| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Karolinska Institutet, Department of Laboratory Medicine, Division of Clinical Chemistry C1–74, Karolinska University Hospital at Huddinge, Stockholm, Sweden
2Correspondence: Karolinska Institutet, Department of Laboratory Medicine, Division of Clinical Chemistry C1–74, Karolinska University Hospital at Huddinge, Stockhold SE-141 86, Sweden. E-mail: mary.hunt{at}ki.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: ß-oxidation • acyl-coenzyme A • peroxisome • fatty acids
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) include oxidation of fatty acids (
, ß, and 
-oxidation), biosynthesis of lipids, and allosteric regulation of a number of enzymes (2
3
4)
and opening of KATP channels in human pancreatic beta cells (5)
. There are, however, many processes that are activated by acyl-CoAs and inhibited by the free fatty acids or vice versa, showing the importance of maintaining the balance of these compounds in vivo. It was recently shown that palmitoyl-coenzyme A (CoA) inhibits AMP-activated protein kinase (AMPK) indirectly by a specific and concentration-dependent inactivation of its upstream kinase (6)
. Another example where the balance of free fatty acids vs. acyl-CoAs is important is in the activation of the peroxisome proliferator-activated receptors (PPARs) by free fatty acids (7
8
9
10)
and, more recently, by CoA esters of long chain fatty acids (acyl-CoAs) (11)
. One way in which the cell can control the intracellular acyl-CoA and free fatty acid levels is via the activity of enzymes named acyl-CoA thioesterases (ACOTs), which can hydrolyze acyl-CoAs to the corresponding free fatty acid and CoA (CoASH) (for reviews, see refs. 12
, 13
). These acyl-CoA thioesterases are expressed in several cellular compartments that coincide with metabolic processes involving acyl-CoAs.
In 1995, a particular family of acyl-CoA thioesterases was identified in rat and was named Type-I acyl-CoA thioesterases (14)
. In the past few years, the cloning of cDNAs and genes has identified the Type-I acyl-CoA thioesterases to be a novel gene family with six members in the mouse: a cytosolic isoform (Acot1, previously CTE-I), a mitochondrial isoform (Acot2, previously MTE-I), and four peroxisomal isoforms (Acot3, 4, 5, and 6, previously PTE-Ia, -Ib, -Ic, and -Id) (15
16
17)
. Both the sequence similarity and the gene organization show a high degree of conservation, suggesting that these genes have evolved by gene duplications. The identification of four peroxisomal ACOTs was surprising, but characterization of their acyl-CoA substrate specificities showed that they catalyze distinct activities and also show individual tissue expression. However, based on the multiple and important functions that peroxisomes play in degrading a variety of lipids, which is reflected in the severity of peroxisomal disorders (for reviews, see 18
, 19
), it is tempting to speculate that these ACOTs participate in regulation of ß-oxidation and release of ß-oxidation products. In the mouse this Type-I gene cluster is located on mouse chromosome 12 D3 within 120 kb of DNA. We have now characterized the genes in the corresponding gene cluster of Type-I ACOTs in humans, which is localized on human chromosome 14q24.3. However, this cluster contains four genes, of which three are transcribed to encode full-length ACOTs, and only one gene encodes a peroxisomal acyl-CoA thioesterase (ACOT4). We show here that the human ACOT4 gene apparently has evolved into a multifunctional protein that catalyzes the combined activities of mouse ACOT3, 4, and 5, while human ACOT1 and ACOT2 have functions similar to the mouse orthologs.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The plasmids containing the human ACOT1, 2, and 4 were then used to transform BL21(DES3)pLysS cells (Novagen Inc., Darmstadt, Germany). Bacteria were cultured in 500 ml Luria-Bertani medium at 37°C with the addition of ampicillin (50 µg/ml) and chloramphenicol (34 µg/ml) until the optical density (OD)600nm was 
0.6. Protein expression was induced by addition of 1 mM isopropyl-1-thio-ß-D-galactopyranoside and cells were grown for 3 h at 37°C (ACOT1 and ACOT2) or at 30°C for 5 h (ACOT4). The recombinant proteins were purified as described previously and eluted using 500 mM imidazole (20)
. The purity of the expressed proteins was examined using SDS/PAGE analysis and Coomassie Brilliant blue staining, and these purified proteins were used to measure acyl-CoA thioesterase activity. Mouse ACOT3, 4, and 5 recombinant proteins were produced as outlined in ref. 16
, 17
and activity was measured with various acyl-CoA substrates.
Northern blot analysis
Human Multiple Tissue blots were purchased from Clontech Inc (Palo Alto, CA, USA) and analysis was carried out according to the manufacturer’s instructions, using -32P-labeled cDNA probes. The cDNA probes for the human ACOTs were constructed using RT-PCR from human liver total RNA. For ACOT1 and ACOT2, a cDNA probe was amplified corresponding to amino acids 167–371 of the open reading frame (ORF) (shown in
Fig. 2
). The ACOT4 cDNA probe was amplified corresponding to amino acids 215 to 405 of the ORF, and a specific probe located in the mitochondrial targeting signal of ACOT2 was amplified corresponding to amino acids 21 to 59 of the ORF.
|
|
Enzyme activity measurements
ACOT activity was measured spectrophotometrically at 412 nm with 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB). The medium contained 200 mM potassium chloride, 10 mM HEPES, and 0.05 mM DTNB (pH 7.4). An E412 = 13,600 M–1 cm–1 was used to calculate the activity. The acyl-CoA substrates used were obtained from Sigma-Aldrich Inc. (St. Louis, MO, USA). The addition of BSA to prevent substrate inhibition with CoA esters was titrated for long chain acyl-CoAs (C16-C20-CoA). A BSA:C18-CoA molar ratio of 
1:10 was used to measure ACOT1 and ACOT2 activities. Otherwise addition of BSA was not required for measurement of acyl-CoA thioesterase activity.
Cellular localization of human ACOTs using green fluorescent fusion protein
The ORFs for the human ACOT1, 2, 4, and 6 were cloned as fusion proteins with green fluorescent protein (GFP) to examine the intracellular localization of the proteins. ACOT6 was amplified using primers 5'-ATGCTGCAGCATCCAAAGGTG-3' and 5'-TCTGTCAAGCACAGCAAAATATAA-3'. The ORFs were amplified using PCR and the PCR products for ACOT1, ACOT4, and ACOT6 were cloned into the pcDNA3.1/NT-GFP vector (Invitrogen Corp., CA, USA), in-frame with the GFP, leaving the carboxyl-terminal end accessible. The ACOT2 was cloned in-frame with the pcDNA3.1/CT-GFP (Invitrogen Corp., CA, USA), leaving the N-terminal end accessible. Sequence analysis was performed using Big Dye Terminator Ready Reaction kit (Perkin Elmer Inc., Wellesley, MA, USA) and sequenced by Cybergene AB (Huddinge, Sweden).
Human skin fibroblasts or HepG2 cells were grown in Eagles MEM (Sigma-Aldrich Inc.), supplemented with 10% fetal calf serum (Invitrogen Corp., CA, USA) and 100 U penicillin/100 µg streptomycin in an atmosphere of 5% CO2. Cells were grown overnight in 60 mm dishes on glass coverslips and transfected with 10 µg of the various plasmids using calcium phosphate method as described (20)
. The localization of ACOT1, ACOT4, and ACOT6 was investigated in fibroblasts, and in the case of ACOT4, skin fibroblasts from a Zellweger patient were also used. The localization of ACOT2 was investigated in HepG2 cells and Mitotracker® Orange CMTMRos (Invitrogen Corp., CA, USA) was used to confirm mitochondrial structures. Transfected cells were grown for 48 h and fibroblasts were treated for immunofluorescence as described (20)
. HepG2 cells treated with 200 nM Mitotracker® Orange for 30 min before immunofluorescence microscopy was carried out as described previously (20)
, but without use of the Tritc-labeled antibody (Ab).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
16
17)
. Database searches using the cDNA sequences for the mouse Type-I Acot genes identified four putative human ACOT genes on chromosome 14q24.3 located within 80 kb of genomic sequence (Fig. 1
). Based on homology to the mouse genes, we identified the putative human orthologs of Acot1, Acot2, Acot4, and Acot6 in the human gene cluster. The distance between the genes is 25.7 kb between ACOT1 and ACOT2, 16.5 kb between ACOT2 and ACOT4, and 16.9 kb between ACOT4 and ACOT6. Comparison of the two gene clusters reveals two striking differences: first, that the order of the ACOT1 and ACOT2 genes are different in the human genome compared to mouse, and second, that the human gene cluster contains only one gene encoding a putative peroxisomal ACOT. Similar to the mouse genes, the open reading frames for three of these genes are coded for by three exons (ACOT1, ACOT2, and ACOT4) whereas the fourth gene identified in the cluster, ACOT6, contained the genomic information for all three exons (which would result in a full-length ACOT protein), but apparently is translated from a methionine at the end of exon 2 through exon 3. This is based on expressed sequence tag (EST) sequences, the estimated size of the transcript detected by Northern blot (data not shown), and by the fact that we could not amplify the predicted full-length open reading frame but could amplify a transcript that covered part of exon 2 and exon 3. Database searches in the entire human genome did not indicate the presence of any other genes that would correspond to mouse Acot3 or Acot5, but another gene was identified on chromosome 19q13.12 that shows 91% sequence identity at nucleotide level to the ACOT4 gene on chromosome 14. However, this gene on chromosome 19 lacks introns and contains several stop codons that will not result in active protein. Searches in EST databases show that the pseudogene is expressed in purified pancreatic islets (EST No. CA843741, CB177282, and CA843493), brain (EST No. BI827748 and AW090080), and placenta (EST No. DB328688 and DA849162).
Alignment of the deduced amino acid sequences for the open reading frames of the ACOT1, ACOT2, and ACOT4 genes shows the striking conservation between the various ACOTs (Fig. 2
). The ACOT1 and ACOT4 genes encode proteins of 421 amino acids, with the latter containing –PKL at its carboxyl terminal, which is a variant of the peroxisomal type-1 targeting signal (PTS1) that can target proteins to peroxisomes (21)
. ACOT2 encodes a protein of 483 amino acids and contains a 62 amino acid leader sequence at the N-terminal that targets the protein to mitochondria (see below). The 421 amino acids of ACOT2 that align to ACOT1 show 98.6% sequence identity, with only 5 amino acids being different. The three amino acids identified to be the catalytic triad of the active site in the mouse ACOT enzymes (22)
, a serine (acting as the nucleophile) in a Gly-X-Ser-X-Gly motif common to serine esterases, a histidine, and an aspartic acid are all conserved in the human enzymes (indicated in Fig. 2
). The serine is located at position 232 of ACOT1, the aspartic acid at 326, and the histidine at 360.
Subcellular localization of the human ACOTs
The subcellular localization of the four human ACOTs identified in this study was examined by expression of the ACOTs in human skin fibroblasts and HepG2 cells as GFP fusion proteins. The rationale for choosing either N-terminal or C-terminal GFP was based on the presence of putative targeting signals at the N- or C-terminal ends, and also on previous data obtained for localization of the mouse orthologs. The ACOT1 (homologous to the mouse ACOT1, which is cytosolic; ref. 23
) contains no apparent targeting signal but contains a PTS1-like sequence of –SKV at its C-terminal end. However, transfection of this GFP fusion protein into human fibroblasts resulted in a protein with a diffuse pattern of expression with no visible punctate staining, in line with a cytosolic localization (Fig. 3
A). ACOT6 also contains a PTS1-like sequence of –SKI at its C-terminal end, but this GFP fusion protein was also localized in cytosol (Fig. 3B
). The ACOT2 protein contains a putative mitochondrial targeting signal at its N-terminal end. This GFP fusion protein showed a typical pattern for mitochondrial localization as judged by immunofluorescence microscopy (Fig. 3C
). Use of a mitotracker confirmed that the ACOT2 fusion protein was localized in mitochondria (Fig. 3D
). The human ACOT4 contains a near-consensus PTS1 of –PKL at its C-terminal, which is also present in C. tropicalis sterol carrier protein (POX18) and has been shown to target proteins to peroxisomes (24)
. We transfected the ACOT4/NT-GFP vector into both control fibroblasts and fibroblasts from a Zellweger patient, which are unable to import peroxisomal matrix proteins. Using immunofluorescence microscopy for GFP detection, ACOT4 showed a punctate pattern of expression in control fibroblasts indicative of a peroxisomal localization (Fig. 3E
). However, in Zellweger fibroblasts, transfection of ACOT4 resulted in a diffuse GFP expression, showing an inability of the protein to be imported into peroxisomes (Fig. 3F
).
|
Recombinant expression and characterization of human and mouse ACOT proteins
Cloning of the open reading frames of ACOT1 and ACOT2 into pET16b and ACOT4 into pET102 vectors resulted in expression of these proteins as His-tagged fusion proteins, to allow for purification using affinity chromatography. After purification on a HiTrapTM column, the purified ACOT1 and ACOT2 were detected as single bands of 46 kDa in mass on SDS/PAGE gel stained with Coomassie Brilliant blue, whereas ACOT4 was detected as an 60 kDa band as a thioredoxin fusion protein (data not shown).
The human ACOT4 gene shows the strongest similarity to ACOT4 in the mouse, and is therefore likely to be the human ortholog of mouse ACOT4. The mouse ACOT4 enzyme was recently shown to be a highly specific succinyl-CoA thioesterase, with only low activity with glutaryl-CoA and no activity with any other tested acyl-CoAs (17)
. In line with this, characterization of recombinant human ACOT4 indeed showed the highest activity with succinyl-CoA (Vmax581 nmol/min/mg protein, Km14 µM), with much lower Vmax and higher Km with glutaryl-CoA (132 nmol/min/mg protein and 147 µM, respectively). Surprisingly, ACOT4 was also active on all long chain saturated acyl-CoAs of >8 carbon atoms as well as on all unsaturated acyl-CoAs tested (C18:1, C18:2, and C20:4). The calculated Vmax and Km values of ACOT4 with C14-CoA (which was the best long chain acyl-CoA substrate) were 137 nmol/min/mg protein and 3.4 µM, respectively. Figure 4
A summarizes the activity of human ACOT4 with all acyl-CoAs tested, demonstrating that human ACOT4 is a succinyl-CoA thioesterase that also hydrolyzes long chain saturated and unsaturated monocarboxylic acyl-CoAs. The pattern of activity of human ACOT4 mirrors the combined activities of the recently characterized mouse ACOT3, ACOT4, and ACOT5 proteins (summarized in Fig. 4B
). Therefore, the human peroxisomal ACOT4 can replace the function of three distinct peroxisomal enzymes in mouse. Although succinyl-CoA (C4-dicarboxylyl-CoA) was the best substrate for ACOT4, the corresponding C4-monocarboxylyl-CoA (butyryl-CoA) was not a substrate, and in contrast the C12 dicarboxylyl-CoA was not a substrate even though the corresponding C12 monocarboxylyl-CoA was a good substrate (data not shown). These data demonstrate that human ACOT4 has not simply evolved into a nonspecific acyl-CoA thioesterase, but rather that human ACOT4 has acquired the combined activities of mouse ACOT3, ACOT4, and ACOT5.
|
Kinetic characterization of recombinant human ACOT1 and ACOT2 showed similar substrate specificities, with both enzymes being mainly active on long chain saturated acyl-CoAs of 12 to 20 carbon atoms and long chain unsaturated acyl-CoAs such as C16:1-CoA, C18:1-CoA (Fig. 5
A, B). The data thus show that ACOT1 and ACOT2 function as long chain acyl-CoA thioesterases with no detectable activity with acyl-CoAs of eight carbon atoms or shorter. Kinetic characterization of ACOT1 and ACOT2 show that the Km for long chain acyl-CoAs is in the low micromolar range, demonstrating that these are probably the in vivo substrates for the enzymes (Table 1
).
|
|
Tissue expression of human ACOTs
Northern blot analysis was carried out to examine tissue expression of the various human ACOTs (Fig. 6
). As ACOT1 and ACOT2 in humans show 99% sequence identity to each other, it was not possible to make a probe that would be specific for the ACOT1. Using a cDNA probe to ACOT1/ACOT2, one transcript of 1.8 kb was detected with highest levels in heart, liver, muscle, and kidney. However, a cDNA probe corresponding to the mitochondrial targeting signal of the ACOT2 was constructed that will distinguish between these two mRNAs. Using the ACOT2-specific probe, the mRNA signal was again strongest in heart, liver, muscle, and kidney and weak in placenta and pancreas, suggesting that ACOT2 is much more strongly expressed than ACOT1. This is further supported by searches in the EST database, which showed that ESTs corresponding to ACOT2 were much more abundant than ESTs corresponding to ACOT1. The ACOT4 was detected as two transcripts of 1.7 and 2.2 kb, with the strongest expression in liver and kidney and weaker expression in placenta, heart, and muscle. These two transcripts were differentially expressed with the lower transcript expressed mainly in liver and kidney; the upper transcript was detectable in heart, liver, muscle, and kidney.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
16
17)
containing six individual genes: Acot1 (cytosolic), Acot2 (mitochondrial), and Acot3, 4, 5, and 6 (all peroxisomal). Here we have characterized the corresponding cluster in human that consists of four ACOT genes on chromosome 14q24.3. Three of these genes—ACOT1, ACOT2, and ACOT4—encode full-length ACOT proteins and one gene, ACOT6, is translated from a methionine at the end of exon 2 into a putative open reading frame of only 207 amino acids. Although the catalytic amino acids would be contained in the protein, it is not clear whether the protein will be active. The mouse ACOT6 ortholog contains a canonical PTS-1 targeting signal of –SKL in the C-terminal end that targets mouse ACOT6 to peroxisomes (M. A. K. Westin et al., unpublished results), whereas human ACOT6 ends with an isoleucine instead of leucine (C-terminal end of –SKI), which results in a cytosolic localization. It is remarkable that the mouse contains so many peroxisomal ACOT enzymes, whereas human contains only one functional peroxisomal Type-I ACOT. However, our finding that human ACOT4 hydrolyzes succinyl-CoA, glutaryl-CoA, and long chain acyl-CoAs, provides a plausible explanation for the difference in the number of peroxisomal ACOTs in humans and mice. In the mouse there are three Type-I ACOT enzymes characterized to date that hydrolyze straight chain acyl-CoA esters, the ACOT3 active on long chain acyl-CoAs and the ACOT5 active on medium chain acyl-CoAs (16)
, together with ACOT4, a newly identified succinyl-CoA/glutaryl-CoA thioesterase (17)
. While mouse ACOT4 only hydrolyzes succinyl-CoA with high activity and glutaryl-CoA with low activity, human ACOT4 also hydrolyzes saturated and unsaturated medium to long chain acyl-CoAs. It is therefore reasonable to suggest that human ACOT4 catalyzes the combined activities of mouse ACOT3, 4, and 5. Also, human ACOT4 is mainly expressed in liver and kidney, similar to mouse Acot3 and 4, further strengthening a functional redundancy. It would therefore be interesting to compare this gene cluster in various species to trace the appearance/disappearance of genes in this cluster and to follow the acquisition of the different activities in ACOT4 in higher vertebrates. A similar convergent evolution from mouse to human is seen among acyl-CoA oxidases, peroxisomal enzymes catalyzing the first reaction in the ß-oxidation of various acyl-CoAs in peroxisomes. While rat and mouse have three genes encoding straight chain acyl-CoA oxidase, trihydroxycoprostanoyl-CoA oxidase, and branched chain (pristanoyl-CoA) acyl-CoA oxidase, (25)
, humans have only two genes encoding the straight chain oxidase and a combined trihydroxycoprostanoyl-CoA and pristanoyl-CoA oxidase (26
, 27)
. In humans there is also an intron-less ACOT pseudogene located on chromosome 19q13.12, which shows a high similarity to ACOT4. Although this pseudogene is expressed at mRNA level in several tissues, the function is not known. However, recently it was shown that a pseudogene of the Makorin 1 gene (Makorin 1-p1) could regulate mRNA stability of its homologous coding gene (28)
. Therefore, the ACOT pseudogene may have a regulatory function on the expression of ACOT4. The fourth gene in the ACOT gene cluster, ACOT6, contains the genetic information for all three coding exons common to the ACOT genes described in this work; however, translation apparently starts from a methionine present at the end of exon 2. We were able to amplify a PCR product containing this truncated open reading frame of ACOT6, but it is not known whether ACOT6 is expressed as an active protein. The corresponding mouse ACOT6 gene encodes a 46 kDa protein that is expressed, but so far the substrate for this enzyme has not been defined (M. A. K. Westin et al., unpublished results).
Cloning and sequence analysis of ACOT1 and ACOT2 revealed open reading frames encoding proteins of 421 amino acids that aligned to each other, with 98.6% sequence identity. While ACOT2 contains a further 62 amino acids in-frame at the N-terminal end that functions as a mitochondrial targeting signal, ACOT1 lacks these amino acids. ACOT2 was previously cloned as a human peroxisomal acyl-CoA thioesterase named PTE2 (29)
, but sequence alignment of PTE2 reveals that it relates to ACOT2 without its mitochondrial targeting signal. The assignment of the name PTE2 was based on the finding that the –SKV tripeptide at the C-terminal end acted as a weak peroxisomal targeting signal, mediating some import of this protein into peroxisomes. Indeed, it has been shown that mutation of the peroxisomal targeting signal from the consensus –SKL to –SKV results in mistargeting to cytosol (30)
. We now show that ACOT2 contains a mitochondrial targeting signal at its N-terminal end, which targets the protein to mitochondria. In line with the apparent lack of targeting signals in ACOT1, the protein remained in the cytosol in the GFP experiments.
Mouse and human peroxisomes contain a further acyl-CoA thioesterase, now renamed ACOT8 (previously called PTE1, hTE, hACTEIII, and PTE-2) in a new acyl-CoA thioesterase nomenclature system (31)
, which can hydrolyze a wide variety of CoA esters, including bile acid CoAs, branched chain CoAs, and long, medium, and short chain acyl-CoAs (20
, 32)
. Therefore it appears that mouse peroxisomes contain five different peroxisomal acyl-CoA thioesterases, hydrolyzing a broad spectrum of CoA esters, whereas human contains only two genes coding for peroxisomal acyl-CoA thioesterases, whose activity cover the same span of CoA ester substrates. The need for several peroxisomal ACOT enzymes with different specificities becomes evident in view of the multiple functions that peroxisomes play in chain shortening of various carboxylic acids (as peroxisomes do not catalyze complete ß-oxidation of lipids), including long and very long chain fatty acids, dicarboxylic acids, bile acid intermediates, and xenobiotic carboxylic acids (19
, 33)
. Furthermore, fatty acids are apparently transported across the peroxisomal membrane as CoA esters (in contrast to mitochondria that utilizes a carnitine-acylcarnitine transport system), whereas fatty acid degradation products may be transported out of the peroxisome as carnitine esters, glycine, or taurine conjugates or as the free acids. This implies that ACOTs may serve essential functions in peroxisomes to regulate ß-oxidation (CoASH levels) as well as in the termination of ß-oxidation at various chain lengths for exit out of the organelle. In view of this, we were surprised to only find one peroxisomal Type-I ACOT in the human genome. However, the finding here that ACOT4 in fact catalyzes the activities of three mouse ACOTs not only explains the lower number of human ACOT genes, but underscores the physiological importance of these ACOTs in peroxisomal lipid metabolism.
ACOTs play key roles in maintaining the intracellular ratio between the CoA esters of various lipids and the free acid (e.g., acyl-CoAs and free fatty acids). Putative functions for these ACOT enzymes in cytosol, mitochondria, and peroxisomes have been described in detail in a recent review, based mainly on work carried out on the mouse enzymes (12)
. ACOT1 and ACOT2 have quite similar substrate specificities, both active on long chain acyl-CoAs, which is not surprising based on their extremely high level of sequence identity (>98%). However, these enzymes show two distinct cellular localizations in cytosol and mitochondria, and therefore have specific functions within these compartments in the metabolism of long chain acyl-CoAs. In mitochondria, ACOT2 has been suggested to be involved in a coordinate shuttle pathway together with uncoupling protein 3 (UCP3) in muscle and brown adipose tissue (34)
. This hypothesis suggests that free fatty acids are shuttled out of the mitochondria by UCP3 when fatty acid ß-oxidation predominates. These free fatty acids would result from the hydrolysis of acyl-CoAs by ACOT2, thus releasing CoASH for further reactions within the mitochondria during ß-oxidation. The free fatty acids transported out of the mitochondria by UCP3 would then be reactivated by acyl-CoA synthetases to form acyl-CoAs, which could subsequently be transported back into the mitochondria as a carnitine ester via the sequential action of carnitine palmitoyl transferase I (CPTI) and the carnitine acylcarnitine (CAC) transporter. This shuttle mechanism would prevent overloading of the mitochondria with acyl-CoA and trapping of CoASH, which is vital for ß-oxidation and the citric acid cycle to proceed. A role for ACOT1 has been suggested in control of ligand supply for the PPAR family of nuclear receptors in the form of acyl-CoAs or free fatty acids and/or channeling fatty acids toward degradation rather than esterification (22)
.
In summary, the identification of a family of human acyl-CoA thioesterases with distinct tissue expression and substrate specificities opens up avenues for further research into these enzymes, which may be important in regulation of numerous intracellular processes, as well as for maintaining robustness in response to rapid changes in availability of energy substrates. Patients with lowered ACOT activity have recently been described (35)
; however, no patients deficient in the Type-I family of ACOT enzymes characterized in this study have yet been identified. The role of ACOT enzymes in lipid metabolism within different intracellular compartments suggests that future studies may reveal their importance in inborn errors of metabolism and lipid metabolism.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Present address: Finnish Genome Center, University of Helsinki, Helsinki 00014, Finland.
4 Present address: Biovitrum AB, Arvid Wallgrens Backe 20, Göteborg SE-413 46, Sweden.
Received for publication March 10, 2006. Accepted for publication April 17, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and 
. Proc. Natl. Acad. Sci. U. S. A. 94,4318-4323 and 
. Proc. Natl. Acad. Sci. U. S. A. 94,4312-4317 (PPAR) interacts with high affinity and is conformationally responsive to endogenous ligands. J. Biol. Chem. 280,18667-18682)-regulated peroxisomal acyl-CoA thioesterases. J. Biol. Chem. 279,21841-21848This article has been cited by other articles:
|
|
 |
|
  A. S. Andrew, V. Bernardo, L. A. Warnke, J. C. Davey, T. Hampton, R. A. Mason, J. E. Thorpe, M. A. Ihnat, and J. W. Hamilton Exposure to Arsenic at Levels Found in U.S. Drinking Water Modifies Expression in the Mouse Lung Toxicol. Sci., November 1, 2007; 100(1): 75 - 87. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Dongol, Y. Shah, I. Kim, F. J. Gonzalez, and M. C. Hunt The acyl-CoA thioesterase I is regulated by PPAR{alpha} and HNF4{alpha} via a distal response element in the promoter J. Lipid Res., August 1, 2007; 48(8): 1781 - 1791. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-J. Reilly, E. M. O'Shea, U. Andersson, J. O'Byrne, S. E. H. Alexson, and M. C. Hunt A peroxisomal acyltransferase in mouse identifies a novel pathway for taurine conjugation of fatty acids FASEB J, January 1, 2007; 21(1): 99 - 107. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
