The Peroxisomal Acyl-CoA Thioesterase Pte1p from Saccharomyces cerevisiae Is Required for Efficient Degradation of Short Straight Chain and Branched Chain Fatty Acids*

The role of the Saccharomyces cerevisae peroxisomal acyl-coenzyme A (acyl-CoA) thioesterase (Pte1p) in fatty acid β-oxidation was studied by analyzing the in vitro kinetic activity of the purified protein as well as by measuring the carbon flux through the β-oxidation cycle in vivo using the synthesis of peroxisomal polyhydroxyalkanoate (PHA) from the polymerization of the 3-hydroxyacyl-CoAs as a marker. The amount of PHA synthesized from the degradation of 10-cis-heptadecenoic, tridecanoic, undecanoic, or nonanoic acids was equivalent or slightly reduced in the pte1Δ strain compared with wild type. In contrast, a strong reduction in PHA synthesized from heptanoic acid and 8-methyl-nonanoic acid was observed for the pte1Δ strain compared with wild type. The poor catabolism of 8-methyl-nonanoic acid via β-oxidation in pte1Δ negatively impacted the degradation of 10-cis-heptadecenoic acid and reduced the ability of the cells to efficiently grow in medium containing such fatty acids. An increase in the proportion of the short chain 3-hydroxyacid monomers was observed in PHA synthesized in pte1Δ cells grown on a variety of fatty acids, indicating a reduction in the metabolism of short chain acyl-CoAs in these cells. A purified histidine-tagged Pte1p showed high activity toward short and medium chain length acyl-CoAs, including butyryl-CoA, decanoyl-CoA and 8-methyl-nonanoyl-CoA. The kinetic parameters measured for the purified Pte1p fit well with the implication of this enzyme in the efficient metabolism of short straight and branched chain fatty acyl-CoAs by the β-oxidation cycle.

Degradation of fatty acids is mediated by the ␤-oxidation cycle, which is located exclusively in the peroxisome in Saccharomyces cerevisae. Numerous enzymes involved in the degradation of both saturated and unsaturated fatty acids have been identified and characterized. These include the three core enzymes of the ␤-oxidation cycle, namely the acyl-CoA oxidase Pox1p, the multifunctional enzyme Fox2p, harboring both an enoyl-CoA hydratase II activity and an (R)-3-hydroxyacyl-CoA dehydrogenase activity, and the 3-ketothiolase Fox3p. Furthermore, the roles of the auxiliary enzymes enoyl-CoA isomerase Eci1p, dienoyl-CoA isomerase Dci1p, and 2,4-dienoyl-CoA reductase Sps19p in the degradation of unsaturated fatty acids have also been defined (1).
YJR019C (PTE1, TES1) has been identified in the S. cerevisiae genome as a gene encoding an acyl-CoA thioesterase (2,3). Acyl-CoA thioesterases are a group of enzymes catalyzing the hydrolysis of acyl-CoAs to fatty acids and unesterified CoA (CoASH). 3 Two types of acyl-CoA thioesterases have been found in Escherichia coli (4). The acyl-CoA thioesterase II (TesB) of E. coli is a tetramer with a molecular mass of 120 kDa and cleaves C6 -C18 acyl-CoA esters as well as 3-hydroxyacyl-CoA esters (5). TesB-like acyl-CoA thioesterases have also been found in eukaryotes, including yeast, mammals, and plants (2,3,6,7). The S. cerevisiae Pte1p is a TesB-like acyl-CoA thioesterase and contains the Asp, Ser/Thr, and Gln catalytic triad expected for thioesterases. The protein contains a peroxisomal targeting signal type I at the carboxylterminal end and was shown to be located in the peroxisome (2,3). Amino acid sequences of acyl-CoA thioesterases belonging to the TesB family show good conservation among the animal, plant, fungi, and bacterial kingdoms (7). The existence of various acyl-CoA thioesterases having activities toward either a narrow or broad range of substrates could provide important control points in the oxidation of many peroxisomal substrates as well as regulate intraperoxisomal levels of CoA esters and CoASH.
Although it seems likely that the yeast Pte1p would be implicated in some aspect of fatty acid ␤-oxidation, its physiological function and enzymatic properties have not yet been clearly defined. Although one study has shown a reduction in the capacity of the pte1⌬ mutant to grow in media containing oleic acid as the main carbon source, a separate study does not reveal any significant differences in phenotype between wild type and pte1⌬ mutant (2,3). In this work, we have studied the role of the Pte1p protein in peroxisomal ␤-oxidation by analyzing the kinetic activity of the purified protein in vitro, as well as by measuring the carbon flux through the ␤-oxidation cycle in vivo using the synthesis of peroxisomal polyhydroxyalkanoate (PHA) from the polymerization of the 3-hydroxyacyl-CoAs as a marker.
The plasmid Yiplac128-PHA containing the gene for the PHAC1 synthase from Pseudomonas aeruginosa, modified at the carboxyl end by the addition of a peroxisomal targeting sequence and placed under * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Recipient of a fellowship of the exchange program between the Fonds National Suisse de la Recherche Scientifique (FNRS) and the Japan Society for the Promotion of Science (83JS-067392). 2  the control of the catalase A (CTA1) promoter and terminator, has been previously described (8). The plasmid Yiplac128-PHA was transformed into the wild type BY4742 as well as the isogenic pte1⌬ by the lithium acetate procedure (9). S. cerevisiae strains harboring the PHA synthase gene were maintained in leucine-deficient medium (0.67% yeast nitrogen base without amino acids (Difco), 0.5% ammonium sulfate, 2% glucose, and 0.69 g/liter of the appropriate amino acid drop-out supplement (Q-Biogen, Basel, Switzerland). For experiments analyzing PHA synthesis in cells growing on medium containing fatty acids, a stationary phase culture was harvested by centrifugation, the cells were washed once in water and resuspended at a 1:10 dilution in fresh medium containing 0.1% (w/v) glucose, 2% Pluronic-127 (w/v) (Sigma), and between 0.01 to 0.1% (w/v) of various fatty acids. The cells were grown for an additional 4 days before harvesting them for PHA analysis. The 10-cis-heptadecenoic, tridecanoic, undecanoic, nonanoic, and heptanoic fatty acids were purchased from Nu-Check-Prep (Elysian, MN), and 9-methyl-undecanoic acid and 10-methyl-undecanoic acid were purchased from Larodan Fine Chemicals (Malmö, Sweden), whereas 8-methyl-nonanoic acid was purchased from Sigma.
For experiments aimed at analyzing the growth of yeast in medium containing fatty acids, BY4742 and the isogenic pte1⌬ were first grown in YPD (1% yeast extract, 2% bacto-peptone, and 2% dextrose) until the exponential phase. The cells were then washed once in water before suspending them at a density of 3.5 ϫ 10 5 cells/ml in fresh medium containing 0.67% yeast nitrogen base with amino acids (Difco), 0.5% ammonium sulfate, and either 0.05% 10-cis-heptadecenoic acid alone or in combination with either 0.01 or 0.02% 8-methyl-nonanoic acid. The cells were counted after 4 days of growth.
Analysis of PHA-PHA was analyzed as previously described (8). Briefly, cells were harvested by centrifugation, washed with water, and lyophilized in a glass tube. The dried material was washed with warm methanol (65°C) to remove free fatty acids, lipids, and acyl-CoAs, and the remaining material, including PHA, was trans-esterified at 94°C for 4 h with 1 ml of methanol/chloroform mixture (1:1) containing 3% sulfuric acid. One ml of 0.9% NaCl was added to the cooled mixture, vortexed vigorously, and centrifuged at 5000 ϫ g for 5 min. The chloroform phase containing the methyl esters of 3-hydroxy-acids was then analyzed and quantified by gas chromatography/mass spectrometry using a Hewlett-Packard 5890 gas chromatograph (HP-5MS column) coupled to a Hewlett-Packard 5972 mass spectrophotometer.
Plasmid Construction, Expression, and Purification of Recombinant Acyl-CoA Thioesterase-The YJR019C open reading frame (EMBL/ DDBJ/GenBank TM accession number NC_001142) was amplified from genomic DNA of the BY4742 strain using Pfu1 DNA polymerase (Stratagene) and oligonucleotides to add 12-and 15-base single-stranded cohesive ends at the 5Ј and 3Ј ends of YJR019c, respectively. These cohesive ends were created by T4 DNA polymerase treatment of the PCR product in the presence of deoxyguanosine 5Ј-triphosphate, and the resulting fragment was cloned into the expression plasmid pET30 Xa/LIC (Novagen, Madison, WI). The resulting clone, designated pET30-YJR019c, created a Pte1 protein that was fused at the amino terminus to 6 histidine residues (Pte1p-His 6 ). A control plasmid encoding only a histidine tag was also created, named pET30-His.
pET30-YJR019c and pET30-His were first sequenced to ensure the absence of mutations and then introduced into E. coli BL21(DE3)pLysS (Novagen). For polypeptide expression, cells were first cultivated at 37°C until an A 600 of 0.4, isopropyl ␤-D-thiogalactopyranoside was then added at a concentration of 0.5 mM, and cells were grown overnight at 25°C. Polypeptides were extracted from cells immediately after the overnight cultivation with the B-PER bacterial protein extraction reagent (Pierce). The extract was loaded onto a column of His⅐bind resin (Novagen) equilibrated with 50 mM NiSO 4 . Polypeptides were purified with a standard protocol of His⅐bind buffer kit (Novagen) by gravity flow.
Kinetic curves were fitted to the Hill equation (11), as follows: where v is the velocity, V max is the maximum velocity, K m is the Michaelis constant, and h is the Hill coefficient. The Hill coefficient and the Michaelis constant were determined from data plot to the following: log Analytical Procedures-Proteins were separated by SDS-PAGE and stained with Coomassie Brilliant Blue G-250. Gel filtration chromatography was performed with an HPLC system equipped with a TSK gel G3000SW column (7.5 ϫ 300 mm; Tosoh, Inc., Tokyo, Japan). Protein samples were eluted with 20 mM Tris-HCl buffer, pH 7.5, including 0.5 M NaCl at 1.0 ml min Ϫ1 . Acyl-CoAs were analyzed at A 260 with an HPLC system equipped with a TSK gel ODS-80TM column (4.6 ϫ 250 mm, 5-m-diameter particle, Tosoh, Inc.). Elution was performed with an acetonitrile gradient of 10 -90% (v/v) in 20 mM potassium phosphate buffer, pH 5.4, at 1.0 ml min Ϫ1 .

Analysis of Carbon Flux through ␤-Oxidation in pte1⌬
Mutant-The quantity and monomer composition of PHA synthesized from intermediates of the ␤-oxidation cycle was first measured in the wild type and pte1⌬ background for S. cerevisiae cells grown in medium containing either 0.1% 10-cis-heptadecenoic acid or 0.1% tridecanoic acid. The amount of PHA synthesized in the pte1⌬ strain represented 57 and 62% of the amount produced in wild type for cells in medium with 10-cisheptadecenoic acid and tridecanoic acid, respectively (Fig. 1, A and B). Furthermore, a significant increase in the proportion of the H5:0 monomer (the 3-hydroxy acid monomers found in PHA are defined with the prefix H followed by the number of carbon and the number of unsaturated bonds; see Table 1) was evident in the pte1⌬ strain. Thus, for PHA produced from 10-cis-heptadecenoic acid, the H5:0 monomer increased from 19 Ϯ 1 mol % in wild type to 26.3 Ϯ 0.8 mol % in pte1⌬, whereas for PHA produced from tridecanoic acid, the H5 monomer increased from 9 Ϯ 1 mol % in wild type to 15.7 Ϯ 0.7 mol % in pte1⌬ (Student's t test, p Ͻ 0.0001). In contrast to the relatively modest decrease in PHA quantity produced in the mutant strain grown in the presence of 10-cis-heptadecenoic acid or tridecanoic acid, a drastic difference was observed for cells grown in medium containing 8-methylnonanoic acid. Wild type cells grown in medium containing 0.1% 8-methyl-nonanoic acid produced 10 ϫ 10 Ϫ4 g of PHA/g of cell dry weight, with the polymer containing only two monomers, namely 3-hydroxy,8-methyl-nonanoic acid (8M-H9:0) and 3-hydroxy,6-methylheptanoic acid (6M-H7:0). In contrast, pte1⌬ cells grown in the same medium produced PHA near the detection limit at Ͻ0.02 ϫ 10 Ϫ4 g of PHA/g of cell dry weight (Fig. 1C). A large decrease in PHA synthesized from the ␤-oxidation of 8-methyl-nonanoic acid was also observed when the quantity of fatty acid in the medium was decreased to 0.01% (Fig. 1D). Thus, under these conditions, wild type produced 11 ϫ 10 Ϫ4 g of PHA/g of cell dry weight, whereas pte1⌬ produced only 0.3 ϫ 10 Ϫ4 g of PHA/g of cell dry weight. The proportion of the 6M-H7:0 monomer in cells grown on 0.01% 8-methyl-nonanoic acid increased from 30 Ϯ 0.3 mol % in wild type to 41 Ϯ 4 mol % in pte1⌬ strain (Student's t test, p Ͻ 0.005).
To determine whether the decrease in the catabolism of 8-methylnonanoic acid observed in the pte1⌬ strain was related to either the presence of an isomethyl group or the length of the main carbon chain, the quantity and monomer composition of PHA was first compared for wild type and pte1⌬ cells grown in tridecanoic, undecanoic, nonanoic, and heptanoic acids (Fig. 2). Because nonanoic and heptanoic acids were found to significantly inhibit growth at concentrations above 0.025%, the experiment was performed with the same concentration for all fatty acids, namely 0.025%. Although no significant differences were observed in the amount of PHA synthesized by wild type and pte1⌬ strains grown on either tridecanoic or undecanoic acid, the amount of PHA synthesized in the pte1⌬ cells grown in medium with nonanoic and heptanoic acids were 70 and 8.9% of the wild type level, respectively. For all fatty acids tested, there was a significant increase in the proportion of the H5 monomer in the pte1⌬ strain compared with wild type. Thus, the H5 monomer was found to increase from 6.2 Ϯ 0.2 to 9.1 Ϯ 0.3, 6.3 Ϯ 0.2 to 8.8 Ϯ 0.3, 4.2 Ϯ 0.6 to 7 Ϯ 1, and from 80 Ϯ 10 to Ͼ99 mol % for cells grown on tridecanoic, undecanoic, nonanoic, and heptanoic acids, respectively (Student's t test, p Ͻ 0.01).
PHA was also analyzed for cells grown in medium containing 9-methyl-undecanoic acid and 10-methyl-undecanoic acid, both at a 0.025% (Fig. 3). Deletion of Pte1 resulted in a significant decrease in PHA synthesized from the ␤-oxidation of the branched chain fatty acid tested, with the PHA amount in pte1⌬ being 34 and 59% of wild type level for cells grown on 9-methyl-undecanoic acid and 10-methyl-undecanoic acid, respectively. There was also a large increase in the proportion of the smallest monomer observed for PHA produced from either branched chain fatty acids in the pte1⌬ mutant. Thus, for PHA synthesized from 9-methyl-undecanoic acid, the 5M-H7:0 monomer increased from 8.7 Ϯ 0.5 mol % in wild type to 24.5 Ϯ 0.4 mol % in pte1⌬ (Student's t test, p Ͻ 0.00001). Similarly, for PHA synthesized from 10-methyl-undecanoic acid, the 6M-H7:0 monomer increased from 25.5 Ϯ 0.5 mol % in wild type to 49.
To examine whether the poor catabolism of 8-methyl-nonanoic acid through the ␤-oxidation observed in the pte1⌬ strain could negatively influence the catabolism of other fatty acids in the same cells, wild type and pte1⌬ cells were grown in medium containing either only 0.05% 10-cis-heptadecenoic acid, only 0.01% 8-methyl-nonanoic acid, or the combination of both 0.05% 10-cis-heptadecenoic acid and 0.01% 8-methyl-nonanoic acid (Fig. 4). For all treatments, the amount of PHA synthesized from the degradation of each external fatty acid was determined separately. Although PHA synthesized from the degradation of 10-cis-heptadecenoic acid in the pte1⌬ strain reached 66% of the wild type level for cells grown in medium containing only 10-cis-heptadecenoic acid, PHA synthesized from the catabolism of the same fatty acid in pte1⌬ decreased to 19% when 8-methyl-nonanoic acid was added to the medium. Conversely, although PHA synthesized from the degradation of 8-methyl-nonanoic acid in the pte1⌬ strain reached only 3% of wild type level for cells grown in medium containing only the branched chain fatty acid, PHA synthesized from the catabolism of the same fatty acid in pte1⌬ reached 15% when 10-cis-heptadecenoic acid was added to the medium.
To assess whether the reduction in the catabolism of 10-cis-heptadecenoic acid observed in the pte1⌬ strain by the addition of 8-methylnonanoic acid in the medium had an impact on the ability of the cells to grow using fatty acids as the external carbon source, wild type and pte1⌬ cells that did not contain the peroxisomal PHA synthase were first grown to an exponential phase in medium containing glucose, 3.5 ϫ 10 5 cells/ml were then inoculated to medium containing 0.05% 10-cis-heptadecenoic acid alone or in combination with 0.01 or 0.02% 8-methylnonanoic acid, and the cells were counted after 4 days (Fig. 5). Although pte1⌬ cells reached a density equivalent to 75% of wild type in medium containing only 10-cis-heptadecenoic acid, this level decreased to 40  and 34% when 0.01 or 0.02% 8-methyl-nonanoic acid was added to the medium, respectively. Enzymatic Activity of Purified Pte1p-A recombinant protein fraction purified by nickel affinity chromatography was analyzed with SDS-PAGE and gel filtration chromatography. A predominant protein migrating at 43.4 kDa was found by SDS-PAGE, whereas the molecular mass of PTE1p-His 6 polypeptide was predicted to be 45.3 kDa (Fig. 6A). Therefore, this nickel affinity protein fraction was used for Pte1p kinetic analysis. In gel filtration chromatography, several peaks emerged on the chromatogram of the PTE1p-His 6 fraction (Fig. 6B). From comparison of the retention time of peak 2 with those of molecular mass marker proteins, it was found that the peak 2 fraction contained a protein with the molecular mass of 45 kDa, which corresponded to the putative molecular mass of PTE1p-His 6 (Fig. 6B). Thioesterase activity toward acetyl-, butyryl-, octanoyl-, lauroyl-, oleoyl-, and 8-methyl-nonanoyl-CoA could be detected only in the peak 2 fraction but not in the peak 1, -3, -4, and -5 fractions (Fig. 6B and data not shown). Although the nature of peaks 3-5 has not been analyzed, they could represent degradation products of PTE1p-His 6 or other contaminants, such as imidazole. As a control, thioesterase activity in a nickel affinity protein fraction extracted and purified from a strain expressing only a His tag (pET30His) was tested. No thioesterase activity toward acetyl-, butyryl-, octanoyl-, lauroyl-, oleoyl-, or 8-methyl-nonanoyl-CoA was found in the fraction, indicating that intrinsic E. coli enzymes do not interfere with the Pte1p kinetic analysis using the nickel affinity protein fraction.
Thioesterase activity with octanoyl-CoA was found to be maximal at pH 8 (Fig. 7A). Thus, all further assays were performed in a potassium phosphate buffer at pH 8.0. The requirement of BSA for Pte1p acyl-CoA thioesterase activity was investigated. BSA did not significantly affect the specific activity with acetyl-, butyryl-, octanoyl-, 8-methyl-nonanoyl-, and nonanoyl-CoA. However, the specific    Bars with different letters (a, b, or c) indicate statistical differences between groups (analysis of variance and Tukey honestly significantly different (HSD) test).

DISCUSSION
Peroxisomal acyl-CoA thioesterase has been identified in mouse, human, plant, and fungi (2, 3, 6, 7). Numerous roles have been postulated for this enzyme in the metabolism of lipids, including the release of free fatty acids from acyl-CoA intermediates of the ␤-oxidation cycle followed by export of the fatty acids out of the peroxisome for either excretion into the urine or re-entry into the mitochondria for further ␤-oxidation, prevention of the accumulation of poorly metabolized CoA esters to promote efficient flux through ␤-oxidation, or catabolism of cholesterol via hydrolysis of bile acid-CoA (12). However, direct experimental evidence of the in vivo importance of the peroxisomal acyl-CoA thioesterase in these proposed activities has been generally lacking. In S. cerevisiae, although one study shows that deletion of the Pte1 gene leads to a small but significant decrease in the ability of cells to grow on oleic acid as the main carbon source, no obvious effects have been observed in a separate study with a similar mutant (2, 3). Overexpression of the human peroxisomal acyl-CoA thioesterase III (hTE/ hACTEIII/PTE-1) in human and murine T-cell lines or in thymocytes of transgenic mouse results in an increase in peroxisome number (13).
In the present study, the role of the S. cerevisae peroxisomal acyl-CoA thioesterase Pte1p in the metabolism of fatty acids via the ␤-oxidation cycle was analyzed using the synthesis of PHA via the polymerization of the ␤-oxidation intermediate 3-hydroxyacyl-CoA as a marker for the state of carbon flux through this pathway. Previous studies using PHA synthesis in S. cerevisae peroxisomes have been used to reveal the existence of a futile cycle of intermediates of the fatty acid biosynthetic pathway toward the ␤-oxidation cycle, to monitor the substrate specificity of mutants of the multifunctional enzyme Fox2p, as well as to study the involvement of auxiliary enzymes in the degradation of transunsaturated and conjugated fatty acids (14 -17).
PHA synthesized from the ␤-oxidation of fatty acids present in the medium was found to be slightly reduced (Ͻ2-fold) in S. cerevisae pte1⌬ mutant compared with wild type when cells were grown in medium containing 0.1% 10-cis-heptadecenoic acid or 0.1% tridecanoic acid. For tridecanoic acid, the difference in PHA amount between wild type and pte1⌬ was abolished when the concentration of the fatty acid in the medium was decreased to 0.025%. No differences in PHA amount were also observed for wild type and pte1⌬ cells grown in medium containing 0.025% undecanoic acid, whereas only 1.4-fold less PHA was observed for pte1⌬ compared with wild type for cells grown on 0.025% nonanoic acid. In contrast, 11-fold less PHA was observed for pte1⌬ compared with wild type for cells grown in medium containing 0.025% heptanoic acid. These data indicate that, although the presence of the peroxisomal acyl-CoA thioesterase is not essential to maintain a normal or near normal level of carbon flux through the ␤-oxidation cycle when the medium or long straight chain fatty acids are metabolized, the same enzyme is important for efficient degradation of short chain fatty acids.
The effect of the absence of the peroxisomal acyl-CoA thioesterase on the ␤-oxidation of fatty acids was found to be even more drastic for the metabolism of 8-methyl-nonanoic acid. Thus, PHA synthesis from the ␤-oxidation of 8-methyl-nonanoic acid was reduced by ϳ30 -500fold in the pte1⌬ mutant compared with wild type when the external 8-methyl-nonanoic acid concentrations were 0.01 and 0.1%, respectively. In contrast, the reduction in PHA synthesized in pte1⌬ compared with wild type was Ͻ3-fold for 9-methyl-undecanoic acid and 1.5-fold for 10-methyl-undecanoic acid. These results indicate that short branched chain fatty acids are less well metabolized via ␤-oxidation than their straight chain counterpart in the pte1⌬ mutant compared with wild type and that this difference is enhanced for shorter branched chain fatty acids such as 8-methyl-nonanoic acid.
The very low amount of PHA synthesized from 8-methyl-nonanoic and heptanoic acids in the pte1⌬ strain compared with wild type highlight the importance of the peroxisomal acyl-CoA thioesterase for the efficient metabolism of these short straight chain or branched chain fatty acids via the ␤-oxidation cycle. It is possible that one or several of the enzymes of the core ␤-oxidation cycle poorly metabolize such fatty acids. Accumulation of poorly metabolized CoA-esters would result in a block of the ␤-oxidation cycle, because adequate concentration of free CoASH is essential for the 3-ketoacyl-CoA thiolase step. Alternatively, it is possible that a CoA-ester derived from 8-methyl-nonanoic and heptanoic acids inhibits one of the enzymes of the ␤-oxidation cycle. Thus, whereas in the pte1⌬ the inhibitory or poorly metabolized acyl-CoAs would accumulate and reduce the flux through the ␤-oxidation cycle, the presence of the acyl-CoA thioesterase in the wild type would decrease the level of such acyl-CoAs and allow ␤-oxidation to proceed (Fig. 9). A corollary of this hypothesis is that accumulation of the inhibitory or poorly metabolized acyl-CoA in the pte1⌬ mutant grown in medium containing 8-methyl-nonanoic acid would not only block the metabolism of 8-methyl-nonanoic acid but also that of other fatty acid targeted for degradation via ␤-oxidation in the same cell. This was verified by showing that, although synthesis of PHA from 10-cis-heptade-   cenoic acid in wild type cells was not significantly affected by the addition of 8-methyl-nonanoic acid in the medium, a 3.5-fold reduction in PHA synthesized from the degradation of 10-cis-heptadecenoic acid in pte1⌬ was observed when 8-methyl-nonanoic acid was added to the medium. The reduced catabolism of 10-cis-heptadecenoic acid in pte1⌬ cells caused by the presence of 8-methyl-nonanoic acid also impacted the ability of cells to grow in medium containing fatty acids as the main carbon source. This was revealed by the reduction in cell density in pte1⌬ cells when either 0.01 or 0.02% 8-methyl-nonanoic acid was added to medium containing 0.05% 10-cis-heptadecenoic acid. Even for fatty acids that are relatively well metabolized in both wild type and pte1⌬, the reduced catabolism of short chain fatty acyl-CoAs in pte1⌬ is revealed by the increased proportion of the H5:0 monomer in PHA synthesized in the mutant from straight chain fatty acids, as well as the 6M-H7:0 and 5M-H7:0 monomer in PHA synthesized from the catabolism of 10-methyl-undecanoic acid and 9-methyl-undecanoic acid, respectively. A similar increase in the proportion of short chain PHA monomers (H5:0 or H6:0) has been observed in a previous study, where the wild type Fox2p encoding the multifunctional enzyme (MFE-2) was replaced by a mutant variant with an inactivation in the B domain of the dehydrogenase (14). This MFE-2(b⌬) mutant showed undetectable dehydrogenase activity toward (R)-3-hydroxybutyryl-CoA, whereas k cat values toward (R)-3-hydroxydecanoyl-CoA and (R)-3-hydroxyhexadecanoyl-CoA was reduced by Ͻ2-fold (18). It was reasoned that the reduced activity of the MFE-2(b⌬) toward short chain (R)-3-hydroxyacyl-CoAs made them more available to the PHA synthase compared with the other intermediates generated by the wild type MFE-2. Thus, it can be also reasoned that the increased proportion of the small straight or branched chain 3-hydroxyacids present in PHA derived from the pte1⌬ strain reflects an increased concentration of the corresponding 3-hydroxyacyl-CoAs in the mutant because of their reduced metabolism via the ␤-oxidation cycle in the absence of the peroxisomal acyl-CoA thioesterase. This suggests that an inappropriate high level of certain acyl-CoAs in the pte1⌬ would have a larger negative effect on the metabolism of short chain ␤-oxidation intermediates compared with longer ones.
The kinetic parameters measured for the purified Pte1p fit well with the implication of this enzyme in the metabolism of short straight and branched chain fatty acyl-CoAs. The enzyme showed a relatively broad range of activity, with lowest activity toward oleoyl-CoA and highest activity toward the short chain butyryl-CoA and intermediate activities toward nonanoyl-, decanoyl-, and 8-methyl-nonanoyl-CoA. The kinetic parameters of the enzyme toward nonanoyl-and 8-methyl-nonanoyl-CoA were almost identical, indicating that the addition of an isomethyl group did not substantially modify enzymatic activity. The substrate preference observed for the purified Pte1p differed from the maximal activity toward lauroyl-CoA measured by Kal et al. (3) in a crude extract from cells overexpressing the Pte1 gene from a 2 multicopy plasmid. Numerous differences exist in the assay conditions between these two studies, including parameters of the assay buffer, such as pH and the presence of detergent. The presence of inhibitors in a crude S. cerevisae extract may potentially result in a shift in the maximal activity of the acyl-CoA thioesterase toward longer acyl-CoAs.
The activity of Pte1p was inhibited by CoASH. Such inhibition has also been reported for the human peroxisomal PTE-2 (6). Thus, in the presence of inhibitory or poorly metabolized acyl-CoAs that would lead to a sequestration of CoASH into esters, such as 8-methyl-nonanoyl-CoA, the acyl-CoA thioesterase would be highly active and would work to decrease the concentration of the acyl-CoA and re-establish a pool of free CoASH required for maintaining a flux through the ␤-oxidation cycle.
Short chain fatty acids and methyl-branched fatty acids are relatively common fatty acids found in nature. Caproic (6:0) and caprylic (8:0) acids are found in both animal and plant fats, such as the milk of cows and goats and the seeds of plants, such as coconut and oil palms. Although, to our knowledge, branched chain fatty acids have not been identified in S. cerevisae, fatty acids with a methyl group in the iso or anteiso position are abundantly found in the membrane lipids of a wide variety of bacteria, including bacteria found in the gut of ruminants, causing the appearance of such fatty acids in the milk and meat of these animals (19,20). Thus, the requirement of the peroxisomal acyl-CoA thioesterase for the efficient degradation of short straight chain and branched chain fatty acids could have implications as far the fitness of organisms, such as S. cerevisae, which must grow and survive in an environment containing such fatty acids.