UV-C irradiation delays mitotic progression by recruiting Mps1 to kinetochores

Abstract

The effect of UV irradiation on replicating cells during interphase has been studied extensively. However, how the mitotic cell responds to UV irradiation is less well defined. Herein, we found that UV-C irradiation (254 nm) increases recruitment of the spindle checkpoint proteins Mps1 and Mad2 to the kinetochore during metaphase, suggesting that the spindle assembly checkpoint (SAC) is reactivated. In accordance with this, cells exposed to UV-C showed delayed mitotic progression, characterized by a prolonged chromosomal alignment during metaphase. UV-C irradiation also induced the DNA damage response and caused a significant accumulation of γ-H2AX on mitotic chromosomes. Unexpectedly, the mitotic delay upon UV-C irradiation is not due to the DNA damage response but to the relocation of Mps1 to the kinetochore. Further, we found that UV-C irradiation activates Aurora B kinase. Importantly, the kinase activity of Aurora B is indispensable for full recruitment of Mps1 to the kinetochore during both prometaphase and metaphase. Taking these findings together, we propose that UV irradiation delays mitotic progression by evoking the Aurora B-Mps1 signaling cascade, which exerts its role through promoting the association of Mps1 with the kinetochore in metaphase.

Keywords: :

• UV-C irradiation
• Mps1 kinase
• Aurora B
• kinetochore association
• mitotic delay

03/21/2013

Introduction

There have been extensive studies of the influence on mammalian cells of UV-A (315–400 nm), UV-B (280–315 nm) and UV-C (< 280 nm) radiation.^1-4 In contrast, only the effect of UV-B on mitotic cells has been studied. UV-B irradiation disturbs the mitotic process by affecting the integrity of mitotic organelles. The dynamics of spindle fiber assembly are affected reversibly by it.^5,6 The centrosomes during early metaphase are also sensitive to UV-B, and 5 seconds of UV-B microirradiation lead to a rapid chromosome shift toward the normal pole and disorganized the spindle.⁷ Interestingly, microirradiation of chromosomes and the cell cytoplasm during middle metaphase may block the metaphase to anaphase transition, but such treatment has no effect on cells in late metaphase.⁷ UV-B irradiation also induces DNA damage in mitosis, resulting in a preliminary DNA damage response.^8-10 Like UV-B, UV-C can also induce the DNA damage response through activating ATR-p53 signaling cascades during interphase, resulting in cell cycle arrest and apoptosis.^11-13 Recent studies have also shown that sub-lethal UV-C irradiation can synergize with other antitumor reagents to kill tumor cells in vitro,^14,15 assigning to UV-C a potential new application in tumor therapy. However, how UV-C irradiation affects mitotic progression remains less well defined.

Faithful partition of sister chromatids during mitosis is essential for genome stability. Chromosome aberrations during segregation cause aneuploidy, which is frequently linked to tumorigenesis, miscarriage and age-related diseases.^16-18 The mitotic defects caused by environmental insults can be monitored by a safety mechanism termed the spindle assembly checkpoint (SAC). The SAC is an evolutionarily conserved mechanism that operates from prophase to metaphase to prevent the premature onset of anaphase.¹⁹ The core components of the SAC signaling pathway in mammalian cells include mitotic arrest deficiency proteins (Mad1, Mad2), budding uninhibited by benomyl proteins (Bub1, Bub3, BubR1), monopolar spindle 1 (Mps1) and the CPC complex (Aurora B, INCENP, survivin and borealin).²⁰ SAC signaling converges on Cdc20, a key activator of the ubiquitin E3 ligase anaphase-promoting complex/cyclosome (APC/C) that targets cyclin B and securin for proteolysis. Upon SAC activation, the mitotic complex (including BubR1, Bub3, Mad2 and Cdc20) binds to and inactivates APC/C.¹⁹ All SAC proteins identified so far are kinetochore-binding proteins, and the loading and unloading of these proteins from the kinetochore are regulated strictly in a cell cycle-dependent way. Mps1 kinase is a classic SAC protein, playing essential roles in the SAC and chromosomal alignment.^21,22 Mps1 binds to kinetochores from prophase to prometaphase and is then disassociated during metaphase.^23,24 Blocking Mps1 release from the kinetochore during metaphase via genetic fusion with a constitutive kinetochore binding protein, Mis12, delays cell cycle progression.²⁵ A similar phenomenon has also been observed in the case of Mad2, the downstream regulator of Mps1 in the SAC signaling pathway.²⁶

Recently, we revealed that Mps1 undergoes nuclear translocation at the G₂/M boundary.²⁷ In an effort to identify the insults affecting Mps1 distribution in unsynchronized cells, we unexpectedly found an increased amount of Mps1 on the kinetochores in metaphase cells upon UV-C irradiation. UV-C irradiation also caused a delay in mitotic progression instead of inducing apoptosis immediately. Importantly, both the kinase activity and kinetochore localization of Mps1 are indispensable for the UV-C-induced mitotic delay. Finally, we found that Aurora B can be activated by UV-C irradiation, and its activity is required for the recruitment of Mps1 to the kinetochore. Our data show for the first time that UV-C radiation can cause mitotic delay by triggering the Aurora B-Mps1 signaling cascade on kinetochores.

Results

UV-C irradiation increases Mps1 binding to kinetochores in the metaphase plate

The subcellular location of Mps1 is tightly regulated throughout the cell cycle. Mps1 resides in the cytoplasm, nuclear envelope and centrosomes during interphase and becomes concentrated in the nucleus at the G₂/M boundary.^21,27,28 After entry into mitosis, Mps1 associates with the kinetochore from prophase to prometaphase and is released upon the achievement of chromosomal pairing (Fig. 1A and B).^22,29

Figure 1. UV irradiation promotes Mps1 binding to kinetochores during metaphase. (A) Subcellular location of Mps1 during the cell cycle. Asynchronous SW480 cells were fixed with 1% paraformaldehyde and stained with Anti-Mps1 antibody (NT, Millipore) and a CREST antiserum (Antibodies, Inc.) to identify Mps1 (green) and kinetochores (red) respectively. DNA was stained with DAPI. (B) Visualizing the binding and release of Mps1 to/ from the kinetochore via live cell imaging. The stable cell line SW480-YFPMps1 was synchronized at G₁/S by the double thymidine procedure and released. The time-lapse microscope video was recorded 7 h after release. Times are given in minutes. The dynamics of Mps1 on the kinetochores were quantified by calculating the intensity of Mps1-on-kinetochores against off-kinetochores. (C) IF and quantification of Mps1 on the kinetochores in metaphase cells upon UV irradiation. Asynchronous SW480 cells were exposed to UV irradiation at a dose of 40J/cm² and then fixed with 1% paraformaldehyde. The subsequent IF staining was conducted exactly as in (A). To measure the Mps1 signal intensity in the metaphase plate, a line scan was conducted by running the plot profile procedure in ImageJ software. (D and E) IF and quantification of Mad2 or Plk1 on kinetochores upon UV irradiation. The kinetochore location of Mad2 or Plk1 in metaphase cells was also analyzed in the same samples as in (C). UV, UV irradiation; DAPI, 4, 6-diamidino-2-phenylindole; IF, immunofluorescence; R.U., random unite

UV-C irradiation has been shown to mediate nuclear retardation of the RelA component of NFκB.³⁰ Similarly, we investigated whether UV-C irradiation affects the location of Mps1 in unsynchronized SW480 cells. In interphase cells, UV-C irradiation did not cause redistribution of Mps1, most of which remained in the cytoplasm (unpublished). In contrast, we found that UV-C irradiation of metaphase cells triggered a significant retargeting of Mps1 to the kinetochores (Fig. 1C). The recruitment of Mps1 to the kinetochore is required for the targeting of Mad2 to kinetochores during prometaphase.^31-33 Mad2 is a downstream regulator of Mps1 in the SAC signaling pathway and shows a similar kinetochore binding pattern to Mps1.³¹ Similarly, we found that UV-C treatment directs Mad2 to the kinetochores in metaphase cells (Fig. 1D). However, the binding of Plk1 to the kinetochore did not increase after UV-C irradiation (Fig. 1E)

UV-C irradiation causes delay of mitotic progression

SAC proteins localize to the unattached kinetochore and are released from the kinetochore once the SAC is passed.¹⁹ The SAC proteins Mps1 and Mad2 relocate to the kinetochore during metaphase upon UV-C irradiation, suggesting activation of a mitotic checkpoint. In view of this, we decided to investigate whether UV-C irradiation affects mitotic progression via monitoring the cyclin B level. SW480 cells were synchronized and subjected to western blotting at the indicated time-points after UV-C irradiation (Fig. 2A). In line with the speculation that UV-C irradiation activates the SAC, we found that the cyclin B degradation was retarded significantly by such radiation (Fig. 2B). This retardation appears to be SAC-dependent, as it was abrogated after co-incubation with Reversine (Fig. 2C), which depletes the SAC through inhibiting Mps1 kinase activity.³⁴

Figure 2. UV irradiation delayed the progression of cell mitosis. (A) A schematic diagram of the experimental procedure utilized in (B and C) to track the mitotic progression of SW480 cells by monitoring the cyclin B level. SW480 cells were synchronized in prometaphase by combining a STA (24 h) plus Noc treatment (12 h). Mitotic cells were then collected by shaking off and subjected to treatments as indicated before release. (B) Effect of UV irradiation on cyclin B degradation. SW480 cells were synchronized as described in (A) and were divided into two samples. For UV irradiation, cells were re-suspended in 1 ml D-PBS, transferred to a 10-cm dish and then exposed to UV-C at a dose of 40 J/cm². They were then collected by centrifugation and re-suspended in fresh DMEM before release into 6-well plates. Cells with/without UV irradiation were collected at the designated time points and the cyclin B level was measured by western blotting. (C) Mps1 kinase activity is required for UV-induced mitotic delay. The cells were all treated with UV-C as in (B) and released into DMEM with/without the Mps1 inhibitors 0.5 μM Reversine or 1 μM MPI-0479605. The cells with UV irradiation were collected at the designated time points and the cyclin B level was measured by western blotting. (D) Live cell imaging of HeLa expressing H2BGFP. Cells were synchronized in prometaphase as in (B). After three washes with D-PBS, they were treated as shown in the representative figures and then a time-lapse microscope video was taken for 2 h at 5 min intervals. (E) Statistical result of metaphase arrest induced by UV-C irradiation (10 and 14 metaphase cells upon UV-C irradiation and untreated were analyzed respectively). Noc, Nocodazole; Cyc B, Cyclin B; STA, single thymidine arrest.

To determine further how UV-C affects mitotic progression, we also traced the behavior of mitotic cells via live cell imaging. HeLa-H2BGFP were synchronized in prometaphase by Nocodazole treatment and then exposed to UV irradiation directly without shaking off. The cells were then released into fresh DMEM and subjected to live cell imaging. As shown in Figure 2D, UV-C induced two types of mitotic delay. About 61% of the mitotic cells in prometaphase remained arrested in prometaphase for about 30 min and then displayed chromosomal over-condensation within 60 min, which is similar to the fate of interphase cells. In the other 39% of mitotic cells, with partially formed metaphase plates, chromosomes became aligned more than 60 min before chromosomal over-condensation occurred (Fig. 2D; Supplemental Movies). The appearance of two types of mitotic cells during UV-C irradiation could be due to the difference in cell progression after release. Consistent with the western blotting result, the delay in mitotic progression induced by UV-C irradiation can be abolished by Reversine (Fig. 2D). As Mps1 inhibition can accelerate the mitotic progression,³⁴ it is likely that UV-C irradiation may still block cell mitotic progression in presence of Mps1 inhibition. To rule out this possibility, we measured the effect of Reversine on cell mitotic progression with or without UV-C irradiation. As showed in Figure S1, no statistical difference in mitotic progression was observed between control and UV-C-irradiated cells in presence of Reversine (p = 0.095). This finding confirms that mitotic delay by UV-C irradiation can be abrogated by Mps1 inhibition.

Mitotic delay by UV-C irradiation is not related to the DNA damage response

In flies, the SAC pathway can be activated by the DNA damage response, thereby causing metaphase/anaphase transition delay.³⁵ To investigate whether UV-C irradiation triggers the SAC via the DNA damage signaling pathway, we analyzed mitotic progression using doxorubicin, a DNA topoisomerase-2 inhibitor that can cause DNA double-strand breaks. As shown in Figure 3A, doxorubicin showed less effect on mitotic progression, although it increased the level of p53 protein significantly (Fig. 3A and B). Further, in comparison to the moderate level of γ-H2AX in response to UV-C irradiation, doxorubicin over a range of doses from 0.25 to 1.0 μM caused significantly more γ-H2AX to accumulate at DNA damage foci. Notably, there was only marginal Mps1 migration to the kinetochore in these metaphase cells with 1.0 μM doxorubicin (Fig. 3C), suggesting that activation of the SAC by UV-C irradiation is likely not due to the DNA damage response.

Figure 3. Mitotic delay by UV irradiation is not due to the DNA damage response. (A) Detecting the progression of mitosis in SW480 cells upon doxorubicin treatment via western blotting. Mitotic cells arrested by Nocodazole were shaken off and released into the medium with/without 1.0 μM doxorubicin. Mitotic progression was measured by evaluating the cyclin B level via western blotting. The experiment was repeated three times independently, and this was followed by the corresponding statistical analysis. (B) Doxorubicin promotes p53 expression. (C) IF and quantification of the cells subjected to DMSO or doxorubicin at the doses shown. SW480 cells were synchronized in metaphase via MG132 (10 μM for 1 h) and then treated with doxorubicin at the indicated dose for another hour. They were fixed and subjected to IF with antibodies against Mps1 and γ-H2AX. The line scan for the representative figure was conducted as in Figure 1. (D) Effect of VP-16 on Mps1 association with kinetochores. SW480 cells were arrested in prometaphase and then released into DMEM with MG132 for 1 h and then subjected to VP-16 at a dose of 50 or 250 μg/ml for another 1 h with/without caffeine. Mps1, CREST and DNA were stained and quantified as in Figure 1. (E) Effect of caffeine on UV-C induced kinetochore recruitment of Mps1. SW480 cells were treated with RO-3306 for 12 h after a single thymidine arrest, followed by wash-out and treatment with MG132 plus caffeine or DMSO for 2 h. They were then UV-C-irradiated as indicated and fixed after 30 min. Cells were then imaged after staining with antibodies or dye as indicated.

To test this speculation, we used VP-16, another DNA damage reagent. In contrast to doxorubicin, a high dose of VP-16 causes metaphase arrest, although at both high doses (250 μg/ml) and low doses (50 μg/ml), it causes massive DNA damage.³⁶ SW480 cells were synchronized during metaphase by Nocodazole plus MG132 in the presence of 50 or 250 μg/ml VP-16. As with doxorubicin, 50 μg/ml VP-16 did not promote the targeting of Mps1 to the kinetochores in metaphase cells. However, when subjected to 250 μg/ml VP-16, the metaphase cells showed a strong Mps1 signal on the kinetochores. Strikingly, the binding of Mps1 to the kinetochore after a high dose of VP-16 was not affected by caffeine (Fig. 3D), which abolishes the DNA damage checkpoint via ATM inhibition.³⁷ Further, caffeine treatment also failed to block the UV-C irradiation mediated kinetochore recruitment of Mps1, although it significantly blocked UV-C induced accumulation of γ-H2AX on mitotic chromosomes (Fig. 3E; Fig. S2). These studies indicated that the mitotic delay induced by UV-C irradiation is not due to the DNA damage response but could be related to the recruitment of Mps1 to the kinetochore, which is evoked by other regulators.

Kinetochore recruitment of Mps1 during metaphase is required for the induction of mitotic delay by UV-C irradiation

The N-terminal phosphorylation of Mps1 (especially T12 and S15 sites) was proven to be indispensable for Mps1 binding to kinetochores in prometaphase in our previous study.²⁴ Likewise, we investigated whether the same phosphorylation events are required for its kinetochore association in metaphase upon UV-C irradiation. The cells stably expressing Mps1 mutants as shown in Figure 4A were arrested at metaphase via MG132 treatment and then irradiated before subject to immunofluorescence staining (Fig. 4A). As shown in Figure 4B, wild type Mps1 was significantly relocated to the kinetochores of metaphase chromosomes upon UV-C irradiation, while the mutants Mps1 NTOP (N-terminal without phosphorylation site) and Mps1T12S15AA were not. The phosphorylation of S80 also partially contributes to Mps1 targeting to the metaphase plate. The Mps1 mutant T33S37AA showed no detectable deficiency in kinetochore binding. Similar to the prometaphase result, the defect in kinetochore recruitment of the mutant Mps1 T12S15AA was largely restored by the phospho-mimetic Mps1T12S15DD (Fig. 4B).²⁴ These data showed that the N-terminal phosphorylation event is indispensable for Mps1 association with the kinetochore during metaphase.

Figure 4. N-terminal phosphorylation of Mps1 is indispensable for UV-induced kinetochore recruitment of Mps1 and metaphase arrest. (A) Schematic map of the −Mps1 mutants used in the following immunofluorescence assay. (B) Stable SW480 cell lines expressing YFP-fused Mps1 or Mps1 mutants with non-phospho-mimetics were arrested in metaphase using the Nocodazole-MG132 procedure. They were exposed to UV irradiation and fixed. IF staining for CREST and DNA were conducted as usual. The corresponding line scan results are shown in the bottom panel. (C) A schematic diagram of the experimental procedure utilized in (D and E) to track the mitotic progression of SW480 cells by monitoring the cyclin B level. (D and E) The effect of UV-C irradiation on the stable SW480 cell line with siRNA resistant Mps1 or Mps1 mutant defective for kinetochore recruitment. The mitotic SW480-YFPMps1siR and SW480-YFPMps1siRT12S15AA cells were treated as described in (C). Mitotic progression was monitored by evaluating the cyclin B level by western blotting (* denotes a degraded product of YFPMps1). Rev, Reversine; Chl, channel.

Recent studies have shown that blocking the release of the checkpoint proteins Mps1 and Mad2 from the kinetochores during metaphase via fusion with the constitutive centromere protein Mis12 causes a delay in the metaphase/anaphase transition.^25,26 Since UV-C irradiation increases the recruitment of both Mps1 and Mad2 to the kinetochore during metaphase, we propose that the delay of mitotic progression by UV-C irradiation is probably due to the relocation of Mps1 to the metaphase kinetochore. As the Mps1 mutant T12S15AA is defective in kinetochore binding, we then set out to determine whether this mutant is required for UV-C-induced mitotic delay. As shown, the depletion of endogenous Mps1 via siRNA had no effect on the UV-C-induced delay of mitotic progression in SW480YFPMps1siR cells (Fig. 4D). However, Mps1 depletion caused failure to respond to UV-C irradiation in SW480YFPMps1T12S15AAsiR cells (Fig. 4E). Since the T12S15AA mutation affects the binding of Mps1 to the kinetochore but not its kinase activity,²⁴ we conclude that association of Mps1 with the kinetochore is essential for UV-C irradiation-induced mitotic delay.

UV-C irradiation promotes kinetochore recruitment of Mps1 via activation of Aurora B kinase

Aurora B/INCENP is required for the kinetochore association of Mps1 in the frog, placing it upstream of Mps1 in kinetochore assembly.³² Recently, Aurora B has also been shown to contribute to full kinetochore recruitment and activation of Mps1 in human tumor cells.^38-40 Consistent with these observations, we found that treating cells with the Aurora B inhibitor ZM447439 significantly reduced the association of Mps1 with the kinetochore in prometaphase cells even in the presence of Reversine (Fig. 5A). Remarkably, the retargeting of Mps1 to the kinetochore upon UV-C irradiation was also blocked (Fig. 5B), suggesting a regulatory role for UV-C in Aurora B activity. Phosphorylation of Aurora B at pT232 is required for its kinase activity.⁴¹ As expected, we found that UV-C irradiation enhances the amount of Aurora B phosphorylated at pT232. This effect of UV-C can be blocked by adding ZM447439 (Fig. 5C), implicating enhancement of Aurora B kinase activity. Collectively, these results suggest that UV-C irradiation enhances Aurora B kinase activity, which subsequently promotes Mps1 association with the kinetochore during metaphase. As a control, the binding of active Plk1 to the kinetochore did not increase after UV-C irradiation (Fig. S3), although it has been indicated that the T12S15 site of Mps1 could be a target for Plk1.³⁸

Figure 5. Kinase activity of Aurora B is indispensable for Mps1 binding to kinetochores in metaphase upon UV-C irradiation. (A) Effect of Aurora B inhibition on Mps1 location in prometaphase. SW480 cells were treated with Nocodazole for 12 h after STA, then MG132, Aurora B inhibitor ZM447439 (ZM) and/or Rev as indicated were added. The cells were then fixed and subjected to IF staining for Mps1, kinetochores and DNA. Representative prometaphase cells and the quantitative results are shown. (B) Effect of Aurora B inhibition on Mps1 location in metaphase upon UV-C irradiation. The SW480 cells were released from Nocodazole and co-incubated with MG132 plus reagents as indicated before UV-C irradiation. They were imaged after staining with anti-Mps1 antibody (NT, Millipore), CREST serum and DAPI. Representative metaphase cells and the quantitative results are shown. The Mps1 level on kinetochore upon ZM or ZM plus Reversine treatment is statistically lower than that with DMSO or Reversine treatment (p < 0.0001) (C) UV-C treatment increases Aurora B phosphorylation at T232. Mitotic SW480 cells were collected by shaking off after Nocodazole treatment and released into fresh DMEM with reagents as indicated. After 1.5 h, the cells were exposed to UV-C and collected for western blotting with antibody as indicated. ZM, ZM447439; Rev, Reversine.

Discussion

During interphase, mammalian cells undergo cell cycle arrest or apoptosis in response to UV-C irradiation.^11,42 However, how mitotic cells respond to UV-C irradiation has not to our knowledge been defined. In this paper, we identified several novel aspects of mitotic cell behavior in response to UV-C irradiation. First, mitotic cells with chromosomes aligned on the spindle equator do not undergo cell death immediately but show delayed mitotic progression upon UV-C irradiation. Second, in contrast to its effect on interphase cells, UV-C-induced mitotic delay is not due to DNA damage but to enhanced recruitment of the SAC protein Mps1 to the kinetochore. Finally, we found that UV-C irradiation promotes Aurora B kinase activity, resulting in the relocation of Mps1 to the kinetochore in both prometaphase and metaphase cells. On the basis of these findings, we propose that UV-C activates Aurora B kinase and thereby transforms Mps1 from a kinetochore-unbound form to a bound form by promoting its phosphorylation at T12S15 and other amino acid residues; then the retargeting of Mps1 to the kinetochore alerts the spindle assembly checkpoint and arrests the cells in metaphase.

Although the DNA damage response (DDR) in mitosis can elicit the SAC and block the metaphase/anaphase transition in yeast,^43,44 we failed to establish this connection in human tumor cells in this paper. In mammalian cells, the DDR shares some cell cycle-regulating components with the SAC, including the SAC proteins BubR1 and Plk1;^9,45 conversely, some DDR proteins such as Chk1 and Chk2 are employed in regulating normal mitotic progression.^46,47 Mps1 is another dual-functional molecule that is necessary for both the SAC and DDR. Mps1 binds to CHK2 and causes intermolecular phosphorylation, which subsequently facilitates the integrity of the G₂ checkpoint.^48,49 Interestingly, a fraction of Mps1 phosphorylated at threonine 288 accumulates in the DNA damage foci in response to X-irradiation during interphase.⁴⁹ Apart from the G₂ checkpoint, it has also been suggested that Mps1 is involved in the G₁ tetraploid checkpoint by association with p53.⁵⁰ However, despite these connections, whether the DNA damage signal can activate the SAC remains controversial.^36,51,52 The idea that the DDR activates the SAC in vertebrate cells originated from studies on flies.³⁵ However, the DNA damage reagent doxorubicin did not induce metaphase arrest in human cancer cells, including HeLa, U2OS, HCT116, CFPAC-1 and transformed hTERT-RPE1 cells.^36,52 In our experiments, we confirmed that the DDR caused by doxorubicin did not affect mitotic progression. We also found that VP-16 at a dose of 250 μg/ml but not 50 μg/ml induces the recruitment of Mps1 to the kinetochore, which is very similar to Mps1 behavior in response to UV-C irradiation. Notably, we did not see Mps1 co-localization with DNA break foci caused by UV-C irradiation, as in the case of doxorubicin during interphase,⁵¹ implying that Mps1 is not involved in the DNA damage response during mitosis as it is in interphase. Recently, two independent investigations have shown that the DDR in mitosis is a primary response that includes the activation of ATM and DNA-PK and the recruitment of γ-H2AX, MDC1 and Mre11-Rad50-Nbs1 to DNA damage sites but without RNF8, RNF168, 53BP1 and BRAC1. The initiation of damage response in mitosis does not affect mitotic progression but is required for DNA repair during the G₁ phase.^10,53 Taking these findings together, we suppose that the SAC is not evoked by the DDR upon UV-C irradiation but by other unknown signal transducers that trigger the Aurora B-Mps1 signaling pathway.

In this paper, we further confirmed that N-terminal phosphorylation is essential for UV-C-induced targeting of Mps1 to the kinetochore during metaphase. Our previous study identified a stretch of auto-phosphorylated serines and threonines at the Mps1 N-terminal, some of which (especially T12 and S15) are important for Mps1 binding to the kinetochore. We also found that the kinase-inactive form of Mps1 fails to bind to the kinetochore in SW480 cells following abrogation of endogenous Mps1 by siRNA, implying that the kinase activity of Mps1 is essential for its kinetochore binding.²⁴ Similarly, a decreased association of Mps1 with the kinetochore was observed in U2OS cells upon treatment with a specific Mps1 inhibitor, NMS-P715.⁵⁴ These observations are also confirmed in a recent study of Drosophila in which the kinetochore recruitment of kinase-inactive Mps1 required endogenous Mps1.⁵⁵ However, there are also conflicting reports. For example, treating HeLa cells with other Mps1 inhibitors, AZ3146 and Reversine, can promote Mps1 binding to the kinetochore,^34,56 implying that the kinase activity is not required for such binding. In this paper, we also found that Reversine with MG132 enhances the association of Mps1 with the kinetochore in prometaphase cells (Fig. S4A). Notably, this effect was not recapitulated in metaphase cells, as Mps1 does not bind to kinetochores in the presence of either Reversine or MPI-0479605 (Fig. S4B). Further, mutant Mps1T12S15AA did not show apparent kinetochore association with/without Reversine treatment in both prometaphase and metaphase cells (Fig. S4C and 4D), contradicting the implication that kinase activity inhibition by Reversine increases Mps1 targeting to kinetochores in prometaphase. To reconcile these conflicting results, a straightforward hypothesis is that loading and unloading of Mps1 on/ off the kinetochore during prometaphase and metaphase is mediated by different Mps1 substrates. In this case, NMS-P715 prevents Mps1 from phosphorylating its kinetochore binding domain and the substrates that direct Mps1 binding to the kinetochore. Conversely, AZ3146 and Reversine could inhibit Mps1 phosphorylation of the substrates that are responsible for unloading Mps1 from the kinetochore. During mitotic progression, the binding of Mps1 to these two types of substrate could be regulated in time and space by other kinases. In line with this speculation, two kinases, Cdk1 and Aurora B, have recently been reported to potentiate Mps1 activation in mitosis.^40,57 Importantly, inhibiting Aurora B kinase activity prevents Mps1 kinetochore recruitment not only in prometaphase both also during metaphase upon UV irradiation, suggesting that Aurora B is an indispensable mediator of UV irradiation that promotes Mps1 association with the kinetochore.

In conclusion, our results showed that UV-C irradiation can promote kinetochore recruitment of Mps1 during metaphase via activation of Aurora B, thereby causing mitotic delay.

Materials and Methods

Generation of stable cell lines

The stable SW480 cell line expressing the retroviral expression vector mutants YFP-Mps1, YFP-Mps1R and YFP-Mps1 mutants defective for N-terminal phosphorylation in the pRex background have been described previously.²⁴

UV-C treatment

Cells for immunofluorescence were grown on coverslips. Prior to staining or western blotting, they were irradiated with UV-C (254 nm) at a dose of about 40 J/cm² from a mercury germicidal lamp installed in a biosafety hood as described.⁵⁸

Immunofluorescence and quantification

Cells were washed three times with D-PBS and fixed for 10 min in D-PBS plus 1% paraformaldehyde. Anti-Mps1 and Anti-Mad2 were prepared at 1:100 dilution. Kinetochores were identified by staining with CREST antisera (Antibody Inc.). Alexa Fluor 488-conjugated goat anti-mouse secondary antibodies or Alexa Fluor 596-conjugated goat anti-human secondary antibodies (Invitrogen) were used depending on the source of primary antibody. After staining, the coverslips were mounted on pre-cleaned microscope slides with D-PBS plus 1.5 μg/ml DAPI and 50% glycerol and sealed with nail oil. Images were acquired on a Nikon TE2000 microscope equipped with a 63 × oil objective lens or a Zeiss LSM 510 equipped with a 100 × objective.

To quantify the proteins located on the kinetochore, two methods were employed: for prometaphase cells, the relative intensity of proteins of interest over CREST was calculated as described in previous reports;²⁴ for metaphase cells, we measured the relative intensity of the proteins on the metaphase plate by running a Plot profile procedure in ImageJ software owing to the overlapping of CREST signals, which make it impossible to measure the intensity of a single kinetochore.

Cell synchronization

Cells were synchronized in prometaphase as described by Zhang et al.²⁷ Basically, a single thymidine synchronization protocol was used to obtain S phase-arrested cells followed by release into growth medium containing 100ng/ml Nocodazole for 12 h. To block cells at metaphase, the cells synchronized in prometaphase were washed out and released into the fresh medium with 10 μM MG132 for 2 h.

siRNA transfection

Control and Mps1 siRNA were purchased from Invitrogen (Invitrogen). The sequences of the control and Mps1 siRNAs have been described previously.²⁴ Briefly, the cells were split into 12-well plates and transfected at 20% confluence using the lipid reagent Transfectin (Bio-Rad).

Immunoblotting

For western blotting, cells were washed with cold D-PBS twice and lyzed in buffer [20 mM Tris-Cl (pH 8.0), 0.2 M NaCl, 0.5% NP40, 1 mM EDTA, 1 mM PMSF, 20 mM NaF, 0.1 mM Na₃VO₄, 1 × protease inhibitors (Roche)] for 5 min prior to scraping. Cell extracts were clarified by spinning at 13,000 rpm for 10 min. The protein concentrations were determined by a BSA assay. Equal amounts of total proteins were resolved on 10% SDS-PAGE gels and transferred to nitrocellulose membranes. Primary and secondary antibodies were applied sequentially. The blots were developed in Super Signal WestDura (Pierce) according to the manufacturer’s instructions. Band density was measured using ImageJ and then plotted via Excel software.

Time-lapse fluorescence imaging

SW480 cells stably expressing YFP-Mps1 and HeLa S3 H2B-GFP were seeded in a DT 0.15 mm dish (Bioptechs) or an eight-chambered cover glass (Lab-Tek Chambered no 1.0 Borosilicate Cover Glass System, Nunc) in DMEM at 50% confluence. For SW480 cells, late G₂ phase synchronization was achieved by a single thymidine arrest plus RO-3306 treatment procedure. For HeLa S3 H2B-GFP, cells were synchronized in prometaphase using a single thymidine arrest plus Nocodazole for 12 h. Prior to imaging, the cells were released into fresh DMEM, and the dish was transferred to an environmental chamber mounted on the microscope. Images were taken at 5 min intervals and representative images of slices were prepared in ImageJ software.

Disclosure of Potential Conflicts of Interest

Acknowledgments

We are very grateful to Drs Daniel Wettstein, Ian McAlexander and Brandi Williams at Myrexis for a generous supply of MPI-0479605. We thank members of Zhong laboratory and Cao laboratory for sharing reagents. We also thank Dr Bao-feng Jin for helpful discussion. This work was supported by grants from the National Natural Science Foundation of China (30971444, 31101013, 31270800).

11/26/12

Np potential conflicts of interest were disclosed