Activation of Rac1 and the p38 mitogen-activated protein kinase pathway in response to arsenic trioxide.

Arsenic trioxide induces differentiation and apoptosis of malignant cells in vitro and in vivo, but the mechanisms by which such effects occur have not been elucidated. In the present study we provide evidence that arsenic trioxide induces activation of the small G-protein Rac1 and the alpha and beta isoforms of the p38 mitogen-activated protein (MAP) kinase in several leukemia cell lines. Such activation of Rac1 and p38-isoforms results in downstream engagement of the MAP kinase-activated protein kinase-2 and is enhanced by pre-treatment of cells with ascorbic acid. Interestingly, pharmacological inhibition of p38 potentiates arsenic-dependent apoptosis and suppression of growth of leukemia cell lines, suggesting that this signaling cascade negatively regulates induction of antileukemic responses by arsenic trioxide. Consistent with this, overexpression of a dominant-negative p38 mutant (p38betaAGF) enhances the antiproliferative effects of arsenic trioxide on target cells. To further define the relevance of activation of the Rac1/p38 MAP kinase pathway in the induction of arsenic-dependent antileukemic effects, studies were performed using bone marrows from patients with chronic myelogenous leukemia. Arsenic trioxide suppressed the growth of leukemic myeloid (CFU-GM) progenitors from such patients, whereas concomitant pharmacological inhibition of the p38 pathway enhanced its growth-suppressive effects. Altogether, these data provide evidence for a novel function of the p38 MAP kinase pathway, acting as a negative regulator of arsenic trioxide-induced apoptosis and inhibition of malignant cell growth.


Arsenic trioxide (As 2 O 3 ) suppresses the growth of malignant cells in vitro and in vivo
. Several studies have shown that this agent exhibits potent growth inhibitory effects on several cell lines of diverse malignant phenotypes, including leukemia, multiple myeloma, prostate carcinoma, and neuroblastoma cells (5)(6)(7)(8)(9)(10). As 2 O 3 exhibits its antineoplastic effects by inducing apoptosis and cell cycle arrest, whereas it also enhances differentiation of leukemia cells when used at lower doses (1)(2)(3)(4). This agent is highly effective in the treatment of patients suffering from acute promyelocytic leukemia refractory to alltrans-retinoic acid, and this has made it an extremely important component in the clinical management of this leukemia (1)(2)(3)(4). Despite the well documented clinical efficacy of arsenic in leukemia therapy, the precise mechanisms regulating arsenicdependent induction of apoptosis of neoplastic cells have not been elucidated.
Mitogen-activated protein (MAP) 1 kinases are a family of widely expressed serine-threonine kinases that regulate important cellular processes. Four MAPK family subgroups exist: extracellular signal-regulated kinases, c-Jun N-terminal or stress-activated protein kinases, extracellular signal-regulated kinase 5/big mitogen activated protein kinase 1 (BMK1), and the p38 group of protein kinases (reviewed in Ref. 11).
Recent studies from our laboratory (37) have shown that all-trans-retinoic acid (RA) induces activation of the p38 MAP kinase pathway and that pharmacological inhibition of p38 potentiates the antiproliferative and differentiating effects of RA on NB-4 cells (37). As arsenic trioxide suppresses the growth of acute promyelocytic leukemia cells in vitro, including RA-resistant cells, we performed studies to evaluate the role of the p38 MAP kinase in the induction of arsenic trioxide-dependent cellular responses. Our data demonstrate that p38 is phosphorylated and activated by arsenic trioxide in an early and sustained manner and that this activation is downstream of redox events activated by arsenic trioxide. Furthermore, they establish that the small G-protein Rac1 and the MAP-KapK-2 kinase are arsenic-dependent upstream and downstream effectors, respectively, of the p38 MAP kinase. Interestingly, our findings indicate that inhibition of p38 by pharmacological inhibitors, or by overexpression of a dominant-negative mutant, enhances arsenic-induced apoptosis. These results strongly suggest that such activation of the p38 MAP kinase signaling cascade provides a negative feedback regulatory mechanism that counteracts the antileukemic effects of arsenic trioxide. In further support of this hypothesis, in experiments in which the effects of pharmacological inhibitors of p38 were directly examined in primary leukemic progenitors from chronic myelogenous leukemia (CML) patients, we found that p38 inhibition strongly enhances arsenic-dependent suppression of leukemic precursor cell growth.

EXPERIMENTAL PROCEDURES
Cells and Reagents-The NB-4 human acute promyelocytic leukemia, the NB-4.007/6 RA-resistant variant (38), and the K562 CML-blast crisis cell lines were grown in RPMI 1640 supplemented with fetal bovine serum and antibiotics. The MCF-7 human breast carcinoma cell line and the LNKAP prostate carcinoma cell lines were grown in Dulbecco's modified Eagle's medium supplemented with fetal bovine serum and antibiotics. Arsenic trioxide and ascorbic acid were purchased from Sigma. An antibody against the phosphorylated form of p38 was obtained from New England Biolabs. A monoclonal antibody against Rac1 was obtained from Transduction Laboratories (Lexington, KY). Polyclonal antibodies against p38␣ and p38␤ were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). A polyclonal antibody against MAPKap kinase-2 was obtained from Upstate Biotechnology. The p38 MAP kinase inhibitor SB203580 was purchased from Calbiochem.
Cell Lysis and Immunoblotting-Cells were stimulated with the indicated doses of arsenic trioxide for the indicated times and were lysed as described previously (39,40). Immunoprecipitations and immunoblotting using an enhanced chemiluminescence method were performed as described previously (39,40).
Rac1 Activation Assays-The activation of Rac1 was determined by a methodology described previously (33,35,41). Briefly, the pGEX-4T3 construct encoding for the GTPase binding domain of human Pak1 (33,35,41) (provided by Dr. Gary Bokoch, Scripps Research Institute, La Jolla, CA) was expressed in Escherichia coli as a GST fusion protein (GST-PBD). The cells were treated with arsenic trioxide as indicated and lysed in phosphorylation lysis buffer. Cell lysates were incubated with GST-PBD, and bound proteins were separated by SDS-PAGE and immunoblotted with a monoclonal antibody against Rac1 to detect GTP-bound Rac1.
In Vitro Kinase Assays-These assays were performed as described previously (32,36). Briefly, cells were treated with arsenic trioxide for the indicated times and lysed in phosphorylation lysis buffer. Cell lysates were then immunoprecipitated with antibodies against p38␤ or MAPKap kinase-2, and the immunoprecipitated proteins were washed three times in phosphorylation lysis buffer and two times in kinase buffer (25 mM Hepes, pH 7.4, 25 mM MgCl 2 , 25 mM ␤-glycerophosphate, 100 M sodium orthovanadate, 2 mM dithiothreitol, 20 M ATP), and resuspended in 30 l of kinase buffer containing 5 g GST-ATF2 (for the p38␤ kinase assays) or 5 g of hsp25 protein (for the MAPKapK-2 kinase assays) and 25 Ci of [␥-32 P]ATP. The reaction was incubated for 30 min at room temperature and was terminated by the addition of SDS sample buffer. Proteins were subsequently analyzed by SDS-PAGE, and the phosphorylated forms of ATF2 or hsp25 were detected by autoradiography.
Cell Proliferation Assays-NB-4 cells were pre-treated for 30 -60 min with the indicated doses of SB203580 and were subsequently treated with arsenic trioxide for the indicated times, in the continuous presence of the pharmacological inhibitors. Cell proliferation assays using the MTT method were performed as in previous studies (35,42).
Hematopoietic Progenitor Cell Assays-The effects of arsenic trioxide on the growth of hematopoietic progenitors from patients with CML was determined in clonogenic assays in methylcellulose, as in previous studies (35,36,43). Bone marrow aspirate specimens were obtained under local anesthesia from patients with chronic myelogenous leukemia, after obtaining informed consent approved by the Institutional Review Board of the University of Illinois at Chicago. Bone marrow mononuclear cells were separated by Ficoll Hypaque sedimentation, and cells were cultured in a methylcellulose mixture containing hematopoietic growth factors (35,36,43), in the presence or absence of arsenic trioxide (2 M) and SB203580 (5 or 10 M). Colony forming units-granulocyte/macrophage (CFU-GM) from the leukemic bone marrows were scored on day 14 of culture.
Evaluation of Apoptosis-Cell lines were exposed to arsenic trioxide in the presence or absence of SB203580 (10 M) as indicated. The percentage of apoptotic cells were determined at various time points by flow cytometry after staining with fluorescein-conjugated annexin-V and propidium iodide, as in previous studies (36,44).
Generation the p38␤AGF Mutant and Stable Transfections-To create the p38␤2AGF mutant, a commercial kit (QuikChange TM site-directed mutagenesis kit; Stratagene Inc., La Jolla, CA) was used according to the manufacturer's instructions. Briefly, a pair of complementary primers with 30 bases (N terminus, 5Ј-ACGAGGAGATGGCCGGCTT-TGTGGCCACGC-3Ј; C terminus, 5Ј-GCGTGGCCACAAAGCCGGCCA-TCTCCTCGT-3Ј) was designed, which included the mutation to change threonine (ACC) to alanine (GCC) and tyrosine (TAT) to phenylalanine (TTT). pCDN-p38␤2 (provided by Dr. Sanjay Kumar, Smith, Kline Beecham) was amplified using Pfu DNA polymerase with these primers for 12 cycles in a DNA thermal cycler (PerkinElmer Life Sciences). After digestion of the parental DNA with DpnI, the amplified DNA incorporated with the nucleotide substitution was transformed into E. coli (DH5␣-strain). The mutations were confirmed by DNA sequencing. MCF-7 cells were transfected using the superfect reagent and were subsequently selected in G418 using standard methodologies.

RESULTS
We first determined whether As 2 O 3 treatment induces phosphorylation/activation of the p38 (also called p38␣) MAP kinase in the NB-4 acute promyelocytic leukemia cell line. NB-4 cells were incubated for 30 or 60 min in the presence or absence of As 2 O 3 . The cells were subsequently lysed in phosphorylation lysis buffer, and total lysates were resolved by SDS-PAGE and immunoblotted with an antibody against the phosphorylated/ activated form of p38. As 2 O 3 treatment induced strong phosphorylation of p38 after either 30 or 60 min of treatment (Fig.  1A). Such phosphorylation of p38 was detectable when low (0.5 M) or high (2 M) doses of As 2 O 3 were used (Fig. 1A). Stripping and reprobing the blots demonstrated that equal amounts of p38 were detectable prior to and after As 2 O 3 treatment (Fig.  1B). We subsequently determined whether arsenic-dependent phosphorylation/activation of p38 (p38␣) occurs in an NB-4 variant cell line, NB-4.007/6 (38). This variant form of NB-4 is refractory to all-trans-RA-induced differentiation and inhibition of cell growth, because of constitutive degradation of PML-RAR (38), and in our previous studies we have shown that the activation of p38 by all-trans-retinoic acid is defective in these cells (37). Treatment of NB-4.007/6 cells with As 2 O 3 resulted in phosphorylation of p38 ( Fig. 1, C and D), consistent with the fact that these cells are responsive to the growth inhibitory effects of As 2 O 3 (data not shown). Thus, As 2 O 3 induces phosphorylation of p38 in cells that are resistant to the growth inhibitory effects of RA, indicating that the early upstream regulatory signals that mediate As 2 O 3 -dependent activation of p38 are different from the ones mediating RA-dependent activation. The activation of p38 by As 2 O 3 in NB-4 cells was sustained and prolonged and could be detected for as long as 5 days of As 2 O 3 treatment of the cells, a time point at which As 2 O 3 induces cell-differentiation (Fig. 1E).
In parallel studies, we examined whether the ␤-isoform of p38 (p38␤) is expressed in acute promyelocytic leukemia cells and whether As 2 O 3 induces its activation. NB-4 cells were incubated in the presence or absence of As 2 O 3 , and after cell lysis, cell lysates were immunoprecipitated with a specific anti-p38␤ antibody. Immune-complex kinase assays were subsequently carried out on the immunoprecipitates, using ATF2 as an exogenous substrate. As shown in Fig. 2B, treatment of the cells with As 2 O 3 resulted in induction of the kinase activity of p38␤, evidenced by the phosphorylation of ATF2 used as an exogenous substrate ( Fig. 2A). Thus, in addition to the p38␣, the p38␤ isoform is also activated by As 2 O 3 in the NB-4 acute promyelocytic leukemia cell line, suggesting that it also participates in the induction of As 2 O 3 responses in these cells.
Previous studies have established that a downstream effector for p38 is the MAPKapK-2 kinase, which is activated in response to stress signals and growth factors (31,45), as well as in response to interferon-dependent activation of the p38 pathway (33,35). We examined whether As 2 O 3 -induced activation of p38 leads to downstream engagement and activation of MAPKapK-2 in NB-4 cells. Cells were incubated in the presence or absence of As 2 O 3 , lysates were immunoprecipitated with an anti-MAPKapK-2 antibody, and immunoprecipitated proteins were subjected to in vitro kinase assays using hsp25 as an exogenous substrate. As 2 O 3 treatment of the cells resulted in activation of MAPKapK-2, evidenced by the phosphorylation of hsp25 in the kinase assay (Fig. 2B). In addition, pre-treatment of cells with the p38-specific inhibitor SB203580 completely abrogated the As 2 O 3 -inducible activation of MAP-KapK-2 (Fig. 2B), demonstrating that the engagement of MAPKapK-2 in an As 2 O 3 -dependent cellular pathway occurs downstream of p38.
In addition to its effects against acute promyelocytic leukemia cells, As 2 O 3 has been shown to exhibit pro-apoptotic and growth inhibitory effects in a variety of different tumor cell lines (5-10). To examine whether activation of the p38-pathway by arsenic is limited to cells of promyelocytic origin or also occurs in cells of other malignant phenotypes, the phosphorylation/activation of p38 was examined in several cell lines of diverse origin. The cell lines used included the K562 acute erythroleukemia cell line, which has been derived from a patient with CML in blast crisis and expresses BCR-ABL, the LNKAP prostate carcinoma cell line, and the MCF-7 breast carcinoma cell line, all of which exhibited sensitivity to the growth inhibitory effects of As 2 O 3 in MTT cell proliferation assays (data not shown). As shown in Fig. 3, As 2 O 3 induced strong phosphorylation/activation of p38 in K562 cells (Fig. 3, A and B), LNKAP cells (Fig. 3, C and D), and MCF-7 cells (Fig. 3, E and F). Thus, the p38-pathway is activated in an As 2 O 3 -dependent manner in a variety of malignant cell phenotypes, indicating that it plays a universal role in the generation of As 2 O 3 responses.
It is well established that the activated, GTP-bound form of the small G-protein Rac1 acts as an upstream regulator of the p38 MAP kinase pathway in response to various stimuli (33, 35, 37, 46 -48). We examined whether As 2 O 3 induces activation of Rac1 to regulate downstream engagement of p38 in NB-4 cells. As 2 O 3 treatment resulted in activation of Rac1, as shown by the increase in the GTP-bound form of Rac1 (Fig. 4A), providing the first evidence for engagement of this small Gprotein in an As 2 O 3 -activated cellular pathway, and strongly suggesting that this GTPase is an upstream regulator of the As 2 O 3 -dependent activation of p38. To better understand the mechanisms of activation of the Rac1/p38 pathway by As 2 O 3 , we examined whether the activation of Rac1 is downstream or upstream of the redox reactions and free radicals generated by arsenic trioxide treatment. Previous studies have shown that ascorbic acid enhances As 2 O 3 -mediated responses, by increasing cellular H 2 O 2 stores (49,50). We therefore determined the effects of ascorbic acid on the As 2 O 3 -dependent activation of Rac1. When cells were pretreated with ascorbic acid prior to As 2 O 3 treatment, there was an increase in the As 2 O 3 -induced Cell lysates were immunoprecipitated with an anti-p38␤ antibody, and immunoprecipitated proteins were subjected to in vitro kinase assays, using ATF2 as an exogenous substrate. Proteins were resolved by SDS-PAGE, and phosphorylated proteins were detected by autoradiography. B, NB-4 cells were pre-incubated for 30 min in the presence or absence of SB203580 and were subsequently incubated for 60 min with As 2 O 3 as indicated, in the continuous presence or absence of SB203580. Cell lysates were immunoprecipitated with an anti-MAPKapK-2 antibody and subjected to an in vitro kinase assay using hsp25 as an exogenous substrate.
GTP-bound form of Rac1 (Fig. 4A), suggesting that the activation of Rac1 is downstream of free redox reactions. Consistent with this, treatment of cells with ascorbic acid also enhanced As 2 O 3 -dependent p38 activation (Fig. 4B). On the other hand, pre-treatment with dithiothreitol, a reducing agent, significantly diminished the phosphorylation/activation of p38 (Fig. 4,  B and C). Altogether these studies demonstrated that a cellular cascade involving Rac1, p38, and MAPKapK-2 is activated by arsenic trioxide downstream of redox reactions.
In subsequent studies we sought to determine the functional relevance of activation of the Rac1/p38 pathway by As 2 O 3 . Experiments were performed to determine the effects of pharmacological inhibition of p38 on the generation of growth inhibitory responses by As 2 O 3 in NB-4 cells. Cells were incubated for 5 days with As 2 O 3 , in the presence or absence of different concentrations of the SB203580 p38 inhibitor, and cell proliferation was determined by MTT assays. As expected, treatment of NB-4 cells with As 2 O 3 suppressed cell growth of NB-4 cells. Such a growth inhibitory effect was noticeable at a 2 M As 2 O 3 concentration, whereas it was not detectable at the lower concentration of 0.5 M (Fig. 5A). This is consistent with other studies that have demonstrated that at this low dose (0.5 M) As 2 O 3 does not inhibit proliferation but induces cell differentiation of acute promyelocytic leukemia cells (1)(2)(3)(4). Interestingly, concomitant treatment of the cells with the SB203580 p38 inhibitor enhanced the effects of As 2 O 3 in a dose-dependent manner and resulted in the generation of growth inhibitory responses at the lower (0.5 M) concentration of As 2 O 3 (Fig.  5A), indicating that the p38 inhibitor exhibits synergistic effects with As 2 O 3 . In a similar manner, treatment of cells with the p38 inhibitor SB203580 enhanced the growth inhibitory effects of As 2 O 3 on LNKAP prostate adenocarcinoma cells (Fig. 5B).
This finding, that p38 inhibition promotes As 2 O 3 -dependent growth suppression, prompted us to perform further studies to define whether such effects are because of enhancement of As 2 O 3 -induced apoptosis. NB-4 cells were incubated with arsenic trioxide for 2 days, in the presence or absence of SB203580, and the percentage of cells undergoing apoptosis was determined by flow cytometry. Treatment of cells with SB203580 alone did not alter the percent of background apoptotic cells compared with untreated cells (Fig. 6). As expected (50), treatment of cells with As 2 O 3 resulted in strong induction of apoptosis (Fig. 6), whereas concomitant treatment of cells with the p38 inhibitor further enhanced As 2 O 3 -induced cell death (Fig. 6).
Altogether, the findings using the SB203580 pharmacological inhibitor of p38 strongly suggested that this pathway exhibits a negative regulatory role on As 2 O 3 -induced apoptosis. To confirm that this is indeed the case, we determined whether overexpression of a phosphorylation-defective mutant (p38␤2AGF) exhibits dominant-negative effects on arsenic-induced cell death. We first generated the pCDN-p38␤2AGF mutant, in which threonine 180 and tyrosine 182 in the TGT motif of p38␤ were mutated to arginine and phenylalanine, respectively, rendering the mutant resistant to dual threonine-tyrosine phosphorylation and activation by upstream MAP kinases. MCF-7 cells were subsequently transfected with either the empty pCDN vector or the pCDN-p38␤2AGF mutant tagged with HA, and the transfectants were selected in G418. Prior to examining the effects of overexpression of this mutant on As 2 O 3 -dependent apoptosis, the expression of the p38␤2AGF protein and its ability to block endogenous p38␤ activation were examined. The p38␤2AGF mutant was expressed abundantly in the transfectants (Fig. 7A), as shown by anti-HA immunoblots. In addition, overexpression of the p38␤2AGF mutant blocked endogenous p38␤ kinase activation in response to arsenic trioxide stimulation, whereas

FIG. 4. As 2 O 3 induces activation of Rac1, and ascorbic acid enhances such an activation.
A, NB-4 cells were pre-incubated in the presence or absence of ascorbic acid for 60 min and were subsequently treated with As 2 O 3 for 30 min, as indicated. The cells were lysed, and total cell lysates were bound to either a GST-PBD fusion protein or GST alone (control). Bound proteins were analyzed by SDS-PAGE, and GTP-bound Rac1 was detected by anti-Rac1 immunoblotting. B, NB-4 cells were preincubated in the presence or absence of ascorbic acid or dithiothreitol for 60 min as indicated and were subsequently treated with As 2 O 3 for 30 min, as indicated. Total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody against the phosphorylated form of p38. C, the blot shown in B was stripped and re-probed with an anti-p38 antibody to control for loading. F) were treated in the presence or absence of the indicated doses of As 2 O 3 . Equal amounts of total cell lysates were resolved by SDS-PAGE and immunoblotted with an anti-phospho-p38 antibody (A, C, and E). The same blots were stripped and reprobed with an anti-p38 antibody (B, D, and F), to control for loading. such activation was intact in cells transfected with the empty pCDN vector (Fig. 7B). In a similar manner, the activation of the MAPKapK-2 kinase downstream of p38 was intact in cells transfected with the empty vector (Fig. 7C) but defective in the p38␤AGF transfectants (Fig. 7D).

FIG. 3. Phosphorylation/activation of p38 (p38␣) by As 2 O 3 . K562 CML-blast crisis cells (A and B) or LNKAP prostate carcinoma cells (C and D) or MCF-7 breast carcinoma cells (E and
Subsequently, experiments were performed in which the growth inhibitory effects of arsenic trioxide were examined in MCF-7 cells stably transfected with the p38␤2AGF mutant or with control empty vector. As 2 O 3 suppressed the growth of MCF-7 cells transfected with the empty vector in a dose-dependent manner (Fig. 8A). This effect was clearly enhanced in cells in which the p38␤2AGF mutant was expressed (Fig. 8A). Similarly, expression of the p38␤2AGF mutant protein enhanced As 2 O 3 -induced growth inhibition in a different clone of MCF-7 cells stably transfected with the p38␤2AGF mutant (Fig. 8B). These findings are consistent with the results of the studies using the SB203580 pharmacological inhibitor of p38 and further support a role for p38 as a negative regulator of arsenic-induced apoptosis.
To further explore the role of the p38 pathway in a more physiologically relevant system, we evaluated the effects of pharmacological inhibition of p38 on the induction of the suppressive effects of arsenic trioxide on primary leukemia progenitors from patients with CML. Bone marrow mononuclear cells from three patients with CML were isolated, and leukemic CFU-GM progenitor colony formation was determined by clonogenic assays in methylcellulose. Addition of As 2 O 3 to the cultures suppressed leukemic progenitor growth from the bone marrows of all three cases studied (Fig. 9), whereas addition of the SB203580 p38 inhibitor alone had no significant effects.  7. Stable overexpression of a p38␤2AGF mutant inhibits endogenous p38 kinase activity. A, MCF-7 cells stably transfected with either the empty pCDN vector or the HA-tagged pCDN-p38␤2AGF construct were lysed, and total cell lysates were analyzed by SDS-PAGE and immunoblotted with an anti-HA antibody. B, MCF-7 cells stably transfected with either the empty pCDN vector or the pCDN-p38␤2AGF construct were treated with As 2 O 3 as indicated, and cell lysates were immunoprecipitated with an anti-p38␤ antibody. Immunoprecipitates were subjected to an in vitro kinase assay, using ATF2 as an exogenous substrate. Proteins were resolved by SDS-PAGE, and phosphorylated proteins were detected by autoradiography. C and D, MCF-7 cells stably transfected with either the empty pCDN vector (C) or the pCDN-p38␤2AGF construct (D) were treated with As 2 O 3 as indicated, and cell lysates were immunoprecipitated with an anti-MAP-KapK-2 antibody. Immunoprecipitates were subjected to an in vitro kinase assay, using hsp25 as an exogenous substrate. Proteins were resolved by SDS-PAGE, and phosphorylated proteins were detected by autoradiography.
However, SB203580 further enhanced the suppressive effects of As 2 O 3 on leukemic CFU-GM (Fig. 9), demonstrating that activation of p38 exhibits negative regulatory effects on the leukemic progenitor cell growth suppression by arsenic trioxide.

DISCUSSION
Arsenic trioxide induces apoptosis of target cells and exhibits growth inhibitory effects against a variety of human cell lines in vitro and in vivo. The important properties that this agent exhibits against leukemia and other malignant cells have led to extensive studies, aimed to understand the mechanisms by which it mediates its effects. There is evidence that this compound induces cell cycle arrest, apoptosis, and/or differentiation of target cells, depending on the doses used and the cellular context (1-4, 8, 52). Previous studies have provided evidence for several mechanisms that may be contributing to arsenic-induced apoptosis. It has been demonstrated that, in acute promyelocytic leukemia cells, arsenic trioxide lowers bcl-2 levels and reduces NF-B translocation to the nucleus, whereas it also induces collapse of mitochondrial transmembrane potential, resulting in cytochrome c release and activation of caspase-3 (53,54). All these events appear to contribute to the induction of apoptosis, but the upstream regulatory mechanisms of such pro-apoptotic signals are not known. Interestingly, generation of reactive oxygen species by arsenic trioxide potentiates induction of cell killing (8,50), and an accumulating body of evidence points toward an important role of reactive oxygen species, particularly H 2 O 2 , on arsenic-induced apoptosis (55). Generation of H 2 O 2 is dependent on cellular glutathione stores, whereas reduced cellular glutathione (GSH) acts as an inhibitor of arsenic-dependent cell death, by either conjugating arsenic in the form of As(GS) 3 complexes or by sequestering the reactive oxygen species induced by arsenic (55). This has prompted studies aimed to determine whether depletion of intracellular GSH stores can enhance the effects of arsenic. Such studies have demonstrated that down-regulation of GSH levels by pre-treatment of myeloma cells with either ascorbic acid (55) or buthionine sulfoximine (5) promotes induction of arsenic-dependent apoptosis. On the other hand, increasing GSH cellular levels by pre-treatment with N-acetylcysteine attenuates arsenic-dependent cytotoxicity (55).
Altogether, these studies have established that induction of apoptotic signals by arsenic trioxide is dependent on the generation of reactive-oxygen species and the cellular redox system. However, it is not known whether during treatment of cells with arsenic trioxide other cellular pathways are activated in a negative-feedback manner. Induction of such cellular signals might counteract arsenic-activated cascades that mediate apoptosis and provide a mechanism by which malignant cells develop resistance to its pro-apoptotic properties. We have demonstrated recently that during differentiation of acute promyelocytic leukemia cell lines in response to all-trans-retinoic acid, there is activation of the p38 MAP kinase pathway (37). This pathway apparently exhibits negative regulatory effects on the induction of cell differentiation and growth inhibition by all-trans-retinoic acid (37), as pharmacological inhibitors of p38 strongly enhance all-trans-retinoic acid-dependent differentiation and growth suppression of acute promyelocytic leukemia cells (37). As arsenic trioxide is a potent inducer of apoptosis of acute promyelocytic leukemia cells, including all-trans-retinoic acid-resistant cells, we performed studies to examine whether it elicits activation of the p38 pathway in acute promyelocytic leukemia. Our studies provide the first evidence that arsenic trioxide activates the ␣ and ␤ isoforms of the p38 MAP kinase in acute promyelocytic leukemia cells. Such activation was detectable in both the NB-4 acute promyelocytic leukemia cell line and in an all-trans-retinoic acid-resistant NB-4 variant (NB-4.007/6) (38). This is of particular interest, as retinoic acid does not activate the p38 pathway in NB-4.007/6 cells, indicating that different early signals mediate arsenic-induced p38 activation, as compared with retinoic acid-dependent activation. In other experiments we demonstrated that p38 is activated during treatment of other cell lines of diverse malignant phenotypes with arsenic, indicating that arsenic-dependent activation of this cascade is not cell-type restricted. Our data also demonstrate that the small GTPase Rac1 is activated by arsenic and that the activation of p38 ultimately leads to engagement and activation of the MAPKapK-2 kinase, which functions as a downstream effector for the p38-pathway. Interestingly, activation of Rac1 and p38 by arsenic was enhanced by pre-treatment of cells with ascorbic acid and was blocked by dithiothreitol, indicating that GSH levels exhibit regulatory effects on its activation.
Altogether, our studies indicate that treatment of cells with arsenic results in sequential activation of a Rac1 3 (MKK) 3 p38(␣/␤) 3 MAPKapK-2 cascade, whose activation is probably downstream of cellular redox changes induced by arsenic. In experiments to define the functional role of this pathway in the induction of arsenic responses we found that inhibition of this cascade enhances arsenic-dependent induction of apoptosis and suppression of cell growth of arsenic-responsive cell lines. This was demonstrated by experiments using SB203580, a specific pharmacological inhibitor of p38, which acts by binding to the ATP site of the molecule to abrogate its kinase activity. The basis of the specificity of this pharmacological inhibitor of p38 has been established previously by mutagenesis studies and x-ray crystallographic structures of p38⅐inhibitor complexes (56 -58). Previous studies have also demonstrated that this pyridinyl imidazole compound blocks activation of both the p38␣ and p38␤ isoforms, which, based on our data, are both activated during arsenic trioxide treatment of target cells but not the p38␥ and p38␦ isozymes (16,23,59). Thus, it appears that important signals for regulation of arsenic-dependent responses are mediated by the ␣ and/or ␤ isoforms of p38. Consistent with this, we were able to demonstrate that overexpression of a phosphorylation-defective p38␤ isoform mutant enhances arsenic-dependent cytotoxicity in MCF-7 cells, confirming the data with the pharmacological inhibitor of p38.
The observation that the combination of arsenic trioxide and SB203580 has more potent cytotoxic effects than arsenic alone in vitro raises the possibility that such combination may result in more potent antileukemic effects in vivo. In addition to acute promyelocytic leukemia, another leukemia in which such combination may prove to be of therapeutic value is CML.
CML is a clonal myeloproliferative disorder of stem cells, characterized by the expression of the BCR-ABL oncoprotein. The BCR-ABL abnormal protein is the product of the bcr-abl oncogene, which is generated by the reciprocal translocation between chromosomes 9 and 22, resulting in the fusion of bcr to c-abl (60 -62). The function of the abnormal BCR-ABL tyrosine kinase is essential for the pathogenesis of CML (63) by mediating mitogenic effects via phosphorylation of protein substrates and activation of multiple downstream mitogenic pathways (64). Studies from other groups have demonstrated previously that arsenic trioxide induces apoptosis of BCR-ABLexpressing cell lines (65). Interestingly, arsenic trioxide-induced apoptosis of such cell lines was associated with a decrease in the levels of the BCR-ABL oncoprotein but no significant effects in the levels of Bcl-x L , Bax, Apaf-1, Fas, and FasL (65). Other studies have suggested that arsenic trioxide may have selective inhibitory effects on the proliferation of BCR-ABL-expressing cells, as it was found to induce apoptosis of Ph ϩ but not Ph Ϫ lymphoblasts, whereas ectopic expression of BCR-ABL in U937 myelomonocytic cells dramatically increased their sensitivity to arsenic trioxide, independently of BCR-ABL kinase activity (66). Also, arsenic trioxide has been shown previously to reduce proliferation of CML leukemic blasts (66). Our data are consistent with these findings, as they demonstrate that arsenic trioxide suppresses leukemic CFU-GM colony formation from CML bone marrows. Most importantly, they demonstrate that such effects can be enhanced by p38 inhibition, raising the possibility that p38 inhibitors may enhance the clinical activity of arsenic trioxide. This may prove to be important in the design of future therapeutic approaches in the treatment of CML, where studies with arsenic are being planned or are already in progress (51,67,68).