Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Identification of Conserved Amino Acid Residues in Rat Liver Carnitine Palmitoyltransferase I Critical for Malonyl-CoA Inhibition

Identification of Conserved Amino Acid Residues in Rat Liver Carnitine Palmitoyltransferase I... THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 11, Issue of March 14, pp. 9058 –9063, 2003 © 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Identification of Conserved Amino Acid Residues in Rat Liver Carnitine Palmitoyltransferase I Critical for Malonyl-CoA Inhibition MUTATION OF METHIONINE 593 ABOLISHES MALONYL-CoA INHIBITION* Received for publication, September 30, 2002, and in revised form, December 17, 2002 Published, JBC Papers in Press, December 23, 2002, DOI 10.1074/jbc.M209999200 Montserrat Morillas‡§, Paulino Go ´ mez-Puertas§¶, Assia Bentebibel‡, Eva Selle ´ s‡, Nuria Casals**, Alfonso Valencia¶, Fausto G. Hegardt‡ ‡‡, Guillermina Asins‡, and Dolors Serra‡ From the ‡Department of Biochemistry and Molecular Biology, University of Barcelona, School of Pharmacy, E-08028 Barcelona, Spain, the ¶Protein Design Group, National Center for Biotechnology, Consejo Superior de Investigaciones Cientı ´ficas, Cantoblanco, E-28049 Madrid, Spain, and the **Department of Biochemistry and Molecular Biology, International University of Catalonia, 08190 Sant Cugat, Spain Carnitine palmitoyltransferase (CPT) I, which cata- The enzyme carnitine palmitoyltransferase (CPT) I cata- lyzes the conversion of palmitoyl-CoA to palmitoylcar- lyzes the conversion of long chain fatty acyl-CoAs to acylcar- nitine facilitating its transport through the mitochon- nitines, which is the first step in the transport of fatty acyl-CoA drial membranes, is inhibited by malonyl-CoA. By using groups from the cytosol to mitochondria where they undergo the SequenceSpace algorithm program to identify -oxidation. This reaction is inhibited by malonyl-CoA, and so amino acids that participate in malonyl-CoA inhibition this enzyme could be the most physiologically important regu- in all carnitine acyltransferases, we found 5 conserved latory step in mitochondrial fatty acid oxidation (1). This proc- 314 464 478 593 608 amino acids (Thr , Asn , Ala , Met , and Cys , ess allows the cell to signal the relative availability of lipid and rat liver CPT I coordinates) common to inhibitable malo- carbohydrate fuels in liver, heart, skeletal muscle, and pancre- nyl-CoA acyltransferases (carnitine octanoyltransferase atic -cell (2). The mechanism of malonyl-CoA inhibition can be and CPT I), and absent in noninhibitable malonyl-CoA acyl- potentially mimicked by pharmacological malonyl-CoA-related transferases (CPT II, carnitine acetyltransferase (CAT) and agents for the treatment of metabolic disorders such as diabe- choline acetyltransferase (ChAT)). To determine the role of tes, insulin resistance, and coronary heart disease (3). these amino acid residues in malonyl-CoA inhibition, we Mammals express two isoforms of CPT I, a liver isoform prepared the quintuple mutant CPT I T314S/N464D/A478G/ (L-CPT I) and a heart/skeletal muscle isoform (M-CPT I), which M593S/C608A as well as five single mutants CPT I T314S, are the products of two different genes (4, 5). The identity in N464D, A478G, M593S, and C608A. In each case the CPT I amino acids residues is high (62%) but they are differentially amino acid selected was mutated to that present in the regulated by malonyl-CoA. The L-CPT I isoform is inhibited by same homologous position in CPT II, CAT, and ChAT. Be- malonyl-CoA to a much lesser extent than the M-CPT I isoform cause mutant M593S nearly abolished the sensitivity to ma- (the IC value for M-CPT I is about 2 orders of magnitude lonyl-CoA, two other Met mutants were prepared: M593A and M593E. The catalytic efficiency (V /K )ofCPTIin lower than for L-CPT I) (6). This property is probably involved max m mutants A478G and C608A and all Met mutants toward in the finer regulation of fatty acid oxidation in heart and carnitine as substrate was clearly increased. In those CPT I skeletal muscle in comparison to liver. proteins in which Met had been mutated, the malonyl- From studies on the pH dependence of the affinity of CPT I CoA sensitivity was nearly abolished. Mutations in Ala , for its substrate and from the ability of palmitoyl-CoA to dis- 608 314 Cys , and Thr to their homologous amino acid residues place [ C]malonyl-CoA bound to skeletal muscle mitochondria in CPT II, CAT, and ChAT caused various decreases in ma- it was hypothesized (7) that the palmitoyl-CoA and malonyl- lonyl-CoA sensitivity. Ala is located in the structural CoA bind at different sites. A number of studies have shown model of CPT I near the catalytic site and participates in that in rat liver CPT I there are two malonyl-CoA binding sites: the binding of malonyl-CoA in the low affinity site (Morillas, one with greater capacity for binding and regulation of the M., Go ´ mez-Puertas, P., Rubı ´, B., Clotet, J., Arin ˜ o, J., inhibitor and not susceptible to competition from acyl-CoA, Valencia, A., Hegardt, F. G., Serra, D., and Asins, G. (2002) which behaves as an allosteric component (8 –12); and a second J. Biol. Chem. 277, 11473–11480). Met may participate in acyl-CoA binding site, which is located near the catalytic the interaction of malonyl-CoA in the second affinity site, site (13). whose location has not been reported. Various groups have attempted to establish the basis of the L-CPT I/malonyl-CoA interactions. The probable binding sites of malonyl-CoA in L-CPT I were deduced to be at the C termi- * This work was supported in part by Direccio ´ n General de Investi- gacio ´ n Cientı ´fica y Te ´ cnica, Spain, Grant BMC2001-3048 and Ajuts de nus after preparation of several L-CPT I chimeras whose IC Suport als Grups de Recerca de Catalunya Grant 2001SGR-00129 (to values for malonyl-CoA corresponded to the C-terminal region F. G. H.) and the Marato ´ de TV3. The costs of publication of this article (14) of the chimera. However, the N terminus of L-CPT I was were defrayed in part by the payment of page charges. This article must also shown to influence the enzyme/inhibitor interaction. Mu- therefore be hereby marked “advertisement” in accordance with 18 3 5 140 U.S.C. Section 1734 solely to indicate this fact. tation of Glu , His ,orHis produced a loss of malonyl-CoA § Contributed equally to the results of this study. Recipient of a fellowship from the Ministerio de Ciencia y Tecnolo- gı ´a, Spain. The abbreviations used are: CPT, carnitine palmitoyltransferase; ‡‡ To whom correspondence should be addressed: Dept. of Biochem- L-CPT I, liver isoform of carnitine palmitoyltransferase I; M-CPT I, istry and Molecular Biology, School of Pharmacy, Diagonal 643, muscle isoform of carnitine palmitoyltransferase I; CAT, carnitine E-08028 Barcelona, Spain. Tel.: 34-93-402-4523; Fax: 34-93-402-4520; acetyltransferase; COT, carnitine octanoyltransferase; ChAT, choline E-mail: [email protected]. acetyltransferase. 9058 This paper is available on line at http://www.jbc.org This is an Open Access article under the CC BY license. Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity 9059 palmitoyltransferase activity was determined by the radiometric sensitivity (15, 16). In addition, the removal of the segment method as described in Ref. 19 with minor modifications. The sub- comprised between amino acids 1 and 18 in L-CPT I and 1–28 strates were L-[methyl- H]carnitine and palmitoyl-CoA. Enzyme activ- in M-CPT I produced a decrease in malonyl-CoA sensitivity, ity was assayed for 4 min at 30 °C in a total volume of 200 l. which emphasizes the importance of the N terminus before the For determination of the K for carnitine, palmitoyl-CoA was fixed at first transmembrane region as a modulator of the malonyl-CoA 135 M (for L-CPT I). For determination of the K for acyl-CoA, carni- inhibition (17, 18). On the basis of these results, it was pro- tine concentration was fixed at 400 M. When malonyl-CoA inhibition was assayed, increasing concentrations of malonyl-CoA were included. posed that the two malonyl-CoA inhibitable domains might be The IC , defined as the malonyl-CoA concentration that produces 50% located at the C terminus as suggested by several kinetic stud- inhibition of enzyme activity, was determined using 50 M palmitoyl- ies. The development of a CPT I catalytic core model (19) CoA and 400 M carnitine. K was estimated by analyzing the data allowed us to assign the low affinity binding site to a domain from three experiments using the program Enzifit (Biosoft), and IC near the catalytic channel in which palmitoyl-CoA is bound was calculated by Excel software using linear regression analysis. containing the catalytic acyl-CoA binding domain (20) Values reported in the text are the means and standard deviations of three to five determinations. Curve fitting was carried out using Excel Here we used the SequenceSpace algorithm program to iden- 314 464 478 593 software. All protein concentrations were determined using the Bio-Rad tify five amino acid residues (Thr , Asn , Ala , Met , protein assay with bovine albumin as standard. and Cys ), which may contribute to the sensitivity of CPT I to Immunological Techniques—Western blot analysis was performed as malonyl-CoA. The proposal is based on the finding that they described (19). The antibody for rat L-CPT I was kindly given by are present in malonyl-CoA-inhibitable CPT I ((isoforms L- and Dr. V. A. Zammit (Hannah Research Institute, Ayr, Scotland, United M-) and COT from various organisms and absent in noninhib- Kingdom) and was directed against peptide 428 – 441, in the cytosolic catalytic C-terminal domain. itable acyltransferases (CPT II, CAT, and ChAT). Mutation of these amino acids to their counterparts in CPT II showed that RESULTS mutation of Met by itself, M593S, or the quintuple mutant Residues Conserved in Malonyl-CoA Inhibited Versus Nonin- containing the M593S mutation, T314S/N464D/A478G/M593S/ hibited Carnitine-Choline Acyltransferases—An exhaustive C608A, or other Met point mutants such as M593A and analysis of the presence of residues shared by all the malonyl- M593E nearly abolished malonyl-CoA sensitivity of L-CPT I. CoA-regulated enzymes of the carnitine-choline acyltrans- The remaining mutated amino acids showed slight, varied sen- ferase family versus the malonyl-CoA nonregulated members of sitivity to malonyl-CoA inhibition. the same family was performed using the algorithm Sequence- EXPERIMENTAL PROCEDURES Space (23, 24). This method uses a vectorial representation of each protein sequence as a point in a multidimensional space Tree-determinants Analysis—Sequences of proteins from the carni- tine-choline acyltransferase family were obtained using BLAST (21). (SequenceSpace) and multivariate statistics, principal compo- Multiple alignment was performed using ClustalW (22). The analysis of nent analysis, to allow reduction of the number of dimensions. conserved differences (tree-determinants) between malonyl-CoA-regu- This representation allows us not only to define clusters of lated (L-CPT I, M-CPT I, and COT) and nonregulated (CPT II, CAT, and proteins according to specific properties by choosing the appro- ChAT) acyltransferases, using multivariate statistics for low-dimen- priate axes defined by the highest corresponding eigenvalues sional representation, was done using the SequenceSpace algorithm (also known as proper values), but also to project the individual (23, 24). Graphics of vectors representing protein sequences and indi- vidual residues from the multiple alignment were performed using the residues on the same axes, and thus trace the positions con- Sequence Space Java-based viewer (www.industry.ebi.ac.uk/SeqSpace). served in the subfamilies defined. The main advantage of this wt Construction of Site-directed Mutants—Plasmids pYESLCPTI and method is the possibility of predicting which residues may be A478G pYESLCPT were obtained as previously described (20). Plasmids responsible for the specific characteristics of each protein sub- T314S N464D M593S M593A pYESLCPT , pYESLCPT , pYESLCPT , pYESLCPT , M593E C608A family or group of subfamilies as has been reported previously pYESLCPT , and pYESLCPT were constructed using the for short- and medium-long substrate specificity for the carni- QuikChange polymerase chain reaction-based mutagenesis procedure wt (Stratagene) with the pYESLCPT plasmid as template. The following tine-choline acyltransferases protein family (19, 20) or effector primers were used: primer T314S.for 5-GGGAGCGACTCTTCAATAG- recognition by some members of the Ras superfamily (25). TTCCCGGATCCCTGGG-3, primer T314A.rev 5-CCCAGGGATCCG- The two-dimensional projection of sequence vectors on the GGAACTATTGAAGAGTCGCTCCC-3, primer N464D.for 5-CACCTT- plane defined by the axes corresponding to eigenvalues 2 and 4 TGTTGTCTTCAAAGACAGCAAGATAGGC-3, primer N464D.rev 5- showed clustering of the enzyme subfamilies according to their GCCTATCTTGCTGTCTTTGAAGACACCAAAGGTG-3, primer malonyl-CoA inhibition properties (Fig. 1A). Proteins whose M593S.for 5-CCTCACATATGAGGCCTCCAGTACCCGGCTCTTCCG AGAAGG-3, primer M593S.rev 5-CCTTCTCGGAAGAGCCGGGTAC- activity is not regulated by malonyl-CoA (CPT II, CAT, and TGGAGGCCTCATATGTGAGG-3, primer M593A.for 5-CCTCAC- ChAT subfamilies) were grouped, whereas the sequences of the ATATGAGGCCTCCGCGACCCGGCTCTTCCGAGAAGG-3, primer proteins regulated by malonyl-CoA (COT, L-CPT I, and M-CPT M593A.rev 5-CCTTCTCGGAAGAGCCGGGTCGCGGAGGCCTCATA I) occupy separate, and opposite, zones. The projection of the TGTGAGG-3, primer M593E.for 5-CCTCACATATGAGGCCTCCGA- individual amino acid residues on the same plane (Fig. 1B) GACCCGGCTCTTCCGAGAAGG-3, primer M593E.rev 5-CCTTCTC- revealed the amino acids responsible for this segregation might GGAAGAGCCGGGTCTCGGAGGCCTCATATGTGAGG-3, primer be responsible for the susceptibility to malonyl-CoA of the C608A.for 5-GAGACTGTACGCTCCGCCACTATGGAGTCCTGC-3, and C608A.rev 5-GCAGGACTCCATAGTGGCGGAGCGTACAGTCT- corresponding enzymes. Five of these amino acids (Thr , 464 478 593 608 C-3 (the mutated nucleotides are underlined). The plasmid Asn , Ala , Met , and Cys ) were present in all malonyl- T314S/N464D/A478G/M593S/C608A pYESLCPTI was obtained by the same CoA inhibitable carnitine acyltransferases and absent in the method, but performing each new mutation stepwise starting on plas- nonmalonyl-CoA inhibitable acyltransferases (CPT II, CAT, T314S mid pYESLCPT . The appropriate substitutions as well as the and ChAT from several species). Fig. 2 shows the sequence absence of unwanted mutations were confirmed by sequencing the alignment of three fragments of the C-terminal region of vari- inserts in both directions with an Applied Biosystems 373 automated DNA sequencer. ous acyltransferases. We can also observe that those enzymes Expression of L-CPT I in Saccharomyces cerevisiae—The expression that are not inhibitable by malonyl-CoA (CPT II, CAT, and of the constructs containing L-CPT I wild type and mutants (see above) ChAT) show the same amino acids in these positions, which are in yeast cells and the preparation of the cell extracts were performed as different from those observed in inhibitable malonyl-CoA acyl- described in Ref. 19. S. cerevisiae was chosen as an expression system carnitines. As an example the positions and amino acids of CPT for L-CPT I wild type and the mutants because it does not have endog- 223 363 377 490 505 II are given: Ser , Asp , Gly , Ser , and Ala (Fig. 2). enous CPT I activity. Determination of Carnitine Acyltransferase Activity—Carnitine Expression of Wild Type and Mutants in S. cerevisiae—We 9060 Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity FIG.1. Sequence space analysis of the carnitine-choline acyltransferase family. A, protein sequences projected onto the plane defined by principle axes 2 and 4. This two-dimensional space allows separation of protein subfamilies according to their malonyl-CoA regulatory characteristics; CPT II, CAT, and ChAT (CACP) enzymes (malonyl-CoA insensitive) are clustered to the lower left corner of the panel, whereas CPT I (L- and M-isoforms) and COT (malonyl-CoA inhibitable enzymes) are projected on the upper and right areas of the vertical and horizontal axes, respectively. B, the sequence of each subfamily is represented as a vector point in a multidimensional space (sequence space), with residue positions and types as the basic dimensions. Single residues completely conserved in CPT I or COT subfamilies are projected in the same position as their corresponding protein sequences. Residues conserved in both groups of malonyl-CoA-regulated enzymes occupy the upper right corner, whereas the residues conserved in the nonregulated cluster of acyltransferases (CPT II, CAT, and ChAT) occupy the opposite one. Residues located in alignment positions present in both opposite corners of the two-dimensional plot are responsible for protein cluster segregation and are predicted to be involved in malonyl-CoA sensitivity. FIG.2. Alignment of representative sequences of mammalian carnitine-choline acyltransferases. Amino acid sequence of 18 repre- sentative members of the malonyl-CoA-insensitive enzymes, CPT II (CPT2) from rat, mouse, and human; CAT (CACP) from human and mouse; ChAT (CLAT) from human, pig, rat, and mouse; and malonyl-CoA inhibitable enzymes L-CPT I (CPT1) from rat, mouse, and human; M-CPT I (CPTM) from human, rat, and mouse; and COT (OCTC) from human, rat, and bovine, were obtained from the SwissProt data bank and aligned using ClustalW (22). A, schematic representation of the position of the tree-determinant residues obtained using the SequenceSpace algorithm (23, 223 314 363 464 377 478 490 593 505 608 24) on the rat CPT II and L-CPTI proteins: Ser /Thr , Asp /Asn , Gly /Ala , Ser /Met , Ala /Cys . Transmembrane regions of 372 473 L-CPT I are also represented (tm1 and tm2). Position of the catalytic histidine (His /His ) as well as the previously three-dimensional modeled core of the proteins, 2dub, (amino acids 368 –567 of L-CPT I) (19, 20), are indicated.B, selected regions of the multiple alignment of the protein family. Subfamily conserved residues according to malonyl-CoA regulation are shadowed. Position of catalytic histidine (arrowhead) is also indicated. 1 1 prepared a quintuple mutant, T314S/N464D/A478G/M593S/ (values ranged between 14 and 20 nmol min mg protein ) C608A, and separately, the point mutants T314S, N464D, when the protein was overexpressed 20 h after galactose induc- A478G, M593S, and C608A and all were expressed in S. cer- tion, showing that the various mutations assayed produce evisiae. After we observed that mutant M593S nearly abolished small changes in L-CPT I activity (Table I). the sensitivity to malonyl-CoA (see below), new point Met mu- All mutants exhibited standard saturation kinetics when the tants were prepared: M593A and M593E. All transformed carnitine concentration was varied relative to a constant con- yeast cells expressed a protein with the same molecular mass centration of the second substrate, palmitoyl-CoA, and when (88 kDa) and the mutant enzymes were expressed in roughly palmitoyl-CoA concentration was varied relative to a constant the same proportion per milligram of protein as the wild type carnitine concentration, a property identical to that of the wild L-CPT I as deduced from immunoblot analysis (data not type L-CPT I (Fig. 3). The quintuple mutant produced small shown). changes in the kinetic constants for carnitine and palmitoyl- Kinetic Properties of CPT I Wild Type and Mutants—L-CPT CoA as substrates (Table I). Catalytic efficiency (V /K ) was max m I activities of the wild type, quintuple mutant variant T314S/ increased by a factor of 2.6 (carnitine) and 2.2 (palmitoyl-CoA). N464D/A478G/M593S/C608S, and point mutants were similar The catalytic efficiency for carnitine as substrate of those point Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity 9061 TABLE I Enzyme activity, malonyl-CoA sensitivity and kinetic parameters of carnitine palmitoyltransferase I in Saccharomyces cerevisiac cells expressing CPI I wild type and point mutants, T314S, N464D, A478G, C608A, M593S, M593A, M593E and quintuple mutant T314S/N464D/A478G/M593S/C608A (QM) Extracts from yeast expressing wild type and several mutants of L-CPT I were assayed for activity, malonyl-CoA sensitivity, and kinetics as described under “Experimental Procedures.” The results are the mean  S.D. of at least three independent experiments with different prepara- tions. In parentheses are shown the increase (in-fold number) of the catalytic efficiency (V /K ) versus to that of the wild type. max m K V Catalytic efficiency m max IC L-CPT I Activity malonyl-CoA Carnitine Palmitoyl-CoA Carnitine Palmitoyl-CoA Carnitine Palmitoyl-CoA 1 1 1 1 nmol min mg protein m nmol min mg protein V /K max m Wild-type 17.7  0.9 12.3 127  4.5 4.9  0.3 6.6  0.8 6.3  0.4 0.05 ( 1) 1.28 ( 1) T314S 14.4  2.1 15.0 88.2  2.4 1.7  0.5 12.8  0.1 6.8  0.1 0.15 ( 2.8) 3.98 ( 3.1) N464D 20.1  3.1 8.7 69.5  8.2 4.1  0.4 19.4  1.4 18.9  3.6 0.28 ( 5.6) 4.63 ( 3.6) A478G 16.7  0.7 39.5 327  41 15.1  4.0 69.8  9.3 50.4  17 0.21 ( 4.1) 3.34 ( 2.6) C608A 17.3  1.7 27.5 51.6  4.0 24.3  2.0 23.7  5.0 67.5  9.0 0.46 ( 8.8) 2.78 ( 2.2) M593S 17.0  0.8 319 124  0.8 7.4  1.2 133  18 20.9  1.6 1.07 ( 21) 2.84 ( 2.2) M593A 17.2  0.9 155 56.3  2.1 6.1  0.2 32.5  4.6 30.3  4.7 0.58 ( 12) 4.81 ( 3.7) M593E 14.1  1.8 220 150  3.4 6.3  0.5 31.3  2.6 27.5  1.8 0.21 ( 4.2) 4.37 ( 3.4) QM 13.6  1.6 258 95.7  2.8 4.6  1.5 13.1  4.7 13.0  6.3 0.14 ( 2.6) 2.84 ( 2.2) Even at concentrations as high as 100 M malonyl-CoA the CPT I quintuple mutant maintained 80% of the activity of the control without malonyl-CoA. We then addressed the individual responsibility of the sep- arate CPT I mutants for the malonyl-CoA sensitivity. Mutants T314S, N464D, M593S, and C608A expressed in S. cerevisiae were incubated with increasing amounts of malonyl-CoA, and CPT I activity was determined. Mutant A478G had been pre- viously studied in Ref. 20 and showed decreased sensitivity to malonyl-CoA (IC of 39.5 versus 12.3 M of the wild type). The kinetics of inhibition by malonyl-CoA depended on the mutant considered. Whereas mutant M593S (Fig. 4A) showed very low sensitivity at malonyl-CoA inhibition (IC of 319 M), the other mutations produced varied changes in malonyl-CoA sensitivity. L-CPT I C608A slightly modified the sensitivity to malonyl-CoA (IC is 27.5 M), the change in IC of mutant 50 50 T314S was small, whereas N464D showed similar sensitivity to malonyl-CoA to the wild type (Fig. 4B and Table I). Because the highest changes in sensitivity to malonyl-CoA and K values for carnitine were observed in the methionine mutants (point and quintuple mutants), we additionally prepared two new FIG.3. Kinetic analysis of wild type and different mutants of L-CPT I. Yeast extracts (10 g of protein) of (A and C) wild type (open mutants: M593A and M593E to examine whether Met was circles) and mutants M593S (open triangles), M593A (black rhombus), essential to the malonyl-CoA interaction in L-CPT I. Results M593E (black squares), and (C and D) T314S (open rhombus), N464D show that the sensitivity to malonyl-CoA was also nearly abol- (open squares), A478G (black squares), C608A (black triangles), and ished in these mutants (Fig. 4A) (IC values of 155 and 220 M, quintuple mutant T314S/N464D/A478G/M593S/C608A (black circles) 50 were incubated at increasing concentrations of carnitine (A and B) and respectively) as in the M593S mutant, confirming the essential palmitoyl-CoA (C and D). role of Met in this interaction. DISCUSSION mutants that altered the sensitivity to malonyl-CoA increased We attempted to identify the amino acids in the C-terminal (see below). The catalytic efficiency of the methionine mutants domain of L-CPT I that are responsible for the inhibition of the increased between 4.2- and 21-fold, C608A increased 8.8-fold, catalytic activity by malonyl-CoA. Over many years much work and A478G increased 4.1-fold. T314S, which produced a small has been done to identify the domains in L-CPT I that may bind change in malonyl-CoA sensitivity (see below), increased the malonyl-CoA. Different groups have tested different empirical V /K value by only 2.8, whereas in N464D, in which the hypotheses and mutated amino acids, mostly in the amino- max m sensitivity to malonyl-CoA was unchanged (see below), terminal region of L-CPT I. The results have shown that this the catalytic efficiency was modified by a factor of 5.6. domain plays a role in the regulation of CPT I by malonyl-CoA, An analogous tendency was also observed in K for palmi- because in some cases the sensitivity to the inhibitor is toyl-CoA but the changes were smaller. K values for palmi- impaired. toyl-CoA were 24.3, 15.1, 7.4, 6.1, and 6.3 for mutants C608A, A different approach was employed by our group very re- A478G, M593S, M593A, and M593E, respectively (K value for cently. This was based on the conservation of two histidine the wild type was 4.9) (Table I). Catalytic efficiencies for palmi- residues, which are present in the inhibitable malonyl-CoA toyl-CoA as substrate increased in all mutants, the values carnitine acyltransferases (CPT I and COT) and absent in ranging between 2.78 and 4.81 (Table I). noninhibitable enzymes (CPT II and CAT). Mutation of both Inhibition of CPT I Wild Type and Mutants by Malonyl- histidines resulted in the abolition of malonyl-CoA sensitivity CoA—When inhibitory kinetics versus increasing concentra- in COT (26). Analogous results were observed in CPT I when its tions of malonyl-CoA was performed, the quintuple mutant concentration at the mitochondrial membranes was not high. practically abolished the sensitivity toward malonyl-CoA (IC Mutation of other amino acids in the domain proximal to the 478 479 of 258 versus 12.3 M of the wild type) (Fig. 4B and Table I). catalytic site (Ala and Pro ) indicated that a malonyl-CoA- 9062 Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity FIG.4. Effect of malonyl-CoA on the activity of yeast overexpressed L-CPT I (wild type) and several mutants. A, L-CPT I wild type (open circles) and point methionine mutants M593S (black circles), M593A (black rhombus), M593E (black squares), and B, quintuple mutant (QM) (black circles) and point mutants T314S (open circles), N464D (open rhombus, broken line), A478G (open triangles, broken line), and C608A (open squares) overexpressed in yeast were incubated with increasing concentrations of malonyl-CoA and the enzyme activity was measured. Data are expressed relative to control values in the absence of inhibitor (100%) as the mean of three independent measurements. inhibitable domain was probably the low-affinity malonyl-CoA case, it appears that Met is critical in the interaction of binding site. Our previous studies showed that the location of malonyl-CoA with L-CPT I. malonyl-CoA in the structural model was compatible with com- It was of interest to measure the kinetic constants of all CPT petition of the inhibitor versus the substrate in the malonyl- I mutants. Several authors reported the competition between CoA low affinity binding site (20). malonyl-CoA and carnitine (27, 28). The tissues in which the The site-directed mutagenesis study used here to identify sensitivity of CPT I to malonyl-CoA is highest are those that amino acids responsible for malonyl-CoA inhibition is based on require the highest concentration of carnitine to drive the re- the comparison of the sequences in a range of carnitine and action and the requirement for carnitine and sensitivity to choline acyltransferases, taking the positive or negative sensi- malonyl-CoA appears to be inversely related. The authors con- tivity to malonyl-CoA as a discriminatory criterion. The bio- cluded that the sites to which the two metabolites bind are computing study has shown that five amino acids are present closely associated (27, 7). Studies by Bird and Saggerson (28) in all CPT I (isoforms L- and M-) and in COT from various showed on the one hand that malonyl-CoA reduced the effec- organisms and that they are absent not only in other nonma- tiveness of carnitine as substrate, and on the other hand, that lonyl-CoA-inhibitable carnitine acyltransferases but also in carnitine might diminish the regulatory effect of malonyl-CoA 314 464 ChAT. In rat L-CPT I these amino acids are Thr , Asn , (29). Although a clear mechanism for this competition could not 478 593 608 Ala , Met , and Cys . The corresponding positional amino be established, the data strongly supported this idea. In the 223 363 377 acids in CPT II, CAT, and ChAT are Ser , Asp , Gly , present study the various CPT I mutants have altered K or 490 505 Ser , and Ala , respectively (coordinates of rat CPT II). V for carnitine. Whereas the K for C608A was half of the max m Therefore, we considered it highly probable that these amino wild type, its V was 3.6-fold higher. The mutant M593S had max acids were involved in the interaction of malonyl-CoA. Results the same K value for carnitine as the wild type but its V m max confirmed in part this supposition. The quintuple mutant re- increased 20-fold. The mutant A478G increased both the K duced malonyl-CoA sensitivity almost completely (80% activity value and the V with respect to the wild type values. The max at 100 M malonyl-CoA (which is outside the physiological relationship between these values and catalysis is best re- range)), supporting the initial hypothesis. The results obtained vealed in the term catalytic efficiency. This term as calculated using separate single mutants indicate that not all of these by the V /K ratio varies considerably among different mu- max m amino acids have the same role in malonyl-CoA inhibition. tants. Carnitine catalytic efficiencies for mutants M593S, Whereas M593S nearly abolished the sensitivity to malonyl- M593A, M593E, C608A, and A478G increased 21-, 12-, 4.2-, CoA like the quintuple mutant, A478G increased the IC from 8.8-, and 4.1-fold with respect to the wild type. This means that 12 to 39.5 M (20). The other amino acids are less responsible mutations designed to decrease malonyl-CoA sensitivity for the inhibition. strongly modified the catalytic efficiency of CPT I mutants The relevance of Met as a critical amino acid for malonyl- measured in the absence of malonyl-CoA. Interestingly, the CoA sensitivity was confirmed by the results of mutation to increase in catalytic efficiency appears to be roughly propor- other two amino acids, Ala and Glu. The mutants equally tional to the extent of the alteration in malonyl-CoA sensitivity. showed diminished sensitivity to malonyl-CoA like mutant The IC values for malonyl-CoA run in the same direction to M593S. Met , when mutated to Ser as it appears in CPT II the catalytic efficiency of the mutants. This indicates that those and CAT, decreased the sensitivity to malonyl-CoA in a stron- mutants that can locate carnitine better at the catalytic site ger fashion than when it was mutated to other amino acids like might displace malonyl-CoA from its site, preventing the bind- Ala and Glu, which were unrelated to this position in other ing of the metabolite and thus the inhibition of CPT I. carnitine acyltransferases. Therefore, we conclude that the oc- Because L-CPT I has not been crystallized, we do not know currence of Ser in this position has probably been evolutionary the proximity of Met to the site of carnitine binding to conserved in nonmalonyl-CoA-sensitive carnitine acyltrans- perform the catalytic event. However, Met is very near the 602– 604 ferases because it prevents sensitivity to malonyl-CoA. In any tripeptide TET , which has been reported to play an Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity 9063 important role in the accommodation of carnitine in catalysis. negative dominant CPT Is, which are expected to be independ- Cronin (30) showed that mutation of the homologous tripeptide ent of malonyl-CoA concentration in a range of tissues such as VDN in choline acetyltransferases to TET greatly increased the liver, muscle, and the -cell, in which the metabolism of fatty catalytic efficiency of the reaction (137-fold) using carnitine as acids plays important roles in ketone body synthesis, resistance 593 602– 604 substrate. This proximity between Met and TET to insulin, and glucose-stimulated insulin secretion, respec- would explain the inverse correlation observed between the tively. Some of these topics are the subject of current investi- catalytic efficiency for carnitine and the IC for malonyl-CoA gations in our laboratory. values of the mutants assayed. A new scenario appears in the Acknowledgment—We are grateful to Robin Rycroft of the Language mutual interaction between carnitine and malonyl-CoA in CPT Service for valuable assistance in the preparation of the manuscript. I. The domain comprised, at least, between amino acid residues 593 and 604 is probably the site of interaction between carni- REFERENCES tine and malonyl-CoA, which exclude each other. Higher cata- 1. McGarry, J. D., and Brown, N. F. (1997) Eur. J. Biochem. 244, 1–14 lytic efficiencies for carnitine in the mutants are followed by 2. Zammit, V. A. (1999) Biochem. J. 343, 505–515 3. Anderson, R. C. (1998) Curr. Pharmaceut. Design 4, 1–16 decreases in the inhibitory sensitivity to malonyl-CoA. 4. Esser, V., Britton, C. H., Weis, B. C., Foster, D. W., and McGarry, J. D. (1993) It is equally interesting to note that all mutants tested show J. Biol. Chem. 268, 5817–5822 5. Yamazaki, N., Shinohara, Y., Shima, A., and Terada, H. (1995) FEBS Lett. higher catalytic efficiency for palmitoyl-CoA as substrate than 363, 41– 45 the wild type. The increase in V /K ranges from 2- to 3-fold. max m 6. Esser, V., Brown, N. F., Cowan, A. T., Foster, D. W., and McGarry, J. D. (1996) Previous work with a partially purified preparation of CPT I J. Biol. Chem. 271, 6972– 6977 7. Mills, S. E., Foster, D. W., and McGarry, J. D. (1984) Biochem. J. 219, 601– 608 had indicated that the kinetics of the reaction with respect to 8. Bird, M. I., and Saggerson, E. D. (1984) Biochem. J. 222, 639 – 647 carnitine concentration could be highly dependent on the con- 9. Nic a’Bhaird, N., and Ramsay, R. R. (1992) Biochem. J. 286, 637– 640 10. Saggerson, E. D., and Carpenter, C. A. (1981) FEBS Lett. 132, 166 –168 centration of the second substrate, palmitoyl-CoA (29). Exper- 11. Cook, G. A., Mynatt, R. L., and Kashfi, K. (1994) J. Biol. Chem. 269, iments carried out by Bird and Saggerson (28) showed that in 8803– 8807 fasted animals, in which carnitine concentration was de- 12. Kashfi, K., Mynatt, R. L., and Cook, G. A. (1994) Biochim. Biophys. Acta 1212, 245–252 creased, the IC values for malonyl-CoA increased up to 17- 13. Grantham, B. D., and Zammit, V. A. (1986) Biochem. J. 233, 589 –593 fold and the binding of [2- -C]malonyl-CoA was reduced by 14. Jackson, V. N., Zammit, V. A., and Price, N. T. (2000) J. Biol. Chem. 275, 38410 –38416 35% at 50 M palmitoyl-CoA and to even lower values at in- 15. Swanson, S. T., Foster, D. W., McGarry, J. D., and Brown, N. F. (1998) creasing palmitoyl-CoA concentrations. Biochem. J. 335, 513–519 Only two of these mutated amino acids are located in the 16. Shi, J., Zhu, H., Arvidson, D. N., and Woldegiorgis, G. (1999) J. Biol. Chem. 274, 9421–9426 three-dimensional CPT I structural model. Ala is one of the 17. Shi, J., Zhu, H., Arvidson, D. N., Cregg, J. M., and Woldegiorgis, G. (1998) amino acids present in the low affinity site of malonyl-CoA Biochemistry 37, 11033–11038 479 483 interaction. This amino acid together with Pro and His 18. Shi, J., Zhu, H., Arvidson, D. N., and Woldegiorgis, G. (2000) Biochemistry 39, 712–717 conform a domain to which malonyl-CoA appears to bind (20). 19. Morillas, M., Go ´ mez-Puertas, P., Roca, R., Serra, D., Asins, G., Valencia, A., Mutation of this amino acid would explain a decrease in sen- and Hegardt, F. G. (2001) J. Biol. Chem. 276, 45001– 45008 20. Morillas, M., Go ´ mez-Puertas, P., Rub´ ı, B., Clotet, J., Arin ˜ o, J., Valencia, A., sitivity to malonyl-CoA, and therefore it would also explain the Hegardt, F. G., Serra, D., and Asins, G. (2002) J. Biol. Chem. 277, increase in catalytic efficiency. On the other hand, Asn is 11473–11480 also present in the catalytic core of the structural model of CPT 21. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) J. Mol. Biol. 215, 403– 410 I (20), but its location does not permit any conclusions about a 22. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22, participation in the malonyl-CoA inhibitory effect. In fact it is 4673– 4680 23. Casari, G., Sander, C., and Valencia, A. (1995) Nat. Struct. Biol. 2, 171–178 located on the opposite site to malonyl-CoA (data not shown). 464 24. Pazos, F., Sanchez-Pulido, L., Garc´ ıa-Ranea, J. A., Andrade, M. A., Atrian, S., Therefore, it is not surprising that its mutation from Asn to and Valencia, A. (1997) in Biocomputing and Emergent Computation Asp does not alter sensitivity to the inhibitor. As a corollary (Lundh, D., Olsson, B., and Narayanan, A., eds) pp. 132–145, World Scien- tific, Singapore of this study, we conclude that the occurrence of the five other 25. Bauer, B., Mirey, G., Vetter, I. R., Garcia-Ranea, J. A., Valencia, A., 223 363 377 490 505 amino acids (Ser , Asp , Gly , Ser , and Ala )atthe Wittinghofer, A., Camonis, J. H., and Cool, R. H. (1999) J. Biol. Chem. 274, 17763–17770 positions, respectively, identical to those amino acids seen in 26. Morillas, M., Clotet, J., Rub´ ı, B., Serra, D., Asins, G., Arin ˜ o, J., and Hegardt CPT I may be sufficient to prevent the sensitivity to malonyl- F. G. (2000) FEBS Lett. 466, 183–186 CoA not only to carnitine acyltransferases such as CPT II and 27. McGarry, J. D., Mills, S. E., Long, C. S., and Foster, D. W. (1983) Biochem. J. 214, 21–28 CAT but also to ChAT. 28. Bird, M. I., and Saggerson, E. D. (1985) Biochem. J. 230, 161–167 The use of either the quintuple mutant or the methionine 29. Bremer, J., and Norum, K. (1967) J. Biol. Chem. 242, 1744 –1748 point mutants may allow studies on the influence of these 30. Cronin, C. N. (1998) J. Biol. Chem. 273, 24465–24469 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Identification of Conserved Amino Acid Residues in Rat Liver Carnitine Palmitoyltransferase I Critical for Malonyl-CoA Inhibition

Download PDF
6 pages

Loading next page...
 
/lp/unpaywall/identification-of-conserved-amino-acid-residues-in-rat-liver-carnitine-AGtvh65g21

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m209999200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 11, Issue of March 14, pp. 9058 –9063, 2003 © 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Identification of Conserved Amino Acid Residues in Rat Liver Carnitine Palmitoyltransferase I Critical for Malonyl-CoA Inhibition MUTATION OF METHIONINE 593 ABOLISHES MALONYL-CoA INHIBITION* Received for publication, September 30, 2002, and in revised form, December 17, 2002 Published, JBC Papers in Press, December 23, 2002, DOI 10.1074/jbc.M209999200 Montserrat Morillas‡§, Paulino Go ´ mez-Puertas§¶, Assia Bentebibel‡, Eva Selle ´ s‡, Nuria Casals**, Alfonso Valencia¶, Fausto G. Hegardt‡ ‡‡, Guillermina Asins‡, and Dolors Serra‡ From the ‡Department of Biochemistry and Molecular Biology, University of Barcelona, School of Pharmacy, E-08028 Barcelona, Spain, the ¶Protein Design Group, National Center for Biotechnology, Consejo Superior de Investigaciones Cientı ´ficas, Cantoblanco, E-28049 Madrid, Spain, and the **Department of Biochemistry and Molecular Biology, International University of Catalonia, 08190 Sant Cugat, Spain Carnitine palmitoyltransferase (CPT) I, which cata- The enzyme carnitine palmitoyltransferase (CPT) I cata- lyzes the conversion of palmitoyl-CoA to palmitoylcar- lyzes the conversion of long chain fatty acyl-CoAs to acylcar- nitine facilitating its transport through the mitochon- nitines, which is the first step in the transport of fatty acyl-CoA drial membranes, is inhibited by malonyl-CoA. By using groups from the cytosol to mitochondria where they undergo the SequenceSpace algorithm program to identify -oxidation. This reaction is inhibited by malonyl-CoA, and so amino acids that participate in malonyl-CoA inhibition this enzyme could be the most physiologically important regu- in all carnitine acyltransferases, we found 5 conserved latory step in mitochondrial fatty acid oxidation (1). This proc- 314 464 478 593 608 amino acids (Thr , Asn , Ala , Met , and Cys , ess allows the cell to signal the relative availability of lipid and rat liver CPT I coordinates) common to inhibitable malo- carbohydrate fuels in liver, heart, skeletal muscle, and pancre- nyl-CoA acyltransferases (carnitine octanoyltransferase atic -cell (2). The mechanism of malonyl-CoA inhibition can be and CPT I), and absent in noninhibitable malonyl-CoA acyl- potentially mimicked by pharmacological malonyl-CoA-related transferases (CPT II, carnitine acetyltransferase (CAT) and agents for the treatment of metabolic disorders such as diabe- choline acetyltransferase (ChAT)). To determine the role of tes, insulin resistance, and coronary heart disease (3). these amino acid residues in malonyl-CoA inhibition, we Mammals express two isoforms of CPT I, a liver isoform prepared the quintuple mutant CPT I T314S/N464D/A478G/ (L-CPT I) and a heart/skeletal muscle isoform (M-CPT I), which M593S/C608A as well as five single mutants CPT I T314S, are the products of two different genes (4, 5). The identity in N464D, A478G, M593S, and C608A. In each case the CPT I amino acids residues is high (62%) but they are differentially amino acid selected was mutated to that present in the regulated by malonyl-CoA. The L-CPT I isoform is inhibited by same homologous position in CPT II, CAT, and ChAT. Be- malonyl-CoA to a much lesser extent than the M-CPT I isoform cause mutant M593S nearly abolished the sensitivity to ma- (the IC value for M-CPT I is about 2 orders of magnitude lonyl-CoA, two other Met mutants were prepared: M593A and M593E. The catalytic efficiency (V /K )ofCPTIin lower than for L-CPT I) (6). This property is probably involved max m mutants A478G and C608A and all Met mutants toward in the finer regulation of fatty acid oxidation in heart and carnitine as substrate was clearly increased. In those CPT I skeletal muscle in comparison to liver. proteins in which Met had been mutated, the malonyl- From studies on the pH dependence of the affinity of CPT I CoA sensitivity was nearly abolished. Mutations in Ala , for its substrate and from the ability of palmitoyl-CoA to dis- 608 314 Cys , and Thr to their homologous amino acid residues place [ C]malonyl-CoA bound to skeletal muscle mitochondria in CPT II, CAT, and ChAT caused various decreases in ma- it was hypothesized (7) that the palmitoyl-CoA and malonyl- lonyl-CoA sensitivity. Ala is located in the structural CoA bind at different sites. A number of studies have shown model of CPT I near the catalytic site and participates in that in rat liver CPT I there are two malonyl-CoA binding sites: the binding of malonyl-CoA in the low affinity site (Morillas, one with greater capacity for binding and regulation of the M., Go ´ mez-Puertas, P., Rubı ´, B., Clotet, J., Arin ˜ o, J., inhibitor and not susceptible to competition from acyl-CoA, Valencia, A., Hegardt, F. G., Serra, D., and Asins, G. (2002) which behaves as an allosteric component (8 –12); and a second J. Biol. Chem. 277, 11473–11480). Met may participate in acyl-CoA binding site, which is located near the catalytic the interaction of malonyl-CoA in the second affinity site, site (13). whose location has not been reported. Various groups have attempted to establish the basis of the L-CPT I/malonyl-CoA interactions. The probable binding sites of malonyl-CoA in L-CPT I were deduced to be at the C termi- * This work was supported in part by Direccio ´ n General de Investi- gacio ´ n Cientı ´fica y Te ´ cnica, Spain, Grant BMC2001-3048 and Ajuts de nus after preparation of several L-CPT I chimeras whose IC Suport als Grups de Recerca de Catalunya Grant 2001SGR-00129 (to values for malonyl-CoA corresponded to the C-terminal region F. G. H.) and the Marato ´ de TV3. The costs of publication of this article (14) of the chimera. However, the N terminus of L-CPT I was were defrayed in part by the payment of page charges. This article must also shown to influence the enzyme/inhibitor interaction. Mu- therefore be hereby marked “advertisement” in accordance with 18 3 5 140 U.S.C. Section 1734 solely to indicate this fact. tation of Glu , His ,orHis produced a loss of malonyl-CoA § Contributed equally to the results of this study. Recipient of a fellowship from the Ministerio de Ciencia y Tecnolo- gı ´a, Spain. The abbreviations used are: CPT, carnitine palmitoyltransferase; ‡‡ To whom correspondence should be addressed: Dept. of Biochem- L-CPT I, liver isoform of carnitine palmitoyltransferase I; M-CPT I, istry and Molecular Biology, School of Pharmacy, Diagonal 643, muscle isoform of carnitine palmitoyltransferase I; CAT, carnitine E-08028 Barcelona, Spain. Tel.: 34-93-402-4523; Fax: 34-93-402-4520; acetyltransferase; COT, carnitine octanoyltransferase; ChAT, choline E-mail: [email protected]. acetyltransferase. 9058 This paper is available on line at http://www.jbc.org This is an Open Access article under the CC BY license. Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity 9059 palmitoyltransferase activity was determined by the radiometric sensitivity (15, 16). In addition, the removal of the segment method as described in Ref. 19 with minor modifications. The sub- comprised between amino acids 1 and 18 in L-CPT I and 1–28 strates were L-[methyl- H]carnitine and palmitoyl-CoA. Enzyme activ- in M-CPT I produced a decrease in malonyl-CoA sensitivity, ity was assayed for 4 min at 30 °C in a total volume of 200 l. which emphasizes the importance of the N terminus before the For determination of the K for carnitine, palmitoyl-CoA was fixed at first transmembrane region as a modulator of the malonyl-CoA 135 M (for L-CPT I). For determination of the K for acyl-CoA, carni- inhibition (17, 18). On the basis of these results, it was pro- tine concentration was fixed at 400 M. When malonyl-CoA inhibition was assayed, increasing concentrations of malonyl-CoA were included. posed that the two malonyl-CoA inhibitable domains might be The IC , defined as the malonyl-CoA concentration that produces 50% located at the C terminus as suggested by several kinetic stud- inhibition of enzyme activity, was determined using 50 M palmitoyl- ies. The development of a CPT I catalytic core model (19) CoA and 400 M carnitine. K was estimated by analyzing the data allowed us to assign the low affinity binding site to a domain from three experiments using the program Enzifit (Biosoft), and IC near the catalytic channel in which palmitoyl-CoA is bound was calculated by Excel software using linear regression analysis. containing the catalytic acyl-CoA binding domain (20) Values reported in the text are the means and standard deviations of three to five determinations. Curve fitting was carried out using Excel Here we used the SequenceSpace algorithm program to iden- 314 464 478 593 software. All protein concentrations were determined using the Bio-Rad tify five amino acid residues (Thr , Asn , Ala , Met , protein assay with bovine albumin as standard. and Cys ), which may contribute to the sensitivity of CPT I to Immunological Techniques—Western blot analysis was performed as malonyl-CoA. The proposal is based on the finding that they described (19). The antibody for rat L-CPT I was kindly given by are present in malonyl-CoA-inhibitable CPT I ((isoforms L- and Dr. V. A. Zammit (Hannah Research Institute, Ayr, Scotland, United M-) and COT from various organisms and absent in noninhib- Kingdom) and was directed against peptide 428 – 441, in the cytosolic catalytic C-terminal domain. itable acyltransferases (CPT II, CAT, and ChAT). Mutation of these amino acids to their counterparts in CPT II showed that RESULTS mutation of Met by itself, M593S, or the quintuple mutant Residues Conserved in Malonyl-CoA Inhibited Versus Nonin- containing the M593S mutation, T314S/N464D/A478G/M593S/ hibited Carnitine-Choline Acyltransferases—An exhaustive C608A, or other Met point mutants such as M593A and analysis of the presence of residues shared by all the malonyl- M593E nearly abolished malonyl-CoA sensitivity of L-CPT I. CoA-regulated enzymes of the carnitine-choline acyltrans- The remaining mutated amino acids showed slight, varied sen- ferase family versus the malonyl-CoA nonregulated members of sitivity to malonyl-CoA inhibition. the same family was performed using the algorithm Sequence- EXPERIMENTAL PROCEDURES Space (23, 24). This method uses a vectorial representation of each protein sequence as a point in a multidimensional space Tree-determinants Analysis—Sequences of proteins from the carni- tine-choline acyltransferase family were obtained using BLAST (21). (SequenceSpace) and multivariate statistics, principal compo- Multiple alignment was performed using ClustalW (22). The analysis of nent analysis, to allow reduction of the number of dimensions. conserved differences (tree-determinants) between malonyl-CoA-regu- This representation allows us not only to define clusters of lated (L-CPT I, M-CPT I, and COT) and nonregulated (CPT II, CAT, and proteins according to specific properties by choosing the appro- ChAT) acyltransferases, using multivariate statistics for low-dimen- priate axes defined by the highest corresponding eigenvalues sional representation, was done using the SequenceSpace algorithm (also known as proper values), but also to project the individual (23, 24). Graphics of vectors representing protein sequences and indi- vidual residues from the multiple alignment were performed using the residues on the same axes, and thus trace the positions con- Sequence Space Java-based viewer (www.industry.ebi.ac.uk/SeqSpace). served in the subfamilies defined. The main advantage of this wt Construction of Site-directed Mutants—Plasmids pYESLCPTI and method is the possibility of predicting which residues may be A478G pYESLCPT were obtained as previously described (20). Plasmids responsible for the specific characteristics of each protein sub- T314S N464D M593S M593A pYESLCPT , pYESLCPT , pYESLCPT , pYESLCPT , M593E C608A family or group of subfamilies as has been reported previously pYESLCPT , and pYESLCPT were constructed using the for short- and medium-long substrate specificity for the carni- QuikChange polymerase chain reaction-based mutagenesis procedure wt (Stratagene) with the pYESLCPT plasmid as template. The following tine-choline acyltransferases protein family (19, 20) or effector primers were used: primer T314S.for 5-GGGAGCGACTCTTCAATAG- recognition by some members of the Ras superfamily (25). TTCCCGGATCCCTGGG-3, primer T314A.rev 5-CCCAGGGATCCG- The two-dimensional projection of sequence vectors on the GGAACTATTGAAGAGTCGCTCCC-3, primer N464D.for 5-CACCTT- plane defined by the axes corresponding to eigenvalues 2 and 4 TGTTGTCTTCAAAGACAGCAAGATAGGC-3, primer N464D.rev 5- showed clustering of the enzyme subfamilies according to their GCCTATCTTGCTGTCTTTGAAGACACCAAAGGTG-3, primer malonyl-CoA inhibition properties (Fig. 1A). Proteins whose M593S.for 5-CCTCACATATGAGGCCTCCAGTACCCGGCTCTTCCG AGAAGG-3, primer M593S.rev 5-CCTTCTCGGAAGAGCCGGGTAC- activity is not regulated by malonyl-CoA (CPT II, CAT, and TGGAGGCCTCATATGTGAGG-3, primer M593A.for 5-CCTCAC- ChAT subfamilies) were grouped, whereas the sequences of the ATATGAGGCCTCCGCGACCCGGCTCTTCCGAGAAGG-3, primer proteins regulated by malonyl-CoA (COT, L-CPT I, and M-CPT M593A.rev 5-CCTTCTCGGAAGAGCCGGGTCGCGGAGGCCTCATA I) occupy separate, and opposite, zones. The projection of the TGTGAGG-3, primer M593E.for 5-CCTCACATATGAGGCCTCCGA- individual amino acid residues on the same plane (Fig. 1B) GACCCGGCTCTTCCGAGAAGG-3, primer M593E.rev 5-CCTTCTC- revealed the amino acids responsible for this segregation might GGAAGAGCCGGGTCTCGGAGGCCTCATATGTGAGG-3, primer be responsible for the susceptibility to malonyl-CoA of the C608A.for 5-GAGACTGTACGCTCCGCCACTATGGAGTCCTGC-3, and C608A.rev 5-GCAGGACTCCATAGTGGCGGAGCGTACAGTCT- corresponding enzymes. Five of these amino acids (Thr , 464 478 593 608 C-3 (the mutated nucleotides are underlined). The plasmid Asn , Ala , Met , and Cys ) were present in all malonyl- T314S/N464D/A478G/M593S/C608A pYESLCPTI was obtained by the same CoA inhibitable carnitine acyltransferases and absent in the method, but performing each new mutation stepwise starting on plas- nonmalonyl-CoA inhibitable acyltransferases (CPT II, CAT, T314S mid pYESLCPT . The appropriate substitutions as well as the and ChAT from several species). Fig. 2 shows the sequence absence of unwanted mutations were confirmed by sequencing the alignment of three fragments of the C-terminal region of vari- inserts in both directions with an Applied Biosystems 373 automated DNA sequencer. ous acyltransferases. We can also observe that those enzymes Expression of L-CPT I in Saccharomyces cerevisiae—The expression that are not inhibitable by malonyl-CoA (CPT II, CAT, and of the constructs containing L-CPT I wild type and mutants (see above) ChAT) show the same amino acids in these positions, which are in yeast cells and the preparation of the cell extracts were performed as different from those observed in inhibitable malonyl-CoA acyl- described in Ref. 19. S. cerevisiae was chosen as an expression system carnitines. As an example the positions and amino acids of CPT for L-CPT I wild type and the mutants because it does not have endog- 223 363 377 490 505 II are given: Ser , Asp , Gly , Ser , and Ala (Fig. 2). enous CPT I activity. Determination of Carnitine Acyltransferase Activity—Carnitine Expression of Wild Type and Mutants in S. cerevisiae—We 9060 Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity FIG.1. Sequence space analysis of the carnitine-choline acyltransferase family. A, protein sequences projected onto the plane defined by principle axes 2 and 4. This two-dimensional space allows separation of protein subfamilies according to their malonyl-CoA regulatory characteristics; CPT II, CAT, and ChAT (CACP) enzymes (malonyl-CoA insensitive) are clustered to the lower left corner of the panel, whereas CPT I (L- and M-isoforms) and COT (malonyl-CoA inhibitable enzymes) are projected on the upper and right areas of the vertical and horizontal axes, respectively. B, the sequence of each subfamily is represented as a vector point in a multidimensional space (sequence space), with residue positions and types as the basic dimensions. Single residues completely conserved in CPT I or COT subfamilies are projected in the same position as their corresponding protein sequences. Residues conserved in both groups of malonyl-CoA-regulated enzymes occupy the upper right corner, whereas the residues conserved in the nonregulated cluster of acyltransferases (CPT II, CAT, and ChAT) occupy the opposite one. Residues located in alignment positions present in both opposite corners of the two-dimensional plot are responsible for protein cluster segregation and are predicted to be involved in malonyl-CoA sensitivity. FIG.2. Alignment of representative sequences of mammalian carnitine-choline acyltransferases. Amino acid sequence of 18 repre- sentative members of the malonyl-CoA-insensitive enzymes, CPT II (CPT2) from rat, mouse, and human; CAT (CACP) from human and mouse; ChAT (CLAT) from human, pig, rat, and mouse; and malonyl-CoA inhibitable enzymes L-CPT I (CPT1) from rat, mouse, and human; M-CPT I (CPTM) from human, rat, and mouse; and COT (OCTC) from human, rat, and bovine, were obtained from the SwissProt data bank and aligned using ClustalW (22). A, schematic representation of the position of the tree-determinant residues obtained using the SequenceSpace algorithm (23, 223 314 363 464 377 478 490 593 505 608 24) on the rat CPT II and L-CPTI proteins: Ser /Thr , Asp /Asn , Gly /Ala , Ser /Met , Ala /Cys . Transmembrane regions of 372 473 L-CPT I are also represented (tm1 and tm2). Position of the catalytic histidine (His /His ) as well as the previously three-dimensional modeled core of the proteins, 2dub, (amino acids 368 –567 of L-CPT I) (19, 20), are indicated.B, selected regions of the multiple alignment of the protein family. Subfamily conserved residues according to malonyl-CoA regulation are shadowed. Position of catalytic histidine (arrowhead) is also indicated. 1 1 prepared a quintuple mutant, T314S/N464D/A478G/M593S/ (values ranged between 14 and 20 nmol min mg protein ) C608A, and separately, the point mutants T314S, N464D, when the protein was overexpressed 20 h after galactose induc- A478G, M593S, and C608A and all were expressed in S. cer- tion, showing that the various mutations assayed produce evisiae. After we observed that mutant M593S nearly abolished small changes in L-CPT I activity (Table I). the sensitivity to malonyl-CoA (see below), new point Met mu- All mutants exhibited standard saturation kinetics when the tants were prepared: M593A and M593E. All transformed carnitine concentration was varied relative to a constant con- yeast cells expressed a protein with the same molecular mass centration of the second substrate, palmitoyl-CoA, and when (88 kDa) and the mutant enzymes were expressed in roughly palmitoyl-CoA concentration was varied relative to a constant the same proportion per milligram of protein as the wild type carnitine concentration, a property identical to that of the wild L-CPT I as deduced from immunoblot analysis (data not type L-CPT I (Fig. 3). The quintuple mutant produced small shown). changes in the kinetic constants for carnitine and palmitoyl- Kinetic Properties of CPT I Wild Type and Mutants—L-CPT CoA as substrates (Table I). Catalytic efficiency (V /K ) was max m I activities of the wild type, quintuple mutant variant T314S/ increased by a factor of 2.6 (carnitine) and 2.2 (palmitoyl-CoA). N464D/A478G/M593S/C608S, and point mutants were similar The catalytic efficiency for carnitine as substrate of those point Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity 9061 TABLE I Enzyme activity, malonyl-CoA sensitivity and kinetic parameters of carnitine palmitoyltransferase I in Saccharomyces cerevisiac cells expressing CPI I wild type and point mutants, T314S, N464D, A478G, C608A, M593S, M593A, M593E and quintuple mutant T314S/N464D/A478G/M593S/C608A (QM) Extracts from yeast expressing wild type and several mutants of L-CPT I were assayed for activity, malonyl-CoA sensitivity, and kinetics as described under “Experimental Procedures.” The results are the mean  S.D. of at least three independent experiments with different prepara- tions. In parentheses are shown the increase (in-fold number) of the catalytic efficiency (V /K ) versus to that of the wild type. max m K V Catalytic efficiency m max IC L-CPT I Activity malonyl-CoA Carnitine Palmitoyl-CoA Carnitine Palmitoyl-CoA Carnitine Palmitoyl-CoA 1 1 1 1 nmol min mg protein m nmol min mg protein V /K max m Wild-type 17.7  0.9 12.3 127  4.5 4.9  0.3 6.6  0.8 6.3  0.4 0.05 ( 1) 1.28 ( 1) T314S 14.4  2.1 15.0 88.2  2.4 1.7  0.5 12.8  0.1 6.8  0.1 0.15 ( 2.8) 3.98 ( 3.1) N464D 20.1  3.1 8.7 69.5  8.2 4.1  0.4 19.4  1.4 18.9  3.6 0.28 ( 5.6) 4.63 ( 3.6) A478G 16.7  0.7 39.5 327  41 15.1  4.0 69.8  9.3 50.4  17 0.21 ( 4.1) 3.34 ( 2.6) C608A 17.3  1.7 27.5 51.6  4.0 24.3  2.0 23.7  5.0 67.5  9.0 0.46 ( 8.8) 2.78 ( 2.2) M593S 17.0  0.8 319 124  0.8 7.4  1.2 133  18 20.9  1.6 1.07 ( 21) 2.84 ( 2.2) M593A 17.2  0.9 155 56.3  2.1 6.1  0.2 32.5  4.6 30.3  4.7 0.58 ( 12) 4.81 ( 3.7) M593E 14.1  1.8 220 150  3.4 6.3  0.5 31.3  2.6 27.5  1.8 0.21 ( 4.2) 4.37 ( 3.4) QM 13.6  1.6 258 95.7  2.8 4.6  1.5 13.1  4.7 13.0  6.3 0.14 ( 2.6) 2.84 ( 2.2) Even at concentrations as high as 100 M malonyl-CoA the CPT I quintuple mutant maintained 80% of the activity of the control without malonyl-CoA. We then addressed the individual responsibility of the sep- arate CPT I mutants for the malonyl-CoA sensitivity. Mutants T314S, N464D, M593S, and C608A expressed in S. cerevisiae were incubated with increasing amounts of malonyl-CoA, and CPT I activity was determined. Mutant A478G had been pre- viously studied in Ref. 20 and showed decreased sensitivity to malonyl-CoA (IC of 39.5 versus 12.3 M of the wild type). The kinetics of inhibition by malonyl-CoA depended on the mutant considered. Whereas mutant M593S (Fig. 4A) showed very low sensitivity at malonyl-CoA inhibition (IC of 319 M), the other mutations produced varied changes in malonyl-CoA sensitivity. L-CPT I C608A slightly modified the sensitivity to malonyl-CoA (IC is 27.5 M), the change in IC of mutant 50 50 T314S was small, whereas N464D showed similar sensitivity to malonyl-CoA to the wild type (Fig. 4B and Table I). Because the highest changes in sensitivity to malonyl-CoA and K values for carnitine were observed in the methionine mutants (point and quintuple mutants), we additionally prepared two new FIG.3. Kinetic analysis of wild type and different mutants of L-CPT I. Yeast extracts (10 g of protein) of (A and C) wild type (open mutants: M593A and M593E to examine whether Met was circles) and mutants M593S (open triangles), M593A (black rhombus), essential to the malonyl-CoA interaction in L-CPT I. Results M593E (black squares), and (C and D) T314S (open rhombus), N464D show that the sensitivity to malonyl-CoA was also nearly abol- (open squares), A478G (black squares), C608A (black triangles), and ished in these mutants (Fig. 4A) (IC values of 155 and 220 M, quintuple mutant T314S/N464D/A478G/M593S/C608A (black circles) 50 were incubated at increasing concentrations of carnitine (A and B) and respectively) as in the M593S mutant, confirming the essential palmitoyl-CoA (C and D). role of Met in this interaction. DISCUSSION mutants that altered the sensitivity to malonyl-CoA increased We attempted to identify the amino acids in the C-terminal (see below). The catalytic efficiency of the methionine mutants domain of L-CPT I that are responsible for the inhibition of the increased between 4.2- and 21-fold, C608A increased 8.8-fold, catalytic activity by malonyl-CoA. Over many years much work and A478G increased 4.1-fold. T314S, which produced a small has been done to identify the domains in L-CPT I that may bind change in malonyl-CoA sensitivity (see below), increased the malonyl-CoA. Different groups have tested different empirical V /K value by only 2.8, whereas in N464D, in which the hypotheses and mutated amino acids, mostly in the amino- max m sensitivity to malonyl-CoA was unchanged (see below), terminal region of L-CPT I. The results have shown that this the catalytic efficiency was modified by a factor of 5.6. domain plays a role in the regulation of CPT I by malonyl-CoA, An analogous tendency was also observed in K for palmi- because in some cases the sensitivity to the inhibitor is toyl-CoA but the changes were smaller. K values for palmi- impaired. toyl-CoA were 24.3, 15.1, 7.4, 6.1, and 6.3 for mutants C608A, A different approach was employed by our group very re- A478G, M593S, M593A, and M593E, respectively (K value for cently. This was based on the conservation of two histidine the wild type was 4.9) (Table I). Catalytic efficiencies for palmi- residues, which are present in the inhibitable malonyl-CoA toyl-CoA as substrate increased in all mutants, the values carnitine acyltransferases (CPT I and COT) and absent in ranging between 2.78 and 4.81 (Table I). noninhibitable enzymes (CPT II and CAT). Mutation of both Inhibition of CPT I Wild Type and Mutants by Malonyl- histidines resulted in the abolition of malonyl-CoA sensitivity CoA—When inhibitory kinetics versus increasing concentra- in COT (26). Analogous results were observed in CPT I when its tions of malonyl-CoA was performed, the quintuple mutant concentration at the mitochondrial membranes was not high. practically abolished the sensitivity toward malonyl-CoA (IC Mutation of other amino acids in the domain proximal to the 478 479 of 258 versus 12.3 M of the wild type) (Fig. 4B and Table I). catalytic site (Ala and Pro ) indicated that a malonyl-CoA- 9062 Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity FIG.4. Effect of malonyl-CoA on the activity of yeast overexpressed L-CPT I (wild type) and several mutants. A, L-CPT I wild type (open circles) and point methionine mutants M593S (black circles), M593A (black rhombus), M593E (black squares), and B, quintuple mutant (QM) (black circles) and point mutants T314S (open circles), N464D (open rhombus, broken line), A478G (open triangles, broken line), and C608A (open squares) overexpressed in yeast were incubated with increasing concentrations of malonyl-CoA and the enzyme activity was measured. Data are expressed relative to control values in the absence of inhibitor (100%) as the mean of three independent measurements. inhibitable domain was probably the low-affinity malonyl-CoA case, it appears that Met is critical in the interaction of binding site. Our previous studies showed that the location of malonyl-CoA with L-CPT I. malonyl-CoA in the structural model was compatible with com- It was of interest to measure the kinetic constants of all CPT petition of the inhibitor versus the substrate in the malonyl- I mutants. Several authors reported the competition between CoA low affinity binding site (20). malonyl-CoA and carnitine (27, 28). The tissues in which the The site-directed mutagenesis study used here to identify sensitivity of CPT I to malonyl-CoA is highest are those that amino acids responsible for malonyl-CoA inhibition is based on require the highest concentration of carnitine to drive the re- the comparison of the sequences in a range of carnitine and action and the requirement for carnitine and sensitivity to choline acyltransferases, taking the positive or negative sensi- malonyl-CoA appears to be inversely related. The authors con- tivity to malonyl-CoA as a discriminatory criterion. The bio- cluded that the sites to which the two metabolites bind are computing study has shown that five amino acids are present closely associated (27, 7). Studies by Bird and Saggerson (28) in all CPT I (isoforms L- and M-) and in COT from various showed on the one hand that malonyl-CoA reduced the effec- organisms and that they are absent not only in other nonma- tiveness of carnitine as substrate, and on the other hand, that lonyl-CoA-inhibitable carnitine acyltransferases but also in carnitine might diminish the regulatory effect of malonyl-CoA 314 464 ChAT. In rat L-CPT I these amino acids are Thr , Asn , (29). Although a clear mechanism for this competition could not 478 593 608 Ala , Met , and Cys . The corresponding positional amino be established, the data strongly supported this idea. In the 223 363 377 acids in CPT II, CAT, and ChAT are Ser , Asp , Gly , present study the various CPT I mutants have altered K or 490 505 Ser , and Ala , respectively (coordinates of rat CPT II). V for carnitine. Whereas the K for C608A was half of the max m Therefore, we considered it highly probable that these amino wild type, its V was 3.6-fold higher. The mutant M593S had max acids were involved in the interaction of malonyl-CoA. Results the same K value for carnitine as the wild type but its V m max confirmed in part this supposition. The quintuple mutant re- increased 20-fold. The mutant A478G increased both the K duced malonyl-CoA sensitivity almost completely (80% activity value and the V with respect to the wild type values. The max at 100 M malonyl-CoA (which is outside the physiological relationship between these values and catalysis is best re- range)), supporting the initial hypothesis. The results obtained vealed in the term catalytic efficiency. This term as calculated using separate single mutants indicate that not all of these by the V /K ratio varies considerably among different mu- max m amino acids have the same role in malonyl-CoA inhibition. tants. Carnitine catalytic efficiencies for mutants M593S, Whereas M593S nearly abolished the sensitivity to malonyl- M593A, M593E, C608A, and A478G increased 21-, 12-, 4.2-, CoA like the quintuple mutant, A478G increased the IC from 8.8-, and 4.1-fold with respect to the wild type. This means that 12 to 39.5 M (20). The other amino acids are less responsible mutations designed to decrease malonyl-CoA sensitivity for the inhibition. strongly modified the catalytic efficiency of CPT I mutants The relevance of Met as a critical amino acid for malonyl- measured in the absence of malonyl-CoA. Interestingly, the CoA sensitivity was confirmed by the results of mutation to increase in catalytic efficiency appears to be roughly propor- other two amino acids, Ala and Glu. The mutants equally tional to the extent of the alteration in malonyl-CoA sensitivity. showed diminished sensitivity to malonyl-CoA like mutant The IC values for malonyl-CoA run in the same direction to M593S. Met , when mutated to Ser as it appears in CPT II the catalytic efficiency of the mutants. This indicates that those and CAT, decreased the sensitivity to malonyl-CoA in a stron- mutants that can locate carnitine better at the catalytic site ger fashion than when it was mutated to other amino acids like might displace malonyl-CoA from its site, preventing the bind- Ala and Glu, which were unrelated to this position in other ing of the metabolite and thus the inhibition of CPT I. carnitine acyltransferases. Therefore, we conclude that the oc- Because L-CPT I has not been crystallized, we do not know currence of Ser in this position has probably been evolutionary the proximity of Met to the site of carnitine binding to conserved in nonmalonyl-CoA-sensitive carnitine acyltrans- perform the catalytic event. However, Met is very near the 602– 604 ferases because it prevents sensitivity to malonyl-CoA. In any tripeptide TET , which has been reported to play an Amino Acids Involved in L-CPT I Malonyl-CoA Sensitivity 9063 important role in the accommodation of carnitine in catalysis. negative dominant CPT Is, which are expected to be independ- Cronin (30) showed that mutation of the homologous tripeptide ent of malonyl-CoA concentration in a range of tissues such as VDN in choline acetyltransferases to TET greatly increased the liver, muscle, and the -cell, in which the metabolism of fatty catalytic efficiency of the reaction (137-fold) using carnitine as acids plays important roles in ketone body synthesis, resistance 593 602– 604 substrate. This proximity between Met and TET to insulin, and glucose-stimulated insulin secretion, respec- would explain the inverse correlation observed between the tively. Some of these topics are the subject of current investi- catalytic efficiency for carnitine and the IC for malonyl-CoA gations in our laboratory. values of the mutants assayed. A new scenario appears in the Acknowledgment—We are grateful to Robin Rycroft of the Language mutual interaction between carnitine and malonyl-CoA in CPT Service for valuable assistance in the preparation of the manuscript. I. The domain comprised, at least, between amino acid residues 593 and 604 is probably the site of interaction between carni- REFERENCES tine and malonyl-CoA, which exclude each other. Higher cata- 1. McGarry, J. D., and Brown, N. F. (1997) Eur. J. Biochem. 244, 1–14 lytic efficiencies for carnitine in the mutants are followed by 2. Zammit, V. A. (1999) Biochem. J. 343, 505–515 3. Anderson, R. C. (1998) Curr. Pharmaceut. Design 4, 1–16 decreases in the inhibitory sensitivity to malonyl-CoA. 4. Esser, V., Britton, C. H., Weis, B. C., Foster, D. W., and McGarry, J. D. (1993) It is equally interesting to note that all mutants tested show J. Biol. Chem. 268, 5817–5822 5. Yamazaki, N., Shinohara, Y., Shima, A., and Terada, H. (1995) FEBS Lett. higher catalytic efficiency for palmitoyl-CoA as substrate than 363, 41– 45 the wild type. The increase in V /K ranges from 2- to 3-fold. max m 6. Esser, V., Brown, N. F., Cowan, A. T., Foster, D. W., and McGarry, J. D. (1996) Previous work with a partially purified preparation of CPT I J. Biol. Chem. 271, 6972– 6977 7. Mills, S. E., Foster, D. W., and McGarry, J. D. (1984) Biochem. J. 219, 601– 608 had indicated that the kinetics of the reaction with respect to 8. Bird, M. I., and Saggerson, E. D. (1984) Biochem. J. 222, 639 – 647 carnitine concentration could be highly dependent on the con- 9. Nic a’Bhaird, N., and Ramsay, R. R. (1992) Biochem. J. 286, 637– 640 10. Saggerson, E. D., and Carpenter, C. A. (1981) FEBS Lett. 132, 166 –168 centration of the second substrate, palmitoyl-CoA (29). Exper- 11. Cook, G. A., Mynatt, R. L., and Kashfi, K. (1994) J. Biol. Chem. 269, iments carried out by Bird and Saggerson (28) showed that in 8803– 8807 fasted animals, in which carnitine concentration was de- 12. Kashfi, K., Mynatt, R. L., and Cook, G. A. (1994) Biochim. Biophys. Acta 1212, 245–252 creased, the IC values for malonyl-CoA increased up to 17- 13. Grantham, B. D., and Zammit, V. A. (1986) Biochem. J. 233, 589 –593 fold and the binding of [2- -C]malonyl-CoA was reduced by 14. Jackson, V. N., Zammit, V. A., and Price, N. T. (2000) J. Biol. Chem. 275, 38410 –38416 35% at 50 M palmitoyl-CoA and to even lower values at in- 15. Swanson, S. T., Foster, D. W., McGarry, J. D., and Brown, N. F. (1998) creasing palmitoyl-CoA concentrations. Biochem. J. 335, 513–519 Only two of these mutated amino acids are located in the 16. Shi, J., Zhu, H., Arvidson, D. N., and Woldegiorgis, G. (1999) J. Biol. Chem. 274, 9421–9426 three-dimensional CPT I structural model. Ala is one of the 17. Shi, J., Zhu, H., Arvidson, D. N., Cregg, J. M., and Woldegiorgis, G. (1998) amino acids present in the low affinity site of malonyl-CoA Biochemistry 37, 11033–11038 479 483 interaction. This amino acid together with Pro and His 18. Shi, J., Zhu, H., Arvidson, D. N., and Woldegiorgis, G. (2000) Biochemistry 39, 712–717 conform a domain to which malonyl-CoA appears to bind (20). 19. Morillas, M., Go ´ mez-Puertas, P., Roca, R., Serra, D., Asins, G., Valencia, A., Mutation of this amino acid would explain a decrease in sen- and Hegardt, F. G. (2001) J. Biol. Chem. 276, 45001– 45008 20. Morillas, M., Go ´ mez-Puertas, P., Rub´ ı, B., Clotet, J., Arin ˜ o, J., Valencia, A., sitivity to malonyl-CoA, and therefore it would also explain the Hegardt, F. G., Serra, D., and Asins, G. (2002) J. Biol. Chem. 277, increase in catalytic efficiency. On the other hand, Asn is 11473–11480 also present in the catalytic core of the structural model of CPT 21. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) J. Mol. Biol. 215, 403– 410 I (20), but its location does not permit any conclusions about a 22. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22, participation in the malonyl-CoA inhibitory effect. In fact it is 4673– 4680 23. Casari, G., Sander, C., and Valencia, A. (1995) Nat. Struct. Biol. 2, 171–178 located on the opposite site to malonyl-CoA (data not shown). 464 24. Pazos, F., Sanchez-Pulido, L., Garc´ ıa-Ranea, J. A., Andrade, M. A., Atrian, S., Therefore, it is not surprising that its mutation from Asn to and Valencia, A. (1997) in Biocomputing and Emergent Computation Asp does not alter sensitivity to the inhibitor. As a corollary (Lundh, D., Olsson, B., and Narayanan, A., eds) pp. 132–145, World Scien- tific, Singapore of this study, we conclude that the occurrence of the five other 25. Bauer, B., Mirey, G., Vetter, I. R., Garcia-Ranea, J. A., Valencia, A., 223 363 377 490 505 amino acids (Ser , Asp , Gly , Ser , and Ala )atthe Wittinghofer, A., Camonis, J. H., and Cool, R. H. (1999) J. Biol. Chem. 274, 17763–17770 positions, respectively, identical to those amino acids seen in 26. Morillas, M., Clotet, J., Rub´ ı, B., Serra, D., Asins, G., Arin ˜ o, J., and Hegardt CPT I may be sufficient to prevent the sensitivity to malonyl- F. G. (2000) FEBS Lett. 466, 183–186 CoA not only to carnitine acyltransferases such as CPT II and 27. McGarry, J. D., Mills, S. E., Long, C. S., and Foster, D. W. (1983) Biochem. J. 214, 21–28 CAT but also to ChAT. 28. Bird, M. I., and Saggerson, E. D. (1985) Biochem. J. 230, 161–167 The use of either the quintuple mutant or the methionine 29. Bremer, J., and Norum, K. (1967) J. Biol. Chem. 242, 1744 –1748 point mutants may allow studies on the influence of these 30. Cronin, C. N. (1998) J. Biol. Chem. 273, 24465–24469

Journal

Journal of Biological Chemistry – Unpaywall

Published: Mar 1, 2003

There are no references for this article.