HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 278, Issue 15, 13159-13165, April 11, 2003

This Article Abstract Full Text (PDF) All Versions of this Article: 278/15/13159    most recent M212356200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Wu, D. Articles by McKenna, R. Search for Related Content PubMed PubMed Citation Articles by Wu, D. Articles by McKenna, R. Social Bookmarking                      What's this?

Structure of Human Carnitine Acetyltransferase

MOLECULAR BASIS FOR FATTY ACYL TRANSFER^*

Donghai Wu, Lakshmanan Govindasamy^§, Wei Lian, Yunrong Gu, Thomas Kukar, Mavis Agbandje-McKenna^§, and Robert McKenna^§¶

From the  Department of Medicinal Chemistry, College of Pharmacy and ^§ Department of Biochemistry and Molecular Biology, College of Medicine, University of Florida, Gainesville, Florida 32610

Received for publication, December 4, 2002, and in revised form, January 12, 2003

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Carnitine acyltransferases are a family of ubiquitous enzymes that play a pivotal role in cellular energy metabolism. We report here the x-ray structure of human carnitine acetyltransferase to a 1.6-Å resolution. This structure reveals a monomeric protein of two equally sized / domains. Each domain is shown to have a partially similar fold to other known but oligomeric enzymes that are also involved in group-transfer reactions. The unique monomeric arrangement of the two domains constitutes a central narrow active site tunnel, indicating a likely universal feature for all members of the carnitine acyltransferase family. Superimposition of the substrate complex of a related protein, dihydrolipoyl trans-acetylase, reveals that both substrates localize to the active site tunnel of human carnitine acetyltransferase, suggesting the location of the ligand binding sites for carnitine and coenzyme A. Most significantly, this structure provides critical insights into the molecular basis for fatty acyl chain transfer and a possible common mechanism among a wide range of acyltransferases utilizing a catalytic dyad.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In eukaryotic cells, the carnitine system plays a vital role in fatty acid -oxidation and maintenance of acyl coenzyme A (acyl-CoA)¹ pools (1, 2). The important physiological functions of carnitine are made possible by the presence of carnitine/acylcarnitine transporters on cellular membranes and a group of enzymes, the carnitine acyltransferases, which catalyze the reversible transfer of acyl groups between CoA and carnitine (1, 2) as shown in Equation 1.

(Eq. 1)

The three known classes of carnitine acyltransferases differ in their acyl group specificity, subcellular localization, tissue distribution, and physiological function (3-5). Carnitine acetyltransferase has a substrate preference for short chain acyl-CoAs and is found in the mitochondrial matrix, the endoplasmic reticulum, and the peroxisome. Carnitine octanoyltransferase (COT) is predominantly localized in peroxisomes with a substrate preference for medium length acyl-CoAs. Carnitine palmitoyltransferases (CPT) are found both in the outer mitochondrial membrane (CPT I) and the mitochondrial matrix (CPT II) with a substrate preference for long chain acyl-CoAs.

Arguably, the most critical member is CPT I, which is responsible for facilitating the transfer of long chain fatty acids into the mitochondria. There are two different types of CPT I in mammalian tissues (1), a liver (L-CPT I) and a muscle isoform (M-CPT I). Both isoforms of CPT I are inhibited by malonyl-CoA, which is the committed intermediate for the biosynthesis of fatty acids. However, each isoform displays a different sensitivity to this metabolite. Malonyl-CoA inhibition regulates and balances -oxidation and the biosynthesis of fatty acids, depending on the metabolic needs of the cell. Interestingly, CPT I has an extended N terminus, which contains two trans-membrane domains, that is not found in the other carnitine acyltransferases and appears to be important not only in anchoring the enzyme in the outer mitochondrial membrane but also in modulating sensitivity of the CPT I isoforms to malonyl-CoA (1). The important metabolic roles of CPT I make it a potential drug target for diabetes, coronary heart disease, and other disorders involving abnormal fatty acid metabolism (6-9).

However, the lack of a molecular structure for any member of this family has hampered our understanding of the biology of these enzymes and the rational design of selective inhibitors. Here we report the x-ray crystal structure of human peroxisomal carnitine acetyltransferase (hpCAT), which represents the first structure for the carnitine acyltransferase family (Table I). To gain more insight into the function of hpCAT, we have also modeled the binding sites of carnitine and CoA using the superimposition of the ternary substrate complex of a structurally and functionally related protein, dihydrolipoyl trans-acetylase (E2p). The results from this work correlate well with previous experimental data and provide further information regarding the catalytic mechanism of hpCAT and ligand binding sites.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Crystallization-- Production, purification, and crystallization of hpCAT were performed as described previously (10) with the exception that selenomethionine was also incorporated into hpCAT, was crystallized, and was used for multi-wavelength anomalous dispersion (MAD) phasing. The crystals belong to orthorhombic space group P2₁2₁2₁, and its unit cell parameters are a = 137.72, b = 84.77, c = 57.86 with one molecule in the crystallographic asymmetric unit.

Data Collection-- All of the x-ray diffraction data were collected at 100 K. Before flash-freezing, the crystals were dipped in the crystallization reservoir solution to which 15% polyethylene glycol 8000 and 40% glycerol had been added. The MAD data sets were collected from a single selenomethionine crystal at the Cornell High Energy Synchrotron Source F-2 beamline using the ADSC Quantum-4 CCD detector system (Table I). The native data set were collected from a single crystal using an "in-house" R-AXIS IV++ image plate system with Osmic mirrors and a Riguku HU-H3R CU rotating anode generator (Table I). All of the diffraction images were indexed, integrated, and scaled by the HKL2000 suite programs DENZO and SCALEPACK (11).

MAD Phasing-- The conventional direct methods programs SHAKE and BAKE (SnB) and SHELXS97 (12, 13) were used to initially locate 18 selenium atom sites. The CNS software suite, version 1.1, was then used to refine the initial sites located by SnB and to identify four additional selenium sites (14). The selenomethionine diffraction data were then phased, and the solvent was flattened using CNS to produce an interpretable electron density map to a 1.6-Å resolution. The initial figure of merit of 0.49 was improved to 0.97 after density modification. The correct handedness of the selenium atom model was determined by careful inspection of electron density maps.

Model Building and Structure Refinement-- Initial model building was carried out using the auto-tracing program MAID (15), which assigned 250 residues in hpCAT. Model building was further carried out using the graphics program O (16).

The refinement of the structure was performed with CNS (14). The model was refined with CNS against MAD data in cycles of simulated annealing, positional, and individual B-factor refinement and alternated with interactive rebuilding with the program O. Once the model was at a reasonable quality as judged by refinement statistics, refinement was continued against the 1.6-Å native data set. Further cycles of refinement and model building were carried out until the R_factor and R_free converged to 0.208 and 0.232, respectively. 5% reflections were selected randomly and excluded from the refinement procedure (17). Solvent water molecules were assigned from difference Fourier maps by the use of the automated scripts implemented in CNS. At present, the model has been built from residue 9 to 599, and 472 solvent molecules have been placed. The Ramachandran plot obtained from PROCHECK (18) shows that 99.8% of the residues are in favorable regions and 0.2% of residues are in disallowed regions. Table I gives the crystallographic data, phasing, and refinement statistics. Figs. 1 and 4-7 were produced using BOBSCRIPT (19).

Sequence Alignment-- ClustalW (20) was used to perform a multiple sequence alignment of all of the known human carnitine acetyltransferases (GenBank^TM accession number P43155), COT (GenBank^TM accession number Q9UKG9), L-CPTI (GenBank^TM accession number P50416), M-CPTI (GenBank^TM accession number Q92523), and CPT II (GenBank^TM accession number P23786). Fig. 3 was generated using ALSCRIPT (21) to format the alignment and include the secondary structural elements of hpCAT. The Dali server (22) was used to identify structural homologues (optimal pairwise alignment of protein structures) between hpCAT and other proteins in the Protein Data Bank.

Structural Superimposition and Substrate Modeling-- The molecular modeling-interactive graphics program O (16) was used to superimpose E2p complexed with lipoamide and CoA (Protein Data Bank code 1EAB) onto domain I of the hpCAT structure and to model the active site location of carnitine and CoA binding sites. Carnitine was modeled into the active site of hpCAT by manually docking the chemical structure of carnitine onto the position of lipoamide and aligning their reactive -hydroxyl and sulfur groups, respectively. The reaction mechanism that we propose for hpCAT involves the formation of a tetrahedral transition state intermediate (see "Discussion" (Fig. 8). Using the program O (16), this putative tetrahedral intermediate was manually docked onto the previously modeled positions of CoA and carnitine, taking into account the bond angles and stereochemistry. Residues in hpCAT were then identified that may interact with the negatively charged oxyanion reaction intermediate.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The overall structure of hpCAT reveals a monomeric protein with dimensions of ~75 × 45 × 45 Å (Fig. 1). The tertiary structure consists of 17 helices (H1-H17) and 14 strands (S1-S14), which are arranged into two equally sized / domains, I (residues 87-385) and II (residues 9-86 and 386-599) (Fig. 2). The domains are tightly associated with each other by interconnecting loops and helices that form a central tunnel of ~10-Å diameter that transverses through the molecule. This tunnel defines a solvent-accessible surface in the center of the protein, which constitutes the putative active site for the enzyme. Seventy-six amino acids accounting for 3400 Å² (15% of the total surface area) are buried at the interface between domains I and II (Fig. 1).

View larger version (63K): [in this window] [in a new window]   Fig. 1.   Overall structure of human peroxisomal carnitine acetyltransferase. Ribbon diagram representation depicts the demarcation of domains I (blue) and II (orange). The N and C termini are indicated.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Schematic of the topology of domains I and II. The secondary structural elements are shown with sequential numbering. Triangles represent strands S1-S14, and squares represent helices H1-H17.

The hpCAT structure is composed of 39% helix (H1-H17), 13% -strand (S1-S14), and 48% loops and turns. The most noticeable common feature in domains I and II is an "open-faced" core five-stranded-mixed -sheet consisting of three centrally located parallel -strands flanked on either side by an anti-parallel -strand that adopt characteristic left-handed twists. In domain I, this motif is defined by strands S5, S4, S6, S7, and S8, and in domain II, this motif is defined by strands S14, S13, S11, S9, and S10. The -sheet of domain I has a sixth strand, S1, that runs anti-parallel to strand S8. Domain II provides an additional seventh strand, S12 (looped out between strands S11 and S13 of its five-stranded -sheet), to the domain I -sheet that is adjacent and anti-parallel to strand S1. Therefore, strands S1 and S12 create an interface between the two domains (Fig. 2).

The N-terminal region of domain II is comprised of four -helices (H1-H4). Helices H1 and H2 form an anti-parallel bundle that leads to an extended structure made up of helices H3 and H4 that cross over to domain I and contribute part of the floor and front surface of the active site tunnel. Domain I is defined from residues 87 to 385. Residues 306-320 constitute an anti-parallel pair of strands (S7 and S8) linked by a type II turn, which forms the rostra lining on the left side of the active site. An active site histidine (residue 322) is located on a turn at the end of strand S8 and is positioned in the center of the active site tunnel. This turn (residues 321-326) leads to helix H10 that lines the active site tunnel, which is followed by a loop (located behind the -sheet in domain I) that forms the upper lip of an extended surface groove. Helix H11 is next lying on the front portion of the -sheet of domain I. A final extended loop after H11 crosses into domain II. This loop leads into helix H12 that runs parallel with the -sheet motif of domain II. -Strands S11-S14 form an intricate hydrogen-bonding network with each other to facilitate the "flipping out" of strand S12 to align it with strand S1 in domain I. Four helices, H13-H16, between strands S10 and S11 create a supporting scaffold that positions the domain II -sheet open face toward the active site and constitutes the top and right lining of the tunnel. Domain II terminates with helix H17 that also supports the -sheet (Figs. 1 and 2).

The structural arrangement of hpCAT appears to be unique for an acyltransferase, although similar structural elements to portions of domains I and II have been observed in some oligomeric enzymes that catalyze similar group-transfer reactions (Fig. 3). The secondary structural elements, helices H5, H6, and H10 and strands S1, S4, S5, S6, S7, and S8 of domain I, share structural similarity to the catalytic domain of dihydrolipoyl trans-acetylase (E2p) (116 C atoms structurally align with an r.m.s. deviation of 2.0 Å) (Protein Data Bank code 1EAB) (for review see Ref. 24). In domain II, helices H12, H13, and H17 and strands S11, S13, and S14 are structurally similar to portions of ornithine decarboxylase (ORD) (55 C atoms structurally align with an r.m.s. deviation of 2.1 Å) (Protein Data Bank code 1ORD) (for review see Ref. 25).

View larger version (60K): [in this window] [in a new window]   Fig. 3.   Structural similarity and superimposition of dihydrolipoyl trans-acetylase (E2p) and ORD with hpCAT. Coil diagram shows the superimposition of E2p (blue) onto domain I and ORD (green) onto domain II of hpCAT.

In addition to structural features, E2p also displays functional similarity to hpCAT. E2p is a component of the multienzyme pyruvate dehydrogenase complex and is responsible for the transfer of acetyl groups between lipoamide and CoA. The catalytically active form of E2p is a trimer with the active site located in the subunit interface. This interaction forms an active site channel for CoA and acetyl lipoamide to bind and is the location of His⁶¹⁰, which serves as a general base for catalysis in E2p (24). A number of studies have also suggested that the carnitine acyltransferases utilize a histidine as a general base for catalysis (1). Based on their structural and functional similarity, the ternary complex of E2p with both substrates (lipoamide and CoA) was superimposed onto domain I of the hpCAT structure (see "Experimental Procedures"). The overlay of these two structures results in the direct superimposition of the catalytic residue His⁶¹⁰ of E2p (23) onto His³²² of hpCAT, which is completely conserved (Fig. 4) and has been previously identified as essential for enzymatic activity in other carnitine acyltransferases (1, 28, 29). In hpCAT, this histidine residue is located on a turn at the end of strand S8 and is positioned in the center of the active site tunnel (Fig. 6).

View larger version (68K): [in this window] [in a new window]   Fig. 4.   Sequence alignment of all human carnitine acyltransferases. The amino acid numbering and secondary structural elements are based on the structure of hpCAT. Strands S1-S14 (arrows) and helices H1-H17 (cylinders). The extended N terminus of CPT I is not shown. Domains I and II are colored blue and orange. Identical and conserved residues are indicated by red and yellow blocks, respectively. Residues known to be associated with carnitine binding and CoA binding are indicated by solid colored circles in blue and orange. Residues essential for the activity of the acyltransferases are shown in green (for review see Ref. 1).

The location of His³²² in the center of the catalytic tunnel allows access to this essential residue from either side, suggesting that the binding sites for CoA and carnitine lie on opposite sides of this tunnel (Fig. 6). Additionally, the superimposition of the ternary complex of E2p places lipoamide and CoA into the active site tunnel of hpCAT. The putative location of the separate binding sites for carnitine (Fig. 7A) and acetyl-CoA (Fig. 7B) were then identified by modeling their positions based on the locations of lipoamide and CoA using the superimposition of the E2p complex onto hpCAT. The orientation of the binding pockets allows each substrate to approach the active site tunnel from opposite sides independently, which is consistent with the rapid equilibrium random order kinetics proposed for this enzyme (30).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The two domain structural architecture of hpCAT represents a general model for the structures of other carnitine acyltransferases based on their sequence homology (Fig. 4). There is a 21-31% identity at the amino acid level among the different human carnitine acyltransferases (1). COT shares the highest homology with hpCAT followed by CPT II and CPT I. Of note are two consensus sequences that have been defined for these acyltransferases (23). These are Prosite PS00439 (residues 14-28 in hpCAT) and PS00440 (residues 299-327 in hpCAT), which form characteristic supersecondary structural motifs of an extended proline rich region in helix H1 of domain II and strands S7 and S8 of domain I, respectively (Fig. 5). Previously, PS00439 was believed to be important for carnitine binding (6); however, our structure shows that this is not the case and instead plays a structural role. On the other hand, PS00440 is located in the center of the catalytic tunnel and contains the active site histidine. The regions of highest homology are clustered in domain II, implying potential locations of common structural and/or functional significance such as the binding sites for shared substrates. On the contrary, domain I is more divergent and may contain structural differences for recognition of variable length acyl-CoAs and the physiological regulator malonyl-CoA.

View larger version (49K): [in this window] [in a new window]   Fig. 5.   Functional residues of carnitine acyltransferases. Coil diagram highlights the locations of the two consensus sequences for the carnitine acyltransferases gene family. Prosites PS00439 (residues 14-28, yellow) and PS00440 (residues 299-327, red) are represented by colored coils. Residues known to be associated with carnitine binding and CoA binding are indicated by solid colored circles in blue and orange. Residues essential for the activity of the acyltransferases are shown in green (as in Fig. 3). His322 is shown in the center of the catalytic tunnel active site.

Active Site Tunnel and Catalytic Histidine-- The side chain of His³²² of hpCAT adopts an unusual conformation (₁ = 140°; ₂ = 39°) to form an intraresidue hydrogen bond (2.7 Å) between the imidazole nitrogen N1 and the backbone carbonyl oxygen. In E2p as well as chloramphenicol acetyltransferase, which also catalyzes an acyl group transfer, a similar conformation of the active site histidine has been observed and may represent a conserved structural feature that facilitates catalysis (23, 26). This unusual angle most probably allows positioning of the imidazole N3 to align with carnitine, enabling the extraction of a proton from its primary alcohol. This is consistent with the finding that the reaction of carnitine acetyltransferase with bromoacetylcarnitine only produces labeled N3 carboxymethyl histidine (27). Additionally, the histidine at this position is highly conserved among this enzyme family, and the removal of the corresponding homologous residues, His⁴⁷³, His³⁷², and His³²⁷ in rat CPT I, CPT II, and COT, respectively (Fig. 4), has been shown to inactivate these enzymes completely (28, 29). When His³²² was changed to an alanine, a serine, or a glutamine, the mutant forms of hpCAT were completely inactive, supporting its catalytic role.²

Catalytic Dyad-- The modeling of carnitine into the active site reveals a number of residues that may be involved in binding and catalysis. Most importantly, the N3 nitrogen of the catalytic His³²² is positioned such that it can form a hydrogen bond and abstract a proton from the -hydroxyl of carnitine. In addition, a functionally conserved glutamate, Glu³²⁶, interacts with His³²² and appears to be critical for catalysis (Fig. 7A). Carbonyl oxygen 1 (O1) of Glu³²⁶ is positioned 4.03 Å from the N3 nitrogen of His³²², whereas O₂ forms a salt bridge with Arg⁴⁴³ (2.78 Å). This interaction probably further potentiates the activity of His³²² and also stabilizes the positive charge that would develop after substrate deprotonation. In almost all of the other members of the family, an aspartate is found at this position, which could functionally substitute for glutamate. In rat CPT II, substituting Asp³⁷⁶, which is homologous to Glu³²⁶ in hpCAT, to an alanine produced a completely inactive protein, illustrating the important role of this negative charge in catalysis (28). Similar interactions have been observed in other enzyme active sites that utilize a Ser-His-Asp catalytic triad including a variety of proteases and lipases as well as acetylcholine esterase (31, 32). Thus, hpCAT as well as the other carnitine acyltransferases appear to utilize a His-Glu/Asp catalytic dyad to carry out catalysis.

Proposed Binding Sites for Carnitine and CoA-- Several conserved residues in carnitine acyltransferases, in particular, Tyr⁴³¹ and Thr⁴⁴⁴ of hpCAT, lie within hydrogen-bonding distance of the O1 and O₂ of carnitine and appear to interact directly with the carnitine carboxylate group (Fig. 7A). Previously, it has been speculated that a positively charged amino acid forms a salt bridge with the carboxylate group of carnitine in the active site. In bovine COT, the mutation of a conserved arginine to asparagine decreased the binding of carnitine with a greater than 1650-fold increase in the apparent K_m(carnitine) (33). In hpCAT, the equivalent residue, Arg⁴⁹⁷, may fulfill this role by forming an electrostatic interaction with carnitine because it is ~5 Å away. Other naturally occurring and site-directed mutations have been shown to affect carnitine binding (Figs. 4 and 5) (for review see Ref. 1).

The modeled binding site of CoA (Figs. 6 and 7B) reveals that the adenine ring binds to the outer portion of the catalytic tunnel with the pantothenic arm extending into the active site. The reactive sulfur group is positioned for catalysis close to both His³²² and the -hydroxyl of carnitine. Some highly conserved residues cluster around this site and may be involved in CoA binding. The positive charge of Lys³⁹⁸ probably interacts with a negatively charged phosphate group, whereas Leu¹⁴² may bind CoA through hydrophobic interactions with the methyl groups of the pantothenic arm. Ser⁴⁰⁷ and Asp⁴⁸⁷ are within hydrogen-bonding distance of the modeled position of CoA and thus may contribute to CoA binding. Additional residues that have been shown to be associated with CoA binding and enzymatic activity in hpCAT and other members of the carnitine acyltransferases cluster in the vicinity of the active site as shown in Figs. 4 and 5 (for review see Ref. 1).

View larger version (50K): [in this window] [in a new window]   Fig. 6.   Modeled binding sites of carnitine and CoA in hpCAT. Coil diagram highlighting the regions of hpCAT shows structural homology with dihydrolipoyl trans-acetylase (red color) and ornithine decarboxylase (yellow color). Ball and stick models of carnitine and CoA are shown on opposite sides of the catalytic His322 in the active site tunnel of hpCAT.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   The active site of human peroxisomal carnitine acetyltransferase and putative binding sites of carnitine (A) and CoA (B). The residues surrounding carnitine are shown (His322, Glu326, Tyr431, Arg443, Thr444, Ser531, and Thr532) (A). The residues surrounding CoA are shown (Ser533, Leu142, and Lys398, Ser407, and Asp487) (B).

Transition State Stabilization-- The structure of hpCAT also allows the identification of another critical class of active site residues, those that contribute to the stabilization of the putative transition state intermediate. In chloramphenicol acetyltransferase, which appears to utilize a similar mechanism as hpCAT for acyl transfer, a serine and a water molecule hydrogen-bonded to a threonine stabilize the transition state (34). Interestingly, a Ser-Thr-Ser motif, which corresponds to Ser⁵³¹, Thr⁵³², and Ser⁵³³ in hpCAT, is completely conserved in this family (Fig. 4). Substitution of the first serine and threonine to an alanine in bovine COT increased the K_m toward carnitine by factors of 15 and 80 without significantly changing k_cat (35). The replacement of the last serine did not affect the K_m for carnitine but reduced k_cat by a factor of 10. This finding suggests that the first serine and threonine are involved in carnitine binding, whereas the last serine contributes to transition state stabilization. In the structure of hpCAT, all three of these residues are in the active site close to the modeled position of carnitine. Ser⁵³¹ approaches the quaternary amine group and Thr⁵³² is near the carboxylate moiety. When a putative transition state analog of carnitine and CoA is modeled into the active site, Ser⁵³³ is in the proper orientation to interact with the oxyanion of this intermediate, supporting the role of Ser⁵³³ in transition state stabilization, although other residues may also be involved.

Comparison to Other Carnitine Acyltransferases-- Determination of the structure of hpCAT also provides a more realistic model for the study of other members of the acyltransferase family. Recently, Morillas et al. (29, 36) reported a model of the active sites of CPT I and COT based on the structure of enoyl-CoA hydratase, which included ~200 amino acids surrounding the putative catalytic histidine. This model predicts that Ala³⁸¹ in CPT I (Ala²³⁸ in COT) is positioned in the vicinity of the active site histidine. This particular alanine residue is of significance because it is conserved in all of the medium and long chain carnitine acyltransferases (Fig. 4), being replaced by glycine in carnitine acetyltransferase (Gly²²⁸). Substitution of this alanine with an aspartate decreased enzymatic activity ~80%. The authors concluded that this result supported their structural model. However, the structure of hpCAT contradicts this proposed model in that Ala³⁸¹ in CPT I (Ala²³⁸ in COT) is predicted to be at the exterior end of domain I and specifically localized on a buried loop between two helices far away from the active site histidine. Thus, the decrease in activity from an aspartate substitution for Ala³⁸¹ in CPT I (Ala²³⁸ in COT) is most probably because of the deleterious effects of a buried negative charge. Therefore, it is clear that the model based on the structure of enoyl-CoA hydratase is limited and may not be relevant to the structures of the carnitine acyltransferases.

When the structure of hpCAT is used to analyze other members of the acyltransferase family, i.e. COT and CPT I and II, it becomes apparent that while the active site region is largely conserved (Fig. 4), significant structural differences are likely to occur on the walls lining the active site tunnel. These differences are predicted as a consequence of amino acid mutations and insertions and deletions within strand S1, a coil region between strands S11 and S12, as well as helix H10 and helix H16 (Figs. 1, 2, and 4). These secondary structural elements may represent the underlying structural basis for selectivity of fatty acid chain length among the carnitine acyltransferase classes. In particular, the most striking difference in the CPT I enzymes is a 13 amino acid insertion between strands S11 and S12 that runs directly through the active site. These changes in CPT I may form a flexible binding pocket that can accommodate long chain acyl groups that hpCAT cannot utilize as substrates.

Structural Basis of Malonyl-CoA Inhibition-- The structure of hpCAT can also be used as a guide to delineate the locations of residues known to be involved in malonyl-CoA inhibition despite the insensitivity of hpCAT to malonyl-CoA (Fig. 4). Studies have identified several conserved residues in the malonyl-CoA-sensitive carnitine acyltransferases. These include His²⁷⁷, His⁴⁸³, and Ala⁴⁷⁸ of rat L-CPT I and His¹³¹, His³⁴⁰, and Ala³³² of rat COT (35). Because carnitine acetyltransferase and CPT II lack these residues, this may be the reason why these enzymes are insensitive to malonyl-CoA inhibition. These residues map to helices H5 and H10 of domain I, which are situated on the same face as the solvent-accessible CoA binding surface. Malonyl-CoA possesses additional interactions with CPT I and COT at these residues and thus is able to bind these enzymes preferentially. The binding of malonyl-CoA at this site would prevent the entry of acyl-CoAs into the active site tunnel and thus inhibit enzymatic activity. Furthermore, it is known that both isoforms of CPT I have additional residues localized on the N terminus before two transmembrane domains, which are absent in hpCAT and COT, that influence malonyl-CoA sensitivity and may function by modulating the interaction of malonyl-CoA at this site (37, 38).

Proposed Reaction Mechanism for the Carnitine Acyltransferases-- Despite these structural differences, the carnitine acyltransferases seem to share a common mechanism for trans-esterification. Because chemical modification and site-directed mutagenesis studies have shown that a highly conserved histidine is essential for catalysis in all members of the carnitine/choline acyltransferase enzyme family, it has been suggested that the histidine functions as a general base to catalyze the acyl transfer reaction (1). It has also been postulated that an aspartate may be involved, leading to the suggestion of a catalytic triad-type mechanism, similar to the serine proteases as mentioned previously (28).

Based on the three-dimensional structure of hpCAT, we propose the following catalytic mechanism for this reaction utilizing a His-Glu/Asp catalytic-dyad (Fig. 8). In the forward direction, His³²² acts as a general base to deprotonate the primary alcohol of carnitine, resulting in a positively charged histidine. Glu³²⁶ serves to polarize the histidine to increase catalytic activity while also stabilizing the positive charge that develops on the histidine ring. Deprotonation of carnitine facilitates the nucleophilic attack of the carbonyl carbon of the thioester bond of acetyl coenzyme A, resulting in the formation of a tetrahedral intermediate. In the active site Ser⁵³³ helps to stabilize the negatively charged oxyanion intermediate. The transition state intermediate then collapses to release acetylcarnitine and free coenzyme A. This catalytic mechanism is likely conserved across the carnitine/choline acyltransferase family because of the high degree of reaction similarity and the nature of these conserved residues.

View larger version (9K): [in this window] [in a new window]   Fig. 8.   Carnitine acetyltransferase catalytic mechanism. The diagram shows the general base-catalyzed nucleophilic attack of the thioester of acetyl-CoA by the -hydroxyl group of carnitine and the formation of the putative transition state tetrahedral intermediate. The catalytic residues His322 and Glu326 as well as carnitine, acetyl-CoA, CoA, and acetylcarnitine are indicated.

In summary, the first structure of a carnitine acyltransferase, hpCAT, provides a number of important structural insights into the catalytic mechanism of this group of enzymes as well as the putative active site residues involved in substrate recognition and potential binding sites for the physiological regulator malonyl-CoA. Additionally, the structure of hpCAT provides a valuable model for structural determination of other members of the carnitine acyltransferases as well as the highly related protein, choline acetyltransferase, which is the biosynthetic enzyme for the neurotransmitter acetylcholine (39). The structure of hpCAT also serves as a model for the rational design of novel and subtype selective drugs for the treatment of heart disease, diabetes, and other related diseases.

	ACKNOWLEDGEMENTS

We thank all of the staff at the Cornell High Energy Synchrotron Source for their help and support, especially Chris Heaton, Marian Szebenyi, and Irina Kriksunov for help at the F2 station during x-ray data collection. We also thank David Levitte for providing MAID program and Nathan Bryant for help with the preparation of figures.

	FOOTNOTES

^* This work was supported in part by the University of Florida, College of Medicine start-up funds (to R. M.) and National Institutes of Health Grant R01 GM58197 (to D. H. W.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and the structure factors (code 1NM8) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

^¶ To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, College of Medicine, P. O. Box 100245, University of Florida, Gainesville, Florida 32610. Tel.: 352-392-5696; Fax: 352-392-3422; E-mail: rmckenna@.ufl.edu.

Published, JBC Papers in Press, January 31, 2003, DOI 10.1074/jbc.M212356200

² Y. Gu, T. Kukar, and D. Wu, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: Acyl-CoA, acyl coenzyme A; hpCAT, human peroxisomal carnitine O-acetyltransferase; CPT, carnitine palmitoyltransferase; L-CPT-I, mitochondrial liver isoform of CPT I; M-CPT-I, mitochondrial muscle isoform of CPT I; r.m.s., root mean square; COT, peroxisomal carnitine O-octanoyltransferase; H, helix; S, strand; MAD, multi-wavelength anomalous dispersion; ORD, ornithine decarboxylase.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. Faye, C. Esnous, N. T. Price, M. A. Onfray, J. Girard, and C. Prip-Buus Rat Liver Carnitine Palmitoyltransferase 1 Forms an Oligomeric Complex within the Outer Mitochondrial Membrane J. Biol. Chem., September 14, 2007; 282(37): 26908 - 26916. [Abstract] [Full Text] [PDF]

				Y.-S. Hsiao, G. Jogl, and L. Tong Crystal Structures of Murine Carnitine Acetyltransferase in Ternary Complexes with Its Substrates J. Biol. Chem., September 22, 2006; 281(38): 28480 - 28487. [Abstract] [Full Text] [PDF]

				S. H. Lawrence, K. B. Luther, H. Schindelin, and J. G. Ferry Structural and Functional Studies Suggest a Catalytic Mechanism for the Phosphotransacetylase from Methanosarcina thermophila J. Bacteriol., February 1, 2006; 188(3): 1143 - 1154. [Abstract] [Full Text] [PDF]

				X. Ma, J. Koepke, S. Panjikar, G. Fritzsch, and J. Stockigt Crystal Structure of Vinorine Synthase, the First Representative of the BAHD Superfamily J. Biol. Chem., April 8, 2005; 280(14): 13576 - 13583. [Abstract] [Full Text] [PDF]

				G. Jogl, Y.-S. Hsiao, and L. Tong Crystal Structure of Mouse Carnitine Octanoyltransferase and Molecular Determinants of Substrate Selectivity J. Biol. Chem., January 7, 2005; 280(1): 738 - 744. [Abstract] [Full Text] [PDF]

				A. G. Cordente, E. Lopez-Vinas, M. I. Vazquez, J. H. Swiegers, I. S. Pretorius, P. Gomez-Puertas, F. G. Hegardt, G. Asins, and D. Serra Redesign of Carnitine Acetyltransferase Specificity by Protein Engineering J. Biol. Chem., August 6, 2004; 279(32): 33899 - 33908. [Abstract] [Full Text] [PDF]

				Y.-S. Hsiao, G. Jogl, and L. Tong Structural and Biochemical Studies of the Substrate Selectivity of Carnitine Acetyltransferase J. Biol. Chem., July 23, 2004; 279(30): 31584 - 31589. [Abstract] [Full Text] [PDF]

				S. Gobin, L. Thuillier, G. Jogl, A. Faye, L. Tong, M. Chi, J.-P. Bonnefont, J. Girard, and C. Prip-Buus Functional and Structural Basis of Carnitine Palmitoyltransferase 1A Deficiency J. Biol. Chem., December 12, 2003; 278(50): 50428 - 50434. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 278/15/13159    most recent M212356200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed