If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Molecular Nutrition Unit, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Québec H2L 4M1, CanadaInstitut du Cancer de Montréal, Université de Montréal, Montréal, Québec H2L 4M1, Canada
Molecular Nutrition Unit, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Québec H2L 4M1, CanadaInstitut du Cancer de Montréal, Université de Montréal, Montréal, Québec H2L 4M1, Canada
Molecular Nutrition Unit, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Québec H2L 4M1, CanadaInstitut du Cancer de Montréal, Université de Montréal, Montréal, Québec H2L 4M1, Canada
Molecular Nutrition Unit, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Québec H2L 4M1, CanadaInstitut du Cancer de Montréal, Université de Montréal, Montréal, Québec H2L 4M1, Canada
Molecular Nutrition Unit, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Québec H2L 4M1, CanadaInstitut du Cancer de Montréal, Université de Montréal, Montréal, Québec H2L 4M1, CanadaDepartment of Nutrition and Biochemistry, Université de Montréal, Montréal, Québec H2L 4M1, Canada
To whom correspondence should be addressed: Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Hôpital Notre-Dame, Y-5603, 1560 Sherbrooke Est, Montréal, Quebec H2L 4M1, Canada. Tel.: 514-890-8000 (ext. 26827); Fax: 514-412-7590
Molecular Nutrition Unit, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, Québec H2L 4M1, CanadaInstitut du Cancer de Montréal, Université de Montréal, Montréal, Québec H2L 4M1, CanadaDépartement de Médecine, Université de Montréal, Montréal, Québec H2L 4M1, Canada
* This work was supported in part by Studentships from the Fonds Québécois de la Recherche sur la Nature et les Technologies (to S. H.), from the Association du Diabète du Québec (to W. E.-A.) from the Fondation Bourgie/Institut du Cancer de Montréal (to E. P.), and by research grants from the Montreal Breast Cancer Foundation, the Canadian Cancer Etiology Research Network, and the Fondation René Malo/Institut du Cancer de Montréal (to M. P., E. J., and Y. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ∥ Recipient of a Distinguished Scientist Award from the Canadian Institutes of Health Research.
Little is known about the biochemical basis of the action of free fatty acids (FFA) on breast cancer cell proliferation and apoptosis. Here we report that unsaturated FFAs stimulated the proliferation of human MDA-MB-231 breast cancer cells, whereas saturated FFAs inhibited it and caused apoptosis. Saturated FFA palmitate decreased the mitochondrial membrane potential and caused cytochrome c release. Palmitate-induced apoptosis was enhanced by the fat oxidation inhibitor etomoxir, whereas it was reduced by fatty-acyl CoA synthase inhibitor triacsin C. The non-metabolizable analog 2-bromopalmitate was not cytotoxic. This indicates that palmitate must be metabolized to exert its toxic effect but that its action does not involve fat oxidation. Pharmacological studies showed that the action of palmitate is not mediated via ceramides, reactive oxygen species, or changes in phosphatidylinositol 3-kinase activity. Palmitate caused early enhancement of cardiolipin turnover and decreased the levels of this mitochondrial phospholipid, which is necessary for cytochrome c retention. Cosupplementation of oleate, or increasing β-oxidation with the AMP-activated protein kinase activator, 5-aminoimidazole-4-carboxamide-1-β-d-ribonucleoside, both restored cardiolipin levels and blocked palmitate-induced apoptosis. Oleate was preferentially metabolized to triglycerides, and oleate cosupplementation channeled palmitate esterification processes to triglycerides. Overexpression of Bcl-2 family members blocked palmitate-induced apoptosis. The results provide evidence that a decrease in cardiolipin levels and altered mitochondrial function are involved in palmitate-induced breast cancer cell death. They also suggest that the antiapoptotic action of oleate on palmitate-induced cell death involves both restoration of cardiolipin levels and redirection of palmitate esterification processes to triglycerides.
Epidemiological studies and experiments using animals have suggested that dietary fatty acids contribute to the risk of breast cancer (
). However, there is a debate as to whether this is attributable to particular types of fatty acids and their ratio in the diet. Free fatty acids (FFAs)
play an important role in numerous biological functions. They serve as a source of energy and as precursors of many signaling and structural molecules (
). Little information exists on the biochemical pathways by which different types of fatty acids influence breast cancer cell growth and death. The observation in breast cancer cells that overexpression of enzymes involved in fatty acid synthesis (fatty acid synthase, acetyl-CoA carboxylase) correlates with poor prognosis strengthens the hypothesis that fatty acids are involved in the regulation of tumor cell growth (
). Also, consistent with this view is the recent observation documenting an interaction of the tumor suppressor protein BRCA1 with acetyl CoA-carboxylase, the rate-limiting enzyme in fatty acid biosynthesis (
). Moreover, we reported that various types of FFAs are not equivalent with respect to their actions on breast cancer cell proliferation and apoptosis: the monounsaturated FFA oleate (C18:1) stimulated the proliferation of breast cancer cells, whereas the saturated palmitate (C16:0) induced apoptosis (
). A possible explanation for the differential effect of these two FFAs came from the surprising observation that oleate and palmitate increased and decreased, respectively, phosphatidylinositol 3-kinase (PI3K) activity. The antagonistic action of saturated versus unsaturated FFAs on breast cancer cell growth control may explain the contradiction reported in the epidemiological literature with respect to fatty acids and cancer by pointing out that different FFAs are not equivalent as far as cell growth and tumor promotion are concerned (
), but, for most of these cell types, the mechanism by which palmitate induces cell death remains elusive. It has been proposed that excess palmitate could induce cell death through increased intracellular concentration of ceramide (
), a metabolite exclusively produced from saturated FFAs. However, this hypothesis has been challenged by work suggesting that apoptosis induced by palmitate could occur through the generation of reactive oxygen species (ROS) and not through ceramide synthesis (
). Alternatively, palmitate-induced apoptosis in rat neonatal cardiomyocytes has been attributed to an important diminution of cardiolipin (CL) synthesis, correlating well with the extent of cytochrome c release into the cytosol (
). CL is an anionic phospholipid containing four unsaturated fatty acids responsible for insertion and retention of cytochrome c in the mitochondrial membrane (
). Whether palmitate influences the CL level in breast cancer cells remains unknown.
The central role of mitochondria, the main intracellular source of energy, in palmitate metabolism raises the possibility that palmitate-induced changes in mitochondria could impact the process of breast cancer cell death. Mitochondria play a crucial role in several processes linked to apoptosis (
). Apoptotic stimuli produce the release of mitochondrial proapoptotic proteins into the cytosol, like cytochrome c, Smac/Diablo, and apoptosis-inducing factor (
). The mechanism that mediates the release of cytochrome c from the mitochondria is unclear. In one view, the release occurs following mitochondrial permeability transition as a consequence of the opening of a large pore called the permeability transition pore (
). Alternatively, cytochrome c release has been attributed to the formation of nonspecific holes in the mitochondria that does not involve mitochondrial permeability transition (
). By whatever mechanism cytochrome c is released, members of the Bcl-2 family of proteins control this key step in the apoptotic process. Some members inhibit cytochrome c release and promote cell survival (Bcl-2, Bcl-xL, and Bcl-w), whereas others, by mediating cytochrome c release, promote apoptosis (Bax, Bak, and Bid) (
In the present study, the mechanism by which the saturated fatty acid, palmitate, induces apoptosis in the breast cancer cell line MDA-MB-231 was investigated. The results suggest that mitochondria and CL play a critical role in palmitate-induced apoptosis in breast cancer cells. In addition, they allow us to propose a model explaining the lipotoxic action of palmitate and other saturated fatty acids.
EXPERIMENTAL PROCEDURES
Materials—The caspase-3 substrate Ac-DEVD-AFC, triacsin C, fumonisin B1, and LY294002 were from BIOMOL (Plymouth Meeting, PA). Fatty acids sodium salts were purchased from Nu-Check Prep (Elysian, MN). Fatty acid-free bovine serum albumin (BSA) (fraction V), etomoxir, and AICAR were obtained from Sigma (St. Louis, MO). 3,3′-dihexyloxacarbocyanine iodide (DiOC6)(3) was from Molecular Probes Inc. (Eugene, OR). [3H]Thymidine (specific activity, 71 Ci/mmol) and [14C]palmitate (specific activity, 55 mCi/mmol) were obtained from PerkinElmer Life Sciences (Boston, MA), and [14C]oleate (specific activity, 60 mCi/mmol) and [32P]H3PO4 were from Amersham Biosciences (Baie d'Urfé, Québec, Canada).
Cell Lines and Culture Conditions—The human breast cancer cell line MDA-MB-231 was obtained from the American Type Culture Collection. Cells were cultured at 37 °C with 5% CO2 in phenol red-free minimal essential medium (MEM) containing non-essential amino acids, 2 mm glutamine, and 5% heat-inactivated fetal bovine serum (standard medium). Albumin-bound fatty acids were prepared by stirring fatty acid sodium salts (≥99% purity) at 37 °C with 5% BSA essentially fatty acid-free as described before (
). After being adjusted to pH 7.4, the solution was filtered through a 0.22-μm filter, and the fatty acid concentration was measured using a NEFA C kit (Wako Chemicals, GmbH). When BSA-bound fatty acids were added to serum-free culture medium, the final concentration of BSA was always adjusted to 0.5%.
[3H]Thymidine Incorporation—For cell growth experiments, a protocol similar to the one already detailed was used (
). Cells were seeded at 5000 cells/well in 96-well plates and incubated for 24 h in standard medium. After a 24-h starvation period in medium without serum but with 0.5% BSA, cells were incubated without or with BSA-bound fatty acids for 24 h. DNA synthesis was then assayed with a pulse of [3H]thymidine (1 μCi/well) during the last 4 h of incubation. Cells were harvested with a PHD cell harvester from Cambridge Technology (Watertown, MA), and the radioactivity retained on the dried glass fiber filters was measured by liquid scintillation.
Caspase-3 Activity Assay—Cells were seeded in 60-mm Petri dishes at 5 × 105 cells/dish and incubated for 24 h in standard medium. After 12 h of serum starvation in medium containing 0.5% BSA, cells were incubated in serum-free medium without or with BSA-bound fatty acids. Caspase activity was determined using the ApoAlert caspase assay kit (Clontech, Palo Alto, CA). Briefly, cells were washed twice with ice-cold phosphate-buffered saline (PBS) and harvested in lysis buffer. After determination of protein concentration using the BCA colorimetric assay (Pierce Biotechnology Inc., Rockford, IL) with BSA as a standard, 50 μg of protein was incubated with the fluorogenic peptide substrate Ac-DEVD-7-amino-4-trifluoromethyl coumarin (AFC). The initial rate of release of free AFC was measured using a FLUOstar OPTIMA microplate reader (BMG Labtechnologies, Inc., Durham, NC) at an excitation wavelength of 380 nm and an emission wavelength of 505 nm.
Mitochondrial Membrane Potential—Cells were cultured and incubated as described above. After treatment, cells were trypsinized, washed two times with PBS and resuspended in PBS containing 0.1% BSA (PBS-BSA) and 0.2 μm DiOC6(3) (
). After 20 min of incubation at 37 °C, cells were washed, resuspended in PBS-BSA, and incubated for 30 min at 37 °C. Cells were collected and resuspended in PBS-BSA containing 1 μm propidium iodide. DiOC6(3) fluorescence was measured by flow cytometry using an Epics XL flow cytometer from Coulter (Miami, FL) at an excitation wavelength of 485 nm and an emission wavelength of 500 nm. The percentage of cells with low mitochondrial membrane potential (Δψm) represents the percentage of gated cells displaying decreased DiOC6(3) fluorescence.
Cytochrome c Release Analysis—Cells were seeded in 100-mm Petri dishes at 1 × 106 cells/dish, cultured, and incubated as described above. At different times, cells were collected and resuspended in extraction buffer containing 220 mm mannitol, 68 mm sucrose, 50 mm PIPES-KOH (pH 7.4), 50 mm KCl, 5 mm EGTA, and protease inhibitors. After 10 min of incubation on ice, cells were homogenized with a syringe needle (25-gauge), and centrifuged at 14,000 × g for 10 min. Supernatants were removed and centrifuged at 100,000 g for 1 h. The supernatant containing the cytosolic fraction was removed and stored at –80 °C until analysis by immunoblot as described previously (
Measurement of Ceramide Levels—Cells were seeded in 60-mm Petri dishes at 5 × 105 cells/dish and incubated as described above. Lipids were collected according to the method of Bligh and Dyer (
Incorporation of Palmitate and Oleate into Various Complex Lipids— Cells seeded in 60-mm Petri dishes at 5 × 105 cells/dish were incubated for 24 h in standard medium. After 12 h of serum starvation in medium containing 0.5% BSA, they were incubated for 1 h in serum-free medium supplemented with BSA-bound palmitate and/or oleate (0.1 mm) in the presence of 0.5 μCi/ml [14C]palmitate or [14C]oleate. Then, total lipids were extracted and separated by thin layer chromatography as described (
). Spots corresponding to triacylglycerol (TG), diacylglycerol (DAG), cholesterol ester, and phospholipids were individually scraped and counted by liquid scintillation. Incorporation in neutral lipids and phospholipids was normalized for lipid recovery.
Measurement of Phospholipid Levels—Phospholipid measurements were performed essentially as described previously (
). Cells were seeded in 60-mm Petri dishes at 5 × 105 cells/dish and incubated for 24 h in standard medium. Cells were labeled for 24 h with [32P]Pi (20 μCi/ml), after which cells were incubated in serum-free medium without or with BSA-bound fatty acids in the presence of [32P]Pi (20 μCi/ml). For cardiolipin (CL) turnover experiments, cells were labeled as described above for 24 h, washed twice in PBS, and incubated in serum-free medium lacking [32P]Pi without or with BSA-bound fatty acids. After treatment, cells were washed, and lipids were extracted using a solution of CH3OH/CHCl3/0.1 m HCl (10:5:4). Lipids were then isolated by acid organic extraction (half volumes each of CHCl3 and 0.1 m HCl, 0.5 m NaCl), dried, and resuspended in an appropriate volume of CHCl3. Phospholipids were separated by thin layer chromatography using silica gel plates (
), and individual species were identified by comigration of standards (Sigma-Aldrich Corp., St. Louis, MO). The spots corresponding to CL and to phosphatidic acid (PA) plus phosphatidylglycerol (PG) were individually scraped, counted by liquid scintillation, and normalized to the total amount of label present in each sample.
Adenovirus Infection—The adenovirus recombinants AdTREx-FLAG-Bcl-xL, AdTREx-FLAG-Bcl-w, and AdTREx-GFP expressing, respectively, the antiapoptotic mouse Bcl-xL and human Bcl-w proteins and the green fluorescent protein (GFP) were provided by Gordon Shore. Cells were seeded in 6-well plates at 1.5 × 105 cells/well 24 h prior infection with the adenovirus recombinants at a multiplicity of infection of 5 plaque-forming units per cell. After 12 h of adsorption in MEM without serum in the presence of 0.5% BSA, cells were incubated in serum-free medium without or with BSA-bound palmitate for 24 h prior harvesting.
RESULTS
Opposite Effects of Unsaturated and Saturated Fatty Acids on the Proliferation and Apoptosis of MDA-MB-231 Cells—We showed previously that the monounsaturated fatty acid oleate stimulated the proliferation of three different breast cancer cell lines (
). In contrast, the saturated fatty acid palmitate induced apoptosis in these cells. Here, the effects of different unsaturated and saturated FFAs on the proliferation of serum-starved MDA-MB-231 cells were studied. All the tested unsaturated FFAs stimulated [3H]thymidine incorporation (Fig. 1A). The extent of stimulation ranged from 250% for oleate (C18:1) and linoleate (C18:2) to 170% for arachidonate (C20:4) and docohexanoate (C22:6). In contrast, the saturated FFAs myristate (C14:0), palmitate (C16:0), and stearate (C18:0) decreased [3H]thymidine incorporation by 50–95% when compared with control. Interestingly, 2-bromopalmitate, a non-metabolizable analog of palmitate, did not affect [3H]thymidine incorporation suggesting that palmitate has to be metabolized to exert its effect. The results indicate that, whereas unsaturated FFAs stimulate DNA replication in breast cancer cells, saturated FFAs inhibit it.
Fig. 1Effects of various fatty acids on the proliferation and apoptosis of MDA-MB-231 breast cancer cells.A, after 24 h of serum starvation in MEM in the presence of 0.5% BSA, cells were incubated for 24 h in serum-free medium supplemented with BSA alone (Cont) or 0.1 mm fatty acids bound to BSA (0.5%); Ole, oleate; Lin, linoleate; Ara, arachidonate; DHA, docohexanoate; Myr, myristate; Pal, palmitate; Stear, stearate; Br-Pal, 2-bromopalmitate. During the last 4 h of incubation, cells were labeled with [3H]thymidine. B, after 12 h of serum starvation in MEM in the presence of 0.5% BSA, cells were incubated for 24 h in serum-free medium supplemented with 0.1 mm fatty acids bound to BSA (0.5%) or BSA alone (Cont). Caspase-3 activity in lysates from cells incubated with fatty acids or BSA alone was measured as described under “Experimental Procedures.” Values represent means ± S.E. of two independent experiments performed in triplicate. All values are significant (p < 0.01) versus their respective control, except for bromopalmitate (not significant).
The saturated fatty acid palmitate has been shown to be proapoptotic in MDA-MB-231 cells, because it induced the proteolysis of poly-ADP ribose polymerase and the cleavage of DNA into nucleosomal fragments, two hallmark features of apoptosis (
). To determine if the cell proliferation inhibitory effect of the other saturated FFAs was related to apoptosis, we examined the activation of caspase-3 in serum-starved cells. Saturated FFAs increased caspase-3 activity by 2- to 7-fold, and this increase was correlated with their chain length (Fig. 1B). In contrast, the unsaturated FFA oleate decreased caspase-3 activity by 80%. Moreover, oleate protected cells against apoptosis triggered by saturated FFA, because the activity of caspase-3 was blocked when oleate was mixed in equimolar ratio with saturated FFAs. The results establish that saturated FFAs induce apoptosis in MDA-MB-231 cells, an effect that can be counteracted by oleate.
Mitochondria Play a Central Role in Palmitate-induced Apoptosis—To examine the mechanism by which saturated FFAs induced apoptosis in MDA-MB-231 cells, we chose palmitate because it is the most abundant saturated FFA in the plasma. An examination of the time course of caspase-3 activation induced by palmitate revealed that the activation first seen at 6 h became significant at 8 h and thereafter continued to increase (Fig. 2A).
Fig. 2Time course of the effects of palmitate on various parameters of the apoptotic process in MDA-MB-231 cells. After 12 h of serum starvation in MEM in the presence of 0.5% BSA, MDA-MB-231 cells were incubated in serum-free medium supplemented with 0.1 mm palmitate bound to BSA for the indicated times. A, caspase-3 activity in lysates from cells incubated with palmitate bound to BSA (0.5%) (Pal) or BSA alone (Cont) was assessed as described under “Experimental Procedures.” The results represent the means ± S.E. of two independent experiments performed in triplicate. *, p < 0.01 versus control. B, Δψm was quantified as described under “Experimental Procedures.” “Δψmlow” represents the percentage of cells with decreased mitochondrial membrane potential in the presence of palmitate (•) or oleate (0.1 mm) alone (▪). Values represent means ± S.E. of three separate experiments. *, p < 0.01 versus control. C, release of cytochrome c in the cytosol was studied using immunoblot analysis as described under “Experimental Procedures.” The figure shows a representative experiment that has been repeated two times.
). The effect of palmitate on Δψm was analyzed by flow cytometry in intact cells with the strong cationic dye DiOC6(3). Damaged mitochondria lose membrane integrity and, as a consequence, are no longer able to maintain their transmembrane potential, resulting in a decreased binding of DiOC6(3). Palmitate induced within6ha significant increase in the percentage of cells with low Δψm, whereas no change occurred after 10 h of treatment with the unsaturated FFA oleate (Fig. 2B). Palmitate-induced depolarization of the membrane potential preceded the activation of caspase-3, suggesting that alteration of the mitochondria could be among the first steps in the apoptotic process induced by the fatty acid.
The time dependence of cytochrome c release in the cytoplasm was examined by immunoblotting analysis. Basal amounts of cytochrome c were detected in cytosolic fraction until 4 h following palmitate treatment (Fig. 2C). A more than 3-fold increase in cytochrome c release was seen 6 h posttreatment, and, thereafter, cytochrome c levels remained elevated in the cytoplasm. Moreover, as for the depolarization of the mitochondrial membrane potential, palmitate-induced cytochrome c release preceded caspase-3 activation, suggesting that the mitochondrion is involved in the mechanism underlying palmitate-induced cell death.
Alterations of Lipid Partitioning Affects Palmitate-induced Apoptosis—To provide further evidence that palmitate must be metabolized to exert its proapoptotic effect, we used the acyl-CoA synthetase specific inhibitor, triacsin C. Acyl-CoA synthetase activates fatty acids to fatty acyl-CoA, which are then channeled to different metabolic pathways (
). To minimize the long term cytotoxic effect of triacsin C, caspase-3 activity was measured after a 12-h treatment. Under these conditions, triacsin C, which slightly increased apoptosis in control cells, reduced by 50% the proapoptotic effect of palmitate (Fig. 3A). Taken together, the results obtained with 2-bromopalmitate and triacsin C demonstrate that palmitate must be metabolized at least to the CoA derivative step (fatty acyl-CoA) to promote apoptosis in MDA-MB-231 cells.
Fig. 3Alterations of lipid partitioning affect palmitate-induced apoptosis.A, after 12 h of serum starvation in the presence of 0.5% BSA in MEM, cells were incubated for 12 h in serum-free medium supplemented with 0.1 mm palmitate bound to BSA (0.5%) (Pal) or BSA alone (Cont) in the absence or presence of 5 μm triacsin C. Apoptosis was evaluated by measuring caspase-3 activity as described under “Experimental Procedures.” Values represent means ± S.E. of two independent experiments performed in triplicate. *, p < 0.01 versus control; **, p < 0.01 versus palmitate. B, after 12 h of serum starvation in MEM in the presence of 0.5% BSA, cells were incubated for 24 h in serum-free medium supplemented with 0.1 mm palmitate bound to BSA (0.5%) (Pal) or BSA alone (Cont) in the absence or presence of 0.5 mm AICAR or 0.2 mm etomoxir. Values of caspase-3 activity represent the means ± S.E. of two (AICAR) or three (etomoxir) independent experiments performed in triplicate. *, p < 0.01 versus palmitate alone.
). Hence, using pharmacological tools to alter fatty acid oxidation, we investigated the importance of mitochondrial β-oxidation for the proapoptotic action of palmitate in MDA-MB-231 cells. Increasing mitochondrial fatty acid oxidation with 5-aminoimidazole-4-carboxamide-1-β-d-ribonucleoside (AICAR), an AMP-activated protein kinase (AMPK) activator (
) carnitine palmitoyltransferase I, the rate-limiting enzyme of fatty acids β-oxidation, increased by 2-fold the proapoptotic effect of palmitate. Control experiments revealed that AICAR produced a 2-fold increase in the rate of fatty acid oxidation, whereas etomoxir produced a 70% decrease (data not shown). These results indicate that changes in lipid partitioning through alterations in fatty acid oxidation affect the proapoptotic action of palmitate and that enhanced fat oxidation is not involved in this process.
De Novo Ceramide Synthesis Is Not Required for Palmitate-induced Apoptosis—Ceramides are metabolites exclusively produced from saturated FFAs (
). Treatment of MDA-MB-231 cells with palmitate increased the level of ceramides by 3-fold (Fig. 4A). As expected, the unsaturated fatty acid oleate and the non-metabolizable analog 2-bromopalmitate did not change the cellular ceramide content. The increase in ceramide induced by palmitate was blocked by fumonisin B1, a specific inhibitor of ceramide synthase (
), suggesting that the elevation in ceramide production was attributable to de novo synthesis from palmitate. To determine whether the increase in de novo ceramide synthesis is critical for palmitate-induced apoptosis, caspase-3 activity was measured in palmitate-treated cells in the presence of fumonisin B1. Inhibition of de novo ceramide synthesis by fumonisin B1 did not prevent the increase in caspase-3 activity caused by palmitate (Fig. 4B). Similar results were obtained with myriocin, another chemically unrelated inhibitor of de novo ceramide synthesis acting on serine palmitoyltransferase (
) (data not shown). These results establish that de novo ceramide synthesis is not required for palmitate-induced apoptosis in MDA-MB-231 cells.
Fig. 4Inhibition of de novo ceramide synthesis does not block palmitate-induced apoptosis in MDA-MB-231 cells. After 12 h of serum starvation in MEM in the presence of 0.5% BSA, cells were incubated for 18 h in serum-free medium supplemented with 0.1 mm fatty acids bound to BSA (0.5%) (Pal, palmitate; Ole, oleate; Br-Pal, 2-bromopalmitate) or BSA alone (Cont) in the absence or presence of 50 μm fumonisin B1 (FB1). A, ceramide levels were quantified using the diglyceride kinase assay and normalized to total lipid phosphate. Values are means ± S.E. of three independent experiments performed in duplicate. *, p < 0.01 versus palmitate. B, apoptosis was evaluated by measuring caspase-3 activity as described under “Experimental Procedures.” Values represent means ± S.E. of two independent experiments performed in triplicate.
Oleate Increases Palmitate Channeling toward Triglyceride Synthesis—The dichotomy between the action of palmitate and oleate on MDA-MB-231 cell growth and apoptosis might possibly be due to different fates and actions of the two fatty acids on lipid metabolism. Therefore, the metabolic fate of palmitate and oleate in MDA-MB-231 cells was examined by assessing the rate of incorporation over a 1-h incubation period of [14C]palmitate or [14C]oleate into phospholipids and neutral lipids (DAG and TG). There was no difference between [14C]palmitate and [14C]oleate uptake in these cells (7.3 ± 0.2 versus 7.1 ± 0.2 nmol/h/mg of protein, respectively, means ± S.E., n = 6). The rates of incorporation of palmitate and oleate into phospholipids differed by less than 8% (Fig. 5A). When [14C]palmitate and [14C]oleate incorporations into DAG were compared, a 2.5-fold higher rate was observed with the saturated fatty acid (Fig. 5B). In contrast, [14C]palmitate incorporation into TG was 2.5-fold lower than that of [14C]oleate (Fig. 5C). These results demonstrate that the fate of oleate and palmitate is quantitatively different in MDA-MB-231 cells.
Fig. 5Oleate alters the fate of palmitate into neutral lipids. After 12 h of serum starvation in MEM in the presence of 0.5% BSA, cells were incubated for 1 h in serum-free medium supplemented with 0.1 mm palmitate (Pal) or/and oleate (Ole) bound to BSA (0.5%) in the presence of trace amounts of [14C]palmitate ( 14Pal) or [14C]oleate ( 14Ole). Lipid levels were analyzed as described under “Experimental Procedures” and expressed as percentage of counts (labeled palmitate or oleate) incorporated into total lipids. Values represent means ± S.E. of two independent experiments performed in triplicate. *, p < 0.01 versus palmitate.
To test whether oleate cosupplementation affects the metabolic fate of palmitate, the incorporation of [14C]palmitate was determined in the presence of equimolar concentrations of palmitate and oleate. Oleate cosupplementation increased the incorporation of [14C]palmitate into TG by 2-fold (Fig. 5C), and this was accompanied by a slight decrease in the channeling of palmitate into DAG (Fig. 5B). There was no difference in the incorporation of [14C]oleate in the different lipid pools when both fatty acids were added together (Fig. 5, A–C). The results show that palmitate is channeled toward the TG synthesis pathway in the presence of exogenous oleate. This may contribute to the protective effect of oleate from palmitate-induced apoptosis as it has recently been proposed (
Palmitate-induced Apoptosis Is Associated with Pronounced Changes in Phospholipid Pools—In cardiomyocytes, the levels of phospholipid species, in particular CL, have been shown to differ between oleate- and palmitate-treated cells (
). After steady-state labeling of MDA-MB-231 cells with 32P, the levels of the anionic phospholipids were determined at different times after treatment with palmitate or oleate. The levels of PA plus PG, the precursors of CL, increased sharply in palmitate-treated cells and reached a maximum after only 2 h of treatment (Fig. 6A). In contrast, in oleate-treated cells, the levels of these phospholipids decreased. The increased levels of PA plus PG in palmitate-treated cells suggest a problem converting these saturated species to CL (
). CL levels in palmitate-treated cells decreased significantly by 20% after only 4 h (Fig. 6B). By 6 h, the CL levels were decreased by 50%. In contrast, the CL levels did not change in control and oleate-treated cells over the 12-h experimental period. Interestingly, the initial decrease in CL levels seen in palmitate-treated cells (4 h) shortly preceded the release of cytochrome c into the cytosol (6 h) (Fig. 2C).
Fig. 6Effect of palmitate and oleate on the cellular anionic phospholipid composition and cardiolipin half-life. After 12 h of serum starvation in MEM in the presence of [32P]Pi and 0.5% BSA, cells were incubated in medium supplemented with 0.5% BSA (▪), 0.1 mm palmitate (•), or oleate (○) bound to BSA (0.5%) for the indicated times in the presence (A and B) or absence (C) of [32P]Pi. Phospholipid levels were analyzed as described under “Experimental Procedures” and quantified as a percentage of radiolabel in the region of chromatograms encompassing PA plus PG (A) or CL (B). C, CL turnover was assessed by measuring labeled CL in cells overtime in the absence and presence of exogenous palmitate or oleate (time 0 = 100%). Values represent means ± S.E. of three independent experiments performed in duplicate. *, p < 0.01 versus oleate (A and B) or control (C).
), a decrease in its synthesis could not be sufficient to account for its rapid diminution in palmitate-treated cells. To determine whether palmitate could influence CL degradation, CL turnover was measured following palmitate or oleate treatment. Fig. 6C shows that palmitate increased CL turnover, whereas oleate decreased it. The increase of the CL turnover (Fig. 6C) observed in palmitate-treated cells paralleled the decrease seen in CL levels (Fig. 6B), suggesting that it contributes significantly to the diminution of CL. Because CL was more stable in oleate-treated cells than in controls and the steady-state levels were similar in both conditions, it may be hypothesized that in oleate-treated cells a reduced de novo synthesis compensates the reduced turnover.
Blocking palmitate-induced apoptosis by increasing fatty acid oxidation with AICAR decreased the accumulation of PA plus PG (Fig. 7A) and restored CL levels in palmitate-treated cells (Fig. 7B). Because oleate protected cells against death triggered by palmitate (Fig. 1B), its effect on phospholipid levels was tested. Cotreatment with equimolar concentration of palmitate and oleate blocked accumulation of PA plus PG and completely restored CL levels. Taken together, these data provide suggestive evidence that the diminution in CL levels in palmitate-treated cells is responsible for the release of cytochrome c and the initiation of the caspase cascade leading to apoptotic cell death.
Fig. 7Oleate supplementation or AICAR treatment restore CL levels in palmitate-treated MDA-MB-231 cells. After 12 h of serum starvation in MEM in the presence of [32P]Pi and 0.5% BSA, cells were incubated for 8 h in medium supplemented with 0.1 mm palmitate (Pal) or/and oleate (Ole) bound to BSA (0.5%) or BSA alone (Cont)inthe absence or presence of 0.5 mm AICAR. PA plus PG (A) and CL (B) levels were analyzed as described under “Experimental Procedures” and quantified as a percentage of radiolabel incorporated into total phospholipids. Values represent means ± S.E. of three independent experiments performed in duplicate. *, p < 0.01 versus palmitate alone (A) or control (B).
Antiapoptotic Proteins Block Palmitate-induced Apoptosis— Most of the prosurvival proteins of the Bcl-2 family are targeted primarily to the mitochondrial outer membrane and stabilize it by preventing the release of proapoptotic molecules, such as cytochrome c (
). To determine whether the prosurvival proteins of the Bcl-2 family protect cells against palmitate-induced apoptosis, caspase-3 activity was evaluated following overexpression of genes encoding Bcl-xL or Bcl-w with recombinant adenovirus (Fig. 8). Both proteins impaired caspase-3 activation induced by palmitate, whereas the GFP used as a control was without effect. This provides additional evidence for a role of the mitochondrial arm of apoptosis in the process of palmitate-induced MDA-MB-231 cell death.
Fig. 8Antiapoptotic Bcl-2 family members decrease palmitate-induced apoptosis in MDA-MB-231 cells. After 12 h of serum starvation in the absence (Mock) or presence of recombinant adenoviruses AdTREx-GFP (GFP), AdTREx-FLAG-Bcl-w (Bcl-w), or AdTREx-FLAG-Bcl-xL (Bcl-xL), cells were incubated for 24 h in serum-free medium supplemented with 0.1 mm palmitate bound to BSA (0.5%) (Pal) or BSA alone (Cont). A, apoptosis was evaluated by measuring caspase-3 activity as described under “Experimental Procedures.” Values represent means ± S.E. of two independent experiments performed in triplicate. *, p < 0.01 versus palmitate. B, protein extracts of infected cells were analyzed for recombinant protein expression (FLAG-Bcl-w and FLAG-Bcl-xL) by immunoblotting with an anti-FLAG antibody.
PI3K Is Not Implicated in the Protective Effect of Oleate against Palmitate-induced Apoptosis—In our previous work, the unsaturated FFA oleate increased PI3K activity, whereas the saturated FFA palmitate decreased it (
). Moreover, the combination of both FFAs restored the basal level of PI3K activity. To determine whether PI3K is implicated in the protective effect of oleate against palmitate-induced apoptosis, we used the PI3K inhibitor LY294002 and measured caspase-3 activity. LY294002 did not alter caspase-3 activity in the presence of both FFAs (Fig. 9). Moreover, the PI3K inhibitor curtailed the protective effect of oleate against serum-starvation (control situation, not shown) and, surprisingly, also markedly reduced palmitate-induced apoptosis. These results indicate that PI3K signaling is not implicated in the protective effect of oleate against palmitate-induced cell death and, moreover, that the decrease in PI3K activity is not responsible of palmitate-induced apoptosis.
Fig. 9PI3K is not implicated in the protective effect of oleate against palmitate-induced apoptosis. After 12 h of serum starvation in MEM in the presence of 0.5% BSA, cells were incubated for 24 h in serum-free medium supplemented with 0.1 mm palmitate (Pal) and/or oleate (Ole) bound to BSA (0.5%) or BSA alone (Cont) in the absence or presence of 25 μm of the PI3K inhibitor LY294002. Apoptosis was evaluated by measuring caspase-3 activity. Means ± S.E. of three independent experiments performed in duplicate. *, p < 0.01 versus palmitate alone; **, p < 0.01 versus oleate alone.
Studying the effects of FFAs on MDA-MB-231 breast cancer cells, we found as a general principle that saturated FFAs were proapoptotic and unsaturated FFAs stimulated proliferation. Investigation on the mechanism by which saturated FFAs induce apoptosis suggests that CL synthesis and its turnover play a key role in the opposite effects of the two types of FFA. The saturated FFA palmitate induced a decrease in the mitochondrial membrane potential and caused a release of cytochrome c in the cytosol, both preceding caspase-3 activation. The release of cytochrome c in palmitate-treated cells correlated in time with a prominent decrease in the levels of the mitochondrial phospholipid CL. In contrast, the monounsaturated FFA oleate and the polyunsaturated FFA linoleate (not shown), which stimulated cell proliferation, did not affect CL levels. Moreover, adding oleate or the AMPK-activator AICAR to palmitate-treated cells, which prevented palmitate-induced apoptosis, restored CL levels. Thus, it is suggested that cell survival critically depends of an adequate pool of unsaturated phospholipids that are necessary for CL synthesis.
The saturated FFA palmitate must be metabolized to exert its proapoptotic effect as evidenced from the fact that the non-metabolizable analog 2-bromopalmitate had no effect and the acyl-CoA synthesis inhibitor triacsin C curtailed palmitate action on apoptosis. However, the results do not support the suggestion that one of these metabolites could be ceramide (
). Thus, an increase in ceramide levels in palmitate-treated MDA-MB-231 cells was observed, and inhibitors of the de novo ceramide pathway blocked this rise. However, because these inhibitors did not block apoptosis, we conclude that ceramide accumulation is not involved in palmitate-induced apoptosis. Recent data obtained with CHO cells and cardiomyocytes also support a non-essential role for de novo ceramide synthesis in palmitate-induced apoptosis (
). Excess palmitate may also lead to increased mitochondrial β-oxidation pathway generating ROS in excess of endogenous antioxidant. Increased production of ROS has been implicated in palmitate-induced apoptosis of CHO cells (
). Blocking β-oxidation with etomoxir increased apoptosis in palmitate-treated cells, whereas promoting fatty acid oxidation with AICAR blocked palmitate-induced apoptosis. Moreover, the antioxidants (N-acetylcysteine, ebselen, and aminoguanidine) did not affect death triggered by palmitate (data not shown). Thus, it appears that the β-oxidation pathway and the generation of ROS are not implicated in the cell death process caused by palmitate in MDA-MB-231 cells. ROS were also shown not to be implicated in palmitate induced-apoptosis in cardiomyocytes (
The temporal correlation between the initial decrease in mitochondrial CL levels and the subsequent cytochrome c release in the cytoplasm and loss of the mitochondrial membrane potential provides suggestive evidence for a role of CL in the proapoptotic action of palmitate. Thus, because cytochrome c is bound to the inner mitochondrial membrane by anionic phospholipids, primarily CL (
), the decrease in CL levels could directly facilitate cytochrome c release into the cytosol. Support for this hypothesis has been provided in palmitate-induced apoptosis of rat neonatal cardiomyocytes by the observation that decreased CL synthesis directly correlated with the release of cytochrome c (
). Large increases in the levels of PA and PG were also seen in these cardiomyocytes where these two phospholipids were nearly all dipalmitoyl (C16:0 and C16:0) PA and dipalmitoyl (C16:0 and C16:0) PG. Because CL synthase activity was unaffected in palmitate-treated cardiomyocytes, it was concluded that, although the myocyte has PG available to increase CL levels, new CL is not synthesized because of the low affinity of CL synthase for dipalmitoyl PG. Interestingly, we show that CL turnover is increased by palmitate, whereas it is decreased by oleate. Because the increase of CL turnover observed in palmitate-treated cells paralleled the decrease seen in CL levels, this suggests that it contributes concomitantly with decreased CL synthesis to the diminution of CL levels and apoptosis. In contrast, the stabilization of CL seen in oleate-treated cells could contribute to the antiapoptotic effect of oleate against palmitate-induced cell death. The mechanism by which palmitate increases CL degradation is unknown. It may involve activation of mitochondrial phospholipase A2 induced by the elevation in PalCoA and/or its derivatives (
Increasing fatty acid oxidation with the AMPK activator AICAR restored CL levels and blocked palmitate-induced apoptosis in MDA-MB-231 cells. The increase in fatty acid oxidation induced by AICAR has been shown to result from a stimulation in the activities of both CPT-1 (
), the enzyme which degrades malonyl-CoA, the allosteric inhibitor of carnitine palmitoyltransferase I. In several systems, CPT-1 can protect from palmitate-triggered cell death (
). The protection from palmitate-induced apoptosis afforded by CPT-1 is most likely to be the consequence of palmitoyl-CoA clearance from the cytoplasm, because the CPT enzyme complex effectively transfers long chain fatty acyl-CoA into the mitochondrial matrix, where they serve as fuel for β-oxidation. Thus, CPT-1 may limit synthesis of saturated phospholipids, like the saturated forms of PA and PG. Palmitate caused a very early and pronounced rise in PA and PG in MDA-MB-231 cells. PA is the precursor of many phospholipidic structural and signaling molecules to which has been ascribed many biological activities, including protein kinase activation (
). Hence, the possibility should also be considered that PA or one of its derivatives such as DAG contribute to palmitate-induced apoptosis.
In the present study, we correlated the channeling of palmitate and oleate to distinct metabolic fates with their ability to induce apoptosis. The accumulation of DAG observed in palmitate-treated cells may participate in the induction of apoptosis by acting on DAG-responsive enzymes like protein kinases C (PKC) (
). In contrast, the channeling of palmitate into TG stores induced by oleate cosupplementation may contribute to rescue cells from palmitate-induced apoptosis by redirecting palmitate away from pathways leading to apoptosis. Direct evidence that accumulation of excess saturated fatty acids in cellular TG stores protects from apoptosis was provided by studies with acyl CoA:diacylglycerol transferase 1 null cells showing that the ability to synthesize triglycerides was critical for protection from lipotoxicity (
). Also, because CHO cells overexpressing stearoyl-CoA desaturase (SCD), the enzyme that converts saturated fatty acid to monounsaturated fatty acid, were protected from palmitate-induced apoptosis, it was proposed that the endogenous pools of unsaturated fatty acids produced by SCD may also rescue cells from palmitate-induced apoptosis (
), it is tempting to speculate that the difference in SCD activity may explain the difference in sensitivity to palmitate-induced cell death. The fact that palmitate did not alter the incorporation of oleate into phospholipids renders unlikely that in palmitate-treated cells, the fatty acyl-CoA pool is so deficient in unsaturated species as to preclude assimilation into phospholipids.
Another important difference observed in our previous works between unsaturated and saturated FFAs was their antagonistic effect on PI3K, a key signal transduction protein that inhibit apoptosis by activating the protein kinase Akt/PKB (
). The unsaturated FFA oleate increased PI3K activity, whereas the saturated FFA palmitate decreased it. Others have reported that the activity of Akt/PKB, the downstream target of PI3K, was reduced by palmitate treatment (
). Because downstream targets of the PI3K/Akt signaling pathway that have been identified include the inhibition of proapoptotic proteins Bad and Bax (
), we initially hypothesized that palmitate-induced apoptosis might involve an activation of the proapoptotic Bcl-2 family proteins, resulting from the inactivation of the PI3K/Akt pathway. However, the present results with the PI3K inhibitor LY294002, which in fact curtailed palmitate-induced cell death, rule out a role for a decrease signaling through PI3K in the proapoptotic effect of palmitate. Moreover, the protection from palmitate-induced cell death afforded by oleate was not decreased by LY294002, indicating that PI3K signaling is not implicated in the protective effect of oleate. The observation that overexpression of the antiapoptotic proteins Bcl-xL and Bcl-w decreased palmitate-induced apoptosis is an indication that activation of proapoptotic Bcl-2 proteins is important for the death process caused by palmitate as it has been shown that antiapoptotic Bcl-2 family members protect cells from proapoptotic members by preventing their oligomerization at the mitochondrial membrane (
). To our knowledge, this is the first evidence that antiapoptotic proteins of the Bcl-2 family protect cell death triggered by saturated fatty acids.
Taken together, the results suggest that fatty acid metabolism, unrelated to β-oxidation, ceramide synthesis, ROS production, and changes in PI3K activity, has an important role in the action of palmitate on cell growth and apoptosis. This study allows the proposition of the following model to explain the lipotoxic action of saturated FFAs on breast cancer cells. A deficiency in mitochondrial CL brought about by both inadequate supply of unsaturated fatty acids for CL synthesis and an increase in CL turnover generates a soluble pool of cytochrome c in the mitochondria. The activation of the proapoptotic proteins belonging to the Bcl-2 family that form lipidic pores in mitochondrial membranes by which soluble cytochrome c can escape may also contribute to the cell death process. In contrast, oleate by channeling palmitate to inert triglyceride stores and by permitting sustained CL synthesis, and decreasing its turnover, not only blocks apoptosis but also permits cell proliferation.
Acknowledgments
We thank Richard Marcellus and Gordon Shore for their generous gift of adenovirus recombinants.
83983-0&id=gr1.jpg){kind=link}
83983-0&id=gr2.jpg){kind=link}
83983-0&id=gr3.jpg){kind=link}
83983-0&id=gr4.jpg){kind=link}
83983-0&id=gr5.jpg){kind=link}
83983-0&id=gr6.jpg){kind=link}
83983-0&id=gr7.jpg){kind=link}
83983-0&id=gr8.jpg){kind=link}
83983-0&id=gr9.jpg){kind=link}