Inhibition by Dithiothreitol of the Utilization of Glutamine by Carbamyl Phosphate Synthetase EVIDENCE FOR FORMATION OF HYDROGEN PEROXIDE*

SUMMARY The glutaminase activity of glutamine-dependent carbamyl phosphate synthetase (Escherichia coli) and that of the separated light subunit of this enzyme, as well as the glutaminedependent synthetase activity catalyzed by the native enzyme, are inactivated by incubation in air with relatively low concentrations (0.025 to 1 mM) of dithiothreitol; oxidized dithiothreitol does not inhibit. Other mercaptans (glutathione, Z-mercaptoethanol, dithioerythritol) also inhibit. Low concentrations of dithiothreitol under nitrogen and high concentrations (5 to 25 mu) of dithiothreitol in air do not inactivate the glutamine-related functions of the separated light subunit or of the native enzyme. However, enzyme preparations that have been inhibited by low concentrations of dithiothreito1 in air are not readily reactivated by treatment with high concentrations of dithiothreitol. Incubation of the native enzyme with hydrogen peroxide (0.2 mM) also inhibits the glutamine-dependent activities. In distinction, treatment of the enzyme with either hydrogen peroxide or low concentrations of dithiothreitol in air does not aBect either the ammonia-dependent synthetase activity or the synthesis of ATP from ADP and carbamyl phosphate, while the bicarbonate-dependent hydrolysis of ATP is stimulated moderately by both of these reagents. The addition of catalase as well as the addition of L-albizziin, L-glutamine, EDTA, or a mixture of ATP, magnesium ions, bicarbonate, and L-ornithine protect against inhibition of the glutamine-dependent synthetase activity by dithiothreitol. Neither L-ornithine (a positive allosteric effector) nor UMP (a negative allosteric effector) significantly affects the inhibition while a high protein concentration (10 mg per ml) protects against inhibition. The glutaminase activity of the separated light subunit, like that of the intact enzyme, is inhibited by the glutamine analog L-2-amino-4-oxo-5-chloropentanoic acid (chloroketone); L-glutamine or L-albizziin can protect against this inhibition. [i4C]Chloroketone binds to the dithiothreitol-inhibited enzyme to the same extent as it does to the native enzyme; such binding is reduced in the presence of glutamine. The most straightforward interpretation of the present results is that metal ion-catalyzed oxidation of

The glutaminase activity of glutamine-dependent carbamyl phosphate synthetase (Escherichia coli) and that of the separated light subunit of this enzyme, as well as the glutaminedependent synthetase activity catalyzed by the native enzyme, are inactivated by incubation in air with relatively low concentrations (0.025 to 1 mM) of dithiothreitol; oxidized dithiothreitol does not inhibit. Other mercaptans (glutathione, Z-mercaptoethanol, dithioerythritol) also inhibit. Low concentrations of dithiothreitol under nitrogen and high concentrations (5 to 25 mu) of dithiothreitol in air do not inactivate the glutamine-related functions of the separated light subunit or of the native enzyme. However, enzyme preparations that have been inhibited by low concentrations of dithiothrei-to1 in air are not readily reactivated by treatment with high concentrations of dithiothreitol. Incubation of the native enzyme with hydrogen peroxide (0.2 mM) also inhibits the glutamine-dependent activities. In distinction, treatment of the enzyme with either hydrogen peroxide or low concentrations of dithiothreitol in air does not aBect either the ammonia-dependent synthetase activity or the synthesis of ATP from ADP and carbamyl phosphate, while the bicarbonate-dependent hydrolysis of ATP is stimulated moderately by both of these reagents. The addition of catalase as well as the addition of L-albizziin, L-glutamine, EDTA, or a mixture of ATP, magnesium ions, bicarbonate, and L-ornithine protect against inhibition of the glutamine-dependent synthetase activity by dithiothreitol. Neither L-ornithine (a positive allosteric effector) nor UMP (a negative allosteric effector) significantly affects the inhibition while a high protein concentration (10 mg per ml) protects against inhibition. The glutaminase activity of the separated light subunit, like that of the intact enzyme, is inhibited by the glutamine analog L-2-amino-4-oxo-5-chloropentanoic acid (chloroketone); L-glutamine or L-albizziin can protect against this inhibition.
[i4C]Chloroketone binds to the dithiothreitol-inhibited enzyme to the same extent as it does to the native enzyme; such binding is reduced in the presence of glutamine. The most straightforward interpretation of the present results is that metal ion-catalyzed oxidation of *This work was supported in part by a grant from the National Institutes of Health, United States Public Health Service. dithiothreitol produces hydrogen peroxide which oxidizes one or more critical residues in the glutamine binding site of the light subunit. Such oxidation does not involve the sulfhydryl group which in the native enzyme reacts with chloroketone, nor does it involve a sulfhydryl group on the light subunit whose reaction with N-ethylmaleimide in the presence of ATP, magnesium ions, and bicarbonate, leads to loss of glutamine-dependent synthetase activity and to enhanced glutaminase activity.
The monomer (mol wt N 163,000) of glutamine-dependent carbamyl phosphate synthetase from Escherichia coli can be reversibly dissociated into a heavy and a light subunit in potassium thiocyanate with retention of catalytic activities (1,2). The separated heavy subunit can catalyze the synthesis of carbamyl phosphate from ammonia (but not from glutamine), the bicarbonate-dependent hydrolysis of ATP, and the synthesis of ATP from ADP and carbamyl phosphate.
The only enzymatic activity associated with the separated light subunit is the capacity to catalyze the hydrolysis of glutamine (or y-glutamyl hydroxamate). Treatment of the native enzyme with the glutamine analog, L-2-amino-4-oxo-5-chloropentanoic acid (3), leads to inactivation of the glutamine-dependent (but not the ammoniadependent) carbamyl phosphate synthetase activity, and also the glutaminase activity (4,5) ; such inactivation is associated with the covalent binding of the 4-oxonorvaline moiety of the analog to a sulfhydryl group at the glutamine binding site of the light subunit (6).
The present communication describes additional studies on the structure and function of the light subunit.
We have found that low concentrations of dithiothreitol lead to inactivation of the glutamine-related functions of the enzyme. This effect of dithiothreitol, which contrasts with its more usual action as a stabilizer of enzyme activity, was examined in detail. The experimental observations lead to the conclusion t.hat dithiothreitol in the presence of oxygen can produce hydrogen peroxide which oxidizes an amino acid residue (or residues) on the light subunit which participates in catalyzing the hydrolysis of glutamine.

Materials
The enzyme was isolated from E. coli B as previously described (7). The light subunit was isolated by Sephadex G-ZOO chromatography in potassium thiocyanate (8).
[aH]Dithiothreitol was a product of Calatomic. 2-Mercaptoethanol was obtained from Mann Research Laboratories, and glutathione from Calbiochem. Hydrogen peroxide (50%) was purchased from Fisher Scientific Co., and L-albizziin was purchased from Aldrich Chemical Co. Catalase (0.64 mg per ml; 59,000 units per ml) was obtained as a sterile solution from Worthington Chemical Co. L-2-Amino-4-oxo-5-chloro-[5-i4C]pentanoic acid was prepared as previously described (6). Nitrogen (prepurified grade) was obtained from Union Carbide Corporation.

Methods
Assays-Glutaminase activity was measured by incubating the enzyme at 37' for 10 to 30 min with 20 mM L-glutamine (native enzyme) or with 150 to 180 mM L-glutamine (light subunit) in 150 mM potassium phosphate buffer (pH 7.8) containing 0.5 mM EDTA.

Inhibition
of Glutaminase Activity of Separated Light Subunit by Dithiothreitol and by L-2-Amino-.&oxo-5-chloropentanoic Acid-When the separated light subunit was preincubated with relatively low concentrations of dithiothreitol (0.025 to 1 mM), there was dramatic inhibition of glutaminase activity (Fig. 1). However, at higher concentrations of dithiothreitol (5 to 25 mM), activation occurred which leveled off at about twice the initial activity.
Such activation seems to reflect the reversal of oxidative changes which may have occurred during preparation of the light subunit in potassium thiocyanate.
A similar interpretation may explain observations on the native enzyme in which it was found that high concentrations of dithiothreitol also activated, but only to the extent of 20 to 25% (see below).
Inhibition of the glutaminase activity of the separated light subunit was also observed after preincubation with 0.1 mM L-2amino-4-oxo-5.chloropentanoic acid (chloroketone). This inhibition, similar to that found after preincubation of the native enzyme with the chloroketone (5), was markedly reduced when either L-glutamine or n-albizziinr was added during preincubation. Although the glutaminase activity of the separated light subunit 1 Schroeder et al. (12) were apparently the first to use albizziin as a glutamine analog; their studies showed that albizziin inhibited formylglycinamide ribonucleotide amidotransferase. After incubation for 30 min at 37" the reaction was stopped by adding 0.05 ml of 1 N HCI and placing the acidified solution at 0". After 10 min, the solution was neutralized by adding 0.05 ml of 1 M Tris, and the amount of glutamate formed was measured by the DPN-glutamate dehydrogenase assay system (8,9). may be inhibited by preincubation with either the chloroketone or low concentrations of dithiothreitol, the data (see below) indicate that these reagents inhibit by different mechanisms.* Eflect oj Dithiothreitol on Various Enzymatic Activities Exhibited by Native Enzyme-In previous studies it was found that treatment of the enzyme with n-2-amino-4-oxo-5-chloropentanoic acid led to virtually complete inhibition of carbamyl phosphate synthesis from glutamine (but not from ammonia) and to a substantial increase in the carbon dioxide-dependent ATPase activity (3,6). Treatment of the enzyme with chloroketone greatly decreased its ability to hydrolyze both y-gluta.myl hydroxamate and glutamine, but did not affect the activity responsible for the synthesis of ATP from ADP and carbamyl phosphate.
In the present work it was found that incubation of the native enzyme with dithiothreitol led to effects which are similar to those observed after t,reatment of the enzyme with the chloroketone. AS indicated in Fig. 2, incubation of the native enzyme with low concentrations of dithiothreitol led to substantial inhibition of the glutaminase activity (as was observed in similar studies on the separated light subunit (Fig. 1)). Higher concentrations of dithiothreitol produced moderate activation of the glutaminase activity.
When the native enzyme was treated with 1.5 mM dithiothreitol for 16 hours at 23' there was marked inhibition of glutamine-dependent carbamyl phosphate synthetase activity, glutaminase activity, and ability of the enzyme to hydrolyze y-glutamyl hydroxamate (Fig. 3). On the other hand, there * It is of interest that the apparent K, value for n-glutamine for the separated light subunit is 130 to 180 mM, while the corresponding value for the native enzyme is about 1 mM (8). On the other hand, 0.1 mM chloroketone inhibits the separated light subunit and the native enzyme about equally well. This apparent difference between the binding of glutamine and that of the chloroketone may be due to the irreversible formation of a covalent linkage in the binding of the chloroketone. The remainder of the assay was performed as described in Fig. 1.
The effect of incubation of the enzyme with dithiothrei-to1 (DTT) and hydrogen peroxide on the various reactions catalyzed by carbamyl phosphate synthetase (CPSase). In the dithiothreitol studies, the enzyme (0.2 mg per ml) was allowed to stand at 23" for 16 hours in a solution containing 150 mM potassium phosphate buffer (pII 7.8) and 0.5 mM EDTA in the presence and in the absence of 1.5 rnM dithiothreitol.
In experiments with hydrogen peroxide, the enzyme (0.37 mg per ml) was incubated at 37" for 30 min in a solution containing 150 mM potassium phospliate buffer (pH 7.8) and 0.5 mM EDTA in the presence and absence of 0.2 mM hydrogen peroxide. The solutions were then placed at 0" and assays were performed for the various enzymatic activities using the procedures described above. The activity of the control (no dithiothreitol or no hydrogen peroxide) is arbitrarily set at 100% for each reaction. The effects produced by incubating the enzyme with chloroketone (3,4,6) are included for comparison.
was no change in the ammonia-dependent carbamyl phosphate synthetase activity or in the ability of the enzyme to catalyze the synthesis of ATP from carbamyl phosphate and ADP.
The treated enzyme exhibited about a 257, increase in bicarbonatedependent ATPase activity.
When the enzyme was incubated with 0.2 mM hydrogen peroxide, effects on the various activities of the enzyme were observed which were virtually identical to The effect of dithiothreitol on the activity during assay was found to be negligible.
those observed with dithiothreitol ( Fig. 3). Similar results were obtained when the enzyme was incubated with the chloroketone (3, 6); these are included in Fig. 3 for comparison.
As indicated in Fig. 4 (control), the glutamine-dependent carbamyl phosphate synthetase is lost rapidly when the native enzyme is incubated at 37" in the presence of 1 mM dithiothreitol. (It was previously observed that some preparations of the enzyme exhibited t,ransient moderate act,ivat.ion when studied under similar conditions (13) ; as stated above, such st.imulation of activity may reflect the reversal of sulfhydryl group oxidation which probably occurred during or after preparat.ion of the enzyme.) The addition of a low concentration of catalase virtually completely protected against inhibition. This result strongly implies the intermediate formation of hydrogen peroxide in the mechanism of inhibition.
The addition of L-albizziin also completely protected against inhibition.
Similarly,L-albizziin was found to protect against the inhibition observed upon the direct addition of hydrogen peroxide which is demonstrated in Fig. '3. Addition of glutamine decreased the rate at which inhibit.ion OCcurred, and when the enzyme was incubated with dithiothreitol in the presence of ATP, magnesium ions, and bicarbonate, there was some initial stimulation of activit.y followed by inhibition. It is notable that ATP, magnesium ions, and bicarbonate also protect against chloroketone inhibition of the enzyme (3, 6). Neither the addition of L-ornithine (a positive effector) nor of UNP (a negative effector) significantly affected the rate of inhibition.
No inhibition was observed in other experiments in which the dithiothreitol was first oxidized by prolonged incubation at 37" in the presence of oxygen prior to incubation with the enzyme.
When the enzyme concentration was increased to 10 mg per ml, the addition of 1.0 mM dithiothreitol did not inhibit the glutamine-dependent synthetase activity. One interpretation of this finding is that a high protein concentration may remove traces of metal ions which can cat,alyze the production of hydrogen peroxide.
This hypothesis is supported by the observation that an increase of the EDTA concentration from 0.5 to 50 rniM completely protects against dithiothreitol inhibition (Fig. 4). This result must reflect an interference with the production of hydrogen peroxide since the same concentrat.ion of EDTA has no effect on the inhibit,ion if hydrogen peroxide is added directly to the enzyme.
Studies with other mercaptans gave substantially similar results. As indicated in Fig. 5, glutathione inhibited after about 100 min of incubation.
2-Mercaptoethanol and dithioerythritol also inhibited; the time-course with dithioeryt.hritol u'as about the same as observed with dithiothreitol.
Experiments in which the native enzyme was incubated with dithiothreitol at several values of pH are described in Fig, 6. It is notable that inhibition was much greater at pH 8.6 than at pH 7.8, and that little inhibition occurred at pH 6.8. In distinction, addition of hydrogen peroxide readily inhibits at pH 6.8. Thus, the failure of dithiothreitol to inhibit at pH 6.8 is probably related to insufficient production of hydrogen peroxide rather than to a titration of residues on the protein.
E$ect of Exclusion of Air on Inactivation of Glutamine-dependent Carbamyl Phosphate Synthetase Activity--When dithiothreitol alone was incubated at 37" at pH 7.8 under the conditions used in the studies on the enzyme, there was a marked decrease in its sulfhydryl titer (Fig. 7, inset); when an effort was made to exclude air by flushing with nitrogen, the sulfhydryl titer was lost much less rapidly.
This oxidation is presumably associated with the formation of hydrogen peroxide.
In a comparable experiment in which the enzyme was incubated with 1 mM dithiothreitol under nitrogen, the glutamine-dependent enzymatic activity was affected only slightly, while a marked reduction in activity was observed when air was not excluded (Fig. 7). In  Enzyme (0.53 mg per ml in 150 mM potassium phosphate buffer (pH 7.8) containing 0.5 mM EDTA) was incubated with 1.0 mM dithiothreitol at 37'. The buffer was previously flushed with moist nitrogen gas and was frequently flushed with nitrogen throughout incubation.
A control experiment was also carried out in air (02). Aliquots (0.02 ml) were assayed for glutamine-dependent carbamyl phosphate synthetase. Inset, effect of incubation at 37" on the sulfhydryl content of dithiothreitol solutions in the presence and absence of air. Dithiothreitol (1.0 mM) in 150 mM pot.assium phosphate buffer (pH 7.8) containing 0.5 mM EDTA was incubated at 37". A second sample was treated in the same way except that the solution was flushed with moist nitrogen during incubation.
Aliquots (0.01 ml) were withdrawn at intervals and the sulfhydryl content was determined by titration with 5,5'dithiobis(2-nitrobenzoic acid) according to Ellman (11). another experiment, in which the enzyme was treated with 5 mM dithiothreitol for 150 min under nitrogen and then exhaustively dialyzed under nitrogen to remove the dithiothreitol, virtually no glutamine-dependent carbamyl phosphate synthetase activity was lost. Exposure of the enzyme to air at this point did not lead to inhibition, but addition of 1 mM dithiothreitol in the presence of air produced marked inhibition. The findings thus indica.te that inhibition of the enzyme requires the simultaneous presence of both molecular oxygen and dithiothreitol. This result is consistent with the conclusion that it is the formation of hydrogen peroxide from the reaction between molecular oxygen and dithiothreitol which causes the inhibition.
The data given in Fig. 2 show that enzymatic activity is maintained in the presence of high concentrations of dit.hiothreitol, while inactivation occurs with low concentrations of dithiothreitol. Experiments were carried out to determine whether or not the inhibition produced by low concentrations of dithiothreitol is reversible. Thus, an enzyme preparation was inhibited by incubation with 1 mM dithiothreitol in the presence of air; about i'5yo inhibition was observed in this study (Fig. 8). When a high concentration (60 mM) of dit.hiothreitol was then added, there was only a small increase in enzymatic activity, indicating that the dithiothreitol-induced inhibit.ion of . the enzyme cannot be readily reversed (although it can be prevented; see  The enzyme (1.3 mg per ml) was incubated with 2 mM dithiothreitol in 150 mu potassium phosphate buffer (pH 7.8) containing 0.5 rnM EDTA for 4 hours at 37" and then dialyzed overnight against two changes of 500 volumes of the same buffer to remove dithiothreitol; the glutaminase activity was completely inhibited.
The inhibited enzyme (Experiments 1 and 2) and the native enzyme (Experiments 3 and 4) were then incubated with ["Clchloroketone (0.2 mM) for 90 min at 37" in the presence and absence of 100 mhl: L-glutamine, and the binding of "C was determined as described (6).

Conditions 1
Dithiothreitol-inactivated enzyme 2 Dithiothreitol-inactivated enzyme + glutamine 3 Native enzyme 4 Native enzyme + glutamine Chlorokefone-In order to determine whether or not the inhibition of the glutamine-dependent carbamyl phosphate synt.hetase activity by dithiothreitol involves the sulfhydryl group that is modified by treatment of the enzyme with t.he chloroketone (6), an esperiment was performed in which the native enzyme and the dithiothreitol-treated enzyme were treated with [YJchloroketone in the presence and absence of glutamine (Table I). The data show that. both enzyme preparations bound equivalent amounts of labeled chloroketone; in addition, glutamine protected each of the enzyme preparations against the binding of about 1 mole of chloroketone per mole of protein. These findings indicate that inhibition of the enzyme by dithiothreitol does not involve oxidation of the sulfhydryl group t.hat interacts with the chloroketone and suggest that the inhibition by dithiothreitol is related to an oxidation of another group (or groups) which may be involved in catalysis rather than in the binding of glutamine or of the chloroket.one.
It was of interest to ascert.ain whether the sulfhydryl group modified by the chloroketone or the oxidation associated with inactivation by dithiothreitol was related to the sulfhydryl group reported by Foley and co-workers (15) to react wit.h N-ethylmaleimide in the presence of ATP, magnesium chloride, and bicarbonate; react.ion of this sulfhydryl group (located on the light subunit) is associated with substantial loss of the glutaminedependent synthetase activity. In contrast, the glutaminase activity of N-ethylmaleimide-treated enzyme is greatly stimulated. This enhanced glutaminase act.ivity can be completely abolished by incubation with 0.25 mM chloroketone, and such inhibition is virtually completely prevented by 100 mM L-albizziin (5). 3 We have also found that incubation with 1 mM dithiothreitol in air inactivated the enhanced glutaminase activity of the N-ethylmaleimide-treated enzyme. The findings thus indicate that the glutamine-dependent synthetase activity of the enzyme can be inhibited by three separat.e modifications of the light subunit.4 a V. P. Wellner and A. Meister, manuscript in preparation.

Incorporation of [aH]
Dithiothreitol into Enzyme-In order to determine whether inactivation of the enzyme is associated with the binding of dithiothreitol, t.he enzyme was incubated with 1 rnivr [VI]dithiothreitol.
After complete inactivation, the enzyme was separated from unreacted reagent by dialysis and found to contain the equivalent of about 0.6 mole of [3H]dithiothreitol per mole of enzyme. However, when the same experiment was performed in the presence of 100 mM L-albizziin, which completely protected against inactivation, the incorporation of tritium into the enzyme was about the same (about 0.7 mole per mole of enzyme).
The findings indicate that incorporation of [3H]dithiothreitol into the enzyme occurs at a site (or sites) different from the glutamine binding site; such incorporation is apparently not directly related to the dithiothreitol-induced inactivation of glutamine-dependent activities. DISCUSSION The findings indicate that the glutamine-related activities exhibited by carbamyl phosphate synthetase (glutaminase, hydrolysis of y-glutamyl hydroxamate, and glutamine-dependent carbamyl phosphate synthesis) are inhibited by low concentrations of dithiothreitol or other mercaptans in the presence of air. This effect of dithiothreitol contrasts with its more usual role as a stabilizer of enzymatic activity.
The ammonia-dependent carbamyl phosphate synthetase activity and the ability of the enzyme to synthesize ATP from carbamyl phosphate and ADP are unaffected while the bicarbonate-dependent ATPase activity is moderately activated.
The glutaminase act,ivity of the isolated light subunit is also subject to t,he same type of inhibition by dithiothreitol observed in studies on the native enzyme. Thus, it may be concluded that the event.s associated with inhibition by dithiothreitol occur exclusively within the light subunit; the present findings offer no evidence for the formation of intersubunit or intermolecular disulfide bonds. The effect of catalase in abolishing the inhibition by dithiothreitol indicates that hydrogen peroxide is the actual oxidizing agents: 6 It is well known that hydrogen peroxide is formed in the autoxidation of thiols (18). That the oxidation of dithiothreitol produces hydrogen peroxide under conditions similar to those used in the studies on carbamyl phosphate synthetase reported here was demonstrated by use of the fluorometric assay for hydrogen peroxide described by Lichtenberg and Wellner (19). Thus, a solution containing 1 mM dithiothreitol, 150 mM potassium phosphate (pH 7.8), 0.5 mM EDTA, 1 MM scopoletin, and horseradish peroxidase (0.2 mg per ml) was placed at 26"; after 20 min more than 80% of the fluorescence due to scopoletin had disappeared.
When this experiment was carried out in the presence of catalase (0.03 mg per ml), only about 40y0 of the scopoletin was oxidized. These findings strongly support the conclusion that hydrogen peroxide is formed in the oxidation of dithiothreitol. mixed disulfides between dithiothreitol and sulfhydryl groups on the protein. ) Consistent with this concept is the finding that addition of hydrogen peroxide to the enzyme duplicates all of the inhibitory effects observed with dithiothreitol.
Thus, the failure of high concentrations of dithiot,hreitol to inhibit can be explained by preferential reaction of the hydrogen peroxide formed with the excess dithiot.hreitol present rather than with a residue on the protein.
Another interpretation is that high concentrations of dithiothreitol removes traces of metal ions, which can catalyze the autoxidation of thiols (18). The participation of metal ions in the oxidation of dithiothreitol reported here seems highly probable since EDTA or high protein concentrations, both of which can remove traces of metal ions, offer virtually complete protection against inhibition by dithiothreitol. The conclusions outlined above are consistent with the findings and interpretations of Cavallini et d. (20), who demonstrated that the luminol chemiluminescence excited by autoxidizing thiols (in alkaline solution in the presence of Cu2+) is related to the formation of hydrogen peroxide.
Hydrogen peroxide is a relatively nonspecific oxidizing agent; sulfur-containing amino acids are possible candidates for its site of reaction (21) but others (e.g. histidine) must be considered. The nature of the modified residue is currently being studied. The data presented here indicate that the group oxidized in the presence of dithiothreitol is neither the sulfhydryl group modified by the chloroketone (6) nor the sulfhydryl group which reacts with N-ethylmaleimide in the presence of ATP, magnesium ions, and bicarbonate (15). Whatever the nature of the group oxidized in the presence of dithiothreitol, it appears to be involved in catalysis rather than in the noncovalent binding of glutamine to the enzyme. This conclusion is consistent with the data which indicate that both glutamine and chloroketone can bind to the dithiothreitol-inhibited enzyme (Table I). An alternative hypothesis which must be considered is that the group (or groups) oxidized does not participate directly in catalysis but that the oxidation caused by dithiothreitol or hydrogen peroxide triggers a conformational change which renders the catalytic site inactive. The fact that glutamine protects against inhibition would then imply that there must also be a glutamine-induced conformational change.
It is of interest that the conditions employed here in the reaction with hydrogen peroxide (0.2 mM; 37"; 30 min at pH 7.8) are much milder than those previously employed for the modification of enzymes and polypeptides (22). This may indicate that the oxidative reaction with carbamyl phosphate synthetase may be highly seIective for a critica residue at the glutamine binding site. It is notable also that the oxidation reaction is not readily reversible by incubation with a high concentration of dithiothreitol. However, in those cases in which reversibility by thiols has been observed (21, 22), more stringent conditions for reduction have generally been employed.
It is tempting to speculate that other glutamine amidotransferases may be susceptible to oxidation of the type found here for carbamyl phosphate synthetase.
Indeed, the present work raises the question as to whether use of dithiothreitol (or other thiols) may produce inhibition of other enzymes by generating hydrogen peroxide and suggests that indiscriminate use of such thiols for enzyme stabilization may be inadvisable.
As noted previously (13) earlier observations on glutamine amidotransferases in which unexpected variability was found in the relative rates of reaction with ammonia and glutamine may have been due to effects, such as those of dithiothreitol described here, on the glutamine binding site.
Note Added in Proof-While this paper was in press, it was reported (23) that prevention of a reaction by catalase is not necessarily evidence for the formation of hydrogen peroxide since certain catalase preparations are contaminated with superoxide dismutase.
Indeed we have confirmed this observation in studies on a number of commercial cat&se preparations, including the Worthington sterile catalase preparation used in the studies reported here. However, when we separated the superoxide dismutase from catalase by Sephadex G-100 chromatography, this purified catalase also completely protected against inhibition of glutamine-dependent carbamyl phosphate synthetase by dithiothreitol.
On the other hand, purified superoxide dismutase (obtained from Truett Laboratories, Dallas, Texas) was found to have no effect on the inhibition by dithiothreitol. These findings show that it is hydrogen peroxide rather than superoxide ion which is produced in the oxidation of dithiothreitol. (The authors wish to acknowledge the valuable suggestions and advice on this problem given to us by Dr. I. Fridovich.)