Mutation for Nonsyndromic Mental Retardation in the trans-2-Enoyl-CoA Reductase TER Gene Involved in Fatty Acid Elongation Impairs the Enzyme Activity and Stability, Leading to Change in Sphingolipid Profile*

Background: The P182L mutation in the trans-2-enoyl-CoA reductase (TER) gene required for very long-chain fatty acid (VLCFA) synthesis causes nonsyndromic mental retardation. Results: This mutation reduces the activity and stability of the TER enzyme. Conclusion: The impaired TER function affects VLCFA synthesis and thereby alters the cellular sphingolipid profile. Significance: Maintenance of a proper VLCFA level may be important for neural function. Very long-chain fatty acids (VLCFAs, chain length >C20) exist in tissues throughout the body and are synthesized by repetition of the fatty acid (FA) elongation cycle composed of four successive enzymatic reactions. In mammals, the TER gene is the only gene encoding trans-2-enoyl-CoA reductase, which catalyzes the fourth reaction in the FA elongation cycle. The TER P182L mutation is the pathogenic mutation for nonsyndromic mental retardation. This mutation substitutes a leucine for a proline residue at amino acid 182 in the TER enzyme. Currently, the mechanism by which the TER P182L mutation causes nonsyndromic mental retardation is unknown. To understand the effect of this mutation on the TER enzyme and VLCFA synthesis, we have biochemically characterized the TER P182L mutant enzyme using yeast and mammalian cells transfected with the TER P182L mutant gene and analyzed the FA elongation cycle in the B-lymphoblastoid cell line with the homozygous TER P182L mutation (TERP182L/P182L B-lymphoblastoid cell line). We have found that TER P182L mutant enzyme exhibits reduced trans-2-enoyl-CoA reductase activity and protein stability, thereby impairing VLCFA synthesis and, in turn, altering the sphingolipid profile (i.e. decreased level of C24 sphingomyelin and C24 ceramide) in the TERP182L/P182L B-lymphoblastoid cell line. We have also found that in addition to the TER enzyme-catalyzed fourth reaction, the third reaction in the FA elongation cycle is affected by the TER P182L mutation. These findings provide new insight into the biochemical defects associated with this genetic mutation.

Very long-chain fatty acids (VLCFAs, chain length >C20) exist in tissues throughout the body and are synthesized by repetition of the fatty acid (FA) elongation cycle composed of four successive enzymatic reactions. In mammals, the TER gene is the only gene encoding trans-2-enoyl-CoA reductase, which catalyzes the fourth reaction in the FA elongation cycle. The TER P182L mutation is the pathogenic mutation for nonsyndromic mental retardation. This mutation substitutes a leucine for a proline residue at amino acid 182 in the TER enzyme. Currently, the mechanism by which the TER P182L mutation causes nonsyndromic mental retardation is unknown. To understand the effect of this mutation on the TER enzyme and VLCFA synthesis, we have biochemically characterized the TER P182L mutant enzyme using yeast and mammalian cells transfected with the TER P182L mutant gene and analyzed the FA elongation cycle in the B-lymphoblastoid cell line with the homozygous TER P182L mutation (TER P182L/P182L B-lymphoblastoid cell line). We have found that TER P182L mutant enzyme exhibits reduced trans-2-enoyl-CoA reductase activity and protein stability, thereby impairing VLCFA synthesis and, in turn, altering the sphingolipid profile (i.e. decreased level of C24 sphingomyelin and C24 ceramide) in the TER P182L/P182L B-lymphoblastoid cell line. We have also found that in addition to the TER enzyme-catalyzed fourth reaction, the third reaction in the FA elongation cycle is affected by the TER P182L mutation. These findings provide new insight into the biochemical defects associated with this genetic mutation.
Most of the saturated and monounsaturated VLCFAs synthesized in the ER are used for sphingolipid synthesis (3). Sphingolipids are one of the major membrane lipids in eukaryotes and consist of a ceramide backbone (i.e. a long-chain base with an amide-linked FA) and a polar head group (8). The polar head group of mammalian sphingolipids is either phosphocholine in sphingomyelin or a sugar chain in glycosphingolipids (8,9). The simplest glycosphingolipids are hexosylceramides, including glucosylceramide and galactosylceramide, the latter of which and its sulfated derivative (sulfatide) are abundant in myelin (10). Sphingolipids containing VLCFAs (mainly C24:0 and C24:1) are ubiquitous among mammalian tissues and are especially abundant in the liver, kidney, and brain (3,11,12). For example, C24 sphingomyelin constitutes ϳ60% of the total sphingomyelin in the liver (11), and C24 galactosylceramide/sulfatide constitutes ϳ70% of the total galactosylceramide/sulfatide in the spinal cord (12). Impairment of VLCFA synthesis is associated with several disorders and dysfunctions (3). For example, the dominant mutation in the ELOVL4 gene causes Stargardt disease type 3, which is a juvenile onset macular dystrophy (13), whereas the recessive ELOVL4 mutation leads to a neurocutaneous disorder of ichthyosis, seizures, mental retardation, and spasticity (14). A recessive mutation in the 3-hydroxyacyl-CoA dehydratase HACD1/PTPLA gene of the dog (15) and human (16) is known to cause myopathy. Moreover, VLCFAs have been shown to be essential for viability of yeast and mammals; for example, deletion of the TSC13 gene (the only yeast trans-2-enoyl-CoA reductase gene, i.e. the yeast homolog of the TER gene) is lethal in yeast (17), and likewise the only 3-ketoacyl-CoA reductase KAR gene knock-out mice result in embryonic lethality due to disruption of organogenesis (18).
Recently, the P182L mutation in the TER gene (TER P182L mutation) has been identified in patients with nonsyndromic mental retardation (NSMR) (19). This mutation causes a substitution of a leucine for a proline residue at residue 182 in the TER enzyme (Fig. 1B). Although the membrane topology of the mammalian TER enzyme has not been determined, it is highly likely that this enzyme is an integral membrane protein with six membrane-spanning domains and cytosolic N and C termini (Fig. 1C) as deduced from the membrane topology of the yeast and Arabidopsis TER homologs (20). In this topology model, the Pro-182 residue is positioned in the second luminal loop.
Like the KAR enzyme, which is the only mammalian 3-ketoacyl-CoA reductase involved in VLCFA synthesis and is essential for embryonic viability (18), the TER enzyme is the only isozyme of mammalian trans-2-enoyl-CoA reductase for VLCFA synthesis; therefore, the TER null mutation should also be considered to be embryonically lethal. Because the TER P182L mutation affects only mental development (19), the TER P182L mutant enzyme appears to retain some residual function. The TER P182L mutation, not being a null mutation, must be useful for understanding the function of VLCFA in various tissues. In the present study, we have investigated the effect of the P182L mutation on the activity, stability, and intracellular localization of the TER enzyme using yeast and mammalian cells transfected with the TER P182L mutant gene. We have also analyzed the FA elongation cycle in the B-lymphoblastoid cell line (B-LCL) derived from NSMR patients (homozygous for the TER P182L mutation, TER P182L/P182L ) to gain some insight into the pathogenesis of NSMR.

EXPERIMENTAL PROCEDURES
Cell Culture and Transfection-HEK 293T and HeLa cells were grown in DMEM (D6429 for HEK 293T cells and D6046 for HeLa cells; Sigma) supplemented with 10% FBS, 100 units/ml penicillin, and 100 g/ml streptomycin; HEK 293T cells were maintained in dishes coated with 0.3% collagen. Transfections were performed using Lipofectamine Plus TM reagent (Life Technologies) according to the manufacturer's instructions.
B-LCLs were generated at the University of Chicago from four NSMR patients, four carriers, and three noncarrier healthy controls. This study was approved by the Institutional Review Board at the University of Chicago. All unaffected subjects provided written informed consent; informed consents were obtained from the parents of the NSMR patients. Peripheral blood mononuclear cells were isolated from whole blood samples of each subject using Ficoll-Paque separation protocol (21). Peripheral blood mononuclear cells were then suspended in RPMI 1640 (Life Technologies) medium (supplemented with 20% fetal bovine serum and 50 g/ml gentamicin) and incubated with Epstein-Barr virus at 37°C and 5% CO 2 for 3-5 days to establish B-LCLs.
In Vitro trans-2-Enoyl-CoA Reductase Assay-The yeast membrane fraction was prepared as follows. Yeast cells suspended in buffer I (50 mM HEPES-NaOH (pH 6.8), 150 mM NaCl, 10% glycerol, 1 mM DTT, 1 mM PMSF, and a 1ϫ Complete TM protease inhibitor cocktail (EDTA-free; Roche Diagnostics)) were lysed by vigorous mixing with glass beads; unlysed cells were removed by centrifugation at 2,000 ϫ g for 3 min; the supernatant was centrifuged at 100,000 ϫ g and 4°C for 30 min; and the resulting pellet (total membrane fraction) was suspended in buffer I.
The membrane fraction of B-LCLs was prepared as follows. Cells were washed with PBS, suspended in buffer I, and lysed by sonication; cell debris was removed by centrifugation at 300 ϫ g for 5 min; the supernatant was centrifuged at 100,000 ϫ g and 4°C for 30 min; and the pellet (total membrane fraction) was suspended in buffer I.
The assay was performed by incubating the membrane fraction with 5 nCi of 3-hydroxy[1-14 C]palmitoyl-CoA (55 mCi/ mmol; American Radiolabeled Chemicals, St. Louis, MO) in 50 l of reaction buffer (buffer I containing 2 mM MgCl 2 , 1 mM CaCl 2 , and 1 mM NADPH) at 37°C for 30 min. The reaction was terminated by adding 25 l of 75% (w/v) KOH-H 2 O and 50 l of ethanol followed by heating at 70°C for 1 h, and then the mixture was acidified with 100 l of 5 M HCl in 50 l of ethanol. Lipids were extracted with 700 l of hexane and dried. The lipid residue was suspended in 20 l of chloroform and separated by TLC on a Silica Gel 60 high performance TLC plate (Merck) with hexane/diethyl ether/acetic acid (30:70:1, v/v) as the solvent system. Radiolabeled lipids were detected by autoradiography and quantified by a bioimaging analyzer BAS-2500 (Fuji Photo Film, Tokyo, Japan).
Pulse-Chase Experiment-HEK 293T cells were transfected with the pCE-puro 3ϫFLAG-TER or pCE-puro 3ϫFLAG-TER P182L plasmid. Forty-eight hours after transfection, the medium was changed to Cys/Met-and serum-free DMEM (D0422, Sigma). Cells were treated with 55 Ci of [ 35 S]Met/ [ 35 S]Cys (EXPRE 35 S 35 S protein labeling mix; PerkinElmer Life Sciences, Ontario, Canada) for 1 h. After changing the medium to normal DMEM medium (D6429, Sigma) containing 10% FCS, cells were incubated at 37°C for 1, 2, 4, and 8 h. At the end of each incubation time, cells were washed with PBS, resuspended in 1 ml of radioimmune precipitation buffer prepared in-house (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.1% sodium deoxycholate, 1 mM DTT, 1 mM PMSF, and 1ϫ Complete TM protease inhibitor cocktail), and disrupted by five passages through a 21-gauge needle. Cell debris was removed by centrifugation at 20,000 ϫ g and 4°C for 5 min, and the supernatant was incubated at 4°C overnight with anti-FLAG M2 agarose beads (Sigma). The beads were washed twice with 1 ml of radioimmune precipitation buffer and once with 1 ml of 10 mM Tris-HCl (pH 8.0), and the bound protein was eluted with 2ϫ SDS sample buffer.
Lipid Analysis Using Mass Spectrometry (MS)-Cells were spiked with an internal standard, C17:0 sphingomyelin or C17:0 ceramide (500 pmol; Avanti Polar Lipids, Alabaster, AL), and homogenized in 250 l of chloroform/methanol (1:2, v/v). Extracted lipids were recovered by centrifugation at 9,000 ϫ g and room temperature for 1 min and treated with 10 l of 4 M KOH in methanol. After incubation at 37°C for 1 h, the mixture was neutralized with 10 l of 12 M formic acid and subjected to phase separation by adding 80 l of H 2 O and 80 l of chloroform followed by centrifugation. The organic phase was recovered and dried, and the lipid residue was suspended in chloroform/methanol (1:2, v/v). Lipids were analyzed as described previously (27), using a 4000 QTRAP MS/MS system (AB Sciex, Framingham, MA) equipped with the nanoflow ion source Tri-Versa NanoMate (Advion BioSystems, Ithaca, NY). Ions for ceramide, hexosylceramide, and sphingomyelin were identified by precursor ion scans of m/z 264.4 (for ceramide and hexosylceramide) and m/z 184.1 (for sphingomyelin) in the positive ion mode and analyzed by Analyst software (version 1.6; AB Sciex).

The TER P182L Mutant Enzyme Exhibits a Reduced Activity-
To investigate the effect of the P182L mutation on the trans-2enoyl-CoA reductase activity of the TER enzyme, we first examined the ability of the TER P182L mutant enzyme to complement the growth defect of the yeast TSC13 mutant deficient in trans-2-enoyl-CoA reductase. Because the TSC13 gene is essential for the growth of yeast (17), we utilized the AID system (23) to generate a conditional yeast TSC13 mutant. Thus, the chromosomal TSC13 gene was replaced with the TSC13 gene tagged with tandemly oriented HA and AID (TSC13-HA-AID). As expected, the TSC13-HA-AID cells harboring the vector only were not able to grow in the presence of the auxin 3-indolacetic acid (IAA) ( Fig. 2A) due to the degradation of the Tsc13-HA-AID protein via the ubiquitin-proteasome pathway in response to IAA (Fig. 2B). The IAA-induced growth defect of the TSC13-HA-AID cells was complemented by the introduction of the plasmid encoding either the wild-type TER gene or the TER P182L mutant gene; however, the growth of cells expressing the mutant gene was slightly slower than those expressing the wild-type TER gene ( Fig. 2A). In addition, the cellular level of the TER P182L mutant enzyme was found to be much lower (over 5-fold) than that of the wild-type TER enzyme (Fig. 2B).
We subsequently performed an in vitro trans-2-enoyl-CoA reductase assay using the total membrane fraction from the TSC13 deletion mutant (tsc13⌬) bearing the plasmid encoding either the wild-type TER gene or the TER P182L mutant gene. In an effort to obtain a similar expression level, each mRNA was transcribed under the control of the weaker ADH or strong GAPDH promoter, respectively. As a result, the level of the mutant enzyme was increased to about half of the wild-type enzyme level (Fig. 3A). An increasing amount of each membrane fraction (i.e. an increasing amount of the 3ϫFLAGtagged wild-type/mutant enzyme) was then subjected to the assay using 3-hydroxy[1-14 C]palmitoyl-CoA, which is the substrate of 3-hydroxyacyl-CoA dehydratase in the third reaction of the FA elongation cycle (Fig. 1A), because the radiolabeled substrate of the TER enzyme, trans-2-enoyl-CoA, is not commercially available. The acyl-CoA products were analyzed after hydrolysis to the corresponding FAs. In the absence of the TER enzyme cofactor, NADPH, both membrane fractions accumulated trans-2-hexadecenoyl-CoA, some of which was however further converted to palmitoyl-CoA at the higher amount of the membrane fraction (Fig. 3B), probably due to the endogenous NADPH present in the membrane preparation. In both membrane fractions, a nearly identical amount of trans-2-enoyl-CoA was produced irrespective of the amount of the membrane fraction, suggesting either that the dehydration reaction by the endogenous yeast 3-hydroxyacyl-CoA dehydratase (Phs1) was saturated under the assay condition or that some 3-hydroxypalmitoyl-CoA was not accessible to the yeast Phs1 enzyme for an unknown reason. In the presence of the cofactor NADPH, the membrane fraction from the yeast tsc13⌬ mutant carrying the wild-type TER gene converted almost all trans-2-hexadecenoyl-CoA to palmitoyl-CoA, whereas, for the yeast tsc13⌬ mutant carrying the TER P182L mutant gene, the degree of the conversion was proportional to the amount of the membrane fraction (Fig. 3, B and C). The TER P182L mutation appears to cause a reduction in the enzyme activity.
The TER P182L Mutant Enzyme Is Unstable-It is possible that the lower cellular level of the TER P182L mutant enzyme as compared with the wild-type TER enzyme may be due to its instability; therefore, we conducted a pulse-chase experiment using HEK 293T cells expressing either the 3ϫFLAG-tagged wild-type TER mRNA or the 3ϫFLAG-tagged TER P182L mutant mRNA from the elongation factor 1␣ promoter. Like in yeast cells, the level of the mutant enzyme was lower than that of the wild-type enzyme (Fig. 4A). Cells were pulse-labeled with [ 35 S]Met/Cys, and the fate of the labeled enzyme was monitored over time (Fig. 4B). The initial amount of the labeled TER P182L mutant enzyme (1 h) was approximately equal to that of the labeled wild-type TER enzyme, indicating normal translation of the mutant enzyme. However, during the chase period of 8 h, the amount of the labeled mutant enzyme was reduced to ϳ20% of the initial level, whereas the amount of the labeled wild-type enzyme remained little changed (Fig. 4, B and C). Immunoblotting of each sample used for the pulse-chase experiment showed that the protein level (cold and radiolabeled) was fairly constant for both enzymes (Fig. 4B). These results suggest that the TER P182L mutation destabilizes the protein.
We also determined the intracellular localization of the TER P182L mutant enzyme using indirect immunofluorescence microscopy. Both the wild-type and the mutant enzymes tagged with 3ϫFLAG displayed a reticular pattern typical for the ER and were merged with the ER marker calreticulin (Fig.  5). The ER localization of the wild-type TER enzyme is consistent with the previous study (6); consequently, the TER P182L mutation does not affect the intracellular localization of the enzyme.

The TER P182L Mutation Has an Indirect Influence on the Third Reaction (Dehydration of 3-Hydroxyacyl-CoA) in the FA
Elongation Cycle-We next analyzed the FA elongation cycle in B-LCLs derived from NSMR patients (homozygous for the TER P182L mutation, TER P182L/P182L ) (19), asymptomatic carriers (heterozygous for the TER P182L mutation, TER ϩ/P182L ), and FIGURE 3. The activity of the TER P182L mutant enzyme is lower than that of the wild-type TER enzyme. A-C, yeast ABY62 cells (tsc13⌬ mutant harboring the pAB114 (3ϫFLAG-TER) plasmid) and ABY58 cells (tsc13⌬ mutant harboring the pAB110 (3ϫFLAG-TER P182L) plasmid) were grown in SC-His medium at 30°C. An increasing amount of the total membrane fraction was subjected to immunoblotting (A) or in vitro trnas-2-enoyl-CoA reductase assay (B and C): for cells expressing the 3ϫFLAG-TER enzyme, 0.625, 1.25, 2,5, 5, and 10 g, and for cells expressing the 3ϫFLAG-TER P182L mutant enzyme, 1.25, 2,5, 5, 10, and 20 g (corresponding respectively to 2.5, 5, 10, 20, and 40 ng of the 3ϫFLAG-TER enzyme as determined by comparison with the 3ϫFLAG-tagged maltose-binding protein standard, which are indicated in the figure). A, each total membrane fraction was separated by SDS-PAGE and immunoblotted with anti-FLAG antibody or, to compare protein loading, with anti-Pma1 (a plasma membrane protein) antibody. B, each total membrane fraction was incubated with 1.8 M 3-hydroxy[1-14 C]palmitoyl-CoA (5 nCi) at 37°C for 30 min. After a sequential work-up (see "Experimental Procedures"), lipids were separated by normal-phase TLC and detected by autoradiography. trans FA, trans-2-enoyl FA; 3-OH FA, 3-hydroxy FA. C, the radioactivities associated with 3-hydroxy-acyl FA, trans-2-enoyl FA, and FA in B were quantified by a bioimaging analyzer BAS-2500. Each bar represents the percentage of the radioactivity of trans-2-enoyl FA (left graph) or FA (right graph) relative to the total radioactivity and the mean Ϯ S.D. of three independent experiments. Statistically significant differences between the activities of the TER enzyme and the TER P182L mutant enzyme at the same protein amount are indicated (*, p Ͻ 0.05, **, p Ͻ 0.01; Student's t test).

FIGURE 4. The TER P182L mutant enzyme is unstable.
A, HEK 293T cells were transfected with the pCE-puro 3ϫFLAG-1 (vector), pCE-puro 3ϫFLAG-TER, or pCE-puro 3ϫFLAG-TER P182L plasmid. Forty-eight hours after transfection, the total lysate (5 g) prepared from each transfected cell population was separated by SDS-PAGE and subjected to immunoblotting with anti-FLAG antibody or, to demonstrate uniform protein loading, with anti-␣-tubulin antibody. B, HEK 293T cells were transfected with the pCE-puro 3ϫFLAG-TER or pCE-puro 3ϫFLAG-TER P182L plasmid. Forty-eight hours after transfection, cells were labeled with [ 35 S]Met/Cys at 37°C for 1 h. After changing the medium to remove [ 35 S]Met/Cys, cells were incubated for another 1, 2, 4, and 8 h. At the end of each incubation time, the total cell lysate was prepared and subjected to immunoprecipitation with anti-FLAG antibody. The precipitate was separated by SDS-PAGE and analyzed by autoradiography and by immunoblotting with anti-FLAG antibody. C, the radioactivities associated with the TER enzyme and the TER P182L mutant enzyme in B were quantified by a bioimaging analyzer BAS-2500. Each value represents the percentage of the radioactivity of the TER enzyme or the TER P182L mutant enzyme relative to its radioactivity at the 1-h chase point and the mean Ϯ S.D. of three independent experiments. Statistically significant differences between the radioactivities of the TER enzyme and the TER P182L mutant enzyme at the same chase point are indicated (*, p Ͻ 0.05, **, p Ͻ 0.01; Student's t test). DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 JOURNAL OF BIOLOGICAL CHEMISTRY 36745 healthy noncarrier controls (TER ϩ/ϩ ). The RT-PCR indicated that the expression of the TER mRNA was comparable among these B-LCLs (Fig. 6A). The total membrane fraction of each B-LCL was incubated with 3-hydroxy[1-14 C]palmitoyl-CoA, and the acyl-CoA products were analyzed after conversion to the corresponding FAs (Fig. 6, B and C). As expected, in the absence of the cofactor NADPH, all the membrane fractions yielded trans-2-enoyl-CoA as the only product in ϳ30% radiochemical yield. In the presence of the cofactor, the TER ϩ/ϩ and TER ϩ/P182L membrane fractions consumed most of trans-2enoyl-CoA, producing acyl-CoA in 72 and 61% radiochemical yield, respectively. The TER P182L/P182L membrane fraction, on the other hand, generated only 28% radiochemical yield of acyl-CoA (39% of acyl-CoA produced by the TER ϩ/ϩ membrane fraction) with a significant amount of unreacted trans-2enoyl-CoA.

Reduced Activity and Stability of TER P182L Mutant Enzyme
Because FA elongation is achieved by a repeated cycle of four successive enzyme-catalyzed reactions (Fig. 1A), we performed an in vitro FA elongation assay to investigate the effect of the TER P182L mutation on the overall FA elongation cycle. The total membrane fraction from each B-LCL was incubated with palmitoyl-CoA (the acyl-CoA substrate) and [2-14 C]malonyl-CoA (the 2-carbon donor), and the radioactive acyl-CoA products were analyzed as the corresponding radioactive FAs (Fig.  7A). Because the first condensation reaction is the rate-limiting step in each round of the FA elongation cycle (4), acyl-CoA was the primary product in all the membrane fractions. In both the TER ϩ/ϩ and the TER ϩ/P182L membrane fractions, 3-hydroxyacyl-CoA was also detected in a small amount, but neither 3-ketoacyl-CoA nor trans-2-enoyl-CoA was detected. Contrary to our expectation, the TER P182L/P182L membrane fraction accumulated a substantial amount of 3-hydroxyacyl-CoA with only a small amount of trans-2-enoyl-CoA (the substrate of the TER enzyme). Similar results were also obtained using HeLa cells with TER knockdown (supplemental Fig. 1). The TER siRNA caused a reduction in the TER mRNA level (supplemental Fig.  1A), accompanied by a significant accumulation of 3-hydroxya-cyl-CoA and, to a lesser extent, trans-2-enoyl-CoA (supplemental Fig. 1, B and C). Therefore, it seems that the third reaction in the FA elongation cycle was indirectly affected by the impairment of the fourth reaction caused by the reduced enzymatic activity of the TER P182L mutant enzyme.
The Levels of C24:1 Sphingomyelin and C24:1 Ceramide Are Decreased in the TER P182L/P182L B-LCL-Because VLCFAs are components of sphingolipids, the decreased activity of the TER P182L mutant enzyme leads to a reduction in VLCFA synthesis,   A and B). A, the TER and GAPDH cDNAs were amplified by RT-PCR from the total RNA prepared from each B-LCL clone using a specific primer. The amplified fragments were separated by agarose gel electrophoresis and stained with ethidium bromide. PL, P182L. B, the total membrane fraction (20 g) prepared from each B-LCL clone was incubated with 1.8 M 3-hydroxy[1-14 C]palmitoyl-CoA (5 nCi) in the presence or absence of 1 mM NADPH at 37°C for 30 min. After a sequential work-up (see "Experimental Procedures"), lipids were separated by normal-phase TLC and detected by autoradiography. trans FA, trans-2-enoyl FA; 3-OH FA, 3-hydroxy FA. C, the radioactivities associated with 3-hydroxy-acyl FA, trans-2-enoyl FA, and FA in B were quantified by a bioimaging analyzer BAS-2500. Each bar represents the percentage of the radioactivity of trans-2-enoyl FA or FA relative to the total radioactivity and the mean Ϯ S.D. of three clones. Statistically significant differences are indicated (*, p Ͻ 0.05, **, p Ͻ 0.01; Student's t test).
which may alter the sphingolipid profile in the cell. To test this possibility, we quantitatively analyzed sphingomyelin, ceramide, and hexosylceramide compositions in the TER ϩ/ϩ , TER ϩ/P182L , and TER P182L/P182L B-LCLs using MS/MS. In all the B-LCLs, C16:0 sphingolipids were the most abundant sphingolipid species followed by C24:1 and C24:0 sphingolipids (Fig. 8). However, in the TER P182L/P182L B-LCL, the levels of C24:1 sphingomyelin and C24:1 ceramide were significantly reduced as compared with those in the TER ϩ/ϩ and TER ϩ/P182L B-LCLs (Fig. 8, A and B). In addition, the levels of C24:0 sphingomyelin and C24:0 ceramide also seemed to be decreased but not statistically significant. Interestingly, there was a concomitant increase in the levels of C16:0 sphingomyelin and C16:0 ceramide. In contrast to sphingomyelin and ceramide, although C16:0 hexosylceramide was significantly increased, no other hexosylceramide showed any noticeable changes in the TER P182L/P182L B-LCL (Fig. 8C). Nevertheless, these findings indicated that the TER P182L mutation also has an influence on the synthesis of sphingolipids.

DISCUSSION
Using yeast and mammalian cells transfected with the TER P182L mutant gene, we have revealed that the TER P182L mutation reduces the activity and stability of the enzyme (Figs.  3 and 4), which seems to explain the decrease in the apparent TER activity observed in the TER P182L/P182L B-LCL (Fig. 6, B and C). In the predicted membrane topology model of the TER enzyme (Fig. 1C), the Pro-182 residue is located in the second luminal loop. Due to its cyclic structure, a proline residue confers a local rigidity to the protein backbone and often plays an important role in protein architecture, producing turns and bends. Indeed, the Pro-182 residue is predicted to constitute a turn structure at the end of a ␤-strand by a protein secondary structure prediction algorithm (28); accordingly, its substitution with a leucine residue could impair the turn and dislocate the secondary structure, resulting in the reduced enzyme activity and stability.  Each FA elongation cycle is a four-reaction process (Fig. 1A) in which the first condensation reaction is the rate-limiting step (4). This is consistent with the results of the FA elongation assay on the B-LCLs showing that the major product was the substrate for the first reaction, acyl-CoA (Fig. 7). Interestingly, in the TER P182L/P182L B-LCL and the TER siRNA-treated HeLa cells, the substrate for the third reaction, 3-hydroxyacyl-CoA, was the second major product followed by the substrate for the fourth reaction (the TER enzyme-catalyzed reaction), trans-2enoyl-CoA ( Fig. 7 and supplemental Fig. 1). A small amount of 3-hydroxyacyl-CoA, but not trans-2-enoyl-CoA, was also detected in both the TER ϩ/ϩ and the TER ϩ/P182L B-LCLs (Fig.  7A). The substrate for the second reaction, 3-ketoacyl-CoA, was not detected in any of the B-LCLs. This observation implies that the third reaction, dehydration of 3-hydroxyacyl-CoA to trans-2-enoyl-CoA, may be the secondary rate-limiting step in the FA elongation cycle and regulated by feedback inhibition of 3-hydroxyacyl-CoA dehydratase by its product accumulation due to the TER P182L mutation. It is noteworthy that, in mammals, multiple isozymes exist for both the first and the third reactions, in contrast to one enzyme each for the second and fourth reactions, suggesting the importance of the first and third reactions in the FA elongation cycle. Alternatively, because the enzymes involved in the FA elongation cycle are associated as a membrane-bound enzyme complex, any reduction in the TER enzyme level caused by either the P182L mutation or siRNA treatment could result in the disruption of the possible interaction of 3-hydroxyacyl-CoA dehydratase with the TER enzyme required for the dehydratase activity and thus the accumulation of its substrate 3-hydroxyacyl-CoA.
The levels of C24:1 sphingomyelin and C24:1 ceramide were significantly decreased in the TER P182L/P182L B-LCL as compared with the TER ϩ/ϩ B-LCL (Fig. 8, A and B). There was also a decrease in the levels of C24:0 sphingomyelin and C24:0 ceramide, but this was not statistically significant, probably due to the small sample size. Ceramide is the common precursor for hexosylceramide and sphingomyelin, and therefore, it was expected that, similar to C24:1 sphingomyelin, the level of C24:1 hexosylceramide would be decreased in the TER P182L/P182L B-LCL; however, it was found to remain virtually unchanged (Fig. 8C). This observation may be explained by the low hexosylceramide content in B-LCL (ϳ0.1 of the sphingomyelin content; data not shown), which could cause measurement deviation. On the other hand, the levels of all three C16:0 sphingolipids were found to be similarly increased in the TER P182L/P182L B-LCL (Fig. 8). Because the TER enzyme is the only trans-2-enoyl-CoA reductase, the mutation in the TER gene must affect every FA elongation cycle; as a consequence, the effect of the TER P182L mutation may become more profound with a progression of the FA elongation cycle (e.g. synthesis of C24 VLCFAs). It is therefore conceivable that, in the TER P182L/P182L B-LCL, the mutation causes a reduction in the production of C24 VLCFAs, which in turn lowers the level of C24 sphingolipids.
In contrast to the TER null mutation, which probably leads to embryonic lethality as in the case of the knock-out mice for the only 3-ketoacyl-CoA reductase KAR gene (18), the TER P182L mutation appears to be a weak mutation and affects only neural function in the brain (19). This weak mutation could be con-sistent with the relatively small changes observed in the level of C24 sphingolipids, which, however, may be enough to harm the nerve system without causing damage to other tissues/organ systems. Myelin plays an important role in the saltatory conduction of action potentials in the nervous system and is enriched in certain lipids, especially C24 sphingolipids (sphingomyelin, galactosylceramide, and sulfatide) (10,12,29). Indeed, mice deficient in C24 ceramide (and hence C24 sphingolipids) are known to exhibit neural defects (30). In this regard, it may be possible that the reduction in C24 sphingolipids as a result of the TER P182L mutation could affect neural development, contributing to the pathogenesis of NSMR.
In summary, we have identified that the TER P182L mutation reduces the activity and stability of the enzyme, leading to a decreased level of C24 sphingolipids, an essential component for proper myelin function, in the TER P182L/P182L B-LCL, which may be a part of the pathogenesis of NSMR. We have also found an indication that the third reaction (dehydration of 3-hydroxyacyl-CoA) is likely to be the secondary rate-limiting step in the FA elongation cycle. Future studies will be needed to determine the effect of the TER P182L mutation on brain lipid composition and brain function as well as to reveal the regulatory mechanism of the probable secondary rate-limiting step.