogenous proteins, plasmids were transfected into cells in the absence or presence of appropriate siRNA duplexes 24 h before analysis or further drug treatment using Lipofectamine 2000 (Invitrogen). All cell lines used exhibited consistent morphology and growth properties and were negative for mycoplasma contamination.

Enrichment of cells in mitosis

BE cells were treated with 200 ng/ml nocodazole for 16 h, followed by shake-off and centrifugation. Fluorescence microscopy analysis showed that approximately 95% of isolated cells were in mitosis as evidenced by their condensed chromatin. Isolated mitotic cells were then washed twice with PBS and plated on microscope slides or 35 mm petri dishes, in the absence or presence of kinase inhibitors, for 1 h. Afterwards, cells were fixed and analysed by fluorescence microscopy, or were lysed and analysed by Western blotting, as appropriate.

Time-lapse imaging and Biotracker staining

HEK 293 T cells were seeded onto 35-mm diameter Petri dishes with a 175 μm-thickness glass base (Greiner, #627860) and an inverted fluorescence microscope (Observer D1; Zeiss) was used. Phase contrast images were taken by using a 20× Plan Neofluor 0.4 NA Ph2 dry objective (Zeiss). Imaging was performed at 37 °C in 5% CO₂ by using a Zeiss AxioCam MRm camera and Zeiss ZEN 2 acquisition software. Drugs were added to the medium immediately before filming, as appropriate. For DNA staining in HEK cells using the Biotracker 488 Green Nuclear Dye (Sigma, Cat# SCT-120), 0.5 μl/ml Biotracker and 100 μM of the efflux pump inhibitor Verapamil were added to the medium 1 h before filming.

Indirect immunofluorescence microscopy and microscope image acquisition

Cells were fixed in 4% paraformaldehyde in cytoskeleton buffer (1.1 M Na₂HPO₄, 0.4 M KH₂PO₄, 137 mM NaCl, 5 mM KCl, 2 mM MgCl₂, 2 mM EGTA, 5 mM Pipes, and 5 mM glucose, pH 6.1) for 5 min at 37 °C, permeabilized in 0.5% Triton X-100 in cytoskeleton buffer, washed twice with PBS at room temperature, and immunostained. For inhibition of protein phosphatases, cells were extracted in prewarmed (37 °C) Phem buffer (60 mM Pipes, 25 mM Hepes, pH 7.0, 10 mM EGTA, and 4 mM MgSO₄) supplemented with 0.5% CHAPS and 100 nM microcystin (Sigma-Aldrich) for 5 min at room temperature, fixed with prewarmed (37 °C) 4% paraformaldehyde in Phem buffer for 10 min at room temperature, permeabilized in 0.5% Triton X-100 in Phem buffer for 2 min at room temperature, washed twice with PBS, and immunostained¹⁰⁷. FITC- or rhodamine-TRITC-conjugated (Jackson ImmunoResearch) or Alexa Fluor 633-conjugated (Thermo Fisher Scientific) secondary antibodies were used as appropriate. DNA was stained with 10 μg/ml DAPI (Biotium) and cells were mounted in Vectashield medium (H-1000; Vector Laboratories).

Images were collected by using a super-resolution SP8 LIGHTNING laser-scanning spectral confocal microscope (Leica Microsystems), LASX software (Leica Microsystems), and a 63×Apochromat 1.40 NA oil objective. The low-fluorescence immersion oil (11513859; Leica Microsystems) was used and imaging was performed at room temperature. Mean projections of image stacks were obtained by using the LASX software.

Quantification of fluorescence signals

Fluorescence intensities of spindle MTs were quantified using the LCS Lite polygon tool by analyzing the image area immediately surrounding the mitotic spindle. To quantify β-tubulin-pT285, Chk1-pS345, ATR, ATRIP, TopBP1 or Zwilch fluorescence intensities at centrosomes or kinetochores, the LCS Lite polygon tool was used to analyze an equal image area around each centrosome/kinetochore and intensity values were normalized versus background values obtained by analyzing a nearby identical area within the cell, by subtracting the background-signal value from the centrosome/kinetochore value¹³¹. After background subtraction, the average values from control or mutant samples were calculated and control or mutant samples values were divided with the average control value to obtain the relative intensity of the values plotted (i.e., relative to control = 1). To analyse cortical DHC:GFP fluorescence in cells with uncentered spindles, the LCS Lite polyline tool was used to analyse an equal image area at the distal or proximal cell cortex, i.e., at the cell cortex close to, respectively, the distal or proximal spindle pole. The ratio of distal/cortical DHC:GFP fluorescence values was plotted for each cell.

Calculation of daughter cell size

Z-sections of daughter cells in late cytokinesis, exhibiting fully formed cell boundaries by actin staining and connected by a midbody, were acquired at 1 μm steps to capture the entire cells, by confocal microscopy. Projected images were generated using the LASX software, the cell area was measured using the LCS Lite stack profile dimension tool and the area ratio (R; R = L/S; large/small) of daughter cells was plotted⁴⁰.

Calculation of spindle rotation

To calculate the angle θ between the spindle axis and the parallel to the substrate (Supplementary Fig. 1j, l), metaphase cells in which chromosomes were imaged from the side were analysed. Z-sections of cells from one spindle pole (marked by pericentrin staining) to the second spindle pole were acquired at 0.5 μm steps and projected images were generated using the LASX software. The vertical distance z between the two spindle poles was calculated from the z-stacks (z = number of stacks between the spindle poles x 0.5 μm), whereas the horizontal distance y between the spindle poles was measured from the projected image using the LCS lite line tool. The tangent of the angle θ (tanθ) was calculated from the ratio z/y; θ was calculated from the inverse tanget (arctan) function and was plotted for each cell. In control spindles without rotation θ ≈ 0^Ο, i.e, both spindle poles are at the same Z-section (slice) and the vertical distance z ≈ 0.

Analysis of spindle positioning

Z-sections of metaphase cells in which chromosomes were imaged from the side were acquired and projected images were generated using the LASX software. The distance between each spindle pole (marked by pericentrin) and the cell cortex (marked by actin staining) along the spindle axis was measured from the projected image using the LCS lite scale bar tool, and the ratio long/short distance values was plotted for each cell. In centered spindles, spindle poles are equidistant from the cell cortex and r ≈ 1.

MT re-growth assays

To determine microtubule nucleation at centrosomes, BE cells were grown on 35 mm culture dishes and microtubules were completely depolymerized by incubation of cells in ice-cold media on ice for 1.5 h. For the 0 min time point, the ice-cold medium was removed and immediately replaced with 4% paraformaldehyde in cytoskeleton buffer, followed by 0.5% Triton X-100. For re-growth, the ice-cold medium was replaced with 37 °C-pre-warmed medium and the culture dishes were returned to the incubator at 37 °C for various times prior to fixation, as appropriate. Cells were then stained for α-tubulin and prometaphase cells were analysed by fluorescence microscopy.

Analysis of cold-stable MT polymers

To induce microtubule disassembly of non-kinetochore microtubules, cell media was replaced with ice-cold media and cells were incubated on ice for 10 min. Subsequently, cells were fixed in 4% paraformaldehyde in cytoskeleton buffer, permeabilized in 0.5% Triton X-100, stained for α-tubulin and prometaphase cells were examined by fluorescence microscopy.

In vitro Chk1 kinase assay

For Chk1 kinase assays, 0.1 μg recombinant Chk1 was incubated with 1 μg GST:tubulin protein substrates in 20 μl Chk1 kinase buffer (20 mM MOPS, pH 7.2, 5 mM EGTA, 10 mM MgCl₂, 25 mM sodium β-glycerophosphate, 1 mM sodium vanadate, 1 mM DTT, 100 μM ATP, and 1 μCi γ-³²P ATP) for 20 min at 30 °C, before analysis by SDS-PAGE. Radioactive labeling of Chk1 substrates was determined by autoradiography. The Chk1 kinase assay in Fig. 3n was as above except that γ-³²P ATP was omitted from the reaction mixture and phosphorylation of Chk1 substrates was determined by Western blotting using an antiserum against phosphorylated T285 of β-tubulin.

Mass spectrometry

After phosphorylation by recombinant Chk1, GST:β1-tubulin (254-444) on agarose beads was digested with Asp-N and the peptides were analyzed by liquid chromatography–mass spectrometry as previously described¹³². The data were searched against an in-house database containing the β-tubulin sequence using Mascot (2.4.1; Matrix Science) and MaxQuant (1.3.8.2). All result files were loaded into Scaffold (4.8.4; Proteome Software Inc.). Protein identification thresholds were set to 95.0%. Peptide identifications were accepted if they could be established at >95.0% probability as specified by the Peptide Prophet algorithm¹³³, resulting in a peptide false discovery rate of 0.76%.

Microtubule polymerization and sedimentration assay

Microtubules were polymerized from porcine brain tubulin (T240, Cytoskeleton) at 37 °C for 30 min in BRB80 buffer (80 mM PIPES, pH 6.9, 8 mM MgCl₂, 0.5 mM EGTA, 1 mM GTP) for 2 min. Samples were then centrifuged at 16,000 g for 1.5 h at 30 °C, and the supernatant and pellet were boiled and separated by SDS polyacrylamide gel electrophoresis.

Microtubule bundling assay

For microtubule bundling assays, microtubules were assembled from rhodamine-labeled porcine brain tubulin (TL590M-A, Cytoskeleton) at 37 °C for 30 min in BRB80 buffer (80 mM PIPES, pH 6.9, 8 mM MgCl₂, 0.5 mM EGTA, 1 mM GTP) and stabilized with 20 mM taxol (Applichem) for 5 min according to manufacturer’s instructions. 1 μM taxol-stabilized microtubules was incubated with 0.125–0.25 μM recombinant Chk1 (14-346, Millipore) in BRB80 buffer at room temperature for 15 min, in 10 μl final volume. Then, 5 μl were squashed in coverslips and observed by fluorescence microscopy.

Microtubule partitioning assay

BE cells were enriched in mitosis by treatment with 200 ng/ml nocodazole for 16 h and mitotic cells were isolated by shake-off and centrifugation. Isolated mitotic cells were then washed twice with PBS, plated on 35 mm petri dishes and returned to the incubator for 30 min. Afterwards, cells were extracted in PEMG buffer supplemented with inhibitors (100 mM PIPES, 1 mM EGTA, 1 mM MgSO₄, 4 M glycerol, pH 6.8, 0.1% Triton X-100, 5 μg/ml leupeptin, 50 μg/ml PMSF, 1 mM benzamidine, 5 μg/ml aprotinin, and 1 mM Na₃VO₄) and centrifuged at 12,000 g for 10 min at room temperature, essentially as described¹³⁴. Supernatant and pellet fractions containing, respectively, soluble or polymerized tubulin were analysed by SDS-PAGE and Western blotting, as appropriate.

Cell proliferation assay

To examine the effect of β-tubulin mutations on cell proliferation, HEK 293 T cells in 35 mm culture dishes were transfected with plasmids coding for siRNA-resistant WT, T285A or T285E GFP:β1-tubulin in the presence of β1-tubulin siRNA for 24 h. Afterwards, cells were trypsinised, counted, plated in 35 mm culture dishes at a density 2 × 105 cells per dish, and left undisturbed for 1–4 days. Dishes were photographed before counting at room temperature using an Axiovert 200 (Zeiss) inverted microscope and an Axiocam ERc5s camera (Zeiss), and the cell number was determined by manual counting at the indicated time-points.

Colony formation in soft agar

To analyze anchorage-independent colony formation, 35 mm culture dishes were coated with 1 ml culture medium containing 1% low melting point agarose (Invitrogen, Cat# 15517-022). Subsequently, 20,000 HEK 293 T cells (transfected with siRNA-resistant WT, T285A or T285E GFP:β1-tubulin in the presence of β1-tubulin siRNA for 24 h) were re-suspended in 1 ml of growth medium containing 0.5% low melting point agarose and were plated on top of the solidified bottom layer. The embedded cells were incubated at 37 °C and 5% CO₂ and 100 μl medium was carefully placed on top of each dish twice per week to avoid dehydration of the soft agar medium. The emerging colonies were stained with 0.01% crystal violet (Sigma, Cat # V5265) thirteen days after seeding and were counted manually using a brighfield stereoscope (Nikon).

Bacteria culture and purification of GST proteins

BL21 (DE3) cells (Agilent Technologies) were grown in 100–200 ml LB Broth (1% w/v tryptone, 0.5% w/v yeast extract, 171 mM NaCl) supplemented with the appropriate selection antibiotic. At OD600 = 0.5, cells were induced with 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) at 16 °C for 16–18 h. Afterwards, cells were spun down at 1000 g for 10 min and the pellet was resuspended in 3 ml NETN (20 mM Tris pH = 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% NP-40) for every 100 ml of culture. The suspension was sonicated for 3 × 10 s and cleared by centrifugation at 12,000 g for 10 min to produce the supernatant (crude extract) containing the GST fusion protein. To prepare GST fusion protein bound to glutathione agarose beads, 150 μl glutathione agarose slurry (sc-2009, Santa Cruz Biotechnology) was added for every 3 ml crude extract, mixed for 1 h at 4 °C using a spiramix, beads were spun down at 1000 g for 5 min, washed three times with 500 μl NETN and an equal volume of NETN (150 μl) was added. GST proteins on glutathione agarose beads were stored at 4 °C for up to 1 week.

Co-immunoprecipitation and GST pull-down

Cells were sonicated three times for 10 s in ice-cold immunoprecipitation/kinase buffer (50 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM EDTA, 2.5 mM EGTA, 10% glycerol, 0.1% Tween 20, 0.1 mM PMSF, 10 μg/ml leupeptin, 10 μg/ml aprotinin, 1 mM sodium fluoride, 10 mM sodium β-glycerophosphate, and 0.1 mM sodium vanadate) and incubated for another 30 min on ice. For co-immunoprecipitations, 1 mg cell lysate was incubated with 1 μg antibody for 16 h followed by the addition of 10 μl protein A/G PLUS-agarose beads (sc-2003, Santa Cruz Biotechnology) for 1 h at 4 °C. For GST pull-downs, 1 mg cell lysate was incubated with 1 μg GST-protein on glutathione-agarose beads for 4 h, at 4 °C. Samples were spun down and washed three times with immunoprecipitation/kinase buffer, and immunoprecipitated proteins on agarose beads were analyzed by SDS PAGE and Western blotting.

Western blotting

Cells were lysed in ice-cold, whole-cell extract buffer (20 mM Hepes, 5 mM EDTA, 10 mM EGTA, 0.4 M KCl, 0.4% Triton X-100, 10% glycerol, 5 mM NaF, 1 mM DTT, 5 μg/ml leupeptin, 50 μg/ml PMSF, 1 mM benzamidine, 5 μg/ml aprotinin, and 1 mM Na₃VO₄) for 30 min on ice. Lysates were cleared by centrifugation at 15,000 g for 10 min at 4 °C. Samples were then analyzed by SDS-PAGE, transferred onto nitrocellulose membrane (Amersham Protran Premium 0.45 NC, Cat #10600003; GE Healthcare), and incubated with the appropriate antibodies. Secondary antibodies were detected by chemiluminescence (Clarity Western ECL Substrate, Cat #1705061; Biorad) and documented by autoradiography (Fig. 2d, e; Supplementary Fig. 1a–d, i; and Supplementary Fig. 5i–k), or by using the Sapphire Biomolecular Imager (Azure Biosystems; in all remaining blots).

Statistics and reproducibility

For fluorescence intensities of spindle MTs, GFP proteins at mitotic spindles or proteins at mitotic centrosomes, (n) cells from at least two independent experiments were analyzed per treatment and SD was calculated. For Zwilch kinetochore intensity, at least 35 kinetochores per cell from cells derived from two independent experiments were analysed per treatment (n > 790) and SD was calculated. For daughter cell size, spindle rotation or positioning, a minimum of 15 cells per experiment from two independent experiments were analyzed per treatment (n > 30). To calculate chromatin misalignment/ missegregation by live-cell imaging (Supplementary Fig. 4o, p), at least 20 cells per experiment from three independent experiments were analysed and SD between experiments was calculated. For cell proliferation or colony formation assays, 3-5 independent experiments were analyzed per treatment and SD was calculated. Statistically significant differences among three or more groups were determined by one-way ANOVA followed by two-tailed Students’ t test between two groups. No statistical method was used to predetermine the sample size. Blots were done twice and representative gels are shown.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

N.B. and E.P. performed the majority of the experiments and analyzed the results. S.L. and S.Z. made the mass spectrometry analysis. S.B. and D.E. provided the live-cell microscopy data, and E.G. performed cell proliferation assays. G.Z. designed and supervised the study and wrote the paper and all authors made comments on the manuscript.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Patrick Meraldi, Manuel Breuer, and George Inglis. [A peer review file is available].

Data availability

All data generated or analysed during this study are included in this published article, in the supplementary information files and in the Supplementary Data 1 files. The raw files for mass spectrometry analysis have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD074543.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-09862-x.