Branched Chain Fatty Acids Induce Nitric Oxide-dependent Apoptosis in Vascular Smooth Muscle Cells*

Clinical observations in patients with peroxisomal disorders and studies employing corresponding mouse models have shown that supraphysiological concentrations of dietary branched chain fatty acids (BCFAs) are associated with a high level of toxicity, which is poorly understood at present. Here we show that phytanic and pristanic acid, two BCFAs that are metabolized in peroxisomes, promote apoptosis in cultured vascular smooth muscle cells of human, rat, and porcine origin. Under the conditions used, the apoptosis-promoting effect of BCFAs was neither shared by saturated or unsaturated straight chain fatty acids nor by artificial peroxisome proliferators, which, like phytanic and pristanic acid, have been shown to activate the peroxisome proliferator-activated receptor α (PPARα). We could demonstrate, however, that BCFA induced tumor necrosis factor α (TNFα) activation and secretion, which is an obligatory step required for induction of apoptosis by BCFAs. Furthermore, incubation of VSMCs with BCFA increased inducible nitric-oxide synthase (iNOS) mRNA and protein concentrations markedly within 2 h of treatment. Correspondingly, apoptosis was significantly reduced when the cells were co-treated with the competitive NOS inhibitors monomethyl-l-arginine monoacetate and aminoguanidine. Moreover, co-incubation with TGFβ1, previously shown to destabilize iNOS mRNA, also abolished apoptosis. These results establish a new signaling cascade in which natural BCFA induced NO-dependent apoptosis, which is apparently triggered by autocrine secretion of TNFα in cultured VSMCs.

Phytanic acid, a degradation product of the chlorophyll side chain, is a prominent natural branched chain fatty acid (BCFA) 1 whose most important dietary sources in humans are fat-rich meat and dairy products. Its catabolism proceeds in peroxisomes via ␣-oxidation yielding pristanic acid followed by ␤-oxidation via the non-inducible peroxisomal ␤-oxidation system (reviewed in Ref. 1). Our group and others (2)(3)(4) showed earlier that phytanic and pristanic acid are high affinity ligands and potent activators of the peroxisome proliferatoractivated receptor ␣ (PPAR␣). PPAR␣ is most prominently expressed in tissues with high fatty acid turnover pertaining to its function as regulator of lipid metabolism (5). In addition, this nuclear hormone receptor is present in cells of the vessel wall, including endothelial cells (6), VSMCs (7), and macrophages (8). Incubation of human VSMCs with the potent artificial PPAR␣ ligand Wy 14,643 suppresses the inflammatory response induced by IL-1␤ (7) thus suggesting a role for this transcription factor in gene regulation during the inflammatory response of the vascular wall. In addition, phytanic but not pristanic acid acts as an activating ligand of RXR, a nuclear receptor that is required as a heterodimerizing partner of PPAR␣ and other nuclear hormone receptors for activation of target gene transcription (9,10).
Phytanic acid accumulates in the cerebro-hepato-renal (Zellweger) syndrome, a peroxisome biogenesis disorder associated with absence of virtually all peroxisomal functions, and in several single peroxisomal enzyme deficiencies leading to severe functional impairment of the peroxisomal pathways required for phytanic and pristanic acid oxidation. Among the latter conditions, Refsum disease, which is caused by mutations in the gene encoding phytanoyl-CoA ␣-hydroxylase (PHH), is associated with excessive storage of phytanic acid (11,12). In humans, PHH deficiency is associated with neurological abnormalities, episodes of weight loss, and sudden cardiac death (1). Secondly, MFP-2 deficiency, is associated with accumulation of phytanic acid along with pristanic acid and bile acid precursors as a result of a block in the ␤-oxidation of branched chain fatty acids and bile acid precursors at the level of the 2-enoyl-CoA hydratase/2-hydroxyacyl-CoA dehydrogenase reaction (13). Compared with Refsum disease, MFP-2 deficiency follows a much more severe clinical course, leading to neonatal death due to functional impairment of multiple tissues and cell types. Finally, we have constructed a null mouse model for SCP2/SCPx deficiency lacking the peroxisomal thiolase (SCPx or pThiol2) required for the ketolytic cleavage step in peroxisomal ␤-oxidation of BCFA and bile acid precursors. Although the condition was mild in the absence of BCFA from the diet, dietary uptake of only moderate amounts of BCFA, which led to excessive storage of phytanic acid along with pristanic acid, were lethal for the SCP2/SCPx-null mice, whereas these diets did not affect genetically normal animals (14). These data led to the assumption that accumulation of pristanic acid, either alone or in combination with phytanic acid, may have been responsible for the observed toxicity of BCFA-containing diets in the MFP-2-deficient patients and the SCP2/SCPx-null mice. In the present study, we therefore investigated potentially toxic effects of phytanic and pristanic acid in vitro in VSMCs, which express PPAR␣ and RXR. We could demonstrate that treatment of VSMCs with relatively moderate concentrations of phytanic and/or pristanic acid can induce apoptosis, which, to our surprise, was not related to RXR/PPAR superstimulation but to activation of a so far unknown signaling pathway leading to activation of autocrine secretion of TNF␣ along with a very significant induction of iNOS gene expression by phytanic and pristanic acid.

MATERIALS AND METHODS
Reagents and Chemicals-If not stated otherwise, chemicals and tissue culture media were obtained from Sigma. Aminohydroxyguanidine was purchased from Calbiochem (Darmstadt, Germany), Wy 14.643 was from Biomol (Hamburg, Germany), fetal calf serum from ICN Biomedicals (Eschwege, Germany), and pristanic acid was obtained from H. J. Tenbrink (University of Amsterdam, NL).
Cell Culture-Rat VSMCs, a gift from Dr. M. Tepel (Herne, Germany), were isolated from thoracic aortas of male Wistar rats by the method of Franks et al. (15). Human and porcine VSMCs were a gift from Dr. G. Plenz (Mü nster, Germany) and prepared as described in Ref. 16. VSMCs were cultured in Dulbecco's modified Eagle's medium with 10% fetal calf serum supplemented with antibiotic/antimycotic supplement (Sigma) and passaged with 1ϫ trypsin/EDTA solution (human and porcine VSMCs) or 10ϫ trypsin/EDTA solution (Sigma) in the case of rat VSMCs. Cells at passages 4 -15 were used for the experiments. Human umbilical vein endothelial cells were a gift from Dr. A. Skaletz-Rorowski (Mü nster, Germany) and cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, epidermal growth factor (5 g/liter), fibroblast growth factor (0.5 g/liter), hydrocortisone (500 g/liter), and endothelial growth supplement (Promocell, Heidelberg, Germany). Human monocyte-derived macrophages were a gift from P. Cullen (Mü nster, Germany) and cultured with RPMI 1640 containing 10% of human serum. For all incubation experiments, the indicated concentrations of the ligand were dissolved in Me 2 SO at a final concentration of 1%, and the serum concentration was reduced to 1% as described in Ref. 17. If not stated otherwise, data are expressed as mean Ϯ S.D. obtained from three independent experiments, each performed in triplicate and analyzed with the Student's t test. Differences with p Ͻ 0.05 were considered significant.
Flow Cytometry Detection of Apoptosis-Subconfluent VSMCs were stimulated with agonists for various time intervals, washed twice with ice-cold phosphate-buffered saline and were collected by trypsinization followed by centrifugation. For the determination of annexin V binding, 5 ϫ 10 5 cells were resuspended in 1 ml of a solution containing 140 mmol/liter NaCl, 10 mmol/liter Hepes pH 7.2, and 2.5 mmol/liter CaCl 2 and annexin V-FITC (Bender Med-Systems Diagnostics, Vienna, Austria) according to the supplier's instructions. The cell suspension was incubated for 30 min at room temperature and then subjected to flow cytometry. For the determination of the cell membrane permeability, 5 ϫ 10 5 cells were resuspended in 1 ml of phosphate-buffered saline and incubated for 30 min with YO-PRO (Molecular Probes, Leiden, The Netherlands) at a final concentration of 0.1 mol/liter as described in Ref. 18. All flow cytometric measurements were performed on a Coulter Epics Elite flow cytometer equipped with a 15 milliwatt argon ion laser (Coulter Corporation, Hialeah, FL) (excitation wavelength, 488 nm; fluorescence emission, 525 nm; data rate, ϳ300 cells/s).
Caspase-3 Activity Measurements-Cells were harvested in a buffer consisting of 25 mmol/liter Hepes pH 7.2, 5 mmol/liter MgCl 2 , 5 mmol/ liter EDTA, 5 mmol/liter dithiothreitol, supplemented with Complete protease inhibitor mixture (Roche Diagnostics) and lysed by 4 freeze/ thaw cycles. After centrifugation at 14,000 ϫ g for 30 min at 4°C, protein concentrations and caspase-3 activities were determined in the supernatant according to the protocol of the CaspACE™ Fluorimetric Assay System (Promega). Data are expressed as picomole of the fluorescent substrate liberated per minute and microgram of protein.
TaqMan TM Real Time Quantitative RT-PCR-Total RNA was isolated with the RNeasy kit (Qiagen, Hilden, Germany) from VSMCs incubated for different time periods with 100 M phytanic or pristanic acid alone or in the presence of 1 ng/ml TNF␣ antibody or 3 ng/ml TGF␤1, respectively. Quantitation of the iNOS mRNA was done by exploiting the 5Ј nuclease activity of the TaqDNA polymerase to cleave a target gene specific TaqMan probe during RT-PCR. All TaqMan probes contained a 6-carboxy-fluorescein moiety as reporter dye at the 5Ј-end and a carboxytetramethyl rhodamine moiety as quencher dye at the 3Ј-end of the probe. Amplifications and sequence detections were carried out on an ABI PRISM 7700 (PE Applied Biosystems, Weiterstadt, Germany). Primers and TaqMan probes were purchased from MWG Biotech (Ebersberg, Germany). Forward and reverse PCR primers were used at a final concentration of 300 nM, TaqMan probes at a final concentration of 150 nM. Relative expression of the iNOS mRNA were evaluated relative to the expression of the ␤-actin mRNA in the same sample. All TaqMan experiments were performed using the Universal Master Mix system from PE Applied Biosystems. Cycling parameters were: 1 min 50°C, 1 min 95°C, 1 min 60°C. To prevent amplification of target sequences from contaminating genomic DNA, all RNA samples were subjected to digestion with RNase-free DNase I prior to the RT-PCR step.
Western Blotting of Inducible NOS-Subconfluent VSMCs were treated with 100 M phytanic or pristanic acid alone or in combination with 3 ng/ml TGF␤1 or 1 ng/ml TNF␣ antibody for various time intervals. Cells were then washed twice with ice-cold phosphate-buffered saline and harvested in a buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 25 mM NaF, 0,5% sodium deoxycholate, 10% SDS, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM orthovanadate, 10 mM sodium pyrophosphate supplemented with Com-plete protease inhibitor mixture. Thereafter, the cells were lysed by four freeze/thaw cycles and centrifuged in an Eppendorf benchtop centrifuge for 10 min at maximum speed. 50 g of supernatant protein were then subjected to Western blotting, performed according to standard procedures. The blots were developed with a 1:4000 dilution of a mouse monoclonal antibody specific for murine iNOS purchased from BD Transduction Laboratories (Lexington, KY).
Reporter Gene Studies-Subconfluent rat VSMCs were grown in standard 6-well dishes and co-transfected with 0.5 g/well of a pCAT3 vector (Promega) containing the NFB response element and the pCMV-␤gal vector together with 1.5 l per well of FuGENE 6 transfection reagent (Roche Diagnostics). 20 h after transfection, the cells were incubated in Dulbecco's modified Eagle's medium containing 1% fetal calf serum and 100 M phytanic or pristanic acid for various time intervals. ␤-galactosidase and CAT activities were measured with an ELISA detection kit supplied by Roche Diagnostics.
TNF␣ ELISA-Subconfluent rat VSMCs were treated with 100 M phytanic or pristanic acid for the indicated time intervals, and the TNF␣ concentration was determined in the culture medium employing a commercially available ELISA kit for rat TNF␣ (Biozol, Eching, Germany). The lower detection limit of the assay was ϳ17 pg/ml. For each experiment, a standard curve was obtained by serial dilutions of a rat TNF␣ stock solution to yield 0 -500 pg/ml.

RESULTS
To evaluate the effect of phytanic acid on exponentially growing rat VSMCs, 70% confluent cells were incubated with increasing concentrations of phytanic acid in the presence of 1% fetal calf serum as a survival factor. After 4 h of incubation in the presence of 100 M phytanic acid, cells rounded, shrunk, and lost contact to their neighbors (Fig. 1). As these observations were suggestive of cell death, we next stained VSMCs treated with phytanic acid with YO-PRO in order to detect plasma membrane permeability. As shown in Fig. 2A, the fraction of YO-PRO-positive cells increased at 6 h of incubation with 100 M phytanic acid compared with the Me 2 SO-treated control. Another well characterized marker for the detection of apoptotic and/or necrotic cells consists of annexin V binding (19). As shown in Fig. 2B, annexin V binding increased upon incubation with the fatty acid. Differentiation of apoptosis from necrosis was performed by three means. (i) Simultaneous incubation with propidium iodide and fluorescently labeled annexin V showed that Ͼ90% of the cells were annexin V-positive/ propidium iodide-negative. (ii) We induced necrotic cell death with a 5-min incubation at 42°C and compared the propidium iodide staining of the necrotic cells with that of the BCFAtreated cells by flow cytometry. Based on the fluorescence intensity per cell, we identified a fraction of only 5.6% of necrotic cells after 6 h of treatment with 100 M phytanic acid (control: 4.0%) (data not shown). (iii) We measured the activity of caspase-3, which is a major effector during the execution phase of apoptosis (20). We found that phytanic acid led to a steep time and dose-dependent increase in caspase-3 activity (Fig.  2C). The maximum effect was observed at 5 h at a concentration of 100 M of the fatty acid. To characterize the species specificity of phytanic acid induced apoptosis, we next incubated VSMCs from porcine and human origin with 100 M of this fatty acid. As is evident from Fig. 3, the effect of phytanic acid was not species-specific, although human cells were somewhat less susceptible in comparison to rat and porcine ones. Neither human umbilical vein endothelial cells nor human monocyte-derived macrophages could be forced to programmed cell death by a 6-h incubation with 100 M phytanic acid (data not shown).
Since phytanic acid represents a potent activator of PPAR␣ (2-4), we next investigated whether it shared its pro-apoptotic property with other known natural or artificial PPAR␣ agonists. We incubated rat VSMCs with palmitic acid, nervonic acid or linoleic acid, as well as with the artificial activators bezafibrate and Wy 14,643. None of these compounds were able to induce morphological changes, which were indicative for apoptosis (not shown). They also did not increase caspase-3 activity when administered to the cells for 6 h at a concentration of 100 M (Fig. 4A) or overnight at concentrations up to 250 M (data not shown). In vivo, phytanic acid is degraded via an ␣-oxidation step, yielding pristanic acid, before entering the peroxisomal ␤-oxidation cycle (21). Pristanic acid activates PPAR␣ more effectively than phytanic acid but it does not activate RXR (4, 10). Therefore, we examined the effect of pristanic acid on VSMC apoptosis. We found that pristanic acid at a concentration of 100 M for 6 h led to a significant activation of caspase-3 (Fig. 4A). The time course of caspase-3 activation differed between the two BCFA. At 100 M, maximal caspase-3 activity was reached already between 2 and 3 h compared with 5 h for phytanic acid (Fig. 4B).
To elucidate the mechanism of phytanic-and pristanic acidinduced apoptosis we characterized key regulators of the process. Since induction of apoptosis in VSMCs can be mediated by the production of NO (22,35), we reasoned that the BCFA effect might be due to excessive production of this compound. Since NO is produced via nitric-oxide synthases, we co-incubated VSMCs with phytanic or pristanic acid and different NOS inhibitors to test this hypothesis. We could demonstrate that the competitive inhibitor of all three NOS isoforms, monomethyl-L-arginine monoacetate (L-NMMA, 250 M), as well as addition of the specific inducible NOS inhibitor, aminohydroxyguanodine (AHG, 200 M), prevent BCFA induced apoptosis (Fig. 5). Our hypothesis that BCFA-induced apoptosis is mediated by NO was furthermore consistent with the observation that VSMCs treated simultaneously with phytanic acid and TGF␤1 were partially protected from programmed cell death (Fig. 5). It is known that TGF␤1 destabilizes iNOS mRNA, which decreases the amount of NO production (23).
As regulation of NO release from iNOS-expressing cells occurs predominantly at the DNA transcription and/or protein stability level we next analyzed the expression pattern of iNOS by quantitative RT-PCR and Western blot analysis. Cultured VSMCs were treated for different time intervals with 100 M phytanic acid alone or in combination with 3 ng/ml of TGF␤1 (Fig. 6). We could demonstrate that the amount of iNOS mRNA and protein markedly increased 2 h after onset of incubation and increased further with the duration of treatment with phytanic acid (Fig. 6, A and B). Co-incubation with TGF␤1 inhibited the increase in iNOS expression significantly at the level of mRNA (Fig. 6A) and also of iNOS protein (Fig. 6B).
Since the promotor region of the iNOS gene includes several binding sites for nuclear factor B (NFB), we next measured NFB activity using a transactivation assay. Subconfluent VSMCs were transfected with a plasmid containing a CAT reporter gene under the control of the NFB responsive element as well as with a ␤-gal normalization vector. 20 h after transfection cells were incubated with 100 M phytanic acid for different time periods and CAT and ␤-galactosidase activities were measured. NFB was activated shortly (between 15 and 30 min) after addition of phytanic acid, and its activity peaked at 60 min of incubation (Fig. 7A).
TNF␣, a membrane-bound cytokine that is cleaved by a specific membrane metalloprotease upon activation, is a well known activator of iNOS. Therefore, we next analyzed whether this cytokine may be involved as upstream factor in phytanic acid-induced apoptosis. To study the involvement of TNF␣, we first inhibited the mature soluble form of TNF␣ in the medium by incubating VSMCs simultaneously with 100 M phytanic acid and a TNF␣ neutralizing antibody. As shown in Fig. 7A, blocking TNF␣ with the antibody prevented the phytanic acidmediated activation of the NFB reporter gene construct almost completely. Moreover, the pro-apoptotic potential of phytanic acid resulting in increased caspase-3 activation was also almost completely abolished in the presence of the TNF␣ antibody (Fig. 7B), whereas incubation with a series of IgG fractions obtained from control sera had no effect (data not shown). To further corroborate this data we investigated whether the TNF␣ antibody was capable of preventing the intensive phytanic acid-induced expression of iNOS. VSMCs were treated with 100 M phytanic acid in the presence or absence of 1 ng/ml of the TNF␣-antibody. As shown in Fig. 8, A and B, the TNF␣ antibody reduced the level of iNOS protein, as detected in Western blots of cell extracts, and the steady state concentrations of the iNOS mRNA in comparison to VSMCs incubated with phytanic acid alone. To show the role of TNF␣ in a more direct manner, we next quantified the amount of functional TNF␣ in the medium. Subconfluent VSMCs were incubated with 100 M phytanic acid for different time intervals between 30 min and 3 h. Subsequently, the concentration of mature TNF␣ was measured in the medium using a rat TNF␣-specific ELISA. As shown in Fig. 8C, the content of TNF␣ in the medium was significantly higher in those media obtained from cells that had been treated with phytanic acid compared with the non-treated controls.
Since these data suggest that TNF␣ induction may be an upstream event involved in apoptosis induction by BCFA, we continued to investigate whether TNF␣ alone is sufficient of induce apoptosis in VSMCs. The cells were incubated in the presence of TNF␣ with or without cycloheximide for 8 h and caspase-3 activity was measured. As shown in Fig. 9, TNF␣ did not induce caspase-3 activation at concentrations of 20 or 40 ng/ml. In contrast, a moderate activation could be observed in the presence of TNF␣ and cycloheximide. However, the effect was distinctively smaller than obtained with phytanic or pristanic acid. DISCUSSION In this study, we show that BCFA induced apoptosis in VSMCs and that they exerted their effect via TNF␣ secretion and iNOS up-regulation. This is the first time that a fatty acid has been reported to induce apoptosis in VSMCs in vitro. However, it is known that various straight-chain fatty acids, like palmitate and stearate, are able to induce apoptosis in other cell types, such as cardiac myocytes or pancreatic ␤ cells (24 -26). Under the conditions of our study, the pro-apoptotic effect was specific for isoprenoid BCFAs like phytanic and pristanic acid and was not mimicked by saturated or unsaturated straight chain fatty acids. To test whether albumin affected the apoptosis-promoting activity of BCFA, we also performed the assays with media supplemented with up to 30 mg/ml of fatty acid-free bovine serum albumin (which would be sufficient to bind ϳ3 mM of fatty acid assuming a binding stoichiometry of 7:1). However, albumin neither inhibited the apoptosis promoting effect nor altered its specificity (data not shown). VSMCs of porcine and rat origin showed similar susceptibility toward BCFA-induced apoptosis whereas this effect was somewhat less pronounced in human VSMCs. Whether this difference was due to more rapid degradation of BCFA, the preponderance of other signaling pathways or to more efficient up-regulation of anti-apoptotic survival genes remains to be elucidated. Regarding the major cell types of the vascular wall, the pro- apoptotic effect of BCFA was specific for VSMCs as these fatty acids failed to cause programmed cell death in human umbilical vein endothelial cells and human monocyte-derived macrophages if subjected to the same conditions as were employed for VSMCs (data not shown).
Our group and others (2)(3)(4) reported previously that phytanic and pristanic acid are activating ligands of PPAR␣. PPAR␣ was first identified in liver cells as an important regulator of fatty acid metabolism but is also expressed within cells of the vascular wall like endothelial cells, smooth muscle cells and macrophages (6,8). It was shown that stimulation of PPAR␣ with the hypolipidmic drug Wy 14,643 causes apoptosis in macrophages (8) whereas it suppresses the IL-1␤-induced inflammatory response in human VSMCs (7). In this study, we demonstrate a pro-apoptotic effect of phytanic and pristanic acid, which is neither shared by the very potent artificial PPAR␣ activator Wy 14,643 nor by several straight chain fatty acids, which have been shown to act as natural PPAR␣ agonists. These findings correspond to results from the group of Staels who demonstrated that an incubation of human VSMCs with high concentrations of Wy 14,643 did not alter the viability of the cells (7). Conclusively, activation of PPAR␣ is most likely not responsible for the programmed cells death caused by BCFAs in VSMCs.
Our study reveals some mechanistic insights into BCFAinduced apoptosis. We could demonstrate that apoptosis induced by BCFA depends on TNF␣ activation and secretion because a TNF␣ neutralizing antibody blocked apoptosis almost completely. In addition, we could show that BCFA induced a steep increase of TNF␣ concentrations in the medium of VSMCs shortly after incubation with BCFA. This implies that BCFA or one of their downstream metabolites stimulate the cleavage of the membrane-bound precursor form of TNF␣, leading to release of the mature soluble form of the cytokine. In addition, our data support that BCFA, either directly or indirectly, trigger the secretion of TNF␣ in an autocrine manner, subsequently leading to programmed cell death. Consistent with previous reports (27), we could show however, that TNF␣ FIG. 8. TNF␣ is a key regulator of phytanic acid-induced apoptosis in VSMCs. A, rat VSMCs were incubated in the presence of Me 2 SO (1%), phytanic acid (100 M) alone, or together with a TNF␣ antibody (1 ng/ml) for 4 h. Subsequently, cell extracts were prepared and subjected to Western blot analysis. The blots were developed with a monoclonal antibody specific for rat iNOS as described under "Materials and Methods." B, rat VSMCs were treated with 100 M phytanic acid, either alone or in combination with the TNF␣ antibody (1 ng/ml) for up to 4 h. The amount of iNOS-specific mRNA was quantified using real-time PCR as described under "Materials and Methods." Data represent means Ϯ S.D. of three independent experiments. C, time course of mature TNF␣ concentration in media collected from VSMCs after treatment with 100 M phytanic acid, Me 2 SO, or medium alone (control) for various time periods. TNF␣ concentrations were quantified by ELISA and data represent means Ϯ S.D. of six independent experiments. alone is not sufficient to induce apoptosis in VSMCs. Thus, induction of apoptosis may require TNF␣ along with a combination of other metabolic changes or cytokines. So far, our attempts to identify the relevant factors that are affected by BCFA along with TNF␣ and thus may act in concert to induce apoptosis in VSMCs were however not successful.
There are several pathways that are known that lead from TNF␣ binding to its receptor to programmed cell death. The most prominent is activation of caspase-8 subsequently resulting in activation of effector caspases mediating the key processes during programmed cell death (28). We could demonstrate that caspase-8 is not involved in BCFA-induced apoptosis as no increase of activity of this enzyme could be detected after incubating VSMCs with these fatty acids for various time periods (data not shown). In addition to the signaling pathway leading to caspase-8 activation, TNF␣ regulates gene expression via activation of several transcription factors. One example for this is the induction of iNOS gene expression as a consequence of cytokine treatment in VSMCs (29). In line with this concept, we could demonstrate that iNOS mRNA as well as protein levels are more than 100-fold up-regulated after incubation of VSMCs with BCFA. The induction of iNOS was obliterated when TNF␣ secretion was blocked by adding a TNF␣-specific antibody to the cells. Furthermore, we were able to show that stimulation of iNOS expression is a necessary step during BCFA-induced cell death as different inhibitors of NOS abolished the pro-apoptotic effect of these fatty acids. Also the effects of TGF␤1, previously shown to destabilize iNOS mRNA (30), are consistent with an important role of NO in VSMC apoptosis. Nishio et al. (22) showed that incubation of VSMCs with various NO donors have the ability to induce apoptosis within this cell type (22). So far, it is not entirely clear, however, whether iNOS induction is sufficient to induce apoptosis in VSMCs.
Because we could show directly that iNOS stimulation depends on TNF␣ secretion, one may speculate that the activation of the first is mediated via NFB. It has been demonstrated by several groups that iNOS gene expression is tightly regulated by NFB with several responsive elements being found in the promoter region of the iNOS gene. Accordingly, we observed a significant 2-fold activation of endogenous NFB when VSMCs were incubated with BCFA. However, it is not clear at present whether a 2-fold up-regulation of the transcription factor may be sufficient to account for the observed more than a hundred-fold induction of iNOS. Moreover, it was demonstrated earlier that NFB may not be the only transcription factor that can mediate the cytokine-induced activation of iNOS gene expression (31,32).
Although the conditions of our current in vitro study differ from the situation in the intact organism, we will be interested to study whether apoptosis may be responsible for some of the toxic effects that result from supraphysiological concentrations of BCFA under in vivo conditions. Serum concentrations of phytanic acid up to 1 mM are present in patients suffering from peroxisomal disorders like Zellweger syndrome or Refsum disease. In addition, other conditions are known which are associated with an accumulation of phytanic acid along with pristanic acid as a result of a block in the peroxisomal ␤-oxidation of BCFA-CoA.
FIG. 9. TNF␣ is not sufficient to induce apoptosis in rat VSMCs. Subconfluent rat VSMCs were incubated for 8 h in the presence of the indicated concentrations of TNF␣ with or without cycloheximide (CHX, 10 mg/ml). Controls were incubated with Dulbecco's modified Eagle's medium containing 1% fetal calf serum for 8 h. Incubations with phytanic acid (6 h, 100 M) and pristanic acid (4 h, 100 M) were performed as described under "Materials and Methods." Caspase-3 activities were measured in cell extracts and expressed as percentage values relative to the value obtained for phytanic acid. They represent means Ϯ S.D. of three independent experiments each performed in triplicate.