Both activating and inactivating mutations in protein tyrosine phosphatase Ptpn11 (encoding Shp2) are associated with tumorigenesis. However, the underlying mechanisms remain unclear. Here, we show that Shp2 plays an important role in mitosis, dysregulation of which results in chromosome instability and cancer predisposition. Depletion of Shp2 compromised the mitotic checkpoint. Shp2-depleted cells exhibited a delay in mitotic entry and an earlier mitotic exit. Moreover, Shp2 deficiency caused defective kinetochore-microtubule attachment, chromosome misalignment, chromosomal congression defects, lagging chromosomes, and chromosome missegregation. Reintroduction of wild-type Shp2, but not a catalytically deficient mutant, restored the checkpoint function and chromosome alignment at metaphase in Shp2-deficient cells, establishing a requirement for the catalytic activity of Shp2 during mitosis. Further analyses revealed that Shp2 was required for the optimal activation of the mitotic kinases PLK1 and Aurora B and thereby the proper kinetochore localization and phosphorylation of BubR1, a core mitotic checkpoint protein that is also critical for chromosome alignment. Together, our findings show a previously unrecognized role for Shp2 in the maintenance of chromosome stability and suggest a new mechanism by which dysregulation of Shp2 signaling contributes to malignancy development. Cancer Res; 72(20); 5296–306. ©2012 AACR.
Mitosis is the biologic process by which a eukaryotic cell precisely separates replicated sister chromatids into 2 daughter cells during cell division. Deregulated mitotic activities are associated with chromosomal instability and aneuploidy, a key feature of many cancers (1–3). Mitosis is an extraordinarily complex process. Many proteins are involved in the regulation of mitosis. Among these proteins, polo-like kinases (PLK), especially, PLK1 is involved in a variety of mitotic events including the G2–M transition, centrosome maturation and separation, mitotic spindle formation, chromosome segregation, and cytokinesis (4–6). It exerts its multiple functions by phosphorylating a wide spectrum of substrates. In addition, members of the Aurora kinase family, such as Aurora B, have emerged as key mitotic regulators. They are involved in centrosome separation and maturation, spindle assembly and stability, chromosomal condensation, congression and segregation, and cytokinesis (5, 7, 8). Aurora kinases function also by phosphorylating many substrate proteins.
Multiple fidelity-monitoring checkpoint systems have evolved to ensure correct temporal and spatial coordination of mitosis. The mitotic checkpoint, also known as the spindle assembly checkpoint, is a cellular surveillance mechanism that acts to inhibit entry into anaphase until all chromosomes have successfully attached to spindle microtubules (1, 2, 9). This checkpoint is mediated by a set of highly conserved proteins including Mad1, Mad2, Mad3/BubR1, Bub1, and Bub3 (2). Proper recruitment of these checkpoint proteins to unaligned kinetochores, the protein structures on chromosomes where the spindle microtubules attach, is essential for the mitotic checkpoint to delay anaphase onset. Displacement of these proteins from kinetochores is associated with mitotic checkpoint defects and chromosomal instability. Among them, BubR1 is a central checkpoint protein in term of its essential role for kinetochore localization of other mitotic checkpoint proteins (10). Aurora B kinase is also involved in the regulation of the mitotic checkpoint (11). Inhibition of Aurora B disturbs kinetochore localization of BubR1, Mad2, and Cenp-E, thereby compromising mitotic checkpoint function (12). In addition, a recent study has shown that Aurora B plays a direct role in initiating the mitotic checkpoint by potentiating Mps1 activation (13).
In addition to the mitotic checkpoint, temporal and spatial regulation of kinetochore-microtubule (KT-MT) attachment is essential for error-free chromosome segregation. All chromosomes/kinetochores must be attached to the spindle microtubules and properly aligned to the metaphase plate for accurate chromosome segregation. Depletion of BubR1 not only disturbs the mitotic checkpoint but also leads to severe chromosome misalignment (12), suggesting that BubR1 plays a dual role in both checkpoint signaling and chromosome alignment. Several lines of evidence show that BubR1 is required for stable chromosome-spindle attachments (14, 15). BubR1 is highly phosphorylated in mitosis, and that phosphorylation of BubR1 is important for maintaining proper KT-MT attachments and chromosome alignment (14, 15). Recent studies have shown that BubR1 is a substrate of PLK1 and that PLK1 regulates chromosome alignment through phosphorylation of BubR1 (14, 16).
Shp2, a ubiquitously expressed Src homology 2 domain-containing protein tyrosine phosphatase, has been implicated in diverse cytoplasmic signaling processes activated by growth factors, hormones, cytokines, and extracellular matrix proteins (17, 18). Mutations in Ptpn11 (encoding Shp2) are involved in several human diseases, including cancers (19–21). Although activating mutations in Ptpn11 are associated with Noonan syndrome, childhood leukemias, and sporadic solid tumors, Leopard syndrome patients with Ptpn11 inactivating mutations also have increased risks of developing malignancies (22–25). In addition, a recent study has shown that Ptpn11 deficiency enhances diethylnitrosamine-induced hepatocellular carcinoma development (26). However, the molecular mechanisms by which deregulated Shp2 signaling is linked to malignancies are not well understood because of lack of complete understanding of Shp2 function. The cytoplasmic function of Shp2 cannot fully explain how both activating and inactivating mutations in Shp2 induce tumorigenesis. Emerging evidence has indicated that Shp2 may also function in other cell organelles, such as the nucleus (27–31) and the mitochondria (32, 33). Understanding of the novel functions of Shp2 in these organelles may shed light on the molecular mechanisms of Shp2-associated tumorigenesis. In our recent studies investigating effects of activating mutations of Ptpn11 on malignant transformation of hematopoietic cells, we have discovered that gain of function of this phosphatase causes aneuploidy (34), which is known to be associated with aberrant mitosis. This unexpected finding led us to further explore the novel function of Shp2 in mitosis.
Materials and Methods
Cell culture, synchronization, and transfection
HeLa cell line used in this study were originally purchased from American Type Culture Collection and maintained 10 to 15 passages in Dulbecco's Modified Eagle's Medium supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. Mouse embryonic fibroblasts (MEF) generated from Ptpn11 conditional knockout (Ptpn11fl/fl) mice were provided by Dr. B. G. Neel. They were previously characterized (35) and were maintained as described above. To synchronize cells at the G1-phase, cells were treated with thymidine (2 mmol/L) for 17 hours, washed with PBS 3 times, and then incubated for 9 hours in fresh medium without thymidine. Thymidine (2 mmol/L) was added again for another 16 hours. To arrest the cells in mitosis, cells were treated with nocodazole (200 ng/mL and 400 ng/mL for HeLa cells and MEFs, respectively) for 16 hours. Mitotic cells were then isolated by “shake-off.” For release experiments, isolated mitotic cells were washed 3 times with PBS and plated into fresh medium for the indicated periods of time. To knockdown Shp2, HeLa cells were transfected with Shp2 siRNA corresponding to 5′-AAG AAU CCU AUG GUG GAA ACA-3′ using an Amaxa Nucleofector and a transfection kit according to the manufacturer's protocol (Lonza Group Ltd.).
Time-lapse video microscopy
Cells were plated in 35 mm culture dishes. Twenty-four hours later, cells were monitored using a Leica DMI600B inverted microscope equipped with an environmental control chamber. Images were taken every 3 or 5 minutes with a Retiga EXI 12 bit camera (Q imaging) and analyzed with MetaMorph software (Universal Imaging Corp.).
Immunostaining and confocal microscopy
For analysis of cold-stable microtubules, cells were fixed for 10 minutes at room temperature with 3.7% formaldehyde in 100 mmol/L Pipes at pH 6.8, 10 mmol/L EGTA, 1 mmol/L MgCl2, and 0.2% Triton X-100. For all other experiments, cells were fixed in 100% methanol at −20°C for 10 minutes. After fixation, cells were permeabilized with 0.25% Triton X-100 in PBS, and then blocked with 2% bovine serum albumin in PBS for 1 hour. All primary antibodies were incubated at room temperature for 1 hour or at 4°C overnight. Cells were then washed 3 times with PBS and incubated with Alexa 488- or Alexa 568-conjugated secondary antibodies for 1 hour. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). After washing, coverslips were mounted onto glass slides and stained cells were examined with a Nikon fluorescence microscope (Nikon). All confocal images were acquired using a Zeiss LSM 510 inverted laser-scanning confocal microscope (Zeiss) and analyzed with Zeiss and MetaMorph software. For staining of chromosome spreads, mitotic cells were obtained by shake-off and swelled in a prewarmed 75 mmol/L KCl hypotonic solution at 37°C for 15 minutes. After spun onto slides using cytospin, chromosomes were fixed and stained with indicated antibodies as above.
For the chromosome alignment assay, control or Shp2 siRNA transfected HeLa cells were synchronized by double thymidine block and released for 10 hours, and then arrested at metaphase by the treatment with MG132 (10 μmol/L) for 2 hours. Cells were fixed and processed for immunostaining. To count unaligned kinetochores, all images were acquired as Z-stack with 0.2 μm spacing using a ×100 objective. Misaligned kinetochores were defined as those without localizing to equatorial plate. At least 5 cells (400 kinetochores) were analyzed for each experiment.
Data are presented as mean ± SD. Statistical significance was determined using unpaired 2-tailed Student t test. *, P < 0.05; **, P < 0.01.
Depletion of Shp2 results in a delay in entering mitosis and causes a failure to sustain the mitotic arrest induced by nocodazole
To test the potential role of Shp2 in mitosis, we transfected HeLa cells with Shp2-specific siRNA and assessed cellular responses to nocodazole, a microtubule-depolymerizing agent that activates the mitotic checkpoint and arrests cells at mitosis. Note that 75% of control cells were arrested at the mitotic stage defined by 4n DNA content and phospho-histone H3 (Ser10)-positive staining. In contrast, knockdown of Shp2 substantially decreased nocodazole-induced mitotic arrest in a Shp2 dose-dependent manner (Fig. 1A and B). Similarly, the percentage of mitotic cells in Shp2-depleted cells treated with taxol, another microtubule poison, was also decreased significantly (Supplementary Fig. S1). To minimize potential influence of the cell-cycle status on cellular responses to mitotic arrest, asynchronized and synchronized G1 HeLa cells were treated with nocodazole and mitotic cells were assessed at multiple time points. Both asynchronized and synchronized Shp2 knockdown cells showed decreased mitotic arrest (Fig. 1C and Supplementary Fig. S2). Consistent with the reduction of mitotic cells, phosphorylation of histone H3 in Shp2-depleted cells was much lower than that in control cells and the inhibitory phosphorylation (Tyr15) of Cdc2 kinase (the kinase activity that is required for mitotic entry) was sustained at higher levels in Shp2 knockdown cells during nocodazole treatment (Fig. 1D).
However, the percentage of phospho-histone H3-negative Shp2-depleted cells with 4n DNA content were increased (Fig. 1A). This reflects a delay in G2 or an accumulation of tetraploid G1 cells escaped from mitosis without dividing. To clarify these possibilities, live synchronized cells at the G1-phase were cultured in the presence of nocodazole and monitored by time-lapse video microscopy. Consistent with the data presented in Fig. 1, Shp2 knockdown cells showed markedly decreased response to this microtubule poison (Fig. 2A and B). By carefully monitoring individual cells, we noticed that about 40% of Shp2 knockdown cells failed to enter metaphase, being arrested at the G2-phase. Furthermore, the percentage of Shp2-depleted cells that escaped from mitotic arrest was increased. In addition, Shp2-depleted mitotic cells underwent a higher incidence of death (Fig. 2C). We next focused on mitotic cells arrested by nocodazole and quantified the duration of mitotic arrest. In the presence of nocodazole, more than 50% of control cells that were arrested in mitosis stayed in mitosis for up to 16 hours, whereas approximately 80% of Shp2-depleted cells escaped from mitotic arrest without dividing and returned to G1 with tetraploid DNA content within 12 hours (Fig. 2D and E). These data suggest that Shp2 ablation results in a delay in entering mitosis as well as a premature mitotic exit despite the spindle defects caused by nocodazole.
A proper control of anaphase onset/mitotic exit depends on ubiquitin-mediated proteolysis of key regulators at the correct time. We collected nocodazole-arrested mitotic cells and examined levels of cyclin B1, PLK1, Mad2, and phospho-histone H3 after release in regular fresh culture medium. Consistent with the earlier mitotic exit found in the presence of nocodazole, degradation of PLK1 and cyclin B1 was accelerated in Shp2 knockdown cells in the absence of nocodazole (Fig. 2F and Supplementary Fig. S3). Phosphorylation of histone H3 was also decreased much faster in Shp2-depleted cells (Fig. 2F). In contrast, Mad2 was not significantly degraded in either control or Shp2-depleted cells (Fig. 2F).
Catalytic activity is required for Shp2 function in mitosis
Shp2 is a protein tyrosine phosphatase that functions in cell signaling in both catalytically dependent and independent manners (36–38). To further elucidate the structural bases for the important role of Shp2 in mitosis, WT Shp2 and mutant Shp2 with C459S mutation that abolishes its catalytic activity were transduced into conditional knockout (Ptpn11fl/fl) MEFs. Transduced cells were infected with adenovirus expressing Cre DNA recombinase (to deplete endogenous Shp2) followed by nocodazole treatment. Consistent with the data shown in Fig. 1, acute depletion of Shp2 decreased mitotic arrest significantly (Fig. 3A). Reintroduction of WT Shp2, but not Shp2 C459S, largely restored responses of Shp2-depleted cells to nocodazole-induced mitotic arrest (Fig. 3B and C). These rescue experiments, while demonstrating cell autonomous effects of Shp2 deficiency on mitosis, provide evidence that the catalytic activity of Shp2 is required for its function in this process.
Depletion of Shp2 causes chromosome misalignment and defective KT-MT attachment
To further determine the role of Shp2 in mitosis, unperturbed mitotic progression was directly visualized in a greater detail. HeLa cells stably expressing a histone H2B-GFP fusion protein that enables monitoring chromosomes in live cells with high resolution were transfected with control or Shp2 siRNA, and then monitored under a video microscope. The percentages of mitotic cells with misaligned and lagging chromosomes were markedly increased after Shp2 depletion (Fig. 4A–C). Moreover, some Shp2-depleted cells entered anaphase without chromosome aligning to the metaphase plate, indicative of chromosome congression defects (Fig. 4B and C), and in some cases, metaphase chromosomes underwent decondensation before the onset of anaphase (Fig. 4B). As a result, the unperturbed mitosis of Shp2-depleted cells was significantly prolonged because of these mitotic defects (Fig. 4D), similar to that of PLK1 deficient cells (39). In agreement with these chromosome analyses results, Shp2 knockdown primary MEFs developed increased aneuploidy compared with control cells (Fig. 4E).
To rule out the possibility that chromosome misalignment in Shp2-depleted cells was because of premature mitotic exit, we blocked the anaphase onset using the proteasome inhibitor MG132. Under these conditions, Shp2-depleted metaphase cells with chromosome misalignment (Fig. 5A) and telophase cells with lagging chromosomes (Supplementary Fig. S4) were still significantly increased. Further rescue experiments showed that add-back of WT Shp2, but not catalytically deficient Shp2 C459S, restored chromosomal alignment at metaphase (Fig. 5B), reaffirming the catalytic-dependent role of Shp2 in this process. Consistent with chromosome misalignment, more detailed visualization of kinetochores illustrated that Shp2-depleted metaphase cells had increased unaligned kinetochores (Fig. 5C). The chromosome misalignment phenotype suggests that Shp2 might be required for attachment of kinetochores to the spindle. Indeed, unattached kinetochores were frequently observed in Shp2-depleted cells (Fig. 5C, insets). To address the underlying cause of chromosome misalignment, we examined the stability of KT-MT attachment by exposing MG132-arrested mitotic cells to low temperature (0°C), which depolymerized unstable microtubules. In control cells, the spindle remained relatively intact, with most CREST (calcinosis, Raynaud's syndrome, esophageal dysmotility, sclerodactyly, telangiectasia) antibody–stained kinetochores clearly attached to microtubule fibers. In contrast, Shp2-depleted cells showed only few cold-stable KT-MT attachments, and most kinetochores had no microtubules attached to them (Fig. 5D).
Kinetochore localization of BubR1 and Aurora B kinase activity are reduced in Shp2-depleted cells
Proper localization to kinetochores is critical for mitotic proteins to execute their functions (40), and displacement of these proteins from kinetochores is associated with mitotic checkpoint defects. To elucidate the mechanisms of the defective mitotic checkpoint caused by Shp2 depletion, we examined the kinetochore localization of BubR1 and Mad2. In control mitotic cells, BubR1 was nicely localized to kinetochores. However, kinetochore signals of BubR1 were diminished in Shp2-depleted cells (Fig. 6A). To confirm this result, we conducted centromeric staining of chromosome spreads. As illustrated in Fig. 6B, the percentage of the metaphase spreads with strong BubR1 signals at kinetochores also was decreased significantly in Shp2-depleted cells. In contrast, the kinetochore localization of Mad2 was unaffected (data not shown). Previous studies have shown that Aurora B activity is important for kinetochore localization of BubR1 (12, 14, 41). We tested whether mislocalization of BubR1 might be associated with abnormal Aurora B activation and found that Aurora B kinase activity determined by phosphorylation of Thr232 (Fig. 6C) and kinase assays (Fig. 6D) was significantly decreased in Shp2-depleted cells. These results suggest that decreased Aurora B activity may be responsible for the diminished kinetochore localization of BubR1 in Shp2 knockdown cells.
Phosphorylation of BubR1 and PLK1 activity are decreased in Shp2-depleted cells
The kinetochore localization of BubR1 has been shown to be important for its phosphorylation, which is essential for KT-MT attachment and proper chromosome alignment (14, 16). We determined phosphorylation levels of BubR1 in mitotic cells. In response to nocodazole and taxol, phosphorylation of BubR1 was substantially decreased in Shp2-depleted cells compared with those in control cells (Fig. 7A). To exclude the possibility that the decreased phosphorylation of BubR1 was because of the reduction of mitosis in Shp2-depleted cells, phosphorylation levels of BubR1 were also determined in synchronized shake-off mitotic cells. Similar results were obtained from Shp2-depleted mitotic cells (Fig. 7B). Recent studies have shown that BubR1 is a substrate of PLK1 (14, 16). We assessed PLK1 kinase activity in Shp2-depleted mitotic cells. As seen in Fig. 7C and D, PLK1 activity determined by phosphorylation of Thr210 and kinase assays was decreased significantly in Shp2 knockdown cells. These results suggest that decreased PLK1 activity, followed by the consequent reduction of BubR1 phosphorylation, leads to chromosome misalignment in Shp2-depleted cells.
Shp2, a ubiquitously expressed protein tyrosine phosphatase, has been well recognized for its important role in the signal transduction of growth factors, cytokines, and extracellular matrix proteins (17, 18). It is implicated in multiple signaling pathways by interacting with numerous signaling proteins. However, the information gathered from previous studies represents the role of Shp2 in interphase, likely G0–G1 cells, because the cells used in those studies were usually growth factor starved. By focusing on mitotic cells we now show that Shp2 also plays an important role in maintenance of chromosome stability. Both activating mutations and inactivating mutations in Ptpn11 (Shp2) are associated with tumorigenesis. Leopard syndrome patients with Ptpn11 inactivating mutations also have increased risks of developing malignancies (23–25). As inactivated Leopard syndrome mutant Shp2 functions in a dominant negative manner because of the increased binding of mutant Shp2 to signaling partners (22), it is possible that the mutant Shp2 contributes to malignancy development by decreasing Shp2 catalytic activity and thereby causing chromosomal instability although it may reduce cell growth.
Shp2 seems to facilitate the mitotic checkpoint and chromosome alignment by promoting PLK1 and Aurora B activation. PLK1 and Aurora B are multifunctional kinases that are critical for all stages of mitosis. Overexpression of these kinases has been observed in tumors, suggesting that these kinases may act as oncogenes (42–45). However, missense mutations in PLK1 have also been identified in several cancer cell lines (46). Moreover, inhibition of PLK1 leads to delayed mitotic entry and cells are arrested in mitosis with a monopolar or disorganized spindle (39, 47). Inhibition of Aurora B in cells causes misaligned chromosomes, perturbed cytokinesis, and resistance to taxol-induced mitotic arrest (12). As both PLK1 and Aurora B kinase activities were decreased in Shp2-depleted cells (Figs. 6C, D, 7C, and D), the phenotypes displayed by these cells may reflect combined effects of PLK1 and Aurora B deficiencies. Delayed entry into mitosis and prolonged duration of unperturbed mitosis are likely the consequences of decreased PLK1 activity while failure to sustain microtubule poisons-induced mitotic arrest, chromosomal misalignment, chromosomal congression defects, and chromosomal missegregation might result from decreased Aurora B activity. Aurora B has been shown to be required for the kinetochore localization of BubR1 (12, 14, 48), a core mitotic checkpoint protein. This accounts for the diminished kinetochore localization of BubR1 in Shp2-depleted cells (Fig. 6A and B). Because recruitment of BubR1 to unattached kinetochores is necessary to maintain the mitotic checkpoint, it is likely that loss of Shp2 compromises this checkpoint by interrupting kinetochore localization of BubR1. PLK1 and Aurora B have been shown to be required for the phosphorylation of BubR1 (12, 14, 16), which is important for chromosome alignment. BubR1 phosphorylation is significantly decreased in Shp2 knockdown cells (Fig. 7A and B). Thus, it seems that Shp2 depletion causes chromosome misalignment also by reducing PLK1 and Aurora B activation. A proposed model of Shp2 function in the maintenance of genomic stability is shown in Fig. 7E.
How Shp2 promotes activation of PLK1 and Aurora B kinases remains to be further determined. Shp2 plays an overall positive role in the activation of receptor and cytosolic kinases despite its direct function in protein dephosphorylation (17, 18). The underlying mechanism is completely unclear. As such, it is not understood why PLK1 activity is decreased in Shp2-depleted mitotic cells. Aurora B kinase activity is also decreased by depletion of Shp2 (Fig. 6C and D). This is likely the result of decreased PLK1 activity because of the well recognized extensive interplay between PLK1 and Aurora B kinases (5). Further studies are needed to define the detailed molecular mechanisms by which Shp2 promotes PLK1 activity. Shp2 seems to function in this context in a catalytic-dependent manner as only WT Shp2, but not catalytically-deficient Shp2 C459S, restores nocodazole-induced mitotic arrest and chromosome alignment in Shp2-depleted cells (Figs. 3 and 5B). It seems that a general mechanism might be involved in the functional interactions between Shp2 and its interacting kinases. Elucidation of rudimental biochemical activities of this phosphatase may shed new light on its acting mechanisms in these signaling pathways.
Disclosure of Potential Conflicts of Interests
No potential conflicts of interest were disclosed.
Conception and design: X. Liu, H. Zheng, C.-K. Qu
Development of methodology: X. Liu, H. Zheng
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X. Liu, H. Zheng
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Liu, H. Zheng
Writing, review, and/or revision of the manuscript: X. Liu, C.-K. Qu
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C.-K. Qu
Study supervision: C.-K. Qu
Technical training: C.-K. Qu
The authors are extremely grateful to Dr. Hongtao Yu of the University of Texas Southwestern Medical Center and Howard Hughes Medical Institute for providing reagents, advice on experimental designs, and critical reading of the manuscript.
This work was supported by the NIH grants HL068212 and HD070716 (C.K. Qu).