Introduction

Recent advances in understanding the molecularmechanism(s) underlying cell death and the identifi-cation of its associated genes have created a widevariety of opportunities for investigating the rolesplayed by these newly identified genes in physiolog-ical and pathological cell death. Among them, Bcl-2,a 26 KDa integral membrane protein, was firstdemonstrated to enhance cell survival by inhibitingapoptosis examplified by membrane shrinkage, bleb-bing, nuclear condensation and DNA fragmenta-tion.1–3 Subsequently, Bcl-2 has been shown to bepart of a growing family, of which can be function-ally subdivided into two categories.4 Some membersof the family, including Bcl-2, exert negative regula-tory effects on cell death induced by a variety ofstimuli whereas others such as Bax, Bad, Bak andBik act as positive regulators of cell death.

A striking feature among Bcl-2 family members is their ability to dimerize. Bax heterodimerizes with Bcl-2 through the so-called Bcl-2 homologydomains.5 The ratio of Bcl-2 to Bax appears to be akey factor in determining the fate of cells whenconfronted with an apoptotic stimulus. Therefore, anexcess of Bax counters the protective effects of Bcl-2 resulting in cell death. Many ensuing studies

support this notion by demonstrating that varioustypes of cellular injury are followed by upregulationof Bax gene expression.6–8 On the other hand, a seriesof recent studies has suggested that Bax may not beessential for determining the anti-cell death activityof the Bcl-2 family9 or that it may actually possess adeath-repressing function.10–11 These contradictingresults raise the possibility that the functional roleand the mechanism by which Bax exerts its actionmay be far more complicated than a simple relation-ship between pro- and anti-apoptotic family mem-bers. In particular, these opposing effects of Bax oncell life and death may depend upon the type of celland the specific cell death-inducing stimulusinvolved.

In order to test this hypothesis and to extend theseobservations into the CNS, we employed a well-defined murine dopaminergic neuronal cell line,MN9D.12–13 This cell line was shown to be very stableafter more than 20 passages in culture as demon-strated by chromosomal analysis.12 Stable MN9D celllines over-expressing murine Bax were first estab-lished using a very early passage of MN9D cells. Cellswere then treated with either staurosporine ornigericin, well known cell death-inducing stimuli.Subsequently, we examined the influence of over-expressed Bax on the death processes induced in

Molecular Neuroscience

1

1

1

1

1

p

© Rapid Science Publishers Vol 8 No 8 27 May 1997 1851

BAX, a member of the Bcl-2 multigene family, is knownto promote apoptosis. To investigate the role of Bax inan experimentally induced cell death of the murinedopaminergic neuronal cell line (MN9D), we establishedMN9D cells stably over-expressing murine Bax (MN9D/Bax) or vector alone (MN9D/Neo). In MN9D/Neo cellstreated with either 1 mM staurosporine or 0.1 mMnigericin, a ladder pattern of DNA fragmentation wasinduced. As expected, over-expression of Bax in MN9Dcells accelerated staurosporine-induced cell death asmeasured by the MTT reduction assay (62.3% survivalin MN9D/Neo vs 27.0% survival in MN9D/Bax). Sur-prizingly, both nigericin-induced cell death and itsaccompanying DNA fragmentation were largely atten-uated in MN9D/Bax cells (22.0% survival in MN9D/Neovs 86.7% survival in MN9D/Bax). Similar patterns wereobserved in two other MN9D/Bax cell lines. Cleavage ofpoly(ADP-ribose)polymerase caused by nigericin wasgreatly attenuated in MN9D/Bax cells suggesting that,like Bcl-2, Bax suppresses nigericin-induced cell death by inhibiting the activation of cysteine proteases. Thus, our data imply that Bax acts as a negative or positiveregulator of cell death depending on the type of deathstimulus applied to the cell.

Key words: Apoptosis; Bax; Cysteine protease; Dopamin-ergic neuron; Neuronal cell death; Nigericin; Staurosporine

Bax acceleratesstaurosporine-induced but suppresses nigericin-inducedneuronal cell death

Jae H. Oh, Karen L. O’Malley,1Stanislaw Krajewski,2 John C. Reed2

and Young J. OhCA

Department of Biology, Yonsei UniversityCollege of Science, 134 ShincheondongSeodaemoonku, Seoul 120-749, Korea;1Department of Anatomy and Neurobiology,Washington University School of Medicine, St Louis, MO 63110; 2Burnham Institute,Program on Apoptosis and Cell DeathResearch, La Jolla, CA 92037, USA

CACorresponding Author

NeuroReport 8, 1851–1856 (1997)

MN9D cells by these drugs and explored some ofthe mechanism(s) involved. Part of this work hasappeared previously in abstract form.

Materials and Methods

Establishment of murine Bax-expressing stable celllines: A full-length murine Bax cDNA in Bluescriptwas obtained (Dr S.J. Korsmeyer, Washington Uni-versity, St Louis), double digested with EcoRI andEheI followed by blunt ending with S1 nuclease. It was subcloned into an EcoRI/S1 nuclease/CIP-treated eukaryotic expression vector containing cyto-megalovirus (CMV) major immediate early enhancer/promoter. MN9D cells (2 ´ 105/well) cultured on 50 mg/ml poly-D-lysine-coated six-well plates (Corn-ing) were transfected with either 2 mg CMV/Bax(MN9D/Bax) or CMV/Neo control vector (MN9D/Neo) using Lipofectamine (GIBCO). These trans-fected cells were cultivated in DMEM supplementedwith 10% fetal bovine serum (FBS; GIBCO) and 500 mg/ml G418 (complete culture medium; CCM)for 7–10 days. Subsequently, single G418-resistantMN9D/Bax colonies were picked and expanded inCCM. For MN9D/Neo, all G418-resistant cloneswere scraped and expanded for experiments. StableMN9D/Bax cells and MN9D/Neo cells were char-acterized by Western blot analysis using rabbitpolyclonal anti-murine Bax.6 Three positive clonesshowing high expression levels of murine Bax wereused for experiments.

Cell culture: MN9D/Bax or MN9D/Neo cells wereplated at a density of 4 ´ 104 cells/well on 50 mg/mlpoly-D-lysine coated 24-well culture dishes (Corn-ing) and maintained in CCM in 10% CO2/90% airat 37ºC for 3 days. The Media were subsequentlyswitched to chemically-defined N2 medium contain-ing 1 mM staurosporine (a non-specific protein kinaseinhibitor) or 0.1 mM nigericin (a potassium-protonionophore) and cultures were further incubated foran additional 12–24 h. The rate of cell survival was assessed using the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) reductionassay.14 Briefly, a final concentration of 1 mg/mlMTT was added to the medium and cells werecultured for 2 h as above. Formazan grains formedby dehydrogenase activity within the surviving cellswere solubilized with 20% SDS in 50% aqueousdimethylformamide overnight. The optical density ofthe dissolved grains was measured at 540 nm usingan ELISA Plate Reader (Molecular Devices). Valuesfrom each treatment were calculated as a percentagerelative to the non-treated matching control (100%survival).

Detection of DNA fragmentation pattern: Formeasuring DNA fragmentation, cells cultured on P-100 dishes at a density of > 5 ´ 106 cells weretreated with 1 mM staurosporine for 12 h or 0.1 mMnigericin for 24 h. Cultures were subsequentlywashed three times with PBS and soluble DNA washarvested by lysing cells in a buffer containing 0.5%Triton X-100, 5 mM Tris pH 7.4, and 20 mM EDTAfor 30 min on ice as described previously.2 Follow-ing microcentrifugation at 4ºC for 15 min, DNA was subjected to phenol/chloroform extraction andethanol precipitation. DNA precipitates were dis-solved in 20 ml 1´ TE and electrophoresed on 1.2%agarose gels. Gels were treated with 1 mg/ml DNase-free RNase (Boehringer Mannheim), stained withethidium bromide and then photographed on a UV-transilluminator using Polaroid 667 film.

Western blot analysis: For measuring the expressionlevel of Bax, cells from MN9D/Bax and MN9D/Neowere extensively washed with ice-cold PBS and lysedin a buffer containing 50 mM Tris pH 7.0, 2 mMEDTA, 1.0% Triton X-100, 2 mM PMSF plus 1´protease inhibitor (GIBCO) for 10 min with occa-sional trituration. Following microcentrifugation at4ºC for 15 min, the protein content of the resultingsupernatant was measured by Bradford assay using a Bio-Rad protein assay kit. Fifty micrograms ofprotein were separated on 12.5% SDS–PAGE gels,blotted on to prewet PVDF-nitrocellulose filter,processed for Western blot analysis, and developedusing ECL kit (Amersham). To elucidate the patternof poly(ADP-ribose)polymerase (PARP) cleavage,both MN9D/Neo and MN9D/Bax cells were treatedwith 0.1 mM nigericin for various time periods andthen processed for Western blot analysis as describedabove. For further normalization of the loadedproteins, the same blots were subjected to Westernanalysis for microtubule-associated protein 2 (MAP-2). Primary antibodies used were rabbit polyclonalanti-murine Bax (1:3000), mouse monoclonal anti-PARP (1:10 000; a generous gift from Dr Poirier) andmouse monoclonal anti-MAP 2 (1:100; BoehringerMannheim).

ResultsTo establish stable MN9D cells over-expressingmurine Bax, cells were transfected with either CMV/Bax or the parental vector CMV/Neo (designatedMN9D/Bax or MN9D/Neo, respectively). Amongrandomly picked 24 G418-resistant MN9D/Baxclones, we chose three clones which expressed thehighest level of Bax mRNA as determined by reversetranscription-polymerase chain reaction (not shown).For further analysis, MN9D/Neo and three MN9D/

J. H. Oh et al.

1

11

11

11

11

11

1p

1852 Vol 8 No 8 27 May 1997

Bax cell lines were subject to Western blot analysis.As shown in Fig. 1, MN9D/Neo demonstrated a lowendogeneous level of Bax protein with a mol. wt of21 KDa. In contrast, all three MN9D/Bax cell linesexpressed at least 10 times more protein thanMN9D/Neo, whereas the levels of MAP-2 weresimilar in all four cell lines. Even though early

passages of MN9D/Bax cell lines were used for thisstudy, these cell lines stably over-expressed Bax aftermore than 20 passages in culture.

To investigate the potential role of Bax in theMN9D dopaminergic neuronal cell line, bothMN9D/Neo and MN9D/Bax cells were treated witha known cell death-inducing drug, 1 mM stau-rosporine (a non-specific protein kinase inhibitor) or0.1 mM nigericin (a potassium-proton ionophore) forvarious times. Treatment of cells with staurosporineinduced neurite arborization within a few hours (notshown). This change was followed by shrinkage ofthe cytoplasmic membrane and nuclear condensationin both MN9D/Neo and MN9D/Bax cells, althoughthe magnitude of this effect was much greater in theMN9D/Bax cells. Staurosporine-induced cell deathwas not blocked in the presence of 0.1–2.0 mg/mlcycloheximide (not shown). Quantitation of thesurvival rate of MN9D/Neo and MN9D/Bax cells bythe MTT assay indicated that over-expression of Baxaccelerated staurosporine-induced cell death, partic-ularly at the 12 h time point (Fig. 2A; 62.3 ± 3.3%survival in MN9D/Neo vs 27.0 ± 4.4% survival inMN9D/Bax). The difference in the rate of cell deathbetween these cells was diminished following 18 h ofincubation as dead cells accumulated in both cultures(not shown).

In contrast to staurosporine, treatment of MN9D/Neo cells with 0.1 mM nigericin did not induceneurite arborization (not shown). Rather, it caused

Role of Bax in neuronal cell death

1

1

1

1

1

p

Vol 8 No 8 27 May 1997 1853

FIG. 1. Characterization of Bax-over-expressing MN9D cells. MN9Dcells were transfected using either CMVBax (MN9D/Bax) or CMVNeo(MN9D/Neo) and selected in culture medium containing 500 mg/mlG-418. Western blot analysis of MN9D/Neo (Lane 1) and three inde-pendent MN9D/Bax clones (Lanes 2–4) was performed as describedin the Materials and Methods using (A) polyclonal anti-murine Baxor (B) mouse monoclonal anti-MAP-2 as primary antibodies. Notethat all three Bax-transfected cells express much higher levels ofBax than MN9D/Neo while the levels of MAP-2 are constant in allfour cell lines.

FIG. 2. Role of over-expressed Bax in staurosporine- and nigericin-induced cell death in MN9D cells. MN9D/Neo and MN9D/Bax cells wereplated at 4 ´ 104 cells/24-well culture dishes and maintained in complete culture medium for 3 days. Cells were subsequently switched toserum-free N2 medium containing (A) 1 mM staurosporine for 12 h or (B) 0.1 mM nigericin for 24 h. The rate of cell survival was measuredusing the MTT reduction assay as described in the Materials and Methods. Values from each of the treatment were expressed as a per centover the non-treated control (100%). Each bar represents the mean ± s.e.m. from three independent experiments in triplicate. Percentagesurvivals of two other MN9D/Bax cell lines after staurosporine treatment were 31.4 ± 6.2% and 36.2 ± 4.1% whereas those after nigericintreatment were 78.8 ± 5.9% and 82.6 ± 7.4%, respectively.

the formation of vacuoles within the cytosol, nuclearcondensation and subsequent cell death in MN9D/Neo. Co-treatment with cycloheximide blockednigericin-induced cell death (not shown). Surpriz-ingly, over-expression of Bax greatly attenuatednigericin-induced cell death (Fig. 2B; 22.0 ± 5.1%survival in MN9D/Neo vs 86.7 ± 5.5% survival in MN9D/Bax). The protective effect of Bax wascomparable to that of Bcl-2, based on comparisonswith MN9D/Bcl-2 cells (not shown). Similar resultswere observed in all three MN9D/Bax cell lines, thusexcluding the possibility that these data may resultfrom clonal variability (not shown). Taken together,these data suggest that over-expression of Bax has atleast two opposing effects within MN9D cellsdepending upon the type of death stimuli applied.

To determine whether over-expressed Bax eitheraccelerated or attenuated the generation of inter-nucleosomal DNA fragmentation, both MN9D/Neoand MN9D/Bax cells were treated with 1 mM stauro-sporine or 0.1 mM nigericin and then assayed forDNA fragmentation. In corroboration with themorphological and quantitative cell viability des-cribed above, significantly more DNA fragmentationwas seen with the staurosporine-treated MN9D/

Bax cells than with MN9D/Neo cells (Fig. 3A). Incontrast, over-expression of Bax attenuated nigericin-induced DNA fragmentation (Fig. 3B).

Recent experiments have suggested that Bcl-2exerts its anti-apoptotic effect upstream of the acti-vation of the interleukin-1-b-converting enzyme(ICE) family. The latter have been shown to be celldeath-inducing cysteine proteases.15–17 To elucidatewhether activation of ICE family proteases isinvolved in nigericin-induced cell death and whetheroverexpression of Bax in MN9D cells affects thisprocess, we assessed the cleavage pattern of a deathsubstrate, poly(ADP-ribose)polymerase (PARP)18 inboth MN9D/Neo and MN9D/Bax cells afternigericin treatment. In MN9D/Neo cells, the 85 kDafragment of PARP appeared 16 h after treatment andwas still present at 24 h (Fig. 4A). However, as shownin Fig. 4B, over-expression of Bax greatly attenuatedthe cysteine protease-mediated cleavage of PARP.These data suggest that, like Bcl-2, Bax may suppressnigericin-induced apoptotic cell death by inhibitingcysteine protease activation.

DiscussionIncreased expression of Bax has been associated witha variety of cell death paradigms. However, recentstudies using bax knock-out mice raise the possibilitythat Bax may have a more diverse role in the processof cell death.11 For example, knock-out mice displaylymphoid hyperplasia consistent with the previouslyproposed role of Bax in promoting cell death.

J. H. Oh et al.

1

11

11

11

11

11

1p

1854 Vol 8 No 8 27 May 1997

FIG. 3. Patterns of DNA fragmentation induced by staurosporineor nigericin in MN9D/Neo and MN9D/Bax cells. Both MN9D/Neo andMN9D/Bax (>5 ´ 106 cells/P-100 Petri dish) were treated with (A) 1mM staurosporine for 12 h or (B) 0.1 mM nigericin for 24 h. Triton X-100-soluble DNA was electrophoresed on 1.2% agarose gels afterRNase digestion and then stained with ethidium bromide. (Lane 1)molecular size markers ranging from 713 bp down to 190 bp; (Lane 2) MN9D/Bax, non-treated; (Lane 3) MN9D/Neo, non-treated;(Lane 4) MN9D/Bax, treated and (Lane 5) MN9D/Neo, treated. Note that overexpression of Bax accerelates staurosporine-induced DNA fragmentation whereas it attenuates nigericin-induced DNAfragmentation.

FIG. 4. Over-expression of Bax attenuates PARP cleavage inducedby nigericin. MN9D/Neo and MN9D/Bax cells were maintained inCCM as described, switched to N2 medium containing 0.1 mMnigericin and then incubated for the time intervals indicated on top of the blots. Western blot analysis of (A) MN9D/Neo and (B) MN9D/Bax was performed using mouse monoclonal anti-PARPantibody.

Surprizingly, loss of Bax in these mice also causeshypoplasia of male germ cells, suggesting that Baxexpression is required to block apoptosis duringspermatogenesis. Moreover, over-expression of Baxhas been reported to promote the survival of sensoryand ciliary neurons deprived of neurotrophic factors.The latter finding expands the possible protective role of Bax into the neuronal development of theperipheral nervous system.10 The data presented here provide a further example of the potential dualrole played by Bax in the mesencephalon-derivedMN9D neuronal cells. Moreover, our findings extendprevious observations by demonstrating that thepotentially dual role of Bax is determined by the typeof cell death-inducing stimulus that is applied to thesame cell. Because our studies have been carried outwith murine Bax in a murine cell type, we can ruleout the possibility that evolutionary divergence in theBax gene might result in an unexpected protectiverole of Bax when over-expressed in a differentspecies.10

Like Bax, other members of the Bcl-2 multigenefamily, have been found to both promote and preventcell death. For example, Bak accelerates apoptosis insympathetic neurons, FL5.12 cell line and Rat-1fibroblasts whereas it prevents apoptotic cell deathin the EBV-transformed WI-L2-729HF2 cell line.19–21

Furthermore, Bcl-2, a well-known anti-apoptoticregulator, was shown to accelerate apoptosis in PC-12 cells and excessive expression of Bcl-2 in theretina of transgenic mice also induced apoptosis in vivo.22–23 We previously reported that Bcl-2 atten-uated MPP+-induced cell death, but significantlyaccelerated 6-hydroxydopamine-induced cell death inMN9D cells.13 Taken together, it is conceivable thatsome members of the Bcl-2 multigene family possessthe ability to control apoptosis both as positive andnegative regulators through far more complicatedmechanisms than previously suggested.

The mechanism(s) by which opposing roles of Baxare achieved within the same cell is yet to be deter-mined. Interestingly, several recent experiments haveindicated that post-translational modification of Bcl-2-related proteins may play an important role indetermining their diverse roles in the process of celldeath.24 However, we have not observed any signsof phosphorylation of Bax as determined by Westernblot analysis (not shown). In addition to phospho-rylation, Bax might interact functionally withdifferent downstream effector proteins in MN9Dcells when treated with staurosporine vs nigericin.Shaham and Horvitz, for example, have proposed thatthe dual protective and killer functions of the ced-9gene product (Bcl-2-like gene in Caenorhaboditiselegans) are mediated by inhibiting either the killerfunction of ced-4S or the protective function of ced-

4L, respectively.25 Since co-treatment of cells withcycloheximide, a protein synthesis inhibitor, rescuedonly nigericin-induced cell death but not stauro-sporine-induced cell death in MN9D cells, it isequally plausible that the types of factors activatedby these two distinct cell death stimuli and theirpossible interactions with Bax may be different inMN9D cells.

Cysteine proteases of the ICE family are requiredto activate the onset of apoptosis and Bcl-2 has been shown to block cell death by acting upstreamof ICE-like protease activation.15–17 Bax delayed thecleavage of poly(ADP-ribose)polymerase, a well-known substrate of ICE-like proteases, in the courseof nigericin-induced cell death. Therefore, we firstpropose that Bax protects MN9D cells from apop-tosis probably by acting at a point similar or iden-tical to Bcl-2. Since we have demonstrated thatover-expression of Bcl-2 in MN9D cells also inhibitsnigericin-induced cell death to the same extent as Bax (not shown), it will be interesting to elucidatewhether both Bcl-2 and Bax share the same mecha-nism throughout the course of nigericin-induced cell death. Careful studies using gene transfer to alterthe ratio of Bcl-2 and Bax may shed some light onthe mechanisms controlling nigericin-induced celldeath and provide insight into the dual role playedby Bax.

Because dysregulation of apoptosis has been spec-ulated to result in a variety of neurodegenerativediseases, studies of Bcl-2-related genes are expectedto contribute to the understanding of these disordersand possibly lead to the development of therapeuticstrategies. Among the many steps that could beimmediately taken in the future, it seems very impor-tant to elucidate why and how the Bcl-2 multigenefamily can have at least two opposite facets in theprocess of cell death regulation. Our dopaminergicsystem will provide a good CNS model to investi-gate the mechanism(s) underlying the dual role of some of the Bcl-2 multigene family membersincluding Bax in neuronal cell death.

ConclusionThe present study demonstrates that over-expressionof Bax both accelerates staurosporine-induced celldeath and suppresses nigericin-induced cell death ina dopaminergic neuronal cell line, MN9D. Theprotective role of Bax seems to be mediated byblocking the activation of a cysteine protease(s).Taken together, these results support the idea thatBax may play positive or negative roles as a regu-lator of cell life and death depending on the cellularcontext.

Role of Bax in neuronal cell death

1

1

1

1

1

p

Vol 8 No 8 27 May 1997 1855

References1. Vaux DL, Cory S and Adams J. Nature 335, 440–442 (1988).2. Hockenbery DM, Nunez G, Milliman C et al. Nature 348, 334–336 (1990). 3. Wyllie AH. Br J Cancer 67, 205–208 (1993).4. White E. Genes Dev 10, 1–5 (1996). 5. Oltvai ZN, Milliman CL and Korsmeyer SJ. Cell 74, 609–619 (1993).6. Krajewski S, Mai JK, Krajewska M et al. J Neurosci 15, 6364–6376 (1995).7. Bargou RC, Bommert K, Weinmann P et al. Eur J Immunol 25, 770–775

(1995).8. Chen J, Zhu RL, Nakayama M et al. J Neurochem 67, 64–71 (1996).9. Cheng EH, Levine B, Boise LH et al. Nature 379, 554–556 (1996).

10. Middleton G, Nunez G and Davies AM. Development 122, 695–701 (1996).11. Knudson CM, Tung KSK, Tourtellotte WG et al. Science 270, 96–99

(1995).12. Choi HK, Won LA, Kontur PJ et al. Brain Res 552, 67–76 (1991).13. Oh YJ, Wong SC, Moffat M et al. Neurobiol Dis 2, 157–167 (1995).14. Shearman MS, Ragan CI and Iversen LL. Proc Natl Acad Sci USA 91,

1470–1474 (1994).15. Miura M, Zhu H, Rotello R et al. Cell 75, 653–660 (1993).16. Boulakia CA, Chen G, Ng FWH et al. Oncogene 12, 529–535 (1996).

17. Chinnaiyan AM, Orth K, O’Rouke K et al. J Biol Chem 271, 4573–4576(1996).

18. Lazebnik YA, Kaufmann SH, Desnoyers S et al. Nature 371, 346–347 (1994).19. Farrow SN, White JHM, Martinou I et al. Nature 374, 731–733 (1995).20. Chittenden T, Harrington EA, O’Conner R et al. Nature 374, 733–736 (1995).21. Kiefer MC, Brauer MJ, Powers VC et al. Nature 374, 736–739 (1995).22. Cortazzo M and Schor NF. Cancer Res 56, 1199–1203 (1996).23. Chen J, Flannery JG, LaVail MM et al. Proc Natl Acad Sci USA 93, 7042–7047

(1996)24. Gajewski TF and Thompson CB. Cell 87, 589–592 (1996).25. Shaham S and Horvitz HR. Cell 86, 201–208 (1996).

ACKNOWLEDGEMENTS: We gratefully acknowledge Drs A. Heller and Lisa Wonfor providing us with the MN9D cell line. This work was supported by theGenetic Engineering Grants from the Ministry of Education, Korea (Y.J.O.), MH45530 (K.L.O.), and California Breast Cancer Research Program 1RB-0093 (J.R.).

Received 28 February 1997;accepted 4 April 1997

J. H. Oh et al.

1

11

11

11

11

11

1p

1856 Vol 8 No 8 27 May 1997

General SummaryBax, a member of the Bcl-2 gene family, is known to promote apoptosis. To date, the exact function of Bax in the central nervoussystem-derived neuronal cell death is not known. Therefore, we investigated the potential role of Bax in two distinct cell death para-digms using a murine dopaminergic neuronal cell line (MN9D). Here, we demonstrate that increased levels of Bax accelerate stau-rosporine-induced cell death but suppress nigericin-induced cell death. Inhibition of cysteine protease activation ascribes to one ofthe possible mechanisms by which Bax attenuates nigericin-induced cell death. Thus, our present work suggests that Bax controlsapoptotic cell death both as a positive and a negative regulator depending upon the cellular context. Furthermore, our dopaminergicsystem will provide a good model to evaluate the mechanism(s) underlying the role of the Bcl-2 gene family in neuronal cell death.This will eventually lead to a better understanding of the pathogenesis of Parkinson’s disease.