Induction of apoptosis in human leukemia cells by the ... - Nature

Oncogene (2004) 23, 1364–1376

& 2004 Nature Publishing Group All rights reserved 0950-9232/04 $25.00 www.nature.com/onc

Induction of apoptosis in human leukemia cells by the tyrosine kinase inhibitor adaphostin proceeds through a RAF-1/MEK/ERK- and AKT-dependent process Chunrong Yu1, Mohamed Rahmani1, Jorge Almenara1, Edward A Sausville2, Paul Dent3 and Steven Grant*,1 1

Department of Medicine, Virginia Commonwealth University, Medical College of Virginia, Richmond, VA 23298, USA; Developmental Therapeutics Program, National Cancer Institute, Bethesda, MD, USA; 3Department of Radiation Oncology, Virginia Commonwealth University, Medical College of Virginia, Richmond, VA 23298, USA 2

Effects of the tyrphostin tyrosine kinase inhibitor adaphostin (NSC 680410) have been examined in human leukemia cells (Jurkat, U937) in relation to mitochondrial events, apoptosis, and perturbations in signaling and cell cycle regulatory events. Exposure of cells to adaphostin concentrations X0.75 lM for intervals X6 h resulted in a pronounced release of cytochrome c and AIF, activation of caspase-9, -8, and -3, and apoptosis. These events were accompanied by the caspase-independent downregulation of Raf-1, inactivation of MEK1/2, ERK, Akt, p70S6K, dephosphorylation of GSK-3, and activation of c-Jun-Nterminal kinase (JNK) and p38 MAPK. Adaphostin also induced cleavage and dephosphorylation of pRb on CDK2- and CDK4-specific sites, as well as the caspasedependent downregulation of cyclin D1. Inducible expression of a constitutively active MEK1 construct markedly diminished adaphostin-induced cytochrome c and AIF release, JNK activation, and apoptosis in Jurkat cells. Ectopic expression of Raf-1 or constitutively activated (myristolated) Akt also significantly attenuated adaphostin-induced apoptosis, but protection was less than that conferred by enforced activation of MEK. Lastly, antioxidants (e.g., L-N-acetylcysteine; L-NAC) opposed adaphostin-mediated mitochondrial dysfunction, Raf-1/ MEK/ERK downregulation, JNK activation, and apoptosis. However, in contrast to L-NAC, enforced activation of MEK failed to block adaphostin-mediated ROS generation. Together, these findings demonstrate that the tyrphostin adaphostin induces multiple perturbations in signal transduction pathways in human leukemia cells, particularly inactivation of the cytoprotective Raf-1/ MEK/ERK and Akt cascades, that culminate in mitochondrial injury, caspase activation, and apoptosis. They also suggest that adaphostin-related oxidative stress acts upstream of perturbations in these signaling pathways to trigger the cell death process. *Correspondence: S Grant, Division of Hematology/Oncology, Virginia Commonwealth University/Medical College of Virginia, MCV Station Box 230, Richmond, VA 23298, USA; E-mail: [email protected] Received 2 September 2003; revised 22 September 2003; accepted 23 September 2003

Oncogene (2004) 23, 1364–1376. doi:10.1038/sj.onc.1207248 Published online 1 December 2003 Keywords: leukemia; ERK; Akt

apoptosis;

adaphostin;

ROS;

Introduction The tyrosine kinase inhibitor adaphostin is a member of the tyrphostin family, a group of small molecules that interfere with peptide binding rather than the kinase ATP-binding site (Levitzki and Gazit, 1995). The tyrphostin AG957 has been under examination as an alternative to STI571 (imatinib mesylate) as an inhibitor of the Bcr/Abl kinase, implicated in the pathogenesis of CML (Kaur et al., 1994). Adaphostin (NSC 680410) is an adamantly ester of AG957 that is more potent than AG957 in vitro and in vivo, and which is currently undergoing preclinical development. Recently, adaphostin has been shown to induce apoptosis more rapidly than imatinib mesylate in Bcr/Abl þ cells in conjunction with Bcr/Abl downregulation as well as STAT5 inactivation, and to trigger cell death in imatinib mesylate-resistant cells (Svingen et al., 2000). Subsequently, adaphostin was reported to induce apoptosis in Bcr/Abl human leukemia lines (e.g., Jurkat, U937), as well as in glioblastoma cells (Avramis et al., 2002). In a very recent communication, Chandra et al. (2003) observed that adaphostin induced apoptosis in various human leukemia cell lines in association with generation of reactive oxygen species. Together, these findings suggest a possible therapeutic role for adaphostin in CML and potentially other leukemias. The mitogen-activated protein kinase (MAPK) pathways consist of three parallel serine–threonine kinase modules involved in the regulation of diverse cellular events, including proliferation, differentiation, and apoptosis, among many others (Chang and Karin, 2001). These consist of the extracellular signal-regulated

Induction of apoptosis in human leukemia cells by tyrosine kinase inhibitor C Yu et al

1365

protein kinase (ERK1/2), the c-Jun-N-terminal kinase (JNK), and the p38 MAPK, along with multiple corresponding activating kinases (e.g., MAPK kinase, MAPK kinase kinase) (Pearson et al., 2001). Although exceptions exist, in most systems, the ERK module plays a cytoprotective role, whereas the JNK and p38 MAPK cascades exert proapoptotic functions (Xia et al., 1995). Links between tyrosine kinases and cytoprotective serine–threonine pathways have been most extensively characterized for growth factor receptors (e.g., EGFR), which transduce signals (via Ras and Raf) to the MAPK kinase (MEK)/ERK1/2 pathway, as well as (via phospholipase C) to the PI3K/Akt survival pathway (Wennstrom and Downward, 1999). It is assumed that interruption of such pathways plays a major role in the lethal effects of tyrosine kinase inhibitors, although which of the downstream targets is most critical for survival remains to be fully elucidated, and may in fact vary with the cell type. Aside from its ability to inhibit the Bcr/Abl kinase, little is known about the effects of adaphostin on other signaling pathways, particularly the MAPK and related modules, in Bcr/Abl leukemia cells. Moreover, the functional relationship between such events, mitochondrial dysfunction, and apoptosis induction by adaphostin has not yet been examined. To address these issues, the effects of adaphostin on various signaling and cell cycle regulatory pathways has been examined in Bcr/Abl human leukemia cells (Jurkat and U937) in relation to mitochondrial injury, caspase activation, and apoptosis. Here we report that in such cells, adaphostin induces multiple perturbations in stress-, survival-, and cell cycle-associated proteins, and that interruption of the Raf-1/MEK/ERK1/2 and Akt pathways appear to play important functional roles in regulating lethality. Evidence is also presented suggesting that adaphostin-mediated oxidative stress operates upstream of these signaling perturbations and mitochondrial injury.

Figure 1 Dose-dependent effects of adaphostin on apoptosis in Jurkat and U937 cells. Jurkat cells (a) or U937 cells (b) were exposed to the designated concentration of adaphostin for 6, 9, or 24 h, after which the extent of apoptosis was determined by monitoring Wright–Giemsa-stained specimens as described in Materials and methods. Values represent the means for three separate experiments performed in triplicate 7s.d.

Results Adaphostin induces apoptosis in Jurkat and U937 leukemia cells in a dose-dependent manner

Adaphostin triggers the dose-dependent release of cytochrome c and AIF, accompanied by caspase activation and PARP cleavage, in human leukemia cells

In initial studies, Jurkat cells were exposed to increasing concentrations of adaphostin for 6, 9, or 24 h, after which apoptosis was monitored. As shown in Figure 1, exposure of Jurkat cells to adaphostin resulted in a timeand dose-dependent induction of cell death that could first be detected after 6 h of drug exposure for adaphostin concentrations X0.75 mM (Figure 1a). Longer (e.g., 24-h) exposures resulted in virtually 100% cell death for adaphostin concentrations X1.25 mM. Roughly similar results were noted when U937 cells were examined, although these cells were slightly less sensitive to adaphostin at the earlier intervals (Figure 1b).

Examination of the S-100 cytosolic cell fraction after 9 h of drug exposure revealed a dose-dependent release of cytochrome c and AIF (Figure 2). These events were associated with cleavage of procaspase-9, -3, and -8, accompanied by PARP degradation. However, none of the adaphostin concentrations tested induced changes in expression of Bcl-2, Bcl-xL, Mcl-1, or XIAP. Together, these findings indicate that submicromolar concentrations of adaphostin effectively and rapidly induce mitochondrial injury, caspase cleavage, and apoptosis in Bcr/Abl- Jurkat leukemia cells. Essentially equivalent findings were obtained when U937 cells were examined (data not shown). Oncogene


1366

a pronounced increase in phosphorylation of the stressrelated kinases JNK and p38 MAPK. The increase in JNK phosphorylation was associated with enhanced phosphorylation of the JNK substrate, c-JUN(ser63). No changes in levels of total ERK1/2 or JNK were observed, although modest declines in levels of total Akt were noted at the higher adaphostin concentrations. Adaphostin induces dephosphorylation of pRb on CDK2- and CDK4-dependent sites, pRb cleavage, and cyclin D1 downregulation in human leukemia cells Parallel effects of adaphostin on several cell cyclerelated proteins were also examined (Figure 3b). Adaphostin exposure resulted in underphosphorylation and cleavage of pRb. Furthermore, the reduction in pRb phosphorylation involved both CDK2- and CDK4specific phosphorylation sites. Adaphostin treatment also induced downregulation of cyclin D1, but had no effect on the phosphorylation status of p34cdc2. Thus, adaphostin-induced apoptosis in Bcr/Abl- Jurkat leukemia cells was associated with multiple perturbations in signaling and cell cycle regulatory pathways, including inactivation of the cytoprotective Raf-1/MEK/ERK1/2 and Akt pathways, activation of the stress-related JNK and p38 MAPK pathways, and dephosphorylation/ cleavage of pRb accompanied by downregulation of cyclin D1. Finally, essentially equivalent results were obtained when effects of adaphostin were examined in U937 cells (Figure 4).

Figure 2 Effect of adaphostin on mitochondrial injury, caspase activation, and expression of Bcl-2 family members in Jurkat cells. Logarithmically growing Jurkat cells were exposed to the designated concentration of adaphostin for 9 h, after which proteins were extracted from whole-cell lysates, separated by SDS–PAGE, and probed with antibodies to cleaved caspase-9, cleaved caspase-3, procaspase-8, PARP, Bcl-2, Bcl-xL, Mcl-1, or XIAP as described in Materials and methods. Alternatively, S-100 cytosolic fractions were obtained as described in Materials and methods, and expression of cytochrome c or AIF monitored by Western analysis as above. CF ¼ cleavage fragment. In each case, lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results

Adaphostin induces downregulation/inactivation of Raf-1, MEK1/2, ERK1/2, Akt, p70S6K, dephosphorylation of GSK-3, and activation of JNK and p38 MAPK, in human leukemia cells The effects of adaphostin were then examined in relation to expression of various signaling and cell cycle regulatory proteins in Jurkat cells. Exposure of cells to increasing concentrations of adaphostin for 6 h resulted in a dose-dependent reduction in expression of Raf-1, and diminished phosphorylation of MEK1/2, ERK1/2, Akt, p70S6K, and GSK-3 (Figure 3a). These events were accompanied by a slight cleavage of MEK1/2, and Oncogene

Caspase inhibition blocks adaphostin-induced procaspase3, -8, and bid processing, but not cytochrome c or aif release or procaspase-9 activation The relationship between caspase activation and mitochondrial injury in adaphostin-treated Jurkat and U937 cells was then examined. In both cell lines, coadministration of Boc-D-fmk with 2.0 mM adaphostin failed to block cytochrome c or AIF release into the cytosolic S100 fraction, or cleavage (activation) of procaspase-9 (Figure 5a and b). In contrast, Boc-D-fmk abrogated adaphostin-induced cleavage of procaspase-3, -8, and Bid. These findings indicate that adaphostin-mediated induction of mitochondrial dysfunction operates independently or upstream of the caspase cascade. Adaphostin-mediated downregulation/inactivation of Raf1, MEK1/2, ERK1/2, p70s6k, Akt, dephosphorylation of GSK-3, activation of JNK and p38 MAPK, and dephosphorylation of pRb are caspase-independent events To determine what role caspase activation might play in adaphostin-mediated perturbations in signaling and cell cycle pathways, Jurkat cells were exposed to 2.0 mM adaphostin in the presence or absence of Boc-D-fmk as above, and Western analysis performed. As shown in Figure 6a and b, caspase inhibition failed to modify adaphostin-mediated downregulation of Raf-1, dephosphorylation of MEK1/2, ERK1/2, p70S6K, Akt, and GSK-3, activation of JNK and p38 MAPK, and


1367

dephosphorylation of pRb on CDK2- and CDK4specific sites. Boc-D-fmk did, however, block cleavage of MEK1/2 and underphosphorylation of pRb as well as downregulation of cyclin D1. Essentially equivalent

results were obtained in U937 cells (data not shown). These results indicate that in human leukemia cells, perturbations in signaling and cell cycle proteins induced by adaphostin in human may be both caspasedependent and independent, but that for the most part, modulation of signaling cascades proceeds in a caspaseindependent manner. Enforced activation of MEK1/ERK1/2 substantially blocks adaphostin-mediated mitochondrial injury, JNK activation, and apoptosis in human leukemia cells Previous studies have indicated that the Raf-1/MEK/ ERK1/2 pathway plays a largely cytoprotective role (Bonni et al., 1999). Consequently, the possibility that adaphostin-mediated downregulation of the Raf/MEK/ ERK cascade might contribute to lethality appeared plausible. To test this possibility, a Jurkat cell line (MEK-30) that inducibly expresses a constitutively active MEK1 construct under the control of a doxycyline-responsive promoter was employed. As shown in Figure 7a, exposure to adaphostin in the absence of doxycycline resulted in apoptosis in nearly 75% of cells, whereas apoptosis was essentially abrogated in the presence of doxycycline. Western analysis revealed that addition of doxycycline to the medium resulted in a substantial increase in expression of phospho-ERK1/2 in both control and adaphostin-treated cells, whereas total ERK1/2 levels were unchanged (Figure 7b). Interestingly, enforced activation of ERK1/2 in adaphostin-treated cells was accompanied by diminished JNK activation, suggesting the possibility of crosstalk between these pathways. Enforced activation of ERK1/2 also diminished adaphostin-mediated caspase3 activation as well as cytochrome c and AIF release. Identical results were obtained when a second MEK1inducible clone (MEK-28) was employed (data not shown). Together, these findings support the notion that inactivation of the MEK/ERK1/2 pathway by adaphostin promotes mitochondrial injury, caspase activation, and apoptosis in Jurkat cells. They also raise the possibility that such an event may promote JNK activation, thereby amplifying the proapoptotic cascade.

Figure 3 (a) Effect of adaphostin on activation of signaling pathways in Jurkat cells. Jurkat cells were exposed to the indicated concentration of adaphostin for 6 h, after which protein lysates were obtained, separated by SDS–PAGE, and probed with the designated antibodies as described in Materials and methods. (b) Effect of adaphostin on cell cycle-related proteins in Jurkat cells. Cells were treated as above, and Western analysis employed to monitor expression of underphosphorylated pRb, CDK2- and CDK4-specific pRb phosphorylation, cyclin D1 levels, and expression of phospho-p34cdc2 (p-cdc2). CF ¼ cleavage fragment. For both (a) and (b), lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results Oncogene


1368

Enforced activation of Raf-1 and Akt partially protect Jurkat cells from adaphostin-mediated lethality A similar analysis was performed in Jurkat cells inducibly expressing Raf-1 or constitutively active Akt. Culture of Raf-1 cells with doxycycline resulted in a clear increase in expression of flag-tagged Raf-1, and a modest but discernible increase in ERK activation (Figure 8a). The extent of ERK activation in Raf-1induced cells was clearly less pronounced than that observed in cells inducibly expressing constitutively active MEK1 (Figure 8). Nevertheless, enforced expression of Raf-1 resulted in a partial, but significant reduction in adaphostin lethality (Po0.02 versus cells cultured in the absence of doxycycline). Comparable results were obtained with an additional Raf-1 clone (Raf-19; data not shown). Similar results were observed in Jurkat cells inducibly expressing myristolated Akt. Thus, culture of these cells in the presence of doxycycline resulted in a marked increase in a Myc tag, expression of phospho-Akt, and a partial but statistically significant reduction in adaphostin-induced apoptosis (Po0.02 versus controls; Figure 9b). Similar results were obtained using two additional Akt clones (Akt-2 and Akt-28; data not shown). Together, these results suggest that adaphostin-mediated downregulation of Raf and inactivation of Akt play significant functional roles in apoptosis induced by this agent. Antioxidants block adaphostin-mediated mitochondrial injury, Raf/MEK/ERK and Akt inactivation, JNK and p38 MAPK activation, and apoptosis in Jurkat cells

Figure 4 Effect of adaphostin on signaling and cell cycleregulatory proteins in U937 cells. U937 cells were exposed to the indicated concentration of adaphostin for 6 h, after which expression of total (a) and phosphorylated signaling proteins or (b) cell cycle-related proteins monitored by Western analysis as described in Materials and methods. CF ¼ cleavage fragment. For both (a) and (b), lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results

Oncogene

Finally, an attempt was made to characterize the hierarchy that might exist between adaphostin-mediated perturbations in redox state and adaphostin-induced mitochondrial dysfunction, caspase activation, and perturbations in pro- and antiapoptotic signaling cascades. To this end, Jurkat cells were exposed to adaphostin (2.0 mM) for 9 h in the presence or absence of the antioxidants L-N-acetylcysteine (L-NAC) (Moldeus and Cotgreave, 1994), Tiron (1 mM) (Wolf et al., 2001) or Trolox (1 mM) (Sparrow et al., 2003). As shown in Table 1, coadministration of L-NAC or Trolox, with adaphostin essentially abrogated adaphostin-mediated induction of the morphologic features of apoptosis, annexin V/PI positivity, and mitochondrial injury (i.e., loss of DCm). The protective effects of Tiron were not quite as great as those of L-NAC or Trolox, but this agent nevertheless significantly attenuated adaphostin lethality compared to controls (Po0.005 versus controls in each case). These results are in accord with recent findings by Chandra et al. (2003). Significantly, coexposure of Jurkat cells to adaphostin and L-NAC abrogated cytochrome c and AIF release, cleavage of caspases-9, -3, and -8, Bid degradation, and apoptosis (Figure 9a). Interestingly L-NAC also blocked adaphostin-mediated Raf-1 downregulation, inactivation and cleavage of MEK1/2, dephosphorylation of ERK1/2, p70S6K, Akt, and GSK-3, and activation of JNK and p38 MAPK, findings compatible with the notion that


1369

these events represent downstream consequences of oxidative stress. Finally, L-NAC blocked adaphostinmediated pRb desphosphorylation and cleavage, diminished pRb phosphorylation on CDK2- and CDK4-

specific sites, and cyclin D1 downregulation. Similar protective effects were obtained with U937 (data not shown). These findings suggest that adaphostin-related oxidative stress acts proximally to induce mitochondrial injury and multiple perturbations in signaling and cell cycle pathways that culminate in leukemic cell death. The cytoprotective effects of ERK1/2 activation operate downstream of adaphostin-induced ROS generation Lastly, the hierarchy of adaphostin-induced ROS generation and perturbations in MEK/ERK activation was investigated. As shown in Figure 10a, coadministration of 10 mM L-NAC very significantly reduced the generation of ROS in Jurkat cells exposed to 1.25 mM adaphostin for 2 h (Po0.01 versus control). In marked contrast, enforced expression of activated MEK1/ ERK1/2, despite largely abrogating adaphostin lethality, had no effect on adaphostin-mediated ROS generation (Figure 10b). This finding suggests that the cytoprotective MEK/ERK1/2 pathway acts to attenuate the lethal consequences of adaphostin-mediated ROS generation, rather than by preventing oxidative injury.

Discussion The tyrphostin AG957 was originally developed as an inhibitor of the Bcr/Abl kinase, and has been shown to induce mitochondrial injury, Bcr/Abl downregulation, and apoptosis in Bcr/Abl kinase þ K562 leukemic cells (Kaur et al., 1994; Svingen et al., 2000). Subsequently, it was shown that adaphostin, an adamantyl ester of AG957, was even more effective than AG957 in downregulating Bcr/Abl, and in addition, effectively induced apoptosis in cells resistant to imatinib mesylate (Chandra et al., 2003). Such findings confirm earlier observations that the Bcr/Abl kinase not only provides a proliferative advantage to hematopoietic cells, but that inhibition of this kinase can trigger the cell death pathway (Dan et al., 1998). While imatinib mesylate is a highly potent inhibitor of Bcr/Abl, it also inhibits other kinases such as c-kit, resulting in antiproliferative effects in tumor cells driven by this oncogene (Krystal et al., 2000). In this context, it is not surprising that adaphostin might also interrupt additional survival pathways, and in so doing, promote apoptosis in human leukemia cells that lack the Bcr/Abl mutation. Consistent with this concept, adaphostin has been shown to Figure 5 Effect of caspase inhibition on adaphostin-mediated mitochondrial dysfunction and caspase activation. Jurkat (a) or U937 cells (b) were exposed to 2.0 mM adaphostin in the presence or absence of 20 mM Boc-D-fmk for 6 h after which either whole-cell lysates or S-100 cytosolic fractions were isolated and subjected to Western analysis to monitor cytochrome c and AIF release (S-100 fractions) and cleavage of caspase-9, -3, -8, and Bid (whole-cell lysates). CF ¼ cleavage fragment. For both (a) and (b), lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results Oncogene


1370

trigger apoptosis in several Bcr/Abl human leukemia cell types (Avramis et al., 2002). Moreover, very recent results suggest that adaphostin induces apoptosis

through an oxidative stress-related mechanism (Chandra et al., 2003). However, the possibility that adaphostin might also induce perturbations in signaling pathways that are capable of modulating mitochondrial responses to these events has not yet been examined. The present findings indicate that relatively low (e.g., submicromolar) concentrations of adaphostin potently induce mitochondrial dysfunction (e.g., cytochrome c and AIF release, loss of DCm) in Bcr/Abl human leukemia cells through an antioxidant-sensitive process. Furthermore, they also demonstrate that adaphostin inactivates several cytoprotective serine/threonine kinase pathways (e.g., Raf-1/MEK/ERK1/2 and Akt) that play significant functional roles in regulating druginduced lethality. There is abundant evidence suggesting a link between constitutive activation of the Raf-1/MEK/ERK1/2 axis and leukemogenesis (Chang et al., 2003; Platanias, 2003), and such evidence has prompted the development of MEK1/2 inhibitors such as PD184352 as antileukemic agents (Sebolt-Leopold et al., 1999; Milella et al., 2001). In most, but not all systems studied, inactivation of MEK1/2 and ERK1/2 is associated with promotion of cell death (Xia et al., 1995). The mechanism by which this occurs is not known with certainty, but may be related to perturbations in downstream ERK1/2 targets, including the Elk and CREB family of transcription factors (Bonni et al., 1999). Alternatively, inactivation of ERK1/2 may prevent phosphorylation of Bad, and in so doing, preserve its proapoptotic capacity (Scheid et al., 1999). Very recently, direct antiapoptotic actions of ERK1/2 related to phosphorylation/inactivation of caspase-9 have been described (Allan et al., 2003). It is therefore noteworthy that exposure of leukemic cells to adaphostin at concentrations that triggered apoptosis was associated with the clear inactivation of MEK1/2 and ERK1/2. However, in contrast to the actions of pharmacologic MEK inhibitors, adaphostin also induced Raf-1 downregulation, suggesting a more proximal site of action for this kinase inhibitor. In this regard, the pattern of adaphostin-induced Raf-1 downregulation and MEK/ERK inactivation is reminiscent of the effects of certain agents (e.g., 17-AAG), which act by interfering with the chaperone function of Hsp90, leading to degradation of client proteins, including Raf-1 (Clarke et al., 2000). Whether adaphostin does in fact disrupt Hsp90 function, or acts by inhibiting tyrosine kinases upstream of the Raf-1/MEK/ERK1/2 Figure 6 Effect of caspase inhibition on adaphostin-mediated perturbations in expression of signaling and cell cycle-related proteins in Jurkat cells. (a) Jurkat cells were exposed to 2.0 mM adaphostin for 9 h in the presence or absence of 20 mM Boc-D-fmk, after which expression of total or phosphorylated signaling proteins was monitored by Western analysis as described in Materials and methods. CF ¼ cleavage fragment. (b) Cells were treated as in (a), after which expression of underphosphorylated pRb, phosphorylation of pRb on CDK2- and CDK4-specific sites, and levels of cyclin D1 were monitored as above. For both (a) and (b), lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results

Oncogene


1371

Figure 7 Effect of enforced activation of MEK/ERK1/2 on adaphostin-mediated lethality. (a) Jurkat cells inducibly expressing a constitutively active MEK1 construct (MEK-30) were incubated in medium alone (Dox) or medium containing 1 mM doxycycline (Dox þ ) for 24 h, after which they were exposed to 1.25 mM adaphostin for an additional 18 h. At the end of the incubation period, the extent of apoptosis was monitored by morphologic assessment or annexin PI analysis by flow cytometry as described in Materials and methods. Values represent the means for three separate experiments performed in triplicate 7s.d. (b) Cells were treated as in (a), after which Western analysis was employed to monitor expression of total and phosphorylated ERK1/2, phospho-JNK, and caspase-3 in total cellular extracts. Alternatively, S100 fractions were isolated as described in Materials and methods, and monitored for expression of cytochrome c and AIF. Lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results

Figure 8 Effects of enforced induction of Raf-1 or activation of Akt on adaphostin lethality in Jurkat cells. Jurkat cells inducibly expressing Raf-1 (a) or constitutively active (myristolated) Akt (b) were incubated in medium alone (Dox) or medium containing 1 mM doxycycline (Dox þ ) for 24 h, exposed to 1.25 mM (a) or 1.0 mM (b) adaphostin for 18 h, after which apoptosis was evaluated by morphologic assessment or annexin V/PI staining by flow cytometry as described in Materials and methods. Values represent the means for three separate experiments performed in triplicate 7s.d. Western blots demonstrating the effects of doxycycline on expression of a Flag-Raf-1 tag and phospho-ERK1/2 (a) or a cMyc Akt tag and phospho-Akt (b) are shown below, along with tubulin controls. *Significantly less than values obtained for cells incubated in medium lacking doxycycline; Po0.02 Oncogene


1372

pathway, remains to be determined. Significantly, constitutive activation of MEK1 largely abrogated adaphostin-mediated lethality, while enforced expression of Raf-1 also protected cells, although to a lesser extent. Collectively, these findings indicate that disruption of the cytoprotective Raf-1/MEK/ERK1/2 cascade by adaphostin plays an important functional role in mediating adaphosin-related antileukemic actions. Exposure of Jurkat and U937 cells to adaphostin also inactivated the cytoprotective serine–threonine kinase Akt, and two of its downstream targets (e.g., p70S6K and GSK-3). The prosurvival function of Akt is well documented, and may involve multiple mechanisms, including phosphorylation of Bad, caspase-9, and inactivation of the forkhead family of transcription

factors, among multiple others (Datta et al., 1999; Franke et al., 1997; Scheid and Woodgett, 2001). Links between receptor tyrosine kinases and the Akt pathway, generally mediated through PI3K and phospholipase C, are well documented (Wu et al., 2001). Whether adaphostin inactivates Akt through tyrosine kinase inhibition or more proximally (e.g., at the level of PI3K) remains to be established. In any case, ectopic expression of constitutively active Akt significantly, although partially, prevented adaphostin-mediated lethality, indicating that disruption of Akt signaling plays an important role in the antileukemic actions of this agent. A corollary of this finding is that Akt inhibitors, which are currently the subject of considerable interest (Hu et al., 2000), might be expected to

Figure 9 Effects of L-NAC on adaphostin-induced apoptosis and perturbations in signaling and cell cycle regulatory proteins. (a) Jurkat cells were exposed to 2.0 mM adaphostin for 9 h in the presence or absence of 10 mM L-NAC, after which cells were lysed, total proteins separated by SDS–PAGE, and cleavage of procaspase-9, -3, -8, and Bid monitored by Western analysis as described in Materials and methods. Alternatively, release of cytochrome c or AIF into the cytosolic S-100 fraction was monitored as described above. (b) Cells were treated as in (a), after which expression of Raf-1, phospho-MEK1/2 and total MEK1/2, phospho-ERK1/2 and total ERK1/2, phospho-JNK, phospho-p38 MAPK, phospho-Akt and total Akt, and phospho-GSK3 was monitored by Western analysis as above. (c) Cells were treated as in (a), after which expression of underphosphorylated pRb, CDK2- and CDK4-specific phosphorylated pRb, and cyclin D1 were monitored as above. For these studies, lanes were loaded with 20 mg of protein; blots were stripped and reprobed with antibodies to tubulin to ensure equivalent loading and transfer. Results of a representative experiment are shown; two additional studies yielded equivalent results

Table 1 Jurkat cells were incubated with 2.0 mM adaphostin for 9 h in the presence or absence of 10 mM L-NAC, 1.0 mM Trolox, or 1.0 mM Tiron, after which apoptosis was assayed by morphologic assessment of Wright–Giemsa-stainted specimens or annexin V/PI uptake by flow cytometry Jurkat cells Control AD-2.0 mM LNAC-10 mM LANC+AD Trolox-1.0 mM Trolox+LNAC Tiron-1.0 mM Tiron+AD

% Apoptotic cells

% Control (annexin V/PI)

% Control (DiOC6)

0.870.3 78.675.4 2.571.6 3.871.8 1.170.8 4.571.9 1.270.8 35.273.8

4.872.4 75.476.1 5.873.1 8.572.1 6.872.4 10.272.9 8.972.3 38.574.1

4.371.8 80.375.8 5.672.5 7.172.8 6.371.8 8.972.5 8.672.1 37.875.3

Alternatively, the percentage of cells exhibiting a reduction in DCm was monitored by flow cytometric analysis of DiOC6 uptake. Values represent the means for three separate experiments performed in triplicate 7s.d. Oncogene


1373

Figure 10 Effects of L-NAC and enforced activation of MEK1/ ERK1/2 on adaphostin-mediated ROS generation. (a) Jurkat cells were exposed to 1.25 mM adaphostin710 mM L-NAC for 2 h after which ROS generation was determined by monitoring CMH2DCFDA uptake by flow cytometry as described in Materials and methods. (b) Jurkat cells inducibly expressing constitutively active MEK in the presence of doxycycline were exposed to adaphostin as above in the presence (Dox þ ) or absence (Dox) of doxycyline, after which ROS production was determined as outlined above. For (a) and (b), values represent the means for three separate experiments performed in triplicate 7s.d. *Significantly less than values for cells exposed to adaphostin in the absence of L-NAC; Po0.01

enhance the antileukemic activity of adaphostin and related agents. Induction of apoptosis in leukemia cells by adaphostin was accompanied by activation of the stress-related kinase JNK, which has been linked to proapoptotic actions in multiple systems (Johnson and Lapadat, 2002). The mechanism by which JNK activation promotes apoptosis is not clear, but may be related to enhanced release of cytochrome c (Tournier et al., 2000), interactions with BH3-only members of the Bcl-2 family (Lei and Davis, 2003) or phosphorylation/inactivation of antiapoptotic proteins (e.g., Bcl-2 and Mcl-1) (Yamamoto et al., 1999; Inoshita et al., 2002). In many settings, the decision of a cell to undergo apoptosis is determined by the net outputs of the cytoprotective Raf1/MEK/ERK1/2 versus the stress-related SEK/JNK

pathways (Xia et al., 1995). In this context, the relationship between adaphostin lethality and its capacity to downregulate/inactivate Raf, MEK1/2, and ERK1/2 while reciprocally activating JNK and p38 MAPK can be better understood. It is worth noting that enforced activation of MEK/ERK had the net effect of attenuating adaphostin-mediated JNK phosphorylation. Crosstalk between the cytoprotective MEK/ ERK1/2 and stress-related JNK pathways has previously been described (Jarvis et al., 1997). Such findings raise the possibility that adaphostin-mediated downregulation of the Raf/MEK/ERK amplifies the apoptotic process by triggering a reciprocal activation of the proapoptotic JNK cascade. Whether this occurs through modulation of the activity of MAP kinase phosphatases (MKPs) (Muda et al., 1996) or other mechanisms is unknown, but is currently the subject of ongoing investigation. Adaphostin induced several changes in cell cyclerelated proteins including dephosphorylation and cleavage of pRb and downregulation of cyclin D1, events previously associated with apoptosis in other systems (Carlson et al., 1999; Chau et al., 2002). While pRb degradation and cyclin D1 downregulation were in large part caspase-dependent, dephosphorylation of pRb on CDK2- and CDK4-specific sites was not. Such a finding suggests that in addition to its other functions, adaphostin may act as a CDK inhibitor, analogous to other pharmacologic kinase inhibitors known to induce apoptosis (e.g., flavopiridol) (Arguello et al., 1998). Thus, it is possible that the complex disruptions in the pRb axis induced by adaphostin (e.g., degradation of cyclin D1 and pRb, putative inhibition of CDK2/CDK4, and desphosphorylation of pRb) may contribute to the apoptotic process. The functional contributions, if any, of such cell cycle perturbations to adaphostin lethality remains to be established. The ability of free radical scavengers block adaphostin-induced apoptosis, as recently described (Chandra et al., 2003), in conjunction with attenuation of related perturbations in signaling pathways suggest that oxidative stress plays a critical and presumably proximal role in adaphostin lethality. Significantly, coadministration of L-NAC blocked not only adaphostin-mediated mitochondrial injury, caspase activation, and apoptosis, but also multiple perturbations in pro- and antiapoptotic signaling pathways, including Raf-1, MEK1/2, and Akt downregulation/inactivation, and JNK activation. The role of stress-related kinases in oxidative injuryrelated cell death is well known (Turner et al., 1998), and has been linked to dissociation of apoptosis signal regulating kinase-1 (ASK-1) from thioredoxin where it is then free to activate JNK (Tobiume et al., 2001). Conversely, the MEK/ERK1/2 module in general attenuates oxidative-stress-induced lethality, and interruption of this pathway (e.g., by pharmacologic MEK1/ 2 inhibitors) lowers the cell death threshold in the face of oxidant injury (Wang et al., 1998). Finally, Akt has been reported to protect cells from oxidative stress, possibly through disruption of the ASK-1/JNK axis (Kim et al., 2001). It is noteworthy that in contrast to the actions of Oncogene


1374 L-NAC,

enforced activation of MEK/ERK1/2 did not attenuate adaphostin-mediated ROS generation, suggesting that this cytoprotective pathway acts distal to oxidative stress signals. In view of these findings, it is tempting to speculate that adaphostin-mediated alterations in redox state may lead to reciprocal perturbations in stress-related (e.g., JNK, p38 MAP kinase) and cytoprotective (e.g., Raf-1/ MEK/ERK1/2, Akt) signaling pathways that render leukemic cells more vulnerable to mitochondrial injury and ultimately, cell death. Thus, a cell death amplification loop can be envisioned in which adaphostin-induced oxidant injury leads to the disabling of compensatory signaling pathways, thereby enhancing lethality. In summary, the present findings indicate that adaphostin effectively induces cell death in human leukemia cells lacking the Bcr/Abl oncogene through a process associated with multiple perturbations in oxidative stress, survival, and cell cycle regulatory pathways. Notably, adaphostin-mediated oxidant injury inactivates the Raf-1/MEK/ERK1/2 and Akt pathways, phenomena that play important functional roles in adaphostin-induced lethality. Further efforts to understand the mechanism(s) by which adaphostin and related tyrphostin tyrosine kinase inhibitors induce apoptosis in human leukemia cells could provide a more rational basis for attempts to incorporate such agents into antileukemic regimens. Accordingly, such efforts are currently underway in our laboratory.

Materials and methods Cells Jurkat and U937 cells were purchased from American Type Culture Collection (Rockville, MD, USA). Cells were cultured in RPMI 1640 supplemented with sodium pyruvate, MEM essential vitamins, L-glutamate, penicillin, streptomycin, and 10% heat-inactivated FCS (Hyclone, Logan, UT, USA). TetOn inducible constitutively active Akt and MEK1 Jurkat clones were maintained and utilized as previously described in detail (Jia et al., 2003; Yu et al., 2003a).

Reagents Adaphostin was kindly provided by Dr Edward Sausville (Division of Cancer Treatment, National Cancer Institute). BOC-fmk was purchased from Enzyme Products, Ltd. (Livermore, CA, USA). All were formulated in sterile DMSO before use. Annexin V/PI was supplied by BD PharMingen (San Diego, CA, USA) and formulated as per the manufacturer’s instructions. L-NAC, Trolox, and Tiron were purchased from Sigma (St Louis, MO, USA). 5- (and 6)chloromethyl-20 ,70 -dichlorodihydrofluorescein diacetate (CM-H2DCFDA) was purchased from Molecular Probes (Eugene, OR, USA). Experimental format Logarithmically growing cells were placed in sterile plastic T-flasks (Corning, Corning, NY, USA) to which were added the designated drugs. The flasks were then placed in a 5% CO2, 371C incubator for various intervals. At the end of the incubation period, cells were transferred to sterile centrifuge tubes, pelleted by centrifugation at 400 g for 10 min at room temperature, and prepared for analysis as described below. Assessment of apoptosis After drug exposures, cytocentrifuge preparations were stained with Wright–Giemsa and viewed by light microscopy to evaluate the extent of apoptosis (i.e., cell shrinkage, nuclear condensation, formation of apoptotic bodies, etc.) as described previously (Yu et al., 2001). For these studies, the percentage of apoptotic cells was determined by evaluating X500 cells/condition in triplicate. To confirm the results of morphological analysis, annexin V/PI staining was used. Annexin V/PI (BD PharMingen, San Diego, CA, USA) analysis of cell death was carried out as per the manufacturer’s instructions. Analysis was carried out using a BectonDickinson FACScan cytofluorometer (Mansfield, MA, USA). Preparation of S-100 fractions and assessment of cytochrome c release Cells were harvested after drug treatment and cytosolic S-100 fractions were prepared as described in detail previously (Yu et al., 2003b). Western blot analysis assessing cytochrome c release was performed as described below. Immunoblot analysis

Tet-On inducible Raf-1 cell line A stable Jurkat lymphoblastic leukemia cell line inducibly expressing a constitutively active human Raf-1 (Y340D), in which tyrosine 340 was mutated to aspartic acid, was generated as follows. Flag-tagged Raf-1 was subcloned into the pTRE2-hygro expression vector (Clontech) according to standard techniques. Jurkat ‘Tet-On’ cells that stably express an rtTA (reverse tet transactivator) regulator protein (Clontech) were transfected with Flag-tagged Raf-1-pTRE2-hygro by electroporation (600 V, 60 ms) using 0.4 mM cuvettes. Stable clones derived from single cells were selected by limiting dilution in RPMI 1640 media supplemented with 10% of FBS (Clontech) in the presence of 500 mg/ml of hygromycin. To test for induced expression of the Flag-Raf, stable clones were left untreated or treated for 24 h with 2 mg/ml doxycycline, after which they were harvested and analysed for Flag-Raf-1 expression by Western blot as described above. Oncogene

Immunoblotting was performed as described previously (Yu et al., 2001). In brief, after drug treatment, cells were pelleted by centrifugation, and lysed immediately in Laemmli buffer (1 ¼ 30 mM Tris-base (pH 6.8), 2% SDS, 2.88 mM bmercaptoethanol, and 10% glycerol), and briefly sonicated. Homogenates were quantified using Coomassie protein assay reagent (Pierce, Rockford, IL, USA). Equal amounts of protein (20 mg) were boiled for 10 min, separated by SDS– PAGE (5% stacker and 10% resolving), and electroblotted to nitrocellulose membrane. After blocking in TBS-T (0.05%) and 5% milk for 1 h at 221C, the blots were incubated in fresh blocking solution with an appropriate dilution of primary antibody for 4 h at 221C. The source of antibodies were as follows. PARP (C-2-10), monoclonal (BioMol Research Laboratories, Plymouth, MA, USA). RB[pT821] CDK2 phosphospecific site, polyclonal; pRB[pS249/pT252] CDK4 phosphospecific site, polyclonal (Biosource International,


1375 Camarillo, CA, USA); cleaved caspase-9, polyclonal; cleaved caspase-3, polyclonal; XIAP, polyclonal; phosphor-MEK, polyclonal; phospho-ERK 1/2 (Thr202/Tyr204), polyclonal; phospho-p70S6K (421/424); phospho-c-JUN(Ser63), polyclonal; phospho-p38, polyclonal; phospho-GSK 3, polyclonal; phospho-cdc 2, polyclonal (Cell Signaling Technology, Beverly, MA, USA); caspase-8, polyclonal; Mcl-1, monoclonal; underphospho-pRb, cyclin D1, monoclonal (Pharmingen, San Diego, CA, USA). Cytochrome c, monoclonal; AIF, monoclonal; Bid, polyclonal; Bcl-xL, polyclonal; Raf 1, monoclonal; phospho-JNK, monoclonal; JNK1, monoclonal; phospho-Akt (Ser473), polyclonal; Akt, monoclonal; cdc 2, polyclonal (Santa Cruz Biotechnology, CA, USA). Smac/DIABLO, polyclonal (Upstate Biotechnology, Lake Placid, NY, USA). Blots were washed 3 15 min in TBS-T and then incubated with a 1 : 2000 dilution of horseradish peroxidase-conjugated secondary antibody (Bio-Rad Laboratories, Hercules, CA, USA) for 1 h at 221C. Blots were again washed 3 15 min in TBS-T and then developed by enhanced chemiluminescence (Pierce, Rockford, IL, USA).

Analysis of ROS generation Following treatment, cells were incubated with 10 mM CMH2DCFDA for 30 min, after which they were washed thoroughly in PBS. The percentage of cells displaying increased fluorescence, reflecting removal of an acetate group from the dye by esterases activated by ROS (Liu et al., 2001) was then determined using a Becton Dickinson FACScan flow cytometer (Hialeah, FL, USA) in conjunction with Cell Quest 3.2 software. Statistical analysis The significance of differences between experimental conditions was determined using the two-tailed Student t-test. Acknowledgements This work was supported by awards CA63753 and CA 93738 from the NIH, award 6045-03 from the Leukemia and Lymphoma Society of America, and award DAMD-17-03-10209 from the Department of Defense.

References Allan LA, Morrice N, Brady S, Magee G, Pathak S and Clarke PR. (2003). Nat. Cell Biol., 5, 647–654. Arguello F, Alexander M, Sterry JA, Tudor G, Smith EM, Kalavar NT, Greene Jr JF, Koss W, Morgan CD, Stinson SF, Siford TJ, Alvord WG, Klabansky RL and Sausville EA. (1998). Blood, 91, 2482–2490. Avramis IA, Christodoulopoulos G, Suzuki A, Laug WE, Gonzalez-Gomez I, McNamara G, Sausville EA and Avramis VI. (2002). Cancer Chemother. Pharmacol., 50, 479–489. Bonni A, Brunet A, West AE, Datta SR, Takasu MA and Greenberg ME. (1999). Science, 286, 1358–1362. Carlson B, Lahusen T, Singh S, Loaiza-Perez A, Worland PJ, Pestell R, Albanese C, Sausville EA and Senderowicz AM. (1999). Cancer Res., 59, 4634–4641. Chandra J, Hackbarth J, Le S, Loegering D, Bone N, Bruzek LM, Narayanan VL, Adjei AA, Kay NE, Tefferi A, Karp JE, Sausville EA and Kaufmann SH. (2003). Blood, [Epub ahead of print]. Chang F, Steelman LS, Lee JT, Shelton JG, Navolanic PM, Blalock WL, Franklin RA and McCubrey JA. (2003). Leukemia, 17, 1263–1293. Chang L and Karin M. (2001). Nature, 410, 37–40. Chau BN, Borges HL, Chen TT, Masselli A, Hunton IC and Wang JY. (2002). Nat. Cell Biol., 4, 757–765. Clarke PA, Hostein I, Banerji U, Stefano FD, Maloney A, Walton M, Judson I and Workman P. (2000). Oncogene, 19, 4125–4133. Dan S, Naito M and Tsuruo T. (1998). Cell Death Differ., 5, 710–715. Datta SR, Brunet A and Greenberg ME. (1999). Genes Dev., 13, 2905–2927. Franke TF, Kaplan DR and Cantley LC. (1997). Cell, 88, 435–437. Hu Y, Qiao L, Wang S, Rong SB, Meuillet EJ, Berggren M, Gallegos A, Powis G and Kozikowski AP. (2000). J. Med. Chem., 43, 3045–3051. Inoshita S, Takeda K, Hatai T, Terada Y, Sano M, Hata J, Umezawa A and Ichijo H. (2002). J. Biol. Chem., 277, 43730–43734.

Jarvis WD, Auer KL, Spector M, Kunos G, Grant S, Hylemon P, Mikkelsen R and Dent P. (1997). FEBS Lett., 412, 9–14. Jia W, Yu C, Rahmani M, Krystal G, Sausville EA, Dent P and Grant S. (2003). Blood, 102, 1824–1832. Johnson GL and Lapadat R. (2002). Science, 298, 1911–1912. Kaur G, Gazit A, Levitzki A, Stowe E, Cooney DA and Sausville EA. (1994). Anticancer Drugs, 5, 213–222. Kim AH, Khursigara G, Sun X, Franke TF and Chao MV. (2001). Mol. Cell. Biol., 21, 893–901. Krystal GW, Honsawek S, Litz J and Buchdunger E. (2000). Clin. Cancer Res., 6, 3319–3326. Lei K and Davis RJ. (2003). Proc. Natl. Acad. Sci. USA, 100, 2432–2437. Levitzki A and Gazit A. (1995). Science, 267, 1782–1788. Liu SX, Athar M, Lippai I, Waldren C and Hei TK. (2001). Proc. Natl. Acad. Sci. USA, 98, 1643–1648. Milella M, Kornblau SM, Estrov Z, Carter BZ, Lapillonne H, Harris D, Konopleva M, Zhao S, Estey E and Andreeff M. (2001). J. Clin. Invest., 108, 851–859. Moldeus P and Cotgreave IA. (1994). Methods Enzymol., 234, 482–492. Muda M, Theodosiou A, Rodrigues N, Boschert U, Camps M, Gillieron C, Davies K, Ashworth A and Arkinstall S. (1996). J. Biol. Chem., 271, 27205–27208. Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K and Cobb MH. (2001). Endocr. Rev., 22, 153– 183. Platanias LC. (2003). Blood, 101, 4667–4679. Scheid MP, Schubert KM and Duronio V. (1999). J. Biol. Chem., 274, 31108–31113. Scheid MP and Woodgett JR. (2001). Nat. Rev. Mol. Cell. Biol., 2, 760–768. Sebolt-Leopold JS, Dudley DT, Herrera R, Van Becelaere K, Wiland A, Gowan RC, Tecle H, Barrett SD, Bridges A, Przybranowski S, Leopold WR and Saltiel AR. (1999). Nat. Med., 5, 810–816. Sparrow JR, Vollmer-Snarr HR, Zhou J, Jang YP, Jockusch S, Itagaki Y and Nakanishi K. (2003). J. Biol. Chem., 278, 18207–18213.

Oncogene


1376 Svingen PA, Tefferi A, Kottke TJ, Kaur G, Narayanan VL, Sausville EA and Kaufmann SH. (2000). Clin. Cancer Res., 6, 237–249. Tobiume K, Matsuzawa A, Takahashi T, Nishitoh H, Morita K, Takeda K, Minowa O, Miyazono K, Noda T and Ichijo H. (2001). EMBO Rep., 2, 222–228. Tournier C, Hess P, Yang DD, Xu J, Turner TK, Nimnual A, Bar-Sagi D, Jones SN, Flavell RA and Davis RJ. (2000). Science, 288, 870–874. Turner NA, Xia F, Azhar G, Zhang X, Liu L and Wei JY. (1998). J. Mol. Cell. Cardiol., 30, 1789–1801. Wang X, Martindale JL, Liu Y and Holbrook NJ. (1998). Biochem. J., 333, 291–300. Wennstrom S and Downward J. (1999). Mol. Cell. Biol., 19, 4279–4288.

Oncogene

Wolf G, Hannken T, Schroeder R, Zahner G, Ziyadeh FN and Stahl RA. (2001). FEBS Lett., 488, 154–159. Wu CJ, O’Rourke DM, Feng GS, Johnson GR, Wang Q and Greene MI. (2001). Oncogene, 20, 6018–6025. Xia Z, Dickens M, Raingeaud J, Davis RJ and Greenberg ME. (1995). Science, 270, 1326–1331. Yamamoto K, Ichijo H and Korsmeyer SJ. (1999). Mol. Cell. Biol., 19, 8469–8478. Yu C, Rahmani M, Dai Y, Conrad D, Krystal G, Dent P and Grant S. (2003a). Cancer Res., 63, 1822–1833. Yu C, Rahmani M, Almenara J, Subler M, Krystal G, Conrad D, Varticovski L, Dent P and Grant S. (2003b). Cancer Res., 63, 2118–2126. Yu C, Wang S, Dent P and Grant S. (2001). Mol. Pharmacol., 60, 143–154.