Cloning and Characterization of the Gene Encoding the IDHl Subunit of NAD+-dependent Isocitrate Dehydrogenase from Saccharomyces cerevisiael

NAD+-dependent isocitrate dehydrogenase from Saccharomyces cerevisiae is composed of two nonidentical subunits, designated IDHl and IDHZ. The gene encoding IDHB was previously cloned and se- quenced (Cupp, J. R., and McAlister-Henn, L. (1991) J. Biol. Chem. 266, 22199-22205), and in this paper we describe the isolation of a yeast genomic clone containing the IDHl gene. A fragment of the IDHl gene was amplified by the polymerase chain reaction method utilizing degenerate oligonucleotides based on tryptic peptide sequences of the purified subunit; this fragment was used to isolate a full length IDHl clone. The nucleotide sequence of the IDHl coding region was determined and encodes a 360-residue polypeptide including an 11-residue mitochondrial targeting pre- sequence. Amino acid sequence comparison between IDHl and IDH2 reveals a 42% sequence identity, and both IDHl and IDHP show -32% identity to Esche-richia coli NAD(P)+-dependent isocitrate dehydrogenase. To examine the function of the IDHl subunit and to determine the metabolic role of NAD+-dependent isocitrate dehydrogenase the IDHl gene was disrupted in a wild type haploid yeast strain and in a haploid strain lacking IDH2. The IDHl disruption strains expressed no detectable IDHl as determined by Western blot analysis, and these strains were found to lack NAD+- dependent isocitrate dehydrogenase activity indicating

NAD+-dependent isocitrate dehydrogenase from Saccharomyces cerevisiae is composed of two nonidentical subunits, designated IDHl and IDHZ. The gene encoding IDHB was previously cloned and sequenced (Cupp, J. R., and McAlister-Henn, L. (1991) J. Biol. Chem. 266,[22199][22200][22201][22202][22203][22204][22205], and in this paper we describe the isolation of a yeast genomic clone containing the IDHl gene. A fragment of the IDHl gene was amplified by the polymerase chain reaction method utilizing degenerate oligonucleotides based on tryptic peptide sequences of the purified subunit; this fragment was used to isolate a full length IDHl clone. The nucleotide sequence of the IDHl coding region was determined and encodes a 360-residue polypeptide including an 11-residue mitochondrial targeting presequence. Amino acid sequence comparison between IDHl and IDH2 reveals a 42% sequence identity, and both IDHl and IDHP show -32% identity to Escherichia coli NAD(P)+-dependent isocitrate dehydrogenase.
To examine the function of the IDHl subunit and to determine the metabolic role of NAD+-dependent isocitrate dehydrogenase the IDHl gene was disrupted in a wild type haploid yeast strain and in a haploid strain lacking IDH2. The IDHl disruption strains expressed no detectable IDHl as determined by Western blot analysis, and these strains were found to lack NAD+dependent isocitrate dehydrogenase activity indicating that IDHl is essential for a functional enzyme. Overexpression of IDHl in a strain containing IDHP restored wild type activity but did not result in increased levels of activity, suggesting that both IDHl and IDHB are required for a functional enzyme. Growth phenotype analysis of the IDHl disruption strains revealed that they grew at a reduced rate on the nonfermentable carbon sources examined (glycerol, lactate, and acetate), consistent with NAD+-dependent isocitrate dehydrogenase performing a critical role in oxidative function of the citric acid cycle. In addition, the IDHl disruption strains grew at wild type rates in the absence of glutamate, indicating that these strains are not glutamate auxotrophs.
* This work was supported by American Cancer Society Grant BE-116. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession numberls) M95203.
$ To whom correspondence should be addressed Dept. of Biological Chemistry, University of California, Irvine, CA 92717.
Isocitrate dehydrogenase catalyzes the conversion of isocitrate to a-ketoglutarate in all cellular organisms. This reaction is an essential and rate-limiting step in the citric acid cycle, and the a-ketoglutarate formed is required for production of glutamate which serves as the precursor for the amino acids arginine, proline, and glutamic acid. Isocitrate dehydrogenase has been purified from a variety of sources, including Escherichia coli (I), Saccharomyces cereuisiae (2), and pig and bovine heart (3,4). Characterization of the prokaryotic NAD(P)+-dependent IDH (NAD(P)+-IDH)' from E. coli has shown it to be a homodimer with a subunit molecular weight of 45,000 (1) which is regulated by phosphorylation ( 5 ) . In contrast, the purified NAD+-dependent IDH (NAD'-IDH) from eukaryotic sources functions as an oligomeric enzyme that is subject to extensive allosteric regulation (6,7).
Characterization of the purified NAD+-IDH enzyme from S. cereuisiue has shown it to be an octamer composed of two nonidentical subunits, designated IDHl and IDH2 (8,9). The apparent molecular weights of IDHl and IDHB were estimated to be -40,000 and -39,000, respectively, by electrophoretic mobility. The enzyme is responsive to cellular energy levels as a result of binding and activation by AMP and NAD' (6). Recently, we reported the cloning and disruption of the gene encoding 1DH2 from S. cereuisiae (10). The amino acid sequence deduced from the gene indicates that IDH2 is synthesized as a precursor of 369 amino acids ( M I 39,694) and is processed upon mitochondrial import to yield a mature protein of 354 amino acids (Mr 37,755). We describe here the cloning and nucleotide sequence of the gene encoding the IDHl subunit of S. cereuisiae NAD+-IDH and compare the deduced amino acid sequence with that of IDH2 as well as NAD(P)+-IDH isozymes from E. coli, S. cereuisiae, and pig. In addition, strains with disruptions in IDHl or both IDHl and IDH2 were constructed and their phenotypes analyzed to determine the metabolic role of NAD'-IDH and to examine the function of the individual IDHl and IDH2 subunits.
Tryptic Peptide Isolation-Mitochondrial NAD+-IDH was partially purified from yeast using methods similar to those described by . Yeast cells were broken and proteins fractionated by ammonium sulfate precipitation. The fraction containing NAD+-IDH was further purified by ion exchange chromatography. Protein samples were electrophoresed in 10% polyacrylamide-sodium dodecyl sulfate gels ( l l ) , and the band migrating at 40,000 Da was excised from the gel and eluted by diffusion using the method described by Hager and Burgess (12). The sodium dodecyl sulfate and Coomassie Blue were removed from the sample by precipitation of the protein with 8 volumes of cold acetone. The protein pellet, containing approximately 1 nmol, was resuspended with 6 M urea in 50 mM KPO4, pH 7.5, and incubated at 37 "C for 1 h. Following denaturation the urea was diluted to 1 M and trypsin added at a 1:50 (w/w) ratio. The sample was incubated at 37 "C for 2 h, after which time a second aliquot of trypsin was added and the incubation continued for an additional 2 h. Tryptic peptides were separated by high performance liquid chromatography (Beckman) with a reversephase C, column (Rainin); a linear H20/acetonitrile gradient was used for peptide elution. Peptides were collected by monitoring at A 2 1 4 n m and used for amino acid sequence analysis (Biotechnology Instrumentation Facility, Dr. Gary Hathaway, University of California, Riverside).
Recombinant DNA Techniques and Nucleotide Sequencing-Plasmids were amplified in the E. coli strain DH5aF' and purified using the alkaline lysis method (13). Separation of plasmids and restriction fragments was achieved using 0.8% agarose gels, and DNA fragments were purified by electroelution using an IBI apparatus. Purified restriction fragments were cloned into specified plasmids using T4 DNA ligase. The shuttle vectors YCp50 (14) and YEp352 (15) were utilized for expression of the ZDHl gene; a 1.8-kb BglII fragment containing the ZDHl gene was inserted into the respective shuttle vector at the BamHI site. The YEp352-ZDHIIIDH2 plasmid was constructed by insertion of the 1.8-kb BglII ZDHl gene fragment at the BamHI site of YEp352 followed by insertion of a 4.8-kb Hind111 fragment containing the IDH2 gene. Plasmids were transformed into yeast using the lithium chloride protocol (16).
Fragments of the IDHl gene were subcloned into Bluescribe plasmids (Stratagene) and used for nucleotide sequencing. Double stranded dideoxynucleotide sequencing as described by Wang (17) was performed using Sequenase Version 2.0 (United States Biochemical Corp.). Commercially available primers corresponding to plasmid sequences (T3 and T7) were used in all sequencing reactions.
Polymerase Chain Reaction-Polymerase chain reaction (PCR) experiments were performed with 0.25 units of Tag DNA polymerase (Boehringer Mannheim), 200 pmol of each primer (Operon, San Pablo, CA), and 0.2 mM dNTPs in a volume of 50 gl. The reactions were cycled 30 times for 45 s at 94 "C, 30 s at 42 "C, and 1 min at 72 "C using an Ericomp thermocycler. DNA fragments amplified by PCR were gel purified and subcloned into Bluescribe plasmids (Stratagene). These constructs were sequenced using the double strand DNA sequencing technique described above.
Gene Disruption-The AZDHI strain was constructed by deletioninsertion (18). The 1.8-kb BglII fragment in Bluescribe was digested with EcoRV and BalI to remove 365-bp from the coding region of the ZDHl gene. A 1.8-kb DNA fragment containing the LEU2 gene was ligated into the plasmid, and recombinant plasmids were selected by screening with a 32P-labeled DNA fragment containing the LEU2 gene. Selected plasmids were amplified and digested with XbaI to yield a linear 2.7-kb fragment. The purified fragment was transformed into yeast strain S173-6B and the AZDH2 strain. Replacement of the wild type gene in Leu+ transformants by integration of the disrupted gene at the ZDHl locus was confirmed by Southern blot analysis. DNA probes used for screening and Southern blot analysis were labeled by the random primer technique (19) using [cP~*P]CTP (Amersham).
Mitochondrial Isolation and Assays-Yeast mitochondria were prepared as described by Daum et al. (20) from strains cultivated overnight in YP medium with 2% glycerol, 2% lactate as carbon sources. Briefly, pelleted cells were washed and resuspended in a buffer containing zymolase to produce spheroplasts. Cells were broken using a Dounce homogenizer and mitochondria isolated by differential centrifugation at 10,000 rpm. Mitochondrial pellets were suspended in a lysis buffer containing 50 mM Tris-HC1, pH 7.5, 1 mM EDTA, and 0.1 mM phenylmethylsulfonyl fluoride. Extracts for use in enzyme and protein assays were obtained by vortexing with glass beads followed by centrifugation in a microcentifuge to remove membranes. Protein concentrations were determined using the Bradford dye binding method (21). NAD+-IDH activity was measured spectrophotometrically as previously described (9). Assays contained 40 mM Tris-HCl (pH 7.5), 4 mM MgCl,, 0.25 mM NAD', and 5-40 r g of mitochondrial protein extracts. All assays were performed at 23 "C and initiated by the addition of 2.5 mM DL-isocitrate.
Western Blot Analysis-Whole cell extracts were prepared by the glass bead lysis technique and were electrophoresed on 10% polyacrylamide-sodium dodecyl sulfate gels (11). Separated proteins were transferred to polyvinylidene fluoride membranes (Immobilon, Millipore) using a graphite dry blotting apparatus (22). The membrane was stained with Coomassie Blue, blocked with 5% bovine serum albumin, and incubated in a 1:300 dilution of anti-IDH antiserum (9). Detection was performed using 1251-labeled protein-A followed by autoradiography.

RESULTS
Cloning and Nucleotide Sequence-Yeast mutants lacking individual subunits of NAD+-IDH were previously described (9), and the ZDH2 gene was cloned by complementation of these mutants (10). Attempts to similarly clone ZDHl by complementation, however, proved unsuccessful. We therefore set out to obtain the ZDHl gene using the PCR technique based on partial amino acid sequence analysis. IDHl was purified as described under "Experimental Procedures" and digested with trypsin to yield peptides for sequence analysis. Amino acid sequence was obtained for four tryptic peptides. These sequences were aligned with the deduced amino acid sequence of IDH2 to predict their relative positions in IDH1.
Degenerate oligonucleotides based on peptides A and D (cf. Fig. 1B) were used as primers for PCR with yeast genomic DNA as the template. A DNA fragment of -900-bp was obtained, consistent with the peptide alignment with IDH2. The nucleotide sequence of this fragment was determined, and the deduced amino acid sequence contained the tryptic peptides B and C obtained from IDHl ( Fig. 1B) and exhibited 43% identity with the aligned amino acid sequence of IDH2. These results strongly suggested that the amplified fragment encodes a portion of the yeast IDHl gene.
To obtain the full length IDHl gene, the 900-bp PCR product was used as a probe to screen bacterial colonies transformed with a yeast genomic DNA library. Restriction mapping conducted with plasmids from positive isolates showed that all contained DNA inserts with overlapping restriction maps from the same region of genomic DNA. A 1.8-kb BglII DNA fragment which was present in all positive isolates was subcloned for further studies. A partial restriction map of the 1.8-kb BglII fragment and the strategy used for nucleotide sequencing of this region are shown in Fig. lA. An open reading frame was found to encode the amino terminus of the mature IDHl polypeptide (nucleotides 34-72, Fig. 1B).
The first methionine codon of the open reading frame is located 11 codons upstream from the codon for the first amino acid of the mature IDHl polypeptide (9). This presumptive mitochondrial targeting peptide is similar in size and properties to the IDH2 mitochondrial import peptide (10). Assuming that translation begins at this methionine codon, the deduced amino acid sequence predicts a precursor protein of 360 amino acids ( M , 39,299); cleavage of an 11-residue amino-terminal peptide upon mitochondrial import would yield the mature protein of 349 amino acids (Mr 38,001). ZDHl Gene Disruption-To confirm the identity of the IDHl gene, the cloned 1.8-kb BglII fragment was disrupted and used for one-step genomic replacement as described under "Experimental Procedures." Fig. 2A illustrates the deletioninsertion construct in which 365-bp within the ZDHl coding region were replaced by a 1.8-kb fragment containing the yeast LEU2 gene. The fragment containing the disrupted Met L e u A s n A r g T h r I l e A l a L y s Arg T h r L e u A l a T h r A l a A l a G l n A l a G l u A r g T h r L e u P r o L y e L y s Tyr 90 105 120 135 150 GGC GGT CGT TTC ACC GTC ACT TTG ATA CCT GGT GAC GGT GTT GGG AAA GAA ATC ACT GAT TCA GTG AGA ACC ATT G l y G l y A r g P h e T h r V a l T h r L e u I l e P r o G l y A s p G l y V a l G l y L y s G l u I l e T h r A s p Ser V a l A r g T h r I l e Ser T h r Met gene was used for genomic replacement in both a LEU2haploid wild type strain and a LEU2-AIDH2 strain, and the replacements confirmed by Southern blot analysis. As shown in Fig. 2B, genomic DNA isolated from the wild type and the AIDH2 strains contained two PuuII fragments of the expected sizes of 5.6-and 1.0-kb which hybridized with a 32P-labeled IDHl probe. The same probe hybridized with a single 8-kb PuuII fragment in DNA from the IDHl disruption strain ( AIDHl) and the IDHl/IDH2 disruption strain (AIDHI/ IDHZ), indicating the expected loss of a PuuII site and an increase in size of 1.4 kb due to the gene replacement. To confirm that the alteration at the IDHl locus was due to integration of the LEU2 disruption construct, the same blot was analyzed by probing with a 32P-labeled LEU2 gene. In addition to the endogenous LEU2 gene found on a 9.5-kb PuuII fragment in DNA from all strains, an 8.0-kb fragment in DNA from the IDHl disruption strains hybridized with the LEU2 probe, indicating integration of the LEU2 gene at the IDHl locus in both the AIDHl and the AIDHlIIDH2 strains.
To determine effects of the disruptions on expression of IDHl and IDH2, mitochondrial extracts from wild type and disruption strains were examined by Western blot analysis using IgG specific for the yeast NAD+-IDH holoenzyme (9). IDHl and IDH2 in extracts from the wild type strain are resolved as polypeptides with apparent molecular weights of -40,000 and -39,000, respectively (Fig. 3, lane 1 ). Mitochondrial extracts from the AIDHI strain lack the 40,000 Da polypeptide (Fig. 3, lane 2 ) , consistent with disruption of the IDHl gene. In addition, the disruption of IDHl does not appear to affect expression of ZDH2, a result similar to that previously obtained following reciprocal disruption of IDH2 (10). This confirms that IDHl and 1DH2 are encoded at separate genomic loci and that the two subunits are independently expressed. Fig. 3, lane 4 also shows that extracts from the AIDHlIIDH2 strain contain no proteins which cross-react with the NAD+-IDH antiserum; this provides evi-NAD+-dependent Isocitrate Dehydrogenase  lane 6 and 100 pg, lane 7). Decreased amounts of cellular protein extracts were loaded in lane 6 because a t higher protein levels (lane 7 ) we experienced saturation of the film for the IDHl signal. Overexpression of IDHl and IDH2 was achieved using the vector YEp352-(IDHI fIDH2) in the AIDHI fIDH2 strain and resulted in increased levels of both subunits (20 pg, lane 8 and 100 pg, lane 9). Lanes [8][9][10] were electrophoresed on a separate gel and aligned using molecular size markers. dence that NAD'-IDH from yeast is composed solely of the two subunits, IDHl and IDH2.
Activity of NAD'-ZDH in Mutant Strains-The contribution of IDHl to the conversion of isocitrate to a-ketoglutarate was determined by assaying NAD'-IDH activity in mitochondrial extracts from strains expressing varying levels of IDHl (Table I). The AZDHl strain, which expresses only IDH2, does not contain detectable levels of NAD'-IDH activity, suggesting that IDH2 alone cannot catalyze the conversion of isocitrate to a-ketoglutarate. Similarly, extracts from the AZDHlIIDH2 strain do not contain NAD'-IDH activity. The absence of activity in these disruption strains is not due to an inhibitory factor in the mitochondrial extracts, because assays conducted with mixtures of wild type and disruption extracts exhibit full activity (Table I).
T o further analyze IDHl expression and activity, single copy and multicopy expression of IDHl was examined in the AZDHl strain. Expression of IDHl using the single copy vector YCp50 resulted in wild type levels of IDHl as shown by Western blot analysis (Fig. 3, lane 5), and restored NAD'-IDH activity to wild type levels ( Table I). Overexpression of IDH1, on the other hand, using the multicopy vector YEp352 resulted in increased levels of IDHl (Fig. 3, lanes 6 and 7) but did not result in elevated NAD+-IDH activity (Table I). These data suggest that IDHl alone is not catalytically active. Similar results were previously obtained for IDH2 suggesting that both subunits are required for a functional enzyme.
Eukaryotes contain an additional mitochondrial isocitrate dehydrogenase isozyme which utilizes NAD(P)' as a cofactor (23, 24). To examine possible effects on activity of the mitochondrial NAD(P)'-IDH isozyme in strains lacking a function NAD'-IDH, mitochondrial extracts from the NAD'-IDH disruption strains were also assayed for NAD(P)+-IDH activity. The three disruption strains, AZDHl, AZDH2, and AIDHl/ ZDH2, exhibited slightly reduced levels of mitochondrial NAD(P)+-IDH activity relative to wild type (Table I). This suggests that activity of the mitochondrial NAD(P)' isozyme is not increased to compensate for the absence of a functional Growth Phenotypes of NAD+-ZDH Mutants-To determine if IDHl or IDH2 retains function in vivo and to confirm the metabolic function of NAD'-IDH, the growth phenotypes of the NAD'-IDH mutant strains were analyzed. We have previously shown that lack of a functional NAD'-IDH in the AZDH2 strain results in decreased growth rates on glycerol or lactate and failure to grow on acetate (IO), and therefore also characterized the growth phenotypes of the AZDHl and AZDHlIZDH2 strains using these nonfermentable carbon sources. As shown in Table  11, all NAD'-IDH disruption strains grew a t a reduced rate relative to wild type on glycerol (-1.7-fold) and lactate (-2.7-fold). It is unclear if the small difference in growth rate of the single disruptions (-8 h) and the double disruption (10 h) with lactate as a carbon source is significant. All NAD+-IDH disruption strains failed to grow brackets. on acetate, doubling no more than once within a 48-h period following dilution from a minimal medium (YNB) starter culture into rich medium (YP) with acetate as a carbon source. An acetate-phenotype was previously reported for the AIDHZ strain (10) and has also been described for yeast strains containing defects in other citric acid cycle enzymes, including mitochondrial malate dehydrogenase (25) and citrate synthase (26). These results indicate that the NAD'-IDH isozyme is the isozyme of isocitrate dehydrogenase which functions in the citric acid cycle. Furthermore, the AIDHI and AIDHZ strains displayed similar growth rates indicating that disruption of IDHl or IDHZ is equivalent in inactivating the NAD'-IDH enzyme. In addition, the single disruption and double disruption strains grew at nearly the same rate on nonfermentable carbon sources, suggesting that IDHl and IDH2 individually do not retain significant activity in uiuo.

NAD+-dependent Isocitrate Dehydrogenase
To examine the role of NAD'-IDH in the production of glutamate, growth rates of the disruption strains were characterized using minimal medium plus and minus glutamate. The AIDHl, AIDHZ, and AIDHIIIDHZ strains were found to grow with rates similar to wild type under both conditions (Table 11), indicating that the NAD'-IDH disruption strains were not glutamate auxotrophs. Two models can explain this result. The a-ketoglutarate produced by NAD'-IDH may not contribute significantly to the biosynthesis of glutamate; on the other hand, the NAD(P)+-IDH isozyme may be compensating for the lack of a functional NAD'-IDH in the glutamate biosynthetic pathway.
Overexpression of NAD'-IDH-The conversion of isocitrate to a-ketoglutarate is a highly regulated step of the citric acid cycle. To determine if protein levels of NAD+-IDH are limiting for growth on nonfermentable carbon sources, the growth phenotype of a yeast strain containing increased levels of NAD+-IDH was characterized. Overexpression of NAD+-IDH using the multicopy vector YEp352 containing IDHl and IDHZ genes in tandem resulted in increased protein expression of both subunits (Fig. 3, lunes 8 and 9). In addition, mitochondria isolated from this strain displayed a 3-fold increase in NAD'-IDH activity (Table I). Growth rate analysis of the strain which contained increased levels of NAD+-IDH, however, indicated that this strain grew at wild type levels on glycerol, lactate, or acetate (Table 11). Because no decrease in doubling time was observed we interpret these findings to indicate that NAD'-IDH protein levels are not rate-limiting for growth on these nonfermentable carbon sources.

DISCUSSION
Metabolic Role of NAD+-IDH-Because of the importance of the conversion of isocitrate to a-ketoglutarate in central metabolic pathways, much effort has been invested in understanding the kinetics and regulation of NAD+-IDH. TO further characterize this complex allosteric enzyme our work has focused on determination of the structure and function of the two individual subunits of the enzyme from S. cerevisiae.
Previously we reported the cloning and nucleotide sequence of the gene encoding the IDH2 subunit and examined how disruption of the IDH2 gene affected NAD'-IDH activity (10). In this paper we present the cloning and nucleotide sequence of the gene encoding the IDHl subunit of NAD'-IDH and results obtained following construction of a double disruption strain, AIDHlIIDHZ, which provides the best background for analysis of the metabolic function of NAD'-IDH.
The presumed role of NAD'-IDH in the citric acid cycle has been based on the complex allosteric regulation of the purified enzyme by adenylate nucleotides (6) and substrates of the citric acid cycle (27). Our results with strains lacking IDHl or both IDHl and IDH2 are consistent with the data previously obtained with strains lacking IDH2, which indicated that NAD'-IDH does have an essential function in the citric acid cycle of eukaryotic cells (10). These strains exhibited decreased growth rates on all nonfermentable carbon sources analyzed and displayed the same acetate-growth phenotype reported for yeast strains with defects in the mitochondrial isozymes of malate dehydrogenase (25) and citrate synthase (26). These results indicate that the normal route of mitochondrial respiration requires delivery of reducing equivalents from the citric acid cycle by NAD'-IDH to the electron transport chain. Furthermore, these results suggest that the mitochondrial NAD(P)+-dependent isozyme does not compensate for the loss of a functional NAD'-dependent isozyme. The inability of mitochondrial NAD(P)'-IDH to compensate for this respiratory function may be due to the absence of transhydrogenase activity in yeast. On the other hand, it has been suggested that NAD+-IDH may specifically interact with other mitochondrial proteins, including citrate synthase and a-ketoglutarate dehydrogenase, to facilitate shuttling of intermediates of the citric acid cycle (28). The mitochondrial NAD(P)'-IDH isozyme may not participate in protein-protein interactions of this type and, therefore, not allow optimal citric acid cycle function. The NAD'-IDH disruption strains are not glutamate auxotrophs, however, suggesting that the NAD(P)'-IDH isozymes, either mitochondrial or cytoplasmic, may be compensating for the lack of a functional NAD+-IDH in the glutamate biosynthetic pathway. Taken together these results indicate that the NAD'-IDH isozyme functions in the citric acid cycle and suggest that the NAD' and NAD(P)+ isozymes function in separate metabolic pathways.
Activities of IDHl and IDH2"Based on earlier ligand binding studies conducted with yeast NAD'-IDH which indicated four isocitrate-and two AMP-and NAD'-binding sites per octamer (291, it has been suggested that the individual subunits of yeast NAD'-IDH may have separate functions in catalysis and regulation. However, our findings with the AIDHl and AIDHZ strains do not support a model of inde-pendent catalytic function. Disruption of either IDHl or IDHZ resulted in a loss of measurable NAD+-IDH activity in mitochondrial extracts, and overexpression of the IDHl subunit in a wild type IDHB background restored only wild type levels of activity. Similar results were observed previously for overexpression of IDHZ (10). These findings imply that both IDHl and IDHZ are required for catalysis. Furthermore, growth phenotype analysis of the single disruption and double disruption strains indicate that they are equally impaired for growth on the nonfermentable carbon sources of glycerol or lactate and all fail to grow with acetate as a carbon source. These results indicate that lack of IDHl or IDHB is equally effective in disrupting NAD+-IDH activity and suggests that the individual subunits do not retain significant activity in uiuo. Further examination of the function of each subunit will require structural studies of the active octamer and individual subunits and alteration of IDHl and IDHB by site-specific mutagenesis.
Amino Acid Sequence Comparisons-Cloning and nucleotide sequence analysis of IDHl allows comparison of the amino acid sequences of IDHl and IDH2. The alignment of IDHl and IDHZ shown in Fig. 4 yields an overall sequence identity of 42% between the two subunits. The overall sizes of the IDHl and IDHZ polypeptides are similar and, although IDHB contains 5 more residues than IDH1, the predicted molecular weight of IDHl (38,001) is greater than that of IDH2 (37,755).
dome predictions about the possible function of the IDHl and IDHS subunits and the role of individual residues can be made from amino acid sequence comparisons with the isocitrate dehydrogenase enzyme from E. coli. The bacterial enzyme has been extensively studied and structural analysis by x-ray crystallography has been performed (30). In contrast to the  yeast NAD+-IDH, which is an octamer of nonidentical subunits, NAD(P)+-IDH from E. coli exists as a dimer of identical subunits (1). Furthermore, the E. coli enzyme is regulated solely by phosphorylation (5), whereas the yeast enzyme is allosterically regulated. Despite these differences, E. coli NAD(P)+-IDH and IDH2 were previously shown to have an overall sequence identity of 33% (10). An amino acid sequence comparison of IDHl and IDH2 with E. coli NAD(P)+-IDH is shown in Fig. 4. Alignment of IDHl with the E. coli enzyme in this manner results in a 31% sequence identity. Interestingly, the yeast NAD+-IDH enzyme shows greater similarity to the E. coli NAD(P)+-IDH enzyme than to the NAD(P)+-IDH enzymes from yeast (24) or pig (31), which exhibit less than 20% sequence identity to either IDHl or IDH2. The x-ray crystal structure of E. coli NAD(P)+-IDH has been solved in the presence of the substrate, isocitrate (32), and cofactor, NAD(P)+ (33), and in its inactive phosphorylated form (34). These studies have established that residues 110-120 of E. coli IDH function in isocitrate binding with Ser-113 forming a hydrogen bond with isocitrate; the phosphorylation which inactivates the enzyme occurs at this serine and inhibits isocitrate binding. Ser-92 of IDHl and Ser-98 of IDHB align with this residue, and the region surrounding these serine residues is highly conserved in all three polypeptides. This high degree of similarity suggests that both IDHl and IDHB subunits could contain isocitrate-binding sites. The NAD(P)+-IDH from E. coli contains two isocitrate-binding sites per dimer in which the subunits contain half-binding sites and both contribute to formation of the active sites. The bacterial model of half-binding sites would predict that the yeast enzyme contains eight binding sites created by the interaction of IDHl and IDH2 to form each complete isocitrate binding pocket. This model, however, is in disagreement with earlier binding studies which indicated four isocitrate binding sites per octamer, suggesting that the isocitrate-binding site of IDHl or IDH2 may be nonfunctional. NAD(P)+-IDH from E. coli binds the nicotinamide cofactor in a cleft between the large and small domains of the enzyme and involves a nucleotide binding fold not previously described for other proteins (33). The residues participating in binding are primarily clustered near the C terminus (Ile-37, Ile-320, Gly-321, His-339, Val-351, Asn-352, Tyr-391, and Asp-392) and are boxed in Fig. 4. Comparison of the sequences of IDHl and IDHZ with E. coli NAD(P)+-IDH in this region shows little conservation of the residues. IDHl contains 2 residues (Gly-255 and Asn-287 of IDH1) and IDHZ contains 4 residues (Ile-32, Gly-263, His-281, and Asn-293 of IDH2) identical to those postulated to be involved in binding the adenine moiety of the nicotinamide cofactor, suggesting that NAD+-IDH from yeast may have a different nucleotide binding fold.
Partial amino acid sequences have also been reported for pig heart NAD+-IDH. Pig NAD+-IDH, like the yeast enzyme, has been shown to be composed of nonidentical subunits of -40,000 Da, but is a tetramer having an a2:p:y subunit structure (35). Several cysteine-containing tryptic peptides from each of the three subunits have been isolated and sequenced (36). We were previously able to align the sequences of the pig NAD+-IDH peptides with IDHZ to give a 43% sequence identity (10). Similar alignment of these peptide sequences with IDHl (Fig. 5) gives identities in 116 positions of the 269 positions aligned, again corresponding to 43% sequence identity. One region of sequence similarity of particular interest is the region encompassing IDHl residues 189-194. Within this sequence Asp-191 of IDHl aligns with an aspartate residue previously found to be chemically labeled by reaction of pig NAD+-IDH with an analogue of the allosteric regulator ADP (37). A similar alignment can be made for IDH2 (10). These results, in addition to the sequence similarities with the E. coli enzyme discussed above, suggests the possibility that both the IDHl and IDHZ subunits may have regulatory as well as catalytic functions.