Address correspondence and reprint requests to Dr. Y. Nomura at Department of Pharmacology, Graduate School of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, Japan.
Abstract : It is well known that caspases are produced as proforms, which are proteolytically cleaved and activated during apoptosis or programmed cell death. We report here that caspases are activated during apoptosis by treatment with NOC18, a nitric oxide (NO) donor. Our present experiments have examined the way in which NO induces neuronal cell death, using a new type of NO donor that spontaneously releases only NO without enzymatic metabolism. NOC18 induced apoptosis in human neuroblastoma SH-SY5Y cells in a concentration-and time-dependent manner as estimated by DNA fragmentation assay, FACScan analysis, and nuclear morphology. Oxyhemoglobin, an NO trapper, suppressed NOC18-triggered DNA fragmentation, indicating that NO from NOC18 is a real activator in this study. Upon the induction of apoptosis, an increase in caspase-3-like protease activity, but not caspase-1, was observed. Procaspase-2 protein, an inactive form of caspase-2, decreased dramatically. In addition, NOC18 also resulted in poly (ADP-ribose) polymerase (PARP) cleavage, yielding an 85-kDa fragment typical of caspase activity. Oxyhemoglobin blocked the decrease of procaspase-2 and the cleavage of PARP by NOC18 in a concentration-dependent manner. Moreover, NO elicited the release of cytochrome c into the cytosol during apoptosis. These results suggest that both stimulation of caspase activity and cytochrome c release are partly involved in NO-induced neuronal apoptosis.
Nitric oxide (NO) is a short-lived, free radical gas synthesized from l-arginine by its synthetic enzyme, NO synthase (NOS). NO has a major physiological function in the nervous system (Garthwaite and Boulton, 1995). Produced in appropriate amounts through a constitutive Ca2+/calmodulin-dependent form of NOS by neurons, it contributes to neuronal communication. On the other hand, excessive and uncontrolled production of NO through a Ca2+-independent and inducible form of NOS by astrocytes or microglia is associated with diseases such as brain ischemia, neurodegeneration, and acute or chronic inflammation (Nathan, 1992 ; Nomura and Kitamura, 1993).
Although some mechanisms regarding NO-induced cytotoxicity in neurons have been proposed, the crucial mechanism of NO-induced cytotoxicity is still unclear (Zhang and Snyder, 1992 ; Brüne et al., 1994 ; Zhang et al., 1994). NO donors are toxic and cause changes in cellular morphology such as condensed and fragmented chromatin, shriveled nuclei, apoptotic bodies, and membrane blebbing (Stuehr and Nathan, 1989 ; Nakazawa et al., 1997). These observations are consistent with the overall description of apoptosis. On the other hand, NO can result in either apoptotic or necrotic cell damage in neuronal cells (Bonfoco et al., 1995). In contrast, NO can block neuronal death due to glutamate cytotoxicity (Lei et al., 1992). There seem to be differences in the effects of NO on cell death, and the diverse results are partly brought about by the kind of NO donor used. For example, sodium nitroprusside (SNP) simultaneously releases both NO and cyanide. The breakdown product mimics the effect of NO and has cytotoxicity. Also, the spontaneous liberation of NO cannot account for the relaxation induced by S-nitroso-N-acetylpenicillamine (SNAP) in vitro (Kowaluk and Fung, 1990). The vascular smooth muscle exhibits catalytic activity toward S-nitrosothiol, as measured by NO generation. This activity appears to be associated with membrane components. Therefore, the NO released from SNAP is not only spontaneous, but it might also be catalyzed at the membrane. It is very difficult to understand whether or not the biological effects using some NO donors are dependent only on spontaneously released NO. Moreover, the half-lives of NO donors such as SNP and SNAP are not well understood. Hence, there are many discrepant results in studies of the physiological effects of NO donors.
NOCs have been developed as a new type of NO-releasing compound. They are very stable solids that release NO without co-factors or enzymatic metabolism and without the production of toxic metabolites. NOC18 can slowly, spontaneously release NO with a half-life of ~20 h at pH 7.4 and 37°C (Mooradian et al., 1995). In the present study, we attempt to further clarify the mechanism of NO-induced neuronal apoptosis using NOC18. NOC18 elicited apoptotic cell death in a concentration-and time-dependent manner in human neuroblastoma SH-SY5Y cells. The cell death induced by NOC18 is blocked by treatment with benzyloxycarbonyl-Asp-CH2-OC(O)-2,6-dichlorobenzene (Z-Asp-CH2-DCB), a nonspecific caspase inhibitor. Furthermore, we report here that NO derived from NOC18 stimulates caspase (caspase-3-like protease) activity but not caspase-1-like protease activity, which is linked to apoptosis accompanying the release of cytochrome c from mitochondria.
MATERIALS AND METHODS
Synthetic peptide-based substrates for cysteine protease p32 (CPP32)-like proteases [N-acetyl-Asp-Glu-Val-Asp-4-methylcoumaryl-7-amide (Ac-DEVD-MCA)] and interleukin-1 β-converting enzyme (ICE)-like protease [N-acetyl-Tyr-Val-Ala-Asp-4-methylcoumaryl-7-amide (Ac-YVAD-MCA)] and non-selective caspase inhibitor (Z-Asp-CH2-DCB) were purchased from the Peptide Institute (Osaka, Japan). NOC18 was from Dojindo Laboratory (Kumamoto, Japan). Oligo-(dT)12-18 primer, 5 × first-strand buffer, and reverse transcriptase (RT) were from GibcoBRL (U.S.A.). The Expand High Fidelity PCR System was purchased from Boehringer-Mannheim (U.K.). The dNTP mixture and RNase inhibitor were from Takara (Kyoto, Japan). Anti-ICE and CED-3 homologue 1 (anti-ICH-1 ; caspase-2) antibody and anti-poly(ADP-ribose) polymerase (anti-PARP) antibody (clone C-2-10) were purchased from Transduction Laboratory (U.S.A.), and Clontech (U.S.A.), respectively. The mouse monoclonal anti-cytochrome c (7H8.2C12) and anti-α-tubulin were from Pharmingen (U.S.A.) and Seikagaku Corp. (Japan), respectively. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Hoechst 33258, Triton X-100, and other reagents were obtained from Sigma (U.S.A.).
Human SH-SY5Y cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% (vol/vol) heat-inactivated fetal calf serum, 50 μg/ml penicillin, and 100 μg/ml streptomycin at 37°C in humidified 5% CO2/90% air.
Assessment of cell viability
Viability was measured in triplicate in 96-well plates by quantitative colorimetric assay with MTT. Viability was expressed as the ratio of the signal obtained from treated cultures and the signal from untouched control cultures multiplied by 100 (% control).
The cells were washed with phosphate-buffered saline (PBS) and lysed [lysis buffer : 10 mM Tris-HCl (pH 7.4), 5 mM EDTA, and 0.5% Triton X-100] for 20 min at 4°C. The samples were centrifuged at 27,000 g for 15 min at 4°C. The supernatant was extracted with equal volumes of phenol, phenol/chloroform (1 : 1 vol/vol), and chloroform, and then DNA was precipitated with 0.1 volume of 3 M sodium acetate (pH 5.2) and 2 volumes of ethanol. The DNA was suspended in a buffer of 10 mM Tris-HCl (pH 8.0) and 1 mM EDTA and treated with 40 μg/ml RNase A for 1 h at 37°C. The concentrations of DNA were determined by the absorbance at 260 nm. A 20-μg sample of DNA was subjected to agarose gel electrophoresis on a 1.5% gel in a buffer of 40 mM Tris-HCl (pH 8.5) and 2 mM EDTA. The gel was then stained with 0.5 μg/ml ethidium bromide for 15 min, and the fragmented DNA was visualized under UV light and then photographed.
For the cell cycle analysis, cells fixed by ethanol were stained with 100 μg/ml propidium iodide (PI) in the presence of RNase. The percentage of apoptotic cells was quantitated using a FACScan Flow Cytometer (Becton Dickinson). Fluorescence data were collected using logarithmic amplification, and necrotic debris was eliminated from the data as described previously (Mannick et al., 1997). Apoptotic cells were distinguished from nonapoptotic, intact cells by their decreased DNA content as determined by their lower PI staining intensity.
Assessment of nuclear morphology
Chromosomal condensation and DNA fragmentation were determined using the chromatin dye Hoechst 33258. Cells were fixed for 30 min in PBS containing 1% glutaraldehyde. After fixation at room temperature, the cells were washed twice with PBS and then exposed to 5 μg/ml Hoechst 33258 in PBS for 30 min at room temperature. After washing, all samples were mounted with glycerol/PBS (1 : 1 vol/vol) and observed using a Nikon fluorescence microscope.
Assay for caspase activity
Cells in 60-mm plates were treated with 250 μM NOC18 for the indicated periods. At the appropriate time, the medium was aspirated, and the cells were washed with PBS and added to 50 μl of lysis buffer [50 mM Tris-HCl (pH 7.4), 1 mM EDTA, 10 mM EGTA, and 10 μM digitonin]. The cells were incubated for 10 min at 37°C. After incubation, the supernatant was centrifuged at 15,000 rpm for 10 min. Next, cleared lysate (50 μg of protein) was incubated at 37°C with 50 μM Ac-DEVD-MCA for 30 min or Ac-YVAD-MCA for 60 min. The amounts of released 7-amino-4-methylcoumarin (AMC) were measured with a spectrofluorometer (Hitachi F-2000 Fluorescence Spectrophotometer) with excitation at 380 nm and emission at 460 nm. One unit was defined as the amount of enzyme required to release 1 pmol of AMC/min at 37°C.
RT-PCR was performed to determine the presence of caspase family mRNA. Total RNA from SH-SY5Y cells was isolated by the guanidinium isothiocyanate/acidic phenol/chloroform method, and RNA concentrations were estimated spectrophotometrically as described previously (Uehara et al., 1998). A 2.0-μg RNA aliquot was reverse-transcribed at 37°C for 1 h in a 20-μl reaction volume containing 100 U of RT, 0.25 μg of oligo-(dT)12-18 primer, 0.5 mM dNTP, 40 U of RNase inhibitor, and 5 mM dithiothreitol. An aliquot (2 μl) of RT product was mixed with 1 mU of DNA polymerase and 0.2 μM of sense and antisense primers in a buffer containing 1 × PCR buffer and 0.2 mM dNTP in a final volume of 20 μl. The primers employed are listed here : caspase-1 (upstream) 5′-AAC CCA GCT ATG CCC ACA TCC-3′ ; caspase-1 (down-stream) 5′-TTA ATG TCC TGG GAA GAG GTA-3′ ; caspase-2 (upstream) 5′-GTT ACC TGC ACA CCG AGT CAC G-3′ ; caspase-2 (downstream) 5′-GCG TGG TTC TTT CCA TCT TGT TGG TCA-3′ ; caspase-3 (upstream) 5′-GAA TAT CCC TGG ACA ACA-3′ ; caspase-3 (downstream) 5′-ACG CCA TGT CAT CAT CAA-3′ ; caspase-4 (upstream) 5′-GGT CAT CAT TGT CCA GGC-3′ ; caspase-4 (downstream) 5′-CCA TTG TGC TGT CTC TCC-3′ ; caspase-7 (upstream) 5′-AGC CTG GGT TTT GAC GTG-3′ ; caspase-7 (downstream) 5′-ACC GTG GAA TAG GCG AAG-3′. The predicted PCR products of caspase-1 (α, β, or γ subtype), -1 (δ subtype), -2, -3, -4, and -7 are 858, 714, 234, 462, 203, and 371 bp, respectively. The number of cycles selected for each primer pair was found to produce a linear relationship between the amounts of input RNA and resulting PCR products. The PCR products were resolved by electrophoresis in a 6% polyacrylamide gel in 0.5 × Tris/borate/EDTA buffer. The gel was stained with ethidium bromide and photographed.
Western blot analysis
Cells (5 × 105) were washed twice with ice-cold PBS and added to the sodium dodecyl sulfate (SDS) sample buffer. The total lysates were then boiled for 5 min. Equal amounts of each sample were subjected to 12% SDS-polyacrylamide gel electrophoresis followed by transfer onto a nitrocellulose filter at 100 V for 1 h at 4°C. The filters were then blocked with 10 mM Tris-HCl (pH 7.5), 100 mM NaCl, and 0.1% Tween 20 containing 5% nonfat milk for 1 h at room temperature. Human anti-ICH-1 (caspase-2), anti-PARP, and anti-α-tubulin were used as primary antibodies, and horseradish peroxidase-labeled goat or mouse Ig was used as a secondary antibody. The antibody-reactive bands were revealed by chemiluminescent detection (ECL western detection kit).
Detection of cytochrome c
Cytochrome c protein was detected by western blot analysis as described previously (Rossé et al., 1998). In brief, cells were disrupted by a Dounce homogenizer with 10 strokes using buffer A [20 mM HEPES (pH 7.4), 1 mM EDTA, 1 mM EGTA, 250 mM sucrose, 2 μg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 1 μg/ml pepstatin]. The homogenates were separated into cytosol and membrane fractions by ultracentrifugation. Equal amounts of protein (5 μg) were subjected to western blot analysis using mouse monoclonal anti-cytochrome c (7H8.2C12 ; Pharmingen) and anti-α-tubulin (Seikagaku Corp., Japan).
NOC18 induces apoptotic cell death in SH-SY5Y cells
To address the ability of NOC18 to induce cell death, we first examined the effect of NOC18 on cell viability by means of an MTT assay. This assay is a quantitative colorimetric assay based on the reduction of a tetrazolium salt, MTT. The salt is reduced to an insoluble blue formazan product in living cells but not in the cellular debris of dead cells. As assessed by the ability of the cells to metabolize MTT, the loss of viability was hardly observed in a quiescent state. The loss of viability occurred in a concentration-dependent manner in response to treatment with NOC18. After treatment with 500 μM NOC18 for 24 h, the viability was reduced to 45% of untreated control cells (Fig. 1A). A small loss of viability was observed within 4-8 h of the NOC18 challenge. At 12 h, the viability was reduced to 30% of control cultures, and by 36 h viability was reduced to 65% (Fig. 1B).
To investigate whether or not the loss of viability in response to NOC18 correlates with a biochemical feature that discriminates between apoptosis and necrosis, we carried out three procedures that are used to characterize apoptosis : internucleosomal DNA fragmentation, FACS analysis, and a determination of nuclear morphology using chromatin dye.
DNA from NOC18-treated cells was examined for oligonucleosomal fragmentation by means of agarose gel electrophoresis. No internucleosomal DNA fragmentation was observed in the control cultures. Low molecular weight DNA extracted from SH-SY5Y cells challenged with NOC18 showed the formation of an oligosomal DNA ladder in a concentration- and time-dependent manner on agarose gels stained with ethidium bromide (Fig. 2A and B). Oxyhemoglobin is known to bind to and inactivate NO. We analyzed the inhibitory effect of oxyhemoglobin on NOC18-induced DNA fragmentation. Treatment with 2 mg/ml oxyhemoglobin alone did not induce the formation of a DNA ladder. The simultaneous addition of various concentrations of oxyhemoglobin and 250 μM NOC18 attenuated the formation of an internucleosomal DNA ladder in a concentration-dependent manner (Fig. 2C).
To verify the induction of apoptosis by NOC18, the cells were labeled with PI and analyzed by flow cytometry (FACS). Figure 3 shows the DNA content histograms obtained after the PI staining of cells that had been treated with various concentrations of NOC18 for 24 h. When the cells were incubated in medium alone, a single peak of nuclei with diploid DNA content was observed. In the presence of NOC18, apoptotic cells with reduced DNA content were distinguishable. A characteristic hypodiploid DNA content peak, which shows sub-G0/G1 apoptotic populations, was observed following the treatment of SH-SY5Y cells with NOC18 in a concentration-dependent manner. The time course for the appearance of this hypodiploid peak of DNA correlated with the appearance of the DNA ladder.
To confirm whether or not the loss of viability was attributable to apoptosis, we performed nuclear staining with the chromatin dye Hoechst 33258 as the last criterion. Nuclear condensation was hardly observed in the untreated cells (Fig. 4A). In cells incubated with 250 μM NOC18 for 24 h, the condensation and fragmentation of chromatin and the shrinkage and fragmentation of nuclei as apoptotic bodies were evident (Fig. 4B). These effects of NOC18 on cell death were significantly blocked by treatment with oxyhemoglobin (Fig. 4C).
NOC18 induces apoptotic cell death via caspase activation
To address whether or not caspases contribute to NOC18-induced apoptosis, the DNA ladder was investigated in SH-SY5Y cells following treatment with NOC18 in the presence or absence of Z-Asp-CH2-DCB, a cell-permeable nonspecific caspase inhibitor. A suppression of DNA ladder formation was observed in the cells 24 h after a 250 μM NOC18 challenge was administered simultaneously with a Z-Asp-CH2-DCB treatment. Treatment with 50 μM Z-Asp-CH2-DCB completely blocked the DNA ladder formation induced by NOC18 (Fig. 5A). Furthermore, no DNA ladder was detected for up to 48 h with the simultaneous addition of 250 μM NOC18 and 100 μM Z-Asp-CH2-DCB (Fig. 5B).
Having established a role for caspases in cell death induced by NOC18, we analyzed the process of caspase activation in this system. Detergent extracts prepared from the cells at various periods after a 250 μM NOC18 treatment were tested for the cleaving activity of two fluorogenic peptide substrates (Ac-DEVD-MCA and Ac-YVAD-MCA). Ac-DEVD-MCA is a fluorogenic, tetrapeptide substrate that is cleaved by caspace-3-like protease (containing caspase-2, -3, -4, and -7). Another peptide, Ac-YVAD-MCA, is cleaved by caspase-1 (Henkart, 1996 ; Nicholson and Thornberry, 1997). Ac-DEVD-MCA cleaving activity was clearly elevated and sustained with NOC18 treatment. Maximal Ac-DEVD-MCA cleaving activity was detected at 24 h following NOC18 challenge. In contrast, SH-SY5Y cells challenged with NOC18 showed only low levels of Ac-YVAD-MCA cleaving activity at any time point examined in this study (Fig. 6).
Caspases are activated, accompanying cytochrome c release from mitochondria, during NOC18-induced apoptosis
RT-PCR was used to determine the presence of mRNA species for caspases. Because Ac-DEVD-MCA is known to be a substrate for caspase-2, -3, -4, and -7 in vitro (Henkart, 1996 ; Nicholson and Thornberry, 1997), we corroborated the existence of each caspase containing caspase-1 using this system. Figure 7 shows a photograph of a gel stained with ethidium bromide. The mRNAs of caspase-2, -3, and -7 but not caspase-1 and -4 were detected in control and NOC18-treated cells. The level of each detected caspase mRNA did not change during the cell death induced by NOC18.
We next investigated the cleavage of PARP and the processing of caspase by treatment with NOC18 in SH-SY5Y cells. Total cell lysates were collected at different time points and subjected to western blot analysis using a specific antibody. It is considered that PARP is a substrate for several subtypes of caspases (caspase-3, -6, -7, -8, and -9) (Lazebnik et al., 1994 ; Martins and Earnshaw et al., 1997 ; Woo et al., 1998). First, to determine whether or not apoptosis induced by NOC18 in SH-SY5Y cells is accompanied by the activation of caspases, we evaluated the cleavage of PARP. As shown in Fig. 8A, treatment with NOC18 resulted in PARP cleavage, yielding an 85-kDa fragment showing typical caspase activity in a time- and concentration-dependent manner. The 85-kDa PARP fragment, which is indicative of proteolytic holoenzyme digestion, is visible after an 18-h incubation period with 250 μM NOC18 (Fig. 8B). PARP cleavage activity was coincident with the appearance of DNA fragmentation (Fig. 2). In addition, oxyhemoglobin prevented the cleavage of PARP by NOC18 (Fig. 8C). Procaspase-2, a 48-kDa precursor of caspase-2, is synthesized, and this precursor is proteolytically cleaved to produce a mature enzyme composed of 18- and 12-kDa subunits. We next examined the cleavage of procaspase-2 in response to NO by western blot analysis. Anti-C-terminal caspase-2 antibody recognized the fulllength procaspase-2. A time- and concentration-dependent decrease in the level of procaspase-2 was observed (Fig. 9A and B). To confirm that the decreased level of procaspase-2 in response to NOC18 is not caused by nonspecific protein degradation, we investigated the levels of α-tubulin, a housekeeping protein, using the same samples. There are equal amounts of α-tubulin in each sample (Fig. 9A and B ; data not shown in Fig. 9C). Treatment with oxyhemoglobin inhibited the decrease in procaspase-2 in a concentration-dependent manner (Fig. 9C). We then measured cytochrome c in the membrane and cytosol fractions by western blot analysis. Without NOC18 challenge, most of the detectable cytochrome c was in the membrane fraction. Cytochrome c in the cytosol fraction increased significantly after 12 h of challenge with NOC18, and it continued to increase for up to 36 h. The amounts of cytochrome c in the membrane fraction showed a corresponding decrease in a time-dependent manner (Fig. 10A). Furthermore, treatment with 1 mg/ml oxyhemoglobin inhibited the release of cytochrome c into the cytosol fractions in response to NOC18 (Fig. 10B).
The aim of this study was to elucidate the mechanism of NO-induced cell death in human SH-SY5Y neuroblastoma cells. We have shown here that NOC18, an NO donor, elicits apoptotic cell death as estimated by inter-nucleosomal DNA fragmentation, chromatin condensation, and FACS analysis. This apoptosis is accompanied by the loss of mitochondrial membrane potential. NOC18 is known to spontaneously release only NO without co-factor or enzymatic metabolism and without producing toxic metabolites (Mooradian et al., 1995). We demonstrated that caspase-3-like proteases are activated during apoptosis by NOC18. Both caspase activation and apoptosis traggered by NOC18 were inhibited by treatment with both a nonselective inhibitor of caspase-3-like proteases and oxyhemoglobin, one of the NO trappers. These findings suggest that the activation of caspases in response to NO is possibly involved in apoptotic cell death in this cell line.
Treatment with NOC18 induced DNA fragmentation in a time- and concentration-dependent manner in SH-SY5Y cells (Fig. 2A and B). These data indicate that exposure to 250 μM NOC18 for >18 h induces apoptotic cell death. Another NO donor, SNAP, also elicits the formation of an internucleosomal DNA ladder in this cell (T. Uehara, Y. Kikuchi, and Y. Nomura, unpublished data). Furthermore, oxyhemoglobin, which can trap and inactivate NO, attenuates NOC18-induced cell death, suggesting that NO derived from NOC18 appears to exert apoptotic effects in SH-SY5Y cells. As oxyhemoglobin cannot penetrate the cell membranes, it seems that extracellular NO derived from NOC18 is important in apoptosis-inducing action. On the other hand, treatment with a high concentration of NG-nitro-l-arginine (100 μM) did not alter the NOC18-induced DNA ladder formation (data not shown), suggesting that NOS activity is not involved in NOC18-triggered apoptosis.
We next examined whether or not caspases, known to be death proteases, are involved in NO-induced apoptosis. Treatment with caspase inhibitory peptide (Z-Asp-CH2-DCB) significantly inhibited the formation of an NO-induced DNA ladder (Fig. 5). Certainly, the enzyme activity of caspase-3-like protease in extracted cytosol using a fluorescent peptide substrate (DEVD-MCA) was observed in response to treatment with NOC18 (Fig. 6), suggesting that NOC18 induces apoptotic cell death through caspase activation. Moreover, we found that PARP, an endogenous substrate for several subtypes of caspases, is also degraded during an apoptotic process induced by NOC18 in an oxyhemoglobin-dependent manner with a similar time course as caspase-3-like protease activation. Hence, we have conclusive evidence that NO derived from NOC18 elicits apoptosis in neuronal cells via caspase activation.
In contrast, we could not detect caspase-1 activity, using another fluorescent peptide substrate (YVAD-MCA), in response to NOC18 (Fig. 6). It has been reported that caspase-1 activity precedes the activation of caspase-3 upon Fas stimulation (Enari et al., 1996). A potential pathway for caspase-3 activation involves the upstream activation of caspase-1 with the proteolytic cleavage of procaspase-3 to the active enzyme. However, caspase-1 activity was hardly detectable in this cell line. Similar results showing no caspase-1 activity during apoptosis in response to treatment with etoposide (Martins et al., 1997), staurosporine (Keane et al., 1997), and cytokine deprivation (Ohta et al., 1997) have previously been reported. In the human brain, neither caspase-1 nor caspase-4 mRNA has been detected using a northern blot analysis (Kamens et al., 1995 ; Van de Craen et al., 1997). In agreement with this observation, RT-PCR analysis has shown that in SH-SY5Y cells, the mRNAs of caspase-1 and -4 cannot be detected (Fig. 7). These results indicate that NO may activate caspase-3-like protease through unidentified pathways different from those seen when the cells are challenged with anti-Fas antibody.
At present we do not know how caspases are activated by NO ; however, some cascades have been proposed. Recently, it has been revealed that a release of apoptogenic proteins such as cytochrome c and apoptosis-inducing factor from the mitochondria to the cytosol is involved in protease activation linked to apoptosis (Liu et al., 1996 ; Susin et al., 1996 ; Yang et al., 1997). Cytochrome c is known to be an essential factor in the mitochondrial respiratory chain and is also released in response to various stimuli that elicit apoptosis, including etoposide, staurosporine, actinomycin D, H2O2, UV, and Ara-C (cytosine β-d-arabinofuranoside ; C. N. Kim et al., 1997 ; Kluck et al., 1997). Ara-C causes a significant release of cytochrome c from the mitochondria to the cytosol with an accompanying decline in the mitochondrial transmembrane potential (Y.-M. Kim et al., 1997). In contrast, Yang et al. (1997) reported that the release of cytochrome c induced by treatment with staurosporine precedes the loss of mitochondrial membrane potential. The precise mechanism for cytochrome c translocation by an apoptosis inducer is still unclear. Once cytochrome c is released into the cytosol from the mitochondria in response to challenge by apoptosis inducers, it activates caspase-3 (Liu et al., 1996). Recently, Li et al. (1997) indicated that cytosolic cytochrome c activates caspase-9. Furthermore, they showed that activated caspase-9 cleaves and activates caspase-3. These results suggested that caspase-9 is the upstream member of the apoptotic protease cascade that is triggered by cytochrome c. In this cell line, we examined whether or not cytochrome c is released into the cytosol in response to NO. Western blot analysis showed that cytochrome c is released significantly after 12 h of challenge with NOC18, and the cytochrome c levels continue to increase for up to 36 h. In contrast, the amounts of cytochrome c in the membrane fraction showed a corresponding time-dependent decrease. The appearance of cytosolic cytochrome c was correlated with the activation of caspase-3-like protease, PARP cleavage, and a loss of the proform of caspase-2. NO is known to result in mitochondrial damage and mitochondrial respiratory chain inhibition (Fig. 1 ; for review, see Bolaños et al., 1997). In particular, NO binds reversibly to cytochrome c oxidase (complex IV) in mitochondrial respiration (Lizasoain et al., 1996). Most of the mitochondrial MTT reduction occurs subsequent to the transfer of electrons from cytochrome c to cytochrome c oxidase (Berridge and Tau, 1993). This leads to the notion that the loss of viability estimated by the MTT assay may be caused by the inhibition of cytochrome c oxidase by NO. On the other hand, Hortelano et al. (1997) have demonstrated that NO induces apoptosis via the triggering of the mitochondrial permeability transition in mouse thymocytes. Interestingly, bongkrekic acid as well as cyclosporine, which are known to interfere with the disruption of the mitochondrial transmembrane potential, prevent NO-induced apoptosis, including nuclear DNA fragmentation and the exposure of phosphatidylserine residues on the cell surface. We found that NO attenuates the loss of MTT reduction preceding caspase activation and DNA fragmentation (Figs. 1, 2, and 8). In other neuronal cells, treatment with NO donors decreased mitochondrial transmembrane potential and energy charge and caspase inhibitors protected cells against apoptosis, suggesting that NO induces apoptosis by a mechanism involving caspase activation and mitochondrial dysfunction (Estévez et al., 1995 ; Leist et al., 1997 ; Keller et al., 1998). Taken together, in light of the loss of mitochondrial function in response to NO, NO may trigger apoptosis through cytochrome c release and the subsequent caspase activation in this system. Moreover, Bcl-2 in mitochondria blocks not only apoptosis but also cytochrome c translocation. We have previously reported that treatment with retinoic acid increases the level of endogenous Bcl-2 in SH-SY5Y cells (Itano et al., 1996). NOC18-induced DNA fragmentation was not observed under these conditions (T. Uehara, Y. Kikuchi, and Y. Nomura, unpublished data). The overexpression of Bcl-2 also prevents NO-mediated apoptosis and PARP cleavage in HeLa G cells (Melkova et al., 1997). These observations suggest that the NO-induced death signal is sensitive to Bcl-2.
In conclusion, we present data indicating that NO can trigger apoptosis via caspase activation in SH-SY5Y neuroblastoma cells. Our data show that NO reduces mitochondrial function, as estimated by MTT assay, proceeding to caspase activation and internucleosomal DNA fragmentation. In particular, we demonstrated that NO can stimulate the release of cytochrome c from the mitochondria. It is well known that inducible NOS, which produces a large amount of NO, is expressed in response to ischemic stresses in the brain and has an essential function in neuronal death in ischemia. Hence, NO derived from inducible NOS in brain ischemia may trigger apoptotic neuronal cell death via caspase activation following mitochondrial damage. Although we demonstrated here that NO induces apoptosis via caspase activation, a detailed mechanism for how NO activates caspase and the relationship between mitochondrial damage and apoptosis remain to be determined in further investigations.