Mercurial-promoted Zn2+ release from Escherichia coli aspartate transcarbamoylase.

The release of Zn2+ from aspartate transcarbamoylase (ATCase; c6r6) upon challenge by p-hydroxymercuriphenylsulfonate (PMPS) has been studied using the sensitive, high-affinity metallochromic indicator 4-(2-pyridylazo)resorcinol at pH 7.0. When the--SH group of each catalytic (c) chain is protected, 1 Zn2+ is released for every 4 eq of PMPS added to ATCase during titration of the 24--SH groups of regulatory (r) chains. Moreover, the release of Zn2+ is a linear function of PMPS added, indicating that the rate-limiting step in Zn2+ release is mercurial attack on the 1st of the 4 r--SH groups bonded tetrahedrally to Zn2+ in an r chain near c:r contacts. Dissociation of ATCase is linked to Zn2+ release and mercaptide formation; e.g. upon addition of 4 eq of PMPS to ATCase in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Hepes) buffer, 1/6th of ATCase is dissociated to c3 and r2 subunits at approximately 83% of the rate of Zn2+ release, with no accumulation of the c6r4 intermediate as is observed in KPO4 buffer. Adding less than or equal to 4 PMPS/ATCase, the release of Zn2+ is first-order in [PMPS] and is virtually independent of [ATCase] with an activation energy of 18 kcal/mol. With large excesses of PMPS, stopped-flow traces show a lag period followed by pseudo first-order release of Zn2+ from ATCase and the reaction order in [PMPS] = approximately 1.3. Under these conditions, PMPS has a chaotropic effect on ATCase; the activation energy for Zn2+ release is much lower than that obtained with limiting PMPS and is increased by the presence of phosphate or active-site ligand from 6.6 to approximately 12 kcal/mol. A reasonable explanation of the observed kinetic data is that the organomercurial reagent binds reversibly to nitrogenous side chain groups in an ATCase molecule prior to the rate-limiting reaction with a sulfhydryl group.

The release. of Zn2+ from aspartate transcarbamoylase (ATCase; cere) upon challenge by p-hydroxymercuriphenylsulfonate (PMPS) has been studied using the sensitive, high-affinity metallochromic indicator 4-(2-pyridy1azo)resorcinol at pH 7.0. When the -SH group of each catalytic (c) chain is protected, 1 Zn2+ is released for every 4 eq of PMPS added to ATCase during titration of the 24 "SH groups of regulatory (r) chains. Moreover, the release of Zn2+ is a linear function of PMPS added, indicating that the rate-limiting step in 2n2+ release is mercurial attack on the 1st of the 4 r-SH groups bonded tetrahedrally to Zn2+ in an r chain near c:r contacts. Dissociation of ATCase is linked to Zn2+ release and mercaptide formation; e.g. upon addition of 4 eq of PMPS to ATCase in 4-(2-hy-droxyethy1)-1-piperazineethanesulfonic acid (Hepes) buffer, %th of ATCase is dissociated to c3 and r2 subunits at -83% of the rate of Zn2+ release, with no accumulation of the car4 intermediate as is observed in KPO, buffer. Adding ~4 PMPS/ATCase, the release of Zn2+ is first-order in [PMPS] and is virtually independent of [ATCase] with an activation energy of 18 kcal/mol. With large excesses of PMPS, stopped-flow traces show a lag period followed by pseudo first-order release of Zn2+ from ATCase and the reaction order in [PMPS] = -1.3. Under these conditions, PMPS has a chaotropic effect on ATCase; the activation energy for Zn2+ release is much Iower than that obtained with limiting PMPS and is increased by the presence of phosphate or active-site ligand from 6.6 to -12 kcal/ mol. A reasonable explanation of the observed kinetic data is that the organomercurial reagent binds reversibly to nitrogenous side chain groups in an ATCase molecule prior to the rate-limiting reaction with a sulfhydryl group.
The regulation of ATCase' (carbamoy1phosphate:L-aspar-*This work was performed under Title IV, Intergovernmental Personnel Act. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' The abbreviations used are: ATCase, aspartate transcarbarnoy-tate carbamoyltransferase, EC 2.1.3.2) from Escherichia coli, which catalyzes the first committed step in pyrimidine biosynthesis, has been reviewed (1, 2). A goal of physical-chemical studies on ATCase is to understand the molecular forces involved in the transition of this enzyme from a low-affinity conformation to a more swollen, less constrained form having a high-affinity for substrates.
Each molecule of ATCase from E. coli contains 6 Zn2+ ions, which play an essential role in maintaining the quaternary structure of this allosteric enzyme (3-5). ATCase contains 6 c and 6 r chains (4, 6-8), structurally arranged as two superimposed C trimers (4, 6) bridged noncovalently by three R dimers (4, 8-10). Each Zn2+ ion is tetrahedrally bonded to all 4 cysteinyl residues of each r chain (residues 109, 114, 137, and 140) near the region of contact between r and c chains The early studies of Gerhart and Schachman (11,12) showed that treatment of ATCase with the organomercurial reagent PMB produces dissociation of the enzyme into C and R subunits. Displacement of the mercurial from unfractionated subunits by the addition of excess 2-mercaptoethanol resulted in reconstitution of ATCase (11). Whereas the Zn2+ in ATCase is not exchanged with added 'j5Zn2+ (3) or removed by chelators (3-4), EDTA will remove Zn2+ from isolated R subunits (3). Aporegulatory subunits do not associate with C trimers to form ATCase (3). However, Zn2+ in r chains has been replaced by Hg2+ (3), Cd2+ (3-4), Ni2+, and Coz+ (13) and these derivatives associate with C trimers to form stable ATCase molecules.
Active-site ligand binding produces a substantial swelling of ATCase as evidenced by decreases in the sedimentation and diffusion coefficients (12, 14-16), changes in low angle xray scattering (17) and x-ray crystallographic studies (9, 18). The sulfhydryl groups of r chains also are made more reactive to PMB by binding active-site ligands to ATCase (12,19), whereas the single "SH group of each c chain is protected by substrates and substrate analogs (2). In addition, activesite ligand binding weakens r:c contact regions in ATCase (20,211 and perturbs the metal ion environment in Ni2+, Go2+, and Cd" enzymes (13, 22).
Previous kinetic studies of the mercurial-promoted disso-(7,9). ciation of ATCase into C trimers and R dimers have shown that the rates of protein dissociation (12,23) and of organomercurial-mercaptide bond formation (19) are closely linked.
In the studies presented here, the release of Zn2+ during the organomercurial reaction with ATCase was monitored by using the very sensitive, high-affinity metallochromic indicator 4-(2-pyridylazo)resorcinol (24)(25)(26). The large change in exctinction coefficient at 500 nm that accompanies Zn2+ binding to this indicator has allowed studies of the reaction of ATCase with substoichiometric concentrations of the mercurial reagent PMPS. In this way it was found that 1 Zn2+ is released for every 4 eq of the mercurial added to the enzyme. Moreover, the use of the indicator PAR facilitated kinetic measurements of Znz+ release from ATCase, which were performed in the absence and presence of inhibitory buffers of mercurial-thiol reactions (27) such as Tris/HCl and KPO, previously used (12,19,23). Evidence is presented that phosphate causes the accumulation of the intermediate lacking one R dimer, C2R2, during the dissociation of ATCase by mercurial attack (23,28).

EXPERIMENTAL PROCEDURES
Materials--Several preparations of wild-type ATCase, purified by the method of Gerhart and Holoubek (29), and the mutant ATCaseasl, isolated and purified as described by Wall et al. (30), were prepared by Y. R. Yang at the University of California, Berkeley, and shipped at 0 "C as protein suspensions in 3.6 M ammonium sulfate containing 5 mM 2-mercaptoethanol. These proteins were stored at 4 "C as received and were collected as needed by centrifugation just prior to dialysis at 4 "C for -2 days against a t least three changes of 100 volumes of either Chelex-treated 40 mM Hepes/KOH, pH 7.0, buffer or 40 mM KPO4, pH 7.0, buffer. Stock solutions of the dialyzed proteins (7-13 mg/ml) were stored<3 months at 4 "C in tightly closed conical tubes to minimize exposure to air.
Water was distilled, then deionized and filtered through a Millipore Milli-Q2 reagent grade system. Hepes, Tris, PMPS, N-ethylmaleimide, and ATP were from Sigma; CTP was from Boehringer Mannheim; PMB was from Aldrich; neobydrin was from ICN Pharmaceutical, Inc; Chelex 100 (100-200 mesh) was from Bio-Rad. Ultrapure grade of urea was from Schwartz/Mann and 4.0 M urea solutions were freshly prepared for addition to ATCase. PALA was a generous gift of Dr. Robert R. Engle of National Institutes of Health. PAR was used as received from Eastman Organic Chemicals. A 5.0 mM PAR stock solution (stored in the dark) was prepared by dissolving the solid dye in deionized water while adding 1 N KOH to maintain the pH a t 8.8 and then diluting to volume. All other chemicals were of analytical reagent grade and were used as received.
Hepes/KOH, pH 7.0, buffer and KPO4, pH 7.0, buffer (if necessary) were purified by passage through a column (5 X 33 cm) of Chelex 100 resin (100-200 mesh) in the K+ form (31). Purified Hepes and KPO, buffers were tested spectrophotometrically for metal ion impurities by the addition of lo" M PAR followed by the addition of 1 mM EDTA, pH 7.0; upon EDTA addition, the absorbance decrease at 500 nm was < 0.003, indicating < 5 X lo-' M Zn2+ if metal ion contamination were Zn2+ only. This measurement also sets an upper limit of 0.1% of the quantity of PAR in the stock solution which was bound to metal ions (assuming a 2:l PAR.Me2+ complex).
Methods-The Perkin-Elmer Model 320 spectrophotometer and black selfmasking semimicro cuvettes of 1.00-cm path length were used for UV difference spectra (ATCase % PALA in different buffer systems, f 0.02 A scale), for spectrophotometric titrations of ATCase with mercurial reagents and for kinetic measurements of reactions when these were > 40 s in duration (0.1 A scale; autorange), and for protein concentration determinations using A!&'", , , = 0.59 (32) and M , = 300,000 (8) for ATCase. Temperature was controlled (k 0.1 "C) in this and other instruments used for kinetic measurements by circulating water (+ 0.05 "C) through thermostable blocks; YSI thermistor probes were used to directly monitor solution temperatures.
Stock solutions of freshly dialyzed ATCase tested negatively for free Zn2+, using M PAR and subsequent 1 mM EDTA addition a t 500 nm. Also, the extent of PALA binding to ATCase preparations in 40 m M Hepes/KOH, pH 7.0, or in 40 mM KPO,, pH 7.0, was measured by spectrophotometric titrations (33) to be the theoretical 6.0 eq of PALA bound per ATCase molecule, monitoring peak-trough absorption changes at 289-285 nm. These measurements checked the concentrations of both the stock PALA solutions (30 mM) and the competent catalytic sites of ATCase.
For kinetic measurements of mercurial-promoted Zn2+ release from ATCase in the spectrophotometer, lo-' M PAR was used (monitoring absorbance changes at 500 nm) and the reactions were initiated by adding a small volume of PMPS to the thermostated ATCase solution (-1.0 ml); mixing was by cell inversion. In the few experiments using PMB, equal volumes (0.50 ml) of protein with PAR and the mercurial reagent with lo" M PAR were equilibrated at the same temperature and then mixed; alternatively, reactions with 0.2 mM PMB were initiated by adding 15 pl of ATCase. At > &fold excess [PAR] to [Zn'+] at pH 7.0) (20 "C), Ac = 6.6 X 10' M" cm-' at 500 nm for (PAR)'Zn*+ formation.' TO obtain first-order rate constants, The stopped-flow spectrophotometer (1.0-cm light path) used for kinetic experiments lasting < 40 s has been described by Rhee and Chock (34). In measurements of the kinetics of ZnZ+ release by stopped-flow, PAR was added at the same concentration to both solutions to be mixed to avoid possible artifacts associated with the formation or dissociation of PMPS .PAR complexes. Difference sedimentation velocity measurements were performed as described previously (21), using dialyzed enzyme samples at a protein concentration of 3.9 mg/ml, Schlieren optics, and a speed of 52,000 rpm. For As measurements of ATCase k PALA in two-cell experiments, the stoichiometry of PALA binding was determined separately by a spectrophotometric titration (see above). Values of As were computed by the procedure of Howlett and Schachman (15); the data Ar/i uersus time and In r uersus time for As and sedimentation coefficient calculations, respectively, were fitted separately by a linear least-squares regression, where r is the radial distance (in centimeters) from the center of the protein boundary to the axis of rotation, f is the average r value for the two boundaries, and Ar is the difference between r values at a given time after reaching speed. Photographs were taken at 4-min intervals and those taken at 12-56 min (after separation of the small amount of dimer in ATCase samples) were used for calculations.
Light scattering measurements at a 90" angle were made at 360 nm using a Perkin-Elmer Model 650-40 fluorescence spectrophotometer (equipped with a Model 057 X-Y recorder) with excitation and emission slits of 2 and 8 nm, respectively.
Polyacrylamide slab gel electrophoresis (7% running gel with 1 cm of 3% stacking gel) was in a Bio-Rad Protean slab gel unit (16-cm cell), using the Tris/glycine system of Jovin et al. (35) as in Ref. 23. Samples of -3.3 pg protein/channel were loaded in -13% glycerol, 0.013% bromphenol blue; after electrophoresis, gels were stained with Coomassie Blue R-250 and photographed after destaining.

RESULTS
The association of Zn2+ with PAR was not rate-limiting in our studies of Zn2+ release from ATCase at pH 7.0 and the affinity of PAR for Zn2+ was such that essentially all Zn2+ released from ATCase was converted to the highly absorbant (PAR)2Zn2+ complex (A6 = 6.6 X lo4 M" cm-' at 500 nm).'   The sensitivity of PAR to the addition of free Zn2+ (yellow to orange color change) and the relative insensitivity of PAR to the presence of mercurial or other reagents also were important properties in the present studies.
Release of Zn'+ during the Titration of AT&S~(PALA)~ with the Mercurial Reagent PMPS-Previous studies (19) showed that when ATCase is saturated with the high-affinity bisubstrate analog PALA, the "SH group of each c chain is protected from reaction with excess PMB. Fig. 1 shows plots of the absorbance changes (recorded after completion of any time-dependent absorbance changes) during the titration of the ATCase(PALA), complex with PMPS. The absorbance increase a t 500 nm in the presence of PAR was a measure of Zn'+ release from the protein and the absorbance increase at 250 nm was indicative of mercaptide bond formation (36). The break in the plot of A A 2 5 0 n m occurs a t 23.5 PMPS/ ATCase or a t -4 PMPS/regulatory chain, indicating that the 4 sulfhydryl groups of each r chain reacted with PMPS but that the sulfhydryl group of each c chain (2) did not react with PMPS in the presence of PALA. The break in the plot of occurred at precisely the same value of PMPS/ ATCase as did the plot of AA2s0nm, and the total quantity of Zn'+ released (based on calibrations with standard Zn'+)' was 97% of that expected for the release of 6 Zn'+/ATCase (based on the ultraviolet absorption of the protein solution).
An interesting feature of Fig. 1 is the linearity of the plot of A &~~~ uersw PMPS added during titration of ATCase "SH groups. This result indicates that even with the protein sulfhydryl groups in great excess (i.e. a t < 6 PMPS/ATCase), the attack of these groups by PMPS is not random. Rather, for each 4 PMPS molecules provided, 1 Zn2+ ion is released. An explanation of such a result is that once the first of the 4 "SH groups in a Zn'+ binding site of an r chain reacts with the mercurial reagent, the other 3 " S H groups of the same Zn'+ cluster rapidly form mercaptides with PMPS. The fact that the mercurial-promoted dissociation of ATCase proceeds 1 molecule at a time (see below and Ref. 12) further suggests that once 1 of the 6 Zn2+ binding sites of ATCase is disrupted, " S H groups of the other 5 Zn2+ sites are much more susceptible to attack by mercurials than are the r chain sulfhydryl groups of Zn'+ sites in intact ATCase.
Kinetics of Reaction of 4.0 eq of P M P S withATCase(PALA)G (%th Mercurial ReagentlAvailable Protein " S H Groups)- Fig. 2A shows first-order rate plots based on absorbance increases at 500 and 250 nm during the reaction of ATCase in Hepes buffer, pH 7.0, with sufficient PMPS to react with only %th of the available r chain sulfhydryl groups. In these experiments the sulfhydryl group of each c chain was protected from attack by PMPS by the binding of PALA. The reaction is clearly first-order in PMPS, the limiting reagent, and the half-time for the reaction measured as mercaptide formation (74 s) or Zn'+ release (77 s) at 19.7 "C is the same within experimental error.
First-order rate constants for the reaction of limiting PMPS with excess sulfhydryl groups of A T C~S~( P A L A )~ are listed in Table I for a wide range of initial PMPS/ATCase ratios. For the experiments with 2.15 PM ATCase(PALA)G, no significant change in the value of k was observed for an 8-fold change in the PMPS/ATCase ratio. This was the case whether Zn'+ release or mercaptide bond formation was monitored. (Of course, if plots such as those of Fig. 2A are truly linear, then changing the PMPS/ATCase ratio merely means looking at different linear segments of the same first-order plot.) However, a remarkable feature of the data in Table I is that the value of k is fairly insensitive to the total ATCase concen- gave a t most a 50% increase in the pseudo first-order rate constant. Fig. 2B shows a first-order rate plot for the light scattering decrease accompanying the reaction of 4.0 eq of PMPS with ATCase(PALAh under the conditions of Fig. 2A. The log of the light scattering decrease (Rr-R,) versus time was linear for a t least 85% of the decrease (data incompletely shown in Fig. 2B). The measured half-time (93 s) for the light scattering decrease was -1.2-fold greater than that for Zn2+ release or mercaptide formation (Fig. 2 A ) . The total light scattering decrease was approximately 12%, which is the value expected for ' 16th dissociation of ATCase (M, = 300,000) into 2 C trimers of M, = 99,000 and 3 R dimers of M, = 34,000 (8). The addition of 5 mM 2-mercaptoethanol at the end of the light scattering decrease in the experiments af Fig. 2B produced (within 1 min) a 97% return to the original light scattering value of intact ATCase.
ATCase samples treated with 1.6-fold excess PMPS were completely dissociated but possibly contained aggregated R subunits since the value of R, was 65% or -88% of that expected for complete dissociation. When 5 mM 2-mercaptoethanol was added to protein samples after complete dissociation with 1.6-fold excess PMPS, the light scattering value   19.7 "C presence of lo-* M PAR or a t 250 nm in the absence of PAR" in 40 Spectrophotometric measurements were made a t 500 nm in the mM Hepes/KOH buffer, pH 7.0, containing a saturating concentration of PALA as described in the legend of Fig. 2.4. The absorptiontime data were analvzed as described under "Experimental Procedures" to obtain the pseudo first-order rate constant.
[ increased within 1 min to 110-120% of the original value and remained constant for at least 10 min. This was consistent with polyacrylamide gel patterns which showed that the addition of 2-mercaptoethanol to mercurial-dissociated ATCase produces some aggregated ATCase species in addition to assembled ATCase. The total light scattering decrease after addition of 4.0 eq of PMPS to ATCase (Fig. 2B) indicates that all of the mercurial reagent is consumed in disrupting the c:r contacts in %th of the ATCase molecules present. This result is in accord with the early ultracentrifuge studies of Gerhart and Schachman (12) on the mercurial-promoted dissociation of ATCase in which the loss of intact ATCase was a linear function of the mercurial reagent added until all of the sulfhydryl groups of r chains had been titrated. Because Subramani and Schachman (23) more recently observed the formation of a relatively stable CZR, intermediate during the early stages of mercurial attack on ATCase in KPOdTris, pH 7.0, buffer, we investigated the effect of phosphate on the mercurial-promoted dissociation of ATCase, using polyacrylamide gel electrophoresis to separate protein components (Fig. 3). An accumulation of the C2Rz intermediate was observed only when the mercurial reactions had taken place in KPO, buffer. After treatment of ATCase(PALA)s in 40 mM KPO,, pH 7.0, buffer with 0, 4, 8, and 12 eq of PMPS, protein patterns show increasing amounts of C2Rz (c6r4) and C trimer (Fig. 3, center  channels). In contrast, the C,RZ intermediate is not visible in samples of ATCase(PALAI6 in 40 mM Hepes, pH 7.0, buffer (& 1 M urea) before or after treatment with 4, 8, or 12 eq of PMPS (Fig. 3).
The mobility of the mercurial reagent in mercaptide linkage to r chain sulfhydryl groups of ' 16th of the ATCase population PMPS (5 pl of 2.64 mM PMPS) to 3.30 p~ ATCase(PALA)e in 0.80 ml of 40 mM Hepes/KOH, pH 7.0, containing 83 p~ PALA at 19.7 "C. Measurements were made at a 90" angle and 360 nm with excitation and emission slits a t 2 and 8 nm, respectively. Prior to the experiment, the protein solution was degassed at room temperature and passed twice through Millex-GS 22-p filters which had been thoroughly rinsed with buffer. For the data in B, a 79% signal offset was used and 2 duplicate experiments were averaged; Rt and R, (as per cent change) is the ratio of the observed light scattering at time t and at infinite time, respectively, to that for undissociated ATCase x 100.
For complete dissociation of CzR3 to 2 C + 3 R, R, = 74% is calculated whereas R, = 65% was measured in a separate experiment by adding 1.6-fold excess PMPS per "SH group to 3.30 p~ ATCase(PALA)+ The initial rapid release of %th of the protein-bound Zn2+ was followed by a very slow release of almost all of the remaining Zn2+ over a period of 3 days, at which time polyacrylamide gel electrophoresis showed that nearly complete dissociation of ATCase had occurred. A control incubation to which 1 mM N-ethylmaleimide was added to ATCase(PALA)fi in the absence of PMPS produced no Zn2+ release or measurable dissociation of the enzyme. Thus, an initial attack on an r chain sulfhydryl group requires the avid mercurial reagent. Once c:r contacts in an ATCase molecule are disrupted, the mercurial reagent can be displaced slowly by the irreversible reaction of sulfhydryl groups with N-ethylmaleimide, thereby freeing the mercurial to attack masked sulfhydryl groups in Zn2+ binding sites in intact ATCase molecules. Table I1 gives first-order rate constants for Zn2+ release at pH 7.0 upon addition of 4.0 eq of PMPS to ATCase at 24.5 or 30.0 "C in the absence and presence of various effectors of the enzyme. In the absence of the bisubstrate analog PALA, smaller total absorbance changes (-20% less) were observed for this reaction, as expected if some of the PMPS reacted with the sulfhydryl group of each c chain in the absence of PALA. The reaction in Hepes buffer was faster in the presence of PALA, but only by a factor of -2, rather than the factor of 6 observed previously for the reaction of excess PMB with ATCase in a mixed Tris/KP04 buffer (19). Also, the rate of Zn'+ release upon the addition of 4.0 eq of PMPS to the inactive mutant ATCase231 in Hepes buffer was about the same as that measured with the wild-type enzyme (Table 11). This is in marked contrast to the 12-fold rate enhancement (compared with the wild-type enzyme) observed by Wall and Schachman (37) for the reaction of excess PMB with ATCasen:31 in Tris/KP04 buffer.

PMPS-
The activation energy for the reaction of ATCase with 4.0 eq of PMPS in Hepes buffer is the same in the absence and presence of PALA (Fig. 4). Furthermore, rate measurements at two temperatures gave the same activation energy of 18

Effects of various ligands on first-order rate constants for Zn2+
release during the reaction of 4.0 eq of PMPS with ATCase Kinetic experiments were performed at 24.5 and 30.0 "C and 500 nm in the presence of lo" M PAR using the Perkin-Elmer Model 320 spectrophotometer. Buffers at pH 7.0 were 40 mM Hepes/KOH, 40 mM KPO4, and a mixture of 25 mM Tris/HCI and 37 mM KPO,. Initial concentrations were 2.15 p~ ATCase and 8.6 p~ PMPS or 8.6 pM PMB (k values in parentheses); concentrations of effectors when present were 30 p M PALA, 0.21 mM CTP, 0.10 m M ATP, and 1.0 M urea. The first-order rate constants given in the table are average values from at least three separate determinations, each with an error in the computer fit of < 0.0007. The ratio of k values (k/$), where $ is without effectors and k is with effectors present, is given for each temperature and buffer system. The mutant enzyme ATCaseZ3, (2.1 p~) was substituted for the wild-type ATCase in this experiment.
PMPS was added to enzyme in 1 M urea after completion of a slow absorbance change at 500 nm, corresponding to the release of -6% of the total protein-bound Zn2+. A disruption of the C2RZ species in ATCase preparations (38) would account for the small absorbance change observed (see Fig. 3).
kcal/mol in KPO, buffer as in Hepes buffer. This is despite the inhibitory action of phosphate, where the substitution of 40 mM KP04, pH 7.0, for Hepes buffer caused a large decrease in the rate of Zn'+ release in the absence or presence of PALA ( Fig. 4; Table 11). The substitution of the mixed KPOJTris, pH 7.0, buffer used previously (19) further inhibited the mercurial reaction (Table 11), but increased the solubility of PMB so that measurements with 4 eq of PMB were possible. Comparable rates of Zn2+ release from ATCase were obtained with 4 eq of PMPS and PMB (Table 11).
The presence of the inhibitor CTP (0.21 mM) or the activator ATP (0.1 mM) in Hepes buffer decreased the first-order k -3-fold for Zn2+ release from ATCase in the absence of PALA (Table 11). However, the bisubstrate analog PALA produced a 3-4-fold increase in the rate of Zn2+ release from ATCase in the presence of CTP, ATP, or phosphate. The inhibitory effects of ligands on the rate of Zn2+ release from ATCase upon challenge with a substoichiometric amount of PMPS in Table I1 can be due to inhibition of the mercurial reaction (27) as well as to effects from ligand-protein interactions. However, the effects of PALA on the first-order k for Zn" release from ATCase (Table 11) must be due to the binding of this high-affinity bisubstrate analog to catalytic sites (2) since the free concentrations of this ligand were negligible. It is possibly significant also that PALA increased the rate of Zn2+ release the most when it was bound in the presence of P,, ATP, or CTP.
The presence of 1 M urea had little effect on the rate of Zn2+ release produced by the addition of 4 eq of PMPS to ATCase in Hepes buffer, increasing k -10% in the absence of PALA and decreasing k -10% in the presence of saturating PALA (Table II).3 Light scattering measurements, performed as in Fig. 2B with 1~ urea present in Hepes buffer, pH 7.0, gave about the same total decrease (R, = 11.4%) as in the absence of urea and the dissociation reaction produced by the addition of 4.0 eq of PMPS to 3.3 PM ATCase(PALA)G was apparently first-order with tL,? = 111 s at 19.7 "C. The firstorder release of Zn2+ measured under the same conditions in the presence of lo-, M PAR (as in Fig. 2 4 ) was -1.3-fold faster than the light scattering change in 1 M urea. Thus, 1 M urea in Hepes buffer decreased slightly the rates of both Zn2+ release ( Table 11)  The PALA-promoted conformational change of ATCase in Hepes/KOH, pH 7.0, buffer was measured by difference sedimentation velocity ( Table 111). The value of -3.6% for AS/S (uncorrected for bound ligand) produced by saturating PALA in 40 mM Hepes/KOH, pH 7.0, buffer was the same as that measured in KPO, buffer (15). Furthermore, the addition of 2 and 4 eq of PALA to ATCase in Hepes buffer produced 54 and 100% of the total negative change in As/s, respectively, as observed previously in KPO, buffer (15,21). Thus, the magnitude of the enhancement in the reactivity of r chain sulfhydryl groups produced by PALA binding of ATCase (Tables I1 and IV below) is not a simple function of the protein conformation, as measured by sedimentation coefficient changes.

Kinetics of Zn2+ Release in the Reaction of ATCase with
Excess PMPS-Previous kinetic studies of the reaction of ATCase with organomercurials (PMB, rather than PMPS) have utilized phosphate or Tris/HCl buffers, or a mixture of these. In the present study, the reaction of ATCase with excess PMPS was found to be much more rapid in Hepes/ KOH than in KPO, or Tris/HCl buffers. Since phosphate, Tris, and C1-are all expected to form complexes with PMPS and the rates of mercurial-sulfhydryl reactions have been shown to be sensitive to the presence of complexing agents Most of the experiments using excess PMPS were performed in the noninhibiting Hepes/KOH, pH 7.0, buffer and were therefore sufficiently rapid to require stopped-flow mixing. The stopped-flow spectrophotometer had the advantage of allowing observation to begin immediately at the time of mixing, whereas about 10 s at the beginning of each reaction were lost in the other kinetics experiments. A typical oscilloscope trace and the corresponding computer-generated firstorder rate plot are shown in Fig. 5A. In experiments over a    to obtain k for each experiment. B, spectrophotometric measurements were made as described in Table I1 at   Pseudo first-order rate constants were obtained by stoppedflow for the reaction of ATCase with excess PMPS under a variety of conditions, using the linear portion of first-order plots such as shown in Fig. 5A. The rate constant for a given concentration of PMPS was found to depend on the buffer used and the effectors present; Table IV, A, presents some representative data, As in the case of the reaction of ATCase with 4 eq of PMPS (excess ATCase sulfhydryl groups), Zn2+ release is more rapid in Hepes/KOH than in KP04, pH 7.0, buffer. In phosphate buffer, the presence of PALA increases the rate 2.2-fold which is less than observed at lower PMPS concentration ( Table 11). The addition of CTP in the presence of PALA decreases the rate 1.3-fold. Surprisingly, the rate of Zn2+ release from ATCase at high [PMPS] in Hepes/KOH, pH 7.0, buffer is decreased by the presence of PALA even though [PMPS]/[PALA] = 20,000.
For the reaction of large excesses of PMPS with ATCase in KPO, buffer with PALA present, the first-order rate constant based on the absorbance-time trace at 250 nm agreed well with that obtained at 500 nm. With PALA absent and in Hepes buffer even with PALA present, however, the absorbance at 250 nm continued to increase after completion of the absorbance change at 500 nm ( i e . Zn2+ release). The most reasonable explanation for the slow absorbance change at 250 nm is mercaptide formation with the active-site sulfhydryl group on each c chain. The PMPS-mercaptide bonds formed with r chain sulfhydryl groups are thermodynamically very stable but kinetically quite labile. These bonds are rapidly disrupted by reaction with 2-mer~aptoethanol.~ Also, the addition of the organomercurial neohydrin (2 or 5 mM) in 10-fold excess over PMPS present at the end of experiments such as that shown in Fig.  5A gave an absorption decrease at 250 nm that was complete within the time of stopped flow mixing (2-3 ms). This indicates that neohydrin rapidly displaces PMPS in mercaptide bond formation.
Pseudo first-order rate constants were obtained at several temperatures for the reaction of 100-fold excess PMPS ([PMPS] = 0.60 mM) with ATCase in Hepes/KOH, pH 7.0, buffer in the absence and presence of PALA. The data from these experiments are presented as Arrhenius plots in Fig.  6A. Below 37 "C, the rate of the reaction of ATCase with excess PMPS is decreased by the addition of PALA, rather than 6-fold increased as under the conditions of Blackburn and Schachman (19) using a mixed KP04/Tris, pH 7.0, buffer (see Table IV).
Considerably lower activation energies were obtained in Hepes buffer for the reaction of excess PMPS with ATCase ( Fig. 6A) than for the reaction of 4 eq of PMPS with ATCase (Fig. 4). Activation energies of -7 and 10 kcal/mol were obtained in the absence and presence of PALA, respectively, for Zn2+ release from ATCase in Hepes buffer upon addition of a 100-fold excess of PMPS (Fig.6A). Under these conditions, PALA appears to stabilize Zn2+ bonding sites, possibly by protecting against thermal melting of c:r contact regions in ATCase.
Arrhenius plots for the reaction of ATCase in 40 mM KPO,, pH 7.0, buffer with 6.7-fold excess PMPS are shown in Fig.  6B. It is clear that the presence of phosphate increased the activation energy (Ea) over that observed in Hepes buffer (Fig.   6A). Under the conditions of Fig. 6B, PALA produced a 4fold increase in the rate of Zn'+ release (Table IV, B) and decreased E, from 13.8 to 12 kcal/mol. For Zn'+ release from the inactive mutant ATCase231 in KPO4 buffer, an activation energy of 14.1 +_ 0.5 kcaljmol was measured as in Fig. 623. Thus, the wild-type and ATCase231 mutant enzymes are indistinguishable by this criterion.
Since previous studies (12,23) of the reaction of excess PMB (-4-fold PMB to ATCase "SH groups) employed a mixed KPO,/Tris buffer at pH 7.0, experiments were conducted under similar conditions using -33-fold PMPS or PMB/available ATCase "SH groups (Table IV, I?). A 6-fold enhancement in the rate of Zn2+ release with PMPS or PMB was produced by saturating PALA in the presence of 25 or 50 mM Tris in 40 mM KPO,, pH 7.0, buffer, as had been observed previously for the enhancement in the rate of ATCase "SH group reactivity by PALA binding (19,23). In fact, the pseudo first-order rate constants in the absence and presence of saturating PALA for Zn2+ release from ATCase in KPO4, 50 mM Tris in Table IV, B, are in excellent agreement with those reported by Blackburn and Schachman (19). NO significant difference was observed using PMPS or PMB as the 'In stopped-flow experiments, the transfer of PMPS from the mercurial complex PMPS-5-thio-2-nitrobenzoate to Z-mercaptoethanol had a second order k of -3 X 10' M" s-' at 15 "C. The results given in Table IV suggest that the magnitude of the increase in the sulfhydryl group reactivity of ATCase promoted by active site ligand binding is dependent on the amount of excess organomercurial and the buffer components present. For example, with 267and 6.7-fold PMPS/ATCase "SH groups in KPO, buffer, the effect of PALA was to increase the rate of Zn2+ release 2-and 4-fold, respectively. The addition of Tris to KP04 buffer further inhibited the mercurial reaction and increased to a factor of 6 the effect of PALA on the rate of Zn2+ release.

Titrations of A T C~S~( P A L A )~
with PMPS in the absence and presence of the high-affinity metallochromic indicator PAR indicated that ZnZ+ release and mercaptide formation are coincidental. Furthermore, the release of Zn'+ from the A T C~S~( P A L A )~ complex was a linear function of the mer-curial added with 1 eq of Zn2+ released for every 4 eq of PMPS added during the titration of the 24 sulfhydryl groups of r chains in the enzyme molecule. At limiting PMPS (with ATCase-SH groups in excess) the release of Zn2+ was first order in mercurial concentration. These observations indicate that the rate-limiting step for Zn2+ dissociation from ATCase is the mercurial attack on one of the 4 "SH groups involved in tetrahedral coordination of Zn2+ in an r chain of an ATCase molecule (7, 9, 13). Once the first of the 4 " S H groups in a Zn2+ binding site reacts with the organomercurial reagent, the other 3 " S H groups react rapidly which gives a linear release of 1 Zn2+ for every 4 eq of PMPS added. The x-ray crystallographic analyses (9, 10) of the ATCase structure places the Zn2+ binding site in each r chain very near the r:c contact region. In fact, Monaco et al. (9) noted that all substructures of loops ending with cysteinyl residues 109, 114, 137, and 140 that are coordinated to Zn2+ in each r chain participate in the interactions between r and c chains. This explains the earlier observations of Gerhart and Schachman (11, 12) who showed that PMB produces dissociation of ATCase into C trimers and R dimers. Furthermore, the mercurial-promoted dissociation of ATCase was shown also in those studies to proceed 1 molecule at a time (with any intact ATCase present containing no mercurial in mercaptide linkage), suggesting that once one r:c contact is disrupted, the other 5 ZnZ+ sites become more susceptible to mercurial attack and consequently to r:c contact disruption. In the present studies, we have corroborated this result by polyacrylamide gel electrophoresis during titration of ATCase with PMPS (Fig. 3) and by light scattering measurements of the reaction of 4 eq of PMPS with A T C~S~( P A L A )~ in Hepes buffer at pH 7.0 (Fig. 2B), in which almost exactly ' 16th dissociation of ATCase was observed. Under these conditions, there appears to be no accumulation of the CzR2 intermediate as there is in KPO,, pH 7.0, buffer (Fig. 3). Thus, the rate-limiting reaction of the 1st of 4 " S H groups in a Zn2+ binding cluster triggers the rapid reaction of the other 23 " S H groups of the 6 r chains in an ATCase molecule. This catastrophic effect of organomercurial reagents on ATCase results in the dissociation of ATCase 1 molecule at a time.
We have used the sensitive metallochromic indicator PAR in the present kinetic studies to monitor the mercurial-promoted release of Zn2+ from ATCase, both under conditions of limiting organomercurial reagent (ATCase sulfhydryl groups in excess) and of limiting ATCase with the mercurial reagent in excess. PMPS rather than PMB was used because PMB has a limited solubility, although PMB when tested gave similar rates to those obtained with PMPS (Tables I1 and   IV). These organomercurial compounds have a single labile ligand which is displaced by reaction with thiols. Khalifah  (27, 39) has shown that the kinetics of organomercurial-thiol reactions can be influenced profoundly by facile replacement of the mercurial labile ligand prior to reaction with the thiol. For this reason, the noninteracting Hepes/KOH buffer at pH 7.0 was chosen for many of our experiments. However, some studies were conducted also in the presence of organomercurial inhibitors at pH 7.0 (Tris/HCl with Ki 0.5 mM, Brwith K, = 0.5 mM, KPO, and mixtures of Tris/HC1/KP04) in order to directly compare the results from this and previous studies (19, 23) . Kahlifah (27, 39) observed that the effect of a particular organomercurial inhibitor is unpredictable, being a function of the stability of the inhibitor-mercurial complex and the nature of the thiol group under attack. In the present studies, the enhancement in the rate of PMPS attack on ATCase produced by binding the high-affinity bisubstrate analog PALA at enzyme catalytic sites also was dependent on the mercurial inhibitor(s) present (Tables 11 and IV). The PALA-promoted conformational change of ATCase (As/s = -3.6% ; Table 111), however, is the same in Hepes/KOH, pH 7.0, buffer as in KP04, pH 7.0, buffer, which acts as a mercurial inhibitor.
Under the conditions of limiting [PMPS] and excess [ATCase " S H groups], the reaction of PMPS with ATCase is clearly first order in [PMPS]. No lag time was noted in these experiments, but observations began after the first 5-10 s of the reaction. With excess [ATCase " S H groups], the reaction rate is insensitive to the concentration of protein (Table I)  and stabilizes an accumulation of the CZR~ intermediate (Fig.  3), which reacts with the mercurial reagent at about the same rate as does intact ATCase, C2R3 (23). Note also that the reaction with excess [ATCase " S H groups] in phosphate buffer is first order in PMPS, the limiting reagent, just as it is in Hepes buffer. Certainly, the major inhibitory effect of phosphate on the mercurial reaction with thiols of ATCase is due to the binding of phosphate to PMPS (27,39).
Several quantitative changes in kinetic behavior were observed as the concentration of PMPS was increased. At high concentrations of PMPS, the reaction becomes first order in [ATCase], after an initial lag period, and the order of the reaction with respect to [PMPS] increases to a value between first and second (Fig. 5). In addition, the apparent Arrhenius activation energy is decreased substantially from that for the case of limiting [PMPS]. Furthermore, the effect of PALA binding to ATCase on increasing the reaction rate with mercurials appears to be reversed at high [PMPS] at temperatures < 37 "C in Hepes buffer (Fig. 6A). PALA and phosphate appear to stabilize ATCase during the attack by excess PMPS, producing an increase in the Arrhenius activation energy over that observed in Hepes buffer. In contrast, the inactive mutant ATCasezal appears to not be stabilized against mercurial attack by the presence of phosphate.
In stopped-flow experiments under the conditions of excess [PMPS] to [ATCase "SH groups], the initial lag observed was a constant fraction of a half-life whether the buffer was in Hepes or phosphate at pH 7.0, or whether the reaction was monitored at 500 nm for Zn2+ release or at 250 nm for mercaptide formation. This initial lag could be due to a steadystate accumulation of a more reactive species. The evidence is against the dissociation intermediate CzRz being a more reactive species: ( a ) Subramani and Schachman (23) found that the rate of mercurial attack on C2R2 and on intact ATCase to be approximately the same; (b) the accumulation of CzR2 in KPO, buffer is not reflected in proportionately longer lag times. We propose below that a more reactive species could be generated in a pre-equilibrium step by mercurial attack at nonthiol groups of ATCase that then enhance the susceptibility of Zn2+-thiol clusters to attack by PMPS.
Many of the discrepant observations may be rationalized by again noting that organomercurials may form complexes with a variety of moieties other than sulfhydryl groups, including nitrogenous side chain groups in the protein itself. Since Hepes apparently binds to PMPS weakly, whereas amines such as Tris bind to PMPS rather well, it seems reasonable to suggest that ATCase may serve simultaneously as a reactant and as an inhibitor in Hepes buffer. For example, the sequestering of PMPS by lysyl e-amino acid groups (without an accompanying absorbance change at 250 nm) would lower the concentration of free [PMPS] available for reaction with sulfhydryl groups (with an accompanying absorbance change a t 250 nm) as illustrated in Reactions 1 and 2: R-HgOH + ATCase-NH; A k, R-Hg-NHZ-ATCase + H 2 0 (1) x, R-HgOH + ATCase " S H -R-Hg-S-ATCase + HzO (2) k, slow where HgOH is the hydroxyl complex of PMPS and ATCase-NHZ and ATCase "SH are simply ATCase molecules written so as to indicate the group with which PMPS reacts.
If Reactions 1 and 2 are occurring simultaneously, then increasing [ATCase] would tend to increase the rate of Reaction 2 directly but at the same time decrease the rate of The proposed inhibitory effect of ATCase would not be observed in any single kinetic experiment even if ATCase were the limiting reagent, since the concentration of competing amino acid side chains would remain constant. The presence of such protein inhibitory groups also would not affect the order of the apparent reaction in a single kinetic experiment.
With [ATCase] limiting at high [PMPS], the mercurial reaction will be pseudo first-order in [ATCase]; Equation 3. It is conceivable that the binding of PMPS to a non-thiol group of ATCase could accelerate the reaction of PMPS with the thiol groups, presumably by means of conformational effects, rather than retarding the reaction by decreasing [PMPS]. This mode of attack by PMB was proposed earlier by Pigiet (41) to explain the second-order path implied by the rate law of Gerhart and Schachman (12), in which the reaction order in [PMB] was greater than 1 a t low [PMB] and was 1 at high [PMB]. However, the earlier studies on the reaction of ATCase with PMB (11, 12) were severely restricted by the sensitivity of measurements of mercaptide formation (36) and by the limited solubility of PMB.
It certainly would not be surprising that the high concentrations of the organomercurial PMPS used in the present studies could have a chaotropic effect. Saturation of "helper" sites in a pre-equilibrium step (reaction 1) could explain the observed initial lag in the pseudo first-order reaction of ATCase with excess [PMPS] and also the change in the apparent reaction order with respect to [PMPS], which indi-cates that more than 1 mercurial molecule is involved in the attack on ATCase thiols when [PMPS] is in excess. In this case, Reactions 1 and 2 are as written and Such a chaotropic effect at high [PMPS] could decrease the Arrhenius activation energy from 18 to -7 kcal/mol. The bisubstrate analog PALA stabilizes ATCase against this action of the organomercurial in Hepes buffer until increasing temperatures melt out such structures. In inhibitory buffers containing KP04, however, PALA promotes a conformational change that enhances the rate of reaction of ATCase thiols with excess PMPS or PMB (19, 23). Indeed, the Arrhenius activation energy for the reaction of excess [PMPS] with ATCase in K P 0 4 buffer (13.8 kcal/mol) is higher than that in Hepes buffer and is decreased to 12 kcal/mol by saturating PALA. The additional effects of amines such as Tris on the fold enhancement in the reaction rate of ATCase with excess [R-Hg] given by PALA binding could relate to differences in facile ligand exchange of the Tris-substituted mercurial reagent with groups exposed on the protein in the R-and Tstate conformations of ATCase (2, 15).
In summary, the studies of the kinetics of Zn2+ release from ATCase using the metallochromic indicator PAR have been found to correlate exactly with reaction rates of organomercurial compounds with ATCase sulfhydryl groups, since the action of the mercurial is necessary to disrupt Zn2+ binding clusters in ATCase. The mercurial-promoted dissociation of ATCase occurs in an "all or none" or "zipper" fashion, suggesting that once one Zn2+ site is disrupted in an ATCase molecule, the other five Zn2+-thiol clusters are destabilized. Thus, the intactness of Zn2+ binding clusters in an ATCase molecule is essential for thermodynamic stabilization of r and c chain interactions responsible for the allosteric properties of this enzyme. Studies on the kinetics of Zn2+ release and uptake by isolated R subunits are in progress.