The reaction of 4-iodoacetamidosalicylic acid with TPN-dependent isocitrate dehydrogenase from pig heart.

Salicylic acid acts as a reversible inhibitor of pig heart TPN-dependent isocitrate dehydrogenase competitive with manganous isocitrate (& = 18.2 m ~ ) but not with TPN. The salicylate derivative, 4-iodoacetamidosalicylic acid (0.5 to 4 m ~ ) , causes an irreversible time-dependent loss of both the isocitrate dehydrogenase and oxalosuccinate decarboxylase activities of the enzyme when incubated at pH 7.0 and 30 "C. The inactivation rate constant is linearly dependent on the 4iodoacetamidosalicylic acid (ISA) concentration, yielding a second order rate constant of 10.7 min" M-', a value 11 times greater than that for iodoacetamide under the same conditions; this result suggests that the salicylate moiety contributes to the effectiveness of ISA in inactivating this enzyme. Total protection against inactivation is provided by manganous ion and isocitrate, but not by either metal ion or isocitrate alone. When the incubation is conducted in the absence of ligands, total inactivation can be related to the incorporation of 4 mol of ['4C]ISA/mol of enzyme; however, when the reaction is carried out in the presence of Mn" and isocitrate, the enzyme retains 100% activity, but still incorporates 2 mol of radioactive ISA/mol of enzyme. Carboxymethyl derivatives of lysine and cysteine are observed in acid hydrolysates derived from inactive enzyme prepared in the absence of ligands; whereas only carboxymethylcysteine is detected in samples derived from active modified enzyme prepared in the presence of Mn" and isocitrate. Modification of a cysteine or partial modification of a cysteine and lysine by LSA appears to be responsible for inactivation. Enzyme reacted first with unlabeled ISA in the presence of protecting substrate plus metal ion, and then with radioactive H A in the absence of ligands yields only radioactive e-dicarboxymethyllysine from the hydrolysate of the inactive enzyme. It is concluded that in this case a lysine residue attacked by 4-iodoacetamidosalicylic acid is critical for the function of isocitrate dehydrogenase and is located within the metal-isocitrate binding site. A new method is also presented for the purification of TPN-dependent isocitrate dehydrogenase using fresh pig hearts as the starting material.

Salicylic acid acts as a reversible inhibitor of pig heart TPN-dependent isocitrate dehydrogenase competitive with manganous isocitrate (& = 18.2 m~) but not with TPN. The salicylate derivative, 4-iodoacetamidosalicylic acid (0.5 to 4 m~) , causes an irreversible time-dependent loss of both the isocitrate dehydrogenase and oxalosuccinate decarboxylase activities of the enzyme when incubated at pH 7.0 and 30 "C. The inactivation rate constant is linearly dependent on the 4iodoacetamidosalicylic acid (ISA) concentration, yielding a second order rate constant of 10.7 min" M-', a value 11 times greater than that for iodoacetamide under the same conditions; this result suggests that the salicylate moiety contributes to the effectiveness of ISA in inactivating this enzyme. Total protection against inactivation is provided by manganous ion and isocitrate, but not by either metal ion or isocitrate alone.
When the incubation is conducted in the absence of ligands, total inactivation can be related to the incorporation of 4 mol of ['4C]ISA/mol of enzyme; however, when the reaction is carried out in the presence of Mn" and isocitrate, the enzyme retains 100% activity, but still incorporates 2 mol of radioactive ISA/mol of enzyme. Carboxymethyl derivatives of lysine and cysteine are observed in acid hydrolysates derived from inactive enzyme prepared in the absence of ligands; whereas only carboxymethylcysteine is detected in samples derived from active modified enzyme prepared in the presence of Mn" and isocitrate. Modification of a cysteine or partial modification of a cysteine and lysine by LSA appears to be responsible for inactivation. Enzyme reacted first with unlabeled ISA in the presence of protecting substrate plus metal ion, and then with radioactive H A in the absence of ligands yields only radioactive e-dicarboxymethyllysine from the hydrolysate of the inactive enzyme. It is concluded that in this case a lysine residue attacked by 4-iodoacetamidosalicylic acid is critical for the function of isocitrate dehydrogenase and is located within the metal-isocitrate binding site. A new method is also presented for the purification of TPN-dependent isocitrate dehydrogenase using fresh pig hearts as the starting material.
Specific modification of amino acid residues of the pig heart TPN-dependent isocitrate dehydrogenase (threO-D,-isocitrate: NADP+ oxidoreductase (decarboxylating), EC 1.1.1.42) has been aimed at identifying those in the catalytic site. Methionyl (11, glutamyl  dues (5-8) have previously been implicated in the region of the substrate or coenzyme binding sites. Related studies of the more complex allosteric DPN-dependent isocitrate dehydrogenase from pig heart have revealed the involvement of a lysine residue in the binding site of the manganous-isocitrate complex (9); however, no comparable lysine has thus far been implicated in the case of the TPN-dependent isocitrate dehydrogenase.
Salicylate and its iodinated derivatives have been found to bind to the adenine binding sites of dehydrogenases and kinases (10)(11)(12). The reagent 4-iodoacetamidosalicylic acid was f i s t proposed by Baker as an active site-directed reagent for glutamate dehydrogenase (13); and the compound has been found to react irreversibly within the substrate binding site of glutamate dehydrogenase, labeling different amino acid residues at pH 7.5 (14)(15)(16) and at pH 6.0 (17). 4-Iodoacetamidosalicylic acid was selected for evaluation as a possible affinity label for the TPN-dependent isocitrate dehydrogenase since it was considered to have the potential to react specifically at either the coenzyme or the isocitrate sites. The present study shows that this reagent inactivates isocitrate dehydrogenase with modification of a limited number of amino acids and provides the first evidence that a lysine residue may be involved in the metal-isocitrate site. dium sulfate from Fisher Scientific Co. All coenzymes and substrates were purchased from Sigma Chemical Co., as were the iodoacetamide and triethanolaminehydrochloride.
Purification of Pig Heart TPN-dependent Zsocitrate Dehydrogenase-The commercial (Boehringer Mannheim Corp.) partially purified TPN-dependent isocitrate dehydrogenase has previously been used in this laboratory as the starting material for the subsequent purification of the enzyme. However, because of the expense and the irregular availability of the commercial enzyme, as well as the limited lifetime of the NADP-agarose featured in the previous purification procedure (8), a new scheme of purification of the enzyme has been devised using fresh pig hearts as the starting material.
Fresh pig hearts were obtained from a local abattoir. Excess fat and connective tissue were trimmed from the hearts. The muscle tissue was cut into 1-inch cubes and frozen at -85 "C overnight.
Batches (800 g) of semithawed tissue were homogenized a t room temperature in a Waring blendor with 2400 ml of Tris/O.Oll M citrate buffer, pH 7.2, containing 10% glycerol and 2 mM MnS04. The homogenate was centrifuged at 8,000 rpm at 4 "C for 20 min in a Sorvall RC2-B refrigerated centrifuge with a GS-3 rotor. The combined supernatants (24,000 u.e./kg of trimmed hearts) were fdtered through cheesecloth, and crystalline ammonium sulfate was added to 60% saturation a t room temperature. After 1 to 2 h at 4 "C, the suspension was centrifuged for 45 min. The pellet was used for isolation of the DPN-dependent isocitrate dehydrogenase.' The supernatant was filtered through cheesecloth and brought to 80% saturation of ammonium sulfate a t room temperature. The solution was allowed to stand at 4 "C overnight and the precipitate was collected by centrifugation for 50 min at 4 "C. The pellet was dissolved (approximately 200 units/ml) in 0.10 M triethanolamine chloride buffer, pH 7.7, containing 10% glycerol and 0.3 M Na2S04 (Buffer A) by stirring overnight at 4 "C. The protein solution was then divided in aliquots and stored at -85 "C. As based on protein determination by the method of Warburg and Christian (18) and enzymatic activity by the standard assay, the enzyme solution typically exhibited a specific activity of about 2 u.e./mg. Approximately 66% of the units present in the initial homogenate were recovered. An additional 11%. of the total initial units could be obtained upon re-fractionation of the "608 ammonium sulfate pellet" after the DPN-dependent isocitrate dehydrogenase was isolated.
The crude enzyme (2000 to 3000 units) was dialyzed for 18 h a t 4 "C against three changes of 2 liters of 0.018 M triethanolamine chloride, pH 7.1, containing 10% glycerol and 1 mM MnS04 (Buffer B). The enzyme was diluted with Buffer B to a final volume of 30 ml prior to application to a column of carboxymethylcellulose (2.5 X 30 cm, Whatman CM-52) equilibrated with Buffer B. After an inactive breakthrough peak was eluted by Buffer B, a linear gradient of Buffer B (300 m l ) tlersus (240 ml of Buffer B + 60 ml of Buffer A) was applied. Fractions of 4 ml/tube were collected at a rate of 40 to 50 ml/h. The enzyme eluted as the fourth or fifth of a group of incompletely resolved protein peaks. Fractions with specific activities greater than about 25 units/mg were pooled, leading to recovery of approximately 625t of the units applied to the column.
Matrix Gel Red-A resin (Amicon Corp.) was precycled by washing with about 50 ml of 7 M urea in 0.1 M NaOH, followed by approximately 1 liter of water and was finally equilibrated with Buffer B. A column (1.0 X 32 cm) was loaded with the pooled enzyme and then eluted with 0.018 M triethanolamine chloride, pH 8.0, containing 10% glycerol and 0.15 M Na2S04 (Buffer C) for several hours. A linear gradient was established from Buffer C (125 ml) to 0.018 M triethanolamine chloride, pH 8.0, containing 10% glycerol and 0.45 M NaS04 (125 ml) and 4-ml fractions were collected at a rate no greater than 20 rnl/h. Fractions with specific activities greater than 35 units/mg were pooled, yielding about 60% of the enzyme units applied to the column.
The enzyme was concentrated a t 4 "C by ultrafiltration using a PM-10 (Amicon) membrane. The concentrated enzyme was dialyzed against Buffer A, centrifuged and stored in convenient aliquots at -85 "C. Typically, about 18 mg of enzyme with specific activity of 40 to 45 units/mg were obtained from an initial 2000 u.e. of crude, dialyzed enzyme, to give an overall yield of approximately 37%. The resultant enzyme preparation exhibited a single band upon electrophoresis in F? polyacrylamide gels containing 2% sodium dodecyl sulfate (19) and had essentially the same amino acid composition as reported for the pig heart TPN-dependent isocitrate dehydrogenase purified by the previous procedure from commercial preparations (8). (It should be noted that for some preparations, generally when the purification ' E. V. Stevens and H. F. Colman, manuscript in preparation. procedure was extended over more than 3 to 4 days, the final specific activity was 30 to 35 u.e./mg. However, these preparations were indistinguishable in terms of electrophoretic behavior, amino acid composition, or kinetic properties from those of higher specific activity.) The protein concentration of the purified enzyme was determined using a value of 10.8 for the E;& (8). A molecular weight of 58,000 was used in all calculations (20).
Assay for Zsocitrate-Dehydrogenase Activity-Enzyme activity was measured a t 25 "c using 0.1 mM TPN, 4 mM isocitrate, and 2 mM MnS04 in 30 mM triethanolamine chloride, pH 7.4, in a total volume of 1.0 ml. Initial velocities were determined spectrophotometricaly a t 340 nm using a Gilford model 240 spectrophotometer with the scale set to 0.1 A units full scale. Specific activity is defined as micromoles of TPN' reduced per min per mg of protein.
Assay for Oxalosuccinate Decarboxylase Actiuity-The oxalosuccinate decarboxylase activity was measured spectrophotometrically at 240 nm in accordance with Grafflin and Ochoa (21). A total volume of 1.0 nd contained 0.23 mM oxdosuccinate, 0.24 mM manganous sulfate, and 0.134 M potassium chloride in 0.2 M sodium acetate buffer, pH 5.5.
Kinetics of Znactiuation-Isocitrate dehydrogenase (0.17 to 2.0 mg/ml) was incubated with 4-iodoacetamidosalicylic acid at 30 "C in 0.05 M triethanolamine chloride buffer, pH 7.0, containing 0.15 M sodium sulfate and 5% glycerol. Unless specified otherwise, the concentration of ISA' was 2 mM. Substrates, metals, or coenzymes were included in the incubation mixture, where indicated, to test for their ability to protect against inactivation. At various times, aliquots were withdrawn and assayed for isocitrate dehydrogenase activity. In all cases controls were run in which enzyme was incubated under the same conditions (i.e., in the presence of the appropriate ligands) but without ISA in order to correct for any minor activity changes. The rates of reaction of isocitrate dehydrogenase with ISA were determined from semilogarithmic plots of E/Eo as a function of time, where E and EO represent activities a t a given time for the experimental and control reaction, respectively.
Incorporation of Radioactive ZSA into Zsocitrate Dehydrogenase-In order to ascertain the stoichiometry of reaction of ISA with isocitrate dehydrogenase, enzyme was incubated with 2 mM [I4C]ISA at 30 "C under the conditions indicated above. At specified times, dithiothreitol was added to yield a concentration of 20 mM in order to decompose the ISA and prevent further reaction. Measurements of enzymatic activity following the addition of dithiothreitol demonstrated that no further change occurred over a t least a 30-min period. After 30 min, the reaction mixture was diluted with a solution of urea to yield a concentration of 5 M, and the enzyme sample was applied to a Sephadex G-25 column (1 X 28 cm) equilibrated with 0.05 M triethanolamine chloride, pH 7.0, containing 5% glycerol, 0.15 M Na2S04, and 6 M urea. Enzyme was recovered in the void volume, completely separated from the excess ISA, which is adsorbed to Sephadex G-25. The radioactivity of the fractions was measured using a Packard Tri-Carb liquid scintillation counter, model 3330. The protein concentration was determined using the Bio-Rad Protein Assay which is based on the method of Bradford (22). Standard solutions were prepared by dilution of native isocitrate dehydrogenase with the same buffer used to equilibrate the Sephadex column.
Identification of Amino Acid Residues-In order to identify the modified amino acids, isocitrate dehydrogenase reacted with ISA under various conditions was separated from excess reagent by gel filtration as described above, dialyzed against distilled water for 24 h and then lyophilized. The protein samples were hydrolyzed with 6 N HCI under vacuum at 110 "C for 20 h after which they were taken to dryness. Each protein sample was applied to a Beckrnan model 12oC amino acid analyzer and fractions of the effluent were collected a t 1min intervals (1.7 ml) after passage through the photometer. The fractions were assayed for radioactivity in the liquid scintillation counter by taking 1.0 ml of each fraction, adding 0.1 ml of concentrated HCI to bleach the ninhydrin, and 10 ml of ACS counting fluid (Amersham Searle). Elution positions of radioactive peaks obtained from modified proteins were compared with those of standard carboxymethyl-amino acids chromatographed on the amino acid analyzer under the same conditions.

RESULTS
Effect of Salicylic Acid on Catalytic Activity of Isocitrate Dehydrogenase-To obtain an initial indication as to what ' The abbreviation used is: ISA, 4-iodoacetamidosalicylic acid. enzyme site might be attacked by 4-iodoacetamidosalicylic acid, the parent compound salicylic acid was tested as a competitive inhibitor with respect to isocitrate or TPN using the isocitrate dehydrogenase assay. While maintaining the concentrations of other ligands given under "Experimental Procedures," the K,,, for DL-isocitrate was determined in the absence or presence of four constant concentrations of salicylic acid from 5 to 22 mM. The maximum velocity was not altered over this range of salicylic acid concentrations, but the apparent K , for total DL-iSOCitrate in the presence of 2 mM MnS04 increased with increasing salicylic acid concentrations. It has been proposed that manganous-isocitrate is the actual substrate for the TPN-dependent isocitrate dehydrogenase (23). Salicylic acid is known to form a chelate with manganous ion, with an association constant of lo5.' M (24). In order to assess whether the apparent behavior of salicyclic acid as a competitive inhibitor could be attributed to its depletion of the manganous ion or the manganous-isocitrate complex, the K,,, for substrate was calculated in terms of manganous-isocitrate using the equations and computer program described previ- enzyme merely by sequestering Mn'+ , but rather acts as a competitive inhibitor. In contrast, salicylic acid does not significantly affect the K,,, for TPN when measured at saturating concentrations of isocitrate (under the standard conditions given under "Experimental Procedures"), indicating that salicylic acid does not bind to the coenzyme binding site of isocitrate dehydrogenase.
Inactivation of Isocitrate Dehydrogenase by 4-Zodoacetamidosalicylic Acid-Incubation of enzyme with 2.9 mM 4iodoacetamidosalicylic acid at pH 7.0 leads to loss of both oxalosuccinate dehydrogenase and isocitrate dehydrogenase activity, as shown in Fig. 1, suggesting that the group(s) attacked are essential for both functions of the enzyme. The reaction obeys pseudo-first order kinetics as far as 98% inactivation, yielding a rate constant of 0.0306 min".
Determination of the pseudo-first order rate constant for inactivation of isocitrate dehydrogenase activity from 0.5 to 4.0 mM ISA reveals an apparently linear dependence on reagent concentration (Fig. a), with a calculated second order rate constant of 10.7 mir" M" for ISA. Saturation kinetics, which might be expected for an affinity label, are not observed; however, it may be notable that the highest ISA concentration used is far lower than the K , value of 18.2 mM calculated for salicylic acid. The salicyclic acid moiety of 4-iodoacetamidosalicylic acid is important in determining the relatively high rate of inactivation of enzyme by 4-ISA since 2 mM iodoacetamide, the corresponding compound which lacks this group, inactivates the enzyme with a pseudo-first order rate constant of only 0.00193 min" under the same conditions. A second order rate constant of 0.963 min" M" may be calculated for the inactivation of isocitrate dehydrogenase by iodoacetamide, a value less than one-tenth that observed for 4-iodoacetamidosalicylic acid.

FIG. 2.
Dependence of pseudo-first order rate constant for inactivation of isocitrate dehydrogenase on ISA concentration. Pseudo-first order rate constants for loss of isocitrate dehydrogenase activity were determined as in Fig. 1 over a  Effect of Substrates on the Rate of Inactivation by 4-Iodoacetamidosalicylic Acid-As shown in Fig. 3, the addition of the substrates isocitrate (4 m M ) and manganous ion (2 mM) to the incubation mixture totally protects the enzyme against loss of isocitrate dehydrogenase activity produced by 2 mM ISA. The activity measurements €or the sample containing isocitrate and manganous ion together with ISA fall on the same line (line B) as those for the two control samples (z.e. enzyme in the absence or presence of the substrates but with no added ISA). These results suggest that reaction occurs in the region of the active site. Table I records the effect of other ligands on the rate constant for inactivation of isocitrate dehydrogenase by 2 mM ISA. All ligands are present at concentrations high relative to their known dissociation or Michaelis constants. Although isocitrate is known to bind to the enzyme in the absence of metal ion (261, isocitrate in the absence of metal does not appreciably decrease the rate constant for inactivation (Table   I, lines 3,4); these results suggest that the manganous-isocitrate complex postulated to be the actual substrate of the TPN-dependent isocitrate dehydrogenase (23) is required to prevent reaction of ISA at the critical site. Several metals are known to interact with the enzyme: manganous, magnesium, and zinc ions serve as activators, whereas calcium ion functions as an inhibitor competitive with respect to Mn2+ (23). The data of Table I (lines 5 to 8) indicate that none of these metals by themselves causes a marked decrease in the rate constant for inactivation by ISA. However, when present together with isocitrate, the activators Mn2+ and Mg2+, as well as the inhibitor Ca2+, completely prevent measurable inactivation (Table I, lines 2,9,10). These results suggest that there are similarities in the binding sites of these three metalisocitrate complexes. Since zinc acetate plus isocitrate (Table  I, line 11) produces only a minima1 decrease in the inactivation rate, the site occupied by the zinc-substrate complex must be distinguishable from those to which the other metal-chelates bind, despite the fact that zinc ion is also an activator of the isocitrate dehydrogenase reaction. The product of the reaction, a-ketoglutarate, fails to alter the inactivation rate in the absence or presence of metal ion and similar results are

Effect of ligands on inactiuution rate by 4-iodoacetamidosalicylic
acid Isocitrate dehydrogenase (0.17 mg/ml) was incubated at pH 7.0 with 2 mM ISA under the conditions described under "Experimental Procedures." Substrates, metals, or coenzymes were included in the incubation mixture as indicated. The rate constants were determined from semilogarithmic plots of E/Eo uersus time as in Fig. 1 where E is determined from the isocitrate dehydrogenase activity. obtained for the oxidized coenzyme, TPN (lines 12 to 15). The reduced coenzyme yields some protection when present together with manganous ion (line 17, 18); however, the fact that the inactivation rate constant does not decrease more than 2.3-fold when the TPNH concentration is increased suggests that this small diminution in the inactivation rate is an indirect effect of binding to a site distinct from that attacked by ISA.
In contrast, the data of Table I1 demonstrate that in the presence of a constant total MnS04 concentration of 2 mM, the rate constant €or inactivation by ISA decreases progressively with increasing isocitrate concentration, reaching an apparent kobs of zero at very high concentrations of isocitrate.
If it is assumed that isocitrate [q binds reversibly at the critical site attacked by ISA, the following equation can be used to describe the relationship between Kohs (the pseudofirst order rate constant for inactivation by ISA in the presence of isocitrate), k, the inactivation rate constant in the absence of ligands, which is 0.021 min" at 2 mM ISA), and Kd (the dissociation constant for the enzyme-isocitrate complex): Incorporation of Radioactive ISA into Isocitrate Dehydrogenase-4-Iodoacetamidosdicylic acid reacts irreversibly and in a limited manner with TPN-dependent isocitrate dehydrogenase. The stoichiometry of the reaction was determined as described under "Experimental Procedures" by measurement of the incorporation of ['4C]ISA as a function of time of incubation with 2 mM reagent in the absence of protecting ligands. A plot of residual enzymatic activity uersus moles of reagent incorporated per mol of protein, shown in Fig. 4-4,  Protection by different concentrations of isocitrate against inactivation by ZSA Enzyme (0.17 m g / d was incubated at pH 7.0 with 2 mM ISA under the conditions described under "Experimental Procedures." MnSO, (2 mM) was added to the reaction mixture together with the indicated total concentrations of DL-isocitrate. The rate constants were determined as described in Table I  indicates that after an initial average incorporation of 0.5 to 1 mol without much change in isocitrate dehydrogenase activity, there is a linear inactivation with increasing number of groups modified. Extrapolation to 100% inactivation yields a value of 4.17 mole reagent incorporated/mol of enzyme, implying that reaction with a maximum of four groups is responsible for loss of activity. Fig. 4 B records the incorporation of radioactive reagent as a function of time of incubation with 2 mM ISA. When the reaction is carried out in the absence of protecting ligands (line a), 3.7 mol of reagent are incorporated at 125 min, when the residual activity has decreased to about 14% of its original value. In contrast, when Mn2+ and isocitrate are included in the incubation mixture (line b), the enzyme retains 100% of its activity, but 2.1 mol of reagent are stiU incorporated at 125 min; extension of the incubation mixture to 255 min does not lead to significant increase in the incorporation of reagent under these conditions. If the four enzymatic sites which react in the absence of ligands include the two which react in the presence of manganous ion and isocitrate, then incorporation of a maximum of 2 mol of ISA causes inactivation.
An experiment was conducted in which isocitrate dehydrogenase was incubated for 180 min with 2 mM nonradioactive ISA in the presence of 2 mM MnS04 and 4 mM isocitrate.
Activity was completely retained, and although the incorporation was not measured directly, presumably an average of 2 mol of reagent should have been incorporated, as indicated from Fig. 4B, Zine b. At the end of that time, the substrates as well as nonradioactive ISA, were removed by gel fitration of Sephadex G-25; the enzymatically active, modified enzyme was then incubated with ['4C]ISA and the radioactive incorporation was measured as described under "Experimental Procedures." This pretreated active enzyme is much more rapidly inactivated by ISA than is native enzyme: after incubation with 0.35 mM [I4C]ISA for 50 min, the residual activity had decreased to 20% of its original value, concomitant with the incorporation of 2.16 mol of radioactive reagent. Although the rates of inactivation are different for native and this pretreated active enzyme (henceforth, termed "protected, then unprotected enzyme"), these results indicate that incorporation of a maximum of 2.7 mol of ISA is required for complete inactivation of isocitrate dehydrogenase under these conditions.
Identification of Modified Amino Acids-In order to identify the types of amino acids which are modified by ISA in isocitrate dehydrogenase, acid hydrolysates were prepared  . Acid hydrolysis of modified protein converts the acetamidosalicylic acid derivatives to their corresponding carboxymethyl derivatives. The distribution of modified residues in these several samples was determined from the radioactivity in the effluent of an amino acid analyzer, as described under "Experimental Procedures", with the results shown in Fig. 5, A to C. The only radioactive peaks observed correspond in elution position to edicarboxymethyllysine, glycolic acid, and carboxymethylcysteine. No radioactivity was detected in the region where carboxymethylhistidine derivatives are expected to elute (immediately following glutamic acid) (27) or where carboxymethylhomocysteine (the major acid degradation product of carboxymethylmethionine) elutes (28). For the unprotected enzyme (Fig. 5A), 76% of the radioactivity or 2.8 mol/mol of protein was accounted for by the product of reaction with cysteine, carboxymethylcysteine; while a minor amount of 5% or 0.2 mol/mol of protein was detected as glycolic acid, which may result from decomposition of a glutamate or asparate derivative (29). Another 19% of the radioactivity or 0.7 mol of carboxymethyl group/mol of protein was measured as the product of reaction with lysine, edicarboxymethyllysine; since this is a disubstituted lysine derivative, the radioactive incorporation represents 0.35 mol of modified lysine/mol of protein.
For the protected enzyme (Fig. 5B), 95% of the radioactive or 2.1 mol/mol of protein appeared as carboxymethylcysteine. These results suggest that the 86% inactivation observed in this unprotected enzyme may result from modification of a cysteine residue or from partial modification of a cysteine and a lysine. For the protected, then unprotected enzyme (Fig.  5 C ) , 92% of the radioactivity (2.0 mol of carboxymethyl groups/mol of protein) or 1.0 mol of edicarboxymethyllysine/ mol of enzyme was observed. This result indicates that complete inactivation can result from modification of only about 1 lysine residue.

DISCUSSION
The reagent 4-iodoacetamidosalicylic acid was initially selected for reaction with the TPN-dependent isocitrate dehydrogenase because of the possibility that it might function as an affinity label of this enzyme. Certain features of the reaction, but not all, accord with the expectation for an affinity label. The parent compound salicylic acid does behave as a weak competitive inhibitor with respect to isocitrate (KI = 18.2 mM). Although the inactivation rate constant exhibits a linear dependence on ISA concentration (0.5 to 4.0 mM) rather than the saturation kinetics expected of an active site-directed reagent, it might be argued that (because of the rapid rate of the covalent reaction of ISA with the enzyme) these studies were restricted to a concentration range so far below the dissociation constant of the enzyme-reagent complex that the concentration dependence would necessarily appear to be linear. The salicylic acid portion of the molecule is important to the efficacy of ISA in reacting with isocitrate dehydrogenase, since the inactivation rate constant for ISA is 11 times higher than that for the simple alkylating agent iodoacetamide. Similarly, the total protection against ISA inactivation provided by metal and isocitrate is consistent with affinity labeling of the substrate binding site of the enzyme. On the other hand, the observation that 4 mol of reagent are incor-porated when total loss of activity occurs indicates that ISA alkylates several sites of isocitrate dehydrogenase and does not act exclusively as an affinity label of a single site on the enzyme. Nevertheless, analysis of this reaction may serve to identify the amino acid residue participants of the active site.
Insight into the location of the essential residue(s) attacked by ISA can be gained by evaluating the ligands which protect against inactivation. Neither TPN nor TPNH causes a marked decrease in the inactivation rate, indicating that reaction of ISA does not occur within the coenzyme binding site. The substrate site is implicated in the ISA attack by the observation of total protection provided by the substrate isocitrate when present together with metal ion in the reaction mixture. Since the product a-ketoglutarate does not influence the rate of inactivation when added alone or with metal ion, it appears that the @-carboxyl group of isocitrate (i.e. the group which is lost during the decarboxylation reaction) is required for the protective effect. It is possible that the reactive iodoacetamido group of ISA is located near the binding site of the P-carboxylate moiety and covalently modifies an amino acid residue which normally strengthens the isocitrate binding. Metal ion is required for the protection by isocitrate, as it is for the enzyme-catalyzed oxidative decarboxylation. The metal ion requirements for isocitrate protection against ISA inactivation can be fulfilled by manganous and magnesium ions, which support the overall catalytic reaction and (in the presence of isocitrate) strengthen the binding of a coenzyme analogue (30), as well as by calcium ion, which acts as an inhibitor competitive with respect to Mn2' and (in the presence of isocitrate) weakens the binding of the coenzyme analogue. Although these metals may have different effects on the function and/or conformation of the enzyme, they appear to occupy similar sites at least with respect to their protection of the critical residue(s) which react with ISA. In contrast to the other metal activators, zinc ion has been found to facilitate the enzyme-catalyzed dehydrogenation of isocitrate, but not the enzyme-dependent decarboxylation of oxalosuccinate (23). Thus, it is not surprising that zinc ion is also distinguished from the other divalent metal ions tested in yielding only a small decrease in the ISA inactivation rate when added with isocitrate. Zinc must bind to a site which is not identical with that at which other metals are bound.
While isocitrate dehydrogenase catalyzes both the pyridine nucleotide-dependent dehydrogenation of isocitrate to form oxalosuccinate and the subsequent decarboxylation of the pketo acid to yield a-ketoglutarate, it might be anticipated that certain amino acid residues would participate in only one of these reactions. For example, alkylation of a methionyl residue by iodoacetate causes a profound disruption of the dehydrogenase activity with a relatively minor effect on the decarboxylase activity (1, 31). Furthermore, incubation of isocitrate dehydrogenase with ethoxyformic anhydride causes a 9-fold greater loss of dehydrogenase than of decarboxylase activity and it has been postulated that this results from modification of a critical histidine in the nucleotide binding site (30). In contrast, an amino acid residue involved in binding the Pcarboxylate group of isocitrate might well be expected to be involved in binding the analogous groups of oxalosuccinate. The observation that 4-iodoacetamidosalicylic acid causes an equal loss of the dehydrogenase and decarboxylase activities of isocitrate dehydrogenase indicates that the group(s) which is modified is important for both phases of the overall oxidative decarboxylation reaction of the enzyme; this observation is consistent with the postulate that the ISA-susceptible group(s) has as its normal function interaction with the pcarboxylate moiety of the substrate.
Although 4 mol of radioactive 4-iodoacetamidosalicylic acid are incorporated into isocitrate dehydrogenase concomitant with complete inactivation, the observation that two groups still react covalently in the presence of isocitrate and Mn2+ when no inactivation occurs suggests that the requirement for inactivation is modification of an average of only 1 to 2 amino acid residues. The prime candidates for the critical residues are a cysteine and a lysine, since these are the extra amino acids which are labeled in the absence of ligands but not in the presence of isocitrate. Johanson and Colman have presented evidence for a cysteine residue in the manganousisocitrate binding site which is attacked by 5,5'-dithiobis(2nitrobenzoic acid) and have isolated a critical cysteine-containing peptide (8). Inactivation by ISA in the absence of ligands (unprotected enzyme) may be caused by alteration of the same cysteine. However, it is also possible that modification of a combination of cysteine and lysine are responsible for inactivation in this case. It is notable that after 2 cysteines have reacted with ISA in the presence of isocitrate and Mn2+ and the protectants are removed, ISA still inactivates isocitrate dehydrogenase, but with modification of only a lysine residue. For this protected then unprotected enzyme, reaction with about 1 lysine appears to be responsible for the inactivation by 4-iodoacetamidosalicylic acid. Carbamylation of 1 lysine per average subunit was reported to irreversibly inactivate the allosteric pig heart DPN-dependent isocitrate dehydrogenase (9) and it was postulated that the reaction locus was in the binding site of the manganousisocitrate complex. Cyanate failed to inactivate the TPNdependent isocitrate dehydrogenase under similar conditions, which suggested that the substrate binding sites were not identical in the two isocitrate dehydrogenases from the same species and tissue. In the present study, the effect of the salicylic acid moiety in directing the reaction of 4-iodoacetamidosalicylic acid may facilitate alkylation of a relatively inaccessible, yet critical lysine residue. The protonated lysine residue may participate in an electrostatic interaction with the P-carboxylate group of isocitrate which strengthens and properly orients the binding of the isocitrate-metal complex.