Toxicology in Vitro 20 (2006) 816–823www.elsevier.com/locate/toxinvit

ERK1/2 activation attenuates TRAIL-induced apoptosis throughthe regulation of mitochondria-dependent pathway

Do Yeon Lee a,1, Myoung Woo Lee a,1, Hyun Jung Lee a, Yoo Hun Noh a, Soon Cheol Park b,Moo Yeol Lee c, Kyung Yong Kim a, Won Bok Lee a, Sung Su Kim a,d,¤

a Department of Anatomy and Cell Biology, College of Medicine, Chung-Ang University, 221 Huksuk-dong, Dongjak-ku, Seoul 156-756, Republic of Koreab Department of Life Science, College of Natural Science, Chung-Ang University, Seoul, Republic of Korea

c Department of Physiology, College of Medicine, Chung-Ang University, 221 Huksuk-dong, Dongjak-ku, Seoul 156-756, Republic of Koread BioGrand Inc., Seoul, Republic of Korea

Received 3 May 2005; accepted 5 January 2006Available online 23 March 2006

Abstract

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) functions as an extracellular signal, which triggers apoptosis intumor cells. In order to characterize the molecular events involved in TRAIL cytotoxic signaling, we attempted to determine the role ofextracellular signal-regulated kinase 1/2 (ERK1/2), as well as its downstream targets in TRAIL-treated HeLa cells. Here we demonstratethat TRAIL exposure resulted in the activation of ERK1/2, and the elevation of anti-apoptotic Bcl-2 protein levels. ERK1/2 inhibitionwith PD98059 promoted cell death via the down-regulation of Bcl-2 protein levels, together with increasing mitochondrial damage,including the collapse of mitochondrial membrane potential, the release of cytochrome c from mitochondria to cytoplasm and caspaseactivity. These results suggest that the ERK1/2 activation is a kind of survival mechanism to struggle against TRAIL-induced stress con-dition in early stage, via activating cellular defense mechanisms like as the up-regulation of the Bcl-2/Bax ratio, as well as several mito-chondrial events.© 2006 Elsevier Ltd. All rights reserved.

Keywords: Tumor necrosis factor-related apoptosis-inducing ligand; Extracellular signal-regulated kinase 1/2; Bcl-2/Bax; Cytochrome c; Apoptosis

1. Introduction

Tumor necrosis factor-related apoptosis-inducing ligand(TRAIL) is a member of the tumor necrosis factor (TNF)family. TRAIL has proven to be capable of inducing apop-tosis in several cell lines (Knight et al., 2001; Matsuda et al.,2005). However, its physiological functions remainunknown. TRAIL is widely expressed in normal cells, andis highly homologous with FasL, another cytotoxic mem-

* Corresponding author. Address: Department of Anatomy and CellBiology, College of Medicine, Chung-Ang University, 221 Huksuk-dong,Dongjak-ku, Seoul 156-756, Republic of Korea. Tel.: +82 2 820 5690; fax:+82 2 826 1265.

E-mail address: sungsu@cau.ac.kr (S.S. Kim).1 Do Yeon Lee and Myoung Woo Lee contributed equally to this work.

0887-2333/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.tiv.2006.01.007

ber of the TNF family (Knight et al., 2001; Matsuda et al.,2005). Several receptors in human have been determined tobind to TRAIL. These include the death receptors DR4(TRAIL-R1) and DR5 (TRAIL-R2), the decoy receptorsDcR1 (TRAIL-R3) and DcR2 (TRAIL-R4), and osteopro-tegerin (Kelley and Ashkenazi, 2004; Yagita et al., 2004).Although some eVorts have been made to elucidate themolecular mechanisms underlying TRAIL signaling, thecomponents of the diVerent TRAIL signaling pathwaysremain largely undeWned.

Mitogen-activated protein (MAP) kinase signal trans-duction pathways in mammalian cells include extracellularsignal-regulated kinase1/2 (ERK1/2), c-Jun N-terminalkinase/stress-activated protein kinase (JNK/SAPK), andp38 MAP kinase (SchaeVer and Weber, 1999; Changand Karin, 2001). ERK1/2 is associated principally with

D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823 817

proliferation and growth factors (Tsukada et al., 2001),while JNK and p38 MAP kinase are induced by stressresponses and cytokines, and are able to mediate diVerenti-ation and cell death (Nagata and Todokoro, 1999). ERK1/2has been conWdently implicated in the regulation of avariety of cellular processes. However, the precise molecu-lar mechanism of ERK1/2 has still remained controversial.For example, ERK1/2 plays a prominent role in ultravio-let (UV)-evoked p53 phosphorylation (She et al., 2000) and1-methyl-4-phenylpyridinium (MPP+)-induced neurotoxi-city (Gomez-Santos et al., 2002). By way of contrast, ERK1/2 activation promotes cell survival in neuronal PC12 cells(She et al., 2000), and ERK1/2 stimulation protects HeLacells from Fas, TNF, TRAIL receptor-induced apoptosis(Tran et al., 2001). Although ERK1/2 has been deWnitivelydemonstrated to be involved in various physiologicalevents, less is understood with regard to its biological rolesin TRAIL signaling.

Downstream targets of ERK1/2 include ribosomal S6kinase (RSK) (Merienne et al., 2000), ETS domain tran-scription factor Elk-1 (Tsai et al., 2000), and the anti-apop-totic Bcl-2 protein family (Desire et al., 2000). The ERK1/2signaling pathway regulates the expression of Bcl-2 andBcl-XL, and promotes the survival of human pancreatictumor cells (Boucher et al., 2000). Furthermore, Bcl-2/Bcl-XL overexpression in cells exerts a protective eVect againsta host of agents, including UV irradiation (Domen et al.,1998), cytotoxic drugs (Srivastava et al., 1999; Takahashiet al., 1999), and p53 (Park et al., 2001). In addition, a greatdeal of evidence suggests that a variety of apoptotic stimuliaVect the formation of mitochondrial permeability transi-tion pores (MPTP), and induce the release of pro-apoptoticmolecules, such as cytochrome c, from the mitochondria(Jiang and Wang, 2000). When present in the cytoplasm,cytochrome c interacts with apoptotic activator factor-1(Apaf-1) and caspase 9 (Purring-Koch and McLendon,2000). This complex then, either directly or indirectly,induces apoptotic cell death via caspase 3 activation (Cain,2003). Recent studies have also reported that Bcl-2 proteinoverexpression can perform an anti-apoptotic function, byinhibiting the release of cytochrome c during TRAIL-induced apoptosis (Sun et al., 2001). However, anotherstudy demonstrated that TRAIL-induced cytochrome crelease is not regulated by the Bcl-2 protein (Keogh et al.,2000). Thus, the relationship between Bcl-2 protein elevatedby ERK1/2 and mitochondrial events in TRAIL-inducedapoptosis remains a matter of some controversy.

In this study, we have evaluated the possibility thatERK1/2 activation, Bcl-2 protein family, and mitochon-drial events are all involved in TRAIL cytotoxic signaling.Therefore, we attempted to more clearly elucidate the rela-tionship between these factors with regard to TRAIL-induced HeLa cell death. Our results suggest that ERK1/2activation plays a protective role, as a cellular defensemechanism to survive, via the regulation of the Bcl-2/Baxratio and several mitochondrial events during TRAIL-induced apoptosis.

2. Materials and methods

2.1. Cell culture

Human adenocarcinoma HeLa cells were cultured at37 °C in Dulbecco’s ModiWed Eagle’s Medium (DMEM;Gibco BRL, NY, USA) supplemented with 10% heat-inac-tivated fetal bovine serum (FBS; Gibco BRL) in a humidi-Wed incubator at an atmosphere of 95% air, and 5% CO2.The cells were transferred to low serum media (1% FBS),2 h before treatment with recombinant human TRAIL(BIOMOL, Plymouth, PA, USA).

2.2. Cell viability assay

Cells were plated onto 96-well plates (Corning, NY,USA) at a density of 5£ 104 cells/well, in 100 �l of 10%FBS/DMEM without phenol red, then incubated for 24 h.Two hours prior to cell stimulation, the cell media was sup-planted with 90 �l of 1% FBS/DMEM without phenol red.TRAIL (4 ng/ml) was added to the plates, which had beenpretreated with either vehicle or PD98059 (Calbiochem,San Diego, CA, USA) 2 h prior to TRAIL treatment. Twohours prior to the end of the TRAIL treatment, 10 �l of10% Triton X-100 was added, followed by the addition of11�l of 3-(4,5-dimethyl-thiozol-2-yl)-2,5-diphenyl-tetrazo-lium bromide (MTT; Sigma, St Louis, MO, USA) 10£solution (10 mg/ml). After 4 additional hours of incubation,all remaining media and MTT solution was suctioned oV,and the cells and crystallized dyes were dissolved via theaddition of 100 �l of 100% DMSO, and 20 min of shaking.Absorbance at 570 nm was measured with an ELISAReader (Molecular Devices, Sunnyvale, CA, USA). Assayvalues obtained immediately prior to vehicle treatmentwere set as 100%, and complete inhibition of MTT reduc-tion (0%) was deWned as the value obtained following theaddition of 10% Triton X-100.

2.3. Western blotting

Stimulated cells were washed with phosphate-buVeredsaline (PBS) and lysed in 300 �l of cold RIPA buVer(20 mM Tris–HCl, pH 7.5 containing 1% Triton X-100,100 mM NaCl, 1 mM Na3VO4, 40 mM NaF, 5 mM EGTA,0.2% SDS, 0.2 mM PMSF, and 100 �M Leupeptin). Celllysates were then centrifuged for 10 min at 15,800g, and4 °C. The supernatants were harvested and analyzed withregard to protein concentration, using a protein assay kit(Bio-Rad, CA, USA). For electrophoresis, 80 �g of proteinswere dissolved in sample buVer (0.1 M Tris, pH 6.8 contain-ing 5% �-mercaptoethanol, 15% glycerol, 3% SDS, and0.1% bromophenol blue), boiled for 5 min, then separatedon 10% SDS gel under reducing conditions. The separatedproteins were then transferred onto polyvinylidene diXuo-ride (PVDF) membranes (Amersham Pharmacia Biotech,UK), using a semidry trans-blot system (Schleicher &Schuell, Germany). The blots were blocked for 1 h with

818 D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823

Tris-buVered saline (TBS) (10 mM Tris, pH 7.5 with100 mM NaCl) containing 5% nonfat dry milk (GibcoBRL), at room temperature. The blots were washed threetimes with TBS, then incubated at 4 °C overnight with anti-ERK1/2 (1:1000, New England Biolabs, Beverly, MA,USA), anti-phospho-ERK1/2 (1:1000, New England Bio-labs), anti-Bcl-2 (1:1000, SantaCruz Biotech, CA, USA),anti-Bax (1:1000, SantaCruz Biotech), or anti-�-actin(1:1000, SantaCruz Biotech) antibodies in TBST (10 mMTris, pH 7.5 containing 100 mM NaCl and 0.05% Tween 20)containing 3% nonfat dry milk. The next day the blots werewashed three times with TBST, then incubated for 1 h withhorseradish peroxidase-conjugated secondary antibodies(1:2000, SantaCruz Biotech) in TBST, containing 3% non-fat dry milk at room temperature. After three washes inTBST, the protein was visualized with the ECL detectionsystem (Amersham Pharmacia Biotech). Relative proteinlevels were determined via densitometry (Vilber Lourmat,France), normalized to the ERK1/2 or �-actin band on aduplicate blot.

2.4. Determination of mitochondrial membrane potential (MMP)

TRAIL-induced changes in mitochondrial membranepotential (MMP) were measured by incubation with tetra-methyl-rhodamine ethyl ester (TMRE; Molecular Probes,Leiden, Netherlands), a cationic potentiometric dye, whichpreferentially accumulates in energized mitochondria, aprocess which is driven by membrane potential. Cellswere stained with 100 nM of TMRE for 15 min, and werethen collected and washed twice with PBS. Cells were posi-tioned on slide-glasses and mounted. Photomicrographs ofmounted cells were taken with a Xuorescent microscopeequipped with a UV supply system (Olympus IX 70, Tokyo,Japan). In order to quantify changes occurring in the MMP,the cells were loaded with 100 nM of TMRE, then incubatedfor 10 min at 37 °C. The intensity of the TMRE Xuorescencewas determined using a Xow cytometer (GENios, Tecan,NC, USA), with an excitation wavelength of 549 nm, and anemission wavelength of 574 nm. The background Xuores-cence intensity in the control cells without TMRE was sub-tracted in all cases. The mean TMRE Xuorescence indiceswere calculated by dividing the mean TMRE Xuorescence ofeach sample by that measured in the cells treated withTRAIL only.

2.5. Analysis of cytochrome c release

Cells were collected and resuspended in 100�l of lysisbuVer (20 mM HEPES, pH 7.5 containing 10 mM KCl,1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 1 mM dithio-threitol, 0.1 mM PMSF, and 100�M Leupeptin). After30 min of incubation on ice, the cells were homogenized by25 strokes of a Dounce homogenizer. Nuclei and unbrokencells were removed by 10 min of centrifugation at 1000g at4 °C, and the supernatants were centrifuged again at

14,000g at 4 °C for 20 min. Cytochrome c levels in theresulting supernatant were analyzed by Western blottingusing anti-cytochrome c antibodies (1:1000, SantaCruz Bio-tech). Relative protein levels were determined using densi-tometry, normalized to the �-actin band on a duplicateblot.

2.6. Caspase substrate cleavage assay

For assaying caspase activity in HeLa cell, monolayersof cultured cells (2§106 cells) were harvested from a 60 mmdish and lysed with 1 ml of cell lysis buVer. Fifty micro-grams of the lysed sample was incubated with 100 �l ofHEPES buVer (100 mM HEPES, pH 7.5, 10% sucrose, 0.1%CHAPS, 10 mM DTT) containing each of 0.5 mM of Ac-DEVD-AMC and Ac-LEHD-AMC (Pharmingen, USA)for 30 min to 1 h. Caspase-3 and -9 activity was measuredwith excitation at 380 nm and emission at 460 nm using aXuorometer (TECAN, GENios). Enzymatic activity isexpressed as arbitrary units of relative value.

2.7. Statistical analysis

All data were expressed as mean§SD. Statistical signiW-cance was determined via the Student’s t-test. A p value of<0.05 was considered to be signiWcant.

3. Results

3.1. ERK1/2 activation as the cellular defense mechanism in TRAIL-induced apoptosis

To examine if the TRAIL could induce cell death inHeLa cells, we Wrst checked the eVect of TRAIL on cell via-bility, using the MTT assay. As shown in Fig. 1(A), TRAILinduced cell death in a time-dependent manner in HeLacells. Treatment with 4 ng/ml of TRAIL induced cell deathin about half of the population at 48 h. In order to deter-mine whether ERK1/2 activation was involved in TRAILsignaling, we assessed ERK1/2 levels by Western blottingwith anti-phospho-ERK1/2 antibodies which recognizeonly the activated, phosphorylated form of ERK1/2. Asshown in Fig. 1(B), TRAIL treatment resulted in a pro-found activation of ERK1/2, the activity of which increasedwithin 3 h, reached maximal levels at 12 h, and returned tobasal levels after 24 h. Also, pretreatment with either 5, 10,or 20 �M of PD98059, a speciWc inhibitor of mitogen-acti-vated protein kinase (MAPK/ERK) kinase (MEK), eVec-tively attenuated TRAIL-induced ERK1/2 activity(Fig. 1(B)). In order to more precisely characterize the roleof activated ERK1/2 in TRAIL-induced cell death, weassessed the eVects of ERK1/2 inhibition on the TRAIL-treated HeLa cells. As shown in Fig. 1(C), the concurrentaddition of 10 �M of PD98059 increased cell death rates byabout 20% more than the cell death rates induced byTRAIL (Fig. 1(C)). Moreover, the inhibition of ERK1/2with PD98059 signiWcantly promoted nuclear condensation

D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823 819

and fragmentation, morphological features of apoptoticcell death (data not shown). In a previous study, wereported that TRAIL-induced neuronal apoptosis occurredvia increasing caspase activity (Lee et al., 2002). Thecaspase activity was increased in a time- and dose-depen-dent manner in TRAIL-induced cell death. Thus, we stud-ied if ERK could eVect of caspase 3 activity. As shown inFig. 1(D), the caspase 3 activity in TRAIL-treated HeLacells was increased several fold compared with untreatedcells, whereas pre-incubation of PD98059 enhanced caspase3 activity. These results indicated that TRAIL profoundlyactivated ERK1/2 during TRAIL-induced apoptosis andERK activation was involved in the negative regulation ofapoptosis through inhibiting activation of caspase 3 in

TRAIL-induced cell death, thereby contributing to cellulardefense against cytotoxic agents, including TRAIL.

3.2. Activated ERK1/2 increased Bcl-2 protein levels during TRAIL-induced apoptosis

In order to evaluate the involvement of the Bcl-2 proteinin TRAIL-induced apoptosis, we assessed changes in thelevels of its expression in TRAIL-treated HeLa cells. Asshown in Fig. 2, Western blotting demonstrated that anti-apoptotic Bcl-2 protein levels increased gradually as theresult of TRAIL treatment (Fig. 2(A)). Fig. 2(B) shows agraph which quantiWed the expression levels of Bcl-2 or Baxat the indicated time points. Also, in the HeLa cells exposed

Fig. 1. Western blot analysis of ERK1/2 activation and eVect of PD98059 on TRAIL-induced apoptosis in HeLa cells. (A) HeLa cells were pretreated with1% FBS/DMEM for 2 h and then treated with 4 ng/ml of TRAIL. Cell viability was determined by MTT assay at indicated time points. Values, presentedas a percentage of control cells incubated with vehicle, are mean § SEM of three separate experiments. CTL D control. HeLa cells were untreated (CTL) ortreated (TRAIL) with 4 ng/ml of TRAIL. (B) HeLa cells were left untreated (CTL; lane 1) or were treated with 4 ng/ml of TRAIL for 3 h (lane 2), 6 h (lane3), 12 h (lane 4) and 24 h (lane 5) (upper panel). HeLa cells were left untreated (CTL; lane 1) and cells were left untreated (lane 2) or incubated with either5 �M (lane 3), 10 �M (lane 4), or 20 �M of PD98059 (lane 5) for 2 h, followed by treatment with 4 ng/ml of TRAIL for 6 h (middle panel). The phosphory-lation state of ERK1/2 was detected by Western blotting with a phospho-speciWc ERK1/2 antibody at the indicated time points (pERK1/2). The samesamples were probed with ERK1/2 antibody as a loading control (ERK1/2). (C) HeLa cells were pretreated with 10 �M of PD98059 for 2 h, then treatedwith 4 ng/ml of TRAIL. Cell viability was determined by MTT assay at indicated time points. Values, presented as a percentage of control cells incubatedwith vehicle, are mean § SEM of three separate experiments. CTL D control. HeLa cells were untreated (CTL) or treated (TRAIL) with 4 ng/ml ofTRAIL. Pretreatment of HeLa cells with 10 �M of PD98059 were treated (PD10+TRAIL) or left untreated (PD10) with 4 ng/ml of TRAIL. (D) Thecaspase-3 activity was detected using Xuorogenic substrate, Ac-DEVD-AMC (Section 2) with TRAIL (4 ng/ml). The eVect of pretreatment of eitherPD98059 (10 �M) or Ac-DEVD-CHO (10 �M) was represented. For caspase 3 activity, 20 �g of cellular extracts incubated 0.5 mM Ac-DEVD-AMC in10 �l of total volume at 37 °C for 30 min. Excitation at 380 nm and emission at 460 nm were measured with a Xuorometer. The data are relative value com-pared with vehicle treated value. The data are the mean § SEM of three separate experiments. ¤ D p < 0.05, compared with vehicle alone, ¤¤ D p < 0.05, com-pared with TRAIL alone.

820 D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823

to TRAIL, pro-apoptotic Bax protein levels increasedwithin 1 h, followed by maintenance of maximum levels(Fig. 2(B)). These results indicate that elevated Bcl-2 pro-

Fig. 2. Expression levels of Bcl-2 or Bax in TRAIL-treated HeLa cells. (A)HeLa cells were untreated (CTL; lane 1) or treated with 4 ng/ml ofTRAIL for 1 h (lane 2), 3 h (lane 3), 6 h (lane 4), 12 h (lane 5), and 24 h(lane 6). (B) QuantiWcation of (A) results. Expression levels of Bcl-2 orBax were detected by Western blotting with anti-Bcl-2 or anti-Bax anti-body at the indicated time points (Bcl-2 or Bax). The same samples wereprobed with �-actin antibody as a loading control (�-actin). Total levelsof Bcl-2 or Bax were used as internal standards in the quantiWcation of�-actin. The levels of Bcl-2 or Bax were expressed as arbitrary units ofrelative value. Values are mean § SEM of three separate experiments.

teins masked the pro-apoptotic eVects conferred by Bax, bybinding to these proteins, and thereby contributing to cellu-lar defenses against TRAIL-induced cytotoxicity. In orderto determine the relationships between activated ERK1/2and anti-apoptotic Bcl-2 protein, we assessed the eVects ofERK1/2 inhibition on Bcl-2 protein levels which had beenincreased by TRAIL treatment. As shown in Fig. 3(A) and(B), ERK1/2 activity induced by TRAIL was eVectivelyinhibited by pretreatment with 10 �M of PD98059. ERK1/2inhibition with PD98059 resulted in signiWcant reductionsof the elevated Bcl-2 protein levels induced by TRAIL(Fig. 3(A)), but did not aVect the increased pro-apoptoticBax protein levels (Fig. 3(B)). These results indicate thatactivated ERK/12 up-regulates Bcl-2 protein levels, but notBax protein levels, during TRAIL-induced apoptosis.

3.3. Activated ERK1/2 attenuated TRAIL-induced cytotoxicity via the regulation of mitochondrial events

In order to characterize the changes in mitochondrialevents induced by TRAIL treatment, we observed the col-lapse of MMP and the release of cytochrome c into thecytoplasm in TRAIL-treated HeLa cells. The TRAIL-induced collapse of MMP was measured via incubationwith TMRE, which preferentially accumulates in energizedmitochondria, a process which is driven by membranepotential. Incubating cells with TRAIL for 6h reduced theXuorescent intensity of TMRE signiWcantly compared withthe control (Fig. 4(A) and (B)). The staining levels of Xuo-rescent TMRE decreased by about 40% of the levels seen inthe non-treated control cells (Fig. 4(C) and (F)), and therelease of cytochrome c into the cytoplasm was alsostrongly induced by 12 h of TRAIL treatment (Fig. 5). In

Fig. 3. EVect of PD98059 on the expression levels of Bcl-2 or Bax. (A) HeLa cells were untreated (CTL) or treated (TRAIL) with 4 ng/ml of TRAIL for12 h. Cells were incubated with 10 �M of PD98059 for 2 h, followed by treatment with 4 ng/ml of TRAIL for 12 h (PD + TRAIL) or no treatment (PD10).(B) QuantiWcation of (A) results. Expression levels of Bcl-2 or Bax were detected by Western blotting with anti-Bcl-2 or anti-Bax antibody at indicatedtime points (Bcl-2 or Bax). The same samples were probed with �-actin antibody as a loading control (�-actin). Total levels of Bcl-2 or Bax were used as aninternal standard in the quantiWcation of �-actin. The levels of Bcl-2 or Bax were expressed as arbitrary units of relative value. Values are mean § SEM ofthree separate experiments.

D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823 821

order to determine the relationship between activatedERK1/2 and mitochondrial events in TRAIL-inducedapoptosis, we examined the eVects of ERK1/2 inhibition onchanges in mitochondrial events as the result of TRAILtreatment. ERK1/2 inhibition with PD98059 resulted in asigniWcant reduction of the staining levels of XuorescentTMRE, as compared with the TRAIL-treated HeLa cells(Fig. 4(D)), and increased the release of cytochrome cinduced by TRAIL data in Fig. 5 show a »2-fold increase(Fig. 5(B)). The other sign of mitochondrial damage is acti-vation of caspase 9. It has been reported that released cyto-chrome c binds to Apaf-1 and that this complex activatescaspase 9 (Purring-Koch and McLendon, 2000). As shownin Fig. 6, caspase 9 activity increased »4.4-fold comparedwith the control group after 6 h of incubation with TRAIL,while pretreatment of PD98059 signiWcantly increased thecaspase 9 activity compared with TRAIL. These resultsindicate that TRAIL-activated ERK1/2 plays a protectiverole, acting as a cellular defense mechanism, via the regula-tion of mitochondrial events in HeLa cells.

4. Discussion

As TRAIL is known to selectively induce apoptosis intumor cells, but not in normal cells, it has evidencedremarkable potential as valuable agent for antitumor ther-apy (Walczak et al., 1999). In spite of the many approachesand eVorts which have been undertaken in the investigationof the biological functions of TRAIL since its discovery,

the molecular mechanisms underlying TRAIL signalingremain poorly understood. Moreover, there have been pre-cious few studies regarding the relationship betweenERK1/2 and mitochondrial events during TRAIL-inducedapoptosis.

Extracellular signals are transduced into the cells via acomplex network of signaling pathways. SpeciWcity of thecellular response is determined by an equilibrium which isreached between a host of distinct pathways. In recentyears, several reports have demonstrated that TRAIL isable to activate MAP kinase pathways, associated withJNK/SAPK (Nakshatri et al., 2004) and p38 MAP kinase(Lee et al., 2002). These pathways have also been implicatedin cellular activities such as stress, and the apoptoticresponse. In addition, ERK1/2 stimulation as the result of avariety of factors has been conWdently implicated in theregulation of a host of cellular physiological processes. Forexample, previous studies have demonstrated the pro-apop-totic role of ERK1/2 in glutamate- and hydrogen peroxide-induced apoptosis (Bhat and Zhang, 1999; Stanciu et al.,2000). Conversely, other studies have reported that ERK1/2activates cytoprotective machinery against serum depriva-tion-induced apoptosis in PC12 cells (Kim et al., 2000), andagainst Fas-induced cytotoxicity in Jurkat cells (Wilsonet al., 1999). In this study, we have demonstrated thatERK1/2 activation is quite active in the mechanisms under-lying protection against TRAIL-induced signaling. Wehave also shown that TRAIL-induced ERK1/2 activation isinhibited by pretreatment with PD98059, a speciWc ERK1/2

Fig. 4. EVect of PD98059 on TRAIL-induced collapse of mitochondrial membrane potential. HeLa cells were left untreated (A) or treated with 4 ng/ml ofTRAIL for 6 h (B) or 12 h (C). Cells were incubated with 10 �M of PD98059 for 2 h, followed by treatment with 4 ng/ml of TRAIL for 12 h (D) or no treat-ment (E). The collapse of mitochondrial membrane potential induced by TRAIL was measured by incubation with Xuorescent probe TMRE. (F) Quanti-Wcation of Xuorescent levels (results of Wgure (A)–(F)). Fluorescent levels are expressed as arbitrary units of relative value. Values are mean § SEM ofthree separate experiments.

822 D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823

inhibitor, and that this inhibition can be correlated withincreases in TRAIL-induced cell death. These results indi-cate that the activation of ERK1/2 is involved in theTRAIL signaling pathway, contributing to cellular defensesagainst exposure to several cytotoxic agents, includingTRAIL. Also, we obtained similar results using anotherERK pathway inhibitor, U126 (data not shown).

ERK1/2 is known to aVect the expression of severalgenes, in which are associated with a variety of cellular con-ditions. Recent reports have indicated that activatedERK1/2 regulates the expression of Bcl-2 and Bcl-XL, andpromotes the survival of a variety of human tumor cells(Boucher et al., 2000; Desire et al., 2000). In this study, weshow that the inhibition of ERK1/2 activities results in thedown-regulation of the expression levels of anti-apoptoticBcl-2 protein. However, Bcl-2 protein overexpression com-pletely blocks TRAIL-induced cytotoxicity, thereby facili-tating the release of cytochrome c (Sun et al., 2001), and themitochondria-mediated caspase signal transduction path-way (Fulda et al., 2002). Our results also indicate that theinhibition of activated ERK1/2 results in a signiWcantincrease in the breakdown of mitochondrial membranepotential, the release of cytochrome c from the mitochon-dria, and the increase of caspase activation, as well as adown-regulation in the expression levels of anti-apoptoticBcl-2 protein. These results suggest that ERK1/2 activation

Fig. 5. EVect of PD98059 on the release of cytochrome c by TRAIL. (A)HeLa cells were left untreated (CTL) or treated (TRAIL) with 4 ng/ml ofTRAIL for 12 h. Cells were incubated with 10 �M of PD98059 for 2 h, fol-lowed by treatment with 4 ng/ml of TRAIL for 12 h (PD + TRAIL) or notreatment (PD10). (B) QuantiWcation of Wgure (A) results. The release ofcytochrome c into the cytoplasm was detected by Western blotting withanti-cytochrome c antibody (cytochrome c). The same samples wereprobed with �-actin antibody as a loading control (�-actin). The levelsof cytochrome c were used as an internal standard in the quantiWcation of�-actin. The levels of cytochrome c are expressed as arbitrary units ofrelative value. Values are mean § SEM of three separate experiments.

attenuates cytotoxicity via the suppression of TRAIL-induced mitochondrial malfunctions. On the other hand,other recent studies showed no changes in ��m coincidentwith cytochrome c release (Annis et al., 2001; Gonzalez-Polo et al., 2003), or Bcl-2 family promotes cytochrome crelease, without ��m reduction as has been reported(Antonsson et al., 1997). However, we observed reductionin ��m coincident with cytochrome c release by TRAILand two mitochondrial events are regulated by ERK1/2, asa survival mechanism to struggle against TRAIL-inducedstress condition in early stage. The details of the mechanismby which ERK1/2 regulates the up-regulation of Bcl-2 pro-tein, or by which it modulates the mitochondrial pathway,remain to be elucidated.

In summary, this report is the Wrst to demonstrate that theERK1/2 activation is a kind of survival mechanism to strug-gle against TRAIL-induced stress condition in early stage,via the up-regulation of Bcl-2 protein levels, and by inhibit-ing mitochondrial malfunctions like as ��m, cytochrome crelease, caspase 9 activation during TRAIL-induced apopto-sis. Understanding the intracellular mechanisms underlyingTRAIL-induced apoptosis may facilitate the optimization ofTRAIL as a therapeutic agent in cancer.

Acknowledgements

This research was supported by grants from the Bio-Green 21 Project, Rural Development Administration ofthe Korean government and Ministry of Commerce, Indus-try and Energy (10009293).

Fig. 6. EVect of PD98059 on TRAIL-induced caspase 9 activation. Thecaspase-9 activity was detected using Xuorogenic substrate, Ac-LEHD-AMC (Section 2) with TRAIL (4 ng/ml). The eVect of pretreatment ofeither PD98059 (10 �M) or Ac-LEHD-CHO (10 �M) was represented.For caspase 9 activity, 20 �g of cellular extracts incubated with 0.5 mMAc-LEHD-AMC in 10 �l of total volume at 37 °C for 30 min. Excitationat 380 nm and emission at 460 nm were measured with a Xuorometer. Thedata are relative value compared with vehicle treated value. The data arethe mean § SEM of three separate experiments. ¤ D p < 0.05, comparedwith vehicle alone, ¤¤ D p < 0.05, compared with TRAIL alone.

D.Y. Lee et al. / Toxicology in Vitro 20 (2006) 816–823 823

References

Annis, M.G., Zamzami, N., Zhu, W., Penn, L.Z., Kroemer, G., Leber, B.,Andrews, D.W., 2001. Endoplasmic reticulum localized Bcl-2 preventsapoptosis when redistribution of cytochrome c is a late event. Onco-gene 12, 1939–1952.

Antonsson, B., Conti, F., Civatta, A., Montessuit, S., Lewis, S., Martinou,I., Bernasconi, L., Bernard, A., Mermod, J.J., Mazzei, G., Maundrell,K., Gambale, F., Sadoul, R., Martinou, J.C., 1997. The Bcl-2 proteinfamily. Science 18, 370–372.

Bhat, N.R., Zhang, P., 1999. Hydrogen peroxide activation of multiplemitogen-activated protein kinases in an oligodendrocyte cell line: roleof extracellular signal-regulated kinase in hydrogen peroxide-inducedcell death. J. Neurochem. 72, 112–119.

Boucher, M.J., Morisset, J., Vachon, P.H., Reed, J.C., Laine, J., Rivard, N.,2000. MEK/ERK signaling pathway regulates the expression of Bcl-2,Bcl-X(L), and Mcl-1 and promotes survival of human pancreatic can-cer cells. J. Cell. Biochem. 79, 355–369.

Cain, K., 2003. Chemical-induced apoptosis: formation of the Apaf-1apoptosome. Drug Metab. Rev. 35, 337–363.

Chang, L., Karin, M., 2001. Mammalian MAP kinase signalling cascades.Nature 410, 37–40.

Desire, L., Courtois, Y., Jeanny, J.C., 2000. Endogenous and exogenousWbroblast growth factor 2 support survival of chick retinal neurons bycontrol of neuronal Bcl-x(L) and Bcl-2 expression through a Wbroblastberowth factor receptor 1- and ERK-dependent pathway. J. Neuro-chem. 75, 151–163.

Domen, J., Gandy, K.L., Weissman, I.L., 1998. Systemic overexpression ofBcl-2 in the hematopoietic system protects transgenic mice from theconsequences of lethal irradiation. Blood 91, 2272–2282.

Fulda, S., Meyer, E., Debatin, K.M., 2002. Inhibition of TRAIL-inducedapoptosis by Bcl-2 overexpression. Oncogene 21, 2283–2294.

Gomez-Santos, C., Ferrer, I., Reiriz, J., Vinals, F., Barrachina, M., Ambro-sio, S., 2002. MPP+ increases alpha-synuclein expression and ERK/MAP-kinase phosphorylation in human neuroblastoma SH-SY5Ycells. Brain Res. 935, 32–39.

Gonzalez-Polo, R.A., Soler, G., Alonso, J.C., Rodriguez-Martin, A., Fuen-tes, J.M., 2003. MPP(+) causes inhibition of cellular energy supply incerebellar granule cells. Neurotoxicology 24, 219–225.

Jiang, X., Wang, X., 2000. Cytochrome c promotes caspase-9 activation byinducing nucleotide binding to Apaf-1. J. Biol. Chem. 275, 31199–31203.

Kelley, S.K., Ashkenazi, A., 2004. Targeting death receptors in cancer withApo2L/TRAIL. Curr. Opin. Pharmacol. 4, 333–339.

Keogh, S.A., Walczak, H., Bouchier-Hayes, L., Martin, S.J., 2000. Failureof Bcl-2 to block cytochrome c redistribution during TRAIL-inducedapoptosis. FEBS Lett. 471, 93–98.

Kim, M.S., So, H.S., Park, J.S., Lee, K.M., Moon, B.S., Lee, H.S., Kim, T.Y.,Moon, S.K., Park, R., 2000. Hwansodan protects PC12 cells againstserum-deprivation-induced apoptosis via a mechanism involving Rasand mitogen-activated protein (MAP) kinase pathway. Gen. Pharma-col. 34, 227–235.

Knight, M.J., RiVkin, C.D., Muscat, A.M., Ashley, D.M., Hawkins, C.J.,2001. Analysis of FasL and TRAIL induced apoptosis pathways in gli-oma cells. Oncogene 13, 5789–5798.

Lee, M.W., Park, S.C., Yang, Y.G., Yim, S.O., Chae, H.S., Bach, J.H., Lee,H.J., Kim, K.Y., Lee, W.B., Kim, S.S., 2002. The involvement ofreactive oxygen species (ROS) and p38 mitogen-activated protein(MAP) kinase in TRAIL/Apo2L-induced apoptosis. FEBS Lett. 512,313–318.

Matsuda, T., Almasan, A., Tomita, M., Uchihara, J.N., Masuda, M., Ohs-hiro, K., Takasu, N., Yagita, H., Ohta, T., Mori, N., 2005. Resistance toApo2 ligand (Apo2L)/tumor necrosis factor-related apoptosis-induc-ing ligand (TRAIL)-mediated apoptosis and constitutive expression of

Apo2L/TRAIL in human T-cell leukemia virus type 1-infected T-celllines. J. Virol. 79, 1367–1378.

Merienne, K., Jacquot, S., Zeniou, M., Pannetier, S., Sassone-Corsi, P.,Hanauer, A., 2000. Activation of RSK by UV-light: phosphorylationdynamics and involvement of the MAPK pathway. Oncogene 19,4221–4229.

Nagata, Y., Todokoro, K., 1999. Requirement of activation of JNK andp38 for environmental stress-induced erythroid diVerentiation andapoptosis and of inhibition of ERK for apoptosis. Blood 94, 853–863.

Nakshatri, H., Rice, S.E., Bhat-Nakshatri, P., 2004. Antitumor agent par-thenolide reverses resistance of breast cancer cells to tumor necrosisfactor-related apoptosis-inducing ligand through sustained activationof c-Jun N-terminal kinase. Oncogene 23, 7330–7344.

Park, S.A., Park, H.J., Lee, B.I., Ahn, Y.H., Kim, S.U., Choi, K.S., 2001. Bcl-2 blocks cisplatin-induced apoptosis by suppression of ERK-mediatedp53 accumulation in B104 cells. Brain Res. Mol. Brain Res. 93, 18–26.

Purring-Koch, C., McLendon, G., 2000. Cytochrome c binding to Apaf-1:the eVects of dATP and ionic strength. Proc. Natl. Acad. Sci. USA 97,11928–11931.

SchaeVer, H.J., Weber, M.J., 1999. Mitogen-activated protein kinases: speciWcmessages from ubiquitous messengers. Mol. Cell. Biol. 19, 2435–2444.

She, Q.B., Chen, N., Dong, Z., 2000. ERKs and p38 kinase phosphorylatep53 protein at serine 15 in response to UV radiation. J. Biol. Chem. 275,20444–20449.

Srivastava, R.K., Sasaki, C.Y., Hardwick, J.M., Longo, D.L., 1999. Bcl-2-mediated drug resistance: inhibition of apoptosis by blocking nuclearfactor of activated T lymphocytes (NFAT)-induced Fas ligand tran-scription. J. Exp. Med. 190, 253–265.

Stanciu, M., Wang, Y., Kentor, R., Burke, N., Watkins, S., Kress, G., Rey-nolds, I., Klann, E., Angiolieri, M.R., Johnson, J.W., DeFranco, D.B.,2000. Persistent activation of ERK contributes to glutamate-inducedoxidative toxicity in a neuronal cell line and primary cortical neuroncultures. J. Biol. Chem. 275, 12200–12206.

Sun, S.Y., Yue, P., Zhou, J.Y., Wang, Y., Choi Kim, H.R., Lotan, R., Wu,G.S., 2001. Overexpression of BCL2 blocks TNF-related apoptosis-inducing ligand (TRAIL)-induced apoptosis in human lung cancercells. Biochem. Biophys. Res. Commun. 280, 788–797.

Takahashi, M., Saito, H., Okuyama, T., Miyashita, T., Kosuga, M., Sumisa,F., Yamada, M., Ebinuma, H., Ishii, H., 1999. Overexpression of Bcl-2protects human hepatoma cells from Fas-antibody-mediated apopto-sis. J. Hepatol. 31, 315–322.

Tran, S.E., Holmstrom, T.H., Ahonen, M., Kahari, V.M., Eriksson, J.E.,2001. MAPK/ERK overrides the apoptotic signaling from Fas, TNF,and TRAIL receptors. J. Biol. Chem. 11, 16484–16490.

Tsai, E.Y., Falvo, J.V., Tsytsykova, A.V., Barczak, A.K., Reimold, A.M.,Glimcher, L.H., Fenton, M.J., Gordon, D.C., Dunn, I.F., Goldfeld,A.E., 2000. A lipopolysaccharide-speciWc enhancer complex involvingEts, Elk-1, Sp1, and CREB binding protein and p300 is recruited to thetumor necrosis factor alpha promoter in vivo. Mol. Cell. Biol. 20, 6084–6094.

Tsukada, Y., Miyazawa, K., Kitamura, N., 2001. High intensity ERK sig-nal mediates hepatocyte growth factor-induced proliferation inhibitionof the human hepatocellular carcinoma cell line HepG2. J. Biol. Chem.276, 40968–40976.

Walczak, H., Miller, R.E., Ariail, K., Gliniak, B., GriYth, T.S., Kubin, M.,Chin, W., Jones, J., Woodward, A., Le, T., Smith, C., Smolak, P., Good-win, R.G.., Rauch, C.T., Schuh, J.C., Lynch, D.H., 1999. Tumoricidalactivity of tumor necrosis factor-related apoptosis-inducing ligandin vivo. Nat. Med. 5, 157–163.

Wilson, D.J., Alessandrini, A., Budd, R.C., 1999. MEK1 activation rescuesJurkat T cells from Fas-induced apoptosis. Cell. Immunol. 194, 67–77.

Yagita, H., Takeda, K., Hayakawa, Y., Smyth, M.J., Okumura, K., 2004.TRAIL and its receptors as targets for cancer therapy. Cancer Sci. 95,777–783.