130
Biochimica et Biophysica Acta, 1040 (1990) 130-133
Elsevier BBAPRO 30271
BBA Report
Studies on the interactions of glycerol dehydrogenase from Bacillus stearothermophilus with Zn 2÷ ions and N A D H P. S p e n c e r 1, A . S l a d e 1, T . A t k i n s o n 2 a n d M . G . G o r e 1 1 Department of Biochemistry, SERC Centre for Molecular Recognition, University of Southampton, Southampton and 2 Biotechnology Division, PHLS Centre for Applied Microbiology Research, Salisbury (U.K.)
(Received 5 January 1990)
Key words: Glyceroldehydrogenase;MetaUo-enzyme;Dissociationconstant; Fluorescence;Binary complex; ( B. Stearothermophilus)
The interactions of the essential divalent cation, Znz+, with the binary complex formed between glycerol dehydrogenase (glycerol:NAD + 2-oxidocednetase, EC 1.1.1.6) and its coenzyme NADH have been examined by fhmceseenee spectroscopy. Both the metallo and mm-metalio form of the enzyme bind the coenzyme NADH. The addition of Znz + ions to a solution of the binary complex formed between metal-depleted enzyme and N A D H results in a rapid increase in fluorescence emission at 430 rim. This has been used to determine the on rate for Zn2+ to the enzyme/binary complex. A ~ . iation constant of 3.02 + 0.25 • 1 0 - 9 M for the equilibrium between Zn2 + ions and the enzyme has been determined.
Two pathways exist in bacteria for the dissimilation of glycerol; one pathway involves the use of an NADP+-linked glycerol dehydrogenase which converts glycerol to glyceraldehyde [1] and the second entails the oxidation of glycerol to dihydroxyacetone with the concomitant reduction of N A D + to N A D H [2]. The glycerol dehydrogenase ( G D H ) (glycerol : N A D + 2-oxidoreductase, EC 1.1.1.6) from Bacillus stearothermophilus falls into the latter category, is N A D + linked and converts glycerol'to dihydroxyacetone [3]. Interest in this group of enzymes has arisen because of their potential role in the estimation of serum triacylglycerols. The enzyme from B. stearothermophilus has been shown to be a tetrameric protein of subunit M r 42000, each binding one equivalent of Zn z+ [4]. The removal of the metal ion causes loss of activity and allows the protein to undergo a conformational change induced by changes in temperature or p H [5]. In solutions at low temperatures (10 o C) or at a p H higher than 7.0, the metal-depleted enzyme favours a form unable to rebind metal ions and hence unable to regain activity immediately (E u) on addition of metal ions. This form of the enzyme
Abbreviations: GDH, glycerol dehydrogenas¢; ADH, alcohol dehydrogenase. Correspondence: M.G. Gore, Department of Biochemistry,University of Southampton, Bassett Crescent East, Southampton, SO9 3TU, U.K.
exists in equilibrium with another (E a, favoured by temperatures above 20 ° C or low pH) which is able to rebind Zn 2+ ions and therefore exhibit spontaneous activity; the transition from one form of the enzyme to the other has a p K of 6.7 [5], is freely reversible and the process shows no hysteresis. Both forms of the enzyme are able to bind stoichiometric amounts of the reduced coenzyme N A D H , although the intensity of fluorescence from the bound N A D H at 430 nm is much higher (approx. 2-fold) when in a complex with metallo-enzyme compared with the complex formed with metaldepleted enzyme [4]. In the present paper we have exploited this difference in fluorescence intensity to measure the rate of binding of the metal ion to the binary complex formed between the e n z y m e / m e t a l depleted enzyme species E a and N A D H . Procedures used for the growth of cells, purification of the enzyme and the assay for the enzyme have all been fully described elsewhere [4]. Protein concentration was determined by the method of Lowry et al. [6]. All calculations of enzyme molarity are based upon the subunit M r of 42 000 [4]. Removal and replacement of Zn 2+ ions and estimation of Zn 2+ content of metalloenzyme were carried out as described in Ref. 4. All titrations were performed on a Perkin Elmer 605S fluorimeter. The excitation wavelength was set at 340 nm (slit width 5 nm) and the emission wavelength at 430 nm (slit width 5 nm), the temperature was maintained at 15 o C. Enzyme (20 #M) was diluted 10 times
0167-4838/90/$03.50 © 1990 ElsevierSciencePublishers B,V. (BiomedicalDivision)
131
into 50 m M potassium phosphate (or triethanolamineHC1) buffer (pH 7.4) at 15°C, N A D H was subsequently added in pl aliquots. The concentration of the coenzyme was determined from absorption measurements at 340 nm (E340 = 6200). The titrations were repeated using 1 ml of buffer alone and the curve generated was subtracted from the saturation curves obtained when enzyme was present. Observed fluorescence readings were corrected for dilution resulting from addition of N A D H or enzyme. Readings were also corrected for the inner filter effect at the excitation wavelength and emission wavelength using the procedure described in Ref. 7. The method of Stinson and Holbrook [8] was used to analyse the fluorescence curves to determine the stoichiometry of binding and the value of the dissociation constant K a. For observation of the binding of Zn 2+ ions to an enzyme-NADH complex, measurements were taken on a Perkin Elmer 605S spectrofluorimeter equipped with a chart recorder. The cell was thermostated at 15 o C. 1 ml samples of solutions of metal-depleted enzyme (2.45 tiM) were mixed with various concentrations of N A D H and the initial fluorescence intensity at 430 nm was recorded. Then an aliquot of a solution of ZnC12 (1 mM) was added to give a known concentration and the change in fluorescence intensity at 430 nm recorded with time. An excitation wavelength of 340 nm was used., Fig. 1 demonstrates typical saturation curves obtained when N A D H is titrated into a solution of metalloor metal-depleted enzyme pre-equilibrated and used at p H 6.0 or 8.0. It can be noted that the fluorescence emission intensity from both solutions containing metallo enzyme is higher at saturating concentrations of the N A D H than either of the solutions containing metal-depleted enzyme. However, whereas the fluorescence intensities obtained at saturating concentrations of N A D H from the former are similar at the two p H values, as indeed are the K d values (0.23 and 0.22 tiM) and the mol equivalent of N A D H bound per subunit (0.95 + 0.023) at p H 6.0 and 8.0, respectively, those obtained in the absence of Zn 2+ ions are noticeably dissimilar, that obtained at p H 6.0 being twice that obtained at p H 8.0. Furthermore, the K a determined at p H 6.0 and 0.36 tiM, close to that obtained when using the metallo enzyme, whereas that obtained at p H 8.0 was much higher at 0.88 #M. Despite the decreased fluorescence emission intensity, no change in stoichiometry of N A D H binding occurs. At p H 6.0 there are 1.01 equivalents of N A D H bound and at p H 8.0 there are 1.05 mol of N A D H bound per mol of subunit. At p H 6.0 the equilibrium which exists between two forms of the metal-depleted enzyme (E u and E ~, see above) favours E a and this experiment suggests that the metal-depleted species E~-interacts with N A D H in a similar manner to metallo enzyme and that the more
/ V " / t j C f
1o
__/¢,,"¢
¢
J
uY
f./ /
O0
I
I
I
I
I
I
I
1
2
3
4
5
6
7
NADH (/.11,4)
Fig. 1. Saturation curves describing N A D H binding to Zn2+-GDH or metal-depleted G D H on pH. Solutions of Zn2+-GDH or metaldepleted G D H (both 13 p M ) pre-equilibrated in 50 m M potassium phosphate buffer at p H 6.0 or 8.0 at 20 ° C were titrated with N A D H after a 10-fold dilution into the same metal free buffers at 15 o C. The curves describe the difference in fluorescence emission intensity at 430 n m between N A D H additions to 1 ml samples of Zn 2 +metallo G D H p H 6.0 (O) and 8.0 (), metal-depleted G D H at p H 6.0 (A) and 8.0 (o) a n d buffer without enzyme present. Excitation wavelength ffi 340 nm.
pronounced differences (fluorescence intensity and K d) noted at p H 8.0 are due to the almost total predominance of the metal depleted form E u. It was noted that if Zn 2+ ions were added to solution of metal-depleted enzyme (pH 6.0) in the presence of a saturating concentration of N A D H a rapid rise in the fluorescence emission occurred over a period of 10-60 s. Fig. 2 shows typical changes recorded at 15 ° C when 9, 19 or 45 # M Zn 2÷ ions were added to samples of enzyme in the presence of 11 p M N A D H (equivalent to the molarity of the enzyme and an excess of N A D H equivalent to 40-times the value of the K d for N A D H under the same conditions). The inset gives semilogarithmic rate analyses for these experiments and show that the processes occurred with apparent single exponential rates (kapp) over this time period of 0.016, 0.028 or 0.054 s -1, respectively. These values of kapp were found to be dependent upon the concentration of N A D H present and increase if the N A D H concentration in the mixture is decreased. Fig. 3A shows the dependence of the rate of change of fluorescence at 430 nm (k~pp) on the concentration of N A D H when 9 p M Zn 2 + ions were added to a solution of binary complex formed between N A D H and metal-depleted enzyme at p H 6.,0, 15 ° C. Analysis of such data obtained using a range of concentrations of Zn 2+ ions (Fig. 3B) demonstrates that the change of k~op correlates with the level
132 110
l
¢
100
.
90
//
/'~"~'~"
b
,
•
11
~
•
u7
70
........
.....
i ~/ ~
I 0
05/ 0.51
I 20
I i 40 eO Time (f~c)
I
I
I
1
2
3
Time
i 80
equilibrium position strongly favours E a. Therefore, the possibility that these rapid changes in fluorescence occur as a result of the transition E" (enzyme unable to bind metal) to E a (enzyme able to bind metal) can be excluded. The amplitude of the change in fluorescence from the bound N A D H when Zn 2+ ions are added decreases if the metal-depleted enzyme is pre-incubated at 4 ° C for 2h before being rapidly adjusted to 15°C, mixed with N A D H and then Zn 2+ ions. Under these pre-incubation conditions the E u form of the metal-de-
A
(min)
0.3
Fig. 2. The changes in fluorescence emission intensity at 430 nm on addition of Zn2+-ions to a solution of binary complex formed between metal-depleted GDH and NADH. The increase in fluorescence at 430 nm on addition of Zn2+ ions to N A D H metal-depleted GDH binary complex (2.45 ~M er~yme+11 ~M NADH) in 50 mM triethanolamine-HCl buffer (pH 6.0 at 15°C), was used to monitor Zn2+ binding. The curves describe the changes in fluorescence intensity at 430 nm after the addition of 9 btM (a), 19/tM (b) or 45/tM (e) Zn2÷ ions. Excitation wavelength = 340 rim. The inset describes the semilogarithmic analysis of the traces obtained in Fig. 2.
02
& o
0.~
0
I
I
2
4
,
NADH
of saturation of the enzyme by N A D H (calculated from N A D H titration experiments as described in Fig. 1). By extrapolation of the curves to a point equivalent to zero saturation of the enzyme by N A D H it is possible to 0 reflecting the interacobtain calculated values for kapp, tion between enzyme and Zn2÷ ions in the absence of N A D H . This was repeated using various concentrations of Zn 2+ ion and the pseudo-first-order rates, k°pp were used to calculate k+ (the second order, on-rate of Zn 2+ ion to the protein) by plotting the k°pp against the combined concentration of the ligand (Zn 2+ ion) and enzyme concentration. The rate determined from graphical analysis of the data (Fig. 4) is 3 8 . 1 0 3 M - t . S -1 at pH 6.0, 15°C. Estimation of the off-rate for Zn 2 + from the metallo protein by fluorimetric means is more difficult because of the very slow rate of dissociation. A kinetic determination was achieved by measuring the rate of loss of activity from a solution of 2.4 # M enzyme as Zn 2+ is removed from the protein by a range of concentrations of EDTA.The pseudo-first-order rates determined at pH 6.0, 15 ° C when using concentrations of 5, 10 or 50 mM EDTA were similar at (1.15 + 0 . 1 0 2 ) . 10 -4 s -t. This value of the off rate, together with the calculated on-rate for the formation of the Zn 2 + enzyme suggests a dissociation constant for the equilibrium of (3.02 + 0.25 • 10 -9 M.
Although a reversible structural transition E u to E" has been shown to take place in the metal-depleted G D H , the rate of this change is independent of the presence of N A D H and is approx. 50-times slower than the rates measured in these studies [5] and at pH 6.0 the
I
I
6
8
10
(~M)
2.0
\ \\
1.5
\
-
\
\
\0
1.0
&
x ~0
0
20
40
60
80
100
o/o S e t u r a t i o n
Fig. 3. Dependence of kapp for Zn2+ binding to metal-depleted G D H on the N A D H concentration. Zn2+ ions were added to give a final concentration of 9 pM to metal-depleted G D H (2.45/tM) in 50 mM triethanolamine-HCl buffer (pH 6.0 at 15°C) in the presence of various concentrations of NADH. The increase in fluorescence at 430 nm was monitored and analysed by a semi-logarithmic plot and the apparent rate constant (kapp) plotted a~ainst the concentration of N A D H present (Fig. 3A). Dependence of k,pp for Zn2+ binding to metal-depleted GDH on the percentage saturation of metal-depleted G D H by NADH. The k,~p of Zn2+ binding to metal-depleted G D H (2.45/tM) in 50 mM triethanolamine-HC! buffer (pH 6.0 at 15°C) at various concentrations of sub-saturating N A D H was determined using Zn2+ concentrations of 9 ttM (@), 19 ;tM (X), 30 tiM ( 0 ) and 45 ~tM (O). The kapp obtained was then plotted against the • saturation of metal-depleted GDH by NADH. This was determined from a standard N A D H saturation curve for metal-depleted GDH (data not shown). The predicted kapp in the absence of N A D H (k ° ) was estimated by extrapolation of the line obtained to zero concentration of N A D H (Fig. 3B).
133 2.0
1.,5
1.0
o i:1 0.,~, O
O-
I 10
I 20 EZn2"+enzyme]
I 30
I 40
I 50
,uM
Fig. 4. Determination of the on rate for Zn2+ binding to metal-depleted GDH in the absence of NADH (k ° ). Values of ka°p were plotted against the (Zn2+ +enzyme) concentration used and the on rate for the Zn2+ binding process to metal-depleted GDH in the absence of NADH (k+) determined from the gradient giving a value of 38.103 M-l.s -]. pleted enzyme is favoured, the relative concentration of the E a form decreased and the rate of the transition from E u to E a is too slow to affect the experiment. This observation therefore supports the conclusion that the change in fluorescence signal of the N A D H arises because of the binding of Zn 2+ ions to the E a form of the enzyme. This change in fluorescence is confined to the fluorescence emission from the nicotinamide ring of the N A D H since no differences in protein fluorescence exist between the E u and E a forms of the enzyme whether free or in the binary complex with N A D H , in the absence or presence of Zn 2+ ions. The metallo enzyme has a 4-fold higher affinity for N A D H than E u and an approx. 2-fold higher affinity for N A D H than E a. This increased affinity may reflect an increased immobilisation of the nicotinamide ring of the N A D H leading to an enhanced quantum yield. It might be envisaged that the presence of bound N A D H hinders the recombination of Zn 2+ with the protein perhaps by partial occlusion of the metal binding site a n d / o r by increasing the rigidity of the metal binding site hindering any slight changes in conformation required to accomodate the metal ion. The kinetics of metal recombination with other apometalloproteins studied to date fall into two groups. Those that exhibit a simple second-order-rate depen-
dence of binding of the metal-depleted enzyme with metal ions e.g., carboxypeptidase [9], carbonic anhydrase [10] and those which exhibit a two step process involving a rapid binding step followed by a slower monomolecular process, e.g., the recombination of metal-depleted alcohol dehydrogenase ( A D H ) with Zn 2÷ ions [11] or concanavilin A with Mn 2÷ ions [12]. The slow monomolecular step ( = 1 . 1 0 - 4 S - 1 ) occurring when metal-depleted A D H binds Zn 2÷ ions results in the appearance of activity. Schneider and Zeppezauer [11] have proposed that this is possibly due to subsequent geometric rearrangement of the metal ligands after initial metal binding or alternatively by subsequent migration of Zn 2+ to the active site from peripheral binding sites [13]. Zn 2÷ has been shown to bind to metal-depleted G D H in the presence of bound N A D H , unlike the metal-depleted A D H [11] and this binding process occurs some 27-times faster in the absence of bound N A D H . The recombination of metal-depleted G D H with Zn 2+ appears to be sufficient to explain the rapid regain in activity of metal-depleted G D H on the addition of metal ions. Therefore metal-depleted G D H appears to fall into the class of metallo-proteins which exhibit simple second-order-rate dependence kinetics for metal recombination to the apo-protein form.
References 1 Viswanath-Reddy, M., Pyle, J.E. and Branch Howe, H. (1978) J. Gen. Microbiol. 107, 289-296. 2 May, J.W. and Sloan, J. (1981) J. Gen. Microbiol. 1123, 183-185. 3 Atkinson, A., Bruton, C.J., Comer, M.J. and Sharp, R.J. (1978) UK. Patent Specifications 21194/78. 4 Spencer, P., Bown, K.J., Scawen, M.D., Atkinson, A. and Gore, M.G. (1989) Biochim. Biophys. Acta 994, 270-279. 5 Spencer, P., Paine, L., Atkinson, A. and Gore, M.G. (1990) FEBS Lett. 259, 297-300. 6 Lowry, O.H., Rosebrough, N.J., Falr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 7 Rooney, E.K., Gore, M.G. and Lee, A.G. (1987) Biochemistry26, 3688-3697. 8 Stinson, R.A. and Holbrook, J.J. (1973) Biochem.J. 131, 719-728. 9 Billo, E.J., Brito, K.K. and Wilkins, R.G. (1978) Bioinorg. Chem. 8, 461-470. 10 Henkens, R.W. and Sturevant, J.M. (1968) J. Am. Chem. Soc. 90, 2669-2676. 11 Schneider, G. and Zeppezauer, M. (1983) J. Inorg. Chem. 18, 59-69. 12 Brown, R.D., Brewer, C.F. and Koeing, S.H. (1977) Biochemistry 16, 3883-3896. 13 Andersson, P., Kvassman, J., Olden, B. and Petterson, G. (1981) Eur. J. Biochem. 113, 425-433.
Biochimica et Biophysica Acta, 1040 (1990) 130-133
Elsevier BBAPRO 30271
BBA Report
Studies on the interactions of glycerol dehydrogenase from Bacillus stearothermophilus with Zn 2÷ ions and N A D H P. S p e n c e r 1, A . S l a d e 1, T . A t k i n s o n 2 a n d M . G . G o r e 1 1 Department of Biochemistry, SERC Centre for Molecular Recognition, University of Southampton, Southampton and 2 Biotechnology Division, PHLS Centre for Applied Microbiology Research, Salisbury (U.K.)
(Received 5 January 1990)
Key words: Glyceroldehydrogenase;MetaUo-enzyme;Dissociationconstant; Fluorescence;Binary complex; ( B. Stearothermophilus)
The interactions of the essential divalent cation, Znz+, with the binary complex formed between glycerol dehydrogenase (glycerol:NAD + 2-oxidocednetase, EC 1.1.1.6) and its coenzyme NADH have been examined by fhmceseenee spectroscopy. Both the metallo and mm-metalio form of the enzyme bind the coenzyme NADH. The addition of Znz + ions to a solution of the binary complex formed between metal-depleted enzyme and N A D H results in a rapid increase in fluorescence emission at 430 rim. This has been used to determine the on rate for Zn2+ to the enzyme/binary complex. A ~ . iation constant of 3.02 + 0.25 • 1 0 - 9 M for the equilibrium between Zn2 + ions and the enzyme has been determined.
Two pathways exist in bacteria for the dissimilation of glycerol; one pathway involves the use of an NADP+-linked glycerol dehydrogenase which converts glycerol to glyceraldehyde [1] and the second entails the oxidation of glycerol to dihydroxyacetone with the concomitant reduction of N A D + to N A D H [2]. The glycerol dehydrogenase ( G D H ) (glycerol : N A D + 2-oxidoreductase, EC 1.1.1.6) from Bacillus stearothermophilus falls into the latter category, is N A D + linked and converts glycerol'to dihydroxyacetone [3]. Interest in this group of enzymes has arisen because of their potential role in the estimation of serum triacylglycerols. The enzyme from B. stearothermophilus has been shown to be a tetrameric protein of subunit M r 42000, each binding one equivalent of Zn z+ [4]. The removal of the metal ion causes loss of activity and allows the protein to undergo a conformational change induced by changes in temperature or p H [5]. In solutions at low temperatures (10 o C) or at a p H higher than 7.0, the metal-depleted enzyme favours a form unable to rebind metal ions and hence unable to regain activity immediately (E u) on addition of metal ions. This form of the enzyme
Abbreviations: GDH, glycerol dehydrogenas¢; ADH, alcohol dehydrogenase. Correspondence: M.G. Gore, Department of Biochemistry,University of Southampton, Bassett Crescent East, Southampton, SO9 3TU, U.K.
exists in equilibrium with another (E a, favoured by temperatures above 20 ° C or low pH) which is able to rebind Zn 2+ ions and therefore exhibit spontaneous activity; the transition from one form of the enzyme to the other has a p K of 6.7 [5], is freely reversible and the process shows no hysteresis. Both forms of the enzyme are able to bind stoichiometric amounts of the reduced coenzyme N A D H , although the intensity of fluorescence from the bound N A D H at 430 nm is much higher (approx. 2-fold) when in a complex with metallo-enzyme compared with the complex formed with metaldepleted enzyme [4]. In the present paper we have exploited this difference in fluorescence intensity to measure the rate of binding of the metal ion to the binary complex formed between the e n z y m e / m e t a l depleted enzyme species E a and N A D H . Procedures used for the growth of cells, purification of the enzyme and the assay for the enzyme have all been fully described elsewhere [4]. Protein concentration was determined by the method of Lowry et al. [6]. All calculations of enzyme molarity are based upon the subunit M r of 42 000 [4]. Removal and replacement of Zn 2+ ions and estimation of Zn 2+ content of metalloenzyme were carried out as described in Ref. 4. All titrations were performed on a Perkin Elmer 605S fluorimeter. The excitation wavelength was set at 340 nm (slit width 5 nm) and the emission wavelength at 430 nm (slit width 5 nm), the temperature was maintained at 15 o C. Enzyme (20 #M) was diluted 10 times
0167-4838/90/$03.50 © 1990 ElsevierSciencePublishers B,V. (BiomedicalDivision)
131
into 50 m M potassium phosphate (or triethanolamineHC1) buffer (pH 7.4) at 15°C, N A D H was subsequently added in pl aliquots. The concentration of the coenzyme was determined from absorption measurements at 340 nm (E340 = 6200). The titrations were repeated using 1 ml of buffer alone and the curve generated was subtracted from the saturation curves obtained when enzyme was present. Observed fluorescence readings were corrected for dilution resulting from addition of N A D H or enzyme. Readings were also corrected for the inner filter effect at the excitation wavelength and emission wavelength using the procedure described in Ref. 7. The method of Stinson and Holbrook [8] was used to analyse the fluorescence curves to determine the stoichiometry of binding and the value of the dissociation constant K a. For observation of the binding of Zn 2+ ions to an enzyme-NADH complex, measurements were taken on a Perkin Elmer 605S spectrofluorimeter equipped with a chart recorder. The cell was thermostated at 15 o C. 1 ml samples of solutions of metal-depleted enzyme (2.45 tiM) were mixed with various concentrations of N A D H and the initial fluorescence intensity at 430 nm was recorded. Then an aliquot of a solution of ZnC12 (1 mM) was added to give a known concentration and the change in fluorescence intensity at 430 nm recorded with time. An excitation wavelength of 340 nm was used., Fig. 1 demonstrates typical saturation curves obtained when N A D H is titrated into a solution of metalloor metal-depleted enzyme pre-equilibrated and used at p H 6.0 or 8.0. It can be noted that the fluorescence emission intensity from both solutions containing metallo enzyme is higher at saturating concentrations of the N A D H than either of the solutions containing metal-depleted enzyme. However, whereas the fluorescence intensities obtained at saturating concentrations of N A D H from the former are similar at the two p H values, as indeed are the K d values (0.23 and 0.22 tiM) and the mol equivalent of N A D H bound per subunit (0.95 + 0.023) at p H 6.0 and 8.0, respectively, those obtained in the absence of Zn 2+ ions are noticeably dissimilar, that obtained at p H 6.0 being twice that obtained at p H 8.0. Furthermore, the K a determined at p H 6.0 and 0.36 tiM, close to that obtained when using the metallo enzyme, whereas that obtained at p H 8.0 was much higher at 0.88 #M. Despite the decreased fluorescence emission intensity, no change in stoichiometry of N A D H binding occurs. At p H 6.0 there are 1.01 equivalents of N A D H bound and at p H 8.0 there are 1.05 mol of N A D H bound per mol of subunit. At p H 6.0 the equilibrium which exists between two forms of the metal-depleted enzyme (E u and E ~, see above) favours E a and this experiment suggests that the metal-depleted species E~-interacts with N A D H in a similar manner to metallo enzyme and that the more
/ V " / t j C f
1o
__/¢,,"¢
¢
J
uY
f./ /
O0
I
I
I
I
I
I
I
1
2
3
4
5
6
7
NADH (/.11,4)
Fig. 1. Saturation curves describing N A D H binding to Zn2+-GDH or metal-depleted G D H on pH. Solutions of Zn2+-GDH or metaldepleted G D H (both 13 p M ) pre-equilibrated in 50 m M potassium phosphate buffer at p H 6.0 or 8.0 at 20 ° C were titrated with N A D H after a 10-fold dilution into the same metal free buffers at 15 o C. The curves describe the difference in fluorescence emission intensity at 430 n m between N A D H additions to 1 ml samples of Zn 2 +metallo G D H p H 6.0 (O) and 8.0 (), metal-depleted G D H at p H 6.0 (A) and 8.0 (o) a n d buffer without enzyme present. Excitation wavelength ffi 340 nm.
pronounced differences (fluorescence intensity and K d) noted at p H 8.0 are due to the almost total predominance of the metal depleted form E u. It was noted that if Zn 2+ ions were added to solution of metal-depleted enzyme (pH 6.0) in the presence of a saturating concentration of N A D H a rapid rise in the fluorescence emission occurred over a period of 10-60 s. Fig. 2 shows typical changes recorded at 15 ° C when 9, 19 or 45 # M Zn 2÷ ions were added to samples of enzyme in the presence of 11 p M N A D H (equivalent to the molarity of the enzyme and an excess of N A D H equivalent to 40-times the value of the K d for N A D H under the same conditions). The inset gives semilogarithmic rate analyses for these experiments and show that the processes occurred with apparent single exponential rates (kapp) over this time period of 0.016, 0.028 or 0.054 s -1, respectively. These values of kapp were found to be dependent upon the concentration of N A D H present and increase if the N A D H concentration in the mixture is decreased. Fig. 3A shows the dependence of the rate of change of fluorescence at 430 nm (k~pp) on the concentration of N A D H when 9 p M Zn 2 + ions were added to a solution of binary complex formed between N A D H and metal-depleted enzyme at p H 6.,0, 15 ° C. Analysis of such data obtained using a range of concentrations of Zn 2+ ions (Fig. 3B) demonstrates that the change of k~op correlates with the level
132 110
l
¢
100
.
90
//
/'~"~'~"
b
,
•
11
~
•
u7
70
........
.....
i ~/ ~
I 0
05/ 0.51
I 20
I i 40 eO Time (f~c)
I
I
I
1
2
3
Time
i 80
equilibrium position strongly favours E a. Therefore, the possibility that these rapid changes in fluorescence occur as a result of the transition E" (enzyme unable to bind metal) to E a (enzyme able to bind metal) can be excluded. The amplitude of the change in fluorescence from the bound N A D H when Zn 2+ ions are added decreases if the metal-depleted enzyme is pre-incubated at 4 ° C for 2h before being rapidly adjusted to 15°C, mixed with N A D H and then Zn 2+ ions. Under these pre-incubation conditions the E u form of the metal-de-
A
(min)
0.3
Fig. 2. The changes in fluorescence emission intensity at 430 nm on addition of Zn2+-ions to a solution of binary complex formed between metal-depleted GDH and NADH. The increase in fluorescence at 430 nm on addition of Zn2+ ions to N A D H metal-depleted GDH binary complex (2.45 ~M er~yme+11 ~M NADH) in 50 mM triethanolamine-HCl buffer (pH 6.0 at 15°C), was used to monitor Zn2+ binding. The curves describe the changes in fluorescence intensity at 430 nm after the addition of 9 btM (a), 19/tM (b) or 45/tM (e) Zn2÷ ions. Excitation wavelength = 340 rim. The inset describes the semilogarithmic analysis of the traces obtained in Fig. 2.
02
& o
0.~
0
I
I
2
4
,
NADH
of saturation of the enzyme by N A D H (calculated from N A D H titration experiments as described in Fig. 1). By extrapolation of the curves to a point equivalent to zero saturation of the enzyme by N A D H it is possible to 0 reflecting the interacobtain calculated values for kapp, tion between enzyme and Zn2÷ ions in the absence of N A D H . This was repeated using various concentrations of Zn 2+ ion and the pseudo-first-order rates, k°pp were used to calculate k+ (the second order, on-rate of Zn 2+ ion to the protein) by plotting the k°pp against the combined concentration of the ligand (Zn 2+ ion) and enzyme concentration. The rate determined from graphical analysis of the data (Fig. 4) is 3 8 . 1 0 3 M - t . S -1 at pH 6.0, 15°C. Estimation of the off-rate for Zn 2 + from the metallo protein by fluorimetric means is more difficult because of the very slow rate of dissociation. A kinetic determination was achieved by measuring the rate of loss of activity from a solution of 2.4 # M enzyme as Zn 2+ is removed from the protein by a range of concentrations of EDTA.The pseudo-first-order rates determined at pH 6.0, 15 ° C when using concentrations of 5, 10 or 50 mM EDTA were similar at (1.15 + 0 . 1 0 2 ) . 10 -4 s -t. This value of the off rate, together with the calculated on-rate for the formation of the Zn 2 + enzyme suggests a dissociation constant for the equilibrium of (3.02 + 0.25 • 10 -9 M.
Although a reversible structural transition E u to E" has been shown to take place in the metal-depleted G D H , the rate of this change is independent of the presence of N A D H and is approx. 50-times slower than the rates measured in these studies [5] and at pH 6.0 the
I
I
6
8
10
(~M)
2.0
\ \\
1.5
\
-
\
\
\0
1.0
&
x ~0
0
20
40
60
80
100
o/o S e t u r a t i o n
Fig. 3. Dependence of kapp for Zn2+ binding to metal-depleted G D H on the N A D H concentration. Zn2+ ions were added to give a final concentration of 9 pM to metal-depleted G D H (2.45/tM) in 50 mM triethanolamine-HCl buffer (pH 6.0 at 15°C) in the presence of various concentrations of NADH. The increase in fluorescence at 430 nm was monitored and analysed by a semi-logarithmic plot and the apparent rate constant (kapp) plotted a~ainst the concentration of N A D H present (Fig. 3A). Dependence of k,pp for Zn2+ binding to metal-depleted GDH on the percentage saturation of metal-depleted G D H by NADH. The k,~p of Zn2+ binding to metal-depleted G D H (2.45/tM) in 50 mM triethanolamine-HC! buffer (pH 6.0 at 15°C) at various concentrations of sub-saturating N A D H was determined using Zn2+ concentrations of 9 ttM (@), 19 ;tM (X), 30 tiM ( 0 ) and 45 ~tM (O). The kapp obtained was then plotted against the • saturation of metal-depleted GDH by NADH. This was determined from a standard N A D H saturation curve for metal-depleted GDH (data not shown). The predicted kapp in the absence of N A D H (k ° ) was estimated by extrapolation of the line obtained to zero concentration of N A D H (Fig. 3B).
133 2.0
1.,5
1.0
o i:1 0.,~, O
O-
I 10
I 20 EZn2"+enzyme]
I 30
I 40
I 50
,uM
Fig. 4. Determination of the on rate for Zn2+ binding to metal-depleted GDH in the absence of NADH (k ° ). Values of ka°p were plotted against the (Zn2+ +enzyme) concentration used and the on rate for the Zn2+ binding process to metal-depleted GDH in the absence of NADH (k+) determined from the gradient giving a value of 38.103 M-l.s -]. pleted enzyme is favoured, the relative concentration of the E a form decreased and the rate of the transition from E u to E a is too slow to affect the experiment. This observation therefore supports the conclusion that the change in fluorescence signal of the N A D H arises because of the binding of Zn 2+ ions to the E a form of the enzyme. This change in fluorescence is confined to the fluorescence emission from the nicotinamide ring of the N A D H since no differences in protein fluorescence exist between the E u and E a forms of the enzyme whether free or in the binary complex with N A D H , in the absence or presence of Zn 2+ ions. The metallo enzyme has a 4-fold higher affinity for N A D H than E u and an approx. 2-fold higher affinity for N A D H than E a. This increased affinity may reflect an increased immobilisation of the nicotinamide ring of the N A D H leading to an enhanced quantum yield. It might be envisaged that the presence of bound N A D H hinders the recombination of Zn 2+ with the protein perhaps by partial occlusion of the metal binding site a n d / o r by increasing the rigidity of the metal binding site hindering any slight changes in conformation required to accomodate the metal ion. The kinetics of metal recombination with other apometalloproteins studied to date fall into two groups. Those that exhibit a simple second-order-rate depen-
dence of binding of the metal-depleted enzyme with metal ions e.g., carboxypeptidase [9], carbonic anhydrase [10] and those which exhibit a two step process involving a rapid binding step followed by a slower monomolecular process, e.g., the recombination of metal-depleted alcohol dehydrogenase ( A D H ) with Zn 2÷ ions [11] or concanavilin A with Mn 2÷ ions [12]. The slow monomolecular step ( = 1 . 1 0 - 4 S - 1 ) occurring when metal-depleted A D H binds Zn 2÷ ions results in the appearance of activity. Schneider and Zeppezauer [11] have proposed that this is possibly due to subsequent geometric rearrangement of the metal ligands after initial metal binding or alternatively by subsequent migration of Zn 2+ to the active site from peripheral binding sites [13]. Zn 2÷ has been shown to bind to metal-depleted G D H in the presence of bound N A D H , unlike the metal-depleted A D H [11] and this binding process occurs some 27-times faster in the absence of bound N A D H . The recombination of metal-depleted G D H with Zn 2+ appears to be sufficient to explain the rapid regain in activity of metal-depleted G D H on the addition of metal ions. Therefore metal-depleted G D H appears to fall into the class of metallo-proteins which exhibit simple second-order-rate dependence kinetics for metal recombination to the apo-protein form.
References 1 Viswanath-Reddy, M., Pyle, J.E. and Branch Howe, H. (1978) J. Gen. Microbiol. 107, 289-296. 2 May, J.W. and Sloan, J. (1981) J. Gen. Microbiol. 1123, 183-185. 3 Atkinson, A., Bruton, C.J., Comer, M.J. and Sharp, R.J. (1978) UK. Patent Specifications 21194/78. 4 Spencer, P., Bown, K.J., Scawen, M.D., Atkinson, A. and Gore, M.G. (1989) Biochim. Biophys. Acta 994, 270-279. 5 Spencer, P., Paine, L., Atkinson, A. and Gore, M.G. (1990) FEBS Lett. 259, 297-300. 6 Lowry, O.H., Rosebrough, N.J., Falr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 7 Rooney, E.K., Gore, M.G. and Lee, A.G. (1987) Biochemistry26, 3688-3697. 8 Stinson, R.A. and Holbrook, J.J. (1973) Biochem.J. 131, 719-728. 9 Billo, E.J., Brito, K.K. and Wilkins, R.G. (1978) Bioinorg. Chem. 8, 461-470. 10 Henkens, R.W. and Sturevant, J.M. (1968) J. Am. Chem. Soc. 90, 2669-2676. 11 Schneider, G. and Zeppezauer, M. (1983) J. Inorg. Chem. 18, 59-69. 12 Brown, R.D., Brewer, C.F. and Koeing, S.H. (1977) Biochemistry 16, 3883-3896. 13 Andersson, P., Kvassman, J., Olden, B. and Petterson, G. (1981) Eur. J. Biochem. 113, 425-433.
