Although the kinase CHK1 is a key player in the DNA damage response (DDR), several studies have recently provided evidence of DDR-independent roles of CHK1, in particular following phosphorylation of its S280 residue. Here, we demonstrate that CHK1 S280 phosphorylation is cell cycle-dependent and peaks during mitosis. We found that this phosphorylation was catalyzed by the kinase PIM2, whose protein expression was also increased during mitosis. Importantly, we identified polo-like kinase 1 (PLK1) as a direct target of CHK1 during mitosis. Genetic or pharmacological inhibition of CHK1 reduced the activating phosphorylation of PLK1 on T210, and recombinant CHK1 was able to phosphorylate T210 of PLK1 in vitro. Accordingly, S280-phosphorylated CHK1 and PLK1 exhibited similar specific mitotic localizations, and PLK1 was co-immunoprecipitated with S280-phosphorylated CHK1 from mitotic cell extracts. Moreover, CHK1-mediated phosphorylation of PLK1 was dependent on S280 phosphorylation by PIM2. Inhibition of PIM proteins reduced cell proliferation and mitotic entry, which was rescued by expressing a T210D phosphomimetic mutant of PLK1. Altogether, these data identify a new PIM–CHK1–PLK1 phosphorylation cascade that regulates different mitotic steps independently of the CHK1 DDR function.
Checkpoint kinase 1 (CHK1, encoded by CHEK1) is a conserved serine/threonine protein kinase, widely known as a master regulator of the DNA damage response (DDR) pathway. Phosphorylation of its substrates, including p53, CDC25 or Wee1, leads to cell cycle arrest and the initiation of DNA repair, thus protecting against cell death induced by genotoxic stress. CHK1 knockout mice are embryonic lethal (Lam et al., 2004; Liu et al., 2000), and Chk1+/– mice exhibit hematopoietic defects (Boles et al., 2010). CHK1-deficient blastocysts and embryonic stem cells also show severe proliferation defects as well as an impaired cell cycle checkpoint response (Liu et al., 2000; Takai et al., 2000). Regulation of CHK1 activity by the upstream kinase ATR occurs through the phosphorylation of CHK1 on S345 and S317. Phosphorylation at these sites is considered to be a hallmark of CHK1 activation in cells. In addition, CHK1 phosphorylates itself on S296 to achieve full activation. Beside these well-described activating mechanisms, CHK1 phosphorylation on S280 by the mitogenic serine/threonine kinases AKT, p90RSK, and PIM1 and PIM2 (PIM1/2), was also recently reported in different models, with diverse consequences on CHK1 activity and its subcellular localization. Our own studies in leukemic cells have suggested that PIM1/2-dependent phosphorylation of this residue is crucial for resistance to DNA-damaging agents (Yuan et al., 2014a) and for cell proliferation in unperturbed conditions (Yuan et al., 2014b).
It is now clear that CHK1 also plays important roles in cell functions that are independent of the DDR pathway, although these pathways have not yet been fully established. CHK1 monitors DNA replication during unperturbed S phase, and there is evidence for its involvement in the control of replication origin firing, elongation and fork stability (Petermann et al., 2010; Guo et al., 2015). In addition to DNA replication, recent reports have documented the involvement of CHK1 in mitosis, particularly in regulating the initiation, progression and fidelity of unperturbed mitosis (Krämer et al., 2004; Enomoto et al., 2009; Tang et al., 2006; Zachos et al., 2007). In line with this, CHK1 haploinsufficiency in mice results in multiple mitotic defects such as cytokinetic regression and increased binucleation (Peddibhotla et al., 2009). CHK1 depletion in primary mammary epithelial cells isolated from Chk1+/– mice also causes apparent defects in kinetochore function, activates the spindle assembly checkpoint, and eventually leads to mitotic catastrophe (Peddibhotla et al., 2009). Moreover, CHK1 appears to be essential for normal mitotic progression in HeLa cells and is involved in the proper alignment of chromosomes to the equatorial plane during metaphase, possibly via the regulation of PLK1, in order to inactivate the spindle assembly checkpoint (Tang et al., 2006). Other reports have described a functional interaction between CHK1 and the Aurora B kinase, leading to increased Aurora B kinase activity (Petsalaki et al., 2011; Zachos et al., 2007). Finally, the mitotic functions of CHK1 also involve the histone-dependent transcriptional regulation of CDK1 and its activating partner cyclin B1 (Shimada et al., 2008). Interestingly, CHK1 is also a target of cyclin-dependent kinase 1 (CDK1). CHK1 can be phosphorylated at S286 and S301 by CDK1 during mitosis, promoting its translocation from the nucleus to the cytoplasm during prophase (Enomoto et al., 2009).
Together, these studies provide evidence for the existence of DNA damage-independent functions of CHK1 during mitosis; however, there are still many questions regarding the precise mechanisms involved and the identity of the molecular targets of CHK1 in this context. In this study, we established that CHK1 phosphorylation at S280 is strongly increased during mitosis and that PIM2 is the kinase controlling this phosphorylation. In turn, S280-phosphorylated CHK1 is able to directly phosphorylate PLK1 on its conserved T210 residue, leading to PLK1 activation. Taken together, these data identify a novel pathway governing CHK1-mediated regulation of PLK1 activity during mitosis.
CHK1 is a target of PIM2 during mitosis
We have previously established that interfering with CHK1 phosphorylation on S280 influences the proliferation rate of leukemic cells (Yuan et al., 2014a). In order to better understand the function of this phosphorylation in the regulation of cell proliferation, we first asked whether CHK1 phosphorylation on S280 was regulated during the cell cycle. For this, we followed the P-S280 level of CHK1 during cell cycle progression by western blot analysis of synchronization experiments. We used an antibody against phosphorylated S280 that we first characterized for its specificity and efficiency by western blot and immunofluorescence (Fig. S1A-C). In H1299 cells released from a thymidine block after serum starvation, the highest level of phosphorylation of S280 was observed during G2/M progression (Fig. 1A). Similar results were obtained with the human osteosarcoma U2OS and the hTERT-immortalized foreskin fibroblast hTERT BJ cell lines (Fig. S1D,E). Accordingly, CHK1 phosphorylation on S280 was strongly increased in HeLa, U2OS and hTERT BJ cell lines blocked in mitosis with nocodazole (Fig. 1B and Fig. S1F). These data indicate that CHK1 is maximally phosphorylated on S280 during mitosis, suggesting that S280 phosphorylation of CHK1 may play an important role in controlling its function during this phase of the cell cycle.
We then asked which kinase was involved in this phosphorylation event. The PIM, AKT and p90RSK family kinases have been reported to directly phosphorylate this residue in various cell types (Puc et al., 2005; Li et al., 2012; Yuan et al., 2014a). Thus, we tested the impact of inhibiting these kinases on CHK1 S280 phosphorylation in U2OS cells blocked in mitosis with nocodazole. As seen in Fig. 1C, inhibition of PIM1 and PIM2 (PIM1/2) for 4 h almost completely abolished S280 phosphorylation under these conditions, whereas inhibition of AKT only had a limited effect and MEK inhibition did not significantly modify S280-phosphorylated CHK1 (P-CHK1) levels. In order to rule out any indirect effects, we verified that the inhibitors did not significantly induce cell death (Trypan Blue staining, Fig. S1G) after 4 h of treatment. These results were confirmed by using RNA interference-mediated down-regulation of PIM1/2 in these cells, which also dramatically decreased CHK1 S280 phosphorylation (Fig. 1D). Altogether, these results strongly suggest that PIM1/2 are important inducers of S280 CHK1 phosphorylation during mitosis.
Cell cycle-dependent expression of PIM2 has never been reported. Therefore, we performed synchronization experiments in UT-7 cells, which are known to express high levels of PIM2 (Adam et al., 2015), in order to assess the expression profile of PIM2 during the cell cycle. As shown in Fig. 1E, PIM2 protein levels progressively increased during S phase and peaked during mitosis in these experiments, correlating with histone H3 phosphorylation levels. In contrast, PIM1 remained almost constant throughout the cell cycle. We then analyzed the ability of PIM2 to phosphorylate CHK1 on S280 in vitro, as we already established previously that PIM1 is involved in this phosphorylation in leukemic cells (Yuan et al., 2014a). Recombinant GST-CHK1 protein was incubated with purified PIM2 kinase in the presence of ATP (Fig. 1F), and CHK1 phosphorylation on S280 was then detected by western blotting. These experiments showed that PIM2 directly phosphorylates CHK1 on S280 as pharmacological inhibition (SGI-1776) of recombinant PIM2 kinase blocked this phosphorylation. We confirmed these results by overexpressing PIM2 in U2OS and HeLa cells, which led to a dramatic increase in CHK1 S280 phosphorylation (Fig. 1G). CHK1 S280 phosphorylation levels were also significantly increased in mitotic HeLa cells expressing an inducible PIM2 kinase (Fig. 1H). Altogether, these data indicate that PIM2 is a mitotic kinase that phosphorylates CHK1 on S280 during mitosis.
Subcellular localization of PIM2 and S280-phosphorylated CHK1 during mitosis
Our data describe for the first time that both PIM2 expression and CHK1 S280 phosphorylation are high during mitosis. We therefore investigated the subcellular localization of PIM2 and CHK1 during this phase. Since endogenous PIM2 is barely detectable by immunofluorescence, we used cells inducibly overexpressing the HaloTag-PIM2 fusion protein to study the subcellular localization of PIM2 in nocodazole-treated HeLa cells. For these experiments, co-staining was performed with an antibody against PLK1, a well-established marker of the different mitotic steps (Strebhardt, 2010). Indeed, PLK1 localization has been well characterized during mitosis, with centrosome, kinetochore and midbody localization consistent with its multiple mitotic functions. We detected PIM2 associated with spindle poles in metaphase, after which it then relocalized to the equatorial plane where spindle microtubules overlap in the midzone as cells go through anaphase, and finally, it was observed in the midbody during cytokinesis (Fig. S2A). Interestingly, apparent co-localization of PIM2 and PLK1 was visible in these experiments, suggesting a possible functional link between these two proteins.
We then determined the distribution of S280-phosphorylated CHK1 during mitosis in H1299 cells (Fig. S2B), after validation of the anti-P-S280 antibody used in these experiments (Fig. S1B). During metaphase, phosphorylated CHK1 was found at the periphery of chromosome arms, as has already been described for the CHK1 protein (Peddibhotla et al., 2009). During anaphase, P-CHK1 was concentrated at the spindle midzone, and during cytokinesis, P-CHK1 accumulated at the midbody where it partly co-localized with citron kinase (Fig. S2C).
Based on these data, we then performed immunolabeling of P-CHK1 and PLK1 in U2OS (Fig. 2A) and HeLa cells (Fig. S2D), in order to investigate the subcellular localization of these two proteins during mitosis. The same localization patterns described above were observed, wherein P-CHK1 localized to the peri-chromosomal layer (PCL) during prometaphase, whereas PLK1 marked kinetochores. At this stage and during metaphase, both partners closely localized with each other (Pearson's coefficients of r=0.612 and r=0.860, respectively). At the onset of sister chromatid separation during anaphase, PLK1 was translocated to the spindle midzone and phosphorylated CHK1 foci partially co-localized with it. Finally, during cytokinesis, a small fraction of phosphorylated CHK1 co-localized with PLK1 at the midbody (Fig. 2A). To validate these observations, proximity ligation assays were performed to detect the co-localization of PLK1 with S280-phosphorylated CHK1 in U2OS cells treated or not with nocodazole. A significant proximity between the two proteins was visualized in nocodazole-treated U2OS cells, consistent with data obtained on normal growing (asynchronous) HeLa cells where we specifically visualized the mitotic cells (Fig. 2C). Similar results were obtained on UT-7 cells following nocodazole treatment (Fig. S2E). Together, these results indicate that a fraction of PIM2, P-CHK1 and PLK1 co-localize at distinct locations at different stages of the mitotic process (Fig. 2 and Fig. S2).
CHK1 phosphorylates PLK1 on T210 during mitosis
To test the possibility that PIM2, P-CHK1 and PLK1 functionally interact during mitosis, we then asked whether CHK1 and/or PIM kinases can regulate PLK1 phosphorylation on T210, a hallmark of PLK1 activation. RNA interference experiments performed in nocodazole-blocked U2OS cells demonstrated that down-regulating either CHK1 or PIM1/2 strongly decreased PLK1 T210 phosphorylation (Fig. 3A). Since cells are blocked with nocodazole in these experiments, it is unlikely that these effects on PLK1 phosphorylation are due to indirect cell cycle modifications. We further verified that RNA interference did not significantly change cell cycle profiles (Fig. S3A) or cyclin B expression (Fig. 3A) in order to rule out these indirect cell cycle distribution effects. Furthermore, we did not detect any cell death in these experiments (Fig. S3B). This observation was confirmed by pharmacological inhibition of CHK1 in H1299 nocodazole-treated cells (Fig. S3C). We then asked whether preliminary phosphorylation of CHK1 on S280 is necessary for its phosphorylation of PLK1. For this, PIM2 and CHK1 were co-transfected into U2OS cells, and the effect on PLK1 phosphorylation was monitored by western blotting. First of all, we noticed that CHK1 overexpression has an effect on overexpressed PIM2 levels probably due to some transfection issues. As shown in Fig. 3B, co-expression of PIM2 and CHK1 induced an important increase in PLK1 phosphorylation on T210. In contrast, when a S280A (SA) mutant of CHK1 was transfected instead of wild-type CHK1, PLK1 phosphorylation was not modified. These data indicate that phosphorylation of CHK1 on S280 is necessary for its kinase activity on PLK1. In addition, we verified that overexpression of wild-type CHK1 or the S280A mutant did not affect cell viability (Fig. S3D). The involvement of PIM2 in PLK1 activation was confirmed by inducibly expressing PIM2 in U2OS and HeLa cells (Fig. S3E), which showed that PIM2 expression led to an increase in PLK1 T210 phosphorylation. We then asked whether phosphorylation of PLK1 affected its co-localization with CHK1 in mitotic cells. For this we used proximity ligation assays to detect the co-localization of CHK1 with T210-phosphorylated PLK1 (Fig. 3C), in U2OS cells treated or not with nocodazole. A close association between CHK1 and PLK1 phosphorylated on T210 (red dots) was significantly increased in nocodazole-treated cells. The association of PLK1 and CHK1 during mitosis was also analyzed by co-immunoprecipitation experiments with a CHK1 P-S280-specific antibody in U2OS cells that had been co-transfected with CHK1 and PIM2, and blocked in mitosis by nocodazole treatment. In these cells, PLK1 co-immunoprecipitated with phosphorylated CHK1, confirming a close association between these two proteins in mitosis (Fig. 3D).
To determine whether CHK1 can directly phosphorylate PLK1, we then performed in vitro kinase assays. As shown in Fig. 3E, recombinant CHK1 was able to directly phosphorylate PLK1 on T210, while PIM2 did not. Interestingly, the presence of PIM2 in these experiments did not change the ability of CHK1 to phosphorylate PLK1 on T210, suggesting that S280 phosphorylation of CHK1 does not modify the catalytic activity of CHK1, but rather alters its capacity to phosphorylate PLK1 through the regulation of its localization and/or interaction with other proteins. To elucidate clearly the role of P-CHK1 in our experiments and after MS/MS analyses (Fig. S4), we observed that the commercial recombinant CHK1 protein is already phosphorylated on residue S280, in consequence, we produced wild-type and S280 mutant forms of CHK1. As shown in Fig. S3F, recombinant proteins used for in vitro kinase assay with one of its well-defined substrates, CDC25C, presented good kinase activity (Fig. S3F). Then, we performed in vitro kinase assay with recombinant CHK1 WT and mutant proteins, PIM2 and PLK1, as shown in Fig. S3G. The results confirm the role of CHK1 in the phosphorylation of PLK1 and suggest that the S280A mutant does not significantly affect the catalytic activity of CHK1. Altogether, these data identify a phosphorylation cascade involving PIM2, CHK1 and PLK1 during mitosis.
PIM2 and CHK1 govern the phosphorylation and behavior of PLK1 substrates during mitosis
Finally, we asked whether interfering with PIM2 or CHK1 expression or activity modifies PLK1 function during mitosis in HeLa cells. For this, we followed the behavior of two well-described mitotic substrates of PLK1, EMI1 and WEE1. EMI1 is an anaphase-promoting complex/cyclosome (APC/C) regulator, and its phosphorylation by PLK1 induces its proteasomal degradation (Margottin-Goguet et al., 2003; Hansen et al., 2007; Moshe et al., 2004). This is also the case for WEE1 kinase, a negative regulator of CDK activity, whose phosphorylation by PLK1 induces its degradation (Watanabe et al., 2004). We found that when CHK1 was down-regulated by RNA interference in HeLa cells, levels of EMI1 and WEE1 increased (Fig. 4A). To reinforce these observations, we performed five independent experiments including two siRNA sequences for CHK1, which also resulted in a significant accumulation of EMI1 and WEE1 proteins (Fig. 4B). Similarly, RNA interference-mediated down-regulation of PIM1/2 also induced the accumulation of EMI1 in HeLa cells (Fig. 4C), and decreased the phosphorylation of nucleophosmin (NPM1), another previously described substrate of PLK1 (Zhang et al., 2004). Moreover, treatment of siRNA-transfected cells with the protein synthesis inhibitor cycloheximide showed that WEE1 and EMI1 degradation was delayed when CHK1 or PIM kinases were downregulated, indicating that CHK1 or PIM1/2 have a negative impact (decreased half-life) on the stability of these two proteins (Fig. 4D). Then, using an inducible PLK1 expression system that allows doxycycline-dependent expression of a constitutively active form of PLK1 (T210D) in HEK 293T cells (3×Flag T210D), we could significantly reverse the accumulation of WEE1 and EMI1 proteins under CHK1 or PIM1/2 downregulation (Fig. 4E). Finally, to substantiate our observations, we evaluated PLK1 activity more directly. For this, we immunoprecipitated PLK1 from HeLa cells treated with CHK1 (500 nM SCH900776) or PIM (5 µM SGI-1776) inhibitors for 16 h, and monitored PLK1 activity by estimating its capacity to phosphorylate two well-characterized substrates CDC25C (S75) (Gheghiani et al., 2017) and NPM1 (S4) (Zhang et al., 2004), in an in vitro kinase assay (Fig. 4F). These experiments showed that pharmacological inhibition of PIM1/2 or CHK1 significantly impairs PLK1 activity. Altogether, these results show that PIM1/2 and CHK1 kinases are involved in PLK1 regulation.
The PIM2-CHK1-PLK1 pathway is involved in cellular proliferation
Finally, we found that the functional interference of PIM1/2 activity either by siRNA or with a potent selective small-molecule inhibitor (SGI-1776) led to a significant inhibition of proliferation in HeLa, U2OS and H1299 cells (Fig. 5A,B). To further specifically investigate the role of the PIM2-CHK1-PLK1 axis in mitotic progression (G2/M transition), we used HEK 293T (3×Flag T210D) and the inducible PLK1 expression system previously described. These cells were synchronized in G2 by treating them with the CDK inhibitor RO3306 (10 µM) for 16 h, after which they were released in fresh medium supplemented or not with the SGI-1776 PIM inhibitor. After 7 h of culture in these conditions, the mitotic index was evaluated by flow cytometry to detect levels of phosphorylated histone H3. Inhibition of PIM1/2 kinase significantly reduced the percentage of mitotic cells (from 60.03±3.33% to 21.9±1.01%), but the concomitant expression of a constitutively active PLK1 mutant (T210D) in part compensated for PIM1/2 inhibition and increased the percentage of cells entering mitosis (to 34.9±1.05%) (Fig. 5C).
These results suggest that the PIM2-CHK1 pathway participates in PLK1 activation by controlling its phosphorylation. Taken together, our results indicate that PIM2 and CHK1 are upstream regulators of PLK1 function during mitosis and proliferation.
In the present work, we have identified the existence of a novel signaling cascade composed of the PIM2 and CHK1 kinases that leads to activation of the serine/threonine kinase PLK1 during mitosis. In addition to its well-described functions in the DNA damage response, several studies have described an independent role for CHK1 in unperturbed cell cycle progression, particularly during mitosis. For instance, CHK1 can phosphorylate and enhance the activity of Aurora B, promoting the functions of this mitotic kinase in chromosome segregation and cytokinesis (Petsalaki et al., 2011; Zachos et al., 2007). A role for CHK1 in the regulation of the BubR1-Mad2-Cdc20 complex and PLK1 have also been described (Zachos et al., 2007; Chilà et al., 2013; Tang et al., 2006), although these different studies have not provided a comprehensive picture of the general impact of CHK1 on mitotic progression. Here, we show that CHK1 directly phosphorylates PLK1 on its conserved activating T210 residue, and thereby positively regulates the diverse functions of PLK1 during mitosis. Since STK10, SLK and Aurora A have also been previously characterized as direct mediators of PLK1 phosphorylation at T210 (Walter et al., 2003; Macůrek et al., 2008; Seki et al., 2008), it remains to be clarified how these different kinases coordinate their activities to phosphorylate and activate PLK1.
Although CHK1 activation has been described as essentially dependent on ATR-mediated phosphorylation on S345 and S317, it remains unclear whether this mechanism is necessary for its DDR-independent functions. Indeed, several lines of evidence suggest that CHK1 can phosphorylate its substrates independently of ATR. For instance, during the G2/M transition, CHK1 is localized to the chromatin where it phosphorylates histone H3 on T11 to trigger the transcriptional expression of the mitotic effectors CDK1 and cyclin B1, allowing the cells to enter into mitosis (Shimada et al., 2008). In addition to its phosphorylation on S345 and S317 by ATR, CHK1 can also be phosphorylated on S280 by AKT, p90RSK or PIM1/2 – serine/threonine kinases that are involved in the activation of the major oncogenic signaling pathways PI3K-AKT, MEK-ERK and STAT5, respectively. The first publication in this field reported that S280 phosphorylation by AKT in PTEN-mutated cells inhibited CHK1 function by sequestering the protein to the cytoplasmic compartment (Puc et al., 2005). More recently, p90RSK was also found to phosphorylate CHK1 on S280 in response to growth factor stimulation of quiescent cells, but surprisingly, in this case it promoted its nuclear accumulation and activation rather than inhibition (Li et al., 2012). Finally, we recently documented that PIM kinases are involved in the phosphorylation of this residue in leukemic cells, in which these kinases are often overexpressed and play an important oncogenic role. In a previous study, we found that this phosphorylation had no detectable effect on CHK1 subcellular localization but did improve the DNA damage-dependent functions of CHK1 (Yuan et al., 2014a), as well as its capacity to stimulate leukemic cell proliferation (Yuan et al., 2014b). The reason for the apparent discrepancies between these different models is not yet clear, but one possibility is that the pathways responsible for CHK1 S280 phosphorylation and the resulting effects on its function differ between cell types and, perhaps, according to growth conditions. Here, we provide an additional and more general account of CHK1 regulation by its phosphorylation on S280, by demonstrating that this is an important parameter of CHK1 function during mitosis, and that PIM1/2 kinases are responsible for this mitosis-specific phosphorylation in different cellular models. Moreover, one important question that remains unresolved is the importance of CHK1 kinase catalytic activation (S296 phosphorylation) versus S280 phosphorylation in terms of regulating changes in its subcellular localization. In a previous work from Li et al. (2012), the authors did not observe any significant modulation of S296 phosphorylation following S280 phosphorylation by p90RSK in vitro. In light of this result, we speculate that this residue may act as a platform to recruit specific partners that may be involved in CHK1 activation and localization.
Although it has been previously involved in the regulation of cell cycle proteins, such as the CDK inhibitor p27Kip1 (Morishita et al., 2008), this is the first time to our knowledge that PIM2 has been described as a mitotic kinase. We found that its expression increased during mitosis, implying that its activity also increased since PIM2 is constitutively active and does not need post-translational (phosphorylation) activating modifications (Adam et al., 2015). In contrast to PIM2, PIM1 kinase has been previously involved in the regulation of mitosis. PIM1 interacts with and phosphorylates the nuclear mitotic apparatus protein NuMa (Bhattacharya et al., 2002) in a process that is important for maintaining a stable complex between NuMA, dynein-dynactin and HP1β, which links chromosomal kinetochores to spindle microtubules. In addition, PIM1 overexpression results in genomic instability by over-riding the mitotic spindle checkpoint (Roh et al., 2003). Finally, the co-localization of PIM1 with PLK1 has been described at the centrosome and midbody (Van der Meer et al., 2014), suggesting a functional interaction between these two proteins for mitotic progression. It is difficult, however, to correctly distinguish the specific cellular functions of PIM1 or PIM2 since these two kinases can compensate for each other in many circumstances. Understanding how PIM1 and PIM2 precisely regulate the different mitotic steps at the molecular level, and how they mutually complement or compensate for each other, still remains a challenging task. Our work implicates PIM2 as a regulator of CHK1 during mitosis, and it will be interesting to determine whether this function can also be performed by PIM1 in specific cellular contexts.
The functional interaction between PIM1/2, CHK1 and PLK1 during mitosis has important implications in the field of cancer biology and therapy. Indeed, it has been shown recently that the inhibition of PLK1, by either shRNA or the pharmacological inhibitor BI2536 in prostate cancer cells overexpressing PIM1, resulted in a dramatic inhibition of tumor progression. Moreover, compared with control cells, PIM1-overexpressing cancer cells are more prone to mitotic arrest followed by apoptosis due to PLK1 inhibition (Van Der Meer et al., 2014). These results suggest that over-activation of this newly identified signaling pathway due to the overexpression of one of its components could sensitize cancer cells and lead to its inhibition. This may be of particular interest in the case of acute myeloid leukemia (AML), since (i) PIM2 kinase is widely overexpressed in AML and is considered to be a potential therapeutic target for this pathology (Tamburini et al., 2009); (ii) we reported that PLK1 is often overexpressed in AML (Renner et al., 2009); and (iii) its pharmacological inhibition blocks proliferation and induces apoptosis in leukemic cell lines, and dramatically inhibits the clonogenic potential of primary cells from patients. Finally, we recently identified CHK1 expression as an independent prognostic marker in AML (David et al., 2016), and we described its regulation by PIM2 phosphorylation in these tumors (Yuan et al., 2014a; Yuan et al., 2014b). Altogether, these data suggest that targeting the PIM2-CHK1-PLK1 pathway in AML may constitute an interesting alternative or may complement chemotherapies for the treatment of these pathologies.
MATERIALS AND METHODS
Cell lines and treatments
Human non-small cell lung carcinoma H1299 cells were maintained in RPMI containing 10% fetal calf serum (FCS). Human bone osteosarcoma epithelial cells (U2OS), the human adenocarcinoma HeLa cell line, human hTERT-immortalized foreskin fibroblasts (hTERT BJ), and U2OS or HeLa cells expressing the HaloTag-PIM2 fusion protein, were cultured in DMEM supplemented with 10% FCS and 1% penicillin-streptomycin (PS). For synchronization, human leukemic cells (UT-7) were grown in DMEM containing 10% FCS and 10 ng ml−1 GM-CSF at 37°C with 5% CO2. The doxycycline-dependent expression system allowing the induction of a constitutively active form of PLK1 (T210D) in HEK 293T cells (3×Flag T210D) was kindly provided by Dr Laurent Créancier (Pierre Fabre Laboratories, Toulouse, France). To establish the HaloTag-PIM2 fusion protein-inducible system, a lentiviral expression vector pLenti PGK Blast DEST (Life Technologies) was modified by the insertion of the HaloTag cDNA (Promega) into the AgeI site. Then, the PIM2 isoform 2 cDNA sequence was placed in phase with HaloTag thanks to the Gateway system. The final construct was used to produce lentiviral particles to transduce HeLa and U2OS cells, and stable clones were selected by using 2 µg/ml blasticidin.
Cytokines, pharmacological inhibitors and reagents
Human granulocyte-macrophage colony-stimulating factor (GM-CSF) was from Miltenyi Biotec (Bergisch Gladbach, Germany). Doxycycline, nocodazole, the CDK1 inhibitor (RO3306) and thymidine were obtained from Sigma-Aldrich (France). The CHK1 inhibitor (SCH900776) was purchased from Active Biochem (Cliniscience, France). The PIM (SGI-1776) and MEK (PD0325901) inhibitors were from Selleckchem (Euromedex, France). The AKT inhibitor VIII was from Calbiochem (Merck Millipore, France). Other reagents were purchased from Sigma-Aldrich (France). The protein synthesis inhibitor cycloheximide was obtained from Sigma-Aldrich (France).
For siRNA transfection into HeLa or U2OS cells, Genome control pool non-targeting siRNA (Dharmacon) or CHK1 siRNA (Sigma-Aldrich) was transfected using 8 μl INTERFERin (Polyplus) in 6-well plates, as described by the manufacturer. For transfection of expression vectors, 2×106 U2OS cells were resuspended in 100 μl Nucleofector Kit V, according to the manufacturer's instructions (program x-001-Amaxa, Cologne, Germany). Experiments were performed 48 h after transfection.
The procedures used for gel electrophoresis and immunoblotting have been described previously (David et al., 2016). Briefly, cells were lyzed directly in NuPAGE LBS sample buffer, then sonicated and heated at 95°C for 5 min. Samples were subjected to electrophoresis in NuPAGE 4-12% Bis-Tris pre-cast gels (Life Technologies). CHK1 was detected with mouse monoclonal antibodies (clone G4, cat. no. sc-8408, Santa Cruz Biotechnology), PLK1 by mouse monoclonal antibodies (clones PL6/PL2, cat. no. 33-1700, Invitrogen), PIM2 by rabbit polyclonal antibodies (clone H-73, cat. no. sc-28778, Santa Cruz Biotechnology), P-PLK1 by rabbit monoclonal antibodies (clone D5H7, cat. no. 9062, Cell Signaling Technology), P-Bad (S112) by rabbit polyclonal antibodies (cat. no. 5284, Cell Signaling Technology), P-histone H3 (S10) by rabbit polyclonal antibodies (cat. no. 9706, Cell Signaling Technology), Cyclin B1 by mouse monoclonal antibodies (clone GNS1, cat. no. sc-245, Santa Cruz Biotechnology), P-Chk1 (S280) by rabbit polyclonal antibodies (cat. no. AP3069a, ABGENT), P-p44-p42 (Erk1/2, T202/Y204) by mouse monoclonal antibodies (clone E10; cat. no. 9106, Cell Signaling Technology), p44-p42 (Erk1/2) by rabbit polyclonal antibodies (cat. no. 9102, Cell Signaling Technology), P-RPS6 by rabbit monoclonal antibody (clone 2F9, cat. no. 4856, Cell Signaling Technology); RPS6 by rabbit monoclonal antibody (clone 5G10, cat. no. 2217, Cell Signaling Technology); P-AKT (S473) by rabbit monoclonal antibodies (clone D9E, cat. no. 4060, Cell Signaling Technology), AKT by rabbit polyclonal antibodies (cat. no. 9272, Cell Signaling Technology), Emi1 by rabbit polyclonal antibodies (cat. no. 38-500, Zymed), WEE1 by mouse monoclonal antibodies (clone B11, cat. no. sc-5285, Santa Cruz Biotechnology), P-NPM (S4) by rabbit monoclonal antibodies (cat. no. 3520, Cell Signaling Technology), P-CDC25C (S75) by rabbit polyclonal antibodies were kindly provided by Dr Olivier Gavet (Gustave Roussy Cancer Campus, Villejuif, France), P-CDC25C (S216) by rabbit polyclonal antibodies (cat. no. 9528, Cell Signaling Technology), CDC25C by rabbit polyclonal antibodies (cat. no. sc-327, Santa Cruz Biotechnology), NPM by mouse monoclonal antibodies (cat. no. AB10530, Abcam), P-NPM (S4) by rabbit polyclonal antibodies (cat. no. 3520, Cell Signaling Technology), and actin by mouse monoclonal antibodies (clone C4, cat. no. MAB1501, Santa Cruz Biotechnology). All antibodies are used at 1:1000 dilution. Secondary antibodies conjugated to HRP were used after incubation with primary antibodies. Immunoreactive bands were visualized by enhanced chemiluminescence (PI32209; Thermo Fisher Scientific) with a Syngene camera. Quantification of chemiluminescent signals was done with the GeneTools software (v.126.96.36.199) from Syngene.
Cells were fixed with 4% formaldehyde in 1× PBS at room temperature for 15 min, then permeabilized with 0.5% Triton X-100 in 1× PBS, as previously described (David et al., 2016). Briefly, cells were stained with the indicated antibodies and slides were mounted with ProLong Gold antifade reagent containing 4′,6-diamidino-2-phenylindole (P36931 from Invitrogen-Life Technologies). Images were acquired using a Zeiss confocal microscope (LSM780) and were subsequently processed using the ImageJ or ZEN software packages.
For proximity ligation assays, cells were washed twice with 1× PBS, fixed with 4% paraformaldehyde then permeabilized. Unspecific proteins were blocked with 3% FBS containing 0.1% Triton X-100 in 1× PBS for 30 min at room temperature. Cells were then incubated with either primary antibodies against CHK1 (DSC-310, cat. no. C-9358, Sigma, 1:400) and P-PLK1 (cat. no. 9062, Cell Signaling Technology, 1:200) or P-CHK1 (cat. no. AP3069A, ABGENT, 1:200) and PLK1 (cat. no. 37-7100, Zymed, 1:400). Then, cells were incubated with the appropriate DNA-linked secondary antibodies (Duolink kit, Sigma), and PCR in situ amplification was performed using the PLA technology, according to the manufacturer's instructions. The PLA signal was detected with a Zeiss confocal microscope. A series of Z-stack confocal microscopy images was taken and quantification of the images was carried out using ImageJ software. The HaloTag System was used according to the manufacturer's recommendations in combination with the previously described immunofluorescence protocol.
Cell synchronization, apoptosis and cell cycle analysis
Cells were synchronized by serum starvation coupled with either a thymidine block (2.5 mM), a double thymidine block (2.5 mM), or nocodazole treatment (200 ng/ml), then cell cycle distribution was analyzed. Briefly, cells were harvested and fixed in ice-cold 70% ethanol at −20°C. Cells were then permeabilized with 1× PBS containing 0.25% Triton X-100, resuspended in 1× PBS containing 10 µg/ml propidium iodide and 1 µg/ml RNase, and incubated for 30 min at 37°C. Staining for mitotic cells was conducted using phospho-histone H3 (S10) antibodies (cat. no. 06-570, Merck Millipore, 1:133). Apoptotic cells were detected with Annexin-V-FITC detection kit from BD PharMingen (San Diego, CA, USA) according to the manufacturer's instructions. Trypan Blue stain (GIBCO, LifeTechnologies, CA, USA) was also used to evaluate cell death. Data were collected on a MACSQuant VYB Analyzer (Miltenyi Biotec, Bergisch Gladbach, Germany) and analyzed using FlowJo v.10 software.
In vitro kinase assays
Human PIM2 (SignalChem), PLK1 (Enzo Life Sciences) and GST-CHK1 (Enzo Life Sciences) recombinant proteins were incubated in 30 μl kinase buffer (50 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 0.1 mM EDTA, 0.01% Brig35, 2 mM DTT) at 30°C for 60 min. The potent selective small-molecule inhibitor (SGI-1776) was used as control of the specificity of the PIM2 kinase activity. Analyses were performed by immunoblotting.
To generate expression vector permitting production of a fusion protein (GST), WT CHK1 sequence was cloned into the pGEX-4T2vector. We generated CHK1 S280 mutated to Ala by using site directed mutagenesis with forward primer 5′-cgagtcactgcaggtggtgtgtcagagtctccca-3′ and reverse primer 5′-tgacacaccacctgcagtgactcggggcctttttgc-3′. Recombinant GST-CHK1 and GST-CHK1 S280A proteins, as well as the maltose binding protein-CDC25C recombinant protein were produced in bacteria, as previously described by Brezak et al. (2005).
HeLa cells were treated with CHK1 (SCH900776-500 nM) or PIM (SGI-1776-5 µM) inhibitors for 24 h, then cells were lysed in buffer containing 25 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EGTA, 1% NP-40, 10 mM N-ethylmaleimide, protease inhibitor cocktail (Roche Applied Science, Basel, Switzerland) and phosphatase inhibitor cocktails 2 and 3 (Sigma-Aldrich). The lysates were used for immunoprecipitation of PLK1 protein (mouse anti-PLK1, Cocktail, Invitrogen). Immunoprecipitates were washed three times with lysis buffer and subjected to in vitro kinases assay with NPM (Abcam, UK) or CDC25C (plasmid provided by Odile Mondésert, ITAV, Toulouse, France). Phosphorylation levels were monitored by immunoblotting using phospho-specific antibodies.
Quantitative PCR (qPCR)
Total RNA was extracted from U2OS cells transfected with control siRNA, siRNA targeting PIM1 and PIM2 or siRNA targeting CHK1, using the RNeasy QIAGEN kit according to the manufacturer's instructions. RNA purity and concentration were monitored with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies Inc., Thermo Fisher Scientific). cDNAs were synthesized from 1 μg total RNA using SuperScript III Reverse transcription (Invitrogen). Real-time qPCRs were performed on a StepOne Real-Time PCR System (Thermo Fisher Scientific) with a TaqMan Gene Expression Master Mix (Applied Biosystems). The primer used was Hs01065498_m1 for PIM1, Hs00179139_m1 for PIM2 and Hs00967506_m1 (Applied Biosystems) for CHEK1. GUSB (Hs00939627_m1) and B2M (Hs00984230_m1) were used as housekeeping genes. Results were analyzed with the stepOnePlus software v.2.2.2 using the conventional ΔΔCt method.
NanoLC-MS/MS analysis and database searches
After SDS-PAGE separation of the in vitro kinase reaction, the silver-stained corresponding band was in-gel digested with trypsin and analyzed by online LC-MS analysis.
CHK1 LC-MS analysis
The peptides digested from CHK1 were measured on an SCIEX 5600+ TripleTOF mass spectrometer operated in DDA mode. A Dionex Ultimate 3000 nanoLC HPLC system and a Hypersil GOLD 150×0.32 mm column (Thermo Fisher Scientific), packed with C18 3 μm 175 Å material were used for peptide separation. For the HPLC method, the buffer A used was 0.1% (v/v) formic acid, and the buffer B was 0.1% (v/v) formic acid, 90% (v/v) acetonitrile. The gradient was 4-45% buffer B in 24 min with a flow rate of 5 µl/min. For MS, a survey scan at the MS1 level (350-1600 m/z) was first carried out with 250 ms per scan. Then, the Top20 most intense precursors, whose charge states are 2-4 were fragmented. Signals exceeding 75 counts per second were selected for fragmentation and MS2 spectra generation. MS2 spectra were collected in the mass range 100-1600 m/z for 80 ms per scan. The dynamic exclusion time was set to 10 s.
LC-MS data analysis
To identify CHK1 peptides, profile-mode.wiff files from data acquisition were centroided and converted to mzML format using the AB Sciex Data Converter v.1.3 and submitted to Mascot (v.2.5) database searches against UniProt SwissProt human database. ESI-Quad-TOF was chosen as the instrument, trypsin/P as the enzyme and 2 missed cleavages were allowed. Peptide tolerances at MS and MS/MS level were set to be 20 ppm and 0.5 Da, respectively. Peptide variable modifications allowed during the search were oxidation of M and phosphorylation of STY. To calculate the false discovery rate (FDR), the search was performed using the ‘decoy’ option in Mascot.
At least three independent experiments were carried out to generate each dataset and statistical analyses were performed with the Student's t-test using the Prism software package (GraphPad Software). Results are expressed as means±s.e.m. Differences were considered significant for the following P-values: *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001.
We thank Laetitia Ligat from the CRCT microscopy facility for helpful discussions on microscopy analyses. We gratefully acknowledge Dr Christine Dozier for critical review of the manuscript. The authors sincerely thank Romain Jugelé for his scientific contribution at the beginning of this project.
Conceptualization: L.Y., P.M., S.M., C.O.D.; Methodology: K.A., M.C., M.L., L.D., S.M., C.O.D.; Software: M.C., L.D., C.O.D.; Validation: K.A., M.C., M.L., L.D., C.O.D.; Formal analysis: K.A., M.C., L.D., S.M., C.O.D.; Investigation: M.C., M.L., L.D., P.M., S.M., C.O.D.; Resources: P.M., S.M.; Data curation: K.A., M.C., P.M., C.O.D.; Writing - original draft: K.A., P.M., S.M., C.O.D.; Writing - review & editing: M.C., A.B., P.M., S.M., C.O.D.; Visualization: S.M., C.O.D.; Supervision: P.M., S.M., C.O.D.; Funding acquisition: P.M., S.M.
This research was funded by the Ligue Contre le Cancer.
The authors declare no competing or financial interests.