Regulation of Diphosphopyridine Nucleotide-Pinked Isocitrate Dehydrogenase from Bovine Heart BIKDING OF INHIBITORY NUCLEOTIDES”

Abstract DPN-linked isocitrate dehydrogenase from bovine heart binds 1 molecule of DPNH per molecule of enzyme of 320,000 daltons with a concomitant 20-fold increase in nucleotide fluorescence. The fluorescence properties of this tightly bound DPNH (Kd = 1.4 µm) are not affected by DPN+, isocitrate, MgSO4, or ADP, either alone or in combinations. However, ATP, an inhibitor of isocitrate dehydrogenase, competitively displaces bound DPNH. Ultrafiltration or gel chromatography of the enzyme in the presence of DPNH reveals a total binding of 3.5 to 4.5 eq of nucleotide with a limiting dissociation constant of 14 µm obtained at near saturating ligand concentrations. Fluorescence titration indicates that isocitrate dehydrogenase also binds 1 eq of TPNH with a Kd of 0.9 µm. In contrast to the enzyme-DPNH complex, bound TPNH is not displaced by ATP and shows no evidence of energy transfer in its fluorescence excitation spectrum. Ultrafiltration experiments reveal at least two additional binding sites for TPNH. The presence of DPNH does not affect the binding of TPNH, and DPNH binding is not altered by TPNH, indicating separate sites on the enzyme for the two types of pyridine nucleotides.

SUMMARY DPN-linked isocitrate dehydrogenase from bovine heart binds 1 molecule of DPNH per molecule of enzyme of 320,000 daltons with a concomitant 20-fold increase in nucleotide fluorescence.
The fluorescence properties of this tightly bound DPNH (KD = 1.4 pM) are not affected by DPN+, isocitrate, MgSO1, or ADP, either alone or in combinations. However, ATP, an inhibitor of isocitrate dehydrogenase, competitively displaces bound DPNH. Ultrafiltration or gel chromatography of the enzyme in the presence of DPNH reveals a total binding of 3.5 to 4.5 eq of nucleotide with a limiting dissociation constant of 14 KM obtained at near saturating ligand concentrations.
Fluorescence titration indicates that isocitrate dehydrogenase also binds 1 eq of TPNH with a KD of 0.9 PM. In contrast to the enzyme-DPNH complex, bound TPNH is not displaced by ATP and shows no evidence of energy transfer in its fluorescence excitation spectrum. Ultrafiltration experiments reveal at least two additional binding sites for TPNH.
The presence of DPNH does not affect the binding of TPNH, and DPNH binding is not altered by TPNH, indicating separate sites on the enzyme for the two types of pyridine nucleotides.
The catalytic activity of DPN-linked is0citrat.e dehydrogenase of bovine heart (three-n,-isocitrate + DPN+ + a-ketoglutarate + COZ + DPNH + H+; EC 1.1. 1.41) is enhanced by ADP. The regulatory implications of adenine nucleotides at the level of isocitrate oxidation were recognized by several investigators in studies with purified enzymes from animal t,issues (2)(3)(4) and microorganisms (5,6). In the latter, the importance of the level of the energy charge of the adenine nucleotide pool has received substantial support by the investigations of Atkinson and associates (for review see Ref. 7). However, the DPN-linked isocitrate dehydrogenases from animal tissues and yeast differ in the severe inhibition of the former (2,4,8) by reduced pyridine nucleotides.
Recent studies of kinetic parameters of purified enzyme (4) and observations with more integrated systems from liver (9,10) suggest that the ratio of DPNH:DPNf may in fluence the in vivo activity of the enzyme more than the level of the positive modifier ilDP.
For example, Williamson et al. (9) found that ethanol slowed the activity of the tricarboxylic acid cycle in perfused liver from starved rats. It was suggested that the increase in reduced pyridine nucleotides observed in the whole organ also occurred within mitochondria causing a shift in the malate to oxaloacet'ate equilibrium and an inhibition by reduced pyridine nucleotides of DPN-specific isocitrate dehydrogenase.
Konig, Nicholls, and Garland (10) have shown that the oxidation of succinate and palmitoyl carnitine in intact rat liver mitochondria resulted in an inhibition of isocitrate oxidation and a concomitant increase of the intramitochondrial DPNH:DPN+ ratio. These observations, while not conclusive, do suggest the importance of reduced pyridine nucleotides in the in vivo regulation of DPN-linked isocitrate dehydrogenase. The effect of the intramitochondrial oxidation-reduction potential on isocitrate oxidation may be mediated by bot,h the DPNH : DPNf ratio and the level of TPNH.
Kinetic experiments with purified enzyme preparations have shown that whereas there is a competitive relationship between DPN+ and DPNH, TPNH (which is neither a substrate nor an inhibitor of DPN-linked isocitrate dehydrogenase) significantly potentiates DPNH irhibition (2,4,8).
DPNH and TPNH thus appear to be important in the regulation of cellular oxidation of isocitrate.
The present investigations with techniques involving fluorescence measurements, ultrafiltration, and gel chromatography provide more direct evidence for the nature and extent of binding of these ligands to the DPN-linked isocitrate dehydrogenase protein.
Two separate sets of binding sites have been determined in t,hese esl)rriments, one specific for DPNH and the other specific for TI'NH. Preliminary reports have appeared on the binding of reduced pyridine nucleotides and of DPNf to purified enzyme prellarations from bovine heart (1) and yeast (1 l), respectively. The following chemicals were purchased from the commercial  sources indicated.  ADP, ATP, DPNf, DPNH, TPN+, TPNH,  EUTA, 'Tris, glucose B-phosphate dehydrogenase, rabbit muscle  lactate  The reduction, monitored at 340 nm, occurred instantaneously.

Materials
The enzyme was denatured by placing the vessel containing the reaction mixture in boiling water for 4 min. The reaction mixture was diluted into 19 ml of 0.01 M NaHC03 and placed on a DEAE-cellulose column (0.9 x 10 cm) previously equilibrated with 0.005 M sodium phosphate buffer at pH 7.2. The column was washed with 0.005 1~ sodium phosphate buffer at pH 7.2, followed by 0.05 M NaCl in the above buffer until in each case the absorbance of the effluent at 340 nm and 260 nm was less than 0.005. The reduced nucleotide was then eluted with 0.25 M NaCl-0.005 M sodium phosphate at pH 7.2. Fractions were combined which had ratios of absorbance at 260 nm: 340 nm of less than 2.6. The combined solutions containing 0.60 to 0.70 pmole of [3H]DPNH usually exhibited a 260 nm:340 nm absorbance ratio of 2.4 to 2.5. This ratio remained constant for 3 to 4 days at 0"; the solution was used in the experiments during this period.
The absorbance at 340 nm was followed until there was no further change. The reaction mixture was then adjusted to pH 10.5, placed in a boiling water bath for 3 min, diluted with 17 ml of 0.01 M NaHC03, and chromatographed on a column of DEAE-cellulose as described above. The product had a 260 nm:340 nm absorbance ratio of 2.5.
The molar concentration of isocitrate dehydrogenase was calculated from the protein content as determined by the biuret method (14)) and a molecular weight of 320,000 (13). Enzyme preparations used in these studies had a specific activity' of 24,000 to 27,000 nmoles of DPN+ reduced per mg of protein per min at 25", migrated as a single band on disc electrophoresis, and contained less then 0.5% carbohydrate as determined by the phenol-sulfuric acid method (16). Fluorescence l!easurements-Fluorescence emission, excitation, and polarization measurements were performed with a recording spectrofluorimeter constructed in this laboratory. The excitation wave length was selected by a monochromator (0.25 meter, Jarrell Ash) from the emission of a xenon lamp (PEK 75 watt powered by a Hewlett Packard 6274 A supply). A quartz lens system (Bausch and Lomb) focused the excitation light on a standard l-cm qua,rtz fluorescence cuvette containing 1 ml of sample.
After isolation by a monochromator or interference filter, the fluorescence emission was monitored with an EM1 photomultiplier (6256SA, EM1 Electronics) and an operational amplifier circuit described previously (17). A voltage proportional to the emission intensity was displayed on a recorder (Bristol) or a digital voltmeter (Heath Instrument Co.). A nominal band width of 5 nm for excitation and emission was used in all experiments.
Fluorescence polarization measurements employed polarizing filters in the excitation and emission beams as outlined by Chen and Bowman (18). Spectra reported are not corrected for lamp, monochromator, and photomultiplier response. The t'emperature of the solution during the fluorescence measurement was held at 25 f 0.1" by a Tamson refrigerated circulating water bath. Absorption spectra were determined in a Cary model 14 spectrophotometer fitted with a thermostated cuvette compartment. Determination of Binding Capacity and Afinity of Enzyme by Ultrafiltration-An ultrafiltration technique described by Paulus (19) was used to measure the binding of radioactive nucleotides to isocitrate dehydrogenase.
A solution (0.5 ml) containing enzyme (44 pg), 0.17 M NaCl, and labeled DPNH or TPNH was filtered through a membrane (UM-10, Amicon Corp.) in an ultrafiltration cell (Metalloglas) at 22-23". The enzyme and bound ligand were retained by the filter while free ligand passed through.
The membrane was removed from the apparatus, placed in a scintillation vial containing 0.2 ml of water, and agitated for approximately 10 min. A solution (10 ml) containing 0.3 g of p-bis[2-(5.phenyloxazolyl)]benzene, 5 g of 2,5-diphenyloxazole, and 50 ml of Biosolv BBS-2 (Beckman Instruments, Inc.) per liter of toluene was added, and the samples were counted in a liquid scintillation spectrometer (Nuclear-Chicago).
The radioactivity retained on the membrane allows a direct calculation of the amount of ligand bound.
The dissociation constant and the number of binding sites were obtained by the Scatchard plot (20).
K rir r=n-G 1 The specific activity of the enzyme used here is essentially the same as that of preparations used previously.
The values reported then (13) were based on protein determined by the method of Warburg and Christian (15) which in the case of this enzyme is about one-half of that found by the biuret method (13, 14). The latter has been adopted as the basis of the protein values in the present report. where r is the moles of bound ligand per mole of enzyme, (A) is the concentration of free ligand, ICD is the dissociation constant for the ligand, and n is the number of sites. The plot of r against r/A is linear if the sites are identical and independent. The slope is -K, and the ordinate intercept is n.
Gel 10 cm) and proportionally reduced amounts of enzyme also gave satisfactory results.
The absorbance at 340 nm, enzyme activity, and protein concentration of the effluent fractions were determined.
The absorbance at 340 nm showed a peak containing the enzyme-DPNH complex followed by a trough which corresponded to the amount of DPNH bound by the isocitrate dehydrogenase.
The difference between the concentration of DPNH in the enzyme-containing fractions and that contained in the equilibrating buffer corresponds to the concentration of bound DPNH.
It The fluorescence emission of reduced pyridine nucleotides was determined in the presence and absence of enzyme at ligand concentration ranging from 0.5 pM to 30 PM. The fluorescent enhancement was used to calculate the amount of enzyme-bound ligand after making corrections for volume changes, internal absorption (if applicable), and protein emission. The slope corresponds to K. while the ordinate intercept is n.  sites. The fluorescence emission resulting from titration of a constant amount of isocitrate dehydrogenase with successive additions of DPNH is shown in Fig. 2A. The initial additions of DPNH show a strong enhancement of fluorescence in accord with the results shown in Fig. 1. As saturation of the nucleotidebinding sites is approached, the fluorescence response becomes characteristic of the free nucleotide. Analysis by the method of Stockell (23) indicated that the data could be described satisfactorily in terms of a single binding site (n = 0.9) with a dissociation constant (K,) of approximately 1 pM (Fig. 3). The binding of DPNH which can be demonstrated fluorimetrically may simply be a manifestation of the reaction product binding at the catalytic site. If isocitrate dehydrogenase, like many other dehydrogenases (25)(26)(27)(28)(29)(30), follows a compulsory mechanism in which DPN+ is the first substrate bound and DPNH is the last product to be released, bound DPNH should be displaced by DPN+.
This prediction was not supported experimentally; fluorimetric titrations with DPNH in the presence of 300 PM DPNf yielded values of n and K. which were identical with those obtained in the absence of the oxidized coenzyme, despite the fact that Ki for DPKH and K, for DPN+ are of the same order of magnitude (2). Similar experiments have shown that isocitrate (2.0 InM), MnS04 (0.5 mM), or EDTA (0.1 mM) were without effect on the fluorescence of bound DPNH. Identical results were obtained with binary combinations of these substances (e.g. DPN+ plus isocitrate in the presence of 1 mM EDTA, isocitrate plus MnS04, or DPN+ plus MnS04) either in the presence or absence of ADP (0.66 mM). The DPNH binding detected by fluorescence thus appears to be unaffected by the presence of substrate or activators either alone or in combinations.
These results may indicate that DPNH is binding at' a site which is separate from the catalytic site.
Since the results consistently showed equimolecular binding of DPNH to isocitrate dehydrogenase, it was desirable to determine whether the conditions selected for fluorescence titration would detect more than one binding site for reduced pyridine nucleotides per molecule of protein.
The validity of the method was confirmed by the demonstration of multiple binding sites in other pyridinoproteins.
For example, the procedure indicated the binding of 3.8 to 4.1 molecules of DPKH per molecule of rab-bit muscle or bovine heart lactate dehydrogenase, values in excellent agreement with those reported in the literature (31).

Displacement of Bound DPNH-While the substrates of iso-
citrate dehydrogenase failed to displace bound DPNH, fluorescence titrations in the presence of ATP revealed that this inhibitor competitively displaced the reduced cofactor (Fig. 4). The dissociation constant for the binding of ATP was calculated to be about 0.1 mM from the following equation which expresses competition by two ligands for identical and independent binding sites on a macromolecule (32). The acetylpyridine derivative of DPNH showed no enhanced fluorescence emission or polarization in the presence of isocitrate dehydrogenase under conditions where a spectrally distinct complex was formed with DPNH.
Acetylpyridine DPNH was ineffective in displacing bound DPNH when the analog was present at 10 times the concentration of DPNH. However, when the ratio of acetylpyridine DPNH to DPNH was 100 to 500, measurable displacement of bound DPNH was observed.
These results are consistent with kinetic data3 indicating that the reduced acetylpyridine analog is only weakly inhibitory to isocitrate dehydrogenase.
Ultrajiltration and Gel Chromatography-While the existence of one binding site for DPNH has been established by fluorescence measurements, the possibility remained that additional DPNH-binding sites may exist which do not lead to an alteration of the fluorescence properties of the bound ligand. Such binding sites should be detectable by procedures which do not rely on a fluorescence end point.
Equilibrium dialysis, frequently used for such determinations, was impractical since this enzyme was denatured extensively during the equilibration period.
Chromatography on columns of Sephadex G-50, equilibrated with buffers containing DPNH, led to minimal inactivation of the enzyme and showed binding of more than 1 molecule of nucleotide.
The results of such experiments as shown in Fig.  5 have been plotted according to the method of Scatchard (20) and indicate, upon extrapolation to infinite concentration of ligand, that 4 molecules of DPNH are bound per molecule of the enzyme. These data are not sufficiently precise to allow firm conclusions about the identity or nonidentity of the binding sites corresponding to r 2 2.5. However, from the slope of the line shown in Fig. 5 a limiting value of Ko of 14 pM can be calculated which is significantly different from a KD of 1.3 PM determined by fluorimetry (Fig. 3) (31) reported that if the activity coefficient of an enzyme is independent of protein concentration, dilution of a solution of the enzyme-ligand will produce no change in n. Under such conditions a linear relationship should obtain bet,ween nucleotide fluorescence and total protein concentration. 4  The fluorescence was measured after each dilution.
All other conditions were as described under "Methods." Fig. 6 shows that this is the case, i.e. DPNH binding determined by fluorescence enhancement is independent of enzyme concentration over a range which includes the low and high protein concentrations used in the fluorescence method and in gel chromatography, respectively.
It can be calculated from the K. of the enzyme monomer-dimer equilibrium (13) that the percentage of monomer varies from 60% to 20% over the range of protein concentrations reported in Acetylpyridine TPNH may not be bound significantly to the enzyme since it failed to exhibit enhanced fluorescence emission or polarization in the presence of isocitrate dehydrogenase and, in contrast to TPNH, did not potentiate the inhibition by DPNH of enzyme activity.3 Binding of TPNH by Isocitrate Dehydrogenase Independence of DPNHand TPA-II-binding Sites Fluorimetry-TPNH is not an inhibitor of the DPN-specific isocitrate dehydrogenase; however, TPNH potentiates the inhibition by DPNH of the enzyme from heart (2) and liver (4) and presumably interacts with the enzyme.
Such an interaction is supported by the observation that the fluorescence emission of TPNH was markedlyenhanced in the presence of isocitrate dehydrogenase (Fig. 1B). In contrast to DPNH, bound TPNH did not show an excitation peak at 285 nm (Fig. lA), i.e. there is no energy transfer through the protein.
Fluorimetric titrations, similar to those described for DPNH, indicated a single binding site for TPNH with a KD of 0.89 pM (Fig. 7). The values of KD and n were the same in the absence or presence of the positive modifier ADP (0.66 mM) or of the inhibitor ATP (1 InM).
UltrajZtration-At least two more binding sites, in addition to t'he one shown fluorimetrically, could be detected by the ultrafiltration technique (Fig. 8). The scatter of the data shown in Fig. 8 makes uncertain the precise value of n; it is possible that, in analogy to DPNH, 4 molecules of TPNH are bound per molecule of protein.
The binding of [W] TPNH (25 PM) to isocitrate dehydrogenase was not altered by the presence of DPNH The interaction of DPNH and of TPNH with isocitrate dehydrogenase is similar in a number of respects.
Thus, while these ligands are bound to multiple sites on the enzyme, fluorescence enhancement corresponding to only one binding site is observed with each of the reduced pyridine nucleotides. Nevertheless, binding of DPKH and TPNH must occur at different sites on the enzyme. 4s shown in Fig. 9, initial partial saturation of the protein with DPNH does not prevent further enhancement of fluorescence by subsequently added TPNH when read at an excitation wave length of 340 nm (Fig. 9A). Furthermore, TPNH does not displace DPNH from the enzyme since the energy transfer from the protein (Fig. 112) which is displayed at the excitation wave length of 290 nm in the presence of DPNH, but not by TPNH, is not diminished in intensity by subsequent addition of TPNH to the DPNH-enzyme complex (Fig. 9B).
Although not shown here, it has also been observed, that when the sequence of addition is reversed from that reported in Fig. 9 However, binding of these molecules by isocitrate dehydrogenase might lead to changes in protein fluorescence.
The effect of various ligands on protein fluorescence is shown in Fig. 10. Isocitrate up to a final concentration of 2.7 mM had no effect. DPN+ lowered fluorescence, but this decrease is almost entirely due to the absorption of the exciting light by DPN+ (inner filter effect).
The decrease in protein fluorescence by DPNH and TPNH shown in Fig. 10 is in excess of that predicted from absorption of the exciting incident light energy. If one assumes that the decrease in protein fluorescence is proportional to the binding of these nucleotides, values of n and Ko can be calculated.

DISCUSSION
Binding of DPNH to isocitrate dehydrogenase leads to a shift in the excitation and emission maxima of the coenzyme and to an increase in fluorescence intensity.
While qualitatively similar changes in spectra have been reported for a number of dehydrogenases, isocitrate dehydrogenase is unique in that the DPNH fluorescence increases lo-to 20.fold on binding to the enzyme. This large increase in quantum yield is accompanied by a pronounced energy transfer from the excited aromatic residues of the protein to the bound DPNH.
A detailed discussion is not yet feasible in the absence of information about the lifetimes of the excited states. However, the fluorimetric data suggest that bound DPNH is shielded from radiationless thermal deactivation in the excited state and that, at least at this site, spatially favored orientations exist between DPNH and one or more aromatic amino acids in the isocitrate dehydrogenase protein.
The binding of DPNH to isocitrate dehydrogenase is also unusual in that only one of the four potential binding sites alters the fluorescence properties of the bound DPNH.
It is not clear whether the DPNH-binding site which can be demonstrated fluorimetrically is functionally different from the three additional binding sites revealed by gel chromatography or ultrafiltration. However, several observations suggest that the DPNH-enzyme interaction accompanied by altered fluorescence does not occur at the catalytically active site. Thus, though kinetic experiments show a competitive relationship between DPNf and DPNH with values of K, and Ki of the same order of magnitude (2), DPNH is not displaced readily by DPN+ from the fluorimetrically detectable site. It may be that this specific site, where binding of the inhibitor ATP also occurs (ATP can displace bound DPNH competitively (Fig. 4)), has a regulatory rather than a catalytic function.
The experimental conditions of binding and kinetic measurements are sufficiently different to make such a conclusion tentative.
Nevertheless, the binding of DPNH to such a regulatory site (instead of product binding at the catalytic site) may have a bearing on the apparently aberrant behavior of DPNH with purified liver DPN-linked isocitrate dehydrogenase.
It was observed in these experiments (4) that, while the apparent K, of DPN+ is changed by varying concentrations of isocitrate and by the positive modifier ADP, the Ki of DPNH is essentially uninfluenced by these substances.
This observation is not consistent with a mechanism in which DPN+ and DPNH are bound to the same site on the enzyme. It is noteworthy that direct interaction of ATP with the enzyme has been demonstrated in the absence of added metal ions (Fig. 4), since there has been some uncertainty whether inhibition by ATP is attributable only to chelation of Mg2+ (33) or to an additional more specific effect of the nucleotide on the enzyme (2).
The binding of TPNH while superficially similar to that of DPNH (e.g. binding at multiple sites of which only one alters the fluorescence of bound reduced pyridine nucleotide) occurs at distinctly different sites (e.g. the failure of TPNH to exhibit energy transfer through the protein (Fig. IA) and to be displaceable by ATP (cf. Fig. 4) or DPNH (Fig. 9)). The previous observations that TPNH potentiates the inhibitory action of DPNH (2,4,8) suggest that there should be an interaction between the TPNH-binding site(s) and at least one DPNH-binding site. However, the presence of TPNH did not markedly alter the stoichiometry or dissociation constant(s) of DPNH, and DPNH appeared to be without effect on TPNH binding either fluorimetrically or by the ultrafiltration technique. It may be that the procedures used in the present studies, especially the ultrafiltration and gel filtration techniques, are not sufficiently precise to detect subtle interactions between the two types of coenzyme sites or that incubation conditions permitting catalytic activity will be necessary to reveal changes in binding characteristics at these sites.
The enzymically active species of bovine heart DPN-linked isocitrate dehydrogenase of molecular weight 330,000 contains eight polypeptide chains which appear equivalent in molecular weight (13), whereas it has been found here that 4 molecules of DPNH and 3 to 4 molecules of TPNH are bound per molecule of