Bcl-2-independent Bcr – Abl-mediated resistance to apoptosis: protection is correlated with up regulation of Bcl-x L Gustavo P Amarante-Mendes 1,3 , Anne J McGahon 1 , Walter K Nishioka 1 , Daniel EH Afar 2 , Owen N Witte 2 and Douglas R Green 1 1 Division of Cellular Immunology, La Jolla Institute for Allergy and Immunology, 10355 Science Center Drive, San Diego; California 92121, USA; 2 Department of Microbiology and Molecular Genetics, Molecular Biology Institute, and Howard Hughes Medical Institute, University of California – Los Angeles, Los Angeles, California 90024 – 1662, USA Bcr – Abl is the molecule responsible for both the transformation phenotype and the resistance to che- motherapeutic drugs found in chronic myelogenous leukemia (CML) cells. Wild-type HL-60, a transformed pro-myelocytic cell line, is very susceptible to apoptosis- inducing agents. We show here that expression of Bcr – Abl in HL-60 cells rendered them extremely resistant to apoptosis induced by a wide variety of agents. The anti- apoptotic eect of Bcr – Abl was found to be independent of the phase of the cell cycle. Treatment with antisense oligonucleotides directed to bcr decreased the expression of the ectopic bcr – abl and restored susceptibility to apoptosis. Double mutations aecting the autophosphor- ylation site and the phosphotyrosine-binding motif (FLVRES) have been previously shown to impair the transforming activity of Bcr – Abl in fibroblasts and hematopoietic cells, however HL-60 cells expressing this double mutant molecule exhibited the same level of resistance to apoptosis as those expressing the wild-type Bcr – Abl. Interestingly, wild type and mutant Bcr – Abl induced in HL-60 cells a dramatic down regulation of Bcl-2 and increased the levels of Bcl-x L . The level of Bax did not change in response to the presence of Bcr – Abl. Antisense oligonucleotides targeted to bcl-x down- regulated the expression of Bcl-x L and increased the susceptibility of HL-60.Bcr – Abl cells to staurosporine. Importantly, HL-60 cells overexpressing Bcl-x L showed higher expression of Bcl-x L but lower resistance to apoptosis when compared to HL-60.Bcr – Abl cells. The results described here show that Bcr – Abl is a powerful mammalian anti-apoptotic molecule and can act in- dependently of Bcl-2. Bcl-x L , however, seems to participate in part in Bcr – Abl-mediated resistance to apoptosis in HL-60 cells. Keywords: Bcr – Abl; Bcl-x; Bcl-2; apoptosis; transfor- mation; tyrosine kinase; HL-60 Introduction For many years, the process of oncogenesis has been described as a deregulation of cellular proliferation leading to the expansion of a particular cell population. More precisely, however, the expansion of a tumor (or any tissue) is directly related to the dierence of the rates of proliferation and cell death (Bresciani et al., 1974). This idea suggests that defects in the apoptosis machinery may contribute to the malignant phenotype found in some forms of cancer. This seems to be true in chronic myelogenous leukemia (CML), where rather than deregulated cell growth, increased resistance to cell death is observed (Koeer and Golde, 1981; Strife and Clarkson, 1988; Stryckmans et al., 1976). This so-called myeloacummulative leukemia is the outcome of a translocation between chromosomes 9 and 22, which generates the Bcr – Abl tyrosine kinase by the fusion of bcr sequences upstream of the second exon of c-abl. This chimeric protein occurs in Philadelphia chromosome-positive leukemias and exists in two dierent forms: a 210 kDa protein is detected in CML and a 185 kDa fusion protein in acute lymphocytic leukemia (ALL). In both cases, Bcr – Abl has been implicated in cellular transforma- tion and inhibition of apoptosis (Carlesso et al., 1994; Cortez et al., 1995). The transforming potential of this oncoprotein is related to its tyrosine kinase activity (Lugo et al., 1990; McLaughlin et al., 1989) and also to the coupling of downstream signal transduction pathways, which is provided by multiple functional domains (Afar et al., 1994; Cortez et al., 1995; Goga et al., 1995; McWhirter et al., 1993; McWhirter and Wang, 1993; Muller et al., 1991; Pendergast et al., 1993). Point mutations in the autophosphorylation site in the SH1 domain or in the conserved FLVRES motif within the SH2 domain impair the potential of the Bcr – Abl molecule to transform fibroblasts (Afar et al., 1994). The two single mutants, but not a double mutant, can be complemented in their transforming ability by over- expression of c-myc (Afar et al., 1994). Despite the vast literature on the biochemical signals triggered by Bcr – Abl, very little is known about the mechanism by which Bcr – Abl inhibits apoptosis. We therefore developed an experimental model in which HL-60, a transformed promyelocytic cell line that is very sensitive to a variety of apoptosis-inducing stimuli, ectopically expresses the p185 isoform of Bcr – Abl. Here, we describe the potent anti-apoptotic eect of Bcr – Abl in HL-60 cells and the possible role of Bcl-2 family members in this eect. We show that ectopic expression of Bcr – Abl leads to the upregulation of Bcl-x L in a variety of cell lines and propose that this contributes to, but does not account for, the overall anti-apoptotic eect of Bcr – Abl. Correspondence: GP Amarante-Mendes. 3 Current address: Departmento de Imunologia, Instituto de Cieˆncias Biome´dicas, Universidade de Sa´o Paulo, Sa´o Paulo 05508-900, Brazil Received 22 May 1997; revised 17 October 1997; accepted 17 October 1997 Oncogene (1998) 16, 1383 – 1390 1998 Stockton Press All rights reserved 0950 – 9232/98 $12.00
Transcript
Bcl-2-independent Bcr ±Abl-mediated resistance to apoptosis: protection iscorrelated with up regulation of Bcl-xL
Gustavo P Amarante-Mendes1,3, Anne J McGahon1, Walter K Nishioka1, Daniel EH Afar2,Owen N Witte2 and Douglas R Green1
1Division of Cellular Immunology, La Jolla Institute for Allergy and Immunology, 10355 Science Center Drive, San Diego;California 92121, USA; 2Department of Microbiology and Molecular Genetics, Molecular Biology Institute, and Howard HughesMedical Institute, University of California ±Los Angeles, Los Angeles, California 90024 ± 1662, USA
Bcr ±Abl is the molecule responsible for both thetransformation phenotype and the resistance to che-motherapeutic drugs found in chronic myelogenousleukemia (CML) cells. Wild-type HL-60, a transformedpro-myelocytic cell line, is very susceptible to apoptosis-inducing agents. We show here that expression of Bcr ±Abl in HL-60 cells rendered them extremely resistant toapoptosis induced by a wide variety of agents. The anti-apoptotic e�ect of Bcr ±Abl was found to be independentof the phase of the cell cycle. Treatment with antisenseoligonucleotides directed to bcr decreased the expressionof the ectopic bcr ± abl and restored susceptibility toapoptosis. Double mutations a�ecting the autophosphor-ylation site and the phosphotyrosine-binding motif(FLVRES) have been previously shown to impair thetransforming activity of Bcr ±Abl in ®broblasts andhematopoietic cells, however HL-60 cells expressing thisdouble mutant molecule exhibited the same level ofresistance to apoptosis as those expressing the wild-typeBcr ±Abl. Interestingly, wild type and mutant Bcr ±Ablinduced in HL-60 cells a dramatic down regulation ofBcl-2 and increased the levels of Bcl-xL. The level of Baxdid not change in response to the presence of Bcr ±Abl.Antisense oligonucleotides targeted to bcl-x down-regulated the expression of Bcl-xL and increased thesusceptibility of HL-60.Bcr ±Abl cells to staurosporine.Importantly, HL-60 cells overexpressing Bcl-xL showedhigher expression of Bcl-xL but lower resistance toapoptosis when compared to HL-60.Bcr ±Abl cells. Theresults described here show that Bcr ±Abl is a powerfulmammalian anti-apoptotic molecule and can act in-dependently of Bcl-2. Bcl-xL, however, seems toparticipate in part in Bcr ±Abl-mediated resistance toapoptosis in HL-60 cells.
For many years, the process of oncogenesis has beendescribed as a deregulation of cellular proliferationleading to the expansion of a particular cellpopulation. More precisely, however, the expansion
of a tumor (or any tissue) is directly related to thedi�erence of the rates of proliferation and cell death(Bresciani et al., 1974). This idea suggests that defectsin the apoptosis machinery may contribute to themalignant phenotype found in some forms of cancer.This seems to be true in chronic myelogenousleukemia (CML), where rather than deregulated cellgrowth, increased resistance to cell death is observed(Koe�er and Golde, 1981; Strife and Clarkson, 1988;Stryckmans et al., 1976).This so-called myeloacummulative leukemia is the
outcome of a translocation between chromosomes 9and 22, which generates the Bcr ±Abl tyrosine kinaseby the fusion of bcr sequences upstream of the secondexon of c-abl. This chimeric protein occurs inPhiladelphia chromosome-positive leukemias andexists in two di�erent forms: a 210 kDa protein isdetected in CML and a 185 kDa fusion protein inacute lymphocytic leukemia (ALL). In both cases,Bcr ±Abl has been implicated in cellular transforma-tion and inhibition of apoptosis (Carlesso et al., 1994;Cortez et al., 1995).The transforming potential of this oncoprotein is
related to its tyrosine kinase activity (Lugo et al., 1990;McLaughlin et al., 1989) and also to the coupling ofdownstream signal transduction pathways, which isprovided by multiple functional domains (Afar et al.,1994; Cortez et al., 1995; Goga et al., 1995; McWhirteret al., 1993; McWhirter and Wang, 1993; Muller et al.,1991; Pendergast et al., 1993). Point mutations in theautophosphorylation site in the SH1 domain or in theconserved FLVRES motif within the SH2 domainimpair the potential of the Bcr ±Abl molecule totransform ®broblasts (Afar et al., 1994). The twosingle mutants, but not a double mutant, can becomplemented in their transforming ability by over-expression of c-myc (Afar et al., 1994).Despite the vast literature on the biochemical
signals triggered by Bcr ±Abl, very little is knownabout the mechanism by which Bcr ±Abl inhibitsapoptosis. We therefore developed an experimentalmodel in which HL-60, a transformed promyelocyticcell line that is very sensitive to a variety ofapoptosis-inducing stimuli, ectopically expresses thep185 isoform of Bcr ±Abl. Here, we describe thepotent anti-apoptotic e�ect of Bcr ±Abl in HL-60cells and the possible role of Bcl-2 family members inthis e�ect. We show that ectopic expression of Bcr ±Abl leads to the upregulation of Bcl-xL in a variety ofcell lines and propose that this contributes to, butdoes not account for, the overall anti-apoptotic e�ectof Bcr ±Abl.
Correspondence: GP Amarante-Mendes.3Current address: Departmento de Imunologia, Instituto de Cieà nciasBiome dicas, Universidade de Sa o Paulo, Sa o Paulo 05508-900, BrazilReceived 22 May 1997; revised 17 October 1997; accepted 17 October1997
Oncogene (1998) 16, 1383 ± 1390 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00
Results
Multi-drug resistance of HL-60 cells expressingBcr ±Abl
Expression of Bcr ±Abl in HL60 cells was achieved byinfection with recombinant retroviruses and selection inG418-containing medium and con®rmed by Westernblot (Figure 1b). Bcr ±Abl expression converted thisapoptosis-sensitive line into cells extremely resistant toapoptosis, regardless of the inducing agent (Figure 2).DNA damaging agents (Ara-C, VP16, VM26, camp-tothecin), protein (CHX) or RNA (Act-D) synthesisinhibitors, cytoskeleton disrupting drugs (CytochalasinB, vincristine sulfate, nocodazol), protein kinasesinhibitors (staurosporine), direct activators of thecaspases (anti-CD95 antibodies), and other agents, werecompared for their e�ects on Bcr ±Abl-positive and-negative HL-60 cells (Figures 2 and 3). Apoptosis wasdetermined after 12 or 24 h, depending on the inducingagent (fast6slow death kinetics). In all cases tested,apoptosis was dramatically inhibited by the expression ofthe Bcr ±Abl tyrosine kinase (Figures 2 and 3). Some ofthese treatments, as expected, changed the cell cyclepro®le of HL-60.Bcr ±Abl cells such that surviving cellswere arrested in di�erent phases of the cell cycle (e.g.CHX-treated cells accumutated in G1 while nocodazol-treated cells arrested in G2 ±Figure 2). Nevertheless,dramatic protection was observed in each of these cases.This suggests that the protection of Bcr ±Abl isindependent of any one cell cycle phase.To determine whether mutations at the autophos-
phorylation site (Y793F), at the conserved FLVRESmotif within the SH2 domain (R552L) of Bcr ±Abl, or atboth of these sites (L552 ±F793; DM) interfere with theresistance to apoptosis conferred by this molecule,mutant Bcr ±Abl proteins were also expressed in HL-60
cells. These mutants have been shown to be unable totransform rodent ®broblasts (Afar et al., 1994). Morerecently, it was shown that the single mutants rendermyeloid and lymphoid cell lines growth factor indepen-dent (Goga et al., 1995). Each of the HL-60 bulk linesobtained were found to express similar levels of Bcr ±Abl(Figure 1b). In every case tested, Bcr ±Abl mutants wereable to protect HL-60 cells from apoptosis to the sameextent observed for the wild-type molecule (Figure 3). Inthe case of apoptosis induced by anti-CD95 antibodies,the expression of CD95 was monitored by FACS analysisand proved to be similar in HL-60.vector and every HL-60.Bcr ±Abl expressing cell population (not shown).
Resistance to apoptosis is inhibited by antisenseoligonucleotides directed against Bcr ±Abl
To con®rm that the anti-apoptotic e�ect observed in HL-60.Bcr ±Abl cells was in fact due to the expression ofBcr ±Abl, these cells were treated with antisense (AS)oligodeoxynucleotides corresponding to the translationstart site of bcr and then examined for resistance toapoptosis. In K562 cells, AS-bcr-induced down regula-tion of Bcr ±Abl protein levels is associated withreduction in the resistance to apoptosis (Martiat et al.,1993; McGahon et al., 1994; Szcylik et al., 1991).Likewise, as exempli®ed for HL-60.Bcr ±Abl DM cells(Figure 4a), down regulation of Bcr ±Abl after treatmentwith AS-bcr restored the susceptibility to VP-16-inducedapoptosis, as detected by the TUNEL assay (Figure 4b).Similar results were obtained with HL-60.Bcr ±Abl WT,HL-60.Bcr ±Abl RL and HL-60.Bcr ±Abl YF cells (datanot shown). As a control for the speci®c e�ect of the AS-bcr on the Bcr ±Abl molecule, we used HL-60 cellsexpressing a temperature-sensitive mutant of v-Abl(RK160) previously described to be resistant toapoptosis at the permissive (328C) but not at therestrictive (398C) temperature (McGahon et al., 1995b,and unpublished results). Treatment of these cells withthe AS-bcr did not alter the expression of v-Abl (notshown) nor a�ect the anti-apoptotic e�ect of v-Ablobserved at 328C (Figure 4c).
Expression of Bcr ±Abl decreases Bcl-2 and increasesBcl-xL levels
Bcl-2 members are well known regulators of apoptosis.Bcl-2 and Bcl-xL protect from apoptosis whereas Baxincreases susceptibility to apoptosis in di�erent systems(Cory, 1995). We therefore compared the levels of theseproteins in HL-60.vector and HL-60.Bcr ±Abl cells. HL-60.vector cells expressed considerable levels of Bcl-2 andBax but not Bcl-xL (Figure 5a). Interestingly, Bcl-2protein was completely absent in HL-60 cells expressingthe wild-type and the mutant Bcr ±Abl molecules (Figure5a). In contrast, Bcl-xL levels were substantially elevatedin HL-60.Bcr ±Abl cells. Bax expression was una�ectedby Bcr ±Abl. Upregulation of Bcl-xL was also observedin Bcr ±Abl-expressing BaF3, DAGM and 3T3 cells(Figure 5b).
Participation of Bcl-xL in Bcr ±Abl-mediated resistanceto apoptosis
Since Bcr ±Abl upregulates the expression of the anti-apoptotic protein Bcl-xL, we decided to investigate
Vect
or
Bcr
–Ab
l WT
Bcr
–Ab
lDM
Bcr
–Ab
l R55
2L
Bcr
–Ab
l Y79
3F
— p185 bcr–abl
— p45 actin
a
b
Figure 1 (a) Scheme of the Bcr ±Abl molecule illustrating itsmost important domains and the point mutations used in thisstudy. (b) Western blot of HL-60 cells expressing wild-type ormutant Bcr ±Abl proteins revealed with anti-c-Abl (upper panel).As a control for protein loading, the same blot was probed foractin (bottom panel)
Anti-apoptotc effect of Bcr ± AblGP Amarante-Mendes et al
1384
whether this e�ect contributed to Bcr ±Abl-mediatedresistance to apoptosis. HL-60.Bcr ±Abl cells weretreated for 60 h with antisense oligonucleotides
corresponding to the translation start site of bcl-xand then examined for resistance to staurosporine-induced apoptosis. Bcl-xL expression was decreased by
Figure 2 Anti-apoptotic e�ect of Bcr ±Abl. HL-60.vector and HL-60p185bcr ± abl WT cells were treated for 12 or 24 h with di�erentdeath-inducing agents and apoptosis estimated by cell cycle analysis. Numbers represent the percentage of cells with subdiploidDNA content
Figure 3 Comparison of the anti-apoptotic e�ect of wild-type and mutant forms Bcr ±Abl. HL-60 cells expressing or not wild-typeor mutant Bcr ±Abl proteins were treated with di�erent concentrations of apoptosis-inducing agents or submitted to U.V.irradiation for di�erent periods of time. The level of apoptosis in these cells were estimated by changes in light scattering properties18 h after the initiated stimuli
Anti-apoptotc effect of Bcr ± AblGP Amarante-Mendes et al
1385
the treatment of cells with the antisense but not withsense or nonsense oligonucleotides (Figure 6a). Downregulation of Bcl-xL correlated with an increase in thesensitivity of HL-60.Bcr ±Abl cells to staurosporine-induced apoptosis (Figure 6b). However, the level ofapoptosis observed under this circumstance (around28%) was less than that observed in the control HL-60.vector cells (always above 75%), suggesting thatBcl-xL is not the only component responsible for theBcr ±Abl-mediated resistance to apoptosis. Similarresults were obtained when HL-60.Bcr ±Abl cells weretreated with a slightly di�erent AS-bcl-x previouslyshown to interfere with the levels of Bcl-xL and topartially block CD40-mediated rescue of sIg-inducedapoptosis in murine B cells (Wang et al., 1995). Tocon®rm this hypothesis, we compared the level ofprotection conferred by overexpression of either Bcr ±
Abl or Bcl-xL in HL-60 cells. HL-60.Bcl-xL cellsexpressed higher levels of Bcl-xL than did HL60.Bcr ±Abl cells (Figure 7a). Nevertheless, HL-60.Bcl-xLcells showed a lower degree of resistance to apoptosisinduced by three di�erent agents than seen inHL-60.Bcr ±Abl cells (Figure 7b). Therefore, over-expression of Bcl-xL did not fully mimic the e�ect ofBcr ±Abl in HL-60 cells.
Discussion
It has been shown in a variety of cell systems thatoncogenic forms of the Abl tyrosine kinase inducetransformation and confer resistance to apoptoticdeath. Substantial knowledge of the mechanismimplicated in Abl-mediated transformation was pro-vided during the past few years and involves an Abltyrosine kinase-dependent activation of multiplebiochemical signals (Cortez et al., 1995; Goga et al.,1995). However, the interaction between oncogenic Abland the cellular apoptotic machinery is far from beingelucidated.In order to investigate the mechanism involved in
the anti-apoptotic e�ect of oncogenic Abl, HL-60 cellswere infected with a retrovirus construct capable ofexpressing Bcr ±Abl. These cells were chosen for tworeasons: ®rst, they are already transformed and do notrequire growth factors (GFs) for survival, and; second,they are very sensitive to a variety of apoptosis-
None AS NSBcr–Abl DM
p185bcr–abl
p45actin
100% 7% 41%
a
b
c
Figure 4 Treatment of HL-60.Bcr ±Abl DM cells with AS-bcrrestores the susceptibility to apoptosis. (a) Western blot of totalcell lysates using anti-c-Abl and anti-actin antibodies. Numbersbelow represent the percentage of Bcr ±Abl protein in relation tocontrol, after normalization to the actin levels. Densitometry wasperformed using the NIH Image software. (b) TUNEL analysisrevealing DNA breaks in cells treated with AS-bcr beforeincubation with VP-16. Note that the concentration of AS useddid not induce cytotoxicity on HL-60.Bcr ±Abl DM cells, andthat NS oligo has no signi®cant e�ect on the resistance toapoptosis in these cells. (c) AS-bcr does not interfere with theanti-apoptotic e�ect of v-Abl. HL-60.v-Ablts treated with AS-bcrretained the resistance to apoptosis at the permissive (328C)temperature as revealed by TUNEL
Vect
or
Vect
or
Vect
or
WT
R/L
Y/F
DM
Bcr–Abl
Bcr
–Ab
l
Bcr
–Ab
l
Bcr
–Ab
l
Bcl-2
Bcl-xL
Bcl-xL
Bax
actin
actin
mo
ck
Ba/F3 DAGM 3T3
a
b
Figure 5 In¯uence of Bcr ±Abl expression on the cellular levelsof Bcl-2 family members. Samples were subjected to SDS±PAGE,transfer to PVDF membranes and probed for the presence of Bcl-2, Bcl-x, Bax and actin. (a) HL-60 cells or (b) BaF3, DAGM and3T3 cells expressing or not Bcr ±Abl were used in the assay
Anti-apoptotc effect of Bcr ± AblGP Amarante-Mendes et al
1386
inducing agents. Therefore, we speculated that thedi�erences in biochemical signals triggered by Bcr ±Ablin these transformed cells and the status quo ofmolecular interactions existing in wild-type HL-60could potentially be related to molecules involved inthe anti-apoptotic pathway.It has been proposed that GF-independence is one
of the multiple steps of tumorigenesis and is directlyrelated to the ability of certain GF to provide anti-apoptotic signals (Adams and Cory, 1992; Canman etal., 1995). Bcr ±Abl was shown to a�ord IL-3-dependent cells, such as 32D and BaF3 cells, GF-independence. This phenomenon was associated withthe ability of Bcr ±Abl to confer resistance toapoptosis and, consequently, it linked a transforma-tion potential to anti-apoptotic signaling. However,Pendergast and colleagues (Cortez et al., 1995) showedthat a mutant Bcr ±Abl (D176 ± 427) that has animpaired transforming ability in these cells stillretained the competence to rescue cells from IL-3withdrawal- or g-irradiation-initiated apoptosis. Thus,Bcr ±Abl-induced resistance to apoptosis can occurindependently from cellular transformation. Werecently observed that the anti-apoptotic e�ect ofBcr ±Abl is independent of PI 3-kinase activity(Amarante-Mendes et al., 1997), which was pre-viously shown to participate in the growth of Bcr ±Abl-positive cells (Skorski et al., 1995) and to beimplicated in the resistance to apoptosis conferred bysome GFs (Minshall et al., 1996; Yao and Cooper,1995). These observations suggest that the biochemicalpathways involved in Bcr ±Abl-mediated transforma-tion and protection from apoptosis may be eitherautonomous or complementary, but are probably notthe same. Also, they suggest that the Bcr ±Abl-
Un
trea
ted
AS
-Bcl
-x
NS
-Bcl
-x
S-B
cl-x
Bcl-xL
actin
a
b
Figure 6 Down-regulation of Bcl-xL by antisense treatmentreduces Bcr ±Abl-mediated resistance to apoptosis. (a) Westernblots of total cell lysates using anti-Bcl-x and anti-actinantibodies. Numbers below represent the percentage of Bcl-xLprotein in relation to control, after normalization to the actinlevels. Densitometry was performed using the NIH Imagesoftware. (b) Cells were treated for 8 h with 1 mM staurosporineand apoptosis was determined by cell cycle analysis of DNAcontent
Vect
or
Bcl
-xL
Bcl-xL
Bcl
-Ab
l
actin
a
b
Figure 7 Comparison of the anti-apoptotic e�ect of Bcl-xL and Bcr ±Abl in HL-60 cells. (a) Western blots of total cells lysatesusing anti-Bcl-x and anti-actin antibodies. (b) Cells were treated for di�erent periods of time with either 100 mM VP16, 1 mMstaurosporine or 100 mM vincristine sulfate. Percentage of apoptosis was determined by cell cycle analysis of DNA content
Anti-apoptotc effect of Bcr ± AblGP Amarante-Mendes et al
1387
dependent anti-apoptotic e�ect probably involvesdi�erent signals from those related to GF-mediatedcell survival.Expression of Bcr ±Abl in HL-60 cells conferred
strong resistance to apoptosis regardless of theinducing agent. This suggests that Bcr ±Abl eitheracts at multiple points to prevent cell death or has onesingle target intimately associated with the centralmechanism of the apoptotic machinery. This mechan-ism, which we refer to as the Executioner (Martin andGreen, 1995), consists of members of a particularfamily of cysteine proteases, namely caspases (Alnemriet al., 1996) (also known as ICE/CED-3 family ofproteases), acknowledged to be responsible for thecleavage of speci®c substrates leading to cellulardemise. Recently, we observed that Bcr ±Abl worksupstream of the activation of the caspases since itblocks proteolytic events associated with apoptosis,including proteolytic activation of caspase-3 andcleavage of PARP, fodrin and others (Amarante-Mendes, GP; Martin SJ and Green DR unpublishedobservations). Bcr ±Abl was shown here to protect HL-60 cells from apoptosis in di�erent phases of the cellcycle (Figure 2), since the protected cells weresometimes arrested in either G1 or G2/M as aconsequence of the agent used.Point mutations in the autophosphorylation site
(Y793F) and/or in the phosphotyrosine binding motif(R552L) did not alter the ability of the Bcr ±Ablmolecule to protect HL-60 cells from apoptosis. It waspreviously reported that both single and doublemutants are defective in transforming ®broblasts(Afar et al., 1994) but the single mutants are fullycapable of transforming hematopoietic cells (Cortez etal., 1995; Goga et al., 1995). In contrast, the doublemutant R552L/Y793F was shown to be uncapable ofbone marrow transformation (Goga et al., 1995) and atriple mutant consisted of the same double mutationplus an additional mutation that inactivates the Grb-2-binding site (Y177F/R552L/Y793F) could not conferIL-3-independency or protect 32D cells fromg-irradiation-induced apoptosis (Cortez et al., 1995).However, the same triple mutant did protect BaF3 cellsfrom g-irradiation and IL-3 withdrawal (Cortez et al.,1995) suggesting that di�erent cells may providealternative pathways for resistance to apoptosisinitiated by Bcr ±Abl. In this regard, HL-60 wouldbehave more like BaF3 than 32D or primary bonemarrow cells. HL-60 cells expressing either of themutant Bcr ±Abl molecules showed similar resistanceto death as those expressing wild-type Bcr ±Abl,regardless of the stimulus and of the method used toevaluate cell death (Figure 3 and unpublishedobservations). These results suggest that in spite ofbeing extremely sensitive to apoptosis, HL-60 cellsprovide a biochemical environment that complementsthe potential de®ciency of some mutant forms of Bcr ±Abl and may be of great value for unraveling themechanisms that operate in Bcr ±Abl-mediated resis-tance to cell death. A possible candidate that couldaccount for this complementation might be c-Myc sincethis molecule is overexpressed in HL-60 cells andshown to collaborate with the single Bcr ±Abl mutantsin transforming ®broblast cells (Afar et al., 1994).All the stable transfectants described here exhibited
a remarkable resistance to apoptosis which could be
overcome by incubation of these cells with antisenseoligonucleotides against the ®rst 18 bases of thetranslation start site of bcr, con®rming the dependencyof the anti-apoptotic e�ect on Bcr ±Abl expression(Figure 4). Such oligonucleotides have previously beenused to show that down regulation of Bcr ±Abl in theCML line K562 renders these cells susceptible toapoptosis (McGahon et al., 1994). It is important tonote that the anti-apoptotic phenotype seems to bedependent on a certain threshold level of Bcr ±Ablsince the treatment with NS-bcr oligonucleotides non-speci®cally down regulated the levels of Bcr ±Abl toaround 40% without a�ecting the resistance toapoptosis.Interestingly, both wild-type and mutant forms of
Bcr ±Abl down regulated Bcl-2 and up regulated Bcl-xL, but did not a�ect the expression of Bax (Figure 5a).Bcl-2 and Bcl-xL have distinguishable biologicalfunctions despite the fact that they share structuraland functional homology. First, it has been shown thatthey have di�erent preferences in binding to other Bcl-2 family members (e.g., Bcl-2 preferentially associateswith Bax (Oltvai et al., 1993) and Bcl-xL with Bad(Yang et al., 1995). Second, while it is stillcontroversial as to whether Bcl-2 prevents CD95/Fas/APO-1-mediated apoptosis (Itoh et al., 1993; Mandalet al., 1996; Memon et al., 1995; Strasser et al., 1995),it is generally accepted that Bcl-xL inhibits this celldeath pathway (Boise and Thompson, 1997). Finally,NunÄ ez and co-workers have recently shown that Bcl-2and Bcl-xL di�erentially block apoptosis induced bychemotherapeutic drugs (Simonian et al., 1997).Therefore, the shift from Bcl-2 to Bcl-xL in HL-60cells expressing Bcr ±Abl might account for, at least,some of anti-apoptotic e�ects of the oncogenic Abl.Indeed, our results with AS-bcl-x support a role forthis anti-apoptotic molecule in Bcr ±Abl-inducedresistance to apoptosis in these cells (Figure 6).In contrast, resistance to apoptosis in HL-60.Bcr ±
Abl cells appears to be independent of Bcl-2 function,since the expression of Bcl-2 is down regulated toundetectable levels in these cells. This result contrastswith a previous observation suggesting that Bcr ±Ablacts through the expression of Bcl-2 (Sanchez Garciaand Grutz, 1995). These authors showed that downregulation of bcl-2 mRNA by transfection of Bcr ±Abl-expressing BaF3 cells with a plasmid containing mousebcl-2 in antisense orientation reverted the IL-3independent phenotype of these cells. It is possiblethat in di�erent cell lines, Bcr ±Abl would require theparticipation of either Bcl-xL or Bcl-2. However, wealso detected reduced levels of Bcl-2 (not shown) andelevated expression of Bcl-xL (Figure 5b) in Bcr ±Abl-positive BaF3 cells suggesting that similar Bcr ±Abl-mediated anti-apoptotic mechanisms operate in bothHL-60 and BaF3 cells. In this regard, it was recentlysuggested that inhibition of apoptosis in BaF3 cells byIL-3 occurs in two distinct levels, one of themindependent of mRNA and protein synthesis and theother dependent on the induction of Bcl-xL expression(Leverrier et al., 1997). Since Bcr ±Abl is capable ofreplacing the IL-3 anti-apoptotic signaling in these cellsit is reasonable to assume that Bcr ±Abl could alsowork by two di�erent mechanisms and therefore onlypart of its anti-apoptotic e�ect would be due to upregulation of Bcl-xL. This hypothesis is supported by
Anti-apoptotc effect of Bcr ± AblGP Amarante-Mendes et al
1388
the observation that overexpression of Bcl-xS in K562cells, which is reportedly antagonistic to the anti-apoptotic e�ect of Bcl-2 and Bcl-xL (Boise et al., 1993),only partially reverts the resistance to apoptosisinduced by chemotherapeutic drugs (Ray et al.,1996). Indeed, we showed here that overexpression ofBcl-xL, despite increasing the resistance to apoptosis inHL-60 cells, did not fully mimic the anti-apoptotice�ect of Bcr ±Abl in this cell line. Therefore, othersignals probably also contribute to the Bcr ±Abl-mediated anti-apoptotic phenotype. One possiblecandidate is the oncogene Ras, which was shown tobe important in Bcr ±Abl-mediated transformation(Pendergast et al., 1993) and also to participate in theresistance to apoptosis observed in the CML line K562(Sakai et al., 1994) and more recently in 32D.Bcr ±Ablcells (Cortez et al., 1996). The fact that Bcr ±Ablmodulated the expression of Bcl-xL in other cell typessuggests a common anti-apoptotic pathway initiated byBcr ±Abl.Thus, there is a remarkable anti-apoptotic e�ect of
Bcr ±Abl in HL-60 cells, independent of both the cellcycle and the expression of Bcl-2 and associated with,but not restricted to, an increase in Bcl-xL proteinlevels. This model may be useful to dissect thebiochemical pathways involved in Bcr ±Abl-mediatedresistance to apoptosis and, consequently, to gaininsights into the molecular mechanisms that coordi-nate the apoptotic machinery of the cells. The potentialknowledge derived from these studies may providemeans to help the development of new strategies fortherapy of Ph1- positive leukemias as well as otherforms of cancer.
Materials and methods
Cell culture and reagents
HL-60 cells were cultured in RPMI-1640 supplementedwith 5% FCS and 2 mM L-glutamine (RPMI-5). HL-60.Bcl-xL line was a generous gift from Dr Kapil Bhalla(Emory University School of Medicine, Atlanta, GA).DAGM cells expressing Bcr ±Abl were described elsewhere(Goga et al., 1995). BaF3 and BaF3.p185bcr ± abl cells werekindly provided by Dr Ann Marie Pendergast (DukeUniversity Medical Center, Durham, NC) and weremaintained in RPMI-5+10% of WEHI-3B conditioningmedia. C2 and PA317 cells were cultured in IMDMsupplemented with 10% FCS and 2 mM L-glutamine.
Etoposide (VP-16) and staurosporine were prepared as100 mM and 1 mM solutions in dimethyl sulfoxide (DMSO),respectively. Actinomycin-D was prepared as as 1 mM stocksolution in RPMI-1640. Cycloheximide (CHX) and ceramide(C6) were prepared in ethanol as 50 mM and 1 mM solutions,respectively. Anti-CD95 IgM mAb (Kamiya Biomedical Co.,Thousand Oaks, CA) was prepared as 0.1 mg ml71 stocksolution in PBS.
Generation of retrovirus and production of Bcr ±Abl-expressingcell lines
HL-60.Bcr ±Abl cells were obtained by retroviral infectionwith pSRaMSVp185bcr ± abltkneo plasmids (Afar et al.,1994). Brie¯y, a combination of 56106 c2 and 56106
PA317 packaging cell lines (c+P) were electroporated(250 mV, 960 mFD) in the presence of 20 mg ml71 ofpSRaMSVtkneo carrying Bcr ±Abl p185 wild-type (WT),p185 R552L (RL; FLVRES mutant), p185 Y793F (YF;
autophosphorylation mutant), or p185 L552 ± F793 (DM;double mutant) genes (Figure 1a). Cells were cultured for 2weeks in DMEM containing 10% iron-supplemented calfserum (Gibco), 2 mM L-glutamine and 8 mg ml71
polybrene, and subsequently selected by their ability togrow in the presence of 0.4 mg ml71 G418 (Gibco BRL).Resistant cells were expanded and served as the source ofretrovirus. HL-60 cells were infected by co-culture with thevirus-producing c+P cells, in the presence of 8 mg ml71
polybrene. G418-resistant cells were maintained in RPMI-1640 supplemented with 5% FCS and 2 mM L-glutamine(RPMI ± FCS). The expression of Bcr ±Abl was con®rmedby Western blot analysis using an anti-c-Abl mAb (Ab-3;Oncogene Science) (Figure 1B).
Treatment with antisense oligonucleotides
The optimal conditions for Bcr ± Abl protein depletion inK562 cells was established previously (McGahon et al.,1994). Bcr ±Abl-positive HL-60 cells were incubated for60 h in the presence of 30 mM of either sense (S), antisense(AS) or nonsense (NS) phosphorothioate derivatisedoligodeoxynucleotides (QCB, USA) corresponding to the®rst 18 bases of the translation start site of bcr or bcl-x.Fresh oligonucleotides were added at 16, 32 and 48 h.Oligonucleotide sequences used were as follows: AS-bcr: 5'-GCCCACCGGGTCCACCAT-3'; NS-bcr: 5'-CGCGCCTC-GTCCCAAGCA-3'; S-bcl-x: 5'-ATGTCTCAGCAACCGG-3'; AS-bcl-x: 5'-CCGGTTGCTCTGAGACAT-3';NS-bcl-x: 5'-CTGAACGGAGACCCTT AG-3'.
Western blot analysis
Cells were harvested, washed once in cold PBS andresuspended in suspension bu�er (SB; 0.1 M NaCl, 0.01 M
Tris-C1 pH 7.6, 0.001 M EDTA, 1 mg ml71 aprotinin, 1 mgml71 leupeptin, 100 mg ml71 PMSF, 1 mM NaNO3). Equalamounts of SDS-sample bu�er 26(5% 2-ME, 4% SDS,20% glycerol, 100 mM Tris-Cl ± pH 6.8) were immediatelyadded and samples were boiled for 5 min. Protein contentin the cell extracts were determined by the Bio-Radmicroassay. Samples were resolved on SDS ± polyacryla-mide gel, transferred to PVDF membranes and immuno-blotted with either anti-c-Abl (c-abl (Ab-3), OncogeneScience) or anti-actin (clone C4 ± ICN Biomedicals, Inc.,Costa Mesa, CA) monoclonal antibodies or with rabbitpolyclonal antiserum raised against Bcl-2, Bcl-x or Bax(kindly provided by Dr John Reed, The Burnham Institute,La Lolla, CA). Reactions were detected with horseradishperoxidase-conjugated goat anti-mouse or mouse anti-rabbit antibody using an enhanced chemiluminescencesystem (ECL, Amersham). The NIH-Image software wasused for densitometry analysis.
Determination of apoptosis
Apoptosis was assessed by several criteria. Percentage ofcell death was calculated by changes in the lightscattering properties due to reduction in cell size andincrease in granularity (McGahon et al., 1995a). DNAfragmentation was quanti®ed by cell cycle analysis oftotal DNA content as described elsewhere (Nicoletti etal., 1991), with slight modi®cations. A total of 26105
cells were washed twice in PBS and resuspended in0.25 ml of hypotonic ¯uorochrome solution (50 mg ml71
propidium iodide and 0.1% Triton X-100 in 0.1% sodiumcitrate). After 30 min at 48C in the dark, the ¯uorescenceof individual nuclei was measured using a FACScan ¯owcytometer (Becton-Dickinson, Mountain View, CA). Thepercentage of hypodiploid nuclei correlates with theextent of apoptosis in the sample. DNA breaks was
Anti-apoptotc effect of Bcr ± AblGP Amarante-Mendes et al
1389
measured by the TUNEL assay as previously described(McGahon et al., 1995a).
AcknowledgementsWe wish to thank Dr Ann Marie Pendergast (DukeUniversity Medical Center, Durham, NC) for providingBaF3 and BaF3.p185bcr ± abl cells, Dr Kapil Bhalla (Emory
University School of Medicine, Atlanta, GA) and Dr JohnC Reed (The Burnham Institute, La Lolla, CA) for the giftof anti-Bcl-2, anti-Bcl-x and anti-Bax polyclonal antibodiesand Dr Andy J Minn (University of Chicago, IL) forhelpful suggestions. GPA-M was a Brazilian ResearchCouncil (CNPq) Fellow. This research was supported by agrant from the American Cancer Society (CB-82) to DRG.This is the publication #189 from the La Jolla Institute forAllergy and Immunology.
References
Adams JM and Cory S. (1992) Cancer Surv., 15, 119 ± 141.Afar DE, Goga A, McLaughlin J, Witte, ON and SawyersCL. (1994). Science, 264, 424 ± 426.
Alnemri ES, Livingston DJ, Nicholson DW, Salvesen G,Thornberry NA, Wong WW and Yuan J. (1996). Cell, 87,171.
Amarante-Mendes GP, Jascur T, Nishioka WK, Mustelin Tand Green, DR. (1997). Cell Death Di�., 4, 548 ± 554.
Boise LH, Gonzalez-Garcia M, Postema E, Ding L, LindstenL, Turka A, Mao X, Nunez G and Thompson CB. (1993).Cell, 74, 597 ± 608.
Boise LH and Thompson CB. (1997). Proc. Natl. Acad. Sci.USA, 94, 3759 ± 3764.
Bresciani F, Paoluzi R, Benassi M, Nervi C, Casale C andZiparo E. (1974). Cancer Res., 34, 2405 ± 2415.
Cory S. (1995). Ann. Rev. Immunol., 13, 513 ± 543.Goga A, McLaughlin J, Afar DE, Sa�ran DC andWitte ON.(1995). Cell, 82, 981 ± 988.
Itoh N., Tsujimoto Y and Nagata S. (1993). J. Immunol., 151,621 ± 627.
Koe�er PH and Golde D. (1981). N. Eng. J. Med., 304,1201 ± 1209.
Leverrier Y, Thomas J, Perkins GR, Mangeney M, CollinsMK and Marvel J. (1997). Oncogene, 14, 425 ± 430.
Lugo TG, Pendergast AM, Muller AJ and Witte ON. (1990).Science, 247, 1079 ± 1082.
Mandal M, Maggirwar SB, Sharma N, Kaufmann SH, SunSC and Kumar R. (1996). J. Biol. Chem., 271, 30354 ±30359.
Martiat P, Lewalle P, Taj AS, Philippe M, Larondelle Y,Vaerman JL, Wildmann C, Goldman JM andMichaux JL.(1993). Blood, 81, 502 ± 509.
Martin SJ and Green DR. (1995) Cell, 82, 349 ± 352.McGahon A, Bissonnette R, Schmitt M, Cotter KM, GreenDR and Cotter TG. (1994). Blood, 83, 1179 ± 1187.
McGahon AJ, Martin SJ, Mahboubi A, Bissonette RP, ShiY, Mogil R, Nishioka W and Green DR. (1995a). CellDeath, Vol. 46: Methods in Cell Biology. Schwartz LM andOsborne BA. (eds). Academic Press, Inc.: San Diego, pp153 ± 186.
McGahon AJ, Nishioka WK, Martin SJ, Mahboubi A,Cotter TG and Green DR. (1995b). J. Biol. Chem., 270,22625 ± 22631.
McLaughlin J, Chianese E and Witte ON. (1989). Mol. Cell.Biol., 9, 1866 ± 1874.
McWhirter JR, Galasso DL and Wang JY. (1993).Mol., CellBiol., 13, 7587 ± 7595.
McWhirter JR and Wang, JY. (1993). EMBO J., 12, 1533 ±1546.
Memon SA, Moreno MB, Petrak D and Zacharchuk CM.(1995). J. Immunol., 155, 4644 ± 4652.
