Mitotic progression is regulated largely by reversible phosphorylation events that are mediated by mitotic kinases and phosphatases. Protein phosphatase 1 (PP1) has been shown to play a crucial role in regulation of mitotic entry, progression, and exit. We previously observed, in Xenopus egg extracts, that phosphatase 1 nuclear targeting subunit (PPP1R10/PNUTS) acts as a mitotic regulator by negatively modulating PP1. This study investigates the role of PNUTS in mitotic progression in mammalian cells, and demonstrates that PNUTS expression is elevated in mitosis and depletion partially blocks mitotic entry. Cells that enter mitosis after PNUTS knockdown exhibit frequent chromosome mis-segregation. Aurora A/B kinase complexes and several kinetochore components are identified as PNUTS-associated proteins. PNUTS depletion suppresses the activation of Aurora A/B kinases, and disrupts the spatiotemporal regulation of the chromosomal passenger complex (CPC). PNUTS dynamically localizes to kinetochores, and is required for the activation of the spindle assembly checkpoint. Finally, PNUTS depletion sensitizes the tumor cell response to Aurora inhibition, suggesting that PNUTS is a potential drug target in combination anticancer therapy.
Delineation of how PNUTS governs the mitotic activation and function of Aurora kinases will improve the understanding of the complex phospho-regulation in mitotic progression, and suggest new options to enhance the therapeutic efficacy of Aurora inhibitors.
Mitotic kinases, particularly cyclin-dependent kinase 1 (CDK1), polo-like kinase 1 (Plk1), and Aurora A/B, are well established as central regulators of mitosis. Activation of these kinases triggers the phosphorylation of various substrates to modulate all aspects of mitotic cell reorganization and progression (1–4). Thus, regulated activation and deactivation of these kinases are defined as molecular events that dictate M-phase entry and exit. Among mitotic kinases, Aurora A and Aurora B kinases phosphorylate a number of substrates to regulate the dynamics of centrosomes, microtubules, and chromatin during mitotic progression. Inactivation of Aurora A and B causes spindle defects, chromosome missegregation, cytokinesis failure, and aneuploidy (4–7). Owing to the crucial role of mitotic kinases in cell proliferation, and the profound toxicity of their inhibition, mitotic kinases are appreciated as potential drug targets for cancer therapy. Over the recent decade, numerous small-molecule inhibitors of Aurora kinases have been identified, and studies using these inhibitors demonstrated promising antitumor activities. Many of these inhibitors have entered clinical trials for breast cancer, prostate cancer, leukemia, and other malignancies (6, 8).
Protein phosphatases antagonize the action of kinases, but the function and regulation of mitotic phosphatases are considerably less studied (1, 2, 4, 9–12). The two major groups of serine and threonine phosphatases, protein phosphatase 1 (PP1) and 2A (PP2A), were both known to play important roles in mitotic regulation. For example, it has been shown that PP1 dephosphorylates numerous mitotic factors and thereby regulates mitotic entry, metaphase–anaphase transition, chromatin condensation, and mitotic exit (13–15). A major gap in knowledge, however, is how PP1 is regulated to achieve these dynamic and specific actions during mitotic progression.
Among the mitotic proteins dephosphorylated by PP1 were Aurora A and B kinases and their substrates, and the general role of PP1 as an antagonizer of Aurora kinases has been shown in multiple experimental systems (16–19). In yeast, the codeletion of PP1/Glc7 with Aurora/Ipl1 rescued mitotic progression and cell proliferation from Aurora/Ipl1 deletion (16). Similarly, PP1 suppression in mammalian cells prevented mitotic defects induced by the inhibition of Aurora B (17–19). The opposing relationship between PP1 and Aurora is perhaps best illustrated in the regulation of spindle assembly checkpoint signaling at kinetochores. The spindle checkpoint is a surveillance mechanism that ensures the proper, bipolar kinetochore–microtubule attachments (20, 21). It has been shown that reversible phosphorylation of kinetochore components plays a pivotal role in governing spindle attachment and the activation/deactivation of the spindle checkpoint (22). On one hand, centromeric Aurora B is responsible for the activation of the spindle checkpoint via both the recruitment of spindle checkpoint kinases, such as Mps1 and Bub1, and the subsequent phosphorylation of BubR1, Mad1, Ndc80/Hec1, Knl1, and other kinetochore proteins (21, 22). On the other hand, several kinetochore components, especially Knl1, yield PP1-binding activities, and were reported to recruit PP1 to kinetochores (19). After the proper spindle–kinetochore attachment, PP1 dephosphorylates Aurora B, Knl1, BubR1, and other kinetochore proteins to silence the spindle checkpoint and trigger metaphase–anaphase transition (22).
Phosphatase 1 nuclear subunit (PNUTS), also known as PPP1R10, was originally described as a nuclear regulator of PP1 that retains a portion of PP1 in the nucleus (23). PNUTS has been implicated in transcription and RNA processing (24, 25), DNA damage response and maintenance of telomere stability (26–30), and modulation of RB and PTEN (31–33). In most cases, with the exception of RNA Pol II, PNUTS acts to inhibit PP1 toward specific substrates. The earliest evidence that linked PNUTS to mitosis came from the observation that PNUTS enhanced the in vitro chromosome decondensation in a PP1-dependent manner (34). Interestingly, our recent study in Xenopus egg extracts suggested a role of PNUTS as an essential regulator of mitotic progression. Overexpression of PNUTS in Xenopus egg extracts inhibited mitotic exit. PNUTS depletion disrupted mitotic maintenance, whereas codepletion of PP1 rescued the defect (35). The level of PNUTS oscillated in cycling extracts and peaked in mitosis. In this study, we sought to examine the role of PNUTS in human cells, and to reveal further mechanistic insights into how PNUTS governs mitotic signaling. Our results confirmed a role of PNUTS in mammalian mitotic progression, characterized PNUTS-mediated regulation of Aurora kinases and spindle checkpoint signaling, and suggested PNUTS as a potential target to enhance the cytotoxicity of Aurora inhibitors.
Materials and Methods
Antibodies and other reagents
Cdc27 antibody was purchased from BD Transduction Laboratories; Aurora A, Aurora B, Bub1, Cenp-E, PP1, and PNUTS antibodies were purchased from Bethyl Laboratories; histone H3, phospho-H3 Ser-10, phospho-Aurora A/B/C, Aurora B, and β-actin antibodies were purchased from Cell Signaling Technology; PP1 antibody was purchased from Santa Cruz Biotechnology. PNUTS siRNA was obtained from Integrated DNA Technologies, with the targeting sequence of GGCGGCUACAAACUUCUU. PNUTS shRNA was generated using a pSUPER vector (Oligoengine), with the targeting sequence of CAGTGGTGGTTTCTGACAA.
Cell culture and treatment
Human cervix carcinoma (HeLa) cells, authenticated by ATCC, were maintained in DMEM (Hyclone) with 10% FBS (Hyclone). Breast cancer ZR75-1 and BT20 cells were authenticated at the University of Colorado Tissue Culture Core, and cultured in DMEM/F12 with 10% FBS. Cell-cycle synchronization at G1–S was performed using two rounds of thymidine treatment (at a final concentration of 2 mmol/L). For mitotic arrest, HeLa cells were treated with nocodazole (100 ng/mL) for 12 hours. Transfection was performed using Lipofectamine (Thermo Fisher Scientific), following a protocol recommended by the manufacturer.
Immunoblotting and immunoprecipitation
SDS-PAGE and immunoblotting were performed as described previously (36). For immunoprecipitation, anti-rabbit magnetic beads (New England Biolabs) were conjugated to the primary antibody, and then incubated in cell lysates for 2 hours. The beads were collected using a magnet, washed, eluted with Laemmli sample buffer, and then analyzed by immunoblotting.
Immunofluorescence and imaging
Immunofluorescence was performed as described previously (37). Briefly, cells were fixed in a fixation buffer (3% formaldehyde with 0.1% Triton X-100), washed, and blocked in a blocking buffer (10% goat serum in PBS). The primary antibodies were diluted in the blocking buffer, and incubated with the cells for 2 hours. The cells were then washed, and incubated with the Alexa Fluor 594 and 488 secondary antibodies (Invitrogen) for 1 hour. Imaging was performed using a Zeiss Axiovert 200M inverted fluorescence microscope at the UNMC Advanced Microscopy Core Facility. Live-cell imaging was performed using the Marianas Live Cell system based around a Zeiss Axiovert 200M microscope stand, and the SlideBook6 software (Intelligent Imaging Innovations, Inc.). Images were collected with 10× objective lens magnification. Once the live-cell microscopy was completed, the captured images were loaded into SlideBook Reader Software (Intelligent Imaging Innovations).
Protein expression, pull-down, and mass spectrometry analysis
MBP-tagged PNUTS was constructed and expressed as in our previous study (35). The recombinant protein was expressed in BL21 bacterial cells and purified on amylose beads. For the pull-down assay, amylose beads conjugated with MBP–PNUTS were incubated in Xenopus egg extracts that were prepared as in our previous study (35). The beads were reisolated, washed, eluted, and then resolved by SDS-PAGE for immunoblotting or mass spectrometry (Taplin mass spectrometry facility, Harvard Medical School, Boston, MA).
PNUTS regulates the mitotic progression of mammalian cells
To investigate the expression level of PNUTS during the cell cycle, HeLa cells were released into cell-cycle progression from thymidine arrest. Interestingly, PNUTS expression is strongly elevated in mitosis (Fig. 1A), consistent with our previous finding in Xenopus egg extracts (35). Compared with cyclin B1, which gradually accumulates prior to mitotic entry, PNUTS is more abruptly upregulated in M-phase (Fig. 1A). We then sought to examine the role of PNUTS in mammalian cell-cycle progression. siRNA-mediated PNUTS knockdown significantly reduced the percentage of mitotic cells (Fig. 1B and C), suggesting a role of PNUTS in mitotic entry. Consistently, a synchronized mitotic entry was not evident in cells treated with PNUTS siRNA, and then released from thymidine arrest (Fig. 1D). In addition to siRNA, the expression of PNUTS was also suppressed by shRNA, which targeted a distinct sequence of PNUTS (Fig. 1E). Real-time microscopy revealed blocked mitotic entry in cells expressing PNUTS shRNA (Fig. 1F). Moreover, although a smaller portion of cells with PNUTS knockdown still entered mitosis, they exhibited high levels of chromosome misalignment in metaphase and chromatin bridges in anaphase (Fig. 2A and B). Collectively, these lines of evidence demonstrated an important role of PNUTS in mitotic regulation.
PNUTS associates with Aurora kinases and kinetochore components
We speculated that PNUTS modulates the action of PP1 toward a subset of mitotic substrates. Thus, to reveal mechanistic insights into the specific function of PNUTS, we used mass spectrometry to identify proteins that copurified with PNUTS. Interestingly, the analysis identified Aurora A/Tpx2, Aurora B/Incenp complexes, and a number of kinetochore proteins, including Cenp-E and Bub1 (Fig. 3A; Supplementary Table S1). Immunoprecipitation in human cell lysates confirmed PNUTS association with Aurora A, Aurora B, Cenp-E, Bub1, and PP1 at the endogenous level (Fig. 3B). Consistently, the reciprocal immunoprecipitation of Aurora B recovered both PNUTS and PP1 (Fig. 3C).
PNUTS modulates PP1-dependent dephosphorylation of Aurora A/B
Identification of Aurora A and B complexes as associated proteins of PNUTS prompted us to examine the potential role of PNUTS in regulation of Aurora kinases. It has been shown that PP1 antagonizes the functions of both Aurora A and Aurora B (14, 15, 22, 38). Interestingly, PNUTS blocked the PP1-dependent dephosphorylation of Aurora B kinase in a reconstitutive phosphatase assay (Fig. 4A and B). A similar impact of PNUTS on PP1-mediated dephosphorylation of Aurora A kinase was also observed (Fig. 4C). Moreover, the collection of mitotic cells with PNUTS knockdown exhibited lower levels of Aurora A and B activation, as judged by autophosphorylation (Fig. 4D). The deficient Aurora phosphorylation was rescued with reexpression of WT, but not W401A, PNUTS (Supplementary Fig. S1). Because the W401A mutation disrupts the PNUTS–PP1 association (24, 30, 35), our findings indicate that PNUTS modulates Aurora kinases in a PP1-dependent manner. Furthermore, immunofluorescent analysis of individual mitotic cells showed those that exhibited metaphase and anaphase defects due to PNUTS knockdown lacked efficient phosphorylation of Aurora A/B (Fig. 5A). Histone H3 Ser-10 is hyperphosphorylated in mitosis, mediated primarily by Aurora B (39). We observed that PNUTS depletion significantly lowered the level of H3 Ser-10 phosphorylation in metaphase and anaphase (Fig. 5B).
PNUTS depletion disrupts the mitotic localization of the chromosomal passenger complex
Upon metaphase-to-anaphase transition, the Aurora B–containing chromosomal passenger complex (CPC) relocates from anaphase chromosomes to the cell equator where it mediates mitotic exit (3, 4, 40). Interestingly, we observed altered localization of Aurora B in anaphase with PNUTS knockdown: in anaphase cells with PNUTS knockdown, Aurora B was trapped on chromosomes rather than being released to spindles at the cell equator (Fig. 5C). The dysregulation of Aurora B localization correlated with mitotic chromosome missegregation after PNUTS silencing (Fig. 5D). Moreover, a similar defect of INCENP localization was observed in PNUTS-depleted cells (Fig. 5E).
The spatiotemporal regulation of PNUTS in mitotic progression
The role of PNUTS in mitosis prompted us to ask how PNUTS itself is spatiotemporally modulated. As expected, PNUTS is exclusively localized in the cell nucleus in interphase. From late prometaphase to early telophase, the majority of PNUTS is removed from chromatin, with a small portion resides on spindles and poles (Supplementary Fig. S2). The reaccumulation of PNUTS on chromosomes begins in telophase (Supplementary Fig. S2), suggesting a potential involvement of PNUTS in the decondensation of mitotic chromosome. Moreover, consistent with the protein association of PNUTS with kinetochore proteins Cenp-E and Bub1 (Fig. 3), a portion of PNUTS colocalizes with kinetochore proteins in interphase cells, as marked by Cenp-E and ACA. However, PNUTS is delocalized from kinetochores in metaphase (Fig. 6A and B), indicating the dynamic dissociation of PNUTS from kinetochores.
PNUTS depletion attenuates the activation of the spindle checkpoint
The role of PNUTS in regulation of Aurora kinases, and the dynamic kinetochore-association of PNUTS suggest a role of PNUTS in kinetochore signaling and the activation of the spindle assembly checkpoint. Interestingly, PNUTS depletion accelerated the mitotic progression from nuclear envelope breakdown (NEBD) to the onset of anaphase (Fig. 6C). Moreover, PNUTS siRNA impaired metaphase arrest after nocodazole treatment (Fig. 6D), and diminished the phosphorylation and kinetochore localization of BubR1 (Fig. 6E and F). Together, these lines of evidence indicate a role of PNUTS in the activation of the spindle checkpoint.
PNUTS is a potential drug target to overcome the cancer resistance to Aurora inhibition
An important lesson learned from existing genetic and cellular studies is that, although Aurora kinases play important functions in mitosis, depletion of Aurora kinases can be well tolerated via downregulation of the counteracting PP1 (16–19). As we discovered PNUTS as a PP1-modulator toward Aurora kinases and their substrates, we speculated that PNUTS targeting would synergize with Aurora inhibition in conferring cytotoxicity. Interestingly, we observed that, compared with the treatment of an Aurora inhibitor, ZM447439, alone, the combination of PNUTS siRNA and ZM447439 more profoundly decreased the viability of breast cancer ZR75-1 cells (Fig. 7A). This breast cancer cell line was selected because it was previously shown to be resistant to Aurora kinase inhibition (41). Consistently, PNUTS siRNA and ZM447439 synergistically induce cell death (Fig. 7B). The synergistic effect between PNUTS siRNA and ZM447439 was also confirmed in BT20, another breast cancer cell line, and HeLa (Supplementary Fig. S3A and B).
PNUTS modulates PP1 during mitotic progression
An evolutionarily conserved role of PP1 in mitotic regulation has been well established in yeast, Drosophila, Xenopus, and mammalian cells (14, 15, 42, 43). Mammalian PP1 exhibits dynamic and isoform-specific localization at kinetochores, centrosomes, chromosomes, and midbodies. Functionally, PP1 modulates the cell-cycle transitions of mitotic entry and exit, and controls various aspects of mitotic progression, including nuclear envelope assembly, kinetochore signaling and spindle checkpoint, chromosome architecture, and cytokinesis (14, 15, 42, 43). The complex functions of PP1 in mitosis rely on many targeting subunits that direct PP1 to specific mitotic substrates. For example, Repo-Man, KNL1, SKA, AKAP149, and Ki67 bind PP1, and recruit PP1 to anaphase chromosome, kinetochore, nuclear lamins, and other structures to govern mitotic progression (19, 43–47).
Interestingly, mitotic regulation of PP1 involves a number of mechanisms that negatively modulate PP1 activities. For example, PP1 is mitotically phosphorylated by CDK1, resulting in suppression of PP1 activity (48). Moreover, a specific inhibitor of PP1, inhibitor-1, binds PP1 during mitosis to reinforce PP1 inhibition (49). Another PP1 inhibitor, inhibitor-2, modulates the action of PP1 toward multiple mitotic kinases, including Nek2A, Aurora, A and Aurora B (18, 50, 51). Our previous study in Xenopus egg extracts suggested a role of PNUTS in the mitotic modulation of PP1 (35). In this study, we first confirmed the role of PNUTS in mammalian mitosis: PNUTS expression oscillates during the cell cycle and peaks in mitosis; PNUTS knockdown reduced mitotic entry, and induced severe chromosome missegregation. We then identified Aurora A, Aurora B-containing complexes, and several kinetochore proteins as specific binding patterns of PNUTS. PNUTS depletion suppressed the activation of Aurora A and Aurora B kinases, inhibited the relocalization of the Aurora B–containing chromosomal passenger complex from anaphase chromosomes to the cell equator, and bypassed the activation of the spindle checkpoint. Thus, our findings revealed a new mechanism involved in the inhibitory regulation of mitotic PP1. In particular, this mechanism is responsible for the efficient activation and function of Aurora kinases in mitotic progression. Although the experiments were primarily carried out in HeLa cells, owing to the high efficiency of siRNA transfection in these cells, a portion of the studies were confirmed in additional breast cancer cell lines. Moreover, the role of PNUTS in mitotic progression was first revealed in Xenopus egg extracts (35), indicating the well-conserved nature of this regulatory mechanism.
It remains a very intriguing question why the cell employs multiple distinct mechanisms to antagonize the activity of PP1 in mitosis. We speculate that the complex pattern of PP1 modulation is necessary to achieve the proper regulation of PP1 during mitotic progression. On one hand, the bulk of PP1 activity needs to be inhibited to allow the phosphorylation of mitotic substrates, which would otherwise be dephosphorylated by PP1. To this end, the antimitotic function of PP1 has been well established as an essential mechanism that promotes mitotic exit (42, 49). As an example, recent studies discovered a role of PP1 in the dephosphorylation of MASTL (also known as Greatwall kinase) during mitotic exit (36, 52–54). On the other hand, portions of PP1 bind specific mitotic structures and substrates, and play an active role in promoting several key steps of mitotic progression. For example, it has been shown that mitotic PP1 governs the proper microtubule–kinetochore attachment, silences the spindle checkpoint, and maintains the proper phosphorylation gradient and chromosome architecture (15, 42, 43). Interestingly, general inhibition of mitotic PP1 via overexpression of nuclear inhibitor of PP1 (NIPP1) in HeLa cells caused prometaphase arrest, spindle-formation and chromosome-congression defects, and hyperactivation of the spindle checkpoint (55). In principle, the involvement of multiple inhibitory mechanisms allows for the dynamic, localization-dependent, and substrate-specific modulation of PP1, although additional molecular details are needed to further delineate this model.
PNUTS mediates kinetochore signaling and spindle checkpoint activation
Our study revealed PNUTS association with kinetochore proteins, Cenp-E and Bub1. A portion of PNUTS colocalizes with Cenp-E and kinetochores in interphase, whereas PNUTS is excluded from kinetochores and chromosomes in metaphase. The dynamic kinetochore association and dissociation of PNUTS is potentially interesting. It is well recognized that reversible phosphorylation of kinetochore components plays a pivotal role in governing spindle attachment and the activation/deactivation of the spindle checkpoint. Several kinetochore components, especially Knl1, yield PP1-binding activities, and were reported to recruit PP1 to kinetochores (19). After the proper spindle–kinetochore attachment, PP1 dephosphorylates Aurora B, Knl1, BubR1, and other kinetochore proteins to silence the spindle checkpoint and trigger metaphase–anaphase transition (22). This switch-like, abrupt activation of PP1 may involve mechanisms that prevent the premature action of PP1. Thus, it is plausible that PNUTS plays a role in suppressing the premature activation of PP1, and that the kinetochore dissociation of PNUTS allows the timely activation of PP1 on kinetochores, which then triggers metaphase–anaphase transition. Further investigations are needed to better delineate the cell cycle–dependent kinetochore localization of PNUTS, as well as the detailed function of PNUTS in kinetochore signaling.
PNUTS is a potential drug target to overcome the resistance to Aurora kinase inhibition
Small-molecule inhibitors of Aurora kinases have been suggested as promising therapeutics for breast, ovarian, lung, and colon cancers, as well as hematologic malignancies (6, 8). Unfortunately, the clinical benefits of these therapeutic agents are often limited by drug resistance. Therefore, revealing mechanisms that allow cells to tolerate Aurora kinase inhibition is a crucial step toward development of strategies that overcome drug resistance. Interestingly, existing genetic and cellular studies showed that, although Aurora kinases play important functions in mitosis, depletion of Aurora kinases can be almost completely tolerated via downregulation of the counteracting PP1 (16–19). As we discovered PNUTS as a PP1-modulator toward Aurora kinases and substrates, we speculated that that mitotic modulation of PP1 by PNUTS may contribute to the cellular resistance to Aurora inhibition. We confirmed that PNUTS depletion sensitized resistant breast cancer cells to Aurora inhibition. Thus, it is plausible that PNUTS upregulation constitutes a mechanism of cancer resistance to Aurora inhibition. As such, PNUTS expression may serve as a prognostic indicator for Aurora resistance, and targeting PNUTS expression or its PP1 interaction, in conjunction with Aurora inhibition, may expand the therapeutic window, particularly in cancer cells that respond poorly to the current clinical inhibitors of Aurora kinases.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Conception and design: F. Wang, A. Peng
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): F. Wang, L. Wang, L.A. Fisher
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): F. Wang, L. Wang, A. Peng
Writing, review, and/or revision of the manuscript: F. Wang, W. Wang, A. Peng
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Li
Study supervision: C. Li, W. Wang, A. Peng
Microscopic analysis was performed at the UNMC Advanced Microscopy Core Facility, supported by the Nebraska Research Initiative, the Fred and Pamela Buffett Cancer Center support grant (P30CA036727), and an Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the NIH (P30GM106397). This work was supported by NIH grant R01CA172574 to A. Peng.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.