Arsenic trioxide (As2O3) exhibits important antitumor activities in vitro and in vivo, but the precise mechanisms by which it induces its effects are not known. We provide evidence that during treatment of BCR-ABL–expressing cells with As2O3, there is activation of a cellular pathway involving the p70 S6 kinase (p70S6K). Our data show that p70S6K is rapidly phosphorylated on Thr421 and Ser424 and is activated in an As2O3-inducible manner. The mammalian target of rapamycin (mTOR) is also phosphorylated/activated in an As2O3-inducible manner, and its activity is required for downstream engagement of p70S6K. p70S6K subsequently phosphorylates the S6 ribosomal protein on Ser235/Ser236 and Ser240/Ser244 to promote initiation of mRNA translation. Treatment of chronic myelogenous leukemia–derived cell lines with As2O3 also results in phosphorylation of the 4E-BP1 repressor of mRNA translation on Thr37/Thr46 and Thr70, sites required for its deactivation and its dissociation from the eukaryotic initiation factor 4E complex to allow cap-dependent mRNA translation. In studies to determine the functional relevance of this pathway, we found that inhibition of mTOR and downstream cascades enhances induction of apoptosis by As2O3. Consistent with this, the mTOR inhibitor rapamycin strongly potentiated As2O3-mediated suppression of primitive leukemic progenitors from the bone marrow of chronic myelogenous leukemia patients. Altogether, our data show that the mTOR/p70S6K pathway is activated in a negative feedback regulatory manner in response to As2O3 in BCR-ABL–transformed cells and plays a key regulatory role in the induction of anti-leukemic responses. [Mol Cancer Ther 2006;5(11):2815–23]
Arsenic trioxide (As2O3) is an arsenic derivative, which exhibits potent growth inhibitory effects against malignant cells (1–3). The remarkable antitumor effects of As2O3 in vitro and in vivo have prompted the development of various clinical trials that established its activity in acute promyelocytic leukemia (1–3). As2O3 is part of the standard treatment for this leukemia and is highly effective in cases that have developed resistance to retinoic acid treatment (1–3). As As2O3 has potent effects in vitro and in vivo against a variety of neoplastic cells, it is also currently under clinical development for the treatment of other hematologic malignancies as well, in particular, chronic myelogenous leukemia (CML) and multiple myeloma (1–4).
Because of the important antitumor properties of As2O3, extensive efforts have been made by several research groups to understand its mechanisms of action in malignant cells. In acute promyelocytic leukemia cells, it has been shown that As2O3 induces degradation of the PML-RARα fusion protein that may account in part for its anti-leukemic effects in these cells (2, 5). However, there is also evidence that acute promyelocytic leukemia cell differentiation can be induced without concomitant degradation of the PML-RARα fusion protein (6), indicating the existence of additional mechanisms. Other cellular events that may contribute to arsenic-induced apoptosis include suppression of bcl-2 levels and decreased nuclear factor-κB translocation to the nucleus (7) and collapse of mitochondrial transmembrane potential, resulting in cytochrome c release and activation of caspase-3 (8). Generation of reactive oxygen species by As2O3 potentiates induction of cell killing, and an accumulating body of evidence points towards a role for reactive oxygen species, particularly H2O2, on arsenic-induced apoptosis (9, 10). Recent work has uncovered an additional mechanism by which As2O3 may generate its anti-leukemic responses, involving inhibition of nuclear receptor function via c-Jun NH2-terminal kinase–mediated retinoid X receptor α phosphorylation (11). Thus, several cellular cascades seem to be engaged in the generation of arsenic-dependent growth inhibition and apoptosis of malignant cells. However, the role of pathways that may be activated in a negative feedback regulatory manner to counteract the induction of arsenic responses is unknown.
In the present study, we provide evidence for a novel signaling cascade activated in response to As2O3 in CML-derived cell lines, involving activation of the mammalian target of rapamycin (mTOR) and sequential downstream engagement of the p70 S6 kinase (p70S6K) and the S6 ribosomal protein. In addition, we show that the 4E-BP1 repressor of mRNA translation is phosphorylated in an As2O3-dependent manner on sites required for its deactivation and dissociation from the eukaryotic initiation factor 4E, to allow initiation of mRNA translation. Our data show that activation of mTOR and its downstream effectors results in negative regulatory effects on As2O3-induced cell death, as evidenced by the promotion of apoptosis during pharmacologic inhibition of mTOR in CML-derived cell lines. In addition, our data establish that inhibition of mTOR activation promotes the generation of the suppressive effects of As2O3 on primitive leukemic granulocyte-macrophage colony-forming unit (CFU-GM) progenitors from CML patients, supporting an important regulatory role for this pathway in the induction of the anti-leukemic effects of arsenic.
Materials and Methods
Cells and Reagents
The CML-derived KT-1, K562, and BV-173 cell lines were grown in RPMI 1640 supplemented with 10% fetal bovine serum and antibiotics. As2O3 was purchased from Sigma (St. Louis, MO). Antibodies against the phosphorylated forms of p70S6K, rpS6, mTOR, and 4E-BP1 were obtained from Cell Signaling Technology, Inc. (Beverly, MA). The FRAP/mTOR inhibitor rapamycin and the phosphatidylinositol 3′-kinase (PI3K) inhibitor LY294002 were obtained from Calbiochem, Inc. (La Jolla, CA).
Cell Lysis and Immunoblotting
Cells were stimulated with the indicated doses of As2O3 for the indicated times and subsequently lysed in phosphorylation lysis buffer as described previously (12). Immunoprecipitations and immunoblotting using an enhanced chemiluminescence method were done as previously described (12). In the experiments in which pharmacologic inhibitors of FRAP/mTOR or the PI3K were used, the cells were pretreated for 60 minutes with the indicated concentrations of the inhibitors and subsequently treated for the indicated times with As2O3, before lysis in phosphorylation lysis buffer. All immunoblotting experiments were highly reproducible, and each experiment shown is representative of at least three independent experiments.
Assays to detect the arsenic-dependent activation of the p70S6K were done as previously described (13). Briefly, KT-1 cells were lysed in phosphorylation lysis buffer, and cell lysates were immunoprecipitated with an antibody against p70S6K or control non-immune rabbit immunoglobulin (RIgG). In vitro kinase assays were done using a synthetic peptide substrate (AKRRRLSSLRA), and p70S6K activity was measured using an S6K assay kit (Upstate Biotechnology, Inc., Lake Placid, NY) according to the manufacturer's instructions. Values were calculated by subtracting nonspecific activity, detected in RIgG immunoprecipitates, from kinase activity detected in anti-p70S6K immunoprecipitates.
Cell Proliferation Assays
KT-1 or K562 cells were treated with the indicated doses of As2O3, in the presence or absence of rapamycin (20 nmol/L), for the indicated time periods. Cell proliferation assays using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide methodology were subsequently done as described previously (14).
Evaluation of Apoptosis
Cells were exposed to As2O3 in the presence or absence of rapamycin (20 nmol/L) or LY294002 (10 or 20 μmol/L). Flow cytometric assays to evaluate apoptosis by Annexin and propidium iodide staining were done essentially as previously described (15).
Human Hematopoietic Progenitor Cell Assays
Bone marrow or peripheral blood was obtained from patients with chronic myelogenous leukemia after obtaining consent, approved by the Institutional Review Board of Northwestern University. Bone marrow or peripheral blood mononuclear cells were used for clonogenic assays in methylcellulose as previously described (16). The cells were cultured in the presence or absence of As2O3 (1 μmol/L), with or without the indicated concentration of rapamycin (10 nmol/L) or LY294002 (10 μmol/L). Leukemic CFU-GM colonies were scored on day 14 of culture.
We initially examined whether treatment of CML-derived cell lines with As2O3 results in phosphorylation and activation of the p70S6K. KT-1 or K562 cells were incubated for different times with As2O3, and total cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody against the phosphorylated form of the p70S6K on Thr421 and Ser424. Treatment of both BCR-ABL–expressing cell lines with As2O3 resulted in rapid phosphorylation of the p70S6K, whereas there was no change in the amounts of p70S6K protein detected after treatment with As2O3 (Fig. 1A and B). We subsequently sought to directly determine whether such As2O3-dependent phosphorylation of the p70S6K results in activation of its kinase domain. KT-1 cells were incubated in the presence or absence of As2O3, and after immunoprecipitation of cell lysates with an anti-p70S6K antibody, immunoprecipitates were subjected to an in vitro kinase assay (13). As shown in Fig. 1C, As2O3 treatment resulted in an activation of the catalytic domain of p70S6K (Fig. 1C), suggesting that this kinase may participate in a cellular cascade that regulates the induction of As2O3 responses in BCR-ABL–expressing cells.
Previous studies have established that mTOR regulates activation of the p70S6K, downstream of the PI3K and PDK-1 (reviewed in refs. 17–19). It is also established that activation of mTOR requires its phosphorylation on Ser2448. To determine whether mTOR is phosphorylated/activated in response to As2O3, KT-1 or K562 cells were treated with As2O3, and after cell lysis, total lysates were resolved by SDS-PAGE and immunoblotted with an anti-phospho mTOR antibody. Some baseline phosphorylation of mTOR was detectable before As2O3 treatment (Fig. 2A and B). However, As2O3 treatment of the cells strongly enhanced phosphorylation/activation of mTOR, showing that this protein is indeed engaged in an As2O3-activated cellular cascade on target cells (Fig. 2A and B). Pharmacologic inhibition of mTOR using rapamycin blocked the As2O3-inducible phosphorylation of p70S6K (Fig. 2C and D), suggesting that such phosphorylation is mTOR dependent. The phosphorylation of p70S6K was also blocked when cells were pretreated with the PI3K inhibitor LY294002 (Fig. 2C and D), suggesting that, as in other systems (17–19), activation of mTOR and p70S6K by As2O3 requires upstream PI3K activity.
In subsequent studies, we sought to identify functional downstream effectors of the As2O3-activated mTOR/p70S6K cascade. One well-known effector of the p70S6K is the 40S ribosomal S6 protein (17–19). This protein is a direct substrate for the activity of p70S6K, and its phosphorylation results in initiation of translation for mRNAs that have oligopyrimidine tracts in the 5′ untranslated region (17–19). We determined whether ribosomal S6 protein is phosphorylated in response to treatment of BCR-ABL–expressing cells with As2O3. As2O3 treatment of several different BCR-ABL–expressing cell lines, including KT-1 (Fig. 3A), K562 (Fig. 3B), and BV-173 (Fig. 3C), resulted in strong phosphorylation of rpS6 on Ser235/Ser236, whereas there was no change in the total amounts of ribosomal S6 detected before and after As2O3 stimulation. As in the case of the phosphorylation of p70S6K, the phosphorylation of the rpS6 protein was abrogated by pretreatment of cells with either the mTOR inhibitor rapamycin or the PI3K inhibitor LY294002 (Fig. 3B and C). Treatment of different CML cell lines with As2O3 also resulted in phosphorylation of the S6 ribosomal protein on Ser240/Ser244 (Fig. 3D and E). Such phosphorylation was also inhibited by pretreatment of cells with either rapamycin or LY294002 (Fig. 3D and E), further establishing that As2O3 engagement of the S6 ribosomal protein occurs downstream of mTOR and the p70S6K. When the phosphorylation of the S6 ribosomal protein was examined after prolonged treatment of the cells with As2O3, we found that such phosphorylation is prolonged and can be detected after 24 and 48 hours of treatment of the cells (Fig. 3F and G). Thus, it seems that the phosphorylation of S6 ribosomal protein in response to As2O3 is biphasic, with one early peak occurring early, within 30 minutes of treatment of the cells, and a second delayed peak occurring after prolonged treatment of the cells (24–48 hours).
Previous work has shown that in response to insulin, cytokines, and retinoids, the 4E-BP1 repressor of mRNA translation is phosphorylated in a PI3K- and mTOR-dependent manner (13, 20–22), and that such phosphorylation leads to its dissociation from the initiation factor eukaryotic initiation factor 4E, resulting in induction of mRNA translation (21). We sought to determine whether As2O3 treatment results in phosphorylation of 4E-BP1 to regulate initiation of translation. KT-1, K562, or BV-173 cells were treated with As2O3, and after cell lysis, total lysates were analyzed by SDS-PAGE and immunoblotted with antibodies against different phosphorylation sites in 4E-BP1. Treatment of all different cell lines with As2O3 resulted in phosphorylation of 4E-BP1 on sites required for its deactivation and dissociation from eukaryotic initiation factor 4E, including Thr70 (Fig. 4A–C) and Thr37/Thr46 (Fig. 4D–E; data not shown). Inhibition of PI3K activity using the LY294002 inhibitor blocked phosphorylation of 4E-BP1 on both sites (Fig. 4B–E). On the other hand, phosphorylation on Thr70 (Fig. 4B and C) was completely blocked by rapamycin, whereas phosphorylation on Thr37/Thr46 was not (Fig. 4D and E). These findings strongly suggest that the activation of 4E-BP1 by As2O3 in CML cell lines is functionally relevant, as it follows the hierarchical phosphorylation on sites essential for deactivation of the protein and its dissociation from eukaryotic initiation factor 4E.
In subsequent studies, we sought to determine the functional relevance of activation of the mTOR/p70S6K pathway in response to As2O3. Experiments were done to determine the effects of pharmacologic inhibition of mTOR on the generation of growth inhibitory responses by As2O3 in KT-1 and K562 cells. Cells were incubated for 5 days with As2O3, in the presence or absence of the mTOR inhibitor rapamycin, and cell proliferation was examined using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assays. As expected, treatment of KT-1 or K562 cells with As2O3 suppressed cell growth (Fig. 5A and B). Such a growth inhibitory response was significant (>50%) only at the high concentration of As2O3 (2 μmol/L; Fig. 5). Interestingly, concomitant treatment of the cells with rapamycin enhanced the effects of As2O3 in a dose-dependent manner and resulted in the generation of growth suppression at the lower (0.5–1 μmol/L) concentrations of As2O3 in KT-1 (Fig. 5A) and K562 cells (Fig. 5B). These findings prompted us to perform further studies, aimed to define whether rapamycin potentiates As2O3-induced apoptosis. KT-1 (Fig. 5C) or K562 (Fig. 5D) cells were incubated with As2O3, in the presence or absence of rapamycin or LY294002, and the percentage of cells undergoing apoptosis were determined by flow cytometry. Treatment of cells with As2O3 resulted in induction of apoptosis (Fig. 5C and D), whereas concomitant treatment of the cells with the mTOR inhibitor rapamycin or the PI3K inhibitor LY294002 further enhanced As2O3-induced cell death (Fig. 5C and D). Thus, pharmacologic inhibition of the mTOR pathway enhances induction of arsenic-dependent apoptosis and growth suppression of BCR-ABL–expressing cell lines, suggesting an important role for this pathway in the control of As2O3 responses.
To further explore the role of such activation in a more physiologically relevant system, we evaluated the effects of inhibition of this cascade on the induction of the suppressive effects of As2O3 on primary leukemia progenitors from a relatively large number of patients with CML. Bone marrow or peripheral blood mononuclear cells from 11 different CML patients were isolated, and leukemic CFU-GM progenitor colony formation was determined by clonogenic assays in methylcellulose. Consistent with our previous studies (23), addition of As2O3 to the cultures suppressed leukemic CFU-GM progenitor growth (Fig. 6). However, concomitant addition of the mTOR inhibitor rapamycin strongly enhanced the suppressive effects of As2O3 on leukemic CFU-GM progenitor growth (two-tailed P = 0.00001), strongly suggesting that the mTOR pathway negatively regulates the generation of As2O3 responses in BCR-ABL cells.
Despite the extensive work in the area of As2O3 signaling, the precise mechanisms required for the generation of As2O3-inducible apoptosis in different cellular backgrounds remain unclear. The well-established role of this agent in the treatment of acute promyelocytic leukemia and its potential applications to other hematologic malignancies have ignited extensive efforts to understand the mechanisms by which it regulates generation of anti-leukemic responses (1–3). The cellular effects of As2O3 are concentration dependent. As2O3 induces differentiation of acute promyelocytic leukemia blasts at low concentrations (<0.5 μmol/L), whereas for the induction of apoptosis, high final concentrations (>2 μmol/L) are required (1–3). Identifying ways to enhance the effects of As2O3 on malignant cells is of particular interest, as it may facilitate the development of novel therapeutic approaches using lower nontoxic doses of arsenic for the treatment of leukemias.
CML is characterized by the expression of the abnormal BCR-ABL oncoprotein. BCR-ABL is the protein product of the bcr-abl oncogene, which results from the reciprocal translocation between chromosomes 9 and 22, and the abnormal fusion of the bcr and c-abl genes (23). Extensive studies over the years have established that the constitutively activated tyrosine kinase activity of BCR-ABL promotes leukemic transformation by activating multiple downstream mitogenic cascades (24). Several signaling elements have been shown to be regulated by BCR-ABL, including the Ras-GAP (25), the Shc oncoprotein (26), the tyrosine phosphatase SHP-2 and phosphatidylinositol polyphosphate 5′-phosphatase SHIP (27), the c-CBL (28) and Vav (29) proto-oncogene products (28), the transcriptional activator Stat5 (30, 31), and the PI3K/Akt signaling pathway (32). In addition, mTOR has been recently implicated as a downstream effector of BCR-ABL (33–35). Consistent with this, the BCR-ABL kinase inhibitor imatinib mesylate (Gleevec) has been shown to block mTOR-dependent signals (33–35), suggesting that inhibition of mitogenic pathways activated downstream of mTOR may be important for the generation of the anti-leukemic properties of this agent.
Although imatinib mesylate is clearly the most potent agent available against CML cells in vitro and in vivo, resistance to its anti-leukemic properties develops in many instances (36, 37), underscoring the importance of developing novel therapeutic approaches to overcome such resistance. As2O3 exhibits potent inhibitory properties against BCR-ABL–expressing cells (4) and has been shown to exhibit synergistic effects with imatinib mesylate (38–41). Moreover, clinically relevant concentrations of As2O3 can induce apoptosis of BCR-ABL cells (38, 42) and suppress endogenous levels of BCR-ABL protein expression (43), further emphasizing the potential of As2O3 as an agent in the treatment of CML.
Despite the well-established anti-leukemic and proapoptotic effects of As2O3 in CML cells, the mechanisms that regulate induction of such effects are not known. In the present study, we show that As2O3 treatment regulates activation of the mTOR-p70S6K pathway in BCR-ABL–expressing cells, and we provide the first evidence for the existence of an As2O3-activated signaling cascade involved in the regulation of mRNA translation. Our data show that As2O3 induces phosphorylation/activation of mTOR in the KT-1 and K562 cell lines and downstream phosphorylation/activation of p70S6K. Such As2O3-dependent activation of p70S6K may play an important role in the regulation of mRNA translation, as it results in the downstream phosphorylation of the S6 ribosomal protein. The S6 ribosomal protein is known to participate in the initiation of translation of mRNAs with oligopyrimidine tracts in the 5′ untranslated region (44–46). Our data establish that As2O3 treatment results in phosphorylation of S6 on Ser235/Ser236 and Ser240/Ser44, in an mTOR-dependent manner, directly implicating it in the regulation of arsenic-inducible responses. We also show that As2O3 induces phosphorylation and deactivation of the translational repressor 4E-BP1, an event required for its dissociation from the eukaryotic initiation factor eukaryotic initiation factor 4E and the start of cap-dependent translation (17, 20, 21).
It should be pointed out that a previous study (43) suggested that As2O3 decreases activation of p70S6K after prolonged treatment (48 hours) of K562 cells. Such down-regulation was proposed as a mechanism for the arsenic-induced decrease in BCR-ABL expression, as the bcr-abl mRNA possesses a 5′ untranslated region containing a TOP sequence. However, in that article (43), the authors did not examine the phosphorylation of the S6 ribosomal protein at earlier time points, as in our study. Independently of that, we failed to reproduce the findings of that study, as we have consistently observed that there is strong phosphorylation of the S6 ribosomal protein after 48 hours of As2O3 treatment of KT-1 and K562 cells, despite some decrease in the total rpS6 protein levels. We cannot account for the differences between our study and that from the study of Nimmanapalli et al., but a recent study showed that As2O3 induces phosphorylation/activation of the Akt kinase in HL-60 cells (47). The results of that study (47) are consistent with our findings, as the Akt kinase is a known regulator of the p70S6K.
Altogether, our studies show that two distinct cellular pathways, known to regulate translation, are activated in response to As2O3 downstream of mTOR. In efforts to understand the functional relevance of these pathways, we found that inhibition of mTOR activation enhances the induction of arsenic-mediated apoptosis and growth suppression in CML-derived cell lines. Moreover, the mTOR inhibitor rapamycin was found to promote the suppressive effects of As2O3 on primary leukemic CFU-GM progenitors from patients with CML. Thus, activation of mTOR-dependent pathways occurs in a negative feedback regulatory manner to counteract the anti-leukemic properties of As2O3, in a manner similar to the previously described activation of the p38 mitogen-activated protein kinase cascade (48). The precise sequence of upstream signaling events that lead to activation of mTOR remains to be determined. Our data raise the possibility of involvement of the PI3K in such regulation, as in addition to mTOR inhibition, PI3K blockade also promotes the suppressive effects of As2O3 on leukemic progenitor cell growth. However, it is also possible that the arsenic-inducible mTOR activation occurs in a PI3K-independent manner, possibly via engagement of tuberous sclerosis complex and Rheb (17), and future studies should address this issue. Independently of the precise upstream regulatory signals, our studies strongly suggest that combination of As2O3 with pharmacologic inhibitors of mTOR may be a novel approach to overcome arsenic resistance in leukemic cells. They also raise the potential of future clinical trials in CML-blast crisis and possibly other leukemias, involving combinations of As2O3 with mTOR inhibitors.
Grant support: Department of Veterans Affairs (L.C. Platanias), NIH grants CA94079 and CA77816 (L.C. Platanias), and NIH/National Cancer Institute training grant T32 CA09560 (P. Yoon).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.