Valinomycin Induces Apoptosis of Ascites Hepatomacells (AH-130) in Relation to Mitochondrial Membranepotential

Home , Trypan blue

CELL STRUCTURE AND FUNCTION 22: 555-563 (1997) © 1997 by Japan Society for Cell Biology Valinomycin Induces Apoptosis of Ascites Hepatoma Cells (AH-130) in Relation to Mitochondrial MembranePotential Yoko Inai, Munehisa Yabuki*, Tomoko Kanno*2, Jitsuo Akiyama**, Tatsuji Yasuda, and Kozo Utsumi*1 Department of Cell Chemistry, Institute of Molecular and Cell Biology, Okayama University Medical School, Okayama 700, Japan, ^Institute of Medical Science, Kurashiki Medical Center, Kurashiki 710, Japan, and **Doonan Institute of Medical Science, Hakodate 041, Japan

Key words: AH-130 cells/valinomycin/mitochondrial membrane potential/Bcl-2 family/caspase-3

ABSTRACT.Valinomycin is a potassium ionophore, and is well knownto cause the collapse of the mitochondrial membranepotential. It has been reported that loss of mitochondrial membranepotential is observed in the early stages of apoptosis induced by various agents. Thus, the effects of valinomycin on tumor cells were examined. Valinomycin induced uncoupling of respiration and depolarization of isolated mitochondria. Depolariza- tion of intact mitochondria in AH-130rat ascites hepatoma cells was also induced by valinomycin. Valinomycin induced apoptosis revealing the typical apoptotic characteristics such as fragmentation and ladder formation of DNA,shrinkage of cells, and formation of pycnotic nucleus. There was a correlation between the depolarization of mitochondria and DNAfragmentation. After depolarization of mitochondria, the activity of caspase-3- like protease but not caspase-1-like protease increased markedly. In contrast, this apoptosis did not involve the release of reactive oxygen species from mitochondria, increase in intracellular calcium concentration, or protein synthesis. In addition, anti-apoptotic members of the Bcl-2 family (Bcl-xL and Bcl-2) were not correlated with apoptosis. These results indicate that valinomycin might induce apoptosis through degradation of the mitochondrial membranepotential. Taken together, these observations suggest that there maybe a mechanismthat trans- mits the signal from mitochondrial depolarization to subsequent apoptosis execution steps.

The removal of surplus cells plays an important role 26, 30). Recently, Shimizu et al. reported that Bcl-2 and in both cellular development and homeostasis. The elim- Bcl-xL protected mitochondria against the loss of mem- ination of such cells is accomplished by apoptosis, a brane potential during apoptosis and certain forms of well-defined mode of cell death with distinct biological necrotic cell death (25). Bcl-2 and Bcl-xL proteins have and morphological features. Misregulation of apopto- been to several intracellular membranesincluding those sis results in various diseases. Apoptosis is regulated by of mitochondria, the endoplasmic reticulum, and the a number of genes, with the bcl-2 gene family encoding nuclear membrane (2, 4, 9). In certain cells, however, several apoptosis-related proteins. Changes in mito- apoptosis was inhibited by Bcl-2 localized in the mitochondrial function have been shownto be involved in chondria but not in the endoplasmic reticulum (31). apoptotic cell death. The mitochondrial membrane These observations indirectly implicate the involvement potential, which is the driving force of mitochondrial of mitochondrial function in the mechanismof apopto- ATP synthesis, decreases during apoptosis, while the sis. Onthe other hand, it has been reported that the loss maintenance of the membranepotential prevents the of mitochondrial membranepotential is not related to apoptosis induced by various stimuli (6, 17, 18, 22, 25, apoptosis induced by certain stimuli (8). Thus, the molecular mechanismof the involvement of mitochondria 1 To whomcorrespondence should be addressed. in apoptotic cell death is not well understood. Tel & Fax: +81-86-426-8616 To determine whether mitochondrial dysfunction is a 2 On leave from the DoonanInstitute of Medical Science. Abbreviations used: AIF, apoptosis-inducing factor; AMC,7-ami- critical event in the apoptotic cascade, it is important to no-4-methyl-coumarin; CHX, cycloheximide; DEVD-CHO, acetyl- investigate the mechanisms of apoptosis induced by vari- Asp-Glu-Val-Asp-aldehyde; DMEM,Dulbecco's modified Eagle's me- ous agents that collapse the mitochondrial membrane dium; KRPbuffer, Krebs-Ringer phosphate buffer; Mit, mitochon- potential. Recently, Marchetti et al. reported that thym- dria; PBS, phosphate-buffered saline; PT, permeability transition; ocyte apoptosis was induced by addition of protopor- ROS,reactive oxygen species; Succ, sodium succinate; Val, valinomy- phyrin IX to the mediumwhich disrupted mitochondri- cin; YVAD-CHO, acetyl-Tyr-Val-Ala-Asp-aldehyde. al membranepotential via interaction with mitochondri- 555 Y. Inai et al. al benzodiazepin receptors and formation of mitochon- treated with 0.4mg/ml RNase A for 1 hour at 37°C and drial permeability transition (PT) pores. Furthermore, treated further by 0.4 mg/ml proteinase K under the same con- they reported that overexpression of Bcl-2 partially in- ditions. The solutions were precipitated at -20°C with 0.2 hibited the mitochondrial PT and apoptosis (17). volume of 5 MNaCl and 1 volume of isopropanol. DNAwas In this study, we investigated the mechanismof apop- pelleted by centrifugation and resuspended in TE (10 mM tosis induced by valinomycin, a potassium ionophore Tris-HCl, pH 7.6, and 1 mMEDTA) buffer. The DNAwas an- and an uncoupler of mitochondrial respiration. Valino- alyzed by electrophoresis in 2% agarose gels. mycin induced apoptosis of AH-130cells (rat ascites Quantification ofDNAfragmentation. The tumor cells (7 hepatoma) through the depolarization of mitochondria x 105 cells) were collected by centrifugation, and the pellets and activation of caspase-3-like protease. Although were dissolved with lysis buffer A and kept on ice for 15 min- Bcl-xL was up-regulated by valinomycin during apopto- utes. Samples were centrifuged at 15,000 rpm for 20 minutes, sis, cycloheximide, a protein synthesis inhibitor, had no and the DNAin supernatants and pellets was measured by the effect on DNAfragmentation. diphenylamine method as fragmented DNAand intact DNA, respectively (3). DNAfragmentation {%) was calculated by MATERIALS AND METHODS the following formula: Chemicals. Valinomycin and cycloheximide were pur- DNAfragmentation {%) = { fragmented DNA/(fragmented DNA chased from WakoCo. Ltd. (Japan) and Sigma Chemicals Co. (USA), respectively. Enhanced chemiluminescence de- +intact DNA)} x 100. tection system was obtained from AmershamCorp. (UK). Measurementof mitochondrial membranepotential in Antibodies were purchased from Santa Cruz Biotechnology cells. Rhodamine 123 was loaded into the cells by incubation (USA). Proteinase K, RNase A, and ATPbioluminescence in medium containing 10 ^M rhodamine 123 for 30 minutes. assay kit CLS II were purchased from Boehringer Mannheim Cells were washed three times with KRPbuffer by centrifuga- (Germany). Fluorogenic peptides and caspase family inhibition. Content of rhodamine 123 was measured after dissolu- tors were obtained from Peptide Institute Inc. (Japan). All tion of cells by addition of 0. \% Triton X-100. Fluorospectro- other reagents were of analytical grade and purchased from metry (Hitachi 650-10LC) was performed at excitation and Nacalai Tesque Co. (Japan). emission wavelengths of 485 nmand 520 nm, respectively. Isolation of liver mitochondria. Male Donryu rats weigh- Western blotting. The tumor cells were washed with PBS ing 200 g were fasted overnight and their liver mitochondria twice, and dissolved in lysis buffer B (0.1% SDS, \% Nonidet were isolated by the method of Hogeboom(10) using medium P-40, 0.5% sodium deoxycholate, and 1 mMphenylmethylsul- containing 0.25 Msucrose, 10 mMTris-HCl, pH 7.4, and 0.1 fonyl fluoride). The cell lysate was separated by SDS-PAGE mMEDTA.The mitochondrial preparations were washed using 10 or 12.5% acrylamide gels and transferred to PVDF twice with respiratory medium (154 mMKC1 containing 5 membranes. After blocking with 5% skimmed milk over- mMK2HPO4, 5 mM KH2PO4, pH 7.4, 3 mM MgCl2, and 0.1 night, the membraneswere incubated with anti-Bcl-xL, or mMEDTA)and resuspended in the same medium. -Bcl-2 antibody. The blots were then, hybridized with second- Preparation ofascites hepatoma cells. About 0.5 ml of as- ary antibody conjugated with horseradish peroxidase. Specific cites containing exponentially growing AH-130cells was inoc- bands for Bcl-xL and Bcl-2 were visualized with an enhanced ulated into the peritoneal cavity of male Donryu rats. After 6- chemiluminescence detection system. 8 days, cells were harvested from the peritoneal cavity. The tu- Hydrogen peroxide measurement. Hydrogen peroxide re- mor cell suspension was hemolyzed with 0.2% NaCl and lease by mitochondria was measured fluorometricaHy by moni- washed by centrifugation in calcium-free Krebs-Ringer phos- toring the oxidation of /7-hydroxyphenylacetic acid coupled phate (KRP) buffer. Membrane integrity was measured by with the enzymatic reduction of hydrogen peroxide using trypan blue dye exclusion. horseradish peroxidase (1 1). The reaction mixture consisted Induction of apoptosis. AH-130cells were suspended in of 0. 15 mg protein/ml rat liver mitochondria, 0.5 mg/mlp-hy- Dulbecco's modified Eagle's medium (DMEM) containing droxyphenylacetic acid, 4 U/ml horseradish peroxidase, and 20% heat-inactivated fetal bovine serum at 1 x 106 cells/ml 7.5 mMsodium succinate in respiratory medium. The oxida- and precultured for 1-3 hours. Valinomycin was added as tion of/7-hydroxyphenylacetic acid was measuredat an excita- ethanol solution and the final concentration of vehicle did not tion wavelength of 320 nmand emission wavelength of 400 exceed 0.1%. Preculture and treatment of the cells werecar- nm. Standard curves were prepared from authentic hydrogen ried out in a humidified incubator at 37°C under a 5% CO2 peroxide. /95% air atmosphere. Preincubation with caspase inhibitors or intracellular calci- DNAextraction and electrophoresis. The tumor cells (1 x umchelator. The tumor cells were suspended in DMEMcon- 106 cells) were washed with phosphate-buffered saline (PBS) taining 20%fetal bovine serum and precultured for 2 hours. and solved in lysis buffer A (10 mMTris-HCl, pH 7.4, 10 mM Thecells were then, preincubated with caspase inhibitors or EDTA, and 0.5% Triton X-100). Soluble fractions were BAPTA-AM(intracellular calcium chelator) for 1 hour be-

556 Valinomycin-induced Apoptosis of AH-130 Cells

fore the addition of valinomycin. Acetyl-Tyr-Val-Ala-Asp-al- A dehyde (YVAD-CHO) and acetyl-Asp-Glu-Val-Asp-aldehyde (DEVD-CHO)were used as inhibitors for caspase-1 and caspase-3, respectively. Measurement ofcaspasefamily activity. Tumor cells incu- 1 2 bated in the presence or absence of valinomycin were collected Mrt Succ Mit Succ by centrifugation and the pellets were resuspended in lysis --v ^ VADP buffer C (50mM Tris-HCl, pH 7.5, 0.5% Nonidet P-40, 0.5 200- Ná" N* mMEDTA, and 150mMNaCl) (24). The suspensions were then, kept on ice for 1 hour and centrifuged at 15,000 rpm for 10 minutes. Soluble fractions were diluted with reaction buff- 3 \ \ er (20mM HEPES, pH7.5, 0.1 M NaCl, and 5mM dithio- LJJ \ \ threitol) and incubated with fluorogenic peptide substrates gioo- \ \ (Acetyl-YVAD-4-methyl-coumaryl-7-amide (YVAD-MCA) specific for caspase-1 or DEVD-MCAspecific for caspase-3) at a final concentration of 10 //M for 1 hour at 37°C. An equal volume of stop solution (0.2M glycine-HCl, pH2.8) was added to samples. The fluorescence of released AMC(7-ami- ol i i i ^-j i-^ no-4-methyl-coumarin) was measured at an excitation wave- 0 4 8 12 length of 355 nmand emission wavelength of 460nm. One TIME (min) unit was denned as the enzyme activity that liberated 1 //mole B of AMCduring 1 hour.

RESULTS Effects of valinomycin on the membranepotential g so, . . , , and respiration of rat liver mitochondria. Isolated intact mitochondria showedslow oxygen consumption in 140" /9^-^_ the presence of sodiumsuccinate as a respiratory sub- jy / ^^^Sl^^--- 0.5 nM Val strate (state 4 respiration). Mitochondria generated an Q I f ~^1nMVal electrochemical proton gradient across the inner mem- Z / I brane by substrate oxidation. The gradient formed a q 30å 7 j ~"~Control - membrane potential and drove ATPsynthesis. In the presence of both substrate and ADP,ATPsynthesis UJ was coupled with substrate oxidation (state 3 respira- oc tion), revealing accelerated oxygen consumption (Fig. 3o 20l 1 1 1 1A). Whenvalinomycin, a potassium ionophore, was "å 0 3 6 9 12 added to mitochondrial suspensions, the membranepo- TIME (min) tential decreased (Fig. IB), and respiration was uncoupled, resulting in an increased oxygen consumption Fig. 1. Effects of valinomycin on the function of isolated rat liver (Fig. 1A). Respiratory control ratio of mitochondria mitochondria. A. Valinomycin induced uncoupled respiration. Oxy- was more than 3.9 in all experiments with sodium succi- gen consumption by mitochondria was measured polarographically at nate (7). 37°C using a Clark type oxygen electrode. Isolated mitochondria (0.5 mgprotein/ml) were suspended in respiratory medium.1. Sodiumsuc- Apoptosis of AH-130 cells induced by valinomycin. cinate induced slow oxygen consumption (state 4 respiration). Valino- WhenAH-130 cells were incubated with 100 nMvalino- mycin (1 nM) caused uncoupled respiration, increasing oxygen con- mycin, DNAfragmentation and ladder formation were sumption in the absence of ADP.2. Respiratory control was observed induced within 16 hours. Appearance of these character- by addition of ADP(state 3 respiration). Mit, 0.5 mg protein/ml mitochondria; Succ, 7.5 mMsodium succinate; Val, 1 nM valinomycin; istics preceded loss of membraneintegrity as assessed ADP, 200 [iM ADP. B. Depolarization of mitochondria was induced by the trypan blue dye exclusion method (Fig. 2). These by valinomycin in respiratory mediumcontaining 7.5 mMsodium suc- effects were observed in AH-130cells treated with over cinate and 250 ng/ml cyanine dye (diS-C3-(5)). The membrane poten- 1 nMvalinomycin. Morphological features of apopto- tial was monitored by changes in fluorescence intensity of diS-C3-(5) sis, such as cell shrinkage and pycnotic nuclei were also by a fluorospectrometer using an excitation wavelength of 622 nmand observed (data not shown). emission wavelength of 670 nm. Mitochondria were suspended in re- Relation between DNAfragmentation and loss of spiratory mediumat 0.1 mgprotein/ml. mitochondrial membrane potential in AH-130 cells

557 Y. Inai et al. A B

^80| 1 ^- -O-Control 0hr24hr | 6j) _ -æf- 100nMVal o

W 60 - z Hi40- / ^40- £20- / 00 20 - UJ 2 0| | | 0 8 16 24 0 8 16 24 TIME (hr) TIME (hr) Fig. 2. Apoptosis of AH-130 cells by valinomycin. AH-130 cells were cultured in DMEMcontaining 20%fetal bovine serum at 1 x 106 cells /ml. Valinomycinwas added to the mediumas an ethanol solution at a concentration of 100 nM.A. Effects of valinomycin on integrity of cytoplasmic membrane of AH-130cells. The integrity of the cytoplasmic membrane of the cells was assessed by the trypan blue dye exclusion method. Data are the means ± s.d. of 3-5 separate experiments. B. DNAfragmentation induced by 100 nMvalinomycin. DNAfragmentation was measured by the diphenylamine method using 7 x 105 cells. Data are the means±s.d. of 3-5 separate experiments. C. DNAladder formation induced by valinomycin. The cells were incubated with 100 nM valinomycin for the indicated periods. The cells were dissolved in lysis buffer A, and DNA in the soluble fraction was collected as fragmented DNA.Then, the DNAwas analyzed by electrophoresis using 2% agarose gels. Each well con- tained fragmented DNAfrom 3 x 105 cells. treated with valinomycin. To assess the changes in AH-130cells. The cells treated with 100 nMvalinomy- mitochondrial membrane potential in intact cells, fluo- cin showed a significant four fold increase in the level rescence intensity of rhodamine 123 was measured. Rho- of Bcl-xL (Fig. 5B). Taken together, these suggest that damine 123 taken up into the mitochondria of intact such apoptosis occurred independently of both the in- cells reflects differences in the magnitude of mitochon- crease in expression of pro-apoptotic membersof the drial membrane potential (13). Relatively high concen- Bcl-2 protein family and decreases in that of anti-apop- trations of valinomycin (100 nM) decreased mitochon- totic members. drial membranepotential rapidly, and resulted in more Noinvolvement of reactive oxygen species and intra- than 45% DNAfragmentation within 24 hours (Fig. cellular calcium concentration. Release of hydrogen 3A). With a lower concentration of valinomycin (1 peroxide from mitochondria was measured, because su- nM), the decrease in mitochondrial membranepotential peroxide is converted to hydrogen peroxide by dismuta- was weak, and DNAfragmentation was less than 20% tion in the mitochondrial matrix. Valinomycin at 1 nM (Fig. 3B). These results indicated a weak correlation be- inhibited the hydrogen peroxide generation by isolated tween the loss of membranepotential and the extent of mitochondria in the presence of 7.5 mMsodium succi- DNAfragmentation in AH-130 cells treated with valino- nate as the respiratory substrate (Table I), This concen- mycin (1-1,000 nM) (Fig. 4). tration of valinomycin (1 nM) was sufficient to cause Protein synthesis and changes in level of Bcl-xL ex- the depolarization and uncoupling of respiration in iso- pression in apoptosis. Weexaminedthe effects of cyclo- lated mitochondria (Fig. 1). heximide on valinomycin-induced apoptosis of AH- To investigate the effects of cytoplasmic calcium 130 cells. The degree of DNAfragmentation was not on apoptosis of AH-130 cells, a cytoplasmic calcium affected by cycloheximide (Fig. 5A) indicating that pro- chelator (BAPTA-AM)was loaded into the cells before tein synthesis was not required in the process of valino- the addition of 100 nMvalinomycin. Consequently, mycin-induced apoptosis. DNAfragmentation was not suppressed by BAPTA- The levels of Bcl-xL and Bcl-2 expression in AH-130 AM(Fig. 6). In addition, intracellular calcium ion con- cells were assessed by western blotting. The expression centration, monitored by the fluorescence dye method, of Bcl-xL but not of Bcl-2 was detected in nontreated was increased slightly by 100 nM valinomycin (data not 558 Valinomycin-induced Apoptosis of AH-130 Cells A. 100 nM Valinomycin B. 1 nM Valinomycin

S1°°* 1 io° ^ gioo^_ . 100 ^ p80A -80§ p80- à"" #^-^(80 O

UJ \ Q < uj < h 60- \ A60 P= g60- -60 H Q. \. / LU Q. UJ z4oå \yå 4oS UJ40- -40 | g20- J)20 £ g20- / \-20 £ HIs \ s^ >cr N>> 2< 2 00^ ' 1 '0 Q 2 OQ--^ ^£- 10 Q 0 8 16 24 0 8 16 24 TIME (hr) TIME (hr) Fig. 3. Loss of mitochondrial membrane potential and DNAfragmentation induced by different concentrations of valinomycin. AH-130 cells were treated with 100 nM(A) or 1 nMvalinomycin (B). Mitochondrial membrane potential was assessed by measurement of rhodamine 123 taken up into mitochondria. Rhodamine123 in mitochondria was measured fruorometricaHy using an excitation wavelength of 485 nmand emission wavelength of 520 nm. Closed circles, mitochondrial membrane potential {% of control); open circles, DNAfragmentation {%). Other experimental conditions were the same as described in Fig. 2 except for concentrations of valinomycin. shown). (ICE) and caspase-3 (CPP32) were added in medium be- A ctivation of caspase-3-like protease during apopto- fore addition of 100 nMvalinomycin. Only the inhibi- sis induced by valinomycin. Inhibitors of caspase-1 tor of caspase-3 suppressed DNAfragmentation in a concentration-dependent manner (IC5O= IS ptM) (Fig. ^60 r^- 1 7). The activities of proteases were examined using fluorogenic substrates specific for each protease. In control t r\~. cells, activities of both proteases were very low. While o \. the activity of caspase-1-like protease did not change, 540 - \. caspase-3-like protease was activated markedly by 100 nMvalinomycin prior to DNAfragmentation (Fig. 8). Maximumactivation of caspase-3-like protease was ob- i .\t served at 16 hours after addition of valinomycin, and §20- \ the activity was 9.83 mU/105 cells. These results indicated that the activation of caspase-3-like protease was required for apoptosis. "å .f(x)=-0.586X+64.0 à" \i j| R=0.812 ^ DISCUSSION Qol 1 1 1 * Recently accumulated evidence has revealed the in- 0 25 50 75 1 00 volvement of mitochondrial function in apoptosis. In MEMBRANEPOTENTIAL (% of control) particular, a numberof studies have shownthat loss of Fig. 4. Relationship between the mitochondrial membrane poten- mitochondrial membranepotential occurs in the early tial and DNAfragmentation. AH-130 cells were incubated with var- stages of apoptosis in various systems (17, 22, 25, 26). ious concentrations of valinomycin. After 8 hours, mitochondrial Therefore, we investigated the apoptosis induced by an membranepotential {%of control) was measured. Twelve hours lat- agent that decreases mitochondrial membrane poten- er, DNAfragmentation was measured. Valinomycin was used at 1- tial. Valinomycin, a potassium ionophore, has been 1 ,000 nM. Experimental conditions were the same as described in Fig. shownto cause depolarization and uncoupling of respir- 3. ation in mitochondria by increasing the membraneper- 559 Y. Inai et al. A B Bcl-xL g80| . Z^ -O- + CHX 0 ng/ml O -#- + CHX 1 ng/ml h 60 - ^ -A-+ CHX10 ng/ml t>\ -= t -æf-+CHX100ng/ml M UJ 40- j/

< 20- JT P ran EHI ^ ^ QQI^ ^ ^ ^ 0 8 16 24 Q 0 4 8 12 16 20 TIME (hr) TIME (hr) Fig. 5. Protein synthesis and changes in Bcl-xL during apoptosis. A. Effects of cycloheximide on DNAfragmentation. The indicated concentrations of cycloheximide and 100 nMvalinomycin were added simultaneously, and AH-130cells were incubated with these agents for the indicated periods. Other experimental conditions were the same as described in Fig. 2. CHX,cycloheximide. B. Western blotting analysis of Bcl-xL during apoptosis induced by valinomycin. AH-130 cells were treated with 100 nM valinomycin for the indicated periods. The proteins of whole-cell ly- sates were separated by SDS-PAGEand specific bands were visualized by western blotting. Specific bands were visualized using monoclonal antibody (upper). Relative expression of Bcl-xL is represented as relative intensity of each band (lower). meability to potassium ions. Valinomycin-induced de- of lymphocytes (17). Protoporphyrin IX is a ligand polarization of the mitochondrial membranewas also of the mitochondrial benzodiazepin receptor and well demonstrated in intact cells using rhodamine 123 taken knownfor its PT-triggering capacity. Valinomycin col- up into mitochondria (13). Valinomycin-induced apop- lapsed the mitochondrial membranepotential directly tosis of AH-130cells showedtypical characteristics of without PT pore formation, and induced apoptosis im- apoptosis such as DNAfragmentation, DNAladder for- mediately. High concentrations of valinomycin (over mation, cell shrinkage, and unclear pycnosis. To examine whether the loss of mitochondrial membrane potential is related to apoptosis, wemeasured the ^5 60 mitochondrial membrane potential and DNAfragmen- Z tation using various concentrations of valinomycin (l- g l,000 nM). There was a correlation betweenthe extent of mitochondrial depolarization and DNAfragmenta- <40 tion. This correlation is considered to imply a direct role for mitochondrial depolarization in apoptosis. Mar- LU chetti et al. have shownthat the formation of mitochon- O20 drial permeability transition (PT) pores by protoporphyrin IX and the subsequent loss of mitochondrial membranepotential are sufficient to induce apoptosis <

Table I. Release of hydrogen peroxide from isolated 0 0.5 5 50 LIVER MITOCHONDRIA. BAPTA-AM (pM) Hydrogen peroxide (pmole/^g protein/min) Fig. 6. Effects of intracellular calcium chelator (BAPTA-AM)on C ontrol 0.1730 the fragmentation of DNAinduced by valinomycin. AH-130 cells V alinom ycin 1 nM 0.0389 were preincubated with BAPTA-AMfor 1 hour. Then, valinomycin was added at 100 nMand cells were collected after 20 hours. Other ex- Data are the meansof two separate experiments. perimental conditions were the same as described in Fig. 2.

560 Valinomycin-induced Apoptosis of AH-130 Cells

gr00 1-// 1 caused apoptosis through elevation of the cytoplas- w -r^- YVAD- mic calcium concentration (20). Thus, it is likely that z «*rV^-à" *^ CHO high concentrations of valinomycin induce apoptosis through a different mechanism, prior to the loss of mito- 72 n. yva d- chondrial membrane potential. On the other hand, 5 40- X -*~CHO h- ^\ + Val apoptosis induced by 100 nMvalinomycin was not due to an increase in cytoplasmic calcium concentration. LUZ ' \ -A- \ DEVD" It has been reported that Bcl-2 was down-regulated g >v -TT- CH0 during various modes of apoptosis (1, 19, 21, 28), and g 20- \ 2 1 k DEVD- that overexpression of Bcl-2 and Bcl-xL protected cells Dl A cho from apoptosis by maintenance of mitochondrial mem- U- . + Val brane potential (6, 17, 25). These reports suggested that Bcl-2 and Bcl-xL could control apoptosis by regulation Q 0 lZ^jTT^J å å ni^ of mitochondrial membrane potential. Therefore, in Q 1 10 100 the present study the changes in contents of Bcl-2 and Bcl-xL during apoptosis were measured. AH-130cells CASPASE INHIBITORS (/iM) showed expression of Bcl-xL but not Bcl-2, and 100 nM Fig. 7. Effects of inhibitors of caspase family on valinomycin-in- valinomycin increased Bcl-xL level by approximately 4- duced DNAfragmentation of AH-130 cells. AH-130 cells were prein- fold for 24 hours (Fig. 3B). Since valinomycin decreases cubated with caspase inhibitors at the indicated concentrations for 1 the mitochondrial membranepotential, it could induce hour, and then 100 nMvalinomycin was added. After 20 hours, the apoptosis without Bcl-xL down-regulation. In addition, cells were collected and DNAfragmentation was measured by the di- an inhibitor of protein synthesis failed to prevent valino- phenylaminemethod. Other experimental conditions were the same mycin-induced DNAfragmentation. Thus, this apopto- as described in Fig. 2. YVAD-CHO, caspase-1 inhibitor; DEVD- sis would not require either the increase in expression of CHO, caspase-3 inhibitor; Val, 100 nM valinomycin. pro-apoptotic members of the Bcl-2 protein family or decreases in that of anti-apoptotic members. 10 juM), however, induced marked DNAfragmentation The relation between reactive oxygen species (ROS) within several hours. Ojcius et al. reported that high and apoptosis has been discussed in various systems. In concentrations of valinomycin (more than 25 /iM) particular, PT pore formation and apoptosis are consid- A B C

uj 10I 1 ui 10I v-I 60I 1 CO -O- Control W _ A 5" HSS^ ©8- -#-100nMVal£^ |=a8- /\ / \ Z~ / V O§ Oo / \ O / £°6 ^S / <4°" J

111 = IJj t / ^20 - / CO> W^ / OC /

0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 TIME (hr) TIME (hr) TIME (hr) Fig. 8. Activation of caspase-3-like protease after treatment with valinomycin. A. Activity of caspase-1-like protease. AH-130cells were cultured with 100 nMvalinomycin for the indicated periods. The cells were dissolved in lysis buffer C. The cell lysate was reacted with fluorogenic substrate specific to caspase-1 at 37°C for 1 hour. Released AMCwas measured fluorometrically at an excitation wavelength of 355 nmand emission wavelength of460 nm. One unit was defined as the enzyme activity that liberated 1 fimol ofAMCin 1 hour. B. Activity of caspase-3-like protease. Experimental conditions were the same as in A except specificity of substrate. C. DNAfragmentation induced by 100 nM valinomycin. Ex- perimental conditions were the same as described in Fig. 2.

561 Y. Inai et al. ered as defense systems preventing ROSformation (26). releasing protein(s) from the mitochondrial intermem- However,1 nMvalinomycin caused the depolarization brane space or decreasing ATPcontent. Further investi- and uncoupling of respiration in isolated mitochondria gations are necessary to determine the actions that asso- and inhibited the generation of hydrogen peroxide of ciate mitochondrial function with activation of caspase- isolated mitochondria in the presence of respiratory sub- 3-like protease. strate. Therefore, it is considered that ROSgeneration is not an important factor in this apoptosis. Acknowledgments. This work was supported in part by a grant The protease encoded by C. elegans ced-3 and its from the Japan Keirin Association. mammalianhomologues represented by the caspase family appear to play key roles in apoptosis. Although our results suggest that the activation of caspase-3-like REFEREN CES protease is a pivotal step in apoptosis induced by valino- Adam, L., Crepin, M., Savin, C, and Israel, L. 1995. Sodi- mycin, the mechanism linking the loss of mitochondrial um phenylacetate induces growth inhibition and Bcl-2 down-reg- membranepotential to caspase activation is unclear. ulation and apoptosis in MCF7rascells in vitro and in nude Igbavboa et al. reported that the PT in isolated mito- mice. Cancer Res., 55: 5156-5160. chondria allowed slow release of matrix proteins (12), Akao, Y., Otsuki, Y., Kataoka, S., Ito, Y., and Tsujimoto, Y. 1994. Multiple subcellular localization of bcl-2: detection and these proteins might mediate between the loss of in nuclear outer membrane,endoplasmic reticulum membrane, mitochondrial membrane potential and subsequent phe- and mitochondrial membranes. Cancer Res. , 54: 2468-2471. nomenain apoptosis. In fact, recently it was reported Burton, K. 1956. A study of conditions and mechanisms of the diphenylamine reaction for the colorimetic estimation of de- that an apoptosis-inducing factor (AIF) in the mito- oxyribonucleic acid. Biochem. J., 62: 315-323. chondrial intermembranespace is released into the cyto- Chen, L.Z., Nourse, J., and Cleary, M.L. 1989. The bcl-2 plasm during the collapse of mitochondrial membrane candidate proto-oncogene product is a 24-kilodalton integral- potential and causes nuclear apoptosis immediately in a membrane protein highly expressed in lymphoid cell lines and cell-free system (22, 26, 27). However, since AIF was lymphomas carrying the t (14; 18) translocation. Mol. Cell inhibited by inhibitors of caspase-1 not caspase-3, it Biol., 9: 701-710. seems unlikely that AIF contributes to apoptosis in- Colofiore, J.R., Stolfi, R.L., Nord, L.D., and Martin, D.S. duced by valinomycin. 1995. Biochemical modulation of tumor cell energy. IV. Evi- In addition, it has also been reported that cyto- dence for the contribution of adenosine triphosphate (ATP) de- chrome c leaks out from the mitochondrial intermem- pletion to chemotherapeutically-induced tumor regression. Bio- chem. Pharmacol., 50: 1943-1948. Decaudin, D., Geley, S., Hirsch, T., Castedo, M., lybrane(14, space29). (15,Although16) andouractivatesresults indicatecaspase-3theimmediate-involve- Marchetti, P., Macho, A., Kofler, R., and Kroemer, G. ment of caspase-3-like protease, it was reported that the 1997. Bcl-2 and Bcl-xL antagonize the mitochondrial dysfunc- release of cytochromec wasnot related to mitochondri- tion preceding nuclear apoptosis induced by chemotherapeutic al depolarization (14, 29). Thus, the loss of mitochon- agents. Cancer Res. , 57: 62-67. drial membranepotential might function in induction Estabrook, R. 1967. Mitochondrial respiratory control and the polarographic measurement of ADP: O rations. In Methods of apoptosis through depletion of ATP(5, 23). Since in Enzymology, volume X (R.W. Estabrook, and M.E. mitochondrial membrane potential is the driving force Pullman, eds.). Academic Press, New York, pp.41-47. of ATP synthesis, mitochondrial depolarization in- Garland, J.M. and Halestrap, A. 1997. Energy metabolism duces the depletion of ATPcontent (data not shown). during apoptosis. /. Biol. Chem. , 272: 4680-4688. Alternatively, an unknownprotein factor(s) released Gonzalez, G.M., Perez, B.R., Ding, L., Duan, L., Boise, from mitochondria by valinomycin-induced mitochon- L.H., Thompson, C.B., and Nunez, G. 1994. bcl-xListhema- drial depolarization might connect the apoptotic signal jor bcl-x mRNAform expressed during murine development and its product localizes to mitochondria. Development, 120: with the following steps including caspase-3-like prote- 3033-3042. ase activation. Hogeboom, G.H. 1955. Fractionation of cell components of In conclusion, our results suggest that valinomycin animal tissues. In Methods in Enzymology, volume I (S.P. Col- induces apoptosis of AH-130 cells (rat ascites hepato- owick, and N.O. Kaplan, eds.). Academic Press, New York, ma cells) dependent on the collapse of mitochondrial pp.16-19. membranepotential at low concentration (1-1,000 Hyslop, P.A. and Sklar, L.A. 1984. A quantitative nM). This apoptosis does not require changes in expres- fluorimetric assay for the determination of oxidant production by polymorphonuclear leukocytes: its use in the simultaneous sion of members of the Bcl-2 family, elevation of intra- fluorimetric assay of cellular activation processes. Anal. Bio- cellular calcium ion concentration, or ROSgeneration. chem., 141: 280-286. Alternatively, the activation of caspase-3-like protease Igbavboa, U., Zwizinski, C.W., and Pfeiffer, D.R. 1989. Re- might be a critical step in this apoptotic mechanism. It lease of mitochondrial matrix proteins through a Ca2+-requir- is considered that the mitochondrial membranepoten- ing, cyclosporin-sensitive pathway. Biochem. Biophys. Res. tial may regulate some actions leading to apoptosis by Commun., 161: 619-625.

562 Valinomycin-induced Apoptosis of AH-130 Cells

Johnson, L.V., Walsh, M.L., Bockus, B.J., and Chen, L.B. 23. Richter, C, Schweizer, M., Cossarizza, A., and 1981. Monitoring of relative mitochondrial membrane poten- Franceschi, C. 1996. Control of apoptosis by the cellular tial in living cells by fluorescence microscopy. /. Cell Biol , 88: ATP level. FEBS Lett., 378: 107-110. 526-535. 24. Sekine, C, Yagita, H., Kobata, T., Hasunuma, T., Kluck, R.M., Bossy, W.E., Green, D.R., and Newmeyer, Nishioka, K., and Okumura, K. 1996. Fas-mediated stimula- D.D. 1997. The release of cytochrome c from mitochondria: a tion induces IL-8 secretion by rheumatoid arthritis synoviocytes primary site for Bcl-2 regulation of apoptosis. Science, 275: independently of CPP32-mediated apoptosis. Biochem. Bio- 1132-1136. phys. Res. Commun., 228: 14-20. Krippner, A., Matsuno, Y.A., Gottlieb, R.A., and Babior, 25. Shimizu, S., Eguchi, Y., Kamiike, W., Waguri, S., B.M. 1996. Loss of function of cytochrome c in Jurkat cells Uchiyama, Y., Matsuda, H., and Tsujimoto, Y. 1996. Bcl-2 undergoing Fas-mediated apoptosis. /. Biol. Chem., 271: blocks loss of mitochondrial membrane potential while ICE in- 21629-21636. hibitors act at a different step during inhibition of death in- Liu, X., Kim, C.N., Yang, J., Jemmerson, R., and Wang, X. duced by respiratory chain inhibitors. Oncogene, 13: 21-29. 1996. Induction of apoptotic program in cell-free extracts: re- 26. Skulachev, V.P. 1996. Whyare mitochondria involved in quirement for dATP and cytochrome c. Cell, 86: 147-157. apoptosis? Permeability transition pores and apoptosis as selec- Marchetti, P., Hirsch, T., Zamzami, N., Castedo, M., tive mechanismsto eliminate superoxide-producing mitochon- Deaudin, D., Susin, S.A., Masse, B., and Kroemer, G. dria and cells. FEBS Lett., 397: 7-10. 1996. Mitochondrial permeability transition triggers lympho- 27. Susin, S.A., Zamzami, N., Castedo, M., Hirsch, T., cyte apoptosis. /. Immunol, 157: 4830-4836. Marchetti, P., Macho, A., Daugas, E., Geuskens, M., and Marchetti, P., Susin, S.A., Decaudin, D., Gamen, S., Kroemer, G. 1996. Bcl-2 inhibits the mitochondrial release of Castedo, M., Hirsch, T., Zamzami, N., Naval, J., Senik, A., an apoptogenic protease. /. Exp. Med., 184: 1331-1341. and Kroemer, G. 1996. Apoptosis-associated derangement of 28. Tangye, S.G. and Raison, R.L. 1996. Leukaemic CD5+ B- mitochondrial function in cells lacking mitochondrial DNA. cell apoptosis: co-incidence of cell death and DNAfragmenta- Cancer Res. , 56: 2033-2038. tion with reduced bcl-2 expression. Br. J. HaematoL, 92: 950- Mentz, F., Mossalayi, M.D., Ouaaz, F., Baudet, S., Issaly, 953. F., Ktorza, S, Semichon, M., Binet, J.L., and Merle, B.H. 29. Yang, J., Liu, X., Bhalla, K., Kim, C.N., Ibrado, A.M., Cai, 1996. Theophylline synergizes with chlorambucil in inducing J., Peng, T.I., Jones, D.P., and Wang, X. 1997. Prevention apoptosis of B-chronic lymphocytic leukemia cells. Blood, 88: of apoptosis by Bcl-2: release of cytochrome c from mitochon- 2172-2182. dria blocked. Science, 275: 1129-1132. Ojcius, D.M., Zychlinsky, A., Zheng, L.M., and Young, 30. Zamzami, N., Marchetti, P., Castedo, M., Hirsch, T., J.D.E. 1991. Ionophore-induced apoptosis: Role of DNA Susin, S.A., Masse, B., and Kroemer, G. 1996. Inhibitors of fragmentation and calcium fluxes. Exp. Cell Res. , 197: 43-49. permeability transition interfere with the disruption of the mito- Petersen, A.J., Brown, R.D., Gibson, J., Pope, B., Luo, chondrial transmembrane potential during apoptosis. FEBS X.F., Schutz, L., Wiley, J.S., and Joshua, D.E. 1996. Nu- Lett., 384: 53-57. cleoside transporters, bcl-2 and apoptosis in CLLcells exposed 31. Zhu, W., Cowie, A., Wasfy, G., Penn, L., Leber, B., and to nucleoside analogues in vitro. Eur. J. Haematol, 56: 213- Andrews, D. 1996. Bcl-2 mutants with restricted subcellular 220. location reveal spatially distinct pathways for apoptosis in differ- Petit, P.X., Susin, S.-A., Zamzami, N., Mignotte, B., and ent cell types. EMBOJ., 15: 4130-4141. Kroemer, G. 1996. Mitochondria and programmed cell death: back to the future. FEBS Lett., 396: 7-13. {Received for publication, June 30, 1997 and in revised form, July 24, 1997)

563