Molecular Identification and Characterization of Two Medium-chain Acyl-CoA Synthetases, MACS1 and the Sa Gene Product*

In this study, we identified and characterized two murine cDNAs encoding medium-chain acyl-CoA synthetase (MACS). One, designated MACS1, is a novel protein and the other the product of the Sa gene (Sa protein), which is preferentially expressed in spontaneously hypertensive rats. Based on the murine MACS1 sequence, we also identified the location and organization of the human MACS1 gene, showing that the human MACS1 and Sa genes are located in the opposite transcriptional direction within a 150-kilobase region on chromosome 16p13.1. Murine MACS1 and Sa protein were overexpressed in COS cells, purified to homogeneity, and characterized. Among C4–C16 fatty acids, MACS1 preferentially utilizes octanoate, whereas isobutyrate is the most preferred fatty acid among C2–C6 fatty acids for Sa protein. Like Sa gene transcript, MACS1 mRNA was detected mainly in the liver and kidney. Subcellular fractionation revealed that both MACS1 and Sa protein are localized in the mitochondrial matrix. 14C-Fatty acid incorporation studies indicated that acyl-CoAs produced by MACS1 and Sa protein are utilized mainly for oxidation.

atively large amounts of medium-chain fatty acid occur in the milk and urine of patients with medium-chain acyl-CoA deficiency, but the metabolism of these fatty acids, utilized by medium-chain acyl-CoA synthetase (MACS), is poorly understood.
The Sa gene is a proposed candidate gene for essential hypertension that was isolated by virtue of its increased expression in the kidney of spontaneously hypertensive rat compared with the kidney of normotensive Wistar-Kyoto rat (10). The Sa gene is highly expressed in the hypertensive kidney (11,12) and its increased expression correlates with increased blood pressure in both spontaneously hypertensive (13)(14)(15)(16) and Dahl hypertensive (17)(18)(19) rat models. However, although recent congenic substitution mapping studies have argued that the Sa gene is a candidate gene locus for hypertension (20), the function of the Sa gene remains unclarified. Based on the amino acid sequence similarity of Sa protein with the ACS/AceCS family (9), we hypothesized that the Sa gene encodes a MACS.
We therefore isolated a murine cDNA for Sa protein, and its expression, purification, and characterization identified it as a MACS. We also isolated and characterized a novel MACS, designated MACS1, using Sa protein cDNA as a probe. In this paper, we describe structural and functional differences, tissue and subcellular distributions, and the gene organization of these two types of MACS.

EXPERIMENTAL PROCEDURES
General Procedures-Standard molecular biology and immunochemical techniques were performed essentially as described by Sambrook and Russell (21) and Harlow and Lane (22), respectively. Nucleotide sequences were determined by the dideoxy chain-termination method with a Dye-Terminator Cycle Sequencing Ready Reaction FS kit (PE Biosystems) and a DNA sequencer (model 310, PE Biosystems). Total RNA was prepared using standard guanidinium thiocyanate lysis buffer and centrifugation over a cesium chloride cushion. To analyze RNA in murine tissues, commercially available Northern blots (CLONTECH, Palo Alto, CA) were used for Northern blot analysis. 32 P-Labeled probes were prepared by priming with random hexanucleotides. For immunoblotting, polyclonal antibodies against murine MACS1 and Sa protein were used as first antibodies. Antibody binding was detected with a chemiluminescence kit (ECL, Amersham Pharmacia Biotech). DNA transfection, subcellular fractionation, and 14 C-fatty acid incorporation studies were performed as described previously (7).
cDNA Cloning-A near full-length cDNA for murine Sa protein was isolated from a murine kidney cDNA library (7) using a murine Sa protein-specific oligonucleotide, 5Ј-CAACACTGATGCTAGCTAGCT-3Ј. A near full-length cDNA for MACS1 was also obtained from the kidney cDNA library using Sa protein cDNA as a probe under reduced hybridization conditions. Human MACS1 cDNAs were obtained by 5Ј-and 3Ј-rapid amplification of cDNA end (RACE) with human MACS1specific primers using a SMART RACE cDNA amplification kit (CLONTECH).
Assay of MACS Activity-Enzyme activities were determined at 37°C by either an isotopic method or by the spectrophotometric method as described previously (7). The latter was used only for the purified enzyme. The standard reaction mixture for the isotopic method contained 100 mM Tris-HCl, pH 8.0, 10 mM MgCl 2 , 10 mM ATP, 1 mM CoA, and 10 mM 14 C-labeled (ϳ1000 cpm/nmol) isobutyrate (for Sa protein) or octanoate (for MACS1) in a total volume of 0.2 ml: octanoate was dissolved in Triton X-100 (final concentration 1% w/v). After 1 min of preincubation at 37°C, the reactions were initiated by the addition of enzyme solution. The reaction was terminated by adding 50 l of 30% (w/v) metaphosphoric acid. The reaction product ([ 14 C]acyl-CoA) was isolated as described previously (7). All assays were carried out within the range where the reaction proceeded linearly with time and the initial rate of reaction was proportional to the amount of enzyme added. The protein content was determined by the Lowry method (23).
Purification of MACS1 and Sa Protein-All purification steps were carried out at 0 -4°C. The results of purification are summarized in Table I. COS-7 cells (80 dishes, 100-mm diameter) transfected with either murine MACS1 or Sa protein were suspended in 30 ml of 20 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, 1 g/ml aprotinin, 1 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride, and disrupted by sonication. The cell extracts were centrifuged at 230,000 ϫ g for 1 h, and the supernatant was used for purification.
For purification of MACS1, the following purification steps were carried out. Protein in the 230,000 ϫ g supernatant was precipitated with (NH 4 ) 2 SO 4 at 70% saturation and centrifuged at 16,000 ϫ g for 15 min. The resulting precipitate was suspended in buffer A (50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, and 10% glycerol (w/v)), dialyzed against buffer A and applied to a DEAE-Sepharose FF (Amersham Pharmacia Biotech) column (1.5 ϫ 5 cm) equilibrated with buffer A. After washing with buffer A, the column was eluted with a linear gradient of 0 to 0.25 M KCl in buffer A. Active fractions were combined, dialyzed against buffer A, and then applied to a Butyl-Sepharose FF (Amersham Pharmacia Biotech) column (1.5 ϫ 5 cm) equilibrated with buffer A containing 30% (NH 4 ) 2 SO 4 . After washing with buffer A containing 30% (NH 4 ) 2 SO 4 , the column was eluted with a decreasing liner gradient of 30 to 0% (NH 4 ) 2 SO 4 in buffer A. Active fractions were collected and dialyzed against buffer A and then applied to a Blue-Sepharose CL-6B (Amersham Pharmacia Biotech) column (1.0 ϫ 3 cm) equilibrated with buffer A. The column was washed with buffer A and eluted with an increasing liner gradient of 0 to 1 M KCl in buffer A. The fractions exhibiting enzyme activities were combined, concentrated by ultrafiltration with Amicon Centriplus 10 (Amicon Inc.), and applied to a Sephacryl S-200 HR column (1.6 ϫ 60 cm). The column was eluted with 150 ml of buffer A containing 150 mM KCl. Active fractions were collected, dialyzed against buffer A, and stored at Ϫ80°C.
For purification of Sa protein, a similar procedure was followed except that DEAE-Sepharose FF, Butyl-Sepharose FF, and Sephacryl S-200 HR were replaced by Red-Toyopearl (Toso, Tokyo), Phenyl-Sepharose FF, and Butyl-Sepharose FF, respectively. After (NH 4 ) 2 SO 4 precipitation, the enzyme was dialyzed against buffer A and applied to a Red-Toyopearl column (1.5 ϫ 2 cm) equilibrated with buffer A. The column was washed with buffer A and eluted with an increasing linear gradient of 0 to 0.25 M KCl containing buffer A. Active fractions were combined, dialyzed against buffer A, and applied to a Phenyl-Sepharose (Amersham Pharmacia Biotech) column (1.5 ϫ 5 cm) equilibrated with buffer A containing 30% (NH 4 ) 2 SO 4 . After washing with buffer A containing 30% (NH 4 ) 2 SO 4 , the column was eluted with a decreasing gradient from 30 to 0% (NH 4 ) 2 SO 4 in buffer A. Other column chromatography steps were carried out as described above. The purified enzyme from the Butyl-Sepharose FF step was dialyzed against buffer A and stored at Ϫ80°C.
Antibodies-To obtain anti-murine MACS1 and Sa protein antibodies, two 15-residue peptides corresponding to the C termini of murine MACS1 (CIKRKELRNKEFGQL) and Sa protein (CVKRNEL-RKKEWVTT) were synthesized by Nippon Gene Research Laboratories (Sendai, Japan). The amino acid composition and sequences were confirmed by the supplier. The peptides were independently coupled to keyhole limpet hemocyanin and injected into New Zealand White rabbits as described previously (3). IgG fractions were prepared by affinity chromatography on protein-A-Sepharose (Amersham Pharmacia Biotech).

Comparison of Amino Acid Sequences of MACS1 and Sa
Protein and Their Human Genes-To identify and characterize murine MACS, we initially isolated a murine cDNA for Sa protein, because (i) ϳ30% of the amino acids in Sa protein are identical to the previously characterized AceCSs and (ii) it contains a sequence motif common to the ACS/AceCS family. Transfection of Sa protein cDNA into COS cells resulted in the induction of MACS activity (see below), indicating that the Sa gene encodes an MACS. Using the Sa protein cDNA as a probe, we isolated a cDNA encoding a novel MACS, designated MACS1, from a murine kidney cDNA library. BLAST searches of the GenBank data bases using the murine MACS1 sequence identified a human expressed sequence tag (GenBank accession number AK025451) encoding a part of human MACS1. 5Ј-and 3Ј-RACE with human kidney RNA-and human MACS1-specific primers revealed the nucleotide and protein sequences of human MACS1. Fig. 1A shows the deduced primary amino acid sequences of murine and human MACS1 compared with those of Sa protein of the two species. There is ϳ53% amino acid identity between murine MACS1 and Sa protein, and the phylogenetic tree of ACSs/AceCSs of various origins indicates that MACS1 and Sa protein belong to the same group (Fig. 1B). Human and murine MACS1 protein consist of 577 and 573 amino acids with calculated molecular weights of 65,258 and 64,759, respectively. Sequence analysis by the PSORT program (24) predicted that MACS1 and Sa protein contain a potential mitochondrial targeting signal at the N terminus (Fig. 1A). We also found two proteins in the GenBank, BAA91273 and rat kidney specific (rKS) protein (25), belonging to the MACS group, suggesting that these proteins also function as MACS.
The BLAST search also identified a bacterial artificial chromosome clone encoding the human Sa gene (CI987SK-44M2) and two overlapping clones (CTD-2152G12 and CIT987K-A-923A4) containing the human MACS1 gene. Comparison of the cDNA sequences and the human genome draft sequence revealed the genomic organization of the human MACS1 gene (Fig. 1, C and D). As shown in Fig. 1C, the human MACS1 and Sa genes are located in the opposite transcriptional direction within a 150-kb region on chromosome 16p13.1. The two genes consist of 13 exons, and their exon/intron boundaries are almost identical (Fig. 1D). The relevant sequence information is available under GenBank accession numbers AB062491 through AB062503.
Purification of MACS1 and Sa Protein-To compare enzymatic properties, cDNAs for murine MACS1 and Sa protein were individually introduced into COS cells, and the resulting enzymes were purified to homogeneity using the 230,000 ϫ g supernatant fractions of the sonicated extracts. The overall purification and yields of MACS1 and Sa protein were 15.3-fold with a yield of 9.7% and 18.7-fold with a yield 17%, respectively ( Table I). The specific activities of the purified MACS1 and Sa protein were, respectively, 15.8 and 11.5 mol/min/mg when assayed with octanoate and isobutyrate, respectively.
The two purified enzymes were essentially homogenous when analyzed by SDS-PAGE (Fig. 2). The molecular masses of the purified MACS1 and Sa protein estimated by SDS-PAGE were ϳ62 kDa: ϳ3 kDa smaller than the calculated molecular masses deduced from the cDNAs. This suggests that the putative mitochondrial targeting signals are cleaved during the transportation of the enzymes into the mitochondria matrix (see below).
Enzymatic Properties-Using the purified enzymes, the fatty acid chain-length preferences of MACS1 and Sa protein were determined. As shown in Fig. 3 (A and B), octanoate is the preferred substrate for MACS1 among fatty acids with 4 -16 carbon atoms. Sa protein showed a preference for isobutyrate among fatty acids with 2-6 carbon atoms (Fig. 3, C and D): the calculated K m value for isobutyrate was 0.05 mM. Both MACS1 and Sa protein require ATP and CoA for their activities (data not shown). When adenylate kinase was omitted from the reaction mixture for the spectrophotometric assay, no oxidation of NADH occurred, indicating that AMP was a reaction product (data not shown).
Tissue Expression and Regulation-Northern blotting of RNA from various mouse tissues revealed hybridization to major transcripts of 2.3 kb for MACS1, expressed abundantly in the liver and kidney (Fig. 4, top panel). A minor transcript of 4.4 kb was also detected in the liver and kidney. This pattern of expression was very similar to that of Sa gene transcripts (middle panel). In contrast to the preferential expression of Sa gene transcripts in spontaneously hypertensive rats (13)(14)(15)(16), the levels of MACS1 transcripts were not increased in transcripts in these hypertensive rats (data not shown). We also analyzed dietary regulation of MACS1 and Sa mRNAs in the liver by Northern blotting. Neither fasting for 48 h, nor refeeding a high carbohydrate diet after 48 h fasting had any effects on the levels of the two mRNAs in the liver (data not shown).
Subcellular Localization-To determine the subcellular distributions of MACS1 and Sa protein, a subcellular fractionation was carried out on mouse kidney homogenate. The tissue homogenate was fractionated into light and heavy mitochondrial, peroxisomal, microsomal, and cytosolic fractions. Among these fractions, anti-MACS1 antibody detected an immunoreactive band of 62 kDa present mainly in the light and heavy mitochondrial fractions (Fig. 5, upper panel). Similarly, a 62-kDa protein detected by anti-Sa protein antibody was present mainly in the light and heavy mitochondrial fractions (Fig. 5,  middle panel). Together with the solubility in hypotonic buffer of MACS1 and Sa protein in transfected COS cells, the cell fractionation data indicate that both MACS1 and Sa protein are localized in the mitochondrial matrix.
Incorporation of 14 (Fig. 6B). In both MACS1-and Sa protein-transfected cells, induction of 14 C incorporation into CO 2 occurred, indicating the 14 C-fatty acids were degraded by oxidation.
These data suggest that the major function of MACS1 and Sa protein is to produce acyl-CoA for oxidation to produce ATP and CO 2 in the mitochondrial matrix. DISCUSSION In the current study, we have identified and characterized two MACSs, designated MACS1 and Sa protein. Both enzymes have a putative mitochondrial targeting signal at the N terminus of their primary amino acid sequence and are localized in

FIG. 3. Substrate specificities and kinetic properties of purified MACS1 and Sa protein.
A and C, the enzyme activity was determined by the spectrophotometric method as described under "Experimental Procedures" with 1 g of the purified enzyme and the standard reaction mixture, except that the indicated fatty acid (final concentration ϭ 1 mM) was used. For measurement of MACS1 activities, Triton X-100 (final 1% w/v) was added in the standard reaction mixture. Enzyme activity is expressed as a percentage of that obtained with octanoate (A, MACS1) or isobutyrate (C, Sa protein). The specific activities of the purified MACS1 (assayed with octanoate) and Sa protein (assayed with isobutyrate) were 15.8 Ϯ 0.16 and 11.5 Ϯ 0.16 mol/min/mg, respectively. The data are means Ϯ S.D. of triplicate determinations. Isobutyrate, 2-methyl butyrate, and isovalerate are labeled isoC4, 2MeC4, and IsoC5, respectively. B and D, Assays were performed with the indicated concentrations of hexanoate and octanoate (B), and butyrate and isobutyrate (D). Data points represent the mean of three determinations obtained in one experiment (S.D. Ͻ 5%). The values for apparent K m were calculated from Lineweaver-Burk plots.
the mitochondrial matrix. A 14 C-fatty acid incorporation study suggested that they play a similar role in the degradation of medium-chain fatty acids for the production of energy. Although the specificities of the two enzymes toward mediumchain fatty acids differ completely, the tissue expression of their transcripts and exon/intron organization of their human genes are very similar. Furthermore, the human genes for the two proteins consist of similar exon/intron organization and are located close together in the same region on chromosome 16p13.1, although in opposite transcriptional directions. These data may suggest that the two proteins arose from a common ancestor by gene duplication.
The Sa gene has been characterized as a proposed candidate gene for essential hypertension (13)(14)(15)(16)(17)(18)(19). Although congenic substitution mapping argued for the Sa gene locus as a candidate for the blood pressure quantitative trait locus (20), the function of the Sa gene product was unclear until now. In this study, we have shown that Sa protein is a unique isobutyratepreferring MACS. The calculated K m value of Sa protein toward isobutyrate is markedly low (Fig. 3D), indicating that Sa protein preferentially utilizes isobutyrate in vivo. Although isobutyrate is generated from (S)-isobutyryl-CoA during the catabolism of valine, little is known about its metabolic fate. The abundant expression of Sa protein in the mitochondrial matrix of the liver and kidney suggest the importance of this enzyme in the metabolism of valine-derived isobutyrate for oxidative degradation.
In contrast to Sa protein, octanoate is the most preferred fatty acid for MACS1 among C4 -C16 fatty acids. Mediumchain fatty acids, including octanoate occur in the liver as a result of chain shortening, by the peroxisomal ␤-oxidation system from long-chain and very long-chain fatty acids (26). These medium-chain fatty acids are believed to undergo further degradation via the mitochondrial ␤-oxidation system after transportation into the mitochondrial matrix. Based on the fatty acid specificity and subcellular distribution of MACS1, it is suggested that MACS1 initiates the degradation of mediumchain fatty acids generated by the peroxisomal ␤-oxidation system from long-chain and very long-chain fatty acids. MACS is also believed to play a role in the conjugation of a series of benzoic acid derivatives with glycine (27). Kasuya et al. (28) have purified a MACS from murine kidney mitochondria, showing that the purified enzyme effectively utilizes not only medium-chain fatty acids but also benzoic acid derivatives. Although purified MACS1 utilizes benzoic acid (Fig. 3A), it had almost no preference toward salicylic acid (data not shown), suggesting the presence of other types of mitochondrial MACS, which play a role in the glycine conjugation of xenobiotic carboxylic acids.
Further studies are necessary to elucidate the precise function and regulation of Sa protein and MACS1, and to determine the disorders caused by the absence of the two enzymes. RNA from the indicated murine tissues was probed with 32 P-labeled murine MACS1 (top panel) and Sa protein (middle panel). The filters were exposed to Kodak XAR-5 film with an intensifying screen at Ϫ80°C for 16 h. The same samples were subsequently hybridized with a control probe for mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH; bottom panel) and exposed to Kodak XAR-5 film with an intensifying screen at Ϫ80°C for 6 h.

FIG. 5. Subcellular localization of MACS1 and Sa protein.
A post-nuclear fraction prepared from murine kidneys was further processed into the indicated fractions. 10 g of protein from the indicated fractions was subjected to SDS-PAGE, blotted onto a nitrocellulose membrane, and detected by anti-MACS1 and anti-Sa protein antibodies. Immunoblots with specific antibodies for AceCS1 (a cytosolic enzyme (7)), NADPH P-450 reductase (a microsome marker), prohibitin (a mitochondria marker), and catalase (a peroxisomal marker) were carried out as controls.