proximate sulfhydryl groups in the acetylglutamate complex of rat carbamylphosphate synthetase i:...

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 241, No. 1, August 15, pp. 200-214, 1985

Proximate Sulfhydryl Groups in the Acetylglutamate Complex of Rat Carbamylphosphate Synthetase I: Their Reaction with the Affinity

Reagent 5’-p-Fluorosulfonylbenzoyladenosine’

MARGARET MARSHALL AND LEONARD A. FAHIEN’

Department of Pharmacology, University of Wisconsin Medical School, Madison, Wisconsin 53706

Received April 29, 1985

A preparation of rat carbamylphosphate synthetase I, isolated in the presence of antipain and stable without glycerol, has been used to investigate the effect of the allosteric activator, N-acetyl-L-glutamate (AcGlu), on the sulfhydryl chemistry of the enzyme. The enzyme * AcGlu complex was rapidly inactivated by several sulfhydryl group reagents and the ATP analog, 5’-p-fluorosulfonylbenzoyladenosine (FS02BzAdo), with the loss of two sulfhydryl groups per monomer. Inactivation was much slower without AcGlu, and ATP/Mp/K+ provided complete protection. Reaction with a 1.1 molar excess of 4,4’-dipyridyldisulfide resulted in an intramonomer disulfide bond between groups that are probably juxtaposed in the activated enzyme, because 1.1 equivalents of the vicinal dithiol reagent, phenylarsine oxide, eliminated the rapid reaction with the disulfide. Evidence is presented that the same disulfide bond was formed in the reactions with 5-thiocyano-2-nitrobenzoic acid and FSOzBzAdo. Inacti- vation by FSOzBzAdo was a pseudo-first-order reaction. The concentration dependence of the rate is consistent with the reaction proceeding through a noncovalent complex (Kt = 67 PM and /ca = 0.23 min-’ at pH 7.0, 30°C). Protection from FSOzBzAdo by ATP required Me in excess of ATP with KM~AT~ = 4.5 PM at saturating free Mg+ (0.1 M K+) and KMgz+ = 6.5 mM. KM$AT~ is close to Kd for the molecule of ATP that contributes the phosphoryl group of carbamylphosphate (H. B. Britton, V. Rubio, and S. Grisolia, (1979) Eur. J. Biochem. 102, 521-5301; KM,z+ agrees with the minimum value for the steady-state kinetic parameter, Ki,Mg~+, obtained under the same conditions. Dissociation constants for adenosine (320 PM), MgADP (110 PM) at 10 mM

Mg2f, and AcGlu (100 PM) were also estimated. o 19% Academic PWS, IW.

Carbamylphosphate synthetase I (EC 6.3.4.16) catalyzes the first step in the biosynthesis of urea (1):

AcClu

NH3 + HCO, + 2ATPe Mg2+,K+

NHzCOzPOi + 2ADP + Pi.

The rat enzyme has negligible activity in the absence of N-acetyl-L-glutamate

i This research was supported by National Insti- tutes of Health Grant Am 17587.

‘Author to whom all correspondence should be addressed.

(AcG~u)~ under the usual assay conditions. Recently, however, Rubio et al. (2) have found that it does have appreciable activity at much higher concentrations of ATP and K+. Also, they found that the slight

3 Abbreviations used: AeGlu, N-acetyl-L-glutamate; FS02BzAdo, 5’-pfluorosulfonylbenzoyladenosine; 4- PyS2, 4,4’-dithiodipyridine; NbSCN, 5-thiocyano-2- nitrobenzoic acid; Mops, 4-morpholinepropanesulfonic

acid, Bes, N,N-bis-(2-hydroxyethyl)-2-aminoethane- suifonic acid; I-PySH, 4-thiopyridone; NbSH, 5-thio- 2-nitrobenzoic acid; SDS, sodium dodecyl sulfate; EGTA, ethylene glycol bis(@aminoethyl ether)-N,iV’- tetraacetic acid; NbSz, 5,5’-dithiobis(2-nitrobenzoic acid).

0003-9X361/85 33.00 Copyright 0 1985 by Academic Press, Inc. All rights of reproduction in any form reserved.

200

PROXIMATE THIOL GROUPS IN CARBAMYLPHOSPHATE SYNTHETASE I 201

lag phase observed in the assay in the absence of AcGlu became much longer at 50 PM AcGlu and then decreased with increasing concentrations of AcGlu, as Guthijhrlein and Knappe (3) had previously reported. An allosteric model was proposed to account for these results in which an e:quilibrium between inactive (T) and active (R) conformations is slowly shifted to a catalytically more efficient conformation on binding of AeGlu to T (2). Consistent with this model, preincu- bation of thle enzyme with 10 mM AcGlu alone shortened the lag phase of an assay at 50 PM AcGlu (2). High (mM) concentrations of AcGlu have also been found to promote dissociation of the dimer (4), which is the predominant species of the rat enzyme above 0.2 mg/ml, and to increase the rate of inactivation of the rat enzyme by elastase (5). Binding of AcGlu at low (PM) concentrations is, however, completely dependent on ATP (6). The affinities of the two functionally distinct ATP sites have been estimated to differ by at least lo-fold in the rat enzyme at saturating AcGlu (‘7). Since binding of ATP to the high-affinity site appears to be independent of AcGlu (2), it is probably the binding of ATP to the low-affinity site that is synergistic with the binding of AcGlu. Alonso and Rubio (6) have proposed that, at least at low concentrations of AcGlu, binding of AcGlu to T. ATP precedes the shift to the active conformation.

We have found inactivation of rat carbamylphosphate synthetase by a variety of sulfhydr:yl group reagents to be considerably more rapid in the presence of AcGlu. In spite of the large number (20/ monomer) of cysteinyl residues, these reactions, as well as that with the ATP analog, 5’-p-fluorosulfonylbenzoyladenosine (FS02BzAd’o), were largely confined to two neighboring sulfhydryl groups in the AcGlu complex. These results are discussed in relation to the allosteric model.

MATERIALS AND METHODS

Reagents. 4:,4’-Dithiodipyridine (4-PySr) was obtained from Aldrich and purified as described (8). 5- Thiocyano-2-nitrobenzoic acid (NbSCN) was synthe-

sized (9). Iodoacetamide was recrystallized from 50% ethanol. The following reagents were purchased from the indicated sources: Ultrogel AcA 34 from LKB;

Bio-Gel P-6 from Bio-Rad; antipain and leupeptin

from Peptide Institutes, Inc. (Osaka, Japan); phenylarsine oxide (>98%) from Eastman-Kodak; ATP, special quality, from Boehringer-Mannheim; and all

other substrates and reagents, including FSOzBzAdo, from Sigma.

Buffer concentrations are given in normality, de-

fined as the concentration of the counterion; pH was measured at 25°C. Potassium 4-morpholinepropane-

sulfonic acid (Mops) was prepared from freshly made KOH and boiled to remove excess COZ.

Rat carbamylphosphate synthetase. The enzyme was isolated (400 mg/lOO g liver) from livers of rats fed a 50% casein diet for 10 days (10). Significant

modifications from published procedures (11-13) included preparation of mitochondria by a high-yield method (14), extraction of the enzyme from the

mitochondria with 0.2% Lubrol WX in 50 mN triethanolamine acetate, 0.5 mM EGTA (pH 7.5), and

addition of antipain or leupeptin (20 pg/ml) at each step of the purification. These proteolytic inhibitors,

but not phenylmethylsulfonylfluoride, prevented the rapid, irreversible inactivation that occurs in aqueous

extracts at pH near neutrality (4°C) (11) and eliminated the requirement for glycerol to stabilize the enzyme (3).

The purified enzyme had the same specific activity

(2.8 units/mg at 37°C) as those of the most active preparations yet reported (11, 12), when the latter were normalized to the same extinction coefficient,

and was homogeneous in polyacrylamide gel electrophoresis with buffers containing sodium dodecyl

sulfate (SDS) or AeGlu. It was stable in dilute buffers (pH 7.0) for several hours +AcGlu at 30°C and for several days at 25°C with NaNa as preser-

vative and without added thiols. It has been stored

in ammonium sulfate (3.0 M, pH 7.5) at 4°C for at least 6 months without loss of activity.

Assays. Carbamylphosphate synthetase was as-

sayed by conversion of carbamylphosphate to citrulline in 0.5 ml of a medium containing 5 mM ATP, 10 mM Mg(OAc)a, 5 mM AcGlu, 50 mM NH,HCOa satu- rated with COz, 50 mM triethanolamine acetate (pH 7.0), 2 mM ornithine-HCl, and 10 pg/ml ornithine

transcarbamylase from Streptococcus fwcalis (15). Citrulline was determined calorimetrically by a

modification of the method of Ceriotti and Gassaniga

(16) that measured lo-120 nmol per 0.5 ml of assay

medium. Incubations were for 10 min a 30°C. A unit is 1 pmol carbamylphosphate/min.

For determination of the kinetic parameters (Fig.

lo), the assay was modified to contain, in 1.0 ml, 50 mN KBes (pH 7.14), 50 mM KHCOa, 10 mM NH&l, 5

mM AeGlu, 0.50 mM phosphocreatine, 20 pg creatine kinase, ornithine and ornithine transcarbamylase as

202 MARSHALL AND FAHIEN

above, 10 pg carbamylphosphate synthetase, and various concentrations of ATP and Mg(OAc)a. At

0.1 mM MgATP, 2.5 mM Me, these concentrations of NH&l, AcGlu, and HCO.$ were found experimen-

tally to be saturating, and rates were linear with time (5-20 min). ATP was included in the citrulline standards to correct for the slight decrease it (>1

pmol) causes in the color reaction. The enzyme was incubated with AcGlu for 5 min before addition of

ATP. Concentrations of Mga+ were confirmed by titration with EDTA (17). Concentrations of Boeh- ringer-Mannheim special-quality NaaATP * 3 Hz0

were based on weight. Extinction coe$in’ents. A value (AgEm = 0.82) was

used for rat carbamylphosphate synthetase. This is

based on a determination of the protein concentration by the biuret reaction (18) with crystalline bovine

serum albumin (A$&‘&,, = 0.66) (19) and P-lactoglob- ulin (AZ:, = 0.97) (20) as standards, both of which

gave the same standard curve. The only published extinction coefficient (0.84) that is based on a reliable

value for the protein concentration (from the re- fractive index) was determined on a preparation with a considerably lower specific activity (21). Molar

concentrations of enzyme refer to the concentration of monomer (M, 160,000) (11).

The extinction coefficient for 4-thiopyridone (4-

PYW (~~nrn = 22,000 at pH 7.0) was determined from the reaction of excess 4-PySa with a limiting

concentration of glutathione, the absorbance having been corrected for that due to unreacted 4-PySa. It

is higher than the reported value (19,800) (22). In 6 M guanidinium chloride, where the absorption max- imum is at 329 nm, taaanm = 22,300 at pH 6.5. The

value for the dianion of 5-thio-2-nitrobenzoic acid

WbW (~~lznrn = 14,000 at pH 7.0), determined by

the same procedure, agrees with that reported by Riddles et aZ. (23) for the pure compound.

Determination of sulfhydryl groups. The enzyme and its derivatives were titrated in 6 M guanidinium chloride (pH 6.5) with 0.15-0.30 mM 4-PySa. Reaction mixtures containing FSOaBzAdo, or the alkylating agents, were diluted directly into the solution of 4-

Py&. Under these conditions there was no detectable reaction of the reagent with I-PySH, and the reaction of the denatured enzyme with 4-PyS2 was rapid

enough that further reaction with the diluted reagent

was insignificant. The absorbance at 329 nm has

been corrected for the absorbance of unreacted 4- PySz and for the slight absorbance of the reaction mixture.

Rates of reactions. All reactions were carried out at 30°C in 0.1 N K- or NaMops (pH 7.03) containing

0.1 mM EDTA, except where indicated otherwise.

Carbamylphosphate synthetase was dissolved (10 rl/mg) in 50 mN triethanolamine acetate (pH 7.5)

and incubated with 5 mM dithioerythritol for 30 min

at 25’C. This solution was then filtered through a

column of Bio-Gel P-6 that had been equilibrated with the Mops buffer.

Stock solutions of all the modifying reagents, except iodoacetamide, were prepared in dimethyl-

formamide. The concentration of this solvent in the reaction mixture did not exceed 0.2% except with FSOaBzAdo, where it was maintained at 1%. Under

the conditions employed (30°C pH 7.0, 1% dimethylformamide) the solubility of FSOaBzAdo exceeded

200 FM, and the enzyme (+AcGlu) was completely stable for at least 2 h. The same results were obtained with FSOaBzAdo in 2% dimethyl sulfoxide.

The enzyme was incubated ?AcGlu (2 mM) for 4

min at 30°C before the reagents were added; these conditions were found to be adequate for temperature equilibration and for maximal activation by AcGlu.

Reactions with 4-PyS2 and NbSCN were followed in

a Gilford 250 spectrophotometer at the appropriate wavelength with the cuvette holder thermostated at

30°C. Rates of inactivation at 30°C were determined by adding aliquots (lo-100 ~1) of the reaction mixture directly to the assay medium at timed intervals. The

combination of dilution and protection by ATP was found to prevent further inactivation by all of the reagents except N-ethylmaleimide; in this case, it

was necessary to include dithioerythritol (0.1 mM)

in the assay. Rates of reaction of glutathione with N-ethylma-

leimide, iodoacetamide, and FSOzBzAdo were determined by titrating aliquots of a reaction mixture

with 4-PySa at timed intervals.

RESULTS

4-P@,. Titration of carbamylphosphate synthetase with 4-PyS2 in 6 M guanidinium chloride at pH 6.5 gave 19.3 + 0.3 sulfhy- dry1 groups per monomer, in agreement with the cysteine content determined as cysteic acid by Clarke (24). In spite of this large number, the enzyme. AcGlu complex was inactivated by 4-PySz approximately in proportion to the molar ratio of 4-PySZ to enzyme (Fig. 1). With a 1.07 molar excess, inactivation and formation of 4-PySH followed essentially the same time course (Fig. 2), with an endpoint at 94% inactivation and 1.98 mol 4- PySH/mol enzyme. The stoichiometry of the reaction (1.85 mol 4-PySH/mol 4- Py&), which did not change over a lo- fold range of concentration of enzyme (l- 10 PM), requires that the reagent have been displaced from the intermediate mixed-disulfide by another sulfhydryl


0 , I I

02 04 06 OS IO 12

Mel Reagent/ Mel Enzyme

FIG. 1. Relat.ionship between inactivation and the

molar ratio of 4-PySZ or phenylarsine oxide to carbamylphosphate synthetase. The enzyme * AcGlu complex (2.5 ~81) in the NaMops buffer was incubated

with the reagent until the activity reached a constant value. 0, 4-Py!s2; A, phenylarsine oxide.

group on the enzyme. Both inactivation and formation of 4-PySH were much slower in the absence of AcGlu (50% inactivated in 16 min versus 50 s), and the endpoints were lower (64% inactivated and 1.62 mol 4-PySH/mol enzyme). Nei- ther ATP nor Me alone affected the reaction with the AcGlu complex appre- ciably, but a combination of ATP (1 mM)

and Me (10 mM) eliminated the initial, rapid form.ation of 4-PySH and reduced the rate of inactivation so that only 5% of the activity had been lost by the time the endpoint was reached in their absence (Fig. 2). Activity was completely recovered from the inactivated enzyme on incubation with dithioerythritol in a pseudo-first- order reaction (Fig. 3).

A disulfide derivative, prepared by reaction of the enzyme * AcGlu complex with 1.1 molar excess 4-Py&, was chromato- graphed on a column of Ultrogel AcA 34 under conditions where the native enzyme is largely monomeric. A small amount of protein (~5%) appeared at the position estimated for the dimer, but the bulk of it emerged at the same position as the control enzyme and later than catalase (Mr 230,000). Also, after the remaining sulfhydryl groups had been rapidly alkylated with IV-ethylmaleimide in 6 M guanidinium chloride, the derivative had the same mobility in SDS-gel electrophoresis (no thiol added) as control enzyme similarly alkylated. These results establish

that the disulfide bond was predominantly intramonomer.

Phenylarsine oxide. The rapid and quantitative formation of this disulfide bond suggests that these sulfhydryl groups are oriented in the AcGlu complex so as to favor disulfide formation rather than being reoriented subsequent to the initial reaction with 4-PySZ. To test this conclu- sion, we investigated the reaction of the enzyme - AcGlu complex with phenylarsine oxide, a reagent that reacts preferentially with dithiols that are capable of forming cyclic dithioarsenite complexes (25). The same results were obtained as with 4- PyS,: (a) inactivation was approximately proportional to the molar ratio of reagent to enzyme (Fig. 1); (b) the enzyme was 94% inactivated by a 1.1 molar excess of reagent in a time-dependent reaction (50% complete in 2.5 min); (c) ATP (1 mM)/

al 20 E 2 c 16

W

$ 12

5 I? 0.8

I * 0.4

0 0 8 16 24 32 40 48 56

Time (min)

FIG. 2. Time course of inactivation (A) and of 4-

PySH formation (B) for the reaction of 4-Py& with carbamylphosphate synthetase. Enzyme in KMops buffer containing the indicated ligands was incubated with a 1.07 molar excess of 4-Py&. Experimental

details are given under Materials and Methods. 0, 2 mM AcGlu (AGA); Cl, no AcGlu; A, 2 mM AcGlu, 1 mM ATP, 10 mM Mg(OAc)p.


I I I 1 I I I I I I I I I 0 IO 20 30 40 50 60 70 00 90 too II0 120

Time (min)

FIG. 3. Comparison of the time course of recovery of activity from carbamylphosphate synthetase inactivated by 4-PySz, NbSCN, or FSOzBzAdo. Derivatives were prepared by incubating the

enzyme 9 AcGlu complex (6.46 PM) in 1.0 ml of 0.1 N KMops, 10 mM MgClz, 1% dimethylformamide (pH 7.03) with ‘7.18 MM I-PySp, 100 pM NbSCN, or 50 FM FSOzBzAdo for 30 min at 30°C. An aliquot (10 ~1) was removed for assay, then 50 ~1 of a solution containing 210 mM dithioerythritol and 21 mM ATP (to prevent further inactivation) was added, and the incubation was continued

at 30°C. Control enzyme was treated similarly. The solid lines were calculated for a first-order reaction with a half-life of 8.5 min. 0, 4-PySz; Cl, NbSCN, A, FS02BzAdo.

Mg+ (10 InM) greatly reduced the rate of inactivation (5% inactivated in the 30 min required to reach the end point in their absence); (d) inactivation in the absence of AcGlu was so slow (67% in 2 h) that it was not feasible to determine the endpoint. Furthermore, prior reaction with phenylarsine oxide (1.1 molar excess) eliminated the rapid reaction with 4-PySz (Fig. 4). Activity was rapidly (90% in 2 min) and completely recovered from this derivative on incubation with dithioerythritol (14 IrIM). Only a cyclic complex would be expected to be stable at these low concentrations (26): for example, a mixture of 19 PM phenylarsine oxide with 1’7 PM

2,3-dimercaptopropanol reacted extremely slowly with 4-PySz at pH 7.0, whereas a seven times molar excess of phenylarsine oxide over glutathione (19 PM) reduced the rate of its reaction with 4-PySz by

only one-half, with no change in the endpoint.

NbscN. We attempted to cyanylate these sulfhydryl groups with NbSCN (100 PM) to test whether this small substituent causes inactivation and to provide a means for locating these cysteinyl residues in the sequence (27, 28). When the slow reaction in the presence of ATP is subtracted (Fig. 5A), the difference conforms to a pseudo-first-order reaction with an endpoint (mol NbSH/mol enzyme) of 0.91 and a rate (0.19 min-‘) close to that for the reaction of glutathione with NbSCN (Ta- ble I). The enzyme was 92% inactivated under these conditions (Fig. 5B). A plot of log& - E,)/(& - E,) is linear and gives a rate (0.15 min-‘) somewhat less than that for the appearance of NbSH. Both the formation of NbSH and the inactivation were much slower in the ab-


5 10 15 20 25 30 35 40 Time (mid

FIG. 4. Reaction of the derivatives of carbamyl-

phosphate synthetase with 4-PyS,. The derivatives were prepared on separate occasions, and their reactions with 4-PyS2 were compared with that of

control enzyme at the same concentration, both in 2 mM AcGlu at 3O’C. Only the control for the FSOzBzAdo derivative is shown. 4-Py& (1.1 molar

excess) was added directly to the reaction mixture. The derivatives were prepared in the NaMops buffer containing AclGlu (2 mM) at 30°C as follows: 0, 50

PM FSOZBzAdo and 5.0 NM enzyme incubated for 45 min; A, 8.6 1tM phenylarsine oxide and 7.85 pM

enzyme incubated for 30 min; A, 20 pM N-ethylmaleimide and 7.85 ELM enzyme incubated for 10 min; 0,

control.

sence of AcGlu, and 0.25 mM ATP/lO mM MS+ provided complete protection. Since recovery o:f activity on incubation with dithioerythritol followed approximately the same time course as that for reacti- vation of the disulfide derivative prepared with 4-PySa (Fig. 3), the derivative must have been predominantly the disulfide.

AlkyJating agents. N-Ethylmaleimide and iodoacetamide also inactivated the enzyme * AcGlu complex. Sulfhydryl groups (2.0 + 0.2) (Table II) were lost on complete inactivation; their alkylation eliminated the rapid phase of the reaction with a 1.1 molar excess of 4-PySZ (Fig. 4). Again, inactivation was much slower in the absence of AcGlu. ATP (2 mM)/M$+ (10 mM) provided complete protection (Fig. 6) and prevented the loss of both sulfhydryl groups (Table II). If, as these results suggest, the same sulfhydryl groups that formed a disulfide bond on reaction with

4-PySZ are the ones that were alkylated, then prior reaction with 4-PySz should protect the enzyme from irreversible inactivation by the alkylating agents. To test this prediction, the enzyme + AcGlu complex was converted to the disulfide derivative (E/E0 = 0.05), incubated with 5 mM iodoacetamide for 18 min at 30°C (E/E0 = 0.17 for the non-PySz-treated control), and subsequently reactivated by treatment with 10 mM dithioerythritol in the presence of ATP/Me for 60 min, conditions adequate to completely reacti- vate the PySz-treated control. Recovery of activity, though appreciable, was incomplete (WE,, = 0.6). Possibly the reduced enzyme was not completely pro- tected from alkylation during the final 60- min incubation.

The time course of inactivation of the enzyme * AcGlu complex by both alkylating agents was biphasic (Fig. 6). The kinetics can be fitted to either of two mathematically equivalent models: (a) The two sulfhydryl groups react independently; alkylation of the more rapidly reacting one causes only partial loss of activity. (b) Alkylation of SHi results in complete loss of activity; alkylation of SH2 does not affect the activity but interferes with reaction with SH1. The relative values of the rate constants calculated for model (b) (see Fig. 6) were about the same for both reagents, although the values for the reaction with N-ethylmaleimide (20 PM) were three times those for the reaction with iodoacetamide (5 mM). The second-order rate constants calculated from kl are close to those for the alkylation of the sulfhydryl group of glutathione (Table I).

FS02BzAdo. This reagent was tested as a potential affinity label for the ATP sites (30). It rapidly inactivated the enzyme * AcGlu complex (Fig. 7A).4 The re-

’ The failure of the lines in Fig. 8A to extrapolate

to log (10 X E/E,) = 1.0 at the higher concentrations of reagent was not always observed, and is probably an artifaet caused by the difficulty of timing the addition of the reagent accurately enough for these fast reactions. Inhibition by the reagent (~4 pM)

present during the assay was ruled out.

206 MARSHALL AND

w I 0

1.1 u

1.0 +a W

0.9 go.6 4

x

2 5: w

0.7 51

z g =

0.6 1

1 -5 2 0.5 Z z

a 0.4 -6 I

0.3 Y W x

0.2 8 - 0.1 2

3

4 6 12 I6 20 24 26 32 Time (min)

FAHIEN FAHIEN

2.0 2.0

I.9 1.9

I8 I6

I7 17

I.6 I.6

I5 I5

14 14

I3 13

I2 12

I I I I

IO IO

09 09

OS OS 0 0 4 4 8 8 I2 I2 I6 I6 20 20 24 24 26 26 32 32

Time (min) Time (min)

FIG. 5. Time course of NbSH formation (A) and inactivation (B) for the reaction of NbSCN with carbamylphosphate synthetase. Enzyme (10.46 ELM in A; 2.4 jiM in B) in 0.1 N KMops, 10 mM MgClz (pH 7.03) was incubated with 100 ELM NbSCN with and without AcGlu and ATP. 0, 2 mM AcGlu; Cl, no AcGlu; A, 2 mM AcGlu + 5 mM ATP in (A) and 0.25 mM ATP in (B); 0 (A), reaction

in the presence of ATP subtracted; 0 (B), first-order plot of (E - E,)/(E, - E,).

action with 50 ~.LM FSOzBzAdo, which was followed until ~2% of the activity re- mained, was found to be accurately pseudo-first-order for at least four half- lives. The inactivated enzyme had 2.0 fewer sulfhydryl groups than the control

TABLE I

COMPARISON OF THE RATES OF REACTION OF SULF- HYDRYL REAGENTS WITH RAT CARBAMYLPHOSPHATE

SYNTHETASE AND GLUTATHIONE

k” (M-l min-‘)

Reagent

N-Ethylmaleimide

NbSCN Iodoacetamide FSOzBzAdo

Glutathione

-3.5 x 10’ 6.2 X 10’

19 12

Enzyme

2.4 X 10db 1.5 x 103

33* 3.4 x lo3

’ Rates were measured at pH 7.0 and 30°C. Values for the enzyme are rates of inactivation, which were assumed to be second-order reactions except in the case of FSOzBzAdo where k = k.JK,.

*The value given is calculated from k1 in model (b) (see Fig. 6).

(Table II) and reacted only slowly with a 1.1 molar excess of 4-PySz (Fig. 4). Activity (90%) was recovered on incubation with

TABLE II

SULFHYDRYL CONTENT OF RAT CARBAMYLPHOSPHATE SYNTHETASE INACTIVATED BY N-ETHYLMALEIMIDE,

IODOACETAMIDE, AND FSOzBzAdo

Experi-

mentO Reagent

SH groups

(mol/mol enzyme)

1 N-Ethylmaleimide 17.65

Control 19.9 2 Iodoacetamide 18.0

+2 mM ATP/lO mM 19.8

Mg(OAc)z Control 19.9

3 FSOzBzAdo 18.1 Control 20.1

‘The enzyme. AcGlu complex in KMops buffer

was incubated at 30°C with the indicated reagent for a sufficient time to inactivate it >90%. Experiment (1) 5 pM enzyme, 25 pM N-ethylmaleimide for 16 min; (2) 5 pM enzyme, 5 mM iodoacetamide for 25 min; (3) 11 pM enzyme, 50 pM FSOzBzAdo for 40 min

(six half-lives).


Time (mln)

dithioerythritol at the same rate as from the enzyme * AcGlu complex inactivated by 4-PySz (Fig. 3). We conclude that inactivation is again predominantly due to formation of the disulfide, in this case by displacement of the analog from an intermediate thiosulfonate (31-34). The first- order kinetics of inactivation require either that the intermediate be inactive or that disulfide formation not be rate limiting.

The hyperbolic dependence of the rate of inactivation on the concentration of FSOzBzAdo is consistent with the reaction proceeding through a noncovalent complex,

E+14 F E.I’Ei,. Dl

FIG. 6. TimIs course of inactivation of carbamyl- A double-reciprocal plot of khd versus the phosphate synthetase by iodoacetamide. The enzyme (3.8 pM) in KMops buffer containing the indicated

concentration of FSOzBzAdo (Fig. 7B)

ligands was incubated with 5 mM iodoacetamide. 0, gives & = 0.23 min-’ and KI = 67 ~.LM which,

2 mM AcGlu; 0, no AcGlu; A, 2 mM AcGlu, 2 mM in this case, can reasonably be expected

ATP, 10 mM Mg(OAc)z. The solid line in (0) was to equal k-Jkl. With Na+ in place of Kf,

calculated from the rate equation for model (b) (see KI (58 PM) was slightly lower. The high text) (29): E/E, = (ks - kl)e-(h+h)* - be-&/b - k, effective concentration of the bound re- - kz where kl (0.165 min-‘) is the rate of alkylation agent must account for the rapid rate of of SHi, b (0.0945 min-‘) that of SHa, and b (0.071 inactivation since rates of inactivation by min-‘) that of SHi following alkylation of SHz. the nonaffinity reagents discussed above

4 S 12 16 20 24 26 32 Time (min)

-20 -10 0 IO 20 30 40

[FS02BzAdoj’(mM)-’

FIG. 7. Inactivation of carbamylphosphate synthetase by FSOzBzAdo. Enzyme (2.5 PM), in the presence of AcGlu (2 mM) except where indicated, was incubated with FSOzBzAdo at the indicated concentrations in 0.1 N KMops (pH 7.03), 10 pM EGTA at 30°C. (A) Time course of inactivation; (B) double-reciprocal plot of the pseudo-first-order rate constants of inactivation versus the concentration of FSOzBzAdo. Rates measured in the presence of 10 mM MgClz (A) or in 0.1 N

NaMops (Cl) are also shown in (B).


are all ~3 times the rate of their reactions with glutathione (Table I).

MgATP (320 PM) and MgADP (3.2 mM),

in the presence of excess Me (10 mM),

and adenosine (10 m&r), even without Mgz+, provided complete protection from 50 PM

FSOzBzAdo. Rates of inactivation were measured at a single concentration of FS02BzAdo (-&I) and several concentrations of protecting ligand (Fig. 8A). A plot of the data according to

PI

is linear for all three ligands (Fig. 8B). The intercept on the abscissa gives

-Expp.~ [31

The value of KL for ATP (7.5 PM) was confirmed at three different concentrations of FSOzBzAdo (0.44, 1.1, and 1.75 KI) by measuring rates of inactivation _t ATP at a concentration equal to the calculated value of Khw,ATP. At each concentration,

the rate in the presence of ATP was reduced by 0.50 + 0.004, as required. K adenosine (320 I@ and KADP (110 @f) were calculated (Eq. [3]) from the apparent values obtained at 50 pM FSOzBzAdo.

The dissociation constants for MgATP and MgADP cited above are designated by a prime because they were obtained at a single concentration of free Me (10 mM). Since free M$+ appears to be an absolute requirement for protection by MgATP and M$+ alone does not affect the rate (Fig. 7B),

FMgATP = KM~ATP( 1 + &) [4]

should apply. K&,MgATp was determined at a single concentration of FSOzBzAdo (50 pM) and three concentrations of Mgz+ (Fig. 9). The association constant for MgATP under the conditions of the experiment (pH 7.0, 0.1 M K+) is about 1.5 X lo4 M-’ (35); thus ATP was >97% satu- rated at the lowest concentration of M8+. Since titration of 0.1 M Mops +- 0.1 M MgClz showed no change in its pK, we have assumed that binding of Mg2+ by the

4 07

zi x 0.6

0 = 05

g J 04

0.3

0.11 ’ I I I I I I I 0 5 IO 15 20 25 30 35 40

Time (mid [ATPI (JJ MI FIG. 8. Protection of carbamylphosphate synthetase by ATP from inactivation by FSOzBzAdo.

The enzyme (1.1 PM) was incubated in 0.1 N KMops (pH 7.03), 10 PM EGTA, 10 mM MgClz, 2 mM AcGlu with 50 PM FSOzBzAdo at 30°C. (A) Time course of inactivation at the indicated concentrations of ATP; (B) plot of the reciprocal of the pseudo-first-order rate constants versus the concentration of ATP.


[Mq2+];:ee,mM-’ I I I I I I I

-30 -20 -10 0 IO 20 30 40 50 60

[Mg ATP] (pM)

FIG. 9. Dependence of K&,MgATP on the concentration of free M$+. Rates (plotted as half-lives) of inactivation of the enzyme (2 pM) by FSO,BzAdo (50 PM) were measured at several concentrations of MgAT:P for each of three fixed concentrations of MgCla (2.5, 5.0, and 10 mM) in 0.1 N KMops (pH 7.03)1, 50 pM EDTA, 2 mM AcGlu. Inset is a plot of KhrnugAr.r versus the reciprocal of the concentration of free Mga+. See text for calculation of the concentration of free Mga+.

buffer was negligible. The concentration of free Mg’!’ has, however, been corrected for Kngcl+ (3.4 M-r) (36) and for binding by EDTA. Whether MgzATP is a significant species at these concentrations of M$+ is uncertain (35); it has been ignored in the present calculations. A plot of %~~,M~ATP versus the reciprocal of the concentration of free MgZ+ is linear and gives KM~z+ (6.5 mM) and Kapp,MgAp (7.4 PM) at saturating Mgz+ (Fig. 9). KMgATp (4.5 FM) can then be calculated from Eq. [31.

Mn2+ is effective at much lower concentrations tlhan Mgz+ in the synthesis of carbamylphosphate (37). It also increased the effectiveness of low concentrations of Mgz+ in supporting the protection by ATP (60 PM), 1 mM Mgzt, 0.1 mM Mn2+ being equivalent to 10 mM Me. Since MgATP constitutes 83% and MnATP only 12% of the total ATP at these concentrations, the role of Mm2+ is probably as a substitute for free M$+. Cerdan et al (38), measuring the binding of Mn2+ to rat carbamylphosphate synthetase directly by EPR under different conditions (37”C, pH 7.8, 30 FM enzyme; no AcGlu) from our experiments, found a single tight site per monomer and

estimated KMgz+ (180 PM) from displacement of Mn2+. This dissociation constant for Me was reported to agree with that determined kinetically. Because it is much less than the value from the protection experiments, we redetermined the steady-state kinetic parameters for Mgz+ and MgATP under the conditions of our experiments (0.1 M K+, 3O”C, pH 7.1), at saturating concentrations of NH:, HCO;, and AcGlu. As others have observed (38,39), plots of [MgATP]& versus [MgATP] at fixed free Mgz+ were concave upward at low [MgATP]. They were, however, linear to significantly lower concentrations than reported previously (38), 30 PM at 10 MM Me and 100 PM at 0.7 InM Me (not shown), and were not parallel, as required for ordered equilibrium binding of Mgz+ (39). Over the limited range of concentrations shown in Fig. 10, our data can be fitted to

(K”,’ + K*) A

K: +M+

K”,‘K”

> AM ’ [51


[ MgATP] (mM1

FIG. 10. Hanes plots of the synthetase activity with MgATP varied at several fixed concentrations of free Mg’+. Mg(OAc)a and a small amount of free

.

IQ?+ not at equilibrium; where A is [MgATP], M is [M$+], and KM is the dissociation constant of the enzyme. AcGlu . Me complex, which is not necessarily the same as KMgz+ for the binary enzyme + Mgz+ complex. Km,app (0.11 mM) for MgATP at 10 mM Mgz+ is considerably lower than the value (1.3 mM)

reported previously at 1.0 mM Mgz+ (38); the minimum value of KM (6.5 mM) for KA,‘PK$ agrees with that estimated in the protection experiments.

Rates of inactivation by FSOzBzAdo (50 PM) were also measured at various concentrations of AcGlu (0.05-5 mM). The reactions, which were followed for at least 2.5 half-lives, appeared to be pseudo-first- order at each concentration. The concentration (2 mM) routinely used was close to saturating. Inactivation also occurred in the absence of AcGlu at l/6 the rate

ATP, where necessary, were added to an equimolar

mixture of Mg(OAc)a and ATP to give the desired concentration of free Mga+. The free ATP required was calculated from the association constant for

MgATP at pH 7.1 and 0.1 M K+ (1.5 X lo4 M-') (34). The concentrations of free Me have been corrected

be first-order but is unlikely to have been due primarily to disulfide formation, because recovery of activity on incubation with dithioerythritol(10 mM) + AcGlu was incomplete and much slower than from enzyme inactivated in the presence of AcGlu. A plot of the data according to

at saturating AcGlu. It, too, appeared to

. . for bmdmg by acetate (KDI,A,+ + = 3.3 M-l) (40); binding by HCO; (KM,,,,; < 0.5 M-l) (41) and AcGlu

have been ignored, since it would reduce all concentrations approximately proportionally and by ~10%.

Binding of Mga+ by 0.5 mM phosphocreatine (KM,p.c, = 40 M-') (35) and by Bes (42) is negligible. Rates

were measured in the modified citrulline assay.

K&Z, and V,, at each fixed concentration of free Me were estimated from direct linear plots of the data according to Eisenthal and Cornish-Bowden

(43). The values of the kinetic parameters were then determined graphically from plots of [M]V&& versus

[Ml and DW&, V& versus [Ml. The lines in the

figure were calculated from Eq [5] with V= 0.93 min-’ mg-‘, KE = 0.30 mM, (Ke + K$) = 70 @M, and K$’ KM = 0.47 mM2. Concentrations of free

Mga+, reading from top to bottom, are 1.98, 2.94, 4.85, and 9.4 mM.

I?, - v- K = 1 + ,;;;;;; ) PI v+ - v-

where V, is the rate at saturating AcGlu and v+ and ?I- are the rates with and without AcGlu, is linear (Fig. 11) and gives Kam,Ac~,u (0.10 mM). Because the fraction of the enzyme containing bound FSOzBzAdo was not negligible in this experiment, Kam,AcGlu equals the dissociation constant for the free enzyme only if binding of FSOzBzAdo does not perturb binding of AcGlu (44). Also, Eq. [6] as- sumes that the binding steps equilibrate rapidly compared to the rate of covalent modification. Whether these assumptions are justified is considered below.

DISCUSSION

derived for the mechanism of Elliott and Under the conditions of the experiments Tipton (39) in which binding of Mg+ reported here, only two proximate sulfhy- precedes binding of ATPI, followed by dry1 groups were modified on inactivation HCO,, ATPz, and finally NH:, but with of the AcGlu complex of rat carbamyl-


- 3.2 I I I I

[AC Glu]-?rnM-‘)

FIG. 11. Dependence of the rate of inactivation by FSOzBzAdo on the concentration of AcGlu. Rates of inactivation of the enzyme (1.5 pM) by FSOzBzAdo

(50 pM) were measured in 0.1 N KMops, 0.1 mM

EDTA (pH 7.1).

phosphate synthetase by a variety of sulfhydryl ,group reagents. Since rates of inactivation were not exceptionally rapid (Table II), the specificity of these reactions must depend on the relative unreactivity or inaccessibility of the other 18 cysteinyl residues. A.dditional groups did react at higher molar ratios of 4-PySz to enzyme. Presumably these account for the reactions of th(e phenylarsine oxide and N- ethylmaleimide derivatives with a 1.1 molar excess of 4-PySz (Fig. 4). In the absence of AcGlu, t.he “nonessential” residues ap- pear to have competed effectively for a limiting amount of reagent, since inactivation by a slight molar excess of 4-PySz and phenylarsine oxide was less complete; both the stoichiometry of 4-PySH formation and the extent of inactivation can be account.ed for if 40% of the reaction occurred with “nonessential” residues to form mixed disulfides with PyS- and the remainder with the “essential” groups to form an intramolecular disulfide bond.

There are at least three possible expla- nations for the increase in the rates of reaction of these sulfhydryl groups in the presence of AcGlu. (i) AcGlu might bind adjacent to them and, thereby, increase their reactivity directly, possibly by low-

ering their pK, values. Although it is not obvious why binding of a divalent anion should have this effect, this possibility cannot be ruled out on the basis of the results presented here. (ii) If these sulfhy- dry1 groups are located close to the inter- face between the subunits of the dimer, then they might become more accessible on dissociation, which is promoted by AcGlu (4). Since, however, rates of inactivation by FSO$zAdo in the absence of AcGlu increased by only 10% with decrease in the concentration of enzyme over a range (0.5-0.1 mg/ml) where monomer becomes an appreciable portion of the total (4), it is unlikely that dissociation per se is responsible for the increased rates. (iii) The sulfhydryl groups might be accessible only in the R conformation postulated in the allosteric model (1) that was discussed in the Introduction. In this case, AcGlu would increase the rates of their reactions by shifting the conformational equilibria to the R form. Reaction of 4-PySz with the small amount of R in equilibrium with T in the free enzyme would account for the much slower formation of the disulfide in the absence of AcGlu. This model would also explain the increased (8-9 times) rates of inactivation of the enzyme by trypsin and elastase in the presence of AcG~u.~ In spite of their different specificities, inactivation by both proteases + AcGlu paralleled hydrolysis at sites approximately 180 residues from one end. Possibly this region is exposed in the R conformation. Changes caused by allosteric effecters in rates of modification of particular residues have been observed and attributed to conformational changes in several allosteric enzymes (45-48).

A value (67 PM) for KAcGlu was also estimated (with a different preparation of enzyme) from the pseudo-first-order rates of inactivation by elastase, measured at a concentration of elastase (3 pg/ml) where rates at saturating AcGlu were linear with concentration and under the conditions given in Fig. 11, except with 2.8 PM

5 M. Marshall and L. A. Fahien, unpublished experiments.


enzyme. In this reaction, carbamylphosphate synthetase was in large molar excess over the protease, and therefore binding of the protease should not have perturbed the binding of AcGlu. If the difference from the value (100 pM) obtained with FSOgBzAdo is significant, then the same assumption may not have been entirely justified for the reaction with FSOzBzAdo. Conformational changes are unlikely to have been limiting in the latter reaction, because rates with elastase were even faster (tl18 5 min versus 8.3 min at saturating AcGlu). These values for A’~~ol~ are not in accord with direct measurements of the binding. Alonso and Rubio (6) reported that binding of AcGlu, determined by flow dialysis at 23”C, was not detectable in the absence of ATP; and even in the presence of ATP/Mp/K+, they estimated a value for Kd of 0.3 mM. Their results in 7% glycerol, where Kd was much lower but still dependent on ATP, extrapolated to a single site per monomer. Kd may be lower at the higher temperature (30°C) and much lower concentration of enzyme used to measure rates of inactivation. Even at 1.5 pM,

however, the enzyme is probably largely dimeric (4). The apparent fit of the rates of inactivation to Eq. [6] may mask a more complex situation where rates for each subunit vary with the state of its neighbor.

The biphasic kinetics of inactivation by the alkylating agents indicates that only one of the sulfhydryl groups is “essential” in the sense that its modification results in complete (>9’7%) loss of activity. Nev- ertheless, ATP/M$+ prevented alkylation of both groups as well as providing complete protection from inactivation by all of the reagents. The question is whether both ATP sites need to be occupied for protection to be complete. Their dissociation constants have been estimated by Britton et al. (7) from pulse-chase experiments with [y-32P]ATP: under the conditions of their experiments (10 mM AcGlu, 18 mM MgZ+, 3’7 mM K+, 6.5% glycerol, pH 7.4, 22’C) Kd g 10 pM for the ATP that contributes the phosphoryl group of carbamylphosphate, and Kd > 0.2 mM (in-

creasing with increase in the enzyme concentration) for the ATP that contributes the Pi. These values provide the only means available for distinguishing between them. Although the complexity of the kinetics of inactivation by the nonaffinity reagents precluded a determination of Kd for the protection, the results with N-ethylmaleimide are consistent with ATP bound to the high-affinity site being primarily responsible: 0.25 mM ATP, 10 mM Mga+ provided complete protection and lower concentrations (25 and 70 PM)

largely eliminated the fast phase of the reaction and reduced the rate of the slow phase (0.22 min-‘) to 0.066 and 0.018 min-‘, respectively.

FSOzBzAdo fulfills the criteria for an affinity reagent for the high-affinity ATP site: (a) it inactivated the enzyme in a pseudo-first-order reaction with the rates hyperbolically dependent on the concentration of reagent; (b) inactivation, which was at least 90% due to reaction with the sulfhydryl groups, must have been facili- tated by its binding because the second- order rate constant (b/K*), relative to that for the reaction of FSOzBzAdo with glutathione, was loo-fold greater than those for the reactions with the nonaffinity reagents; (c) ATP was a competitive in- hibitor with a dissociation constant somewhat less than that estimated for the high-affinity site. Provided the competition was for the same site, as is reasonable for an ATP analog, then these results place the proximate sulfhydryl groups close to the active site. A similar value (5 pM at 10 mM Mgz+) has been determined for the dissociation constant of the ATP that provided complete protection from elastase and trypsin.5 Possibly binding of ATP to the high-affinity site induces a change to a conformation in which both the proximate sulfhydryl groups and the sites that are hydrolyzed are inaccessible.

Adenosine, which is a competitive in- hibitor of ATP in the synthetase reaction (39), also provided complete protection from FSOaBzAdo. Furthermore, it mark- edly reduced the rate of inactivation by the alkylating agents and by trypsin and elastase (4.4 times). Since the same kd


(300 PM) was calculated (Eq. [6]) for the incomplete protection from elastase as was found in the reaction with FS02BzAdo, binding of adenosine to the purine subsite of the high-affinity ATP site appears to be responsible for the protection in each case.

The reaction of rat carbamylphosphate with tritium-labeled FSOzBzAdo in the absence of AcGlu has been studied by Powers et ai. (49). The solvent used (10% dimethyl sulfoxide) may have effected either the con.formation of the reagent (50) or the enzyme because, in spite of the higher temperature (37°C) and pH (7.5), the times for 50% inactivation at 50 and 100 WM FSOzBzAdo were 2.5 times the half-lives under our conditions. Neverthe- less, our more limited results agree with their findings that 8.7 mM ATP, 10 mM

M$+ provided only partial protection and that treatment with dithioerythritol only partially reactivated the modified enzyme. We found aidenosine (6 mM) to be at least as effective as 5 mM ATP, 10 mM M$+ in protecting the enzyme; it reduced the rate of inactivation by 100 PM FSOzBzAdo 4.5 times. The specificity of the reaction with FSOzBzAdo in the AcGlu complex must be largely a consequence of the more rapid reaction of the sulfhydryl groups in the activated enzyme.

Earlier results of Novoa et al. (51) on the reaction of frog carbamylphosphate synthetase I with 5,5’-dithiobis(2-nitrobenzoic acid) (NbSz) suggest that proximate sulfh:ydryl groups close to an ATP site in the AcGlu complex may be a con- served feature of this enzyme. They found that AcGlu greatly increased both the rate of inactivation of this enzyme by a low molar excess of NbSz and the rate of NbSH formation, and that ATP, together with excess; M8+ and K+, provided nearly complete protection. They concluded that the enzyme had two “essential” sulfhydryl groups per monomer.

A comment is in order on the determination of iUi for Mg2+ from steady-state kinetic measurements. This requires knowl- edge of the kinetic mechanism. Cerdan et ah (38) ass,umed the ordered binding proposed by Elliott and Tipton (39) but with

binding of Mp and metal nucleotides all at equilibrium. In this case, rates at saturating HCO;, NH:, and AcGlu should be independent of the concentration of free Me over the range of concentrations of MgATP where double-reciprocal plots of v versus the concentration of MgATP are linear. Our results do not conform to this requirement nor do those for the bovine enzyme (39). Others (11, 39) have found the Kf term in Eq. [5] to be negligible. Possibly this difference from our results can be ascribed to the lower concentrations of free Mg2+ at which rates were measured by these investigators. We have not included measurements at free M$+ < 2.0 mM, because the uncomplexed ATP required to maintain a constant concentration of MgATP may be inhibitory, or at MgATP > free M2+ because, under these conditions, small errors in the concentration of ATP result in considerably larger errors in the concentration of free Mgz+.

1.

2.

3.

4. 5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

REFERENCES

METZENBERG, R. L., HALL, L. M., MARSHALL, M.,

AND COHEN, P. P. (1957) J. Biol. Chem 229,

1019-1025.

RUBIO, V., BRITTON, H. G., AND GRISOLIA, S.

(1983) Eur. J. Biochem. 134. 337-343. GUTH~HRLEIN, G., AND KNAPPE, J. (1968) Eur. J.

Biochem 7,119-127. LUSTY, C. 3. (1981) Biochemistry 20,3665-3674. GUADALAJARA, A. M., RUBIO, V., AND GRISOLIA,

S. (1983) B&hem. Biophys. Res. Commun. 117,

238-244. ALONSO, E., AND RIJBIO, V. (1983) Eur. J. Bioch,em.

135, 331-337. BRITTON, H. G., RUBIO, V., AND GRISOLIA, S.

(1979) Eur. J. Biochem 102, 521-530.

MARSHALL, M., AND COHEN, P. P. (1980) J. Biol. Chem. 255, 7291-7295.

DEGANI, Y., AND PATCHORNIK, A. (1971) J. Org.

Chem. 36, 2727-2728. SCHIMKE, R. T. (1962) J. Biol. Chem. 237, 459-

468.

LUSTY, C. J. (1978) Eur. J. Biochem. 85, 373-383. MCINTYRE, P., AND HOOGENRAAD, N. (1981) FEBS

Lett. 135, 65-69.

POWERS, S. G. (1981) J. Biol. Chem. 256, 11160-

11165.

BUSTAMANTE, E., SOPER, J. W., AND PEDERSEN,

P. L. (1977) Anal. B&hem. 80.401-408.


15. MARSHALL, M., AND COHEN, P. P. (1972) J. Biol Chem. 247, 1641-1653.

16. CERIOTTI, G., AND GASSANIGA, A. (1968) C&Z. Chim. Acta 14, 57-62.

17. SKOOG, D. A., AND WEST, D. M. (1969) in Fun- damentals of Analytical Chemistry, 2nd Ed., pp. 355-35’7, Holt, Reinhardt and Winston,

New York. 18. LAYNE, E. (1957) in Methods in Enzymology

(Colowick, S. P., and Kaplan, N. O., eds.), Vol.

3, pp. 450-451, Academic Press, New York. 19. WILLIAMS, E. J., AND FOSTER, J. F. (1959) J.

Amer. Chem. Sot. 81, 865-870.

20. WETLAUFER, D. B., AND LOUVRIEN, R. (1964) J. Biol. Chem. 239, 596-603.

21. VIRDEN, R. (1972) B&hem J. 127,503-508.

22. GRASSETTI, D. R., AND MURRAY, J. F. (1967) Arch. B&hem. Biophys. 119,41-49.

23. RIDDLES, P. W., BLAKELEY, R. L., AND ZERNER,

B. (1979) Anal. B&hem 94, 75-81. 24. CLARKE, S. (1976) J. Biol. Chem. 251,950-961. 25. WEBB, J. L. (1966) in Enzyme and Metabolic

Inhibitors, Vol. 3, pp. 599-656, Academic Press, New York.

26. LIU, T.-Y. (1978) in The Proteins (Neurath, H.,

and Hill, R. L., eds.), Vol. 3, pp. 298-302, Academic Press, New York.

27. DEGANI, Y., AND PATCHORNIK, A. (1974) Biochem-

istrg 13, l-11. 28. JACOBSON, G. R., SCHAFFER, M. H., STARK, G. R.,

AND VANAMAN, T. C. (1973) J. BioL Chem 248, 6583-6591.

29. RAY, W. J., AND KOSHLAND, D. E. (1961) J. Biol

Chem 236, 1973-1979. 30. WYATT, J. L., AND COLMAN, R. F. (1977) Biochem-

isby 16, 1333-1342. 31. ANNAMALAI, A. E., AND COLMAN, R. F. (1981) J.

Biol. Chem. 256, 10276-10283. 32. TOMICH, J. M., MARTI, C., AND COLMAN, R. F.

(1981) Biochemistry 20, 6711-6720. 33. TOGASHI, C. T., AND REISLER, E. (1982) J. Biol

Chem. 257, 10112-10118.

34. TAKATA, Y., AND FUJIOKA, M. (1984) Biochemists 23, 4357-4362.

35. O’SULLIVAN, W. J., AND SMITHERS, G. W. (1979)

in Methods in Enzymology (Purich, D. L., ed.),

Vol. 63, pp. 294-336, Academic Press, New York.

36. BLAIR, J., McD. (1970) Eur. J. B&hem. 13, 384-

390.

37. LUSTY, C. J. (1978) J. Biol. Chem. 253,4270-4278.

38. CERDAN, S., LUSTY, C. J., DAVIS, K. N., JACOBSON,

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

J. A., AND WILLIAMS, J. R. (1984) J. Biol Chem. 259, 323-331.

ELLIOTT, K. R. F., AND TIPTON, K. F. (1973)

Biochem J. 141, 807-816.

MARTELL, A. E., AND SMITH, R. M. (1977) in

Critical Stability Constants, Vol. 3, p. 3, Plenum, New York.

RAAFLAUB, J. (1960) Helv. Chim. Acta 43, 629- 634.

GOOD, N. E., WINGET, G. D., WINTER, W., CON-

NALLY, T. N., IZAWA, S., AND SINGH, R. M. M.

(1966) Biochemistry 5,467-477.

EISENTHAL, R., AND CORNISH-BOWDEN, A. (1974)

B&hem. J 139, 715-720.

HORIIKE, K., AND MCCORMICK, D. B. (1980) J.

Theor. Biol. 84, 691-708.

TERAOKA, H., IZUI, K., AND KATSUKI, H. (1974)

Biochemistry 13, 5121-5128.

MARCUS, F. (1976) FEB.? ktt. 70, 159-162.

WINSLOW, J. W. (1981) J. Biol. Chem. 256, 9522-

9531.

DREYFUS, M., VANDENBUNDER, B., AND But, H.

(1980) Biochemistry 19, 3634-3642.

POWERS, S. G., MULLER, G. W., AND KAFKA, N.

(1983) J. Biol Chem. 268, 7545-7549.

JACOBSON, M. A., AND COLMAN, R. F. (1984) J.

Biol. Chem. 259,1454-1460.

NOVOA, W. B., TIGIER, H. A., AND GRISOLIA, S.

(1966) Biochim. Biophys. Acta 113, 84-94.

proximate sulfhydryl groups in the acetylglutamate complex of rat carbamylphosphate synthetase i:...

Documents