Homologous Binding Sites in Yeast Isocitrate Dehydrogenase for Cofactor (NAD+) and Allosteric Activator (AMP)
2003; Elsevier BV; Volume: 278; Issue: 15 Linguagem: Inglês
10.1074/jbc.m300154200
ISSN1083-351X
AutoresAn-Ping Lin, Lee McAlister-Henn,
Tópico(s)Amino Acid Enzymes and Metabolism
ResumoYeast NAD+-specific isocitrate dehydrogenase (IDH) is an allosterically regulated octameric enzyme composed of two types of homologous subunits designated IDH1 and IDH2. Based on sequence comparisons and structural models, both subunits are predicted to have adenine nucleotide binding sites. This was tested by alanine replacement of residues in putative sites in each subunit. Targets included adjacent aspartate/isoleucine residues implicated as important for determining cofactor specificity in related dehydrogenases and a residue in each IDH subunit in a position occupied by histidine in other cofactor binding sites. The primary kinetic effects of D286A/I287A and of H281A replacements in IDH2 were found to be a dramatic reduction in apparent affinity of the holoenzyme for NAD+ and a concomitant reduction inVmax. Ligand binding assays also showed that the H281A mutant enzyme fails to bind NAD+ under conditions that are saturating for the wild-type enzyme. In contrast, the primary effect of corresponding D279A/D280A and of R274A replacements in IDH1 is a reduction in holoenzyme binding of AMP, with concomitant alterations in kinetic and isocitrate binding properties normally associated with activation by this allosteric effector. These results suggest that the nucleotide cofactor binding site is primarily contributed by the IDH2 subunit, whereas the homologous nucleotide binding site in IDH1 has evolved for regulatory binding of AMP. These results are consistent with previous studies demonstrating that the catalytic isocitrate binding sites are comprised of residues primarily contributed by IDH2, whereas sites for regulatory binding of isocitrate are contributed by analogous residues of IDH1. In this study, we also demonstrate that a prerequisite for holoenzyme binding of NAD+ is binding of isocitrate/Mg2+ at the IDH2 catalytic site. This is comparable to the dependence of AMP binding upon binding of isocitrate at the IDH1 regulatory site. Yeast NAD+-specific isocitrate dehydrogenase (IDH) is an allosterically regulated octameric enzyme composed of two types of homologous subunits designated IDH1 and IDH2. Based on sequence comparisons and structural models, both subunits are predicted to have adenine nucleotide binding sites. This was tested by alanine replacement of residues in putative sites in each subunit. Targets included adjacent aspartate/isoleucine residues implicated as important for determining cofactor specificity in related dehydrogenases and a residue in each IDH subunit in a position occupied by histidine in other cofactor binding sites. The primary kinetic effects of D286A/I287A and of H281A replacements in IDH2 were found to be a dramatic reduction in apparent affinity of the holoenzyme for NAD+ and a concomitant reduction inVmax. Ligand binding assays also showed that the H281A mutant enzyme fails to bind NAD+ under conditions that are saturating for the wild-type enzyme. In contrast, the primary effect of corresponding D279A/D280A and of R274A replacements in IDH1 is a reduction in holoenzyme binding of AMP, with concomitant alterations in kinetic and isocitrate binding properties normally associated with activation by this allosteric effector. These results suggest that the nucleotide cofactor binding site is primarily contributed by the IDH2 subunit, whereas the homologous nucleotide binding site in IDH1 has evolved for regulatory binding of AMP. These results are consistent with previous studies demonstrating that the catalytic isocitrate binding sites are comprised of residues primarily contributed by IDH2, whereas sites for regulatory binding of isocitrate are contributed by analogous residues of IDH1. In this study, we also demonstrate that a prerequisite for holoenzyme binding of NAD+ is binding of isocitrate/Mg2+ at the IDH2 catalytic site. This is comparable to the dependence of AMP binding upon binding of isocitrate at the IDH1 regulatory site. isocitrate dehydrogenase nitrilotriacetic acid Mitochondrial NAD+-specific isocitrate dehydrogenase (IDH)1 catalyzes a rate-limiting step in the tricarboxylic acid cycle and is subject to complex allosteric regulation. In particular, because of allosteric activation of the mammalian enzyme by ADP (1Chen R.F. Plaut G.W.E. Biochemistry. 1963; 2: 1023-1032Crossref PubMed Scopus (122) Google Scholar) and of the yeast enzyme by AMP, IDH is proposed to regulate metabolic flux in response to energy needs of the cell (2Hathaway J.A. Atkinson D.E. J. Biol. Chem. 1963; 238: 2875-2881Abstract Full Text PDF PubMed Google Scholar). Saccharomyces cerevisiae IDH is an octamer composed of four each of two homologous subunits, IDH1 and IDH2 (3Keys D.A. McAlister-Henn L. J. Bacteriol. 1990; 172: 4280-4287Crossref PubMed Google Scholar). The mature polypeptides are similar in size (349 and 354 amino acid residues, respectively) and share 42% residue sequence identity (4Cupp J.R. McAlister-Henn L. J. Biol. Chem. 1991; 266: 22199-22205Abstract Full Text PDF PubMed Google Scholar, 5Cupp J.R. McAlister-Henn L. J. Biol. Chem. 1992; 267: 16417-16423Abstract Full Text PDF PubMed Google Scholar). Both subunits are essential for holoenzyme structure and function, although, as described below, catalytic function has been primarily attributed to IDH2 whereas regulatory functions have been assigned to IDH1 (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar). Mammalian IDH contains three different types of subunits that share significant homology with those of yeast IDH, an α-subunit with catalytic functions similar to those of yeast IDH2, and β- and γ-subunits that are presumed to impart regulatory properties (9Nichols B.J. Hall L. Perry A.C.F. Denton R.M. Biochem. J. 1993; 295: 347-350Crossref PubMed Scopus (36) Google Scholar, 10Nichols B.J. Perry A.C. Hall L. Denton R.M. Biochem. J. 1995; 310: 917-922Crossref PubMed Scopus (29) Google Scholar).Although crystallographic data are unavailable for yeast and mammalian IDHs, these enzymes share substantial similarity in sequence with several bacterial decarboxylating dehydrogenases for which three-dimensional structures are available. As a family, these dehydrogenases are unique in that they lack the classic Rossman fold described for other enzymes that bind NAD(P)+ (11Rossman M.G. Moras D. Olsen K.W. Nature. 1974; 250: 194-199Crossref PubMed Scopus (1161) Google Scholar). Particularly useful for analyses of yeast IDH are structures reported for Escherichia coli isocitrate dehydrogenase (12Hurley J.H. Thorsness P.E. Ramalingam V. Helmers N.H. Koshland Jr., D.E. Stroud R.M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 8635-8639Crossref PubMed Scopus (206) Google Scholar, 13Hurley J.H. Dean A.M. Koshland Jr., D.E. Stroud R.M. Biochemistry. 1991; 30: 8671-8678Crossref PubMed Scopus (237) Google Scholar, 14Stoddard B.L. Dean A. Koshland Jr., D.E. Biochemistry. 1993; 32: 9310-9316Crossref PubMed Scopus (111) Google Scholar) and for Thermus thermophilus 3-isopropylmalate dehydrogenase (15Imada K. Sato M. Tanaka N. Katsube Y. Matsuura Y. Oshima T. J. Mol. Biol. 1991; 222: 725-738Crossref PubMed Scopus (217) Google Scholar). The former is a homodimeric NADP+-specific enzyme that functions in the bacterial tricarboxylic acid cycle, but that is regulated by phosphorylation rather than by allostery (16Thorsness P.E. Koshland Jr., D.E. J. Biol. Chem. 1987; 262: 10422-10425Abstract Full Text PDF PubMed Google Scholar). The latter is a homodimeric NAD+-specific enzyme in the leucine biosynthetic pathway. Consistent with catalysis of similar reactions, these bacterial enzymes share some similarity in primary structure (∼25% residue identity) and substantial similarity in three-dimensional structure. Differences between residues in catalytic sites of the enzymes have been instructive for analyses of substrate and cofactor specificity (17Chen R. Greer A. Dean A.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 11666-11670Crossref PubMed Scopus (99) Google Scholar, 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar, 19Doyle S.A. Fung S.Y. Koshland Jr., D.E. Biochemistry. 2000; 39: 14348-14355Crossref PubMed Scopus (21) Google Scholar).Both yeast IDH1 and IDH2 subunits share residue sequence identities of ∼32% with E. coli isocitrate dehydrogenase. Based on sequence alignments and modeling, the catalytic isocitrate/Mg2+ binding site of yeast IDH was predicted to be primarily composed of residues from IDH2 (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar), because IDH2 contains identities for nine key residues in the catalytic site of the E. coli enzyme (12Hurley J.H. Thorsness P.E. Ramalingam V. Helmers N.H. Koshland Jr., D.E. Stroud R.M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 8635-8639Crossref PubMed Scopus (206) Google Scholar). IDH1, however, contains identities for only five of these nine residues, and was proposed to bind but not catalytically alter isocitrate. These predictions have been confirmed by results summarized in Table I of site-directed mutagenesis studies (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar, 22Panisko, E. A., Subunit Interactions of Saccharomyces cerevisiae NAD+-dependent Isocitrate Dehydrogenase.Doctoral Dissertation, 2000, University of Texas Health Science Center, San Antonio, TX.Google Scholar). Among shared residues in isocitrate binding sites, IDH1 and IDH2 each contain a serine residue analogous to bacterial Ser-113. The latter is the site for phosphorylation of theE. coli enzyme in vivo (16Thorsness P.E. Koshland Jr., D.E. J. Biol. Chem. 1987; 262: 10422-10425Abstract Full Text PDF PubMed Google Scholar), a modification that inactivates the enzyme by preventing binding of isocitrate (23Dean A.M. Lee M.H.I. Koshland Jr., D.E. J. Biol. Chem. 1989; 264: 20482-20486Abstract Full Text PDF PubMed Google Scholar). Alanine replacement of the analogous Ser-98 in IDH2 was found to profoundly reduce catalysis. Similar replacement in Ser-92 of IDH1 had much less of an effect on catalytic capacity; however, it dramatically reduced cooperativity and allosteric activation by AMP (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar). Replacement of the serine residue in either yeast subunit eliminated half of the holoenzyme isocitrate binding sites, and the combination of these residue replacements in both subunits prevented isocitrate binding (Ref. 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar and Table I). Thus, both the IDH1 and IDH2 sites bind isocitrate but for different kinetic functions. In other studies, the four of nine residues that differ in each of the putative IDH1 and IDH2 isocitrate binding sites were replaced by the corresponding residues in the other subunit site (22Panisko, E. A., Subunit Interactions of Saccharomyces cerevisiae NAD+-dependent Isocitrate Dehydrogenase.Doctoral Dissertation, 2000, University of Texas Health Science Center, San Antonio, TX.Google Scholar). Mutant enzymes containing these reciprocal residue replacements in IDH1 (A108R/F136Y/T241D/N245D) and/or in IDH2 (R114A/Y142F/D248T/D252N) were found to retain the wild-type number of isocitrate binding sites (Ref. 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar and Table I), indicating that these replacements were permissive for binding at each site. However, the mutant enzyme with residue replacements in IDH2 retained very little catalytic activity, and the mutant enzyme with residue replacements in IDH1 exhibited no allosteric activation by AMP nor any binding of AMP (20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). Thus, the unique residues in each subunit isocitrate binding site are essential for different functions in catalysis (IDH2 site) or in allosteric regulation (IDH1 site).Table ISummary of effects of residue replacements in isocitrate sites of IDHEnzyme1-aResults of residue replacements were reported by Linet al. (20) and Lin and McAlister-Henn (21).RelativeVmaxIsocitrate binding sitesAMP effect1-bThe AMP effect is the fold reduction inKD values for isocitrate binding measured in the presence versus the absence of 100 μm AMP (21).IDH1/IDH2—43.5IDH1S92A/IDH2↓∼11x20IDH1/IDH2S98A↓∼150x21.5IDH1S92A/IDH2S98AND1-cND, no measurable catalytic activity or ligand binding.NDNDIDH1A108R,F136Y,T241D,N245D/IDH2↓∼17x40IDH1/IDH2R114A,Y142F,D248T,D252N↓>150x41.5IDH1A108R,F136Y,T241D,N245D/IDH2R114A,Y142F,D248T,D252NND401-a Results of residue replacements were reported by Linet al. (20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar) and Lin and McAlister-Henn (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar).1-b The AMP effect is the fold reduction inKD values for isocitrate binding measured in the presence versus the absence of 100 μm AMP (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar).1-c ND, no measurable catalytic activity or ligand binding. Open table in a new tab Based on these and other results, the homologous yeast IDH subunits appear to be an exceptional model for divergent evolution. The catalytic isocitrate/Mg2+ binding site in IDH2 has been highly conserved, and a similar isocitrate binding site in IDH1 has evolved for regulatory function. In the current study, we use mutagenesis to investigate putative nucleotide binding sites in each subunit. Our hypothesis is that residues important for cofactor NAD+ binding are primarily contributed by IDH2, and that the AMP binding site is comprised of analogous residues of IDH1. A corollary is that residues in the putative NAD+ binding site of IDH2 are likely to be evolutionarily conserved with residues in other cofactor binding sites, whereas analogous residues in the putative AMP binding site in IDH1 are likely to have diverged for binding of the structurally related allosteric activator.Yeast IDH also displays complex interdependencies among various ligands for binding to the enzyme (24Kuehn G.D. Barnes L.D. Atkinson D.E. Biochemistry. 1971; 10: 3945-3951Crossref PubMed Scopus (34) Google Scholar). For example, the presence of isocitrate is a prerequisite for AMP binding, and we have shown the specific nature of this requirement is binding of isocitrate by the regulatory IDH1 site (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). In this report, we also use mutant and wild-type enzymes to investigate the dependence upon citrate or isocitrate to obtain NAD+ binding and the dependence upon Mg2+ for binding of other ligands of the enzyme.DISCUSSIONKinetic and ligand binding analyses described in this report suggest that yeast IDH has distinct but homologous nucleotide binding sites for NAD+ and for AMP. The cofactor binding site contains residues of the IDH2 subunit that are homologous with those in cofactor binding sites of other NAD+-specific decarboxylating dehydrogenases (Fig. 1). Adjacent Asp-286 and Ile-287 residues of IDH2 apparently correspond with adjacent Asp-279 and Ile-280 residues of T. thermophilus 3-isopropylmalate dehydrogenase, residues implicated as important for cofactor specificity (17Chen R. Greer A. Dean A.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 11666-11670Crossref PubMed Scopus (99) Google Scholar, 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar). In NADP+-specific isocitrate dehydrogenases, equivalent positions are occupied by adjacent or nearby lysine (or arginine) and tyrosine residues. Thus, replacement of IDH2 Asp-286 and Ile-287 with alanine residues produces a dramatic reduction in apparent affinity for NAD+ and a concomitant reduction in the catalytic capacity of IDH. Similar effects on affinity for NADP+ and on velocity were reported for mutant enzymes containing replacements for Arg-314 and Tyr-316 residues of mammalian mitochondrial NADP+-specific isocitrate dehydrogenase (30Lee P. Colman R.F. Arch. Biochem. Biophys. 2002; 401: 81-90Crossref PubMed Scopus (14) Google Scholar). In addition, His-281 of IDH2 is important for cofactor binding, since the most dramatic effect of alanine replacement of this residue is a reduction in affinity for NAD+. His-281 thus appears to be the functional homologue of a specific histidine residue in the cofactor binding sites of the T. thermophilus enzyme (His-274, Ref. 15Imada K. Sato M. Tanaka N. Katsube Y. Matsuura Y. Oshima T. J. Mol. Biol. 1991; 222: 725-738Crossref PubMed Scopus (217) Google Scholar), of the mammalian enzyme mentioned above (His-309, Ref. 31Huang Y.C. Colman R.F. Biochemistry. 2002; 41: 5637-5643Crossref PubMed Scopus (13) Google Scholar), and of E. coli isocitrate dehydrogenase (His-339, Refs. 13Hurley J.H. Dean A.M. Koshland Jr., D.E. Stroud R.M. Biochemistry. 1991; 30: 8671-8678Crossref PubMed Scopus (237) Google Scholar and 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar). Unlike the aspartate/isoleucine or lysine/tyrosine pairs described above, this histidine residue apparently functions in binding NAD+ or NADP+ but is not a determinant of cofactor specificity.The yeast IDH1 subunit also contains an aspartate/isoleucine pair at residue positions 279 and 280, but contains an arginine in residue position 274 that aligns with the histidine in position 281 of IDH2. We have shown that the primary effect associated with alanine replacements for both types of residues in IDH1 is a significant reduction in holoenzyme affinity for AMP. The consequence is a defect in allosteric activation, i.e. for these mutant enzymes, isocitrate binding is unaffected by the presence of AMP. Despite similar effects on holoenzyme affinity for AMP, the D279A/I280A and R274A replacements in IDH1 have different kinetic effects. The latter replacement has a greater effect on apparent Vmax values and significantly reduces cooperativity with respect to isocitrate. These results suggest that, in addition to participating in binding of AMP, IDH1 Arg-274 may function in communication between regulatory and catalytic sites.Overall, results obtained in this study suggest that related but unique sites have evolved in the homologous subunits to facilitate binding of different nucleotide ligands of IDH. The differential functions of IDH1 and IDH2 subunits in nucleotide binding are consistent with previous results showing that both subunits also contribute isocitrate binding sites (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar), but that the site comprised primarily of residues from IDH2 is catalytic whereas the site comprised primarily of residues from IDH1 supports regulatory properties of the holoenzyme. Thus, the catalytic isocitrate/Mg2+- and NAD+ binding sites are contributed by IDH2, whereas the regulatory isocitrate- and AMP binding sites are contributed by IDH1. As might be expected, the IDH2 catalytic site(s) exhibits a more significant conservation of residues in catalytic sites of related enzymes (20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar). In the case of IDH1, some residue conservation is observed for binding of structurally related ligands, but key residue differences appear to be important elements in the evolution of regulatory properties of this complex allosteric enzyme.A key to the communication between catalytic and regulatory sites in IDH is that, while each ligand binding site is primarily comprised of residues from one type of subunit, a few residues are apparently contributed by the other type of subunit. For example, results of previous mutagenesis studies (8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) are consistent with the conclusion that of nine residues in the IDH2 catalytic isocitrate/Mg2+binding site, two are contributed by IDH1 and, reciprocally, the analogous two residues of nine in the IDH1 regulatory isocitrate binding site are contributed by IDH2. These and other results from yeast two-hybrid studies (8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) support a model for a heterodimer of IDH1 and IDH2 subunits as the basic structural/functional unit of the holoenzyme. Consistent with reciprocal subunit contributions to isocitrate binding sites, we have also reported kinetic analyses of other mutant enzymes that indicate similar potential intersubunit contributions to nucleotide binding sites (7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). These studies will be expanded by replacement of other residues with similar putative intersubunit functions and by direct ligand binding analyses.In current and previous studies (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar), we have also evaluated some of the complex interrelationships among various ligands for binding by yeast IDH. Cumulative results suggest the following conclusions: (A) Despite residue differences in the isocitrate binding sites, the catalytic site in IDH2 and the regulatory site in IDH1 bind isocitrate with similar affinity. Binding at catalytic and regulatory sites is independent, since isocitrate binding at either site is not affected by loss of isocitrate binding at the other site (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). (B) Nucleotide binding sites are also independent of each other, because residue replacements examined in this study primarily affect either NAD+ or AMP binding (Fig. 4). (C) Binding of isocitrate by the catalytic IDH2 site, but not by the regulatory IDH1 site, requires Mg2+. Thus, the substrate bound by the enzyme is a complex of isocitrate/Mg2+. Furthermore, that isocitrate binding by the IDH1 site is independent of Mg2+ is additional evidence for evolutionary divergence of this site for regulatory rather than catalytic function. (D) A prerequisite for binding of NAD+ is binding of the (iso)citrate/Mg2+ complex by the catalytic IDH2 site. No binding of cofactor is observed in the absence of either (iso)citrate or Mg2+. (E) A prerequisite for binding of AMP is binding of isocitrate by the IDH1 regulatory binding site. An additional prerequisite for binding of AMP is some subsequent change(s) in the holoenzyme elicited by isocitrate binding at the IDH1 site, since replacement of the four non-identical of nine residues in the IDH1 isocitrate binding site with corresponding residues from the IDH2 site is permissive for isocitrate binding but not for AMP binding (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar).Collectively, these results suggest that only isocitrate binding by the regulatory IDH1 site occurs in the absence of other ligands of the enzyme. The complex prerequisites for binding of other ligands may reflect mechanisms for tight control of IDH in vivo,e.g. to ensure that binding and sequestering of a common tricarboxylic acid cycle cofactor occurs only in the presence of sufficient substrate (isocitrate/Mg2+), and to ensure that allosteric activation by AMP occurs only when concentrations of isocitrate are sufficiently elevated. Mitochondrial NAD+-specific isocitrate dehydrogenase (IDH)1 catalyzes a rate-limiting step in the tricarboxylic acid cycle and is subject to complex allosteric regulation. In particular, because of allosteric activation of the mammalian enzyme by ADP (1Chen R.F. Plaut G.W.E. Biochemistry. 1963; 2: 1023-1032Crossref PubMed Scopus (122) Google Scholar) and of the yeast enzyme by AMP, IDH is proposed to regulate metabolic flux in response to energy needs of the cell (2Hathaway J.A. Atkinson D.E. J. Biol. Chem. 1963; 238: 2875-2881Abstract Full Text PDF PubMed Google Scholar). Saccharomyces cerevisiae IDH is an octamer composed of four each of two homologous subunits, IDH1 and IDH2 (3Keys D.A. McAlister-Henn L. J. Bacteriol. 1990; 172: 4280-4287Crossref PubMed Google Scholar). The mature polypeptides are similar in size (349 and 354 amino acid residues, respectively) and share 42% residue sequence identity (4Cupp J.R. McAlister-Henn L. J. Biol. Chem. 1991; 266: 22199-22205Abstract Full Text PDF PubMed Google Scholar, 5Cupp J.R. McAlister-Henn L. J. Biol. Chem. 1992; 267: 16417-16423Abstract Full Text PDF PubMed Google Scholar). Both subunits are essential for holoenzyme structure and function, although, as described below, catalytic function has been primarily attributed to IDH2 whereas regulatory functions have been assigned to IDH1 (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar). Mammalian IDH contains three different types of subunits that share significant homology with those of yeast IDH, an α-subunit with catalytic functions similar to those of yeast IDH2, and β- and γ-subunits that are presumed to impart regulatory properties (9Nichols B.J. Hall L. Perry A.C.F. Denton R.M. Biochem. J. 1993; 295: 347-350Crossref PubMed Scopus (36) Google Scholar, 10Nichols B.J. Perry A.C. Hall L. Denton R.M. Biochem. J. 1995; 310: 917-922Crossref PubMed Scopus (29) Google Scholar). Although crystallographic data are unavailable for yeast and mammalian IDHs, these enzymes share substantial similarity in sequence with several bacterial decarboxylating dehydrogenases for which three-dimensional structures are available. As a family, these dehydrogenases are unique in that they lack the classic Rossman fold described for other enzymes that bind NAD(P)+ (11Rossman M.G. Moras D. Olsen K.W. Nature. 1974; 250: 194-199Crossref PubMed Scopus (1161) Google Scholar). Particularly useful for analyses of yeast IDH are structures reported for Escherichia coli isocitrate dehydrogenase (12Hurley J.H. Thorsness P.E. Ramalingam V. Helmers N.H. Koshland Jr., D.E. Stroud R.M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 8635-8639Crossref PubMed Scopus (206) Google Scholar, 13Hurley J.H. Dean A.M. Koshland Jr., D.E. Stroud R.M. Biochemistry. 1991; 30: 8671-8678Crossref PubMed Scopus (237) Google Scholar, 14Stoddard B.L. Dean A. Koshland Jr., D.E. Biochemistry. 1993; 32: 9310-9316Crossref PubMed Scopus (111) Google Scholar) and for Thermus thermophilus 3-isopropylmalate dehydrogenase (15Imada K. Sato M. Tanaka N. Katsube Y. Matsuura Y. Oshima T. J. Mol. Biol. 1991; 222: 725-738Crossref PubMed Scopus (217) Google Scholar). The former is a homodimeric NADP+-specific enzyme that functions in the bacterial tricarboxylic acid cycle, but that is regulated by phosphorylation rather than by allostery (16Thorsness P.E. Koshland Jr., D.E. J. Biol. Chem. 1987; 262: 10422-10425Abstract Full Text PDF PubMed Google Scholar). The latter is a homodimeric NAD+-specific enzyme in the leucine biosynthetic pathway. Consistent with catalysis of similar reactions, these bacterial enzymes share some similarity in primary structure (∼25% residue identity) and substantial similarity in three-dimensional structure. Differences between residues in catalytic sites of the enzymes have been instructive for analyses of substrate and cofactor specificity (17Chen R. Greer A. Dean A.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 11666-11670Crossref PubMed Scopus (99) Google Scholar, 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar, 19Doyle S.A. Fung S.Y. Koshland Jr., D.E. Biochemistry. 2000; 39: 14348-14355Crossref PubMed Scopus (21) Google Scholar). Both yeast IDH1 and IDH2 subunits share residue sequence identities of ∼32% with E. coli isocitrate dehydrogenase. Based on sequence alignments and modeling, the catalytic isocitrate/Mg2+ binding site of yeast IDH was predicted to be primarily composed of residues from IDH2 (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar), because IDH2 contains identities for nine key residues in the catalytic site of the E. coli enzyme (12Hurley J.H. Thorsness P.E. Ramalingam V. Helmers N.H. Koshland Jr., D.E. Stroud R.M. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 8635-8639Crossref PubMed Scopus (206) Google Scholar). IDH1, however, contains identities for only five of these nine residues, and was proposed to bind but not catalytically alter isocitrate. These predictions have been confirmed by results summarized in Table I of site-directed mutagenesis studies (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar, 22Panisko, E. A., Subunit Interactions of Saccharomyces cerevisiae NAD+-dependent Isocitrate Dehydrogenase.Doctoral Dissertation, 2000, University of Texas Health Science Center, San Antonio, TX.Google Scholar). Among shared residues in isocitrate binding sites, IDH1 and IDH2 each contain a serine residue analogous to bacterial Ser-113. The latter is the site for phosphorylation of theE. coli enzyme in vivo (16Thorsness P.E. Koshland Jr., D.E. J. Biol. Chem. 1987; 262: 10422-10425Abstract Full Text PDF PubMed Google Scholar), a modification that inactivates the enzyme by preventing binding of isocitrate (23Dean A.M. Lee M.H.I. Koshland Jr., D.E. J. Biol. Chem. 1989; 264: 20482-20486Abstract Full Text PDF PubMed Google Scholar). Alanine replacement of the analogous Ser-98 in IDH2 was found to profoundly reduce catalysis. Similar replacement in Ser-92 of IDH1 had much less of an effect on catalytic capacity; however, it dramatically reduced cooperativity and allosteric activation by AMP (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar). Replacement of the serine residue in either yeast subunit eliminated half of the holoenzyme isocitrate binding sites, and the combination of these residue replacements in both subunits prevented isocitrate binding (Ref. 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar and Table I). Thus, both the IDH1 and IDH2 sites bind isocitrate but for different kinetic functions. In other studies, the four of nine residues that differ in each of the putative IDH1 and IDH2 isocitrate binding sites were replaced by the corresponding residues in the other subunit site (22Panisko, E. A., Subunit Interactions of Saccharomyces cerevisiae NAD+-dependent Isocitrate Dehydrogenase.Doctoral Dissertation, 2000, University of Texas Health Science Center, San Antonio, TX.Google Scholar). Mutant enzymes containing these reciprocal residue replacements in IDH1 (A108R/F136Y/T241D/N245D) and/or in IDH2 (R114A/Y142F/D248T/D252N) were found to retain the wild-type number of isocitrate binding sites (Ref. 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar and Table I), indicating that these replacements were permissive for binding at each site. However, the mutant enzyme with residue replacements in IDH2 retained very little catalytic activity, and the mutant enzyme with residue replacements in IDH1 exhibited no allosteric activation by AMP nor any binding of AMP (20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). Thus, the unique residues in each subunit isocitrate binding site are essential for different functions in catalysis (IDH2 site) or in allosteric regulation (IDH1 site). Based on these and other results, the homologous yeast IDH subunits appear to be an exceptional model for divergent evolution. The catalytic isocitrate/Mg2+ binding site in IDH2 has been highly conserved, and a similar isocitrate binding site in IDH1 has evolved for regulatory function. In the current study, we use mutagenesis to investigate putative nucleotide binding sites in each subunit. Our hypothesis is that residues important for cofactor NAD+ binding are primarily contributed by IDH2, and that the AMP binding site is comprised of analogous residues of IDH1. A corollary is that residues in the putative NAD+ binding site of IDH2 are likely to be evolutionarily conserved with residues in other cofactor binding sites, whereas analogous residues in the putative AMP binding site in IDH1 are likely to have diverged for binding of the structurally related allosteric activator. Yeast IDH also displays complex interdependencies among various ligands for binding to the enzyme (24Kuehn G.D. Barnes L.D. Atkinson D.E. Biochemistry. 1971; 10: 3945-3951Crossref PubMed Scopus (34) Google Scholar). For example, the presence of isocitrate is a prerequisite for AMP binding, and we have shown the specific nature of this requirement is binding of isocitrate by the regulatory IDH1 site (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). In this report, we also use mutant and wild-type enzymes to investigate the dependence upon citrate or isocitrate to obtain NAD+ binding and the dependence upon Mg2+ for binding of other ligands of the enzyme. DISCUSSIONKinetic and ligand binding analyses described in this report suggest that yeast IDH has distinct but homologous nucleotide binding sites for NAD+ and for AMP. The cofactor binding site contains residues of the IDH2 subunit that are homologous with those in cofactor binding sites of other NAD+-specific decarboxylating dehydrogenases (Fig. 1). Adjacent Asp-286 and Ile-287 residues of IDH2 apparently correspond with adjacent Asp-279 and Ile-280 residues of T. thermophilus 3-isopropylmalate dehydrogenase, residues implicated as important for cofactor specificity (17Chen R. Greer A. Dean A.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 11666-11670Crossref PubMed Scopus (99) Google Scholar, 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar). In NADP+-specific isocitrate dehydrogenases, equivalent positions are occupied by adjacent or nearby lysine (or arginine) and tyrosine residues. Thus, replacement of IDH2 Asp-286 and Ile-287 with alanine residues produces a dramatic reduction in apparent affinity for NAD+ and a concomitant reduction in the catalytic capacity of IDH. Similar effects on affinity for NADP+ and on velocity were reported for mutant enzymes containing replacements for Arg-314 and Tyr-316 residues of mammalian mitochondrial NADP+-specific isocitrate dehydrogenase (30Lee P. Colman R.F. Arch. Biochem. Biophys. 2002; 401: 81-90Crossref PubMed Scopus (14) Google Scholar). In addition, His-281 of IDH2 is important for cofactor binding, since the most dramatic effect of alanine replacement of this residue is a reduction in affinity for NAD+. His-281 thus appears to be the functional homologue of a specific histidine residue in the cofactor binding sites of the T. thermophilus enzyme (His-274, Ref. 15Imada K. Sato M. Tanaka N. Katsube Y. Matsuura Y. Oshima T. J. Mol. Biol. 1991; 222: 725-738Crossref PubMed Scopus (217) Google Scholar), of the mammalian enzyme mentioned above (His-309, Ref. 31Huang Y.C. Colman R.F. Biochemistry. 2002; 41: 5637-5643Crossref PubMed Scopus (13) Google Scholar), and of E. coli isocitrate dehydrogenase (His-339, Refs. 13Hurley J.H. Dean A.M. Koshland Jr., D.E. Stroud R.M. Biochemistry. 1991; 30: 8671-8678Crossref PubMed Scopus (237) Google Scholar and 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar). Unlike the aspartate/isoleucine or lysine/tyrosine pairs described above, this histidine residue apparently functions in binding NAD+ or NADP+ but is not a determinant of cofactor specificity.The yeast IDH1 subunit also contains an aspartate/isoleucine pair at residue positions 279 and 280, but contains an arginine in residue position 274 that aligns with the histidine in position 281 of IDH2. We have shown that the primary effect associated with alanine replacements for both types of residues in IDH1 is a significant reduction in holoenzyme affinity for AMP. The consequence is a defect in allosteric activation, i.e. for these mutant enzymes, isocitrate binding is unaffected by the presence of AMP. Despite similar effects on holoenzyme affinity for AMP, the D279A/I280A and R274A replacements in IDH1 have different kinetic effects. The latter replacement has a greater effect on apparent Vmax values and significantly reduces cooperativity with respect to isocitrate. These results suggest that, in addition to participating in binding of AMP, IDH1 Arg-274 may function in communication between regulatory and catalytic sites.Overall, results obtained in this study suggest that related but unique sites have evolved in the homologous subunits to facilitate binding of different nucleotide ligands of IDH. The differential functions of IDH1 and IDH2 subunits in nucleotide binding are consistent with previous results showing that both subunits also contribute isocitrate binding sites (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar), but that the site comprised primarily of residues from IDH2 is catalytic whereas the site comprised primarily of residues from IDH1 supports regulatory properties of the holoenzyme. Thus, the catalytic isocitrate/Mg2+- and NAD+ binding sites are contributed by IDH2, whereas the regulatory isocitrate- and AMP binding sites are contributed by IDH1. As might be expected, the IDH2 catalytic site(s) exhibits a more significant conservation of residues in catalytic sites of related enzymes (20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar). In the case of IDH1, some residue conservation is observed for binding of structurally related ligands, but key residue differences appear to be important elements in the evolution of regulatory properties of this complex allosteric enzyme.A key to the communication between catalytic and regulatory sites in IDH is that, while each ligand binding site is primarily comprised of residues from one type of subunit, a few residues are apparently contributed by the other type of subunit. For example, results of previous mutagenesis studies (8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) are consistent with the conclusion that of nine residues in the IDH2 catalytic isocitrate/Mg2+binding site, two are contributed by IDH1 and, reciprocally, the analogous two residues of nine in the IDH1 regulatory isocitrate binding site are contributed by IDH2. These and other results from yeast two-hybrid studies (8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) support a model for a heterodimer of IDH1 and IDH2 subunits as the basic structural/functional unit of the holoenzyme. Consistent with reciprocal subunit contributions to isocitrate binding sites, we have also reported kinetic analyses of other mutant enzymes that indicate similar potential intersubunit contributions to nucleotide binding sites (7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). These studies will be expanded by replacement of other residues with similar putative intersubunit functions and by direct ligand binding analyses.In current and previous studies (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar), we have also evaluated some of the complex interrelationships among various ligands for binding by yeast IDH. Cumulative results suggest the following conclusions: (A) Despite residue differences in the isocitrate binding sites, the catalytic site in IDH2 and the regulatory site in IDH1 bind isocitrate with similar affinity. Binding at catalytic and regulatory sites is independent, since isocitrate binding at either site is not affected by loss of isocitrate binding at the other site (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). (B) Nucleotide binding sites are also independent of each other, because residue replacements examined in this study primarily affect either NAD+ or AMP binding (Fig. 4). (C) Binding of isocitrate by the catalytic IDH2 site, but not by the regulatory IDH1 site, requires Mg2+. Thus, the substrate bound by the enzyme is a complex of isocitrate/Mg2+. Furthermore, that isocitrate binding by the IDH1 site is independent of Mg2+ is additional evidence for evolutionary divergence of this site for regulatory rather than catalytic function. (D) A prerequisite for binding of NAD+ is binding of the (iso)citrate/Mg2+ complex by the catalytic IDH2 site. No binding of cofactor is observed in the absence of either (iso)citrate or Mg2+. (E) A prerequisite for binding of AMP is binding of isocitrate by the IDH1 regulatory binding site. An additional prerequisite for binding of AMP is some subsequent change(s) in the holoenzyme elicited by isocitrate binding at the IDH1 site, since replacement of the four non-identical of nine residues in the IDH1 isocitrate binding site with corresponding residues from the IDH2 site is permissive for isocitrate binding but not for AMP binding (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar).Collectively, these results suggest that only isocitrate binding by the regulatory IDH1 site occurs in the absence of other ligands of the enzyme. The complex prerequisites for binding of other ligands may reflect mechanisms for tight control of IDH in vivo,e.g. to ensure that binding and sequestering of a common tricarboxylic acid cycle cofactor occurs only in the presence of sufficient substrate (isocitrate/Mg2+), and to ensure that allosteric activation by AMP occurs only when concentrations of isocitrate are sufficiently elevated. Kinetic and ligand binding analyses described in this report suggest that yeast IDH has distinct but homologous nucleotide binding sites for NAD+ and for AMP. The cofactor binding site contains residues of the IDH2 subunit that are homologous with those in cofactor binding sites of other NAD+-specific decarboxylating dehydrogenases (Fig. 1). Adjacent Asp-286 and Ile-287 residues of IDH2 apparently correspond with adjacent Asp-279 and Ile-280 residues of T. thermophilus 3-isopropylmalate dehydrogenase, residues implicated as important for cofactor specificity (17Chen R. Greer A. Dean A.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 11666-11670Crossref PubMed Scopus (99) Google Scholar, 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar). In NADP+-specific isocitrate dehydrogenases, equivalent positions are occupied by adjacent or nearby lysine (or arginine) and tyrosine residues. Thus, replacement of IDH2 Asp-286 and Ile-287 with alanine residues produces a dramatic reduction in apparent affinity for NAD+ and a concomitant reduction in the catalytic capacity of IDH. Similar effects on affinity for NADP+ and on velocity were reported for mutant enzymes containing replacements for Arg-314 and Tyr-316 residues of mammalian mitochondrial NADP+-specific isocitrate dehydrogenase (30Lee P. Colman R.F. Arch. Biochem. Biophys. 2002; 401: 81-90Crossref PubMed Scopus (14) Google Scholar). In addition, His-281 of IDH2 is important for cofactor binding, since the most dramatic effect of alanine replacement of this residue is a reduction in affinity for NAD+. His-281 thus appears to be the functional homologue of a specific histidine residue in the cofactor binding sites of the T. thermophilus enzyme (His-274, Ref. 15Imada K. Sato M. Tanaka N. Katsube Y. Matsuura Y. Oshima T. J. Mol. Biol. 1991; 222: 725-738Crossref PubMed Scopus (217) Google Scholar), of the mammalian enzyme mentioned above (His-309, Ref. 31Huang Y.C. Colman R.F. Biochemistry. 2002; 41: 5637-5643Crossref PubMed Scopus (13) Google Scholar), and of E. coli isocitrate dehydrogenase (His-339, Refs. 13Hurley J.H. Dean A.M. Koshland Jr., D.E. Stroud R.M. Biochemistry. 1991; 30: 8671-8678Crossref PubMed Scopus (237) Google Scholar and 18Hurley J.H. Chen R. Dean A.M. Biochemistry. 1996; 35: 5670-5678Crossref PubMed Scopus (86) Google Scholar). Unlike the aspartate/isoleucine or lysine/tyrosine pairs described above, this histidine residue apparently functions in binding NAD+ or NADP+ but is not a determinant of cofactor specificity. The yeast IDH1 subunit also contains an aspartate/isoleucine pair at residue positions 279 and 280, but contains an arginine in residue position 274 that aligns with the histidine in position 281 of IDH2. We have shown that the primary effect associated with alanine replacements for both types of residues in IDH1 is a significant reduction in holoenzyme affinity for AMP. The consequence is a defect in allosteric activation, i.e. for these mutant enzymes, isocitrate binding is unaffected by the presence of AMP. Despite similar effects on holoenzyme affinity for AMP, the D279A/I280A and R274A replacements in IDH1 have different kinetic effects. The latter replacement has a greater effect on apparent Vmax values and significantly reduces cooperativity with respect to isocitrate. These results suggest that, in addition to participating in binding of AMP, IDH1 Arg-274 may function in communication between regulatory and catalytic sites. Overall, results obtained in this study suggest that related but unique sites have evolved in the homologous subunits to facilitate binding of different nucleotide ligands of IDH. The differential functions of IDH1 and IDH2 subunits in nucleotide binding are consistent with previous results showing that both subunits also contribute isocitrate binding sites (6Cupp J.R. McAlister-Henn L. Biochemistry. 1993; 32: 9323-9328Crossref PubMed Scopus (62) Google Scholar, 7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar, 20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar, 21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar), but that the site comprised primarily of residues from IDH2 is catalytic whereas the site comprised primarily of residues from IDH1 supports regulatory properties of the holoenzyme. Thus, the catalytic isocitrate/Mg2+- and NAD+ binding sites are contributed by IDH2, whereas the regulatory isocitrate- and AMP binding sites are contributed by IDH1. As might be expected, the IDH2 catalytic site(s) exhibits a more significant conservation of residues in catalytic sites of related enzymes (20Lin A.-P. McCammon M.T. McAlister-Henn L. Biochemistry. 2001; 40: 14291-14301Crossref PubMed Scopus (24) Google Scholar). In the case of IDH1, some residue conservation is observed for binding of structurally related ligands, but key residue differences appear to be important elements in the evolution of regulatory properties of this complex allosteric enzyme. A key to the communication between catalytic and regulatory sites in IDH is that, while each ligand binding site is primarily comprised of residues from one type of subunit, a few residues are apparently contributed by the other type of subunit. For example, results of previous mutagenesis studies (8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) are consistent with the conclusion that of nine residues in the IDH2 catalytic isocitrate/Mg2+binding site, two are contributed by IDH1 and, reciprocally, the analogous two residues of nine in the IDH1 regulatory isocitrate binding site are contributed by IDH2. These and other results from yeast two-hybrid studies (8Panisko E.A. McAlister-Henn L. J. Biol. Chem. 2001; 276: 1204-1210Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar) support a model for a heterodimer of IDH1 and IDH2 subunits as the basic structural/functional unit of the holoenzyme. Consistent with reciprocal subunit contributions to isocitrate binding sites, we have also reported kinetic analyses of other mutant enzymes that indicate similar potential intersubunit contributions to nucleotide binding sites (7Zhao W.-N. McAlister-Henn L. J. Biol. Chem. 1997; 272: 21811-21877Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). These studies will be expanded by replacement of other residues with similar putative intersubunit functions and by direct ligand binding analyses. In current and previous studies (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar), we have also evaluated some of the complex interrelationships among various ligands for binding by yeast IDH. Cumulative results suggest the following conclusions: (A) Despite residue differences in the isocitrate binding sites, the catalytic site in IDH2 and the regulatory site in IDH1 bind isocitrate with similar affinity. Binding at catalytic and regulatory sites is independent, since isocitrate binding at either site is not affected by loss of isocitrate binding at the other site (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). (B) Nucleotide binding sites are also independent of each other, because residue replacements examined in this study primarily affect either NAD+ or AMP binding (Fig. 4). (C) Binding of isocitrate by the catalytic IDH2 site, but not by the regulatory IDH1 site, requires Mg2+. Thus, the substrate bound by the enzyme is a complex of isocitrate/Mg2+. Furthermore, that isocitrate binding by the IDH1 site is independent of Mg2+ is additional evidence for evolutionary divergence of this site for regulatory rather than catalytic function. (D) A prerequisite for binding of NAD+ is binding of the (iso)citrate/Mg2+ complex by the catalytic IDH2 site. No binding of cofactor is observed in the absence of either (iso)citrate or Mg2+. (E) A prerequisite for binding of AMP is binding of isocitrate by the IDH1 regulatory binding site. An additional prerequisite for binding of AMP is some subsequent change(s) in the holoenzyme elicited by isocitrate binding at the IDH1 site, since replacement of the four non-identical of nine residues in the IDH1 isocitrate binding site with corresponding residues from the IDH2 site is permissive for isocitrate binding but not for AMP binding (21Lin A.-P. McAlister-Henn L. J. Biol. Chem. 2002; 277: 22476-22483Google Scholar). Collectively, these results suggest that only isocitrate binding by the regulatory IDH1 site occurs in the absence of other ligands of the enzyme. The complex prerequisites for binding of other ligands may reflect mechanisms for tight control of IDH in vivo,e.g. to ensure that binding and sequestering of a common tricarboxylic acid cycle cofactor occurs only in the presence of sufficient substrate (isocitrate/Mg2+), and to ensure that allosteric activation by AMP occurs only when concentrations of isocitrate are sufficiently elevated. We thank Dr. Karyl I. Minard for technical advice, and Drs. Minard and Mark T. McCammon for critical reading of this manuscript.
Referência(s)