The p38 MAPK Pathway Mediates the Growth Inhibitory Effects of Interferon- a in BCR-ABL-expressing Cells*

The mechanisms by which interferon- a (IFN- a ) mediates its anti-leukemic effects in chronic myelogenous leukemia (CML) cells are not known. We determined whether p38 MAPK is activated by IFN- a in BCR-ABL-expressing cells and whether its function is required for the generation of growth inhibitory responses. IFN- a treatment induced phosphorylation/activation of p38 in the IFN- a -sensitive KT-1 cell line, but not in IFN- a -re-sistant K562 cells. Consistent with this, IFN- a treatment of KT-1 (but not K562) cells induced activation of the small GTPase Rac1, which functions as an upstream regulator of p38. In addition, IFN- a -dependent phospho-rylation/activation of p38 was induced by treatment of primary granulocytes isolated from the peripheral blood of patients with CML. To define the functional role of the Rac1/p38 MAPK pathway in IFN- a signaling, the effects of

Interferons are potent regulators of malignant hematopoiesis and exhibit growth inhibitory effects in leukemia cells in vitro and in vivo (1)(2)(3)(4)(5)(6). Extensive studies have established the efficacy of interferon-␣ (IFN-␣) 1 in the treatment of leukemias, and this cytokine is currently the treatment of choice for patients with chronic myelogenous leukemia (CML) that are not eligible for bone marrow transplantation (7,8). It is of particular interest that, among several other hematologic malignancies, CML exhibits very high sensitivity to the growth inhibitory effects of IFN-␣ in vivo (7,8). CML is a clonal myeloproliferative disorder of hematopoietic stem cells, and the hallmark of the disease is the expression of the BCR-ABL oncoprotein in the malignant cells. BCR-ABL 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 and the formation of the abnormal bcr/abl proto-oncogene (9 -11). The abnormal bcr/abl proto-oncogene encodes the constitutively active BCR-ABL tyrosine kinase, which plays an essential role in the pathogenesis of the disease (12) via phosphorylation of protein substrates and activation of multiple downstream mitogenic pathways (13).
Recent studies have established that, in addition to the STAT pathway, type I IFNs activate MAPKs, including ERK kinases (15,16) and p38 MAPK (17)(18)(19). p38 MAPK plays a critical role in IFN-␣ signaling, as its function is required for IFN-␣-dependent gene transcription via both IFN-stimulated response elements (17,18) and IFN-␥-activated sites (19). Such regulatory effects of p38 MAPK on IFN transcriptional activation are unrelated to tyrosine or serine phosphorylation of STAT proteins and apparently involve activation of downstream signaling pathways that function independently of the STAT pathway (19). Thus, coordination of the STAT and p38 MAPK pathways is required for transcriptional activation of IFN-␣-sensitive genes.
The mechanisms of action of IFN-␣ in CML are not understood. Very little is known of the signaling events induced by IFN-␣ in BCR-ABL-expressing cells, primarily due to a lack of IFN-sensitive, BCR-ABL-expressing cell line model systems. In this study, we used the recently established IFN-␣-sensitive KT-1 cell line (20) to determine whether p38 and its upstream effector, Rac1, are engaged in IFN signaling in BCR-ABLexpressing cells. We also examined whether the p38 pathway is engaged in IFN-␣ signaling in primary cells obtained from the peripheral blood of patients with CML. Our data indicate that IFN-␣ activates the Rac1/p38 MAPK pathway in CML cells. Most importantly, pharmacological blockade of p38 reverses the inhibitory effects of IFN-␣ in KT-1 cells and primary leukemic bone marrow-derived hematopoietic progenitors, providing the first direct evidence that the p38 MAPK pathway is essential for the generation of the growth inhibitory and antileukemic effects of IFN-␣.

EXPERIMENTAL PROCEDURES
Cells and Reagents-The KT-1 and K562 cell lines were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum and antibiotics. Human recombinant IFN-␣2 was provided by Hoffmann-La Roche. Human recombinant consensus IFN-␣ was provided by Amgen Inc. Antibodies against the phosphorylated forms of p38 and ERK2 were obtained from New England Biolabs Inc. and used for immunoblotting. A polyclonal antibody against p38 was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal antibodies against STAT1, Rac1, and ERK2 were obtained from Transduction Laboratories (Lexington, KY). An antibody that specifically recognizes the tyrosine-phosphorylated form of STAT1 at tyrosine 701 and an antibody that recognizes anti-phosphotyrosine (4G10) were obtained from Upstate Biotechnologies, Inc. and used for immunoblotting. The p38 MAPK inhibitors SB203580 and SB202190 and the MEK kinase inhibitor PD098059 were purchased from Calbiochem.
Rac1 Activation Assays-The activation of Rac1 by IFN-␣ was determined using a recently described methodology (23). Briefly, the pGEX-4T3 construct encoding the GTPase-binding domain of human PAK1 were analyzed by SDS-PAGE and immunoblotted with an antibody against the phosphorylated/activated form of p38. B, the blot shown in A was stripped and reprobed with an antibody against p38 to control for protein loading. C, KT-1 or K562 cells were incubated in the presence or absence of IFN-␣ for the indicated times (min). Equal amounts of total cell lysates (100 g/lane) were analyzed by SDS-PAGE and immunoblotted with an antibody against the phosphorylated/activated form of p38. D, the blot shown in C was stripped and reprobed with an antibody against p38 to control for protein loading. p38 MAPK Pathway Mediates Growth Inhibitory Effects of IFN-␣ (23) was expressed in Escherichia coli as a GST fusion protein (GST-PBD). The cells were treated with IFN for the indicated times and lysed in phosphorylation lysis buffer. Cell lysates were incubated with 5 g of GST-PBD, and bound proteins were separated by SDS-PAGE and immunoblotted with a monoclonal antibody against Rac1 to detect GTPbound Rac1.
In Vitro Kinase Assays-In vitro kinase assays to detect the activation of the MAPKAPK-2 and MAPKAPK-3 kinases by IFN-␣ were performed as previously described (17).
Cell Proliferation Assays-Cells were pretreated for 30 -60 min with SB203580 (20 M), SB202190 (10 M), or PD098059 (20 M) as indicated and subsequently incubated for 7 days with the indicated doses of IFN-␣ in the continuous presence of the pharmacological inhibitors. Cell proliferation assays using the MTT method were performed as described previously (24).
IFN-␣-induced Antiviral Activity Assay-The antiviral effects of IFN-␣ in KT-1 cells were determined by assaying its activity against encephalomyocarditis virus infection as previously described (25).
Isolation of Peripheral Blood Granulocytes from Patients with CML-Informed consent was obtained from patients with CML according to the guidelines established by the Institutional Review Board of the University of Illinois at Chicago. Polymorphonuclear leukocytes were separated from peripheral venous blood using Mono-Poly Resolving Medium (ICN Biomedicals, Aurora, OH). After centrifugation at 300 ϫ g for 30 min at room temperature, the plasma and mononuclear leukocyte band were discarded, and the polymorphonuclear leukocyte band was transferred to individual tubes. Cells were washed and resuspended in culture medium prior to stimulation with IFN-␣.
Hematopoietic Progenitor Cell Assays-The effects of IFN-␣ on the growth of hematopoietic progenitors from patients with CML was determined in methylcellulose assays as described previously (26). Bone marrow aspirate specimens were obtained under local anesthesia from patients with CML after obtaining informed consent. Bone marrow mononuclear cells were separated by Ficoll-Hypaque sedimentation, and cells were cultured in methylcellulose mixture containing hematopoietic growth factors (26) in the presence or absence of IFN-␣ and SB203580 (10 M), SB202190 (10 M), or PD098059 (2 M). Granulocyte/macrophage colony-forming units from leukemic bone marrow samples were scored on day 14 of culture.

RESULTS
We sought to determine whether the p38 MAPK pathway is activated in response to IFN-␣ treatment in BCR-ABL-expressing cells of CML origin. We first performed studies with the CML-derived human leukemia cell line KT-1, which expresses BCR-ABL and is sensitive to the growth inhibitory effects of IFN-␣ (20,27). Cells were incubated in the presence or absence of IFN-␣; and after cell lysis, total lysates were resolved by SDS-PAGE and immunoblotted with an antibody against the phosphorylated/activated form of p38. IFN-␣ induced strong phosphorylation/activation of p38 MAPK in these cells (Fig. 1,  A and B). On the other hand, IFN-␣ treatment of the CMLderived K562 cell line, which is resistant to the growth inhibitory effects of IFN-␣ (27), failed to induce phosphorylation/

FIG. 4. IFN-␣-dependent tyrosine phosphorylation of STAT2 and STAT1 in KT-1 and K562 cells.
A, KT-1 cells were treated with IFN-␣ as indicated. Cell lysates were immunoprecipitated with an antibody against STAT2 or control nonimmune rabbit immunoglobulin (RIgG) as indicated, and immunoprecipitated proteins were analyzed by SDS-PAGE and immunoblotted with an anti-phosphotyrosine (anti-pTyr) antibody (4G10). B, K562 cells were treated with IFN-␣ as indicated. Cell lysates were immunoprecipitated with an antibody against STAT2 or control nonimmune rabbit IgG as indicated, and immunoprecipitated proteins were analyzed by SDS-PAGE and immunoblotted with an anti-phosphotyrosine antibody (4G10). C, KT-1 cells were treated with IFN-␣ as indicated. Cell lysates were immunoprecipitated with an antibody against STAT1 or control nonimmune rabbit IgG as indicated, and immunoprecipitated proteins were analyzed by SDS-PAGE and immunoblotted with an anti-phosphotyrosine antibody (4G10). D, the blot shown in C was stripped and reprobed with an anti-STAT1 monoclonal antibody. E, K562 cells were treated with IFN-␣ as indicated. Cell lysates were immunoprecipitated with an antibody against STAT1 or control nonimmune rabbit IgG as indicated, and immunoprecipitated proteins were analyzed by SDS-PAGE and immunoblotted with an anti-phosphotyrosine antibody (4G10). F, the blot shown in E was stripped and reprobed with an anti-STAT1 monoclonal antibody.

FIG. 5. Tyrosine phosphorylation/activation of STAT1 occurs independently of p38 activation in KT-1 cells.
A, KT-1 cells were pretreated with SB203580 for 30 min and subsequently treated with IFN-␣ for 10 min at 37°C as indicated. Cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody against the tyrosinephosphorylated form of STAT1 on tyrosine 701. B, the blot shown in A was stripped and reprobed with an anti-STAT1 monoclonal antibody to control for loading. C, KT-1 cells were pretreated with SB202190 for 30 min and subsequently treated with IFN-␣ for 10 min at 37°C as indicated. Cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody against the tyrosine-phosphorylated form of STAT1 on tyrosine 701. D, the blot shown in C was stripped and reprobed with an anti-STAT1 monoclonal antibody to control for loading.
p38 MAPK Pathway Mediates Growth Inhibitory Effects of IFN-␣ activation of p38 (Fig. 1C). It should be pointed out that the inability to detect signals in anti-phospho-p38 immunoblots using lysates from K562 cells was not due to lack of p38 expression, as the p38 protein was abundantly expressed in these cells (Fig. 1D). Thus, p38 is phosphorylated/activated in an IFN-␣-dependent manner in KT-1 (but not K562) cells, indicating that the upstream components of this pathway are functional in the KT-1 cell line, but defective in K562 cells. Interestingly, the ERK2 kinase, another member of the MAPK family that has also previously been shown to be activated in response to IFN-␣ in non-BCR-ABL-expressing cells (15,16), was phosphorylated in an IFN-␣-dependent manner in both KT-1 and K562 cells (Fig. 2).
In parallel studies, we sought to determine the activation of other IFN-␣-inducible signaling elements in KT-1 and K562 cells. Consistent with our previous report (27), the TYK2 tyrosine kinase was phosphorylated/activated in an IFN-␣-dependent manner in KT-1 cells (Fig. 3A). TYK2 was also phosphorylated/activated in K562 cells (Fig. 3B), indicating that defective activation of p38 in these cells does not result from lack of TYK2 activation. In studies to determine whether STAT proteins are activated in KT-1 and K562 cells, we found that the IFN-␣-dependent tyrosine phosphorylation of STAT2 and STAT1 was clearly detectable in both KT-1 and K562 cells (Fig.  4). This indicates that the lack of activation of p38 in K562 cells is not due to a common early upstream defect at the receptor level that also results in defective STAT2 and STAT1 activa-tion. Furthermore, in studies to determine the effects of p38 inhibition on the IFN-␣-dependent tyrosine phosphorylation of STAT1, we found that inhibition of p38 activation using the inhibitor SB203580 or SB202190 did not affect STAT1 phosphorylation on Tyr 701 (Fig. 5). Thus, as in the case of non-BCR-ABL-expressing cells (17,19), the IFN-␣-inducible p38 MAPK and STAT pathways appear to function independently of each other in BCR-ABL-expressing cells, and activation of the p38 kinase does not play a regulatory role in phosphorylation/activation of STAT1.
We subsequently determined whether the small GTPase Rac1, which functions as an upstream effector for p38 MAPK (19, 28 -30), is activated by IFN-␣ in BCR-ABL-expressing cells. KT-1 or K562 cells were treated with IFN-␣ for the indicated times; and after cell lysis, lysates were bound to a GST fusion protein encoding the GTPase-binding domain of PAK1 (23) to detect GTP-bound Rac1. IFN-␣ treatment induced strong activation of Rac1 in KT-1 cells (Fig. 6A). In K562 cells, some base-line activation of Rac1 was detectable, but there was no further increase in the amount of GTP-bound Rac1 in lysates from IFN-␣-treated cells (Fig. 6B), indicating that IFN-␣-inducible activation of Rac1 is defective in these cells. We

p38 MAPK Pathway Mediates Growth Inhibitory Effects of IFN-␣
also performed studies to determine whether the activation of Rac1 in KT-1 cells is tyrosine kinase-dependent. The IFN-␣inducible activation of Rac1 was blocked by pretreatment of cells with the tyrosine kinase inhibitor genistein, but not the phosphatidylinositol 3Ј-kinase inhibitor wortmannin (Fig. 6C). Thus, engagement of Rac1 in IFN-␣ signaling in BCR-ABLexpressing cells requires upstream activation of a tyrosine kinase(s), but not the function of the phosphatidylinositol 3Јkinase, which is also activated by the type I IFN receptor (16).
Taken together, our data established that the Rac1/p38 MAPK pathway is activated independently of the STAT pathway in the IFN-␣-sensitive, BCR-ABL-expressing KT-1 cell line. To obtain information on the functional relevance of p38 activation in the induction of IFN-␣ responses in CML cells, we performed experiments in which the effects of pharmacological inhibition of p38 on the generation of the antiproliferative and antiviral activities of IFN-␣ were examined. KT-1 cells were preincubated in the presence or absence of the p38 MAPK inhibitor SB203580 or SB202190 and subsequently treated with IFN-␣ in the continuous presence or absence of the p38 MAPK pharmacological inhibitors. Cell proliferation was subsequently assessed using an MTT assay. As expected (20,27), treatment of KT-1 cells with IFN-␣ suppressed the growth of KT-1 cells (Fig. 7, A and B). However, concomitant treatment of cells with SB203580 (Fig. 7A) or SB202190 (Fig. 7B) reversed the growth inhibitory effects of IFN-␣. On the other hand, treatment of cells with PD098059, which selectively inhibits activation of ERK kinases (but not p38), did not abrogate the IFN-␣-inducible growth inhibition (Fig. 7C). On the contrary, it slightly enhanced the generation of an IFN-␣ antiproliferative response in these cells.
In other studies, we determined whether KT-1 cells are sensitive to the antiviral effects of IFN-␣ and whether p38 plays a role in the induction of an antiviral state by IFN-␣. As shown in Fig. 8, KT-1 cells were susceptible to the cytopathic effects of encephalomyocarditis virus infection, whereas treatment with IFN-␣ inhibited virus replication in a dose-dependent manner. Treatment with SB203580 partially inhibited the induction of IFN-␣-regulated antiviral responses, suggesting a role for p38 in the IFN-␣ induction of antiviral effects.
We subsequently sought to identify putative downstream effectors of p38 that may mediate the induction of IFN-␣-dependent growth inhibitory responses in KT-1 cells. We determined whether the MAPKAPK-2 and MAPKAPK-3 kinases, which are activated downstream of p38 in other systems (31,32), are engaged in IFN-␣ signaling in BCR-ABL-expressing cells. KT-1 cells were treated with IFN-␣ in the presence or absence of the inhibitor SB203580. The cells were lysed and immunoprecipitated with specific antibodies against MAPK-APK-2 or MAPKAPK-3. Subsequently, in vitro kinase assays were carried out on the immunoprecipitates using HSP25 as an exogenous substrate. As shown in Fig. 9, treatment of KT-1 cells with IFN-␣ resulted in strong activation of MAPKAPK-2 and MAPKAPK-3. Such activation was blocked by pretreatment of cells with SB203580, indicating that activation of the kinase domains of MAPKAPK-2 and MAPKAPK-3 is p38-dependent. Thus, pharmacological inhibition of p38 reverses the biological effects of IFN-␣ in the KT-1 cell line, and such reversal correlates with inhibition of MAPKAPK-2/3 kinases (but not STAT) activation (Fig. 5), raising the possibility that MAPKAPK-2 and MAPKAPK-3 are downstream targets of p38 that mediate IFN-␣-dependent growth inhibition in BCR-ABLexpressing cells.
Based on the data from the studies with the KT-1 cell line, the p38/MAPKAPK-2/3 pathway appeared to play an important role in the generation of the antiproliferative effects of IFN-␣ in BCR-ABL-expressing cells. This prompted us to ex- in the presence or absence of SB203580 as indicated. The cells were subsequently incubated in the presence or absence of IFN-␣ for 30 min as indicated in the continuous presence or absence of SB203580. The cells were subsequently lysed; cell lysates were immunoprecipitated (IP) with an antibody against MAPKAPK-2; and in vitro kinase assays were carried out on the immunoprecipitates using HSP25 as an exogenous substrate. Proteins were analyzed by SDS-PAGE, and the phosphorylated form of HSP25 was detected by autoradiography. B, KT-1 cells were preincubated for 30 min at 37°C in the presence or absence of SB203580 as indicated. The cells were subsequently incubated in the presence or absence of IFN-␣ for 60 min as indicated in the continuous presence or absence of SB203580. The cells were subsequently lysed, and cell lysates were immunoprecipitated with an antibody against MAPKAPK-3 and subjected to in vitro kinase assays using HSP25 as an exogenous substrate. Proteins were analyzed by SDS-PAGE, and phosphorylated HSP25 was detected by autoradiography.

p38 MAPK Pathway Mediates Growth Inhibitory Effects of IFN-␣
tend these studies to determine the biological relevance of this pathway in the generation of the anti-leukemic effects of IFN-␣ in primary cells from patients with CML. We first determined whether p38 is phosphorylated/activated in an IFN-␣-dependent manner in isolated granulocytes obtained from the peripheral blood of four different patients with CML. The granulocytes were treated with IFN-␣; and after cell lysis, total lysates were analyzed by SDS-PAGE and immunoblotted with an antibody against the phosphorylated/activated form of p38. Treatment of cells with IFN-␣ induced phosphorylation/activation of p38 MAPK (Fig. 10, A, C, E, and G), whereas there was no change in the levels of p38 protein expression (B, D, F, and H).
We subsequently determined whether the p38 MAPK pathway plays a role in inducing the antiproliferative effects of IFN-␣ on primary leukemic hematopoietic progenitors. We determined whether two specific inhibitors of p38 (SB203580 and SB202190) reversed the generation of the growth inhibitory effects of IFN-␣ on leukemic progenitors in clonogenic assays in methylcellulose. Bone marrow samples from four CML patients were obtained and studied. IFN-␣ inhibited colony formation of myeloid progenitors isolated from the bone marrow samples of all patients with CML (Fig. 11, A-D). Concomitant treatment of cells with SB203580 reversed the growth inhibitory effects of IFN-␣ on the leukemic progenitors (Fig. 11, A-D). In a similar manner, SB202190 also reversed the inhibitory effects of IFN-␣ on leukemic progenitor colony formation, whereas treatment with PD098059 had no effects (Fig. 11, A and B). Thus, pharmacological inhibition of the p38 MAPK pathway abrogates the growth inhibitory effects of IFN-␣ on clonogenic hematopoietic progenitor growth in the bone marrow samples of CML patients, indicating that the function of p38 is essential for the generation of the anti-leukemic effects of IFN-␣. DISCUSSION Despite the well documented clinical effects of IFN-␣ in CML, very little is known regarding the mechanisms of IFN signaling in cells expressing BCR-ABL. In this study, we provide the first evidence that IFN-␣ activates the p38 MAPK pathway in CML-derived IFN-sensitive cells. Most importantly, our data provide direct evidence that this signaling cascade plays a critical role in the induction of the anti-leukemic activities of IFN-␣ in BCR-ABL-expressing cells. This is revealed by the finding that pharmacological inhibition of p38 reverses the biological effects of IFN-␣ in an IFN-␣-sensitive CML line and primary leukemic bone marrow-derived myeloid progenitors. The SB203580 and SB202190 inhibitors of the p38 MAPK pathway used in our studies have been previously shown to exhibit specificity for p38 MAPK. They act by binding to the ATP site and by inhibiting the kinase activity of p38 MAPK, and the basis for their selectivity has been determined by mutagenesis and x-ray crystallographic structures of p38inhibitor complexes (33)(34)(35). Both SB203580 and SB202190 have similar target specificities; and in addition to inhibiting p38 (also called p38␣), they inhibit the p38␤2 isoform (but not the p38␥ and p38␦ isoforms) of the same family (36 -39). Thus, our data demonstrating reversal of the growth inhibitory effects of IFN-␣ by treatment with these pyridinyl imidazole compounds provide strong evidence for an important role of p38 (p38␣) and possibly p38␤2 in the induction of the growth inhibitory effects of IFN-␣ in CML.
Our data clearly establish that the small GTPase Rac1, which functions as an upstream effector of p38 in various systems, including the IFN system (19, 28 -30), is rapidly activated during IFN-␣ treatment of the IFN-sensitive KT-1 CML cell line, but not the IFN-resistant K562 cell line. This activation of Rac1 in KT-1 cells is inhibited by the tyrosine kinase inhibitor genistein, but not the phosphatidylinositol 3Ј-kinase inhibitor wortmannin. Thus, the function of a tyrosine kinase(s) is required for Rac1 activation in KT-1 cells, but such activation does not require upstream engagement of the IFN-␣-dependent insulin receptor substrate/phosphatidylinositol 3Ј-kinase pathway (16,21,24). It is therefore likely that acti- FIG. 10. IFN-␣-dependent activation of p38 in isolated granulocytes from the peripheral blood of patients with CML. Granulocytes isolated from the peripheral blood of four different patients with CML (A, C, E, and G) were treated with IFN-␣ for the indicated times (min). Total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody against the phosphorylated form of p38 MAPK to detect activated p38 MAPK (A, C, E, and G). The blots shown in A, C, E, and G were subsequently stripped, and each of them was reprobed with an antibody against p38 MAPK to control for protein loading (B, D, F, and H, respectively).

p38 MAPK Pathway Mediates Growth Inhibitory Effects of IFN-␣
vation of the TYK2 and/or JAK1 kinase, both of which are associated with the type I IFN receptor, regulates downstream activation of Rac1 and p38. Such a regulation should involve an intermediate protein substrate that links JAK kinase activation to Rac1, and the identity of such a Rac1 regulator remains to be established.
In previous studies, we have established that the function of p38 is required for transcriptional regulation of IFN-sensitive genes that express IFN-stimulated response elements or IFN-␥-activated sites in their promoters (17,19), including the PML gene (19), which mediates IFN-regulated growth inhibitory responses (41). It is possible that the requirement for p38 MAPK in the induction of the growth inhibitory effects of IFN-␣ is mediated by up-regulation of expression of this gene and possibly other related genes with tumor suppressor activity. However, activation of the p38 MAPK pathway by IFN-␣ may have additional effects that mediate growth inhibitory responses such as regulation of signals that modify cell cycle progression in CML cells. Our data indicate that the MAPK-APK-2 and MAPKAPK-3 kinases are downstream effectors for p38, activated by IFN-␣ in BCR-ABL-expressing cells. It is therefore possible that MAPKAPK-2 and MAPKAPK-3 play important roles in the induction of the anti-leukemic effects of IFN-␣ in CML cells either via regulation of gene transcription or via engagement of other downstream effectors that regulate cell cycle progression. Regardless of the precise mechanisms involved, p38 appears to play a critical role in the generation of the antiproliferative effects of IFN-␣ on leukemic progenitors as well as on normal bone marrow erythroid and myeloid progenitors. 2 Thus, the p38 MAPK pathway may participate in the generation of both the anti-leukemic effects of IFN-␣ as well as the documented hematologic toxicity that this cytokine exhibits when administered to humans (7,8).
It remains to be determined whether the p38 pathway acts in cooperation with other IFN-activated pathways such as the STAT pathway to regulate induction of growth inhibition by IFN-␣ in CML cells. Interestingly, it was recently demonstrated that IFN-␣ treatment induces formation of STAT5-CrkL complexes in KT-1 cells and that such complexes bind to one of the IFN-␥-activated sites of the promyelocyte leukemia gene promoter (27). The promoter of the PML gene also contains IFN-stimulated response elements regulated by the IFNstimulated gene factor-3 complex, which involves STAT2-STAT1 complexes. The fact that the p38 MAPK pathway is required for transcriptional regulation of the promoter of this gene (19), which is also regulated by different STAT-binding complexes, suggests a coordination of signaling functions between the p38 MAPK and STAT pathways. Such a coordination may be important in the induction of the direct anti-leukemic effects of IFN-␣ in CML cells.
It has been previously shown that IFN-␣ down-regulates expression of the bcr-abl oncogene, which causes the malignant transformation (42). Our data suggest the existence of a direct mechanism, distinct from inhibition of BCR-ABL expression, that mediates the antiproliferative effects of IFN-␣. It is possible that the p38 MAPK pathway also contributes to the induction of the anti-leukemic effects of other pharmacological agents such as STI571 that exhibit selective growth inhibitory effects in CML cells (40). A potential scenario may be that BCR-ABL exhibits constitutive negative regulatory effects on the activation of the growth inhibitory p38 MAPK pathway. Activation of the IFN-␣-dependent tyrosine kinases or inhibition of BCR-ABL kinase activity by STI571 may overcome such negative regulatory effects of BCR-ABL on p38 MAPK and 2 A. Verma and L. C. Platanias, manuscript in preparation. result in inhibition of malignant cell growth. Future studies in that direction are warranted and may provide further insights into the mechanisms of BCR-ABL-mediated leukemogenesis as well as clarification of the mechanisms by which drugs and cytokines block BCR-ABL-mediated cell growth.