Palmitate-induced Apoptosis Can Occur through a Ceramide-independent Pathway

Cytotoxic accumulation of long chain fatty acids has been proposed to play an important role in the pathogenesis of diabetes mellitus and heart disease. To explore the mechanism of cellular lipotoxicity, we cultured Chinese hamster ovary cells in the presence of media supplemented with fatty acid. The saturated fatty acid palmitate, but not the monounsaturated fatty acid oleate, induced programmed cell death as determined by annexin V positivity, caspase 3 activity, and DNA laddering. De novo ceramide synthesis increased 2.4-fold with palmitate supplementation; however, this was not required for palmitate-induced apoptosis. Neither biochemical nor genetic inhibition of de novo ceramide synthesis arrested apoptosis in Chinese hamster ovary cells in response to palmitate supplementation. Rather, our data suggest that palmitate-induced apoptosis occurs through the generation of reactive oxygen species. Fluorescence of an oxidant-sensitive probe was increased 3.5-fold with palmitate supplementation indicating that production of reactive intermediates increased. In addition, palmitate-induced apoptosis was blocked by pyrrolidine dithiocarbamate and 4,5-dihydroxy-1,3-benzene-disulfonic acid, two compounds that scavenge reactive intermediates. These studies suggest that generation of reactive oxygen species, independent of ceramide synthesis, is important for the lipotoxic response and may contribute to the pathogenesis of diseases involving intracellular lipid accumulation. Cytotoxic accumulation of long chain fatty acids has been proposed to play an important role in the pathogenesis of diabetes mellitus and heart disease. To explore the mechanism of cellular lipotoxicity, we cultured Chinese hamster ovary cells in the presence of media supplemented with fatty acid. The saturated fatty acid palmitate, but not the monounsaturated fatty acid oleate, induced programmed cell death as determined by annexin V positivity, caspase 3 activity, and DNA laddering. De novo ceramide synthesis increased 2.4-fold with palmitate supplementation; however, this was not required for palmitate-induced apoptosis. Neither biochemical nor genetic inhibition of de novo ceramide synthesis arrested apoptosis in Chinese hamster ovary cells in response to palmitate supplementation. Rather, our data suggest that palmitate-induced apoptosis occurs through the generation of reactive oxygen species. Fluorescence of an oxidant-sensitive probe was increased 3.5-fold with palmitate supplementation indicating that production of reactive intermediates increased. In addition, palmitate-induced apoptosis was blocked by pyrrolidine dithiocarbamate and 4,5-dihydroxy-1,3-benzene-disulfonic acid, two compounds that scavenge reactive intermediates. These studies suggest that generation of reactive oxygen species, independent of ceramide synthesis, is important for the lipotoxic response and may contribute to the pathogenesis of diseases involving intracellular lipid accumulation. Zucker diabetic fatty Chinese hamster ovary bovine serum albumin propidium iodide BOC-asp(OMe)-fluoromethylketone reactive oxygen species 6-carboxy-2′,7′-dichlorodihydrofluorescein diacetate, di(acetoxymethyl ester) pyrrolidine dithiocarbamate 4,5-dihydroxy-1,3-benzene-disulfonic acid fluorescein isothiocyanate Intracellular accumulation of long chain fatty acids in nonadipose tissues is associated with cellular dysfunction and cell death and may ultimately contribute to the pathogenesis of disease. For example, lipotoxic accumulation of long chain fatty acids in the pancreatic β-cells of the Zucker diabetic fatty (ZDF)1 rat leads to the development of diabetes caused by β-cell death (1Shimabukuro M. Zhou Y.-T. Levi M. Unger R.H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 2498-2502Crossref PubMed Scopus (1012) Google Scholar). The ZDF rat also develops cardiomyopathy secondary to cardiomyocyte lipid accumulation (2Zhou Y. Grayburn P. Karim A. Shimabukuro M. Higa M. Baetens D. Orci L. Unger R. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 1784-1789Crossref PubMed Scopus (1072) Google Scholar). Similarly, human diabetic cardiomyopathy is associated with increased myocardial triglyceride content, which has been proposed to contribute to susceptibility to arrhythmia and reduced contractile function (3Rodrigues B. Cam M.C. McNeill J.H. J. Mol. Cell Cardiol. 1995; 27: 169-179Abstract Full Text PDF PubMed Scopus (183) Google Scholar). Patients with inherited defects of the mitochondrial fatty acid oxidation pathway also show signs of lipid accumulation in the heart. This may contribute to the development of cardiomyopathy or sudden death in these patients (4Kelly D.P. Strauss A.W. N. Engl. J. Med. 1994; 330: 913-919Crossref PubMed Scopus (325) Google Scholar). Lastly, triglyceride accumulation in liver and muscle of the A-ZIP/F-1 "fatless" mice has been proposed to induce the insulin resistance of these peripheral tissues (5Kim J.K. Gavrilova O. Chen Y. Reitman M.L. Shulman G.I. J. Biol. Chem. 2000; 275: 8456-8460Abstract Full Text Full Text PDF PubMed Scopus (361) Google Scholar). Although intracellular long chain fatty acid accumulation is associated with numerous pathophysiologic states, the mechanism of this lipotoxicity is not fully understood. In mouse and human fibroblasts and in cultured human endothelial cell monolayers, high concentrations (70–300 μm) of long chain saturated fatty acids inhibit cell proliferation and lead to cell death (6Chun-Lan Z. Lyngmo V. Nordoy A. Thromb. Res. 1992; 65: 65-75Abstract Full Text PDF PubMed Scopus (23) Google Scholar, 7Gordon G.B. Exp. Mol. Pathol. 1977; 27: 262-276Crossref PubMed Scopus (38) Google Scholar, 8Doi O. Doi F. Schroeder F. Alberts A.W. Vagelos P.R. Biochim. Biophys. Acta. 1978; 509: 239-250Crossref PubMed Scopus (85) Google Scholar, 9Rosenthal M.D. Lipids. 1981; 16: 173-182Crossref PubMed Scopus (33) Google Scholar). Evidence is emerging that long chain fatty acids induce cell death through apoptosis. Cultured neonatal rat cardiomyocytes, pancreatic β cells of the ZDF rat, and the hematopoietic precursor cell lines LyD9 and WEHI-231 demonstrate signs of apoptosis, including DNA laddering and caspase activation, following fatty acid supplementation (1Shimabukuro M. Zhou Y.-T. Levi M. Unger R.H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 2498-2502Crossref PubMed Scopus (1012) Google Scholar, 10Paumen M.B. Ishida Y. Muramatsu M. Yamamoto M. Honjo T. J. Biol. Chem. 1997; 272: 3324-3329Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar, 11Hickson-Bick D.L. Buja M.L. McMillin J.B. J. Mol. Cell Cardiol. 2000; 32: 9-511Abstract Full Text PDF Scopus (121) Google Scholar, 12de Vries J.E. Vork M.M. Roemen T.H.M. de Jong Y.F. Cleutjens J.P.M. van der Vusse G.J. van Bilsen M. J. Lipid Res. 1997; 38: 1384-1394Abstract Full Text PDF PubMed Google Scholar). Notably, fatty acid-induced apoptosis is specific for the saturated fatty acids palmitate (C16:0) and stearate (C18:0) and does not occur with saturated fatty acids of carbon chain length ranging from C4-C14 or with unsaturated fatty acids (10Paumen M.B. Ishida Y. Muramatsu M. Yamamoto M. Honjo T. J. Biol. Chem. 1997; 272: 3324-3329Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar,12de Vries J.E. Vork M.M. Roemen T.H.M. de Jong Y.F. Cleutjens J.P.M. van der Vusse G.J. van Bilsen M. J. Lipid Res. 1997; 38: 1384-1394Abstract Full Text PDF PubMed Google Scholar). Because palmitate and stearate, but not unsaturated fatty acids, are precursors for de novo ceramide synthesis, it has been hypothesized that fatty acid-induced apoptosis occurs through this pathway. Ceramide is a lipid second messenger involved in the apoptotic response induced by tumor necrosis factor α, ionizing radiation, and heat shock (13Kolesnick R.N. Kronke M. Annu. Rev. Physiol. 1998; 60: 643-665Crossref PubMed Scopus (728) Google Scholar). These stimuli are thought to increase ceramide by hydrolysis of sphingomyelin rather than de novobiosynthesis. The downstream signaling pathways through which ceramide initiates apoptosis remain unclear, but several possible components have been identified. Direct targets of ceramide include ceramide-activated protein kinase (CAPK, KSR), protein kinase Cζ, and ceramide-activated protein phosphatase (14Mathias S. Pena L.A. Kolesnick R.N. Biochem. J. 1998; 335: 465-480Crossref PubMed Scopus (618) Google Scholar). Further downstream, ceramide signaling can affect the mitogen-activated protein kinase and c-Jun N-terminal kinase signaling cascades or the activation of NF-KB and ultimately lead to growth arrest and apoptosis (14Mathias S. Pena L.A. Kolesnick R.N. Biochem. J. 1998; 335: 465-480Crossref PubMed Scopus (618) Google Scholar,15Obeid L.M. Hannun Y.A. J. Cell. Biochem. 1995; 58: 191-198Crossref PubMed Scopus (235) Google Scholar). In the present study, we explored the role of de novoceramide synthesis in fatty acid-induced lipotoxicity. Specifically, we utilized Chinese hamster ovary cells (CHO), a cell line amenable to genetic manipulation, to determine the mechanism whereby palmitate causes cell death. Lipotoxicity in CHO cells is specific for the saturated fatty acid palmitate and does not occur with the monounsaturated fatty acid oleate. We demonstrate that CHO cells do not require de novo ceramide synthesis for palmitate-induced apoptosis. Rather, our studies suggest that palmitate supplementation leads to the generation of reactive intermediates that initiate apoptosis. Cellular damage and death from reactive intermediates generated by saturated fatty acids may contribute to the pathogenesis of diseases such as diabetes mellitus. Palmitic acid, oleic acid, and cholesteryl oleate were purchased from Nu-Check Prep, BOC-asp(OMe)-fluoromethylketone (BAF) was from Enzyme Systems Products, fumonisin B1 from Biomol, [3H]serine was from Amersham Pharmacia Biotech, [14C]cholesteryl oleate from PerkinElmer Life Sciences, and egg ceramide was from Avanti Polar Lipids.l-Cycloserine, fatty acid-free bovine serum albumin (BSA), propidium iodide (PI), pyrrolidine dithiocarbamate (PDTC), and 4,5-dihydroxy-1,3-benzene-disulfonic acid (DBDA) were purchased from Sigma. 6-carboxy-2′,7′-Dicholorodihydrofluorescein diacetate, di(acetoxymethyl ester) (C-2938) was from Molecular Probes, and H2O2 was from Fisher Scientific. Chinese hamster ovary cells (American Type Culture Collection) and LY-B cells (gift from K. Hanada, National Institute of Infectious Diseases, Tokyo, Japan) were cultured in Dulbecco's modified Eagle's medium and Ham's F-12 medium (1:1) with 5% fetal bovine serum supplemented with 2 mml-glutamine, 50 units/ml penicillin G sodium, 50 units/ml streptomycin sulfate, and 1 mm sodium pyruvate. Where indicated, medium was supplemented with 500 μm palmitate or oleate. Fatty acid supplemented medium was prepared by modification of the method of Spector (16Spector A.A. Methods Enzymol. 1986; 128: 320-339Crossref PubMed Scopus (115) Google Scholar). Briefly, a 20 mm solution of fatty acid in 0.01 m NaOH was incubated at 70 °C for 30 min. Dropwise addition of 1 n NaOH facilitated solubilization of the fatty acid. Fatty acid soaps were complexed with 5% fatty acid-free BSA in phosphate-buffered saline at an 8:1 fatty acid to BSA molar ratio. The complexed fatty acid was added to the serum-containing cell culture medium to achieve a fatty acid concentration of 500 μm. The final fatty acid concentration in the medium was measured using a semimicroanalysis kit (Wako Chemicals). The final BSA concentration was measured using the Albumin Reagent (BCG, Sigma). The pH of the medium did not differ significantly with the addition of complexed fatty acid. PDTC, DBDA, BAF (40 mm stock in dimethyl sulfoxide), fumonisin B1 (10 mm stock in water), or l-cycloserine (0.5m stock in phosphate-buffered saline) was added to the cell culture medium where indicated. The pH was corrected when addition of the compound significantly altered the pH of the medium. Annexin V-FITC (Pharmingen 65874X) binding and PI staining were performed according to the recommended protocol and the cells were analyzed by flow cytometry (Becton Dickinson FACScan). Apoptotic cells were defined as 1) PI negative (indicating an intact plasma membrane) and 2) annexin V-FITC positive relative to cells incubated in the absence of palmitate. Each data point represents fluorescence analysis from 104 cells. Activity of the caspase 3 class of cysteine proteases was determined with the Colorimetric Caspase 3 activation assay (R&D Systems) according to the manufacturer's protocol. Ability of the cell lysate to cleave the reporter molecule was quantified spectrophotometrically at a wavelength of 415 nm using a microplate reader (Bio-Rad). The level of caspase enzymatic activity was normalized to cell lysate protein concentration (BCA assay, Pierce). DNA Laddering was assessed by modification of the protocol of Bialik et al. (17Bialik S. Geenen D.L. Sasson I.E. Cheng R. Horner J.W. Evans S.M. Lord E.M. Koch C.J. Kitsis R.N. J. Clin. Invest. 1997; 100: 1363-1372Crossref PubMed Scopus (347) Google Scholar). Briefly, 5 × 106 cells/sample were resuspended in 425 μl of lysis buffer (10 mm Tris, pH 8.0, 100 mm NaCl, 25 mm EDTA, 0.5% SDS). Proteinase K was added (25 μl of 20 mg/ml solution), and the samples were incubated overnight at room temperature. Protein was precipitated by the dropwise addition of 200 μl of 4 m NaCl. The samples were centrifuged at 10,000 × g for 30 min at 4 °C, and the supernatant was extracted with phenol:chloroform (1:1) and phenol:chloroform:isoamyl alcohol (25:24:1). The DNA was precipitated with EtOH and resuspended in Tris-EDTA buffer containing 70 μg/ml RNase. Equal quantities (5–30 μg) of each DNA sample were run on 1.4% agarose gels in TBE buffer (45 mm Tris, 45 mm boric acid, 1 mm EDTA, pH 8.0). Bands were detected by ethidium bromide staining. To measure ceramide synthesis, cells were plated at 2 × 104 cells/35-mm well. The following day, the cells were incubated with serine-free medium supplemented with 2.5–3 μCi of [3H]serine for 20 h to facilitate the incorporation of the tritium label into cellular serine and sphingomyelin pools (18Hanada K. Hara T. Fukasawa M. Yamaji A. Umeda M. Nishijima M. J. Biol. Chem. 1998; 273: 33787-33794Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar, 19Fukasawa M. Nishijima M. Hanada K. J. Cell Biol. 1999; 144: 673-685Crossref PubMed Scopus (151) Google Scholar). Following overnight labeling, the cells were fed serine-free medium with 2.5 μCi of [3H]serine ± 500 μm palmitate ± fumonisin or l-cycloserine for 4.25–5 h. Lipids were extracted (20Morand O. Fibach E. Dagan A. Gatt S. Biochim. Biophys. Acta. 1982; 711: 539-550Crossref PubMed Scopus (46) Google Scholar) and resuspended in chloroform containing 60 μg of egg ceramide and 40 μg of cholesteryl oleate. The lipids were separated by thin layer chromatography on silica gel plates (Whatman 4410 221) using CHCl3:MeOH:NH4OH (200:25:2.5) as the solvent. Samples were visualized by iodine vapor staining and radioactivity incorporated into the ceramide or cholesteryl oleate spots was determined by scintillation counting. The presence of radiolabeled ceramide was normalized to protein concentration (BCA assay, Pierce) and corrected for lipid recovery during extraction using 0.01 μCi of [14C]cholesteryl oleate as a recovery standard. Cells were plated at 1.4 × 105 cells/35-mm well. The following day, cells were supplemented with fatty acid media for 14 h. Prior to C-2938 loading, control cells were supplemented with medium containing 5 mm H2O2 for 1 h at 37 °C. Cells were washed with phosphate-buffered saline and incubated with 0.5 μm C-2938 in phosphate-buffered saline supplemented with 0.5 mm MgCl2 and 0.92 mmCaCl2 for 1 h at 37 °C. Cells were collected and resuspended in media containing 1 μm PI. C-2938 fluorescence was measured by flow cytometry on 104cells/sample (Becton Dickinson FACScan). Cells were gated for cell size and intact plasma membranes (PI negative). The fold increase in median fluorescence over unsupplemented cells was determined. The values reported are the average fold increase for three independent experiments. Differences among groups were compared by one-way analysis of variance in conjunction with the post hoc Scheffe test. To determine whether long chain saturated fatty acids induce cell death in CHO cells, cells were incubated in medium supplemented with 500 μmpalmitate complexed to BSA. The final fatty acid and albumin concentrations in the media were measured and yielded a molar ratio of fatty acid to albumin of 6.6:1. Although the normal physiologic ratio of fatty acid to albumin is ∼2:1, serum fatty acid levels in disease states (e.g. acute coronary syndromes) are elevated yielding ratios as high as 7.5:1 (21Kleinfeld A.M. Prothro D. Brown D.L. Davis R.C. Richieri G.V. DeMaria A. Am. J. Cardiol. 1996; 78: 1350-1354Abstract Full Text PDF PubMed Scopus (87) Google Scholar). Thus, our experimental system was designed to evaluate mechanisms of palmitate toxicity relevant to pathophysiologic states. CHO cells incubated with palmitate supplemented medium showed signs of growth arrest and cell death by 5 h. By 11 h ∼80% of CHO cells displayed cell shrinkage. The cells began to detach from the plate after 16 h in palmitate. Cytotoxicity was observed using medium supplemented with as little as 100 μm palmitate, and the degree of cell death correlated with the amount of palmitate supplementation from 100 to 500 μm (data not shown). We assayed palmitate supplemented CHO cells for annexin V binding, caspase 3 activity, and DNA laddering to determine whether cell death was occurring through apoptosis (Fig. 1). Early in apoptosis, phosphatidylserine is translocated from the inner to the outer leaflet of the plasma membrane. Annexin V, a membrane-impermeable protein, binds phosphatidylserine on intact cells only if phosphatidylserine is present on the outer leaflet. Flow cytometry was used to measure the binding of FITC-labeled annexin V to the surface of CHO cells after incubation in medium supplemented with 500 μm palmitate (Fig. 1 A). Cells with permeabilized plasma membranes were excluded from measurements of annexin V positivity with PI, a fluorescent DNA binding dye. CHO cells began to show annexin V binding after 10 h in palmitate. The percentage of cells binding annexin V increased until 70% of CHO cells were annexin V-positive after 16 h in palmitate. Notably, PI staining, an indication of cell death, followed the appearance of annexin V positivity and steadily increased after 16 h in palmitate supplemented medium (Fig. 1 A). Because activation of the cysteine protease, caspase 3, has been implicated as a common downstream effector of diverse apoptotic pathways, we measured cleavage of a colorimetric substrate specific to the caspase 3 class of cysteine proteases following 25 h of palmitate feeding. Caspase activity increased 9.2-fold with palmitate supplementation and was inhibited by BAF, a pan-caspase inhibitor (Fig. 1 B). DNA laddering, an end stage apoptotic event, was evident after 28 h in 500 μm palmitate and was inhibited when caspases were inhibited with BAF (Fig. 1 C). Taken together, the detection of phosphatidylserine externalization, membrane permeabilization, caspase activation, and DNA laddering following palmitate supplementation supports the hypothesis that the saturated fatty acid palmitate induces programmed cell death in CHO cells. To demonstrate the specificity of fatty acid-induced apoptosis in CHO cells, CHO cells were also incubated with medium containing 500 μm oleate. In contrast to the effects of palmitate, oleate, an 18-carbon monounsaturated fatty acid did not cause CHO cell death. In addition to the absence of morphological changes associated with palmitate feeding, oleate supplementation did not induce phosphatidylserine externalization (Fig. 1 A), caspase 3 activity (Fig. 1 B), or DNA laddering (Fig. 1 C). The inability of oleate to induce cell death is consistent with the hypothesis that this response is specific to saturated fatty acids as reported previously (10Paumen M.B. Ishida Y. Muramatsu M. Yamamoto M. Honjo T. J. Biol. Chem. 1997; 272: 3324-3329Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar, 12de Vries J.E. Vork M.M. Roemen T.H.M. de Jong Y.F. Cleutjens J.P.M. van der Vusse G.J. van Bilsen M. J. Lipid Res. 1997; 38: 1384-1394Abstract Full Text PDF PubMed Google Scholar). Thus, these results demonstrate that CHO cells are an appropriate model system in which to study the mechanism through which saturated fatty acids specifically induce programmed cell death. Prior studies have implicated that de novo synthesis of ceramide, a known inducer of apoptosis, is critical for palmitate-induced apoptosis (10Paumen M.B. Ishida Y. Muramatsu M. Yamamoto M. Honjo T. J. Biol. Chem. 1997; 272: 3324-3329Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar, 22Shimabukuro M. Higa M. Zhou Y.-T. Wang M.-Y. Newgard C.B. Unger R.H. J. Biol. Chem. 1998; 273: 32487-32490Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar). Serine palmitoyltransferase catalyzes the first and rate-limiting step ofde novo ceramide synthesis (Fig.2 A, adapted from Ref. 23Hannun Y.A. Luberto C. Trends Cell Biol. 2000; 10: 73-80Abstract Full Text Full Text PDF PubMed Scopus (643) Google Scholar). This enzyme has high specificity for palmitoyl CoA, the activated form of palmitate, whereas saturated fatty acids such as stearate are the preferred substrates for ceramide synthase (24Riboni L. Viani P. Bassi R. Prinetti A. Tettamanti G. Prog. Lipid Res. 1997; 36: 153-195Crossref PubMed Scopus (189) Google Scholar). Serine palmitoyltransferase and ceramide synthase are specifically inhibited by l-cycloserine and fumonisin B1, respectively. To determine whether palmitate supplementation induces ceramide synthesis in CHO cells, we labeled cells to equilibrium with [3H]serine to incorporate the label into the cellular serine and sphingomyelin pools (19Fukasawa M. Nishijima M. Hanada K. J. Cell Biol. 1999; 144: 673-685Crossref PubMed Scopus (151) Google Scholar, 25Tepper A. Boesen-de Cock J. de Vries E. Borst J. van Blitterswijk W. J. Biol. Chem. 1997; 1997: 24308-24312Abstract Full Text Full Text PDF Scopus (115) Google Scholar). Then, the cells were supplemented with [3H]serine and palmitate for 4.25 h before measuring the production of radiolabeled ceramide. Palmitate feeding for 4.25 h increased the synthesis of labeled ceramide by 2.4-fold (Fig. 2 B). This increase in ceramide production was attributable to de novo ceramide synthesis and not cleavage of sphingomyelin because it was completely inhibited by the inclusion of fumonisin B1 or l-cycloserine. To determine whether the increase in de novoceramide synthesis is critical for palmitate-induced apoptosis, CHO cells were incubated with medium supplemented with 500 μmpalmitate and 100 μm fumonisin B1 or 1 mml-cycloserine. These concentrations of inhibitors completely blocked the increase in ceramide synthesis associated with palmitate supplementation (Fig. 2 B). Apoptosis was assessed by caspase 3 activity and DNA laddering. Surprisingly, inhibition ofde novo ceramide synthesis did not rescue the morphological changes (cell shrinkage and detachment) associated with palmitate feeding. Inhibition of de novo ceramide synthesis blunted but did not completely prevent caspase activity (Fig.3 A), with reduced relative levels of caspase activity at each time point measured. Biochemical inhibition of de novo ceramide synthesis also did not prevent DNA laddering induced by 28 h of palmitate supplementation (Fig.3 B). To verify that de novo ceramide synthesis is not required for cell death, we assayed for palmitate-induced apoptosis in a mutant CHO cell line incapable of de novo ceramide synthesis. LY-B cells lack serine palmitoyltransferase activity, the rate-limiting step in de novo ceramide synthesis (18Hanada K. Hara T. Fukasawa M. Yamaji A. Umeda M. Nishijima M. J. Biol. Chem. 1998; 273: 33787-33794Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar). We verified that LY-B cells failed to stimulate ceramide synthesis in response to palmitate supplementation (Fig. 4 A). Despite the absence of de novo ceramide synthesis in these cells, palmitate supplementation induced the morphological changes associated with cell death. Additionally, caspase 3 was activated in LY-B cells with palmitate supplementation, although the magnitude of this increase was diminished compared with wild-type CHO cells (Fig. 4 B). DNA laddering occurred in response to palmitate feeding in a manner indistinguishable from wild-type cells (Fig. 4 C). Taken together, the studies with l-cycloserine, fumonisin B1, and LY-B cells demonstrate by both biochemical and genetic means thatde novo ceramide synthesis is not required for palmitate-induced apoptosis in CHO cells. We next attempted to identify the mechanism whereby palmitate supplementation induces apoptosis. Evidence is emerging that free fatty acids can stimulate the production of reactive oxygen species (ROS) to a level that exceeds the intrinsic capacity of the cell to detoxify these molecules (26Carlsson C. Borg L.A. Welsh N. Endocrinology. 1999; 140: 3422-3428Crossref PubMed Google Scholar, 27Inoguchi T. Li P. Umeda F., Yu, H.Y. Kakimoto M. Imamura M. Aoki T. Etoh T. Hashimoto T. Naruse M. Sano H. Utsumi H. Nawata H. Diabetes. 2000; 49: 1939-1945Crossref PubMed Scopus (1272) Google Scholar). Moreover, ROS have been implicated as important regulators of apoptotic pathways (28Gottlieb E. Vander Heiden M.G. Thompson C.B. Mol. Cell. Biol. 2000; 20: 5680-5689Crossref PubMed Scopus (309) Google Scholar) and thus may play a role in palmitate-induced apoptosis. To determine whether reactive intermediates were generated with palmitate supplementation, we measured cell fluorescence following C-2938 loading as a marker for oxidative intermediates. C-2938 is a nonfluorescent, membrane permeable probe that becomes fluorescent upon reaction with ROS within the cell and can be detected by flow cytometry. Although there is some debate regarding the specificity of this assay (29Rota C. Fann Y.C. Mason R.P. J. Biol. Chem. 1999; 274: 28161-28168Abstract Full Text Full Text PDF PubMed Scopus (199) Google Scholar), fluorescence of C-2938 has been widely used as a measure of oxidative stress and as a marker for ROS in cells (30Nishikawa T. Edelstein D. Du X.L. Yamagishi S. Matsumura T. Kaneda Y. Yorek M.A. Beebe D. Oates P.J. Hammes H.P. Giardino I. Brownlee M. Nature. 2000; 404: 787-790Crossref PubMed Scopus (3658) Google Scholar, 31Hockenbery D.M. Oltvai Z.N. Yin X.M. Milliman C.L. Korsmeyer S.J. Cell. 1993; 75: 241-251Abstract Full Text PDF PubMed Scopus (3292) Google Scholar, 32Quillet-Mary A. Jaffrezou J.P. Mansat V. Bordier C. Naval J. Laurent G. J. Biol. Chem. 1997; 272: 21388-21395Abstract Full Text Full Text PDF PubMed Scopus (445) Google Scholar). Supplementation of CHO cells with palmitate resulted in an increase in C-2938 fluorescence. Increased C-2938 fluorescence was observed as early as 5 h, and the levels of fluorescence increased with increasing periods of palmitate supplementation. Fig.5 A shows that following 14 h of palmitate supplementation, the level of fluorescence was 3.5-fold higher than the level in unsupplemented cells. Notably, the level of fluorescence with 14 h of palmitate supplementation was similar to that detected when CHO cells were supplemented with 5 mmH2O2 for 1 h prior to C-2938 loading. Including 5 mm of the antioxidant PDTC (28Gottlieb E. Vander Heiden M.G. Thompson C.B. Mol. Cell. Biol. 2000; 20: 5680-5689Crossref PubMed Scopus (309) Google Scholar, 33Joshi B. Li L. Taffe B.G. Zhu Z. Wahl S. Tian H. Ben-Josef E. Taylor J.D. Porter A.T. Tang D.G. Cancer Res. 1999; 59: 4343-4355PubMed Google Scholar) with palmitate supplementation decreased the fluorescence to 1.3-fold the level detected in unsupplemented cells. Similarly, 20 mmDBDA, a membrane-permeable nonenzymatic superoxide scavenger (34Liao D.F. Jin Z.G. Baas A.S. Daum G. Gygi S.P. Aebersold R. Berk B.C. J. Biol. Chem. 2000; 275: 189-196Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar, 35Moreno-Manzano V. Ishikawa Y. Lucio-Cazana J. Kitamura M. J. Biol. Chem. 2000; 275: 12684-12691Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar), reduced fluorescence to 1.6-fold that detected in unsupplemented cells (Fig. 5 A). Importantly, the oxidative stress observed in palmitate supplemented CHO cells was not dependent on de novo ceramide synthesis. LY-B cells, similar to CHO cells, showed increased C-2938 fluorescence following palmitate supplementation (Fig.5 A). Additionally, supplementation with 500 μmoleate did not induce C-2938 fluorescence, indicating reactive intermediates were not produced (Fig. 5 A). This is consistent with the inability of oleate to induce cell death. Taken together, these findings suggest that palmitate supplementation leads to accumulation of reactive oxygen intermediates. To determine whether the generation of reactive intermediates is essential for palmitate-induced apoptosis, we measured the ability of the antioxidants PDTC and DBDA to inhibit caspase activation and DNA laddering. PDTC (5 mm) effectively blocked caspase 3 activity after 24 h of palmitate supplementation (Fig. 5 B). Similarly, 20 mm DBDA significantly reduced caspase 3 activity from 10.0- to 2.2-fold over untreated cells. The failure of DBDA to completely inhibit caspase 3 activation may be due to the low level of reactive intermediates that remained with 20 mm DBDA as shown in Fig. 5 A, or the observation that 20 mm DBDA alone caused an increase in caspase 3 activity (2.1 ± 0.2-fold increase over untreated cells,n = 7). In addition to the effect on caspase activation, PDTC and DBDA both effectively blocked DNA laddering (Fig.5 C). Thus, antioxidants prevent both the generation of reactive intermediates following palmitate supplementation and the induction of apoptosis. Our studies implicate a novel mechanism through which palmitate supplementation leads to apoptosis. Specifically, we propose that CHO cells do not require de novo ceramide synthesis for palmitate-induced cell death. This conclusion is supported by the observations that CHO cells treated with biochemical inhibitors ofde novo ceramide synthesis and CHO cells with a mutation in serine palmitoyltransferase continue to undergo apoptosis in response to palmitate supplementation. In contrast to the hypothesis thatde novo ceramide synthesis is required, our data suggest that palmitate-induced apoptosis occurs through oxidative stress. We observed an increase in reactive intermediates with palmitate supplementation that is independent of de novo ceramide synthesis. Antioxidants inhibited both C-2938 fluorescence and palmitate-induced caspase activation and DNA laddering. Thus, our data support an integral role for the generation of reactive intermediates in palmitate-induced lipotoxicity. Ceramide is generated by de novo biosynthesis following palmitate supplementation and may serve to amplify the apoptotic response in CHO cells. We observed a decrease in the magnitude of caspase 3 activity when de novo ceramide synthesis was inhibited. This reduction was evident when ceramide synthesis was blocked by either the mutation in serine palmitoyltransferase or with the biochemical inhibitors. The reduction in caspase 3 activity occurred at every time point measured (from 17–28 h), indicating that we did not simply observe a delay in caspase 3 activation. Despite this decrease in caspase 3 activity, we continued to observe DNA laddering and cell death, indicating that the remaining caspase 3 activity was sufficient for the induction of apoptosis. These findings are most consistent with a model in which ceramide serves to amplify but not induce the apoptotic response to palmitate supplementation. Recent data showing that cytochrome C release and caspase 3 activation precedes ceramide accumulation in palmitate-treated cardiac myocytes also support a nonessential role for ceramide synthesis (36Sparagna G.C. Hickson-Bick D.L. Buja L.M. McMillin J.B. Am. J. Physiol. 2000; 279: H2124-H2132Crossref PubMed Google Scholar). The mechanism of cellular lipotoxicity likely depends on cell type-specific processes for channeling fatty acids to particular metabolic fates. Our observation that ceramide synthesis is not required for palmitate-induced apoptosis is in contrast to published data showing palmitate-induced apoptosis in hematopoietic precursor cell lines (LyD9 and WEHI-231 cells) and in pancreatic β-cells of the ZDF rat is blocked by inhibitors of de novo ceramide synthesis (1Shimabukuro M. Zhou Y.-T. Levi M. Unger R.H. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 2498-2502Crossref PubMed Scopus (1012) Google Scholar, 10Paumen M.B. Ishida Y. Muramatsu M. Yamamoto M. Honjo T. J. Biol. Chem. 1997; 272: 3324-3329Abstract Full Text Full Text PDF PubMed Scopus (312) Google Scholar, 22Shimabukuro M. Higa M. Zhou Y.-T. Wang M.-Y. Newgard C.B. Unger R.H. J. Biol. Chem. 1998; 273: 32487-32490Abstract Full Text Full Text PDF PubMed Scopus (341) Google Scholar). Our results suggest that fatty acids may be targeted to different metabolic fates in CHO cells as compared with LyD9 and WEHI-231 cells and cells from the ZDF rat. Consistent with this notion, the ZDF rat harbors a mutation in the leptin receptor that is associated with alterations in handling of intracellular fatty acids as shown by an increased capacity to accumulate triglycerides in nonadipose tissues (37Unger R.H. Zhou Y.T. Orci L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 2327-2332Crossref PubMed Scopus (374) Google Scholar). Future studies into the mechanisms that control channeling of fatty acids to specific metabolic fates will provide insight into these cell-specific differences. Our studies with an oxidant-sensitive probe and agents that scavenge oxidants suggest that generation of ROS is essential for the induction of apoptosis in response to palmitate supplementation. Palmitate-induced caspase 3 activity and DNA laddering are inhibited by both PDTC and DBDA. Depending on concentration and cell background, PDTC may function to increase cellular glutathione or directly inhibit the NF-κB pathway, both of which actions may protect against oxidative stress (38Moellering D. McAndrew J. Jo H. Darley-Usmar V.M. Free Radic. Biol. Med. 1999; 26: 1138-1145Crossref PubMed Scopus (58) Google Scholar). DBDA has been used as a scavenger of superoxide but has no known effects on NF-κB signaling (34Liao D.F. Jin Z.G. Baas A.S. Daum G. Gygi S.P. Aebersold R. Berk B.C. J. Biol. Chem. 2000; 275: 189-196Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar, 35Moreno-Manzano V. Ishikawa Y. Lucio-Cazana J. Kitamura M. J. Biol. Chem. 2000; 275: 12684-12691Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar). Therefore, we believe the observed inhibition of palmitate-induced apoptosis is due to the ability of both PDTC and DBDA to decrease ROS. Although production of ROS can be a coincident finding in cell death, evidence from this and other studies suggest that reactive intermediates play a primary role in the activation stage of apoptosis (reviewed in Refs.39Buttke T. Sandstrom P. Free Rad. Res. 1995; 22: 389-397Crossref PubMed Scopus (139) Google Scholar, 40Jacobson M. Trends Biochem. Sci. 1996; 21: 83-86Abstract Full Text PDF PubMed Scopus (729) Google Scholar, 41Mignotte B. Vayssiere J. Eur. J. Biochem. 1998; 252: 1-15Crossref PubMed Scopus (708) Google Scholar). ROS can initiate signaling pathways that affect protein phosphorylation or activate nuclear transcription factors such as NF-κB (30Nishikawa T. Edelstein D. Du X.L. Yamagishi S. Matsumura T. Kaneda Y. Yorek M.A. Beebe D. Oates P.J. Hammes H.P. Giardino I. Brownlee M. Nature. 2000; 404: 787-790Crossref PubMed Scopus (3658) Google Scholar, 42Schreck R. Albermann K. Baeuerle P.A. Free Radic. Res. Commun. 1992; 17: 221-237Crossref PubMed Scopus (1300) Google Scholar, 43Staal F.J. Anderson M.T. Staal G.E. Herzenberg L.A. Gitler C. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 3619-3622Crossref PubMed Scopus (210) Google Scholar). In our studies, two observations support a role for ROS in the induction rather than execution of palmitate-induced apoptosis. First, we observed a 2.2-fold increase in reactive intermediates following 5 h of palmitate supplementation. This early time point corresponds to the time at which we began to see morphological changes caused by palmitate supplementation but before caspase activation and DNA laddering. Secondly, antioxidants inhibited both caspase activation and DNA laddering, suggesting that ROS are acting upstream of these events. Our data is most consistent with a primary role for ROS in the induction of apoptosis following palmitate supplementation. Studies are underway in our laboratory to further characterize the mechanism whereby palmitate supplementation leads to the generation of reactive intermediates. Our observation of palmitate-induced C-2938 fluorescence in LY-B cells suggests that reactive intermediates can be generated independent of de novo ceramide synthesis. ROS may be generated from lipid peroxidation, but this mechanism would require fatty acid desaturation and is unlikely to occur directly from supplementation of a saturated fatty acid. Alternatively, excess palmitate may lead to increased cycling through mitochondrial β-oxidation pathways generating ROS in excess of endogenous cellular antioxidants. However, it is unlikely that this effect would be specific for the saturated fatty acid palmitate and not occur with the unsaturated fatty acid oleate. Finally, evidence is emerging that palmitate can induce the formation of reactive oxygen species through protein kinase C-dependent activation of NAD(P)H oxidase (27Inoguchi T. Li P. Umeda F., Yu, H.Y. Kakimoto M. Imamura M. Aoki T. Etoh T. Hashimoto T. Naruse M. Sano H. Utsumi H. Nawata H. Diabetes. 2000; 49: 1939-1945Crossref PubMed Scopus (1272) Google Scholar). Furthermore, the generation of ROS may cause further cell damage through the production of reactive nitrogen species by the reaction of ROS with nitric oxide, a compound that has been shown to increase with palmitate supplementation (44Shimabukuro M. Ohneda M. Lee Y. Unger R.H. J. Clin. Invest. 1997; 100: 290-295Crossref PubMed Scopus (254) Google Scholar). We are currently exploring whether the toxicity associated with palmitate supplementation is affected by independent perturbation of fatty acid metabolism, NAD(P)H oxidase or NO synthase. In conclusion, our studies indicate that the saturated free fatty acid palmitate induces the formation of reactive intermediates and leads to programmed cell death. Fatty acid-induced apoptosis may contribute to cardiac myocyte death in diabetic cardiomyopathy, cardiomyopathy associated with inherited disorders of mitochondrial fatty acid oxidation, and pancreatic β cell loss in diabetes. Additionally, palmitate-mediated production of ROS may cause significant cellular dysfunction that contributes to the pathogenesis of these diseases prior to cell death. Our findings suggest novel approaches to pharmacologic and genetic rescue strategies in animal models of human heart disease and diabetes. We thank K. Hanada for the gift of the LY-B cells and members of the Schaffer and Ory laboratories for helpful discussions. We are grateful to D. Kelly, M. Linder, and J. Heinecke for critical evaluation of this manuscript.