Activation of the Mitogen- and Stress-activated Kinase 1 by Arsenic Trioxide*

Arsenic trioxide (As2O3) is a potent inducer of apoptosis of leukemic cells in vitro and in vivo, but the precise mechanisms by which it mediates such effects are not well defined. We provide evidence that As2O3 induces activation of the mitogen- and stress-activated kinase 1 (MSK1) and downstream phosphorylation of its substrate, histone H3, in leukemia cell lines. Such activation requires upstream engagement of p38 MAPK, as demonstrated by experiments using pharmacological inhibitors of p38 or p38α knock-out cells. Arsenic-induced apoptosis was enhanced in cells in which MSK1 expression was decreased using small interfering RNA and in Msk1 knock-out mouse embryonic fibroblasts, suggesting that this kinase is activated in a negative feedback regulatory manner to regulate As2O3 responses. Consistent with this, pharmacological inhibition of MSK1 enhanced the suppressive effects of As2O3 on the growth of primary leukemic progenitors from chronic myelogenous leukemia patients. Altogether, these findings indicate an important role for MSK1 downstream of p38 in the regulation of As2O3 responses.

limiting factor for the use of As 2 O 3 in the treatment of various hematological malignancies is the requirement of high cellular concentrations for the induction of antitumor effects in different malignant phenotypes. It is well established that the effects of As 2 O 3 are dose-dependent, with higher concentrations (Ն2 M) leading to apoptosis and lower concentrations (Յ0.5 M) inducing differentiation (1)(2)(3)(4)(5)(6). Thus, the potential development of future translational approaches would be facilitated by the identification of the means to enhance arsenic-dependent apoptosis at lower final concentrations of As 2 O 3 .
We have previously shown that p38 MAPK and its downstream effector, MAPKAPK2 (MAPK-activated protein kinase-2), are activated during treatment of cells with As 2 O 3 (23). Such activation of p38 MAPK appears to occur in a negative feedback regulatory manner, as pharmacological inhibitors of p38 were found to enhance the generation of pro-apoptotic responses by As 2 O 3 in target cells. In the present study, we sought to identify the downstream effectors of p38 that may account for the negative regulatory properties of the p38 MAPK pathway in the generation of As 2 O 3 responses. We found that the nucleosomal kinase MSK1 is also activated in an As 2 O 3 -inducible manner. Pharmacological or small interfering RNA (siRNA)-mediated knockdown of MSK1 resulted in enhanced induction of apoptosis in leukemia cell lines and primary leukemic progenitors from the bone marrow of CML patients. Moreover As 2 O 3 -dependent apoptosis was enhanced in cells with targeted disruption of Msk1 and the related Msk2 gene. Altogether, these data identify MSK1 as a negative regulator of As 2 O 3 responses.

MATERIALS AND METHODS
Cells and Reagents-The KT-1 CML and the NB-4 human acute promyelocytic leukemia cell lines were grown in RPMI 1640 medium supplemented with fetal bovine serum and antibiotics. Immortalized mouse embryonic fibroblasts (MEFs) were cultured in Dulbecco's modified Eagle's medium supplemented with fetal bovine serum and antibiotics. Immortalized MEFs from p38␣ knock-out mice (11)  Cell Lysis, Immunoprecipitation, and Immunoblotting-Cells were incubated with As 2 O 3 for the indicated times and lysed in phosphorylation lysis buffer as described previously (17,18). In experiments in which pharmacological inhibitors were used, SB 203580 (10 M) or H-89 (10 M) was added 1 h prior to treatment with As 2 O 3 . Immunoprecipitations and immunoblotting using enhanced chemiluminescence method were performed as described previously (24,25).
Kinase Assays-Cells were incubated with As 2 O 3 for the indicted times. Total cell lysates were immunoprecipitated with antibody against MSK1 or non-immune rabbit IgG, which was used as a control. In vitro kinase assays were performed as described previously (26 -29). For the MSK1 kinase assay, the values were calculated by subtracting the activity in the rabbit IgG immunoprecipitates from the kinase activity in the anti-MSK1 immunoprecipitates.
Evaluation of Apoptosis-Evaluation of apoptosis by annexin V/propidium iodide staining was performed using an apoptosis detection kit (Pharmingen) as described previously (23,26). Briefly, cells were plated in 100-mm plates and treated  for 48 h with the indicated concentrations of As 2 O 3 . The cells were harvested, washed with cold phosphate-buffered saline, and then incubated for 15 min with fluorescein isothiocyanate-conjugated annexin V and propidium iodide prior to flow cytometric analysis. In the experiments in which the effects of siRNA-mediated targeting of MSK1 were evaluated, KT-1 cells were electroporated with an MSK1 siRNA duplex or control mixture siRNA (New England Biolabs, Ipswich, MA) as recommended by the manufacturer. Twenty-four hours later, the cells were incubated with the indicated concentrations of As 2 O 3 for 48 h and analyzed by flow cytometry.
Hematopoietic Cell Progenitor Assays-Bone marrow and peripheral blood from CML patients were collected. The effects of As 2 O 3 on the growth of hematopoietic progenitors from CML patients were determined in clonogenic assays in methylcellulose as described previously (26,30). Briefly, mononuclear cells were separated by Ficoll-Hypaque sedimentation, and cells were cultured in a methylcellulose mixture containing hematopoietic growth factors (26,30) in the presence or absence of As 2 O 3 (1 M) and H-89 (10 M). Colony forming units-granulocyte/macrophage (CFU-GM) from the leukemic bone marrow were scored on day 14 of culture.
Acid-soluble Protein Extraction-KT-1 cells were treated with H-89 for 1 h prior to treatment with As 2 O 3 (2 M) for 20 min. Extraction of proteins was performed as described previously (31,32). Briefly, acid-soluble proteins were extracted with lysis buffer (10 mM HEPES (pH 7.9), 1.5 mM MgCl 2 , 10 mM KCl, and 0.5 mM dithiothreitol), and 0.2 M sulfuric acid was added, followed by incubation on ice for 30 min. The acid-soluble proteins contained in the supernatant were retained after centrifugation at 11,000 ϫ g for 10 min at 4°C and precipitated on ice for 30 min with a final concentration of 25% trichloroacetic acid. The proteins were pelleted at 12,000 ϫ g for 10 min and washed once with 100% acetone and 0.05 M HCl and once with 100% acetone. The dried pellets were resuspended in acetic acid/urea buffer and resolved by SDS-PAGE.

RESULTS
We initially determined whether the kinase MSK1 is activated during treatment of malignant hematopoietic cell lines with As 2 O 3 . The derived KT-1 CML cell line and the NB-4 acute promyelocytic leukemia cell line were studied. Cells were incubated in the presence or absence of As 2 O 3 , and after cell lysis, equal amounts of lysates were analyzed by SDS-PAGE and immunoblotted with antibody against MSK1 phosphorylated at Ser 376 . As 2 O 3 induced strong phosphorylation of MSK1 in both KT-1 cells (Fig. 1, A and B) and NB-4 cells (Fig. 1, C and D). We subsequently directly examined whether As 2 O 3 treatment of cells results in activation of the kinase domain of MSK1. KT-1 cells were incubated in the presence or absence of As 2 O 3 ; cell lysates were immunoprecipitated with anti-MSK1 antibody; and in vitro kinase assays were performed on the immunoprecipitates. As shown in Fig. 1E, treatment of KT-1 cells with As 2 O 3 resulted in activation of MSK1, whereas such activation was blocked by pretreatment of the cells with the p38 pharmacological inhibitor SB 203580. On the other hand, in similar experiments in which KT-1 cells were pretreated with the MEK/ERK inhibitor PD 98059, we found no significant inhibi-tion of such activation (data not shown), suggesting that p38 is the major kinase that regulates activation of MSK1 by As 2 O 3 in these cells. Previous studies established that activation of MSK1 is regulated by both the p38 and ERK MAPK signaling pathways in different systems (33)(34)(35)(36). Inhibition of MSK1 activity by the p38 inhibitor strongly suggested that p38 plays an important regulatory role in activation of MSK1 by As 2 O 3 . To further explore the requirement of the p38 kinase for activation of MSK1 by As 2 O 3 , MEFs with targeted disruption of the p38␣ gene (68) were used. As expected, As 2 O 3 treatment induced phosphorylation of MSK1 in wild-type MEFs, whereas such phosphorylation was defective in p38␣ Ϫ/Ϫ cells (Fig. 2, A and  B). Consistent with this, the induction of MSK1 kinase activity was defective in cells lacking p38␣ compared with parental cells (Fig. 2C). Altogether, these findings established that phosphorylation/activation of MSK1 by As 2 O 3 is p38-dependent.
In subsequent experiments, we sought to determine the functional role of MSK1 in the generation of As 2 O 3 -mediated apoptosis. In a previous study, we showed that pharmacological inhibition of p38 results in enhanced induction of apoptosis by As 2 O 3 (23), but the downstream effectors that mediate such anti-apoptotic effects have not been identified to date. When the induction of apoptosis by As 2 O 3 was determined in immortalized MEFs with targeted disruption of the Msk1 gene, we found enhanced As 2 O 3 -induced apoptosis in cells lacking Msk1 or in double knockout MEFs for Msk1 and the related Msk2 gene compared with parental cells (Fig. 3A). Consistent with this, poly(ADP-ribose) polymerase cleavage was enhanced in Msk1 Ϫ/Ϫ and Msk1/2 Ϫ/Ϫ MEFs compared with wild-type MEFs (Fig. 3, B and C).
Thus, it appears that MSK1 is a key mediator of the anti-apoptotic effects of p38 MAPK during its activation in response to As 2 O 3 treatment of cells.
A major substrate for the MSK1 kinase is histone H3 (36), the phosphorylation of which is associated with immediate-early gene induction (37). We examined whether As 2 O 3 induces phosphorylation of histone H3 in KT-1 cells. As shown in Fig. 4 (A and B), treatment of KT-1 cells with As 2 O 3 induced phosphorylation of histone H3 at Ser 10 , and such phosphorylation was inhibited by concomitant treatment of the cells with either SB 203580 or H-89, a pharmacological inhibitor to which MSK1 is known to exhibit high sensitivity (37). Pretreatment of KT-1 cells with H-89 also inhibited the arsenic-dependent induction of MSK1 kinase activity (Fig. 4C) and further enhanced the induction of apoptosis by As 2 O 3 in KT-1 cells (Fig. 4D).
To further establish the relevance of MSK1 in the negative regulation of As 2 O 3 -induced apoptosis, we used siRNA interference to block MSK1 expression in cells of hematopoietic origin. KT-1 cells were transfected with an MSK1specific siRNA, and after 72 h, cell extracts were prepared and immunoblotted with anti-MSK1 antibody. As in our previous study (38), transfection of cells with the MSK1specific siRNA resulted in knockdown of MSK1 (Fig. 5, A and  B). The induction of cell death by As 2 O 3 was subsequently assessed after 48 h of treatment. As illustrated in Fig. 5C, knockdown of MSK1 in the presence of As 2 O 3 further potentiated the induction of apoptosis compared with As 2 O 3 alone.
Altogether, these data established that pharmacological or molecular inhibition of MSK1 expression potentiates the generation of the inhibitory effects of As 2 O 3 on leukemic cells. To further explore the role of MSK1 in a more physiologically relevant system, we evaluated the effects of pharmacological inhibition of MSK1 on the induction of the suppressive effects of As 2 O 3 on primary leukemic progenitors from CML patients. Bone marrow and peripheral blood mononuclear cells from six patients with CML were isolated, and leukemic CFU-GM progenitor colony formation was  determined by clonogenic assays in methylcellulose. As expected, addition of As 2 O 3 to the cultures suppressed leukemic CFU-GM progenitor growth (Fig. 6, A-F). On the other hand, addition of H-89 to the cultures alone had no significant effects. However, concomitant addition of H-89 to the cultures strongly enhanced (twotailed p value ϭ 0.001974) the suppressive effects of As 2 O 3 on leukemic CFU-GM progenitor growth.

DISCUSSION
Over the last few years, extensive efforts have been undertaken to understand the mechanisms by which As 2 O 3 induces apoptosis of target cells and exhibits antitumor properties. Treatment of malignant cells with As 2 O 3 is known to result in elevation of reactive oxygen species, loss of mitochondrial membrane potential, and release of cytochrome c (39 -42). These events are followed by activation of the caspase cascade and programmed cell death (39 -42). There is also evidence that As 2 O 3 can influence cell death through activation of JNK (43,44) and inhibition of the transcription factor NF-B (45)(46)(47). It should also be noted that the generation of reactive oxygen species depends on cellular glutathione stores (48) and that a decrease in reduced cellular GSH levels by pretreatment of arsenic-sensitive cells with ascorbic acid (48) or buthionine sulfoximine (49) enhances their sensitivity to arsenic-dependent apoptosis. On the other hand, increased cellular GSH levels result in attenuation of the cytotoxic effects of As 2 O 3 (49).
Despite the advances in our understanding of the mechanisms by which As 2 O 3 induces apoptosis, little is known about the role of pathways that negatively regulate the generation of the anti-leukemic properties of As 2 O 3 . There is evidence that the phosphatidylinositol 3-kinase/ Akt pathway mediates As 2 O 3 resistance in human leukemic cells (50,51) and that pharmacological inhibitors of phosphatidylinositol 3-kinase potentiate As 2 O 3 -induced apoptosis via glutathione depletion and increased peroxide accumulation in myeloid leukemia cells (51). Thus, one pathway that appears to negatively regulate the generation of the anti-leuke- mic properties of As 2 O 3 is the phosphatidylinositol 3-kinase pathway.
We have also previously shown that activation of the p38 MAPK pathway in response to As 2 O 3 treatment of cells exhibits negative regulatory effects on the induction of arsenic-dependent apoptosis (23). This was evidenced by experiments demonstrating that pharmacological inhibitors of p38 or overexpression of a p38 dominant-negative mutant promotes the effects of As 2 O 3 . These findings have suggested that the p38 pathway is activated in a negative feedback regulatory manner to regulate arsenic-induced apoptosis, but the downstream p38 effectors that mediate such effects are not known. Interestingly, p38 MAPK appears to be also activated in a negative feedback regulatory manner in response to all-trans-retinoic acid in acute promyelocytic leukemia cells and to negatively regulate leukemic cell differentiation (52).
In this study, we have provided the first evidence that the kinase MSK1 is activated by As 2 O 3 in leukemia cell lines. Such activation requires upstream engagement of the p38 MAPK pathway, as evidenced in studies using pharmacological inhibitors of p38 or p38␣ knock-out cells. We have also demonstrated that pharmacological inhibition of MSK1 or siRNA-mediated disruption of its expression results in enhanced arsenic-induced apoptosis, suggesting that this p38 effector is a primary mediator of the anti-apoptotic effects of the p38 MAPK pathway on leukemic cells. Similar effects were also seen in cells with targeted disruption of the Msk1 gene or double knock-outs for both Msk1 and the related Msk2 gene. Notably, pharmacological inhibition of MSK1 activity was found to enhance the suppressive effects of As 2 O 3 on primary leukemic CFU-GM progenitors from CML patients, indicating that such effects occur in physiologically relevant systems.
MSK1 and MSK2 are serine kinases that are activated downstream of the p38 and MEK/ERK signaling cascades (36,37,(53)(54)(55). Studies using Msk1 and Msk2 knock-out MEFs have demonstrated that these kinases regulate phosphorylation of histone H3 at Ser 10 and Ser 28 as well as phosphorylation of HMGNI (HMG-14) at Ser 6 , Ser 20 , and Ser 24 (36,53), but they are not required for acetylation of histone H3 (36). Such functions of MSK1 and MSK2 are critical for stress-induced chromatin remodeling and immediate-early gene transcription of a variety of genes (36,37). Notably, phosphorylation of histone H3 at Ser 10 is elevated in oncogene-transformed fibroblasts, suggesting that MSKs are involved in aberrant gene expression observed in oncogene-transformed cells (56). In addition to histone H3 and HMGNI, there are additional targets for the activity of MSKs (53). In response to activation by certain stress signals, MSK1 and MSK2 mediate phosphorylation of cAMPresponsive element-binding protein and ATF1 (57)(58)(59), which results in the regulation of c-fos and junB transcription. MSK1 also phosphorylates NF-B and the ER81 transcription factor, which is involved in oncogenesis (53,60,61), whereas its function is critical for interleukin-1-induced c-fos gene expression in keratinocytes and promotes the growth of both keratinocyte and human epidermoid carcinoma cell lines (62). Recent evidence also suggests important roles for MSK1/2 in epidermal growth factor (63) and vascular endothelial growth factor (64) signaling. Moreover, MSK1 and MSK2 regulate the transcrip-tion of the Nur77, Nurr1, and Nor1 nuclear orphan receptor genes of the NR4A subfamily (65), the up-regulation of which has been implicated in cellular transformation (66). Thus, it appears that MSK1 and MSK2 regulate engagement of multiple downstream signals to mediate anti-apoptotic and mitogenic responses. Our study has established that As 2 O 3 phosphorylates histone H3 at Ser 10 in an MSK1-dependent manner, suggesting that one mechanism by which this kinase ameliorates As 2 O 3 -induced apoptosis may involve the regulation of early expression of anti-apoptotic genes. Altogether, our data strongly suggest that the kinase MSK1 (alone or in combination with As 2 O 3 ) is a highly attractive target for the design of novel therapeutic approaches for the treatment of leukemias.