- Email: [email protected]
Inactivation of yeast phosphoglycerate kinase by Cr-ATP complexes and its implications on the conformation of the enzyme active site
Inactivation of Yeast Phosphoglycerate Kinase by Cr-ATP Complexes and Its Implications on the Conformation of the Enzyme Active Site Engin H. Serpersu, Linda L. Summitt, and Jay D. Gregory Department of Biochemistry, University of Tennessee, Knoxville, Tennessee
ABSTRACT Exchange-inert p,y-bidentate Cr(H,O),(NH,),ATP complexes inactivate yeast phosphoglycerate kinase (PGK) by forming a coordination complex at the enzyme active site. The observed inactivation rates ranged from 0.019 min-’ to 0.118 min- ’ for Cr(NH,),ATP and Cr(H,O),ATP, respectively. Incorporation of one mol of Cr-ATP to the enzyme was sufficient for complete inactivation of the enzyme. The presence of Mg-ATP protected the enzyme against inactivation by Cr-ATP. The other substrate 3-phosphoglycerate (3-PGA), when present, reduced the observed inactivation rates. The reduction of the k,, by 3-PGA was proportional to the number of NH, ligands present in the coordination sphere of Cr3+ in the Cr-ATP complex, suggesting that in the ternary enzyme-Cr-ATP-3-PGA complex 3-PGA may be coordinated to the metal ion. When the effector sulfate ion was present, the presence of 3-PGA did not cause any further effects on the observed inactivation rates. This suggests that bound substrates are in a different arrangement at the active site when sulfate is present and therefore 3-PGA may not need to displace a ligand from Cr3+. Additionally, PGK exhibited a stereoselectivity for the binding of Cr(H,O),ATP. A diastereomer of Cr(H,O),ATP yielded an order of magnitude smaller Ki value compared to the value observed with the A isomer. The recovery of enzyme activity was observed over a period of a few hours upon removal of excess Cr-ATP. The presence of substrates and/or effector ion sulfate did not alter the observed reactivation rate. There was no difference in the reactivation rates of the enzyme which was inactivated with Cr(H,O),ATP or Cr(NH,),ATP with and without 3-PGA. Increasing the ligand exchange rates of Cr3+ of Cr-ATP by increasing the pH value of the recovery medium from 5.9 to 6.8 increased the rate of recovery by a factor of 8. The pH dependence of the reactivation indicated that one hydroxyl group is involved in the recovery of the enzyme activity in enzyme. CrATP and enzyme. CrATP .3-PGA complexes.
Address reprint requests and correspondence to: Engin H. Serpersu, Department of Biochemistry, Walters Life Sciences Building M-408, University of Tennessee, Knoxville, TN 37996-0840. Journal of Inorganic Biochemistry, #3,203-215 (1992) 0 1992 Elsevier Science Publishing Co., Inc., 655 Avenue
203 of the Americas,
NY, NY 10010
0162-0134/92/%5.00
204
E. H. Sepersu
et al.
ABBREVIATIONS 3-PGA, 3-phospho-D-glycerate; PGK, yeast phosphoglycerate kinase (EC2.7.2.3); MES, 2-(N-morpholino)ethanesulfonic acid; 1,3-dPGA, 1,3-diphospho-D-glycerate; ATP, 5’-adenosine triphosphate.
INTRODUCTION Phosphoglycerate
kinase (EC 2.7.2.3) catalyzes the phosphoryl transfer reaction
[II: Mg2+ 1,3-diphospho-D-glycerate
+ ADP + 3-phospho-D-glycerate
+ ATP.
binds two moles of ATP per mole [2-51, one of which is bound to the anion binding site and has very low affinity to Mg2’ [4]. Catalytically required Mg2+ is presumably bound to ATP at the active site. Multivalent anions and particularly sulfate ion are known to interact with phosphoglycerate kinase and affect the kinetics of the reaction [5-71. Sulfate ion in the range l-20 mM causes an increase in the observed reaction rate with an increase in K, of both substrates. The presence of a second binding site distinct from the active site has been suggested by previous studies. However, the nature and proximity of this binding site to the active site are not known. Crystallographic studies have been performed with the enzyme crystals obtained in the presence of 2 M or higher concentrations of ammonium sulfate 18, 91. These studies are tentative because ofothe problem that the phosphor-y1 group to be transferred from Mg-ATP is N 10A from the accepting carboxylate group of 3-PGA [lo]. Small-angle x-ray scattering studies of 3-PGK in solution revealed that the radius of gyration of the enzyme decreases when both substrates are bound to the enzyme [ll]. Additional solution studies also suggested that binding of both substrates to the enzyme induces large conformational changes on the enzyme [ll-131. Therefore, a hinge-bending motion of the enzyme to bring two substrates together in a “closed” active site was suggested for the catalytic mechanism of the enzyme [lo, 111. However, the nature and the extent of the interactions between the substrates and the enzyme is not known. We have started structural studies in solution with binary and ternary enzyme-substrate(s) complexes by using stable exchange-inert Cr 3+-ATP complexes as Mg-ATP analogs 114, 151. Cr-ATP complexes have been shown to be classical competitive inhibitors and inactivators of various kinases and ATPases [14-191. This paper describes inactivation of yeast phosphoglycerate kinase by various Cr-ATP analogs. Our results suggest that Cr-ATP mimics the substrate Mg-ATP and forms a ternary enzyme-Cr-ATP-3-PGA complex at the active site. The slow inactivation of the enzyme by various Cr-ATP complexes allowed more detailed characterization of the ternary enzyme-metal-ATP-3-PGA complex in the presence and absence of effector sulfate ion. The enzyme
SUBSTRATE
EXPERIMENTAL
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
205
PROCEDURES
Materials Various P,ybidentate Cr(H,O),(NH,),,ATP complexes were prepared as described earlier [15, 20-221. The separation of A and A diastereomers of Cr(H,O),ATP was done as described 1231. Experiments were performed with the racemic mixture except where specified. Yeast phosphoglycerate kinase was purchased from Boehringer. The enzyme was also isolated from yeast strain 20B-12 containing multi-copy plasmid pCGY219. The plasmid and the yeast strain were kindly provided by Dr. Hitzeman of Genentech Inc. Both enzymes had k,,, of N 180 see-’ at pH 5.9. Enzyme concentration was determined from its absorbance at 280 nm (Elc,‘.i% = 0.5). Methods Inactivation of the Enzyme. The enzyme (30-40 pg) was incubated in 50 mM MES pH 5.9 at 30°C in a total volume of 200 ~1 containing 150 mM NaCl and various concentrations of Cr-ATP. This pH was selected because of the instability of Cr-ATP complexes at higher pH 115, 201. In some cases, the second substrate 3-phosphoglycerate (2.7 mM) and/or sulfate ion (20 mM) were also present in the incubation medium. When sulfate ion was present ionic strengths of the solutions were adjusted by lowering NaCl concentration as necessary. At indicated time points, 5-8 @’ aliquots were withdrawn and the remaining enzyme activity was determined spectrophotometrically at 340 nm by measuring the initial rates in a coupled enzyme assay in a total volume of 1 ml at 25°C. The assay medium contained 150 mM NACI, 0.45-0.51 mM ATP, 1.5 mM MgCl,, 8 mM 3-phosphoglyceric acid, 0.2 mM NADH, and 150 pg glyceraldehyde 3-phosphate dehydrogenase in 50 mM MES pH 5.9. With the indicated amount of coupling enzyme, from which (NH&SO, was removed by dialysis or desalting, there was no lag in the reaction and the rate observed was linearly dependent on the concentration of phosphoglycerate kinase present. The recovery of enzyme activity upon dilution into the assay cuvette was insignificant and no deviation from linearity was observed for the duration of the assays (45-60 set). Studies with [y 32P]Cr(H,0)4 ATP. Inactivation of phosphoglycerate kinase with [y-32P]Cr(H20)4ATP (25-35 cpm/pmole) was performed in a similar manner as described above with the exception that the enzyme concentrations were 216-250 PM. At indicated times, two aliquots were withdrawn from the inactivation mixture and diluted in a medium containing 50 mM MES pH 5.9, 150 mM NaCl, 7-8 mM 3-PGA, 6 mM ATP, and 7 mM MgCl,; 90 ~1 of 11-fold diluted sample was then applied to a centrifuge column [24] containing Sephadex G-50 in a 1 ml tuberculin syringe and centrifuged at 180 X g for 90 sec. Enzyme-bound radioactivity was then determined by liquid scintillation counting of the effluent. The highest concentration of free [y-32P]Cr-ATP used in the assays did not penetrate more than the upper third of the column length in our conditions. Thus no background counts due to unbound Cr-ATP were observed. Another aliquot from an appropriately diluted second sample was used for the determination of the residual enzyme activity as described above.
206
E. H. Serpersu et al.
Reactivation Studies. Phosphoglycerate kinase (170-190 PM) was inactivated with Cr-ATP (4-5 mM) in the presence or absence 2.70 mM 3-PGA as described above. When the steady-state level (15-20% of the original activity) was reached, 100 ~1 aliquots were applied to the centrifuge columns to separate the enzyme from the excess free Cr-ATP. This enzyme solution was then immediately diluted into 50 mM MES buffer at indicated pH to give an enzyme concentration of 2.57 PM. This solution contained various combinations of substrates/ions as indicated in the figure legends. Diluted enzyme solution was then left at 23°C and at indicated time the enzyme activity was determined as described above.
RESULTS Inactivation of Phosphoglycerate Kinase by Cr-ATP. Incubation of 1 PM enzyme with various concentrations of Cr(H,O),ATP showed a time dependent pseudo-first-order loss of enzyme activity when the residual enzyme activity was determined in the presence of saturating 3-PGA and 3.3 K,, Mg-ATP or 20 K, Mg-ATP (which yielded identical results-data not shown). The inactivation is slow and reaches a steady-state after a few hours. Increasing concentrations of Cr(H,O),ATP resulted in increased rate of inactivation and in decreased levels of residual enzyme activity. Since there is no appreciable dissociation of enzymeCr(H,O),ATP complex during the assay for residual activity (45-60 XC), a long-lived enzyme-Cr(H,O),ATP complex would account for this kinetic behavior. Since Cr3+ exchanges its ligands very slowly [14, 151, it is likely that this long-lived complex may form by donation of a ligand from the enzyme to the coordination sphere of Cr- ‘+ in enzyme-Cr(H20),ATP complex. The pseudofirst-order rate constants (kobs) for each concentration of Cr(Hz0)4ATP were determined and plotted in double reciprocal form (l/k,,h, vs l/[Cr-ATP]) which is linear and yields a finite inactivation rate at saturating Cr(H20),ATP concentration (Fig. 1) which is consistent with the following mechanism [25, 261: E+IsEI k,
2 EI* k,
(I)
in which enzyme (E) and inhibitor (I) rapidly form EI which then slowly isomerizes to EI* [25, 261. The observed K, (161 PM) is in reasonable agreement with the Ki value obtained from the initial rate studies which showed that Cr(H,O),ATP is a linear competitive inhibitor of phosphoglycerate kinase with a Ki = 70 PM with respect to Mg-ATP (J. D. Gregory and E. H. Serpersu, manuscript submitted). As expected, the presence of increasing concentrations of Mg-ATP provided increasing levels of protection of the enzyme against inactivation by Cr(H,O),ATP. The presence of 125 PM Mg-ATP was enough to reduce the observed inactivation rate by a factor of two with 246 PM Cr(Hz0)4ATP present (data not shown). To test the hypothesis that Cr 3+ indeed acquires a ligand from the enzyme in the enzyme-Cr(H20),ATP complex, we have prepared a series of Cr(H,O),(NH,),ATP complexes [20-221. Since displacement of NH, from the coordination sphere of Cr3+ is more difficult than the displacement of HzO, these analogs should inactivate the enzyme with progressively decreasing koh\
SUBSTRATE
-24
BINDING
-16
TO YEAST
PHOSPHOGLYCERATE
0
-8
8
l/[Cr-ATPI
16
KINASE
207
24
mM--1
FIGURE 1. Inactivation of Yeast PGK by various Cr-ATP complexes. 1.0 PM enzyme was incubated at 30°C in 50 mM MES, pH 5.9, containing 150 mM NaCl and various concentrations of the Cr-ATP complexes. The data is presented in double reciprocal form: (0) Cr(H,O),ATP, (0) Cr(H,O),(NH,),ATP, (A> Cr(H,OXNH,),ATP, and (+> Cr(NH,),ATP. The lines represent weighted fits to the data points. The inset shows the three-dimensional plot of the effects of NH, coordination to Cr3+ and 3-PGA on the observed inactivation rates of PGK by Cr-ATP complexes.
values proportional to the increased number of NH, ligands in the coordination sphere of Cr 3+. As is shown in Table 1, this is indeed the case, where the inactivation of PGK by various Cr(H,O),(NH,) ATP complexes revealed that decreases systematically the determined kobs = 0.118 min-’ with Cr(H,O$,ATP as the amino to a 6.2-fold lower value of k,, = 0.019 min-’ with Cr(NH,),ATP groups replace water molecules for coordination to Cr3+ in Cr-ATP complexes. Small changes in the determined Ki values suggest that replacement of H,O with NH, did not have significant effects on formation of the initial EI complex in Eq. (1). However, a long-lived enzyme * Cr-ATP *substrate complex can form without the formation of a covalent bound [22]. To distinguish these possibilities,
TABLE 1. Inactivation of Yeast Phosphoglycerate Kinase by Various Cr-ATP Complexes in the Absence of 3-Phosphoglycerate at pH 5.9 Cr-ATP Complex CXH20,4 ATP Cr(H,O),(NH&4’II’ Cr(H,OXNH,)@F’ Cr(NH,),ATP
kobs
(min-‘) ,118 .068 ,040 .019
+ + + *
,012 .OlO ,006 .003
161 161 145 93
+ f f f
22 31 20 12
208
E. H. Serpersu et al.
we incubated the enzyme with y-[32P]-Cr-ATP. When the enzyme was 76.4% inactivated (23.6% remaining activity) the reaction was stopped by the addition of HClO,, at a final concentration of 4%. The denatured enzyme was then washed several times with buffer solution to remove unbound Cr-ATP. The radioactivity associated with the enzyme yielded 0.71 mole Cr-ATP/mol enzyme stoichiometry. This strongly suggests formation of a bond between the enzyme and Cr-ATP. The Effect of 3-Phosphogfyceric Acid. As can be seen from Table 2, when the inactivation of phosphoglycerate kinase was performed in the presence of 2.70 mM 3-PGA, the observed rates of inactivation were lower compared to the rates observed in its absence, with a small effect on K,. Additionally, the reduction in with the increasing number of NH, in the coordinak obs was more pronounced tion sphere of Cr 3+ in the Cr-ATP, which is presented in Table 2 as the ratios of the rates of inactivation in the absence and presence of 3-PGA (kobs/kobsPGA). is 1.36 which increases to 2.5 with The ratio of kobs/kobsPGA with Cr(H,O),ATP Cr(NH,),ATP. The effects of 3-PGA and NH, coordination to Cr3’ is shown in Figure 1 as a three-dimensional plot. These observations suggest that during the formation of the ternary enzyme-3-PGA-Cr-ATP complex, either a liganding group from 3-PGA molecule or another ligand from the enzyme is entering to the coordination sphere of the metal ion. This would necessitate displacement of another ligand from Cr3+ which would then reduce the observed inactivation rate due to the slower ligand exchange rates of Cr”+. The Effects of Sulfate and Mg Ion. The kinetics of the phosphoglycerate kinase reaction have been found to be affected by the presence of sulfate ion and the observed K, values for the substrate increase [6, 71. Therefore, we have tested the effect of sulfate ion on the inactivation of PGK by Cr-ATP complexes. The presence of sulfate ion resulted in a decrease in kobs to values similar to those observed with 3-PGA when the enzyme was inactivated with Cr-ATP. In the presence of sulfate ion, the addition of 2.7 mM 3-PGA did not cause any further changes in the observed rate of inactivation (data not shown). This was the case with all of the Cr(H,O),(NH,),ATP complexes, which may suggest that in the presence of sulfate ion bound substrates have different arrangements at the active site. Hence no additional ligands of the metal have to be replaced by a liganding group from 3-PGA or from the enzyme. Alternatively, sulfate ion and 3-PGA may bind to the same site on the enzyme and prevent the inactivation of
TABLE 2. Inactivation of Yeast Phosphoglycerate Kinase by Various Cr-ATP Complexes in the Presence of 3-Phosphoglycerate at pH 5.9 Cr-ATP Complex Cr(H,O),ATF’ Cr(H,0)2(NH,),ATP Cr(H20XNH,),ATP Cr(NH,),ATP The column k ,,,,/k,,,(PGA) by Cr-ATP in the absence tively.
k ohs (min-‘1 ,087 ,030 .018 .008
+ .Ol1 + ,005 + ,004 + .OOl
kobs/kobs
(PGA) 145 + 465 f 267 f 92+
21 106 41 18
1.36 2.24 2.26 2.50
shows the ratio of the observed inactivation rates (kubJ of the enzyme and presence of saturating concentration of 3-PGA (2.7 mM), respec-
SUBSTRATE
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
209
the enzyme by different mechanisms. It was previously shown that sulfate activates the enzyme at low concentration and it is a competitive inhibitor at high concentrations 171. The presence of Mg2+ (up to 10 mM) did not alter the observed inactivation rates regardless of the composition of the inactivation medium. This observation suggests that in the formation of the ternary complex no additional metal ion is needed, when there is a metal ion already bound to ATP (Cr3+ in this case). Stoichiometry of Cr(H,O), ATP Binding to PGR Incubation of phosphoglycerate kinase with [ y-32P]Cr(H20),ATP resulted in the stoichiometric labeling of the enzyme (Fig. 2). As shown in Figure 2, labeling of the enzyme was proportional to the inactivation of the enzyme and no further incorporation of radioactivity was observed when steady-state levels of enzyme activity were reached. The inset in Figure 2 shows the correlation of the loss of activity with the stoichiometry of the incorporation of [32P]Cr(H20)4ATP, which indicates binding of one mol of Cr(H,O), ATP to one mol of enzyme. When Mg-ATP was present, incorporation of radioactivity to the enzyme was prevented (Fig. 2). This is consistent with binding of Cr(H,O),ATP to the active site of the enzyme. In all cases the presence of 20 mM sulfate ion in the inactivation medium did not alter the observed Cr(H,O),ATP/enzyme stoichiometry (data not shown).
--
0.8
e
80
z 0.6 L
3
60:’
2
0
50
100
Incubation
Time
150
200
(mid
FIGURE 2. Binding of [-Y-~*P]C~(H,O),ATP to Yeast PGK during inactivation. The inactivation conditions were as described in Experimental Procedures. Open symbols represent the time dependent incorporation of radioactivity to the enzyme (right axis) and filled symbols represent the relative enzyme activity with respect to the initial activity. Activity measurements of some samples were omitted from the plot for simplicity. (O,O) 0.28 mM Cr(H,O),ATP with 4.05 mM 3-PGA, (0, H) 1.12 mM Cr(H,O),ATP with 4.05 mM 3-PGA, (A) 1.08 mM Cr(H,O),ATP, (0) 1.08 mM Cr(H,Ol,ATP with 6.48 mM MgATP. Curve fitting was done as described in Figure 1. In the inset the data for the inactivation of the enzyme in the presence of 3-PGA is plotted in relative activity vs stoichiometry of the incorporated label. V, and V, represent the measured enzyme activity at time t and “zero”, respectively; (0) .28 mM Cr(H,O),ATP and (0) 1.12 mM Cr(H,O),ATP. The lines represent the least-square fits to the data points.
210
E. H. Serpersu et al.
Stereoselectivity of Phospho&cerate Kinase. When individual diastereomers of /3,-y-bidentate Cr(H,O),ATP were used to inactivate phosphoglycerate kinase, both diastereomers yielded kobs values similar to the value observed with racemic mixture. However, the A diastereomer yielded a Ki value of 7.0 PM which is about an order of magnitude smaller than that of the A diastereomer (62.6 PM). In addition, we have inactivated the enzyme with a racemic mixture of Cr(H,O),ATP and removed the enzyme with bound nucleotide by ultrafiltration. The analysis of the remaining solution by CD spectroscopy revealed that the A diastereomer was enriched in the solution, indicating a preferential binding of A diastereomer to the enzyme. These observations are in agreement with the earlier findings with sulfur substituted (at the P-phosphorus of ATP) Mg-ATP analogs, which yielded that PGK is highly specific for S, isomer [271. Screw-sense A isomer of Cr(H,O),ATP corresponds to the S, isomer of Mg-ATPP S [28]. Recovery of the Enzymatic Activity After Preincubation with Cr-ATP. Although we observed no recovery of the catalytic activity for the duration of the enzymatic assays, when inactivated enzyme is separated from the excess Cr(H20),ATP and diluted in buffer with or without the substrates, a slow recovery of the enzymatic activity was monitored over a several-hour period. In the presence of various combinations of substrates and effecters at pH 5.9, the recovery was 70 to 100% of the activity of the enzyme treated in exactly the same manner with the omission of inhibitor (Fig. 3A). The samples without Cr(H,Ol,ATP showed less than 5% loss of enzyme activity for the duration of the experiments t- 16 hr). Analysis of the data for the slow recovery of Cr(H,O),ATP-inactivated PGK activity by fitting to the first-order rate equation yielded an average rate constant of 5.5 x 1O--3 min-’ with t,,, = 140 min. The rate of regain of enzyme activity was not affected when substrates and/or sulfate was present in the recovery medium, suggesting that the conformational changes induced by substrates and/or sulfate ion have no effect on the recovery of enzyme activity. This would be expected if the recovery of the enzyme activity is the result of the displacement of the coordinated enzyme side chain from Cr” by water. Hence, the rate constant for the reactivation will remain unchanged regardless of the constituents of the recovery medium. As shown in Figure 3B, the rate of regain of enzyme activity was much faster when the inactive enzyme-Cr-ATP complex was diluted in buffer solutions at higher pH. The rate of reactivation increased g-fold from 5.5 x 10mm3min- ’ at pH 5.9 to 4.42 X 1O-.2 mini ’ at pH 6.8. Ligand exchange rate of Cr” A in Cr(H,O),ATP also increases sharply with increasing pH 1201. The inset in Figure 3 shows the logarithm of the rate of reaction plotted against pH with the enzymes previously inactivated in the presence and absence of 3-PGA, which yielded lines with slopes of 0.84 and 1.0, respectively. These observations suggest that Cr-ATP is attached to the enzyme by a single coordination both in binary enzyme-Cr-ATP and ternary enzyme-Cr-ATP3-PGA complexes. Although Cr(NH,),ATP inactivated the enzyme with a much lower rate compared to Cr(H,O),ATP, the reactivation rate remained unchanged with the Cr(NH,),ATP-inactivated enzyme. This finding strongly suggests that Crii acquires the same ligand from the enzyme and, therefore, regain of activity occurs with the same rate when compared to the rate observed with
SUBSTRATE
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
211
75
3 ,> -3
50
Pp
25
-I
0
400 Incubation
800 Time
1200
1600
(min)
FIGURE 3. (A) Reactivation of Cr(H,O),ATP inactivated-phosphoglycerate kinase. 187.5 PM PGK was inactivated with 5.07 mM Cr(H,O),ATP in the presence of 3.37 mM 3-PGA as described in Figure 1. Inactivated enzyme was separated from excess. reagents by centrifuge column as described in Experimental Procedures and diluted 51-fold into 50 mM MES, pH 5.9. The activity of the enzyme was measured as a function of time and expressed as the percent of the enzyme activity treated exactly the same way with the omission of Cr(H,O),ATP. Dilution media contained: (0) no additions, (0) 2.79 mM 3-PGA, (A) 40 mM Na,SO,, (0) 40 mM Na,SO, + 2.79 mM 3-PGA, (+) 9.98 mM MgCl,, (X) 2.26 mM MgATP. The curves represent fits to the first-order equation for the recovery of the enzyme activity. (B) pH Dependent reactivation of Cr(H,O),ATP inactivated-PGK. Inactivation of PGK was performed as described in Figure 2. After removal of excess reagents, the enzyme was diluted in 50 mM MES buffer at indicated pH and the recovery of enzyme activity was followed as a function of time; (0) pH 5.9, (0) pH 6.4, (A) pH 6.8. The inset shows the plot of logk,,, for reactivation vs pH of the enzyme inactivated in the presence ( q) and absence (0) of 3-PGA which yielded lines with the slopes of .84 and 1.0, respectively.
Cr(H,O),ATP-inactivated enzyme. Additionally, the presence of 3-PGA during the inactivation of enzyme with both of the Cr-ATP complexes did not cause any change of rate in the recovery of enzyme activity, suggesting an unchanged number of enzyme ligands to Cr3+ in the ternary enzyme-3-PGA-Cr-ATP complex when compared to the binary enzyme-Cr-ATP complex. Since binding of both substrates causes significant conformational changes in the enzyme [ll-131, these findings also suggest that reactivation is not dependent on the conformational state of the enzyme. DISCUSSION Exchange-inert /3,y-bidentate Cr3+ -ATP complex is an analog of Mgzf-ATP complex and binds to Mg-ATP sites in enzymes [14-201. The results presented in this paper strongly suggest that exchange-inert complexes of Cr-ATP inacti-
212
E. H. Serpersu et al.
80 70 60 50 .$ .-
40
-G 6
30
Ki
20
PH
0 FIGURE 3.
50 100 Incubation
150 200 Time (min)
250
(Continued)
vate phosphoglycerate kinase by forming an inner sphere complex with the protein at the active site. The observed K, values for the inactivation of PGK with various Cr-ATP complexes are in agreement with the Ki value measured from the initial rates. The slow inactivation of PGK by Cr(H20),ATP suggests that in the inactive enzyme-Cr-ATP complex, Cr’ ’ of the Cr-ATP complex has acquired a ligand from the enzyme. In bidentate Cr-ATP complex, when the four remaining from Hz0 to NH,, the observed ligands of Cr3 * are changed progressively inactivation rates become proportional to the number of NH, ligands (Fig. 1 and Table 1). Since NH, is much harder to remove from Cr’+. compared to HzO, substitution of NH, for H,O would make it harder for the enzyme to donate a liganding group to Cr’ ’ . A similar trend was observed in the inactivation of Na,K ATPase by the same Cr-ATP complexes [18]. Also studies with various kinases using Cr(H,O),(NH,),ATP showed that K, increased in direct proportion to y/x [lo]. It was suggested that when Cr-ATP does not act as substrate, inner sphere coordination of enzyme to Cr3- is likely to occur [IS]. Studies using U-[‘4C]-PGA showed that bidentatc Cr-ATP is not a substrate for yeast phosphoglycerate kinase [27]. EPR studies using MnADP also suggested that cr and /3 phosphate groups of ADP are coordinated to Mn’ r and an additional oxygen ligand is in the coordination sphere of Mn” in the ternary enzyme. Mn’ ’ . ADP complex [2Y]. However, as the authors stated, metal-ADP complexes do not induce closure of the enzyme active site when 3PGA is present and therefore Mn-ADP may not mimic metal-ATP in the enzyme-substrate(s) complexes. Perhaps most convincing evidence for the coordination of metal to the enzyme is the observation of a stable Cr-ATP-enzyme complex in the acid denatured enzyme as described in this work. Another less likely explanation may be the conversion of bidentate Cr-ATP into tridentate Cr-ATP
SUBSTRATE
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
213
at the active site. This process would also require a ligand displacement from Cr3+ and it would be slower with ammonia ligands. However, the fact that stoichiometric Cr-ATP remains bound to the acid denatured enzyme argues strongly against this possibility. Since the recovery of the enzyme activity is independent of the presence of substrates and/or the effector sulfate ion (Fig. 3A), a slow conformational change in the enzyme is also unlikely to be the reason for the observed inactivation of the enzyme by Cr-ATP. The inactivation of phosphoglycerate kinase by Cr-ATP complexes can occur in the absence or presence of the second substrate 3-PGA. However, the presence of 3-PGA in the inactivation medium lowers the observed inactivation rates. The effect of 3-PGA on k,, increases proportional to y/x in Cr(H,O),(NH,),ATP (Table 2). This observation suggests that either another ligand may be entering the inner coordination sphere of Cr3+ or both substrates are in close contact which causes an overlap when the active site is in “closed” conformation. We have previously observed such changes in the coordination sphere of metal ion at the active site of staphylococcal nuclease upon binding of substrate analog [301. It was suggested that binding of both substrates to phosphoglycerate kinase induces the formation of the “closed” active site [lo, 111. If this is true, then within the smaller volume of the “closed” active site the substrates may have to be closely packed. This may require displacement of a water molecule from the metal during the formation of the “closed” active site to permit the approach of 3-PGA to metal-ATP. Since Cr3+ exchanges its ligand very slowly [14, 151 the exchange of a ligand of Cr3+ may be reflected in the slower inactivation rates observed in the presence of 3-PGA. Indeed, since the displacement of NH, ligand of Cr3+ is more difficult when compared to water ligand, the ratio kobs/kobsPGA increases proportionally to the y/x ratio in Cr(H,O),(NH,),ATP (Fig. 1). However, other explanations are also possible. One may suggest that binding of 3-PGA to the enzyme may alter conformation of the active site such that it will affect inactivation by Cr-ATP. Examination of Tables 1 and 2 indicates that there are no systematic changes in the K, values determined with and without 3-PGA which argues against significant conformational changes at the Cr-ATP binding site. Another interesting possibility is the formation of a single turnover product CrADP-1,3-dPGA (or CrADP. Pi * 3-PGA) complex at the active site, but in an earlier study M. Cohn, et al. failed to observe formation of such complexes with Cr(H,O),ATP and U-[i4C]-PGA [27]. Sulfate ion, when present with 3-PGA, did not alter the observed inactivation rate of enzyme by Cr-ATP. Since the presence of sulfate ion decreases the affinity of substrates to the enzyme and affects the kinetics of the reaction, it must alter the conformation of the active site. Our findings suggest that in the presence of sulfate ion 3-PGA may form a second sphere complex with Cr3+ of Cr-ATP. The enzyme-active site then must be in such a conformation that does not necessitate such close approach of substrates to each other. This conformation may leave the active site partially open with substrates further apart from each other. Indeed, Rao et al. [41 have observed a decreased rate of phosphoryl transfer from ATP to 3-PGA in the presence of sulfate ion by 3’P NMR studies of enzyme-bound substrates. These observations are in excellent agreement with our findings. An alternative explanation is that sulfate and 3-PGA may bind to the same site and inhibit the inactivation by a different mechanism since it was shown that
214 E. H. Serpersu
et al.
sulfate ion is a competitive inhibitor with respect to both substrates at high concentrations [7]. Although the use of higher concentration of 3-PGA in the inactivation didn’t show any additional effect, our data cannot rule out this possible alternative. However, our binding studies also suggest that the closing of the active site may not be complete when sulfate is present (J. D. Gregory, and E. H. Serpersu, manuscript submitted). Previous binding studies with PGK and substrates and substrate analogs yielded 2 moles of bound nucleotide per mole of enzyme [2-51. One of the enzyme-bound nucleotide molecules displayed much higher affinity for Mg’+ [4, 311. It was therefore suggested that this was the ATP molecule at the active site. In our study, the stoichiometric binding of Cr(H20),ATP to the enzyme during the inactivation clearly indicates that metal-ATP complex is binding to the active site of PGK (Fig. 2). An additional piece of evidence for the binding of Cr-ATP complexes to the active site is the observation of the stereoselectivity of PGK for A diastereomer of Cr(H,O),ATP which agrees well with the previous observation that corresponding S, isomer of MgATP/?S is highly preferred over R, isomer by the enzyme for catalysis [27]. The rates observed from the reactivation studies yielded no difference between the enzymes inactivated with tetraammine or tetraaquo Cr-ATP complexes. This finding suggests that Cr3+ in all Cr-ATP complexes acquires the same type of ligand from the enzyme, therefore during the reactivation its displacement by water molecules will occur at the same rate with all of the Cr-ATP complexes. Although the presence of 3-PGA in the inactivation medium affected the inactivation rate, it did not alter the reactivation rate of enzyme when the enzyme was inactivated in the presence of 3-PGA. Therefore, we suggest that 3-PGA is coordinated to the metal ion. Otherwise, if the presence of 3-PGA at the active site causes another ligand from the enzyme to enter in the coordination sphere of Cr” +, this should slow the reactivation of the enzyme simply due to the necessity to replace an additional ligand from the enzymebound metal ion. This was not the case however; therefore 3-PGA is most likely to be the second ligand entering the coordination sphere of Cr” ’ in the ternary enzyme-Cr-ATP-3-PGA complex. Additionally, the slopes of the lines in Figure 3B (inset) clearly suggest that the reactivation of the enzyme requires only one hydroxyl ion in both cases. This also suggests that Cr’ has only one ligand from the enzyme in both binary enzyme-Cr-ATP and ternary enzyme-Cr-ATP-3-PGA complexes and interacts closely with 3-PGA at the active site. The fact that Cr-ATP complexes can bind to the enzyme in the absence of the second substrate 3-PGA with an essentially unaltered Ki value agrees with the random binding of substrates to the enzyme. These findings clearly indicate differences in the arrangement and/or interactions of the substrates at the active site of PGK in the presence and absence of activator sulfate ion. In the absence of sulfate, it seems that substrates are very close to each other, which could not be predicted from the x-ray studies because the enzyme crystals were obtained in the presence of a high concentration of sulfate. We thank Dr. JoTe Churchich of this department for allowing 14s to use his HPLC and we also thank Mr. Oh-Shin Kwon for help in using the HPLC. We are also grate@ to Dr. Jefiey Becker and Mr. Jack R. Perry of the Microbiology department for help in handling yeast strains and plasmids. This work was supported b_y National Institutes of Health Grant No. R29 GM422661 (to EHS).
SUBSTRATE BINDING TO YEAST PHOSPHOGLYCERATE
KINASE
215
REFERENCES 1. T. Bucher, Natwwissenschaflen 30, 756 (1942). E. Wicksell and M. Larsson-Raznikiewicz, .I. Biol. Chem. 262, 14472 (1987). J. A. Wrobel and R. A. Stinson, Eur. J. Biochem. 85, 345 (1978). B. D. Nageswara Rao, M. Cohn, and R. K. Scopes, J. Biol. Chem. 253, 8056 (1978). R. K. Scopes, Eur. J. Biochem. 91, 119 (1978). R. K. Scopes, Eur. J. Biochem. 85, 503 (1978). M. M. Khamis and M. Larsson-Raznikiewicz, Biochim. Biophys. Acta 657, 190
2. 3. 4. 5. 6. 7.
(1981). 8. 9.
10.
11.
H. C. Watson, P. L. Wendell, and R. K. Scopes, J. Mol. Biof. 57, 623 (1971). H. C. Watson, N. P. C. Walker, P. J. Shaw, T. N. Bryant, P. L. Wendell, L. A. Fothergill, R. E. Perkins, S. C. Conroy, M. J. Dobson, M. F. Tuite, A. J. Kingsman, and S. M. Kingsman, EMBO J. 1, 1635 (1982). R. D. Banks, C. C. F. Blake, P. R. Evans, R. Haser, D. W. Rice, G. W. Hardy, M. Merrett, and A. W. Phillips, Nature 279, 773 (1979). C. A. Pickover, D. B. McKay, D. M. Engelman, and T. A. Steitz, J. Biol. Chem. 254,
12.
11323 (1979). C. Roustan, A. Fattoum, (1980).
13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
P. Tanswell, E. W. Westhead, and R. J. P. Williams, Eur. J. Biochem. 63,249 (1976). W. W. Cleland and A. S. Mildvan, Adu. Znorg. Biochem. 1, 163 (1979). W. W. Cleland, Methods Enzymol. 63, 103 (1979). C. A. Janson and W. W. Cleland, J. Biol. Chem. 249, 2567 (1974). E. H. Serpersu, U. Kirch, and W. Schoner, Eur. J. Biochem. 122, 347 (1982). E. H. Serpersu, S. Bunk, and W. Schoner, Eur. J. Biochem. 191, 397 (1990). D. Dunaway-Mariano and W. W. Cleland, Biochemistry 19, 1506 (1980). D. Dunaway-Mariano and W. W. Cleland, Biochemistry 19, 1496 (1980). M. L. DePamphilis and W. W. Cleland, Biochemistry 12, 3714 (1973). K. D. Danenberg and W. W. Cleland, Biochemistry 14, 28 (1975). K. J. Gruys and S. M. Schuster, Anal. Biochem. 125, 66 (1982). H. S. Penefsky, J. Biol. Chem. 252 2891 (1977). J. F. Morrison and C. T. Walsh, Adu. Enzymol. 61, 201 (1988). R. Shapiro and J. F. Riordan, Biochemistry 23, 5234 (1984). E. K. Jaffe, J. Nick, and M. Cohn, J. Biol. Chem. 257, 7650 (1982). F. Eckstein, Ann. Rev. Biochem. 54, 367 (1985). J. M. Moore and G. H. Reed, Biochemistry 24, 5328 (1985). E. H. Serpersu, J. McCracken, J. Peisach, and A. S. Mildvan, Biochemistry 27,8034
R. Jeanneau,
and L. A. Pradell, Biochemistry
19, 5168
(1988).
31.
B. D. Ray, J. M. Moore, and B. D. Nageswara Rao, J. Znorg. Biochem. 40,47 (1990).
Received March 9, 1992; accepted April 28, 1992
ABSTRACT Exchange-inert p,y-bidentate Cr(H,O),(NH,),ATP complexes inactivate yeast phosphoglycerate kinase (PGK) by forming a coordination complex at the enzyme active site. The observed inactivation rates ranged from 0.019 min-’ to 0.118 min- ’ for Cr(NH,),ATP and Cr(H,O),ATP, respectively. Incorporation of one mol of Cr-ATP to the enzyme was sufficient for complete inactivation of the enzyme. The presence of Mg-ATP protected the enzyme against inactivation by Cr-ATP. The other substrate 3-phosphoglycerate (3-PGA), when present, reduced the observed inactivation rates. The reduction of the k,, by 3-PGA was proportional to the number of NH, ligands present in the coordination sphere of Cr3+ in the Cr-ATP complex, suggesting that in the ternary enzyme-Cr-ATP-3-PGA complex 3-PGA may be coordinated to the metal ion. When the effector sulfate ion was present, the presence of 3-PGA did not cause any further effects on the observed inactivation rates. This suggests that bound substrates are in a different arrangement at the active site when sulfate is present and therefore 3-PGA may not need to displace a ligand from Cr3+. Additionally, PGK exhibited a stereoselectivity for the binding of Cr(H,O),ATP. A diastereomer of Cr(H,O),ATP yielded an order of magnitude smaller Ki value compared to the value observed with the A isomer. The recovery of enzyme activity was observed over a period of a few hours upon removal of excess Cr-ATP. The presence of substrates and/or effector ion sulfate did not alter the observed reactivation rate. There was no difference in the reactivation rates of the enzyme which was inactivated with Cr(H,O),ATP or Cr(NH,),ATP with and without 3-PGA. Increasing the ligand exchange rates of Cr3+ of Cr-ATP by increasing the pH value of the recovery medium from 5.9 to 6.8 increased the rate of recovery by a factor of 8. The pH dependence of the reactivation indicated that one hydroxyl group is involved in the recovery of the enzyme activity in enzyme. CrATP and enzyme. CrATP .3-PGA complexes.
Address reprint requests and correspondence to: Engin H. Serpersu, Department of Biochemistry, Walters Life Sciences Building M-408, University of Tennessee, Knoxville, TN 37996-0840. Journal of Inorganic Biochemistry, #3,203-215 (1992) 0 1992 Elsevier Science Publishing Co., Inc., 655 Avenue
203 of the Americas,
NY, NY 10010
0162-0134/92/%5.00
204
E. H. Sepersu
et al.
ABBREVIATIONS 3-PGA, 3-phospho-D-glycerate; PGK, yeast phosphoglycerate kinase (EC2.7.2.3); MES, 2-(N-morpholino)ethanesulfonic acid; 1,3-dPGA, 1,3-diphospho-D-glycerate; ATP, 5’-adenosine triphosphate.
INTRODUCTION Phosphoglycerate
kinase (EC 2.7.2.3) catalyzes the phosphoryl transfer reaction
[II: Mg2+ 1,3-diphospho-D-glycerate
+ ADP + 3-phospho-D-glycerate
+ ATP.
binds two moles of ATP per mole [2-51, one of which is bound to the anion binding site and has very low affinity to Mg2’ [4]. Catalytically required Mg2+ is presumably bound to ATP at the active site. Multivalent anions and particularly sulfate ion are known to interact with phosphoglycerate kinase and affect the kinetics of the reaction [5-71. Sulfate ion in the range l-20 mM causes an increase in the observed reaction rate with an increase in K, of both substrates. The presence of a second binding site distinct from the active site has been suggested by previous studies. However, the nature and proximity of this binding site to the active site are not known. Crystallographic studies have been performed with the enzyme crystals obtained in the presence of 2 M or higher concentrations of ammonium sulfate 18, 91. These studies are tentative because ofothe problem that the phosphor-y1 group to be transferred from Mg-ATP is N 10A from the accepting carboxylate group of 3-PGA [lo]. Small-angle x-ray scattering studies of 3-PGK in solution revealed that the radius of gyration of the enzyme decreases when both substrates are bound to the enzyme [ll]. Additional solution studies also suggested that binding of both substrates to the enzyme induces large conformational changes on the enzyme [ll-131. Therefore, a hinge-bending motion of the enzyme to bring two substrates together in a “closed” active site was suggested for the catalytic mechanism of the enzyme [lo, 111. However, the nature and the extent of the interactions between the substrates and the enzyme is not known. We have started structural studies in solution with binary and ternary enzyme-substrate(s) complexes by using stable exchange-inert Cr 3+-ATP complexes as Mg-ATP analogs 114, 151. Cr-ATP complexes have been shown to be classical competitive inhibitors and inactivators of various kinases and ATPases [14-191. This paper describes inactivation of yeast phosphoglycerate kinase by various Cr-ATP analogs. Our results suggest that Cr-ATP mimics the substrate Mg-ATP and forms a ternary enzyme-Cr-ATP-3-PGA complex at the active site. The slow inactivation of the enzyme by various Cr-ATP complexes allowed more detailed characterization of the ternary enzyme-metal-ATP-3-PGA complex in the presence and absence of effector sulfate ion. The enzyme
SUBSTRATE
EXPERIMENTAL
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
205
PROCEDURES
Materials Various P,ybidentate Cr(H,O),(NH,),,ATP complexes were prepared as described earlier [15, 20-221. The separation of A and A diastereomers of Cr(H,O),ATP was done as described 1231. Experiments were performed with the racemic mixture except where specified. Yeast phosphoglycerate kinase was purchased from Boehringer. The enzyme was also isolated from yeast strain 20B-12 containing multi-copy plasmid pCGY219. The plasmid and the yeast strain were kindly provided by Dr. Hitzeman of Genentech Inc. Both enzymes had k,,, of N 180 see-’ at pH 5.9. Enzyme concentration was determined from its absorbance at 280 nm (Elc,‘.i% = 0.5). Methods Inactivation of the Enzyme. The enzyme (30-40 pg) was incubated in 50 mM MES pH 5.9 at 30°C in a total volume of 200 ~1 containing 150 mM NaCl and various concentrations of Cr-ATP. This pH was selected because of the instability of Cr-ATP complexes at higher pH 115, 201. In some cases, the second substrate 3-phosphoglycerate (2.7 mM) and/or sulfate ion (20 mM) were also present in the incubation medium. When sulfate ion was present ionic strengths of the solutions were adjusted by lowering NaCl concentration as necessary. At indicated time points, 5-8 @’ aliquots were withdrawn and the remaining enzyme activity was determined spectrophotometrically at 340 nm by measuring the initial rates in a coupled enzyme assay in a total volume of 1 ml at 25°C. The assay medium contained 150 mM NACI, 0.45-0.51 mM ATP, 1.5 mM MgCl,, 8 mM 3-phosphoglyceric acid, 0.2 mM NADH, and 150 pg glyceraldehyde 3-phosphate dehydrogenase in 50 mM MES pH 5.9. With the indicated amount of coupling enzyme, from which (NH&SO, was removed by dialysis or desalting, there was no lag in the reaction and the rate observed was linearly dependent on the concentration of phosphoglycerate kinase present. The recovery of enzyme activity upon dilution into the assay cuvette was insignificant and no deviation from linearity was observed for the duration of the assays (45-60 set). Studies with [y 32P]Cr(H,0)4 ATP. Inactivation of phosphoglycerate kinase with [y-32P]Cr(H20)4ATP (25-35 cpm/pmole) was performed in a similar manner as described above with the exception that the enzyme concentrations were 216-250 PM. At indicated times, two aliquots were withdrawn from the inactivation mixture and diluted in a medium containing 50 mM MES pH 5.9, 150 mM NaCl, 7-8 mM 3-PGA, 6 mM ATP, and 7 mM MgCl,; 90 ~1 of 11-fold diluted sample was then applied to a centrifuge column [24] containing Sephadex G-50 in a 1 ml tuberculin syringe and centrifuged at 180 X g for 90 sec. Enzyme-bound radioactivity was then determined by liquid scintillation counting of the effluent. The highest concentration of free [y-32P]Cr-ATP used in the assays did not penetrate more than the upper third of the column length in our conditions. Thus no background counts due to unbound Cr-ATP were observed. Another aliquot from an appropriately diluted second sample was used for the determination of the residual enzyme activity as described above.
206
E. H. Serpersu et al.
Reactivation Studies. Phosphoglycerate kinase (170-190 PM) was inactivated with Cr-ATP (4-5 mM) in the presence or absence 2.70 mM 3-PGA as described above. When the steady-state level (15-20% of the original activity) was reached, 100 ~1 aliquots were applied to the centrifuge columns to separate the enzyme from the excess free Cr-ATP. This enzyme solution was then immediately diluted into 50 mM MES buffer at indicated pH to give an enzyme concentration of 2.57 PM. This solution contained various combinations of substrates/ions as indicated in the figure legends. Diluted enzyme solution was then left at 23°C and at indicated time the enzyme activity was determined as described above.
RESULTS Inactivation of Phosphoglycerate Kinase by Cr-ATP. Incubation of 1 PM enzyme with various concentrations of Cr(H,O),ATP showed a time dependent pseudo-first-order loss of enzyme activity when the residual enzyme activity was determined in the presence of saturating 3-PGA and 3.3 K,, Mg-ATP or 20 K, Mg-ATP (which yielded identical results-data not shown). The inactivation is slow and reaches a steady-state after a few hours. Increasing concentrations of Cr(H,O),ATP resulted in increased rate of inactivation and in decreased levels of residual enzyme activity. Since there is no appreciable dissociation of enzymeCr(H,O),ATP complex during the assay for residual activity (45-60 XC), a long-lived enzyme-Cr(H,O),ATP complex would account for this kinetic behavior. Since Cr3+ exchanges its ligands very slowly [14, 151, it is likely that this long-lived complex may form by donation of a ligand from the enzyme to the coordination sphere of Cr- ‘+ in enzyme-Cr(H20),ATP complex. The pseudofirst-order rate constants (kobs) for each concentration of Cr(Hz0)4ATP were determined and plotted in double reciprocal form (l/k,,h, vs l/[Cr-ATP]) which is linear and yields a finite inactivation rate at saturating Cr(H20),ATP concentration (Fig. 1) which is consistent with the following mechanism [25, 261: E+IsEI k,
2 EI* k,
(I)
in which enzyme (E) and inhibitor (I) rapidly form EI which then slowly isomerizes to EI* [25, 261. The observed K, (161 PM) is in reasonable agreement with the Ki value obtained from the initial rate studies which showed that Cr(H,O),ATP is a linear competitive inhibitor of phosphoglycerate kinase with a Ki = 70 PM with respect to Mg-ATP (J. D. Gregory and E. H. Serpersu, manuscript submitted). As expected, the presence of increasing concentrations of Mg-ATP provided increasing levels of protection of the enzyme against inactivation by Cr(H,O),ATP. The presence of 125 PM Mg-ATP was enough to reduce the observed inactivation rate by a factor of two with 246 PM Cr(Hz0)4ATP present (data not shown). To test the hypothesis that Cr 3+ indeed acquires a ligand from the enzyme in the enzyme-Cr(H20),ATP complex, we have prepared a series of Cr(H,O),(NH,),ATP complexes [20-221. Since displacement of NH, from the coordination sphere of Cr3+ is more difficult than the displacement of HzO, these analogs should inactivate the enzyme with progressively decreasing koh\
SUBSTRATE
-24
BINDING
-16
TO YEAST
PHOSPHOGLYCERATE
0
-8
8
l/[Cr-ATPI
16
KINASE
207
24
mM--1
FIGURE 1. Inactivation of Yeast PGK by various Cr-ATP complexes. 1.0 PM enzyme was incubated at 30°C in 50 mM MES, pH 5.9, containing 150 mM NaCl and various concentrations of the Cr-ATP complexes. The data is presented in double reciprocal form: (0) Cr(H,O),ATP, (0) Cr(H,O),(NH,),ATP, (A> Cr(H,OXNH,),ATP, and (+> Cr(NH,),ATP. The lines represent weighted fits to the data points. The inset shows the three-dimensional plot of the effects of NH, coordination to Cr3+ and 3-PGA on the observed inactivation rates of PGK by Cr-ATP complexes.
values proportional to the increased number of NH, ligands in the coordination sphere of Cr 3+. As is shown in Table 1, this is indeed the case, where the inactivation of PGK by various Cr(H,O),(NH,) ATP complexes revealed that decreases systematically the determined kobs = 0.118 min-’ with Cr(H,O$,ATP as the amino to a 6.2-fold lower value of k,, = 0.019 min-’ with Cr(NH,),ATP groups replace water molecules for coordination to Cr3+ in Cr-ATP complexes. Small changes in the determined Ki values suggest that replacement of H,O with NH, did not have significant effects on formation of the initial EI complex in Eq. (1). However, a long-lived enzyme * Cr-ATP *substrate complex can form without the formation of a covalent bound [22]. To distinguish these possibilities,
TABLE 1. Inactivation of Yeast Phosphoglycerate Kinase by Various Cr-ATP Complexes in the Absence of 3-Phosphoglycerate at pH 5.9 Cr-ATP Complex CXH20,4 ATP Cr(H,O),(NH&4’II’ Cr(H,OXNH,)@F’ Cr(NH,),ATP
kobs
(min-‘) ,118 .068 ,040 .019
+ + + *
,012 .OlO ,006 .003
161 161 145 93
+ f f f
22 31 20 12
208
E. H. Serpersu et al.
we incubated the enzyme with y-[32P]-Cr-ATP. When the enzyme was 76.4% inactivated (23.6% remaining activity) the reaction was stopped by the addition of HClO,, at a final concentration of 4%. The denatured enzyme was then washed several times with buffer solution to remove unbound Cr-ATP. The radioactivity associated with the enzyme yielded 0.71 mole Cr-ATP/mol enzyme stoichiometry. This strongly suggests formation of a bond between the enzyme and Cr-ATP. The Effect of 3-Phosphogfyceric Acid. As can be seen from Table 2, when the inactivation of phosphoglycerate kinase was performed in the presence of 2.70 mM 3-PGA, the observed rates of inactivation were lower compared to the rates observed in its absence, with a small effect on K,. Additionally, the reduction in with the increasing number of NH, in the coordinak obs was more pronounced tion sphere of Cr 3+ in the Cr-ATP, which is presented in Table 2 as the ratios of the rates of inactivation in the absence and presence of 3-PGA (kobs/kobsPGA). is 1.36 which increases to 2.5 with The ratio of kobs/kobsPGA with Cr(H,O),ATP Cr(NH,),ATP. The effects of 3-PGA and NH, coordination to Cr3’ is shown in Figure 1 as a three-dimensional plot. These observations suggest that during the formation of the ternary enzyme-3-PGA-Cr-ATP complex, either a liganding group from 3-PGA molecule or another ligand from the enzyme is entering to the coordination sphere of the metal ion. This would necessitate displacement of another ligand from Cr3+ which would then reduce the observed inactivation rate due to the slower ligand exchange rates of Cr”+. The Effects of Sulfate and Mg Ion. The kinetics of the phosphoglycerate kinase reaction have been found to be affected by the presence of sulfate ion and the observed K, values for the substrate increase [6, 71. Therefore, we have tested the effect of sulfate ion on the inactivation of PGK by Cr-ATP complexes. The presence of sulfate ion resulted in a decrease in kobs to values similar to those observed with 3-PGA when the enzyme was inactivated with Cr-ATP. In the presence of sulfate ion, the addition of 2.7 mM 3-PGA did not cause any further changes in the observed rate of inactivation (data not shown). This was the case with all of the Cr(H,O),(NH,),ATP complexes, which may suggest that in the presence of sulfate ion bound substrates have different arrangements at the active site. Hence no additional ligands of the metal have to be replaced by a liganding group from 3-PGA or from the enzyme. Alternatively, sulfate ion and 3-PGA may bind to the same site on the enzyme and prevent the inactivation of
TABLE 2. Inactivation of Yeast Phosphoglycerate Kinase by Various Cr-ATP Complexes in the Presence of 3-Phosphoglycerate at pH 5.9 Cr-ATP Complex Cr(H,O),ATF’ Cr(H,0)2(NH,),ATP Cr(H20XNH,),ATP Cr(NH,),ATP The column k ,,,,/k,,,(PGA) by Cr-ATP in the absence tively.
k ohs (min-‘1 ,087 ,030 .018 .008
+ .Ol1 + ,005 + ,004 + .OOl
kobs/kobs
(PGA) 145 + 465 f 267 f 92+
21 106 41 18
1.36 2.24 2.26 2.50
shows the ratio of the observed inactivation rates (kubJ of the enzyme and presence of saturating concentration of 3-PGA (2.7 mM), respec-
SUBSTRATE
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
209
the enzyme by different mechanisms. It was previously shown that sulfate activates the enzyme at low concentration and it is a competitive inhibitor at high concentrations 171. The presence of Mg2+ (up to 10 mM) did not alter the observed inactivation rates regardless of the composition of the inactivation medium. This observation suggests that in the formation of the ternary complex no additional metal ion is needed, when there is a metal ion already bound to ATP (Cr3+ in this case). Stoichiometry of Cr(H,O), ATP Binding to PGR Incubation of phosphoglycerate kinase with [ y-32P]Cr(H20),ATP resulted in the stoichiometric labeling of the enzyme (Fig. 2). As shown in Figure 2, labeling of the enzyme was proportional to the inactivation of the enzyme and no further incorporation of radioactivity was observed when steady-state levels of enzyme activity were reached. The inset in Figure 2 shows the correlation of the loss of activity with the stoichiometry of the incorporation of [32P]Cr(H20)4ATP, which indicates binding of one mol of Cr(H,O), ATP to one mol of enzyme. When Mg-ATP was present, incorporation of radioactivity to the enzyme was prevented (Fig. 2). This is consistent with binding of Cr(H,O),ATP to the active site of the enzyme. In all cases the presence of 20 mM sulfate ion in the inactivation medium did not alter the observed Cr(H,O),ATP/enzyme stoichiometry (data not shown).
--
0.8
e
80
z 0.6 L
3
60:’
2
0
50
100
Incubation
Time
150
200
(mid
FIGURE 2. Binding of [-Y-~*P]C~(H,O),ATP to Yeast PGK during inactivation. The inactivation conditions were as described in Experimental Procedures. Open symbols represent the time dependent incorporation of radioactivity to the enzyme (right axis) and filled symbols represent the relative enzyme activity with respect to the initial activity. Activity measurements of some samples were omitted from the plot for simplicity. (O,O) 0.28 mM Cr(H,O),ATP with 4.05 mM 3-PGA, (0, H) 1.12 mM Cr(H,O),ATP with 4.05 mM 3-PGA, (A) 1.08 mM Cr(H,O),ATP, (0) 1.08 mM Cr(H,Ol,ATP with 6.48 mM MgATP. Curve fitting was done as described in Figure 1. In the inset the data for the inactivation of the enzyme in the presence of 3-PGA is plotted in relative activity vs stoichiometry of the incorporated label. V, and V, represent the measured enzyme activity at time t and “zero”, respectively; (0) .28 mM Cr(H,O),ATP and (0) 1.12 mM Cr(H,O),ATP. The lines represent the least-square fits to the data points.
210
E. H. Serpersu et al.
Stereoselectivity of Phospho&cerate Kinase. When individual diastereomers of /3,-y-bidentate Cr(H,O),ATP were used to inactivate phosphoglycerate kinase, both diastereomers yielded kobs values similar to the value observed with racemic mixture. However, the A diastereomer yielded a Ki value of 7.0 PM which is about an order of magnitude smaller than that of the A diastereomer (62.6 PM). In addition, we have inactivated the enzyme with a racemic mixture of Cr(H,O),ATP and removed the enzyme with bound nucleotide by ultrafiltration. The analysis of the remaining solution by CD spectroscopy revealed that the A diastereomer was enriched in the solution, indicating a preferential binding of A diastereomer to the enzyme. These observations are in agreement with the earlier findings with sulfur substituted (at the P-phosphorus of ATP) Mg-ATP analogs, which yielded that PGK is highly specific for S, isomer [271. Screw-sense A isomer of Cr(H,O),ATP corresponds to the S, isomer of Mg-ATPP S [28]. Recovery of the Enzymatic Activity After Preincubation with Cr-ATP. Although we observed no recovery of the catalytic activity for the duration of the enzymatic assays, when inactivated enzyme is separated from the excess Cr(H20),ATP and diluted in buffer with or without the substrates, a slow recovery of the enzymatic activity was monitored over a several-hour period. In the presence of various combinations of substrates and effecters at pH 5.9, the recovery was 70 to 100% of the activity of the enzyme treated in exactly the same manner with the omission of inhibitor (Fig. 3A). The samples without Cr(H,Ol,ATP showed less than 5% loss of enzyme activity for the duration of the experiments t- 16 hr). Analysis of the data for the slow recovery of Cr(H,O),ATP-inactivated PGK activity by fitting to the first-order rate equation yielded an average rate constant of 5.5 x 1O--3 min-’ with t,,, = 140 min. The rate of regain of enzyme activity was not affected when substrates and/or sulfate was present in the recovery medium, suggesting that the conformational changes induced by substrates and/or sulfate ion have no effect on the recovery of enzyme activity. This would be expected if the recovery of the enzyme activity is the result of the displacement of the coordinated enzyme side chain from Cr” by water. Hence, the rate constant for the reactivation will remain unchanged regardless of the constituents of the recovery medium. As shown in Figure 3B, the rate of regain of enzyme activity was much faster when the inactive enzyme-Cr-ATP complex was diluted in buffer solutions at higher pH. The rate of reactivation increased g-fold from 5.5 x 10mm3min- ’ at pH 5.9 to 4.42 X 1O-.2 mini ’ at pH 6.8. Ligand exchange rate of Cr” A in Cr(H,O),ATP also increases sharply with increasing pH 1201. The inset in Figure 3 shows the logarithm of the rate of reaction plotted against pH with the enzymes previously inactivated in the presence and absence of 3-PGA, which yielded lines with slopes of 0.84 and 1.0, respectively. These observations suggest that Cr-ATP is attached to the enzyme by a single coordination both in binary enzyme-Cr-ATP and ternary enzyme-Cr-ATP3-PGA complexes. Although Cr(NH,),ATP inactivated the enzyme with a much lower rate compared to Cr(H,O),ATP, the reactivation rate remained unchanged with the Cr(NH,),ATP-inactivated enzyme. This finding strongly suggests that Crii acquires the same ligand from the enzyme and, therefore, regain of activity occurs with the same rate when compared to the rate observed with
SUBSTRATE
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
211
75
3 ,> -3
50
Pp
25
-I
0
400 Incubation
800 Time
1200
1600
(min)
FIGURE 3. (A) Reactivation of Cr(H,O),ATP inactivated-phosphoglycerate kinase. 187.5 PM PGK was inactivated with 5.07 mM Cr(H,O),ATP in the presence of 3.37 mM 3-PGA as described in Figure 1. Inactivated enzyme was separated from excess. reagents by centrifuge column as described in Experimental Procedures and diluted 51-fold into 50 mM MES, pH 5.9. The activity of the enzyme was measured as a function of time and expressed as the percent of the enzyme activity treated exactly the same way with the omission of Cr(H,O),ATP. Dilution media contained: (0) no additions, (0) 2.79 mM 3-PGA, (A) 40 mM Na,SO,, (0) 40 mM Na,SO, + 2.79 mM 3-PGA, (+) 9.98 mM MgCl,, (X) 2.26 mM MgATP. The curves represent fits to the first-order equation for the recovery of the enzyme activity. (B) pH Dependent reactivation of Cr(H,O),ATP inactivated-PGK. Inactivation of PGK was performed as described in Figure 2. After removal of excess reagents, the enzyme was diluted in 50 mM MES buffer at indicated pH and the recovery of enzyme activity was followed as a function of time; (0) pH 5.9, (0) pH 6.4, (A) pH 6.8. The inset shows the plot of logk,,, for reactivation vs pH of the enzyme inactivated in the presence ( q) and absence (0) of 3-PGA which yielded lines with the slopes of .84 and 1.0, respectively.
Cr(H,O),ATP-inactivated enzyme. Additionally, the presence of 3-PGA during the inactivation of enzyme with both of the Cr-ATP complexes did not cause any change of rate in the recovery of enzyme activity, suggesting an unchanged number of enzyme ligands to Cr3+ in the ternary enzyme-3-PGA-Cr-ATP complex when compared to the binary enzyme-Cr-ATP complex. Since binding of both substrates causes significant conformational changes in the enzyme [ll-131, these findings also suggest that reactivation is not dependent on the conformational state of the enzyme. DISCUSSION Exchange-inert /3,y-bidentate Cr3+ -ATP complex is an analog of Mgzf-ATP complex and binds to Mg-ATP sites in enzymes [14-201. The results presented in this paper strongly suggest that exchange-inert complexes of Cr-ATP inacti-
212
E. H. Serpersu et al.
80 70 60 50 .$ .-
40
-G 6
30
Ki
20
PH
0 FIGURE 3.
50 100 Incubation
150 200 Time (min)
250
(Continued)
vate phosphoglycerate kinase by forming an inner sphere complex with the protein at the active site. The observed K, values for the inactivation of PGK with various Cr-ATP complexes are in agreement with the Ki value measured from the initial rates. The slow inactivation of PGK by Cr(H20),ATP suggests that in the inactive enzyme-Cr-ATP complex, Cr’ ’ of the Cr-ATP complex has acquired a ligand from the enzyme. In bidentate Cr-ATP complex, when the four remaining from Hz0 to NH,, the observed ligands of Cr3 * are changed progressively inactivation rates become proportional to the number of NH, ligands (Fig. 1 and Table 1). Since NH, is much harder to remove from Cr’+. compared to HzO, substitution of NH, for H,O would make it harder for the enzyme to donate a liganding group to Cr’ ’ . A similar trend was observed in the inactivation of Na,K ATPase by the same Cr-ATP complexes [18]. Also studies with various kinases using Cr(H,O),(NH,),ATP showed that K, increased in direct proportion to y/x [lo]. It was suggested that when Cr-ATP does not act as substrate, inner sphere coordination of enzyme to Cr3- is likely to occur [IS]. Studies using U-[‘4C]-PGA showed that bidentatc Cr-ATP is not a substrate for yeast phosphoglycerate kinase [27]. EPR studies using MnADP also suggested that cr and /3 phosphate groups of ADP are coordinated to Mn’ r and an additional oxygen ligand is in the coordination sphere of Mn” in the ternary enzyme. Mn’ ’ . ADP complex [2Y]. However, as the authors stated, metal-ADP complexes do not induce closure of the enzyme active site when 3PGA is present and therefore Mn-ADP may not mimic metal-ATP in the enzyme-substrate(s) complexes. Perhaps most convincing evidence for the coordination of metal to the enzyme is the observation of a stable Cr-ATP-enzyme complex in the acid denatured enzyme as described in this work. Another less likely explanation may be the conversion of bidentate Cr-ATP into tridentate Cr-ATP
SUBSTRATE
BINDING
TO YEAST PHOSPHOGLYCERATE
KINASE
213
at the active site. This process would also require a ligand displacement from Cr3+ and it would be slower with ammonia ligands. However, the fact that stoichiometric Cr-ATP remains bound to the acid denatured enzyme argues strongly against this possibility. Since the recovery of the enzyme activity is independent of the presence of substrates and/or the effector sulfate ion (Fig. 3A), a slow conformational change in the enzyme is also unlikely to be the reason for the observed inactivation of the enzyme by Cr-ATP. The inactivation of phosphoglycerate kinase by Cr-ATP complexes can occur in the absence or presence of the second substrate 3-PGA. However, the presence of 3-PGA in the inactivation medium lowers the observed inactivation rates. The effect of 3-PGA on k,, increases proportional to y/x in Cr(H,O),(NH,),ATP (Table 2). This observation suggests that either another ligand may be entering the inner coordination sphere of Cr3+ or both substrates are in close contact which causes an overlap when the active site is in “closed” conformation. We have previously observed such changes in the coordination sphere of metal ion at the active site of staphylococcal nuclease upon binding of substrate analog [301. It was suggested that binding of both substrates to phosphoglycerate kinase induces the formation of the “closed” active site [lo, 111. If this is true, then within the smaller volume of the “closed” active site the substrates may have to be closely packed. This may require displacement of a water molecule from the metal during the formation of the “closed” active site to permit the approach of 3-PGA to metal-ATP. Since Cr3+ exchanges its ligand very slowly [14, 151 the exchange of a ligand of Cr3+ may be reflected in the slower inactivation rates observed in the presence of 3-PGA. Indeed, since the displacement of NH, ligand of Cr3+ is more difficult when compared to water ligand, the ratio kobs/kobsPGA increases proportionally to the y/x ratio in Cr(H,O),(NH,),ATP (Fig. 1). However, other explanations are also possible. One may suggest that binding of 3-PGA to the enzyme may alter conformation of the active site such that it will affect inactivation by Cr-ATP. Examination of Tables 1 and 2 indicates that there are no systematic changes in the K, values determined with and without 3-PGA which argues against significant conformational changes at the Cr-ATP binding site. Another interesting possibility is the formation of a single turnover product CrADP-1,3-dPGA (or CrADP. Pi * 3-PGA) complex at the active site, but in an earlier study M. Cohn, et al. failed to observe formation of such complexes with Cr(H,O),ATP and U-[i4C]-PGA [27]. Sulfate ion, when present with 3-PGA, did not alter the observed inactivation rate of enzyme by Cr-ATP. Since the presence of sulfate ion decreases the affinity of substrates to the enzyme and affects the kinetics of the reaction, it must alter the conformation of the active site. Our findings suggest that in the presence of sulfate ion 3-PGA may form a second sphere complex with Cr3+ of Cr-ATP. The enzyme-active site then must be in such a conformation that does not necessitate such close approach of substrates to each other. This conformation may leave the active site partially open with substrates further apart from each other. Indeed, Rao et al. [41 have observed a decreased rate of phosphoryl transfer from ATP to 3-PGA in the presence of sulfate ion by 3’P NMR studies of enzyme-bound substrates. These observations are in excellent agreement with our findings. An alternative explanation is that sulfate and 3-PGA may bind to the same site and inhibit the inactivation by a different mechanism since it was shown that
214 E. H. Serpersu
et al.
sulfate ion is a competitive inhibitor with respect to both substrates at high concentrations [7]. Although the use of higher concentration of 3-PGA in the inactivation didn’t show any additional effect, our data cannot rule out this possible alternative. However, our binding studies also suggest that the closing of the active site may not be complete when sulfate is present (J. D. Gregory, and E. H. Serpersu, manuscript submitted). Previous binding studies with PGK and substrates and substrate analogs yielded 2 moles of bound nucleotide per mole of enzyme [2-51. One of the enzyme-bound nucleotide molecules displayed much higher affinity for Mg’+ [4, 311. It was therefore suggested that this was the ATP molecule at the active site. In our study, the stoichiometric binding of Cr(H20),ATP to the enzyme during the inactivation clearly indicates that metal-ATP complex is binding to the active site of PGK (Fig. 2). An additional piece of evidence for the binding of Cr-ATP complexes to the active site is the observation of the stereoselectivity of PGK for A diastereomer of Cr(H,O),ATP which agrees well with the previous observation that corresponding S, isomer of MgATP/?S is highly preferred over R, isomer by the enzyme for catalysis [27]. The rates observed from the reactivation studies yielded no difference between the enzymes inactivated with tetraammine or tetraaquo Cr-ATP complexes. This finding suggests that Cr3+ in all Cr-ATP complexes acquires the same type of ligand from the enzyme, therefore during the reactivation its displacement by water molecules will occur at the same rate with all of the Cr-ATP complexes. Although the presence of 3-PGA in the inactivation medium affected the inactivation rate, it did not alter the reactivation rate of enzyme when the enzyme was inactivated in the presence of 3-PGA. Therefore, we suggest that 3-PGA is coordinated to the metal ion. Otherwise, if the presence of 3-PGA at the active site causes another ligand from the enzyme to enter in the coordination sphere of Cr” +, this should slow the reactivation of the enzyme simply due to the necessity to replace an additional ligand from the enzymebound metal ion. This was not the case however; therefore 3-PGA is most likely to be the second ligand entering the coordination sphere of Cr” ’ in the ternary enzyme-Cr-ATP-3-PGA complex. Additionally, the slopes of the lines in Figure 3B (inset) clearly suggest that the reactivation of the enzyme requires only one hydroxyl ion in both cases. This also suggests that Cr’ has only one ligand from the enzyme in both binary enzyme-Cr-ATP and ternary enzyme-Cr-ATP-3-PGA complexes and interacts closely with 3-PGA at the active site. The fact that Cr-ATP complexes can bind to the enzyme in the absence of the second substrate 3-PGA with an essentially unaltered Ki value agrees with the random binding of substrates to the enzyme. These findings clearly indicate differences in the arrangement and/or interactions of the substrates at the active site of PGK in the presence and absence of activator sulfate ion. In the absence of sulfate, it seems that substrates are very close to each other, which could not be predicted from the x-ray studies because the enzyme crystals were obtained in the presence of a high concentration of sulfate. We thank Dr. JoTe Churchich of this department for allowing 14s to use his HPLC and we also thank Mr. Oh-Shin Kwon for help in using the HPLC. We are also grate@ to Dr. Jefiey Becker and Mr. Jack R. Perry of the Microbiology department for help in handling yeast strains and plasmids. This work was supported b_y National Institutes of Health Grant No. R29 GM422661 (to EHS).
SUBSTRATE BINDING TO YEAST PHOSPHOGLYCERATE
KINASE
215
REFERENCES 1. T. Bucher, Natwwissenschaflen 30, 756 (1942). E. Wicksell and M. Larsson-Raznikiewicz, .I. Biol. Chem. 262, 14472 (1987). J. A. Wrobel and R. A. Stinson, Eur. J. Biochem. 85, 345 (1978). B. D. Nageswara Rao, M. Cohn, and R. K. Scopes, J. Biol. Chem. 253, 8056 (1978). R. K. Scopes, Eur. J. Biochem. 91, 119 (1978). R. K. Scopes, Eur. J. Biochem. 85, 503 (1978). M. M. Khamis and M. Larsson-Raznikiewicz, Biochim. Biophys. Acta 657, 190
2. 3. 4. 5. 6. 7.
(1981). 8. 9.
10.
11.
H. C. Watson, P. L. Wendell, and R. K. Scopes, J. Mol. Biof. 57, 623 (1971). H. C. Watson, N. P. C. Walker, P. J. Shaw, T. N. Bryant, P. L. Wendell, L. A. Fothergill, R. E. Perkins, S. C. Conroy, M. J. Dobson, M. F. Tuite, A. J. Kingsman, and S. M. Kingsman, EMBO J. 1, 1635 (1982). R. D. Banks, C. C. F. Blake, P. R. Evans, R. Haser, D. W. Rice, G. W. Hardy, M. Merrett, and A. W. Phillips, Nature 279, 773 (1979). C. A. Pickover, D. B. McKay, D. M. Engelman, and T. A. Steitz, J. Biol. Chem. 254,
12.
11323 (1979). C. Roustan, A. Fattoum, (1980).
13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
P. Tanswell, E. W. Westhead, and R. J. P. Williams, Eur. J. Biochem. 63,249 (1976). W. W. Cleland and A. S. Mildvan, Adu. Znorg. Biochem. 1, 163 (1979). W. W. Cleland, Methods Enzymol. 63, 103 (1979). C. A. Janson and W. W. Cleland, J. Biol. Chem. 249, 2567 (1974). E. H. Serpersu, U. Kirch, and W. Schoner, Eur. J. Biochem. 122, 347 (1982). E. H. Serpersu, S. Bunk, and W. Schoner, Eur. J. Biochem. 191, 397 (1990). D. Dunaway-Mariano and W. W. Cleland, Biochemistry 19, 1506 (1980). D. Dunaway-Mariano and W. W. Cleland, Biochemistry 19, 1496 (1980). M. L. DePamphilis and W. W. Cleland, Biochemistry 12, 3714 (1973). K. D. Danenberg and W. W. Cleland, Biochemistry 14, 28 (1975). K. J. Gruys and S. M. Schuster, Anal. Biochem. 125, 66 (1982). H. S. Penefsky, J. Biol. Chem. 252 2891 (1977). J. F. Morrison and C. T. Walsh, Adu. Enzymol. 61, 201 (1988). R. Shapiro and J. F. Riordan, Biochemistry 23, 5234 (1984). E. K. Jaffe, J. Nick, and M. Cohn, J. Biol. Chem. 257, 7650 (1982). F. Eckstein, Ann. Rev. Biochem. 54, 367 (1985). J. M. Moore and G. H. Reed, Biochemistry 24, 5328 (1985). E. H. Serpersu, J. McCracken, J. Peisach, and A. S. Mildvan, Biochemistry 27,8034
R. Jeanneau,
and L. A. Pradell, Biochemistry
19, 5168
(1988).
31.
B. D. Ray, J. M. Moore, and B. D. Nageswara Rao, J. Znorg. Biochem. 40,47 (1990).
Received March 9, 1992; accepted April 28, 1992