Functional Groups of Diphosphopyridine Nucleotide-linked Isocitrate Dehydrogenase from Bovine Heart STUDIES OF CYSTEINE RESIDUES*

SUMMARY DPN-linked isocitrate dehydrogenase from bovine heart contains 6 half-cystine residues per subunit of molecular weight of 42,000. All of these residues are present as cysteine since six sulfhydryl groups per subunit can be modified by 5,5’-dithiobis(2-nitrobenzoate) (DTNB). Spectrophoto-metric measurements indicate that the modilication of three thiol groups which react preferentially with DTNB is directly proportional to loss of activity, up to 70% inactivation. 2-Mercaptoethanol and dithiothreitol completely restored activity of enzyme preparations where less than two sulfhy-dry1 groups per subunit had been modified by DTNB, but less than 10% reversal was obtained when more than four sulfhydryl groups had been modified. No reversal with Na2S03 and only partial reversal with KCN is obtained, even though both reagents release 2-nitro&mercaptobenzoate in stoichiometric amounts from the modified enzyme. Manganous isocitrate and, to a lesser extent, DPN+ and DPNH protect the enzyme against inactivation by DTNB. In the presence of 0.8 M (NH&S04, one sulfhydryl


DPN-linked
isocitrate dehydrogenase from bovine heart contains 6 half-cystine residues per subunit of molecular weight of 42,000.
All of these residues are present as cysteine since six sulfhydryl groups per subunit can be modified by 5,5'-dithiobis(2-nitrobenzoate) (DTNB).

Spectrophotometric measurements
indicate that the modilication of three thiol groups which react preferentially with DTNB is directly proportional to loss of activity, up to 70% inactivation.

2-Mercaptoethanol
and dithiothreitol completely restored activity of enzyme preparations where less than two sulfhy-dry1 groups per subunit had been modified by DTNB, but less than 10% reversal was obtained when more than four sulfhydryl groups had been modified. No reversal with Na2S03 and only partial reversal with KCN is obtained, even though both reagents release 2-nitro&mercaptobenzoate in stoichiometric amounts from the modified enzyme. Manganous isocitrate and, to a lesser extent, DPN+ and DPNH protect the enzyme against inactivation by DTNB.
In the presence of 0.8 M (NH&S04, one sulfhydryl group per subunit can be modified by DTNB and in the presence of manganous isocitrate, one additional group can be modified with full retention of activity. These results suggest that only one of the three reactive sulfhydryl groups is required for activity. The pH inactivation profile with DTNB shows that the activity-related thiol group has a pK, of 7.6. Increasing electrolyte concentration does not &ect the rate of enzyme inactivation by DTNB; however, it does increase the reactivity of a sulfhydryl group which is not required for activity and decreases the protection given by manganous isocitrate. These results suggest that high salt concentrations interfere with binding of manganous isocitrate to the enzyme, possibly at an active center amino group (FAN, C. C It was shown previously that DPN-linked isocitrate dehydrogenase (three-n,-isocitrate + DPN+ -cY-ketoglutarate f CO2 + DPNH + H+; EC 1.1.1.41) from bovine heart and porcine liver was inhibited by low concentrations of p-mercuribenzene sulfonate and Ellman's reagent (l-3).
The inactivation by 67 PM p-mercuribenzene sulfonate could be partially prevented by a combination of isocitrate and Mn2+. DPN+ or ADP did not afford such protection.
In this paper, the reaction of bovine heart enzyme with a number of sulfhydryl group modifying agents, especially 5,5'-dithiobis(2-nitrobenzoate) is reported. The protection by substrates and cofactors against these inactivating agents is also described.
The present investigation demonstrates that 3 of the 6 cysteine residues per subunit of enzyme can be modified without drastic alteration of protein structure and that loss of enzyme activity correlates with modification of one of these three reactive thiol groups.
Manganous isocitrate protects this activity-related sulfhpdryl group against modification by DTNB.' The intercept on the X-axis of the straight line portion of this plot is (0.2 mM)-l.
Inactivation by DTNB-Pseudo-first order rate constants of enzyme inactivation were determined with varying reagent concentration and constant enzyme under the conditions shown (11). When isocitrate dehydrogenase was treated with a relatively large excess of reagent ([DTNB] :[subunit] ratio = 833) in the presence of 8 M urea or 0.1 y0 sodium dodecyl sulfate, approximately six sulfhydryl groups per subunit of enzyme were modified rapidly (Fig. 2, Curves a and b) ; about the same number reacted in the absence of denaturants although the rate of reaction was slower (Fig. 2, Curves c and d). The number of groups modified agrees well with the 6 half-cystine residues per subunit determined by amino acid analysis (Table II) and indicates that all of this amino acid is present as cysteine.  only approximately three sulfhydryl groups per subunit were modified after 80 min (Fig. 2, Curve e).
A pseudo-first order plot of the data from an experiment (Fig.  2, Curve d) in which the mixture was preincubated with a relatively high concentration of DTNB in 50 mM sodium-Hepes at pH 7.2 is shown in Fig. 2 (inset). The biphasic nature of this plot indicates a difference in the reactivities of the 6 sulfhydryl residues, groups modified in the initial time period being more reactive than those modified later.
The intercept of the lines occurs at 38 min (Fig. 2, &set), a point indicating modification of 3.4 thiol groups per subunit (Fig. 2, Curve d).
Inactivation a?zd iModijcation---Incubation of the enzyme with a moderate excess of DTNB ([DTNB]: [subunit] = 9.6) in 0.1 M sodium-Hepes buffer at pH 7.06 resulted in progressive modification and inactivation of the enzyme with time (Fig. 3A).
A linear correlation between enzyme inactivation and modification by DTNB was observed when up to 2.8 sulfhydryl groups had been modified and 70% of the activity had been lost (Fig. 3A, inset). Interpolation of the line in Fig. 3A (inset) to complete inactivation indicates that loss of activity is correlated with the modification by DTNB of three to four sulfhydryl groups on the enzyme. A similar relationship between number of sulfhydryl groups modified and inactivation was observed in the reaction of the enzyme with N-[14C]ethylmalcimide.
Changes in the composition of the preincubation mixture altered the pattern of inactivation and modification of the enzyme by DTNB (Fig. 3, B and c). In the presence of buffer containing 0.8 M (NH&SO4 and a small excess of reagent ([DTNB]: [subunit] = 5) about 1 sulfhydryl group per subunit was modified in 60 min; however, there was essentially no change in enzyme act'ivity during this period (Fig. 3B).
The addition of a second and larger amount of DTNB ([DTNB]: [subunit] = 50) at the end of the 60.min incubation period resulted in loss of enzyme activity with time, almost complete inactivation occurring when two additional sulfhydryl groups per subunit had been modified (Fig. 3B).
Even though manganous isocitrate protected against inactivation, reaction of the enzyme with DTNB ([DTNB]: [subunit] = 21) occurred. Under these conditions, approximately two sulfhydryl groups could be modified with retention of 90% of the catalytic activity (Fig. 3C).
Protection against Inactivation-The effects of substrates and cofactors on rates of inactivation by two levels of DTNB in a buffer containing 50 mM Tris-HCl, 10% glycerol, and 0.2 mM EDTA at pH 7.2 is shown in Table III.
The combination of Mn2f and isocitrate gave the most effective protect,ion as noted with other modifying agents (Table I) ; equivalent concentrations of total nL-isocitrate or total Mn*+ were ineffective when added singly.
Some protection was afforded by 0.2 mM DPNH and by DPN+ at 0.5 and 5 mM. However, ADP at a concentration (0.25 mM) where it is a positive modifier of the enzyme (3), was ineffective when added singly or in combination with DPN+.2 Manganous isocitrate has been observed to give protection against inactivation in a number of situations (Tables I and III). Under certain conditions the presence of elevated concentrations of electrolytes had a differential effect on enzyme modification and inactivation (Fig. 3B). The effect of the ionic composition of the preincubation medium was, therefore, examined in more detail in the absence and presence of manganous isocitrate, the protector against inactivation.
At the concentration of DTNB used in the experiments shown in  (Table I) about the same protection by DPN+ and DPNH against inhibition by the mercurial could be obtained as with DTNB (Table III). Preincubation was carried out in 50 mM Tris.HCl-100/o glycerol-O.2 mru EDTA buffer at pH 7.2 and 23". Aliquots were taken and assayed at varying time intervals.
Inactivation was expressed as the pseudo-first order rate constants (k.,,).
* The concentrations of compounds added is shown in parentheses.
Hepes buffer alone; however, 0.8 M LizSO decreased the rate of inactivation appreciably.
Protection by manganous isocitrate against inactivation by DTNB was influenced markedly by the ionic composition of the incubation medium (Table IV).
The protection was more pronounced in 50 mM Tris .HCl-10% glycerol buffer than in 50 mM sodium-Hepes buffer. The protection declined upon addition of (NH&SO4 or L&SO4 to the Hepes buffer media, and variations in salt concentration had a greater effect on protection against inactivation by manganous isocitrate than on the rate of inactivation by DTNB. Enzyme inactivation was not affected by manganous isocitrate in preincubation mixtures containing 0.8 M L&Sod. The differences in concentrations of (NH&S04 and L&SO4 used in experiments shown in Table IV  o Isocitrate dehydrogenase (0.01 mM) was incubated with 0.25 mM DTNB at pH 7.2 and 23" in the buffer systems described.
Inactivation was expressed as the apparent pseudo-first order rate constants (k,&. * nn-Isocitrate (20 mM) and MnSOd (2 mM) were present in the preincubation systems. c The enzyme preparation used had been pretreated with 0.05 mM DTNB in 5 mM sodium phosphate and 0.8 M (NH,)&SOd at pH 7.2, resulting in modification of 0.95 thiol groups per subunit and retention of 95% of the activity.
The treated enzyme was freed from DTNB by gel filtration on Bio-Gel P-10 before the modification described here.
d The enzyme preparation used had been pretreated with 0.10 mM DTNB in a buffer containing 5.3 mM nn-isocitrate, 1.3 mbf MnSOn, 14yc glycerol, and 50 mM sodium-Hepes at pH 7.07. This resulted in the modification of 1.9 thiol groups per subunit with retention of 98yc of the activity.
The treated enzyme was freed from DTNB by gel filtration before the modification described here. be more effective than sodium or ammonium salts in slowing the interaction of enzyme with a number of group specific reagents. This may indicate relatively specific binding of Li+ to a cationic site(s) located near a sulfhydryl group associated with enzyme activity.  Enzyme-It is known that TNB-mixed disulfide bonds can react with a number of nucleophiles.
The reaction mechanisms and kinetics of thiolysis, cyanolysis, and sulfitolysis have been thoroughly investigated with model compounds and proteins containing TNB-mixed disulfide linkages (12, 13).
The effects of a number of nucleophiles on isocitrate dehydrogenase modified initially in different degrees by DTNB are shown in Table VI. Complete regeneration of activity was obtained with 2-mercaptoethanol and dithiothreitol with enzyme preparations where less than two sulfhydryl groups per subunit had been modified.
Dithiothreitol was more effective than 2-mercaptoethanol; with enzyme in which three sulfhydryl groups have been modified, 5 mM dithiothreitol led to essentially complete reversal of inhibition while only partial restoration was obtained with 70 mM 2-mercaptoethanol.
The effectiveness of dithiothreitol may be due to its efficiency in cleaving the disulfide bond of the TNB-enzyme derivative to form the reduced enzyme without appreciable accumulation of intermediary (and presumably inactive) mixed disulfide forms of the reagent. This would be in accord with the previous observation that in the reduction of disulfide bonds the reagent favors formation of a stable internal six-membered cyclic disulfide (14). However, it should be noted that the concentration of 2-mercaptoethanol used here was high enough in other systems to cause quantitative reduction of protein disulfide bonds (15). tative liberation of TNBa and, presumably, formation of the S-sulfo analoguc of the enzyme. However, this reaction was not accompanied by restoration of activity (Table VI) indicating that the product of sulfitolysis is enzymatically inactive. The effect of sulfite is not an inactivation of the enzyme per se since treatment of the DTNB-modified enzyme preparations did not lead to a decrease in activity (Table VI) and reaction of native isocitrate dehydrogenase with the reagent did not result in inactivation.

Modification
of more than three sulfhydryl groups per subunit led to irreversible denaturation of the enzyme as indicated by protein precipitation and the observation that only a 39% reversal of inhibition was obtained with 2-mercaptoethanol when 3.5 sulfhydryl groups per subunit had been modified.
An almost complete lack of reversibility was observed in separate experiments when more than four sulfhydryl groups per enzyme subunit had been modified.
It was shown in separate experiments that reaction of cyanide with DTNB modified enzyme led to stoichiometric liberation of TNB and corresponding formation of the enzyme thiocyanate compound as determined by the method of Schneider and Westley (16). Cyanolysis gave 20 to 30% reversal of inhibition when two to 3.6 sulfhydryl groups had been modified by DTNB; however, a much smaller effect was observed when 1.5 groups had been modified initially.
Replacement of the sulfhydryl group at or near the active center by thiocyanate with retention of activity has been reported for aspartic transcarbamylase   Fig. 4 could be fitted to sigmoidal curves and the infiection points were calculated from a polynomial equation which gave the best fit to these curves. The rates of inactivation differed at the two levels of DTNB used; however, identical values of pK, 7.6 were obtained with both concentrations of inhibitor.

DISCUSSION
The 6 cysteine residues per subunit of DPN-linked isocitrate dehydrogenase can be divided into two groups differing in reactivity with DTNB.
Under mild conditions only three sulfhy-dry1 groups were modified (Fig. 2, Curve e). However, the remaining residues reacted with prolonged incubation time when the concentration of DTNB was raised (Fig. 2, Curve c) or when the ionic strength of the buffer was lowered (Fig. 2, Curve d and inset). All six sulfhydryl groups are easily accessible to DTNB in the presence of protein denaturants (Fig. 2, Curves a and b), and it is likely that the 3 less reactive residues are buried inside of the protein matrix and have a role in maintaining the structure of the protein.
The possibility that modification of these residues leads to unfolding of the polypeptide chains and destruction of the proper protein conformation is supported by the observations that reaction of more than 3 sulfhydryl groups with DTNB led to protein precipitation and that less than 10% of the original activity could be restored by thiol containing compounds after more than four sulfhydryl groups per subunit of isocitrate dehydrogenase had been modified (cf. Table VI).
Loss of activity can be correlated with modification by DTNB of the three more reactive sulfhydryl groups on the enzyme ( Fig   3A, inset). The extensive reversal of inactivation by 2-mercaptoethanol and dithiothreitol when fewer than three sulfhy-dry1 groups per subunit had been modified (Table VI) suggests that modification by DTNB does not destroy the protein structure irreversibly.
However, only one of the three reactive cysteine residues appears to be related to activity since under selective conditions and in the presence of the substrate manganous isocitrate, two sulfhydryl groups can be modified by DTNB with almost complete retention of activity (Fig. 3C). The relative lack of conformational distortion of enzyme in which two sulfhydryl groups per subunit have been modified is also supported by the observation that the rate of inactivation by DTNB in the absence or presence of protector is identical for such a TNB-enzyme derivative and the nat,ive enzyme (Table  IV).
The location of the single activity-related cysteine residue at or near the active center of the enzyme is favored by the observation that manganous isocitrate and, to a lesser extent, DPN+ and DPNH protect the enzyme against inactivation by DTNB (Table III).
Nevertheless, it is not completely certain whether this thiol group per se is necessary for catalysis since partial restoration of activity was obtained upon cyanolysis of DTNB-inactivated enzyme in which more than two t,hiol groups had been modified (Table VI).
However, the fact that cyanide brought about practically no reversal of inactivation when 1.5 residues of the enzyme had been modified (Table VI) suggests that conversion of a sulfhydryl to a thiocyanate group at the catalytic center leads to loss of activity.
It was shown in separate experiments with enzyme preparations at the lower level of inactivation, freed from unreacted DTNB by gel filtration, that there is a displacement by cyanide of TNB accompanied by a stoichiometric appearance of thiocyanate groups on the protein as determined by the method of Schneider and Westley (16). The partial restoration of activity of enzyme preparations in which 2 to 3.6 thiol groups had been modified (Table VI) may be related to the divergence from linearity of enzyme modification and inactivation when more than 70% of the activity has been lost (Fig. 3A, inset). It may be that subsequent to the initial reaction with DTNB of 1 to 2 sulfhydryl residues there is increased modification of groups, other than the activity-related cysteine residue, which are involved in maintaining the conformation of the enzyme essential for activity. Replacement of the bulky TNB substituents by the smaller thiocyanate groups at the latter cysteine residues of the protein may lead to release of steric changes or conformational strains which allow a partial restoration of activity in enzyme molecules in which unmodified activity-related sulfhydryl groups are retained. The situation is further complicated by the fact that active DI'N-linked isocitrate dehydrogenase contains eight apparently identical subunits (4). It is presently unknown whether recombination of subunits modified to different degrees can lead to formation of active partially derivatized enzyme.
The pH rate profile of enzyme inactivation by DTNB suggests that the pK, of the active center is at 7.6 and that the reactive species is a mercaptide ion (Fig. 4). This value of pK, is lower by 2.6 pH units than might be expected for a cysteine residue of a protein (18,19). The relatively low pK, values of the activityrelated sulfhydryl group and of a specific amino group described previously (5 ,6) suggest that these residues of DPN-linked isocitrate dehydrogenase are located in a hydrophobic environment. Support for the presence of a hydrophobic region in DPN-linked isocitrate dehydrogenase which is related to enzyme activity has been obtained in recent studies with fluorescent probes.* The influence of a hydrophobic environment on the reactivity of thiol groups has been demonstrated in studies of a series of N-acyl cysteine derivatives in micelles (20). The values predicted for the free energy change for formation of thiol hydrophobic bonds in this model system agreed with those calculated previously by Nemethy and Scheraga (21) and Tanford (22).
The effects of changes in electrolyte concentration and temperature give further insights into the reactivity of the enzyme with DTNB and substrate.