Leaky (cid:1) -Oxidation of a trans -Fatty Acid INCOMPLETE (cid:1) -OXIDATION OF ELAIDIC ACID IS DUE TO THE ACCUMULATION OF 5- TRANS -TETRADECENOYL-CoA AND ITS HYDROLYSIS AND CONVERSION TO 5- TRANS -TETRADECENOYLCARNITINE IN THE MATRIX OF RAT MITOCHONDRIA*

The degradation of elaidic acid (9- trans -octadecenoic acid), oleic acid, and stearic acid by rat mitochondria was studied to determine whether the presence of a trans double bond in place of a cis double bond or no double bond affects (cid:1) -oxidation. Rat mitochondria from liver or heart effectively degraded the coenzyme A derivatives of all three fatty acids. However, with elaidoyl-CoA as a substrate, a major metabolite accumulated in the mitochondrial matrix. This metabolite was isolated and identified as 5- trans -tetradecenoyl-CoA. In contrast, little or none of the corresponding metabolites were detected with oleoyl-CoA or stearoyl-CoA as substrates. A kinetic study of long-chain acyl-CoA dehydrogenase (LCAD) and very long-chain acyl-CoA dehydrogenase revealed that 5- trans -tetradecenoyl-CoA is a poorer substrate of LCAD than is 5- cis -tetradecenoyl-CoA, while both unsaturated acyl-CoAs are the mixture, and, after preincubation for 2 min, respiration was initiated by the addition of 15 (cid:2) M of the indicated fatty acyl-CoA and 1 m M ADP to achieve state 3 respiration. Rates of respiration were measured polarographically with a Clark oxygen electrode attached to a YS-oxygraph. Analysis of Acyl-CoAs Present in the Mitochondrial Matrix— When fatty acyl-CoAs were analyzed that are present in the mitochondrial matrix, the incubation mixture used for respiration measurements was scaled up to 20 ml. The reaction was terminated by the addition of 25 ml of methanol at the indicated time. After addition of 15 nmol of penta- decanoyl-CoA as internal standard, the suspension was centrifuged at 17,500 (cid:1) g for 15 min. The supernatant was diluted 5-fold before it was passed slowly through a C 18 Sep-Pak column. The bound CoA derivatives were eluted with 2.5 ml of methanol. The extraction process was repeated, and 1 ml of 50 m M ammonium phosphate (pH 5.5) was added to the combined methanolic extracts. After removal of methanol under N 2 or reduced pressure, samples were applied to a (cid:2) Bondapak C 18 reverse-phase column (30 cm (cid:1) 3.9 mm) attached to a Waters gradient HPLC system. The absorbance of the effluent was monitored at 254 nm. Separation was achieved by linearly increasing the acetonitrile/H 2 O (9:1, v/v) content of the 50 m M ammonium phosphate elution buffer (pH 5.5) from 20% to 80% in 30 min at a flow rate of 2 ml/min. Metabolites were quantified by integrating areas under the peaks by use of the Millennium software from Waters Corp. and by using the determined recovery ( (cid:3) 35%) of the internal standard to calculate concentrations of acyl-CoAs. To identify the major metabolite of elaidoyl-CoA, the appro-priate HPLC fractions were collected and freed of acetonitrile by evap- oration under reduced pressure. The resultant aqueous solution was containing 1% at (cid:4) °C. Enzyme and Protein Assays— Thioesterase and CPT II activities were determined spectrophotometrically by measuring the release of coenzyme A from acyl-CoAs with Ellman’s reagent (13) at 412 nm. A typical thioesterase assay mixture contained 0.175 M KP i (pH 8), 0.2 m M 5,5 (cid:6) dithiobis (2-nitrobenzoic acid) (Ellman’s reagent), and 20 (cid:2) M acyl-CoA. The CPT II assay mixture contained 115 m M Tris-HCl (pH 8), 0.1% Triton X-100, 0.11 m M 5,5 (cid:6) -dithiobis(2-nitrobenzoic acid), 1.1 m M L - carnitine, and 35 (cid:2) M acyl-CoA. An extinction coefficient of 13,600 M (cid:4) 1 cm (cid:4) 1 was used to calculate rates. The need to correct rates was most pronounced when partially purified enzyme was used, which generally contained proteins that have sulfhydryl groups. Acyl-CoA dehydrogen- ase was assayed spectrophotometrically at 600 nm with phenazine methosulfate as the primary electron acceptor and dichlorophenolindo- phenol as the final electron acceptor (17). Kinetic parameters ( K m and V max ) were determined for LCAD and VLCAD with tetradecanoyl-CoA, 5- cis -tetradecenoyl-CoA, and 5- trans -tetradecenoyl-CoA as substrates by the fluorometric assay method with 1 m M electron transferring flavoprotein as electron acceptor as detailed by Mohsen and Vockley (18). Data were analyzed by nonlinear curve fitting using the Sigma plot program. One unit of enzyme activity is defined as the amount of enzyme that converts 1 (cid:2) mol of substrate to product per minute. Pro- tein concentrations were determined by the dye-binding

Unsaturated fatty acids with trans double bonds, also referred to as trans fatty acids, are part of the human diet, because they are present in dairy products, meat of ruminants, and partially hydrogenated vegetable oils (1). Clinical studies have prompted the idea that consumption of trans fatty acids has an adverse effect on human health because of an increased risk of cardiovascular diseases (2).
We have extended our study of the ␤-oxidation of unsaturated fatty acids to include unsaturated fatty acids with trans double bonds because of an interest in understanding the consequences of trans fatty acid consumption and to fully explore the molecular mechanisms of double bond metabolism during ␤-oxidation. Elaidic acid, 9-trans-octadecenoic acid, was chosen as a substrate because it contains only one double bond and because it is the geometric isomer of oleic acid whose degradation by ␤-oxidation has been studied in detail (3). Moreover, it has been reported that elaidic acid is partially converted to 5-trans-tetradecenoic acid when it serves as a substrate in the perfused rat heart (4). That observation is surprising, because mitochondrial ␤-oxidation is thought to go to completion without the accumulation of significant amounts of intermediates (5). The unusual leakage of an intermediate during the ␤-oxidation of a trans fatty acid and the potential for gaining a better understanding of how ␤-oxidation is coordinated in intact mitochondria have prompted this investigation. liver dienoyl-CoA isomerase (9) were purified by published procedures.
Isolation of Mitochondria from Rat Liver and Respiration Measurements-Rat liver mitochondria were isolated as described by Nedergaard and Cannon (14) from male Sprague-Dawley rats (240 -260 g) kept on a standard chow and then fasted for 24 h before the isolation of mitochondria. For respiration measurements, 1.5 mg of rat liver mitochondria were incubated in 1.9 ml of incubation buffer containing 20 mM Tris-HCl (pH 7.4), 4 mM KP i , 0.1 M KCl, 4 mM MgCl 2 , 0.1 mM EGTA, bovine serum albumin (0.5 mg/ml), and 0.5 mM L-malate. L-Carnitine (1 mM) was added to the mixture, and, after preincubation for 2 min, respiration was initiated by the addition of 15 M of the indicated fatty acyl-CoA and 1 mM ADP to achieve state 3 respiration. Rates of respiration were measured polarographically with a Clark oxygen electrode attached to a YS-oxygraph.
Analysis of Acyl-CoAs Present in the Mitochondrial Matrix-When fatty acyl-CoAs were analyzed that are present in the mitochondrial matrix, the incubation mixture used for respiration measurements was scaled up to 20 ml. The reaction was terminated by the addition of 25 ml of methanol at the indicated time. After addition of 15 nmol of pentadecanoyl-CoA as internal standard, the suspension was centrifuged at 17,500 ϫ g for 15 min. The supernatant was diluted 5-fold before it was passed slowly through a C 18 Sep-Pak column. The bound CoA derivatives were eluted with 2.5 ml of methanol. The extraction process was repeated, and 1 ml of 50 mM ammonium phosphate (pH 5.5) was added to the combined methanolic extracts. After removal of methanol under N 2 or reduced pressure, samples were applied to a Bondapak C 18 reverse-phase column (30 cm ϫ 3.9 mm) attached to a Waters gradient HPLC system. The absorbance of the effluent was monitored at 254 nm. Separation was achieved by linearly increasing the acetonitrile/H 2 O (9:1, v/v) content of the 50 mM ammonium phosphate elution buffer (pH 5.5) from 20% to 80% in 30 min at a flow rate of 2 ml/min. Metabolites were quantified by integrating areas under the peaks by use of the Millennium software from Waters Corp. and by using the determined recovery (ϳ35%) of the internal standard to calculate concentrations of acyl-CoAs. To identify the major metabolite of elaidoyl-CoA, the appropriate HPLC fractions were collected and freed of acetonitrile by evaporation under reduced pressure. The resultant aqueous solution was mixed with an equal volume of 4 N KOH and kept at room temperature for 1 h to hydrolyze the thioester bond. The mixture was adjusted pH to 1 with concentrated H 2 SO 4 and extracted three times with ether. The combined extracts were dried over anhydrous Na 2 SO 4 and filtered. After removal of ether by evaporation under a stream of N 2 , the residue was dissolved in 1 ml of boron trichloride in methanol (12% w/w) and 2,2-dimethoxypropane (100 l). After this mixture was heated at 60°C for 10 min, 1 ml of H 2 O was added to terminate the reaction. The mixture was extracted three times with 1 ml of hexane each. The combined hexane extracts were dried over anhydrous Na 2 SO 4 and filtered. The filtrate was concentrated under N 2 , and the residue was subjected to GC-MS analysis.
For the spectrophotometric identification of the major metabolite of elaidoyl-CoA, the HPLC-purified compound, presumably 5-trans-tetradecenoyl-CoA, was sequentially reacted in phosphate buffer (pH 8) with 0.1 unit of acyl-CoA oxidase, 8 milliunits of enoyl-CoA isomerase, and 4 milliunits of dienoyl-CoA isomerase. Absorbance spectra were recorded between 200 and 400 nm to follow the progress of the reactions and to document the final UV absorbances.
Identification of 5-trans-Tetradecenoic Acid and 5-trans-Tetradecenoyl-L-carnitine after Incubation of Rat Liver Mitochondria with Elaidoyl-CoA-For identification of 5-trans-tetradecenoic acid, the incubation mixture used for respiration measurements with rat liver mitochondria was scaled up to 20 ml. After incubation for 5 min with elaidoyl-CoA, oleoyl-CoA, or stearoyl-CoA and ADP, the reaction was terminated by the addition of 4 ml of 1.2 N HCl. Twenty micrograms of pentadecanoic acid were added as an internal standard. Ether of the highest purity was used to extract free fatty acids. The ether extract was dried over anhydrous sodium sulfate and evaporated under a stream of nitrogen. The sample was dissolved in CH 2 Cl 2 and directly analyzed by GC-MS, or was analyzed after conversion of acids to methyl esters by treatment with boron trichloride in methanol. Aliquots of 1 l of the fatty acid methyl esters were injected at 250°C into a GC/MS instrument (Shimadzu Scientific Instruments) consisting of a gas chromatograph (model GC-17A) interphased with a mass spectrometer (QP-5000) and equipped with a capillary column (30 m, inner diameter: 0.25 mm; film thickness, 0.25 m; EC-5, Alltech Associates Inc., Deerfield, IL). The oven temperature was raised from 100 to 230°C at 5°C/min, to 300°C at 20°C/min and then held constant for 6 min. The mass spectrometer served as a detector and was operated at 280°C. For the identification of 5-trans-tetradecenoyl-L-carnitine, the incubation mixture used to measure respiration was scaled up to 10 ml. After incubation for 5 min with ADP and 15 M elaidoyl-CoA, oleoyl-CoA, or stearoyl-CoA, the reaction was terminated by the addition of 2 ml of 1.2 N HCl and 2.5 l of 2 mM myristoyl-L-carnitine in CH 3 OH/H 2 O (3:1, v/v) as internal standard. After keeping the mixture for 5 min at 4°C, the pH was adjusted to ϳ5, and the incubation was continued for another 5 min. The resulting suspension was centrifuged at 17,500 ϫ g for 15 min, and the supernatant was collected for acylcarnitine measurements. Acylcarnitines were analyzed as butyl esters by stable isotope dilution electrospray ionization tandem mass spectrometry using a modification of the method of Rashed et al. (15). Deuterium-labeled acylcarnitines (0.2 pmol of C4-C14 and 0.4 pmol of C16-C18) were added as internal standards to 15 l of neutralized extract and centrifuged. The supernatant was dried under a stream of nitrogen then butylated by reacting it with 0.1 ml of 3 N HCl in butanol at 55°C for 15 min. The sample was dried under a stream of nitrogen and dissolved in 0.15 ml of acetonitrile/ water (4:1, v/v). 20 l was used for flow injection analysis by electrospray ionization tandem mass spectrometry. Data were acquired by precursor ion scanning of the common product ion at m/z 85 for labeled and unlabeled acylcarnitines. The concentrations of acylcarnitines were calculated from the peak intensities of unlabeled acylcarnitines relative to labeled internal standards.
Expression and Purification of Rat Long Chain Acyl-CoA Dehydrogenase-A cDNA insert corresponding to the mature form of rat LCAD and its expression vector were prepared by PCR as described (16). Escherichia coli strain XL1-Bla transformed with this plasmid was grown in LB media at 30°C in the presence of ampicillin to an absorbance of about 1.0 at 600 nm and then induced in the presence of 0.6 mM isopropyl-␤-D-thiogalactoside for 6 h. Cell were harvested by centrifugation for 20 min at 8000 ϫ g, and pellets were suspended in 10 mM Tris-HCl (pH 7.5) buffer containing 10% glycerol, 0.3 mM EDTA, and 1 mM FAD (buffer A). The suspension was sonicated 10 times for 20 s each at 4°C and then centrifuged at 100,000 ϫ g for 1 h. The supernatant was loaded onto a DEAE-Sepharose column (2.5 ϫ 40 cm) that had been equilibrated overnight with buffer A. After washing the column with buffer A, the column was eluted with a linear gradient from 0 to 300 mM NaCl in buffer A. Fractions were assayed for long-chain acyl-CoA dehydrogenase. Active fractions were pooled, concentrated in an Amicon concentrator with a YM-10 membrane, and dialyzed against 50 mM KP i (pH 7.6) containing 10% glycerol, 0.3 mM EDTA, and 1 M FAD (buffer B). The sample was applied to a hydroxylapatite column (2.5 ϫ 25 cm), equilibrated with buffer B, and eluted with a linear gradient from 50 to 400 mM KP i (pH 7.6) in buffer B. Active fractions were identified, collected, and concentrated. After overnight dialysis against 10 mM KP i (pH 8.0), 0.3 mM EDTA, 15% glycerol (buffer C), the sample was applied to a Blue Sepharose CL-6B column (1.5 ϫ 10 cm) equilibrated with buffer C. The column was developed with a linear gradient from 0 to 0.6 M KCl in buffer C. Active fractions were combined, concentrated, and dialyzed against 10 mM KP i (pH 7.5), 0.3 mM EDTA, and 20% glycerol (buffer D). The sample was loaded onto a Q-Sepharose column (0.5 ϫ 5 cm) previously equilibrated with buffer D. The column was developed with a linear gradient from 0 to 0.5 M KP i in buffer D. Active fractions were pooled, concentrated to 0.6 ml, diluted with glycerol to ϳ1 ml, and stored at Ϫ80°C.
Partial Purification of Mitochondrial Thioesterase and CPT II from Rat Liver-For the partial purification of thioesterase, rat liver mitochondria were purified by gradient density centrifugation in a selfgenerated gradient of iodixanol (OptiPrep). Equal volumes of iodixanol (50%, w/v) containing 0.25 M sucrose, 1 mM EDTA, and 10 mM Mops-NaOH (pH 7.4) and a suspension of heavy mitochondria were mixed (final iodixanol concentration ϭ 25%; ϭ 1.150 g/ml) and then transferred to 10-ml tubes and centrifuged at 180,000 g av for 3 h in a T865-1 fixed angle rotor at 4°C using the slow acceleration and braking modes. Fractions were collected from the bottom after slowly inserting a thin glass tube through the bottom of the tube. Catalase and malate dehydrogenase were assayed as marker enzymes for mitochondria and peroxisomes, respectively. Fractions containing most mitochondria were combined and diluted 2-fold with MST isolation buffer (210 mM mannitol, 70 mM sucrose, 10 mM Tris-HCl, pH 7.4, and 0.1 mM EDTA) before they were harvested by centrifugation at 17,500 ϫ g for 20 min. The mitochondrial pellet was suspended, dialyzed against MST isolation buffer for at least 4 -6 h, and then sonicated 6 times for 20 s each at intervals of 30 s to keep the temperature of the suspension at 4°C. The suspension of sonicated mitochondria was centrifuged for 1 h at 100,000 ϫ g, and the resultant supernatant was brought to 70% saturation with ammonium sulfate. Precipitated proteins were collected by centrifugation at 15,000 ϫ g for 10 min, dissolved in 3 ml of 20 mM KP i (pH 7.0) containing 0.5 mM benzamidine, 0.5 mM dithiothreitol, 10% glycerol, 30 mM NaCl, and applied to a Sephacryl S-200HR column (2.5 ϫ 40 cm). After elution with the same buffer, active fractions were identified, pooled, concentrated to 0.6 ml, and stored at Ϫ80°C. For the purification of CPT II, mitochondria (1.3 mg/ml) in 20 mM KP i (pH 7.4) containing 1 M KCl (buffer A) were treated with 0.2% CHAPS for 10 min at 4°C. The resultant suspension was centrifuged at 1,500 ϫ g for 10 min, and the supernatant was applied to an Octyl-Sepharose CL-4B column (1 ϫ 4 cm) equilibrated with buffer A containing 0.2% CHAPS. CPT II was eluted with buffer A containing 1% CHAPS, and active fractions were pooled, concentrated, and stored at Ϫ80°C.
Enzyme and Protein Assays-Thioesterase and CPT II activities were determined spectrophotometrically by measuring the release of coenzyme A from acyl-CoAs with Ellman's reagent (13)  Ϫ1 cm Ϫ1 was used to calculate rates. The need to correct rates was most pronounced when partially purified enzyme was used, which generally contained proteins that have sulfhydryl groups. Acyl-CoA dehydrogenase was assayed spectrophotometrically at 600 nm with phenazine methosulfate as the primary electron acceptor and dichlorophenolindophenol as the final electron acceptor (17). Kinetic parameters (K m and V max ) were determined for LCAD and VLCAD with tetradecanoyl-CoA, 5-cis-tetradecenoyl-CoA, and 5-trans-tetradecenoyl-CoA as substrates by the fluorometric assay method with 1 mM electron transferring flavoprotein as electron acceptor as detailed by Mohsen and Vockley (18). Data were analyzed by nonlinear curve fitting using the Sigma plot program. One unit of enzyme activity is defined as the amount of enzyme that converts 1 mol of substrate to product per minute. Protein concentrations were determined by the dye-binding assay as described by Bradford (19) with bovine serum albumin as standard.

Respiration Rates of Rat Mitochondria with cis and trans
Monounsaturated Fatty Acyl-CoAs as Substrates-The perfused rat heart has been reported to produce 5-trans-tetradecenoic acid from elaidic acid, which appears to be a sufficient energy source for sustaining the heartbeat (4). In an effort to compare the effectiveness of elaidic acid with oleic acid and stearic acid as substrates of mitochondrial ␤-oxidation, we measured respiration rates of coupled rat liver and heart mitochondria with several fatty acyl-CoAs as substrates. As shown in Fig. 1, stearoyl-CoA and elaidoyl-CoA supported equal rates of respiration in rat liver mitochondria, whereas oleoyl-CoA sustained a rate that was ϳ50% higher than the rates observed with the other two substrates. Rates obtained with myristoyl-CoA, 5-cis-tetradecenoyl-CoA, and 5-trans-tetradecenoyl-CoA were higher than the rates supported by any of the three longer chain acyl-CoAs. In agreement with a previous report (20), elaidoyl-CoA was degraded in rat heart mitochondria at a rate that was 50 and 33% lower than rates observed with oleoyl-CoA and stearoyl-CoA, respectively (results not shown). These data indicate that elaidic acid is an adequate substrate of ␤-oxidation in both rat liver and heart mitochondria.
Detection and Identification of 5-trans-Tetradecenoyl-CoA in Rat Mitochondria Incubated with Elaidoyl-CoA-In an effort to detect possible differences between the ␤-oxidation of elaidoyl-CoA and oleoyl-CoA, coupled rat liver mitochondria were incubated with either elaidoyl-CoA, oleoyl-CoA, or stearoyl-CoA at state three respiration and fatty acyl-CoAs were analyzed by HPLC. Because no CoA was added to the incubation mixture, fatty acyl-CoAs other than the substrate were overwhelmingly, if not completely, present in the mitochondrial matrix. Shown in Fig. 2 are HPLC chromatograms of fatty acyl-CoAs that were isolated from rat liver mitochondria after incubation with stearoyl-CoA ( Fig. 2A), oleoyl-CoA (Fig. 2B), or elaidoyl-CoA (Fig.  2C) for 5 min. One major peak in each chromatogram corresponds to the fatty acyl-CoA that served as substrate. A second peak marked "IS" is due to pentadecanoyl-CoA that was added as an internal standard after the incubation was completed to facilitate the quantification of metabolites. However, in Fig.  2C, a third prominent peak is visible, which, based on its elution time, was tentatively identified as 5-trans-tetradecenoyl-CoA (t⌬ 5 C14). After further purification by HPLC, this material was hydrolyzed and converted to its methyl ester. Analysis of the methyl ester by gas chromatography/mass spectrometry proved this material to be a homogeneous compound with a molecular ion at m/z of 258 as expected of methyl tetradecenoate upon chemical ionization in the presence of ammonia (data not shown). The material presumed to be 5-trans-tetradecenoyl-CoA had a UV spectrum with an absorbance maximum around 260 nm that is typical of an acyl-CoA (see spectrum 1 in Fig. 3). Upon treatment with acyl-CoA oxidase, spectrum 1 changed to spectrum 2. The observed increase in the absorbance around 260 nm agrees with the expected conversion of 5-trans-tetradecenoyl-CoA to 2,5-trans-tetradecadienoyl-CoA. Treatment of the latter compound with enoyl-CoA isomerase yielded spectrum 3 that is characteristic of 3,5-dienoyl-CoAs. Finally, when the compound, presumed to be 3,5tetradecadienoyl-CoA, was incubated with dienoyl-CoA isomerase, spectrum 3 changed to spectrum 4 with absorbance maxima at 260 and 300 nm, which is characteristic of 2,4dienoyl-CoA. Taken together, the data demonstrate that 5-tetradecenoyl-CoA accumulated in the matrix of mitochondria that ␤-Oxidation of Elaidic Acid in Rat Mitochondria authentic 5-cis-tetradecenoyl-CoA was observed (see Fig. 2B). However, the putative cis metabolite was present at a level 10-times lower than the trans compound (Fig. 2, compare B and  C), and therefore no attempt was made to further identify it. Altogether, it is very likely that the metabolite of oleoyl-CoA was 5-cis-tetradecenoyl-CoA. Tetradecanoyl-CoA (myristoyl-CoA), the corresponding metabolite of stearoyl-CoA, was not detected and therefore did not accumulate at a level that was higher than the background noise.
Kinetics of the Dehydrogenation of 5-trans-and 5-cis-Tetradecenoyl-CoA and Myristoyl-CoA by LCAD and VLCAD-The observed accumulation of 5-trans-tetradecenoyl-CoA in the matrix of rat liver mitochondria raised the question as to the cause for the build-up of this metabolite. We hypothesized that 5-trans-tetradecenoyl-CoA may be a relatively poor substrate of LCAD and/or VLCAD, both of which dehydrogenate long-chain acyl-CoAs. This prediction is supported by the data shown in Table I. The catalytic efficiency (k cat /K m ) of LCAD was 4-times lower with 5-trans-tetradecenoyl-CoA as substrate than with 5-cis-tetradecenoyl-CoA or tetradecanoyl-CoA (myristoyl-CoA). This lower catalytic efficiency is due to a 4-fold higher K m for 5-trans-tertatedecenoyl-CoA as compared with the K m values for the other two substrates. In contrast VLCAD acted equally well on the cis and trans isomers of 5-tetradecenoyl-CoA, which, however, were poorer substrates of this enzyme than was tetradecanoyl-CoA. The latter observation agrees with a previous report showing that saturated acyl-CoAs are better substrates of VLCAD than the corresponding unsaturated substrates with 4,5-or 5,6-double bonds (21).

Identification of 5-trans-Tetradecenoic Acid as a Product of Elaidate ␤-Oxidation in Rat Liver
Mitochondria-The reported formation of 5-trans-tetradecenoic acid in rat hearts perfused with elaidic acid (4) suggested that 5-trans-tetradecenoyl-CoA, a metabolite of elaidate ␤-oxidation, may be hydrolyzed in mitochondria and the resultant free fatty acid may exit from cells. To test this idea, isolated rat liver mitochondria were incubated with elaidoyl-CoA, oleoyl-CoA, or stearoyl-CoA for 5 min, and the resultant acidic products were extracted and analyzed by gas chromatography/mass spectrometry (GC/MS). Shown in Fig. 4 are the gas chromatograms of the acids extracted after incubating mitochondria with elaidoyl-CoA or oleoyl-CoA. Only the incubation of mitochondria with elaidoyl-CoA yielded a compound, marked C14:1 acid, with an elution time identical to that of 5-trans-tetradecenoic acid (see Fig. 4, A  and B). The mass spectrum of this compound was virtually identical with that of 5-trans-tetradecenoic acid (Fig. 4, compare C and D). Hence 5-trans-tetradecenoyl-CoA, which is formed by ␤-oxidation of elaidoyl-CoA in mitochondria and accumulates in the matrix, was hydrolyzed. It was estimated that 1-2% of the elaidoyl-CoA present in the incubation mixture was converted to 5-trans-tetradecenoic acid. The hydrolysis of 5-trans-tetradecenoyl-CoA in the mitochondrial matrix requires an acyl-CoA thioesterase that was detected in the soluble extract from rat liver mitochondria. Shown in Fig. 5 is the substrate profile of a partially purified preparation of one or more of these thioesterases. The enzyme was most active with substrates having acyl chains with 12 and 14 carbon atoms and hence is best classified as a long-chain acyl-CoA thioesterase. The enzyme was highly active with 5-tetradececenoyl-CoA, although slightly more so with the cis than the trans isomer. Intermediates of ␤-oxidation, e.g. 2-trans-tetradecenoyl-CoA, 3-hydroxytetradecanoyl-CoA, and 3-ketohexadecanoyl-CoA, were poorer substrates than regular fatty acyl-CoAs of equal chain length. K m and V max values for the thioesterase-catalyzed hydrolysis of tetradecanoyl-CoA (myristoyl-CoA) were determined to be 12.6 Ϯ 0.8 M and 22.8 Ϯ 0.6 CoA (A), oleoyl-CoA (B), or elaidoyl-CoA (C). Abbreviations: IS, internal standard (pentadecanoyl-CoA); C18, stearoyl-CoA; c⌬5C14, 5-cis-tetradecenoyl-CoA; c⌬9C18, oleoyl-CoA; t⌬5C14, 5-transtetradecenoyl-CoA; and t⌬9C18, elaidoyl-CoA.
Formation of 5-trans-Tetradecenoylcarnitine during the ␤-Oxidation of Elaidic Acid in Rat Liver Mitochondria-The observed hydrolysis of 5-trans-tetradecenoyl-CoA in mitochondria raised the question: Is the 5-trans-tetradecenoyl residue also transferred to carnitine? To answer this question, rat liver mitochondria were incubated with elaidoyl-CoA or oleoyl-CoA for 5 min, and the resultant aqueous phases after removal of protein were analyzed by tandem mass spectrometry to identify acylcarnitines. The mass spectrum of acylcarnitines that were detected after incubating mitochondria with elaidoyl-CoA (see Fig. 6A) shows a peak labeled C14:1 that corresponds to tetradecenoylcarnitine with an m/z of 426.4. Also detected was a peak that corresponds to elaidoylcarnitine with an m/z of 482.5. All other acylcarnitines, including myristoylcarnitine (C14) with an m/z of 428.4, and a mixture of deuterated acylcarnitines marked by asterisks were added as internal stand-ards at the end of the incubation period or prior to derivatizing the acylcarnitines. Most important was the demonstration that tetradecenoylcarnitine (C14:1), presumably 5-trans-tetradecenoylcarnitine, was formed during the ␤-oxidation of elaidoyl-CoA (see Fig. 6B), whereas ␤-oxidation of oleoyl-CoA did not yield such a metabolite (see Fig. 6C). It was estimated that 4 -5% of elaidoyl-CoA present in the incubation mixture was converted to 5-trans-tetradecenoylcarnitine. The latter compound was most likely formed from 5-trans-tetradecenoyl-CoA by carnitine palmitoyltransferase II (CPT II) in the mitochondrial matrix. To confirm this idea, CPT II was partially purified from rat liver mitochondria under conditions that resulted in the inactivation of CPT I (22). This preparation of CPT II was used to determine kinetic properties of this enzyme at a fixed concentration of 1 mM carnitine. The apparent K m values for tetradecanoyl-CoA and 5-trans-tetradecenoyl-CoA as substrates were found to be 21.6 Ϯ 2 M and 10.6 Ϯ 1.3 M, respectively, whereas the corresponding V max values were 286 Ϯ 10 and 387 Ϯ 17 milliunits/mg, respectively. Thus, the catalytic efficiency of CPT II is ϳ3-times greater with 5-trans-  4. Identification of 5-trans-tetradecenoic acid as a product of elaidate ␤-oxidation in rat liver mitochondria. A, gas chromatogram of acidic products extracted from rat liver mitochondria after incubation with elaidoyl-CoA or oleoyl-CoA. B, expanded region of the gas chromatogram where 5-trans-tetradecenoic acid was eluted. C, mass spectrum of the material corresponding to the peak detected at 18 min in the gas chromatogram. D, mass spectrum of authentic 5-trans-tetradecenoic acid. Abbreviations: C14:1 acid, 5-trans-tetradecenoic acid; IS, internal standard (pentadecanoic acid). tetradecenoyl-CoA than with tetradecanoyl-CoA as substrate.
An attempt was made to estimate the relative rates at which 5-trans-tetradecenoyl-CoA was hydrolyzed and converted to the carnitine derivative in the mitochondrial matrix. For this purpose, activities of CPT II and thioesterase were measured in extracts of rat liver mitochondria at one concentration of 5-trans-tetradecenoyl-CoA. The values thus obtained together with the kinetic parameters determined for the partially purified enzymes with the same substrate were used to calculate activities of the two enzymes as a function of the concentration of 5-trans-tetradecenoyl-CoA (Fig. 7). The data show that the conversion of 5-trans-tetradecenoyl-CoA to 5-trans-tetradecenoylcarnitine is favored by a factor of 1.5 over the hydrolysis of 5-trans-tetradecenoyl-CoA to 5-trans-tetradecenoic acid (see Fig. 7). This conclusion agrees with the observed greater accumulation of 5-trans-tetradecenoylcarnitine than 5-trans-tetradecenoic acid.

DISCUSSION
This study was initiated with the aim of analyzing the mitochondrial ␤-oxidation of trans fatty acids at the molecular level. The reported formation of 5-trans-tetradecenoic acid from elaidic acid (9-trans-octadecenoic acid) in perfused rat hearts (4) prompted the idea that ␤-oxidation of trans fatty acids may be an atypical process, because the substrate or part of the substrate was incompletely degraded. In contrast, oleic acid did not give rise to a partially degraded substrate. Despite its incomplete degradation, elaidic acid seemed to be a sufficient energy source for the beating heart (4). Moreover, elaidoyl-CoA was observed to support high rates of respiration in isolated rat heart and liver mitochondria (see Ref. 20 and this study). Hence, the energy production in rat mitochondria does not seem to be compromised when elaidic acid serves as substrate even though part of it is not completely degraded. The premature termination of elaidate ␤-oxidation is most likely related to the accumulation of 5-transtetradecenoyl-CoA in the mitochondrial matrix. Apparently 5-trans-tetradecenoyl-CoA is more rapidly formed than degraded by ␤-oxidation, whereas the corresponding intermediates of more common dietary fatty acids, including 5-cis-tetradecenoyl-CoA derived from oleic acid, are degraded without accumulating to a significant extent. The most obvious reason for the accumulation of 5-trans-tetradecenoyl-CoA would be its slower dehydrogenation by LCAD or/and VLCAD as compared with its formation from elaidoyl-CoA by two cycles of ␤-oxidation (Scheme 1). A kinetic analysis revealed that 5-trans-tetradecenoyl-CoA is a poorer substrate of LCAD than is 5-cis-tetradecenoyl-CoA or myristoyl-CoA. Moreover, both unsaturated fatty acyl-CoAs are poor substrates of VLCAD when compared with myristoyl-CoA. These data support the conclusion that the presence of a 5-trans double bond in place of a 5-cis or no double bond in the substrate reduces the catalytic efficiency of LCAD. However, because the lower catalytic efficiency of LCAD with 5-trans-tetradecenoyl-CoA compared with its efficiency with the 5-cis isomer is due only to a higher K m for the former substrate, an increase in the concentration of 5-trans-tetradecenoyl-CoA should eliminate or minimize the difference between the rates at which the two isomeric substrates are dehydrogenated. Thus, the accumulation of 5-trans-tetradecenoyl-CoA in the mitochondrial matrix is expected to cause the rate of its dehydrogenation to increase to a level that is achieved with a lower concentration of 5-cis-tetradecenoyl-CoA. This conclusion is supported by the observation that 5-trans-tetradecenoyl-CoA supports a rate of respiration that is only 20% lower than rates obtained with 5-cis-tetradecenoyl-CoA or myristoyl-CoA as substrates. Any remaining difference between the rates of elaidate and oleate ␤-oxidation may be due to the inhibitory effects of accumulated 5-trans-tetradecenoyl-CoA or to the adverse effect of the trans double bond on other reactions of ␤-oxidation.
Another consequence of the accumulation of 5-trans-tetradecenoyl-CoA is the effective competition of other enzymes for this intermediate, which thereby will be diverted from ␤-oxidation. As summarized in Scheme 1, 5-trans-tetradecenoyl-CoA in the mitochondrial matrix is not only a substrate of LCAD and VLCAD but also of thioesterase and carnitine palmitoyltransferase II (CPT II). The hydrolysis by thioesterase yields 5-trans-tetradecenoic acid, which is assumed to pass through the mitochondrial and cellular membranes to enter the circulation where it has been observed previously (4). Alternatively, it will move from the mitochondrial matrix into the cytosol where it can be activated by conversion to its CoA derivative and utilized for lipid synthesis, mitochondrial ␤-oxidation, or perhaps for protein acylation in place of myristoyl-CoA. The long-chain thioesterase activity present in rat liver mitochondria seems to be due to one enzyme that has been purified to apparent homogeneity and shown to be highly active with palmitoyl-CoA and myristoyl-CoA (23). The transfer of the 5-trans-tetradecenoyl residue from CoA to carnitine is catalyzed by CPT II that normally operates in the opposite direction to supply substrates for ␤-oxidation. The formation of 5-transtetradecenoylcarnitine is favored when the intramitochondrial concentration of medium-chain and long-chain acyl-CoAs is high (22). Both the hydrolysis of 5-trans-tetradecenoyl-CoA and its conversion to 5-trans-tetradecenoylcarnitine will only contribute significantly to the metabolism of elaidic acid at an elevated concentration of the intermediate, because the K m values of thioesterase and CPT II for 5-trans-tetradecenoyl-CoA are 4-to 10-times higher than the K m values of LCAD and VLCAD for the same substrate while the intramitochondrial activities of thioesterase and CPT II are lower than those of LCAD and VLCAD (see Ref. 21 and data from this study).
The incomplete degradation of elaidic acid raises the specter of other monounsaturated and polyunsaturated trans fatty acids yielding novel fatty acids by partial ␤-oxidation. Because of continuing health concerns about the consumption of trans fatty acids, it seems prudent to study the ␤-oxidation of major trans fatty acids and to assess the biological effects of products that are formed by partial ␤-oxidation. Such evaluation should include conjugated linoleic acids, because they contain trans double bonds and are important constituents of the human diet due to their presence in dairy products, meat of ruminants, and partially hydrogenated vegetable oils.
The observed incomplete degradation of a fraction of elaidic acid contradicts the general conclusion that under normal conditions mitochondrial ␤-oxidation facilitates the complete breakdown of fatty acids and proceeds without the accumulation of extensive amounts of intermediates (reviewed in Ref. 5). The accumulation of substantial quantities of partially degraded fatty acids has only been observed in cases of enzyme deficiencies (27,28) or when inhibitors of ␤-oxidation enzymes or respiration were added to mitochondria (29). However, under such conditions ␤-oxidation usually is severely or completely inhibited. The prevailing view of mitochondrial ␤-oxi-FIG. 6. Acylcarnitine profiles. Acylcarnitines were identified and quantified after incubating rat liver mitochondria with (A and B) elaidoyl-CoA or (C) oleoyl-CoA plus carnitine. At the end of the incubation period, tetradecanoylcarnitine (myristoylcarnitine) labeled C14 (m/z ϭ 428.4) was added as an internal standard. Carnitine derivatives (marked with asterisks) of trideuterated straight-chain, even-numbered carboxylic acids from acetate (C2) to stearate (C18) and nonadeuterated isovalerylcarnitine were added immediately before butylation and analysis of acylcarnitines present in the incubation mixture. For experimental details see "Experimental Procedures." FIG. 7. Activities of thioesterase and CPT II in rat liver mitochondria as a function of the concentration of 5-trans-tetradecanoyl-CoA. Specific activities of thioesterase (TE) and CPT II in rat liver mitochondria were calculated based on activities measured with 20 M tetradecanoyl-CoA, and the kinetic parameters were determined with the partially purified enzyme preparations. Also plotted is the ratio of CPT II to thioesterase activities. For experimental details see "Experimental Procedures." dation is that of a system operating in a highly integrated fashion, perhaps as the result of intermediate channeling due to the intramitochondrial organization of the enzymes of ␤-oxidation. It was assumed that the accumulation of intermediates would interfere with a high flux of fatty acids through ␤-oxidation. This concern does not seem to apply to the build-up of 5-trans-tetradecenoyl-CoA and, by extension, may not be a problem when acyl-CoAs that are substrates of acyl-CoA dehydrogenases accumulate in the mitochondrial matrix. The presumed reason for avoiding an accumulation of significant amounts of intermediates in mitochondria during ␤-oxidation was their potential for inhibiting the process by inhibiting individual enzymes and/or depleting free CoA. This argument may apply to 2-enoyl-CoA, 3-hydroxyacyl-CoAs, and 3-ketoacyl-CoAs that have been shown to strongly inhibit certain enzymes of ␤-oxidation at micromolar concentrations (reviewed in Ref. 30). However, fatty acyl-CoAs that are substrates of acyl-CoA dehydrogenase seem to be less toxic. Total depletion of free CoA would compromise ␤-oxidation by preventing the formation of fatty acyl-CoAs from their carnitine derivatives by CPT II and the thiolase-catalyzed cleavage of 3-ketoacyl-CoAs in the mitochondrial matrix. However, the likelihood of such a situation developing may be small due to the regeneration of free CoA via the tricarboxylic acid cycle and by hydrolysis of acyl-CoAs. In conclusion, this study demonstrates that the accumulation of 5-trans-tetradecenoyl-CoA and possibly of other saturated fatty acyl-CoA intermediates does not reduce the effectiveness of mitochondrial ␤-oxidation. This conclusion changes our understanding of how this pathway operates and is essential for interpreting observations made when ␤-oxidation is compromised by inhibitions or deficiencies of enzymes.