Abstract
Abstract. Ceramide has been implicated to play an important role in the cell signaling pathway involved in apoptosis. Most studies that have used the apoptotic model of cellular injury have suggested that enhanced ceramide generation is the result of the breakdown of sphingomyelin by sphingomyelinases. However, the role of ceramide synthase in enhanced ceramide generation in response to oxidant stress has not been previously examined in any tissue. Hydrogen peroxide (H2O2) (1 mM) resulted in a rapid increase in ceramide generation (as measured by in vitro diacylglycerol kinase assay) in LLC-PK1 cells. The intracellular ceramide level was significantly increased at 5 min after exposure of cells to H2O2 and thereafter continuously increased up to 60 min. H2O2 also resulted in a rapid increase (within 5 min) in ceramide synthase activity (as measured by incorporation of [14C] from the labeled palmytoyl—CoA into dihydroceramide) in microsomes. In contrast, the exposure of cells to H2O2 did not result in any significant change in sphingomyelin content or acid or neutral sphingomyelinase activity. An increase in ceramide production induced by H2O2 preceded any evidence of DNA damage and cell death. The specific inhibitor of ceramide synthase, fumonisin B1 (50 μM), was able to suppress H2O2-induced ceramide generation and provided a marked protection against H2O2-induced DNA strand breaks, DNA fragmentation, and cell death. Taken together, these data provide the first evidence that H2O2 is a regulator of ceramide synthase rather than sphingomyelinases and that ceramide synthase—dependent ceramide generation plays a key role in DNA damage and cell death in oxidant stress to renal tubular epithelial cells.
Ceramide, a metabolite of sphingolipids, has been implicated to play an important role in the cell signaling pathway involved in apoptotic cell death in response to a variety of stimuli (1,2,3). Ceramide is generated by the two major pathways: condensation of sphingosine or sphinganine and fatty acyl—CoA by ceramide synthase or by hydrolysis of sphingomyelin by sphingomyelinases (1,2,3). Most studies that have used the apoptotic model of cellular injury have shown that increased ceramide generation is mediated by the activation of sphingomyelinases (2,4,5). More specifically, in relation to oxidant stress it was shown that hydrogen peroxide (H2O2) results in ceramide generation, with a concomitant quantitative reduction in sphingomyelin content in human monoblastic leukemia cells and bovine endothelial cells (6), which suggests a role of sphingomyelinase activation for ceramide generation. Similarly, a recent study has shown that ceramidase, rather than ceramide synthase, plays a role in iron-mediated enhanced ceramide generation (7). There is limited information on the role of ceramide synthase in enhanced ceramide generation in tissue injury. In response to Fas, ceramide generation has been shown to be enhanced in cells that are deficient in sphingomyelinases (8), which suggests the role of a pathway other than sphingomyelinases for ceramide generation—namely, ceramide synthase. More direct evidence is provided by the observation that daunorubicin (9) and 12-O-tetradecanoylphorbol-13-acetate (10) result in ceramide synthase activation, with a concomitant increase in sphingomyelin content.
This study was therefore undertaken to examine the role of ceramide synthase for ceramide generation in renal tubular epithelial cells exposed to H2O2. We also examined the cause-effect relationship among ceramide generation, DNA damage, and cell death in oxidant injury to renal tubular epithelial cells. Our data provide strong evidence that H2O2 increases ceramide synthase activity, resulting in enhanced ceramide generation, which is a key modulator of DNA damage and cell death in oxidant injury to renal epithelial tubule cells.
Materials and Methods
Materials
Dihydrosphingosine (sphinganine) was purchased from Biomol Research Laboratory, Inc. (Plymouth Meeting, PA). Diacylglycerol kinase of Escherichia coli strain N4830/pJW10 and fumonisin B1 were purchased from Calbiochem-Novabiochem Corp. (La Jolla, CA). [γ-32P] ATP and [N-methyl 14C] sphingomyelin were obtained from Amersham Life Science, Inc. (Arlington Heights, IL). All other agents were purchased from Sigma Chemical Co. (St. Louis, MO). Silica gel 60 plates for thin-layer chromatography (TLC) were purchased from Whatman (Hillsboro, OR).
Cell Culture
LLC-PK1 cells obtained from the American Type Culture Collection (Rockville, MD) were cultured as in our previous studies (11,12,13,14,15). Cultures were maintained in a humidified incubator gassed with 5% CO2-95% air at 37°C and fed at intervals of 48 to 72 h. Cells were used 2 days after confluence.
Experimental Protocol
To examine the effect of H2O2 on ceramide generation, cells were rinsed with serum-free Dulbecco's modified Eagle's medium (DMEM) at pH 7.4 and exposed to 1 mM H2O2 for different time intervals as indicated, and ceramide generation was measured as described below. The dose of H2O2 used and the time of exposure in the study were based on our previous studies (11,12). To determine which pathway—namely, hydrolysis of sphingomyelin by sphingomyelinases or ceramide generation by ceramide synthase—is involved in oxidant-induced ceramide generation, we measured the activities of ceramide synthase and sphingomyelinases as well as sphingomyelin content in the cells exposed to H2O2, as described below. To examine the role of ceramide synthase in oxidant injury, we used a specific inhibitor of ceramide synthase, fumonisin B1, which blocks sphinganine N-acyltransferase, resulting in a decrease in ceramide (9,10,16,17,18). To examine the effect of fumonisin B1 on H2O2-induced ceramide generation as well as DNA damage and cell death, LLC-PK1 cells were washed with serum-free DMEM, preincubated with 50 μM fumonisin B1 for 30 min, and then exposed to H2O2. The dose of fumonisin B1 was chosen on the basis of our recent studies (14).
Assay for Ceramide
The lipids were extracted according to the method of Bligh and Dyer (19). Lipids in the organic-phase extract were dried under nitrogen and subjected to mild alkaline hydrolysis (0.1 N methanolic KOH for 1 h at 37°C) to remove glycerophospholipids (9). Samples were re-extracted, and the organic phase was dried under nitrogen. Ceramide was quantified by in vitro diacylglycerol (DG) kinase assay, as described elsewhere (20) and used in our recent studies (14), with minor modification. In brief, the dried extracted lipids under nitrogen were dissolved in 20 μl of 7.5% n-octyl-β-glucopyranoside, 25 mM dioleoylphosphatidylglycerol, and1 mM diethylenetriaminepentaacetic acid (DETAPAC), pH 7.0. Then the following solutions were added in order: 50 μl of reaction buffer containing 100 mM imidazole HCl (pH 6.6), 100 mM NaCl, 25 mM MgCl2, 2 mM ethyleneglycolbis(B-aminoethyl ether)-NN-tetraacetic acid (pH 6.6), 2 μl of 100 mM dithiothreitol in 1 mM DETAPAC (pH 7.0), 10 μl of DG kinase (5 μg/10 μl) from E. coli strain N4830/pJW10 diluted in 100 mM imidazole-HCl and 1 mM DETAPAC (pH 6.6), and 8 μl of water. The reaction was initiated by the addition of 10 μl each of unlabeled 10 mM ATP and [γ-32P] ATP (4.5 μCi per sample) in 20 mM imidazole (pH 6.6), 1 mM DETAPAC, and incubation at 25°C for 45 min. The reaction was stopped by addition of 0.7 ml 1% (wt/vol) perchloric acid. The lipids were extracted with 3 ml of methanol/chloroform (2/1 vol/vol), 1 ml of chloroform, and 1 ml of 1% perchloric acid. The lower chloroform phase was washed with 2 ml of 1% perchloric acid/methanol (7/1 vol/vol) and then used for measurement of ceramide. The standard ceramide type III (from bovine brain) were also treated in the same manner. Ceramide 1-phosphate was resolved by TLC on silica gel 60 plates by use of a solvent system of chloroform/acetone/methanol/acetic acid/water (10.5/5/2/2/1, vol/vol/vol/vol/vol). The TLC plate was dried and autoradiographed overnight. The band corresponding to ceramide 1-phosphate was scraped, and incorporation of 32P into ceramide 1-phosphate was quantified by Cerenkov scintillation counting. The level of ceramide was determined by comparison with a standard curve generated concomitantly of known amounts of ceramide. Data are expressed as pmol ceramide/nmol of total lipid phosphate, determined according to the method of Ames and Dubin (21).
Measurement of Sphingomyelin Content
The lipids were extracted, dried under nitrogen as described above, and resolved on TLC plate by use of chloroform/methanol/acetic acid/water (50/30/8/5, vol/vol/vol/vol), as described elsewhere (9). The standards of known amounts of sphingomyelin were also run on TLC. The spot corresponding to sphingomyelin was visualized by use of iodine vapor, scraped, and transferred into glass tube. The lipids from silica were extracted three times with 500 μl of chloroform/methanol (2/1 vol/vol), and the phosphate content in the combined extracts was measured. The sphingomyelin content was quantified by use of a standard curve of phosphate amount for sphingomyelin and normalized to total phosphate.
Preparation of the Subcellular Fractions to Measure Ceramide Synthase
Two subcellular fractions were tested: mitochondrial and microsomal. The fractions were prepared as described elsewhere (22), with minor modification. Cells were scraped and pelleted by centrifugation at 1000 × g at 4°C. Cells were washed twice with cold phosphate-buffered saline and resuspended in 300 μl of homogenizing buffer (20 mM Hepes [pH 7.4], 0.25 M sucrose, and 10 μg/ml leupeptin and 10 μg/mg antipain). Cells were homogenized and spun at 800 × g for 10 min to remove nuclei and debris. The supernatant was centrifuged at 10,000 × g for 30 min to precipitate mitochondria, and the pellet was resuspended in the homogenizing buffer. The supernatant was further centrifuged at 100,000 × g for 60 min. The transparent microsomal membrane pellet was resuspended in the homogenizing buffer, which contained 0.2% Triton X-100, and used for the ceramide synthase assay as described below. Protein contents were determined by use of the BCA protein assay kit, according to the manufacturer's instructions (Pierce, Rockford, IL).
Assay for Ceramide Synthase
Ceramide synthase activity in LLC-PK1 cells was measured as described elsewhere (9), with minor modifications. If not specifically changed, the conditions of the reaction were as follows. The same amount of protein (50 to 100 μg) from each sample was incubated in a final volume of 1-ml reaction mixture that contained 2 mM MgCl2, 20 mM Hepes (pH 7.4), 20 μM defatted bovine serum albumin, 20 μM dihydrosphingosine (sphinganine), 70 μM unlabeled palmitoyl—CoA, and 0.2 μCi of [1-14C] palmitoyl—CoA (specific activity, 53.6 mCi/mmol) for 60 min at 37°C. The reaction was stopped by extraction of lipids by use of 2 ml of chloroform/methanol (1/2 vol/vol). The lower phase was dried under nitrogen and applied to a 20-cm silica gel TLC plate. Radiolabeled dihydroceramide was resolved by use of a solvent system of chloroform/methanol/1 N ammonium hydroxide (40/10/1 vol/vol/vol), identified by iodine vapor staining on the basis of comigration with ceramide type III standards, and quantified by a liquid scintillation counting. Data are expressed as the percentage of control at time point 0.
Assay for Acid or Neutral Sphingomyelinase
At the end of experiment, serum-free medium was removed, and the cells were washed two times with phosphate-buffered saline at 4°C. Protein extract from cells was prepared as follows. Cells were lysed in 200 μl of either neutral buffer containing 20 mM Hepes (pH 7.4), 10 mM MgCl2, 2 mM ethylenediaminetetraacetic acid, 5 mM dithiothreitol, 100 mM Na3 VO4, 100 mM NaMO4, 10 mM β-glycerophosphate, 750 μM ATP, 1 mM phenylmethylsulfonyl fluoride, 10 μM leupeptin, 10 μM of pepstatin, and 0.5% CHAPS or acid buffer that contained 0.2% Triton X100, 0.5% CHAPS, 1 mM phenylmethylsulfonyl fluoride, 10 μM leupeptin, and 10 μM pepstatin. After incubation for 20 min at 4°C, cells were homogenized. To obtain the protein extracts for neutral sphingomyelinase assay, centrifugation at 800 × g was used (23). Protein extracts for acid sphingomyelinase assay were obtained by centrifugation at 16,000 × g for 5 min, as suggested by Wiegmann et al. (23) and Jaffrézou et al. (24). The supernatants were kept at -80°C until used for the assay.
The activity of acid or neutral sphingomyelinase was measured as described elsewhere (23,24,25,26), with modification (7). For the acid sphingomyelinase assay, 10 nmoles of unlabeled sphingomyelin and 0.02 μCi of [N-methyl 14C] sphingomyelin (specific activity, 56.0 mCi/mmol) were added to a glass tube, and the lipids were dried under N2. Then 50 μl of acidic buffer containing 250 mM sodium acetate, 1 mM ethylenediaminetetraacetic acid (pH 5.0), and protein extract (30 to 50 μg) were added, and the volume was brought up to 150 μl. The samples were then incubated for 90 min and 2 h at 37°C for acid and neutral sphingomyelin assay, respectively. For the neutral sphingomyelinase assay, the reaction was carried out in neutral buffer that contained 20 mM Hepes (pH 7.4) and 1 mM MgCl2 (pH 7.4). At the end of incubation, the reaction was stopped by the addition of 0.8 ml of chloroform/methanol (2/1 vol/vol) and 0.25 ml of H2O. The samples were then vortexed, centrifuged, and the upper phase (0.3 ml), which contained radioactive phosphatidyllcholine, a product of hydrolysis of [14C] sphingomyelin, was counted by scintillation counting. The data are expressed as pmol of [14C] sphingomyelin hydrolyzed/mg of protein per h. In the used cells, the activity of acid sphingomyelinase in 16,000 × g supernatants was not different from the activity in 800 × g supernatants, as had been observed in some studies (27,28,29,30,31,32).
Determination of DNA Strand Breaks
The residual double-strand DNA was measured by the alkaline unwinding assay and determination of ethidium bromide fluorescence used in our previous studies (11,13,14,15).
In Situ Detection of DNA Fragmentation
The terminal deoxynucleotidyl transferase nick end labeling technique was used to detect in situ DNA strand breaks, as described in our previous studies (15). Cells were grown to confluence on two-chamber tissue culture slides. At the end of the experiment, the cells were fixed in 4% paraformaldehyde and then treated in 1% H2O2 in phosphate-buffered saline for 10 min to remove endogenous peroxidase activity. After labeling with digoxigenin-labeled nucleotides, the cells were treated with peroxidase-conjugated anti-digoxigenin antibody and stained with peroxide substrate by use of the in situ ApopTag Plus labeling method (Oncor, Gaithersburg, MD).
Determination of Cell Injury
Cell viability was determined using trypan blue exclusion, as described elsewhere (11,12,13,14,15). Irreversible cell death was measured by use of lactate dehydrogenase (LDH) release (LDH kit, DG-1340-K, Sigma).
Statistical Analyses
Results are means ± SEM. Statistical significance was determined by unpaired t test and ANOVA. P < 0.05 was considered to be statistically significant.
Results
Effect of H2O2 on Ceramide Generation and Sphingomyelin Content in LLC-PK1 Cells
In our previous studies, we showed that exposure of LLC-PK1 cells to H2O2 results in DNA strand breaks, DNA fragmentation, and cell death (11,12). However, the precise mechanisms of cellular injury remained to be elucidated. To examine whether oxidant stress results in increased ceramide synthase activation, leading to ceramide increase, we first examined the time course of the effect of H2O2 on ceramide generation in LLC-PK1 cells. The exposure of cells to H2O2 (1 mM) resulted in a rapid increase in ceramide generation (Figure 1). The intracellular ceramide level was significantly increased from 15.9 ± 1.5 to 24.0 ± 1.7 pmol/nmol phosphate (n = 10 to 13, P < 0.002) at 5 min after exposure to H2O2 and thereafter increased continuously up to 60 min (38.5 ± 2.5 pmol/nmol phosphate, n = 10 to 13, P < 0.0001). The value for ceramide in control cells was slightly increased from 12.9 ± 1.5 pmol/nmol phosphate at time point 0 to 20.3 ± 1.0 pmol/nmol phosphate (n = 10 to 13) at 60 min incubation. However, compared with the control value at 60-min incubation, H2O2 resulted in significant increase in ceramide level at the same time point (P < 0.0001). Thus, our data showed that exposure of cells to H2O2 indeed resulted in significant increase in intracellular ceramide level in LLC-PK1 cells. In our previous study, we showed that H2O2 caused significant DNA fragmentation in LLC-PK1 cells in a time-dependent manner, which started 30 min after exposure of cells to H2O2, and, under those conditions, it caused significant cell death at 60 min after exposure (12). Taken together, these data indicate that H2O2-induced enhanced ceramide generation occurs very early, before any evidence of DNA damage or cell death (see below).
Time course of the effect of hydrogen peroxide (H2O2) on ceramide generation. Cells were incubated with 1 mM H2O2 in serum-free Dulbecco's modified Eagle's medium (DMEM) for different time periods (0, 1, 5, 15, 30, and 60 min). Ceramide was measured by the in vitro diacylglycerol kinase assay, as described in the Materials and Methods section. Results are means ± SEM, n = 10 to 13. *P < 0.002 and **P < 0.0001, compared with time point 0. ▪, control cells; ▴, H2O2-treated cells.
Effect of H2O2 on Ceramide Synthase Activity in LLC-PK1 Cells
Preliminary experiments were performed to test the optimal conditions (concentration of substrate and incubation time) for the ceramide synthase activity assay in LLC-PK1 cells. We observed that the maximum rate of the enzymatic reaction was reached with 20 μM dihydrosphingosine and 60 min incubation, as had been suggested elsewhere by Bose et al. (9) (Figure 2A). Under these conditions, the enzyme activity was analyzed in mitochondrial and microsomal fractions (Figure 2B), where high ceramide synthase activity is usually observed (29,30). Apparent kinetic parameters were calculated by use of the Lineweaver-Burk equation. The ceramide synthase reaction in mitochondrial fraction was characterized by Vmax = 0.62 pmol/mg per min and Km = 0.19 μM. In microsomal fractions, Vmax was 1.07 pmol/mg per min and Km was 0.09 μM. Thus, in further experiments we used microsomal fraction because it contained a higher activity of ceramide synthase.
Incorporation of 14C-palmitoyl CoA to form ceramide by ceramide synthase. (A) The optimal conditions for ceramide synthase assay in LLC-PK1 cells (75 μg microsomal protein, 0.2 μCi 14C-palmitoyl CoA) were tested by use of several incubation times (1, 5, 15, 30, and 60 min) and different concentration of dihydrosphingosine (1 [▪], 5 [▴], 10 [▾], and 20 [⋄] μM). (B) Incorporation of 14C-palmitoyl CoA to form ceramide by ceramide synthase was compared by use of mitochondrial (▴) and microsomal (▪) subcellular fractions, as described in the Materials and Methods section (20 μM of dihydrosphingosine, 60 min incubation). Regression analysis was made by use of the Prism 2 program (GraphPad Software).
We next examined whether H2O2-induced enhanced generation of ceramide is due to activation of ceramide synthase or sphingomyelinases. To examine whether oxidant stress results in ceramide synthase activation, we determined the time course of the effect of H2O2 on ceramide synthase activation in LLC-PK1 cells. The exposure of cells to H2O2 resulted in a rapid increase of ceramide synthase activity in microsomes at 5 min (119 ± 3% of control, n = 8, P < 0.005) that continuously increased up to 60 min (145 ± 7% of control, n = 8, P < 0.0001, Figure 3). These data indicate that the H2O2-induced ceramide generation is due to the activation of ceramide synthase in LLC-PK1 cells.
Time course of H2O2 effect on ceramide synthase activity in microsomes. Cells were incubated with 100 μM H2O2 in serum-free DMEM for different time periods (0, 1, 5, 30, and 60 min). Ceramide synthase in microsomes was measured as described in the Materials and Methods section. Results are means ± SEM, n = 8. *P < 0.005 and **P < 0.0001, compared with control at time point 0. ▪, control cells; ▴, H2O2-treated cells.
Effect of H2O2 on Sphingomyelin Content and Sphingomyelinase Activity in LLC-PK1 Cells
To determine the contribution of sphingomyelinases to H2O2-induced increase in ceramide, we measured the sphingomyelin content in LLC-PK1 cells exposed to H2O2. As shown in Figure 4A, there was no difference in sphingomyelin content between control cells and cells exposed to H2O2, which suggests little contribution of sphingomyelinases to oxidant-induced ceramide generation in renal tubular epithelial cells. We also measured the activity of acid or neutral sphingomyelinase in cells exposed to H2O2. In contrast to previous studies, which suggested that the contribution of sphingomyelinases to H2O2-induced ceramide generation (6), exposure of LLC-PK1 cells to H2O2 did not result in any significant change in acid or neutral sphingomyelinase activity (Figure 4, B and C). Taken together, these data strongly indicate that H2O2 is a regulator of ceramide synthase but not of sphingomyelinase in renal tubular epithelial cells. The data also indicate that the pathway that triggers ceramide generation varies with cell types even in response to the same stimuli.
Effect of H2O2 on sphingomyelin content (A) and activity of acid (B) or neutral (C) sphingomyelinase. Cells were exposed to 1 mM H2O2 for 60 min, and then sphingomyelin content (n = 7 to 8) and activity of acid or neutral sphingomyelinase (n = 5 to 6) were measured as described in the Materials and Methods section. Results are means ± SEM.
Effect of Inhibitor of Ceramide Synthase, Fumonisin B1, on H2O2-Induced Ceramide Generation
In a separate study, we examined the ability of a specific inhibitor of ceramide synthase, fumonisin B1, to suppress H2O2-induced ceramide generation in LLC-PK1 cells. As shown in Figure 5, there was no significant difference in the ceramide level between control cells pretreated with fumonisin B1 and those without it (19.6 ± 0.4 and 20.4 ± 4.2 pmol/nmol phosphate, respectively, n = 3). However, fumonisin B1 did indeed prevent H2O2-induced ceramide generation (at 60 min, control: 20.4 ± 4.2 pmol/nmol phosphate; H2O2 alone, 45.2 ± 3.3 pmol/nmol phosphate; H2O2 + fumonisin B1, 21.2 ± 2.8 pmol/nmol phosphate; n = 3, P < 0.005).
Effect of inhibitor of ceramide synthase fumonisin B1 (fumo) on H2O2-induced ceramide generation. Cells were preincubated with 50 μM fumonisin B1 in serum-free DMEM for 30 min and then exposed to 1 mM H2O2 for 60 min. Results are means ± SEM, n = 3. *P < 0.0001 compared with control, and **P < 0.005, compared with cells exposed to H2O2 alone.
Effect of Fumonisin B1 on H2O2-Induced DNA Damage
To determine the role of ceramide/ceramide synthase in oxidant-induced DNA damage, we examined the effect of the inhibition of ceramide synthase using fumonisin B1 on H2O2-induced DNA strand breaks. For that study, we used a 60-min time point after exposure of cells to H2O2 (11). As shown in Figure 6, fumonisin B1 was highly protective against H2O2-induced DNA strand breaks (residual double-strand DNA at 60 min: control, 86 ± 1%; H2O2 alone, 53 ± 2%; H2O2 + fumonisin B1, 81 ± 7%; n = 3 to 4, P < 0.01). In a separate study, we examined the effect of fumonisin B1 on H2O2-induced DNA fragmentation (Figure 7). Fumonisin B1 also provided a marked protection against H2O2-induced DNA fragmentation in LLC-PK1 cells. Taken together, these data suggest that the inhibition of ceramide synthase is highly protective against H2O2-induced DNA damage to renal tubular epithelial cells.
Effect of fumonisin B1 on H2O2-induced DNA strand breaks. Cells were preincubated with fumonisin B1 (50 μM) in serum-free DMEM for 30 min and then exposed to 1 mM H2O2 for 60 min. DNA strand breaks were determined by the alkaline unwinding assay. Results are means ± SEM, n = 3. *P < 0.0002, compared with cells exposed to H2O2 alone.
Effect of fumonisin B1 on H2O2-induced in situ DNA fragmentation. Cells were preincubated 50 μM with fumonisin B1 in serum-free DMEM for 30 min and then exposed to 1 mM H2O2 for 120 min. In situ DNA fragmentation was detected by the terminal deoxynucleotidyl transferase nick end labeling assay. (A) control, (B) H2O2 alone, and (C) H2O2 + fumonisin B1. Results are representative of four experiments.
Effect of Fumonisin B1 on H2O2-Induced Cell Death
Finally, we examined the effect of fumonisin B1 on H2O2-induced cell death. In our previous study (12), we showed that 60 min of H2O2 treatment resulted in a significant cell death of LLC-PK1 cells. Thus, we used a 60-min time point after exposure to H2O2. Irreversible cell death was accessed by use of LDH release. In preliminary experiments, we measured LDH release by LLC-PK1 cells treated with 1 mM H2O2 within 24 h after treatment. Our observation was that the release started to increase after 1 h of treatment (10%) and reached its apparent plateau (30% to 40%) between 3 and 24 h after treatment. As shown in Figure 8, fumonisin B1 did not increase the death of untreated cells (1.05 ± 0.35% versus 1.50 ± 0.35% in control). However, it prevented H2O2-induced cell death (H2O2 alone, 10.24 ± 0.89%; H2O2 + fumonisin B1, 2.90 ± 0.64%, n = 10 to 16). A similar pattern was obtained when cell death was measured by use of trypan blue staining (data not shown). Therefore, fumonisin B1 provided a significant protection against cell death in the used model. Taken together, our data provide the first evidence that H2O2 is a regulator of ceramide synthase rather than sphingomyelinases and that H2O2-induced enhanced generation of ceramide through the activation of ceramide synthase is a determinant of DNA damage and cell death in oxidant injury to renal tubular epithelial cells.
Effect of fumonisin B1 on H2O2-induced cell death. Cells were preincubated with fumonisin B1 (50 μM) for 30 min and then exposed to 1 mM H2O2 for 60 min. Cell viability was measured by lactate dehydrogenase (LDH) release (% of total LDH). Results are means ± SEM, n = 8 to 13. *P < 0.0001 compared with control cells, and **P < 0.04 compared with cells exposed to H2O2 alone.
Discussion
Currently, there is little information available on a role of ceramide in oxidant injury, and it is not known whether activation of either sphingomyelinases or ceramide synthase is involved in oxidant-induced ceramide generation in any tissue. Most studies that have used the apoptotic model of cell injury have shown that increased ceramide generation is mediated by activation of sphingomyelinases (2,4,5). Exposure of leukemic or endothelial cells to H2O2 has been shown to result in increased ceramide generation with concomitant decrease in sphingomyelin content, which is suggestive of activation of sphingomyelinase (6). In an in vitro model of hypoxic injury, in which reactive oxygen metabolites have been implicated, to PC12 cells (4), ceramide has been shown to be increased due to activation of neutral sphingomyelinase. Thymocytes, concavalin A—activated T cells, and lipopolysaccharide-activated B cells from knock-out mice to acid sphingomyelinase (aSMase -/-) treated in vitro with anti-Fas antibody exhibited apoptosis like wild-type cells (aSMase [+/+]). Both aSMase (-/-) and aSMase (+/+) cells had increase of ceramide after anti-Fas treatment, which suggests the involvement of other ceramide formation pathway (31). Along these lines of evidence, in our recent study we showed that hypoxia results in an increase in ceramide generation through ceramide synthase activation in LLC-PK1 cells (14).
In the present study, we have shown that exposure of LLC-PK1 cells to H2O2 resulted in a rapid increase in ceramide generation and activation of ceramide synthase but not sphingomyelinases, which preceded any evidence of DNA damage and cell death. Exposure of LLC-PK1 cells to H2O2 did not result in any significant change in activity of sphingomyelinases. This is supported by the finding that there was no difference in sphingomyelin content between control cells and cells exposed to H2O2. Taken together, these data indicate that the target of the enzyme responsible for ceramide generation in oxidant stress is ceramide synthase rather than sphingomyelinases. Our observation that H2O2 resulted in an increase in ceramide level without any significant increase in activity of sphingomyelinases is similar to that reported by Zager et al. (7), in which iron loading resulted in increase in ceramide level in human proximal tubular cells. However, in their study, pretreatment with fumonisin B1 failed to prevent ceramide generation, and ceramidase activity was inhibited under their condition. On the basis of those data, they concluded that the inhibition of ceramidase is responsible for the rise in ceramide in response to iron-mediated oxidant stress. Although the reasons for the differences are not known, the different response of ceramide generation in response to various stimuli by different cell types or conditions studied have been reported. For example, H2O2 has been shown to activate sphingomyelinases (6) in other cells, and the anti-cancer agent, daunorubicin (9), in which reactive oxygen metabolites have been implicated (32), has been shown to result in ceramide synthase activation with a concomitant increase in sphingomyelin content.
To examine a role of ceramide synthase in regulation of ceramide, which results in DNA damage and cell death in oxidant injury, we used a specific inhibitor of ceramide synthase, fumonisin B1. This compound has very similar structure to sphingosine or sphinganine, which is a substrate for ceramide synthase, and thus blocks ceramide synthase, resulting in a decrease in intracellular ceramide level (9,10,16,17,18). Because of the nature of the action of fumonisin B1, it is highly unlikely for this compound to inhibit sphingomyelinases, which is supported by the inability of fumonisin B1 to prevent cell death in a model in which sphingomyelinases have been implicated (4,9). Recent studies have shown the protective effect of fumonisin B1 against apoptotic cell death in response to stimuli (9,10,16) in which ceramide synthase is involved for ceramide generation. In fact, our data confirmed the ability of fumonisin B1 to suppress H2O2-induced ceramide generation in LLC-PK1 cells and showed that fumonisin B1 provided a marked protection against H2O2-induced DNA damage and cell death. In a recent study, we showed that exogenous ceramide results in DNA damage and cell death in LLC-PK1 cells (14) similar to the effect of H2O2. Taken together, these data strongly indicate that ceramide synthase plays a major role of oxidant-induced ceramide generation, which results in DNA damage and cell death in oxidant injury to renal tubular epithelial cells.
Oxidant stress has been shown to result in DNA fragmentation in various types of cells (33), and scavengers of reactive oxygen metabolites prevent oxidant-induced DNA fragmentation (34,35). We have previously shown that oxidant stress induces endonuclease activation as an early event, leading to DNA damage and cell death in renal tubular epithelial cells (11). Although the sequential consequences of ceramide on endonuclease activation and cell death still remain to be elucidated, our data indicate that activation of the ceramide synthase-dependent pathway is one of the mechanisms by which oxidant stress causes endonuclease activation, leading to DNA damage and cell death in renal tubular epithelial cells.
In summary, the present study has demonstrated that H2O2 increased ceramide synthase activity but not sphingomyelinases, which resulted in enhanced ceramide production before any evidence of DNA damage and cell death in LLC-PK1 cells. The inhibition of ceramide synthase was able to suppress oxidant-induced ceramide production and prevented DNA damage and cell death. Our data provided the first evidence that oxidant stress is a mediator of ceramide synthase activity, resulting in enhanced ceramide generation in oxidant injury to renal epithelial tubule cells.
Acknowledgments
This work was supported in part by the VA Merit Review (0002) and the Department of Defense (DAMD 17-00-1-0131). The fellowship grant to Simone M. R. Camargo was supported by the Fundação de Amparo a Pesquisa do Estado de São Paulo, Brazil. The authors thank Ellen Satter for her secretarial assistance.
- © 2001 American Society of Nephrology
References
