The Kinetics of Schiff-Base Formation during ... - Wiley Online Library

Eur. J. Biochem. 94, 207-214 (1979)

The Kinetics of Schiff-Base Formation during Reconstitution of D-Serine Apodehydratase from Escherichia coli with Pyridoxal 5'-Phosphate Thomas A. REED and Klaus D. SCHNACKERZ

Physiologisch-Chemisches Institut der Universitat Wiirzburg (Received September 25, 1978)

Schiff base formation during reconstitution of D-serine dehydratase (Escherichia coli) from its apoenzyme and pyridoxal 5'-phosphate (pyridoxal-P) has been studied by rapid kinetic techniques using absorbance changes at 436 nm. Three distinct reaction phases have been observed. The first is a very rapid change during which pyridoxal-P is initially bound to the apoenzyme. This step has an equilibrium constant of 1500 M-' and a forward reaction rate of the order of 2.6 x lo6 M-' sC1. The second phase shows a first-order rate constant with a value dependent on pyridoxal-P and corresponds to a first-order step with a forward rate constant of 3.04 s-' interacting with the initial equilibrium. The final phase is a slow first-order reaction, the rate constant of which is approximately 0.01 s- and is independent of pyridoxal-P concentration. The active pyridoxal species has been shown to be the free pyridoxal-P as opposed to hemiacetal or hemimercaptal forms.

'

Enzymes using pyridoxal5'-phosphate (pyridoxal-

P) as cofactor are widespread and of basic interest in biochemistry. Since Snell and DiMari [l] suggested that studies of reconstitution of such enzymes from pyridoxal-P and apoenzyme might be of particular interest, only a relatively small number of cases have been investigated [2- 51, none of them by rapid kinetic techniques which might conceivably provide greater insight into individual steps within the reactivation. D-Serine dehydratase from Escherichia coli [6] is a particularly appropriate vehicle for studies of this reconstitution in that it is a monomer with only one active site [6]. Thus, all aspects of the reconstitution kinetics can be attributed to the reactivation of the single site and corrections need not be made for multisite or subunit interaction. Preliminary reports of investigations of various aspects of D-serine dehydratase reactivation by rapid kinetic methods have already been presented [7,8]. Pyridoxal-P in D-serine dehydratase is bound to the &-aminogroup of a lysine residue of the enzyme as a Schiff base which plays a specific role in enzyme Abbreviation. Pyridoxal-P, pyridoxal 5'-phosphate. Enzymes. D-Serine dehydratase or D-serine hydro-lyase (deaminating) (EC 4.2.1.14); tryptophanase (EC 4.1.1.28); gfutamate decarboxylase or L-glutamate 1-carboxy-lyase (EC 4.1.1.15).

activity [6]. In this paper we present a detailed kinetic study of the binding of pyridoxal-P to D-serine apodehydratase and the development of the Schiff base between pyridoxal-P and the specific lysine residue. MATERIALS AND METHODS Pyridoxal-P and 2-mercaptoethanol were purchased from Serva (Heidelberg, F.R.G.). Lactate dehydrogenase from rabbit muscle was obtained from Boehringer Mannheim Corp. D-Serine was a product of Fluka (Buchs, Switzerland). All other chemicals were of analytical reagent grade. Solutions were prepared in deionized, double-distilled water. D-Serine dehydratase and the corresponding apoenzyme were prepared according to the method of Dowhan and Snell [6]. Protein concentrations of apoenzyme and holoenzyme were determined spectrophotometrically at 280nm and molar concentrations of apoenzyme were calculated on the basis of a molecular weight of 45 800 [ 5 ] . Enzymatic activity of D-serine dehydratase was measured according to published procedures [6].

Stopped-Flow Measurements Stopped-flow measurements were performed in a Gibson-Durrum model 13000 stopped-flow spectrom-

208

eter using two different detector systems. In one modification the instrument was equipped with an RCA 1P28 side-window photomultiplier. Transmittance changes were recorded on a Tektronixs model 547 storage oscillograph, photographed on Polaroid film and evaluated manually. In the other modification the instrument was equipped with a Hamamatsu R-376 end-window photomultiplier. The multiplier output was recorded on a Nicolet model 1090 digital oscillograph and transferred via an interface to punched paper tape using an ITT teletype equipped with a tape punch unit. The data were then read from the punched tape into a Wang model 720 C desk computer using a Zerpfel mechanical tape reader. Combined with the necessary control data, tables of time and absorbance were printed or absorbance/ time curves plotted automatically. The instrument dead time was estimated to be 2.5 ms as measured by the reaction of myoglobin with sodium cyanide. The slit width was held between 0.65 mm and 0.3 mm. A cuvette with 2-cm path length was used in all cases. The rate constants for the absorption changes were determined by the method of Guggenheim [9,10] to avoid interference stemming from the biphasic nature of the reaction.

Simulation The relative concentrations of free pyridoxal-P, pyridoxal-P hemimercaptal and pyridoxal-P hydrate (hemiacetal) were determined in the course of a series of kinetic simulation studies (our unpublished results). These were run using the system CHEK [12] on an IBM 360/44 computer at the Max-PlanckInstitut fur Ernahrungsphysiologie (Dortmund, F.R.G.). The forward and reverse rate constants used for the equilibrium between pyridoxal-P and water with M-' S - l the pyridoxal-P hemiacetal were 5.1 x and 10.19 s-', respectively. They were calculated by combining the data of Morosov et al. [13] and Ahrens et al. [14]. The forward and reverse rate constants used for the equilibrium between pyridoxal-P and 2-mercaptoethanol with the related hemimercaptal of pyridoxal-P were 112000 M-' s-l and 3400 s-', respectively. These values were determined by the method of Schuster and Winkler [15] in conjunction with H. Winkler at the Max-Planck-Institut fur Biophysikalische Chemie (Gottingen, Nikolausberg, F.R.G.). Reaction conditions were 100 mM potassium phosphate at pH 7.8 to reflect the conditions of reaction during stopped-flow and stopped-flow measurements combined with temperature-jump. The equilibrium constant for this latter reaction was confirmed by calculations using the method of Benesi and Hildebrand [16].

Reconstitution of D-Serine Apodehydratase with Pyridoxal-P 50 /-

Wavelength

(nm)

Fig. 1. Absorption spectra of the constituents of' the stopped-flow D-serine apodehydratase experiments. D-Serine dehydratase (-), (----) and pyridoxal-P (-.-.-), all in 100 mM potassium phosphate buffer, pH 7.8, containing 25 mM 2-mercaptoethanol; pyridoxal-P (. . . .) in the same buffer without mereaptoethanol

Measurements by Temperature-Jump after Stopped-Flow Mixing

These measurements were made in conjunction with L. Veil [8] at the Max-Planck-Institut fur Biophysikalische Chemie (Gottingen, Nikolausberg, F.R.G.), on an instrument of the type described by Eigen and DeMayer [17]. Critical for these measurements was the instrument dead time of 20 ms which prevented the observation of any rapid relaxation phenomena associated with the equilibrium between pyridoxal-P, 2-mercaptoethanol and the related pyridoxal-P hemimercaptal.

RESULTS The absorption spectrum of D-serine dehydratase (Fig. 1, curve a) shows a peak at 415 nm associated with the presence of a Schiff base formed between the bound cofactor pyridoxal-P and a lysine of the protein [6], as well as a large maximum at 278 nm characteristic of aromatic amino acid residues. As reported by Dowhan and Snell [6], treatment of the enzyme with imidazole citrate followed by successive dialysis against cysteine and buffer causes complete removal of the pyridoxal-P and hence loss of the Schiff base

209

T. A. Reed and K. D. Schnackerz flow stops

flow StODs

A

B

1 F -m

L

control

-

light increase enzyme formation

1

5 m

5-minline

Time

-

200 rns

Fig. 2. Stopped-flow traces (fthe transmittance changes at 436 nni during the reconstitution of D-serme dehydrutase f r o m the apodehydratase andpyridoxal-P. Conditions: 5 pM apodehydratase, 100 mM potassium phosphate buffer, pH 7.8, total pyridoxal-P concentration 400 pM, 25 mM 2-mercdptoethanol. In (A) the time constant is 2 ms, In (B) it 1s 0 5 ms

peak without significant alteration of the 278-nm maximum (Fig. 1, curve b). Comparison of the absorption curves of the holoenzyme and apoenzyme with those of pyridoxal-P in aqueous solution in the presence and absence of 2-mercaptoethanol (Fig. 1, curves c and d) show that for optimum observation of the redevelopment of the Schiff base peak during reconstitution of apoenzyme with a large excess of pyridoxal-P characteristic of stopped-flow studies, a wavelength higher than the Schiff base maximum (415 nm) must be used. 436 nm was chosen as the best compromise between maximal sensitivity and minimum interference by unbound cofactor. The overall course of transmittance changes at 436 nm using an 80-fold excess of total pyridoxal-P in 25 mM mercaptoethanol is shown in Fig.2. The initial change is a slight decrease in transmittance (Fig.2B), so rapid as to be at the limit of resolution of the stopped-flow apparatus. This change, though difficult to reproduce quantitatively, is normally present in reconstitutions containing 2-mercaptoethanol, but absent in reconstitutions conducted in buffer lacking this reducing agent. The major transmittance change, corresponding to an increase in absorbance of about 0.2, is seen as the second phase of Fig. 2 B and the initial phase of the extended-time-scale presentation of the same reaction (Fig. 2A). This reaction is found with the same absorbance change and magnitude of time course in systems with and without 2-mercaptoethanol. The final phase of the reconstitution reaction is best observed on an extended time scale (Fig.2A) and represents a much slower and much smaller increase in absorption coefficient at 436 nm, apparent here as a decrease in transmittance. This slow phase is also observed both in the presence and absence of 2-mercaptoethanol. After approximately five minutes no further reaction is observed, as is shown by the final reaction line superimposed on the

reaction trace with the other bench marks used in absorbance and rate calculations. D-Serine dehydratase has been shown to have an additional very weak binding site for pyridoxal-P (Kd = 1 mM) [18]. To show that the final two reaction phases do not involve secondary binding to this site, stopped-flow experiments in which the holoenzyme was reacted with large excesses of pyridoxal-P (10- 100-fold) were run under the same conditions as the reconstitution experiments and observed with identical sensitivity for five minutes. No change in transmittance was observed, thus precluding secondary binding as playing a role in the transmittance changes observed during reconstitution. The kinetic nature of the major transmittance change and the slower, smaller change which follows is apparent in the reaction trace and its first-order plot shown in Fig. 3. The concentration conditions and time scale were chosen to give a particularly clear view of the final two phases (Fig. 3A). Conversion of trnnsmittance to absorbance and first-order plotting using the five-minute line as the end of the reaction (Fig. 3 B), shows a sharply biphasic curve with an abrupt change in slope between the two ‘legs’ which permits simplified treatment of these stages of the overall reaction. The linearity indicates that the phases may be considered to be first order or pseudo first order. The behavior of these reactions as a function of increasing concentration of pyridoxal-P is a primary consideration in interpreting their nature. Stopped-flow studies were conducted at a series of increasing concentrations of pyridoxal-P. The first 80% of the transmittance change was considered to belong to the first phase of the reaction with little or no interference from the final phase. This was used to obtain an apparent first-order rate constant. The same procedure was undertaken with

210

Reconstitution of D-Serine Apodehydratase with Pyridoxal-P

1.o

f

o,5

light i n c r e a s e

enzyme

formation

~

-I.

-0

0 0

10

20

40

30 Time

50

(5)

Fig. 3. Stopped-flow experiment showing the two phases of the transmittance change at 436 n m and the apparent first-order rate of each phase. (A) Stopped-flow trace of the transmittance change at 436 nm during reconstitution. Conditions: 10 pM apodehydratase, total pyridoxal-P 100 pM, 100 mM potassium phosphate buffer, pH 7.8, 25 mM 2-mercaptoethanol; time constant 2 ms. (B) First-order plot of data from (a) after conversion of transmittance into absorbance

l2

1

A

- 0 0

300 Total pyridoxal - P

500

100

(FM)

0

100 303 500 Total pyridoxal-P ( p M )

Fig. 4. Plots of apparent first-order rate constant versus total pyridoxal-P concentration. In all cases, the measurements were made in 100 mM potassium phosphate buffer, pH 7.8, in the presence of 25 mM 2-mercaptoethanol. The measurement wavelength was 436 nm. (A) Apparent first-order rate constants obtained from the first phase of the reaction curve. (B) Apparent first-order rate constants obtained from the second phase of the reaction curve

the last 10% of change which was assumed to correspond to the final phase of the reaction. In both cases the five-minute line was used as the end point. Plots of the apparent rate constants of the two reactions as a function of total pyridoxal-P concentration are shown in Fig. 4. Within the concentration limits of the measurements, the constants corresponding to the major absorbance change are shown to be linearly proportional to the total concentration of the cofactor used in reconstitution. The case of the final reaction phase is quite different. Here, the apparent first-order rate constants are independent of the total pyridoxal-P concentration and the plot is a straight line parallel to the concentration axis. Thus, while both phases of the reaction are plotted as first-order reactions, the major change shows a pseudo first-order rate constant while the second shows a true first-order rate constant. In order to obtain more information about the nature of the pseudo first-order part of the reconsti-

tution, a further series of stopped-flow studies were run both with and without 2-mercaptoethanol. Calculation of the rate constants of the first 80% of the reactions was accomplished by the method of Guggenheim [9,10]. This method is independent of 'drift' or a slow secondary reaction and does not require a knowledge of the final.end value of the reaction. The computerized stopped-flow system was used for these studies with 1024 data points for each curve stored in a Nicolet 1090 digital oscillograph and processed on a Wang 720C programmable desk calculator. The data points used for calculation of the rate constant were the average of five data points of the initial measurement, thus providing the level of precision necessary for the application of a difference method such as that of Guggenheim. Calculations of the concentrations of free pyridoxal-P, pyridoxal-P hemiacetal, and pyridoxal-P hemimercaptal were made by the kinetic simulation program CHEK [12] in the

21 1

T. A . Reed and K. D. Schnackerz

C

A

9000 rn

1.0 0.5

1 -

8500

Po

8000

0' m

7500

. I

v T..

7000

10 I-

6500

6000

5500

5000

'

0

I / [Pyridoxal-P It,,,

(pM-')

Fig. 5. ( A ) Plot of the apparent first-order rate constants (first phase) versus the concentration of free pyridoxal-P in the presence (m) and absence of 2-mercaptoethanol ( 0 ); ( B ) double-reciprocal plot of the same data. In all cases the reactions were run in 100 mM potassium phosphate buffer, pH 7.8. The limiting rate constant is 3.04 s - l based on linear regression of 2-mercaptoethanol-containing samples. Coefficient of confidence is 0.999

course of simulation studies conducted in connection with other aspects of this project. The rate constants of the individual reactions were obtained from the literature [13,14] or by the extension of already existing measurements [15] into the pH range appropriate for these studies. Plots of the apparent first-order rate constant as a function of the concentration of the various pyridoxal-P species present were prepared. In the case of the plot involving free pyridoxal-P, the rate constants from both series of experiments were superimposed (Fig. 5 A). The total pyridoxal-P concentration was extended beyond that used in the series of experiments shown in Fig.4A, passing from the region of linear dependence of the rate constant into a region where the apparent rate constant showed saturation kinetics. A double-reciprocal plot of the data [19] gives a straight line (Fig. 5 B) with the slope corresponding to a true first-order rate constant of 3 . 0 4 ~ ~ and ' an equilibrium constant for the requisite preequilibrium of 1500 M-'. The data for both the 2-mercaptoethanol-containing system and the mercaptoethanol-free system imply the existence of a preequilibrium with a constant of the same value in each case. To examine the nature of the presumably transient prereaction equilibrium more thoroughly, experiments were conducted using the temperature-jump method

I

0.4

I

0.8 Free pyridoxal

I

I

1.2

1.6

- P (mM) Fig. 6. ( A ) Plot of reciprocal relaration titnr versus j i e e pyridoxal-P concentration f o r temperature-jump following stopped-flow mixing; ( B ) plot of' relative amplitude versus lapse time f r o m initial mixing for successive temperature-jumps. The system contained 625 pM 2-mercaptoethanol, 0.4 M KCI in 100 mM potassium phosphate buffer, pH 1.9

in combination with stopped-flow mixing [19].The measurements were made in solutions containing 0.4 M KCl to decrease the ohmic resistance of the measured solution. In the presence of 2-mercaptoethanol a monophasic relaxation curve was obtained. Plots of the reciprocal of the relaxation time versus the concentration of free pyridoxal-P as calculated previously gave a forward rate constant of 2.6 x lo6 M-' s-l and a rate constant for the reverse reaction of 5.0 x lo3 s-l corresponding to an equilibrium constant of 500 M-' (Fig. 6A). This prereaction equilibrium was shown to be transient by a succession of temperature jumps of the system at various intervals after mixing. The amplitude of the relaxation curve becomes markedly smaller over a period of minutes (Fig. 6B). No relaxation curve was observed in the system free of pyridoxal-P. A stopped-flow reaction run in the normal fashion on a system containing 0.4 M KC1 in addition to 2-mercaptoethanol showed the rate constant of the major transmittance change to be greatly reduced at that ionic strength.

DISCUSSION Removal of the cofactor from pyridoxal-P-utilizing enzymes does not always require a protein-

212

Reconstitution of D-Serine Apodehydratase with Pyridoxal-P

deforming or protein-denaturing agent as is the case for tryptophanase [20]. For D-serine dehydratase, however, a protein-deforming reagent such as imidazole citrate must be present to facilitate removal of cofactor by forming a thiazolidine derivative with cysteine. After removal of the resolution buffer, an apoenzyme which is at least partially unfolded would be expected to regain its original or a related welldefined structure [21]. Experimental evidence for rapid refolding of several enzymes on removal of denaturants was obtained in the classical studies of Ikai and Tanford [22,23]. Pyruvate kinase from Saccharomyces cerevisiae was found to refold on removal of guanidine hydrochloride at a much faster rate than the reactivation occurs [24]. Circular dichroism studies on D-serine dehydratase and apodehydratase, revealing the same a-helix content for both species, indicate that upon removal of the resolving agent the apoenzyme rapidly folds into a well-defined structure irrespective of its lack of enzymatic activity (R. Potsch, K. D. Schnackerz and T. A. Reed, unpublished results). These background data establish two important starting points in the study of D-serine dehydratase. First, in the holoenzyme the cofactor is not only bound covalently to lysine, but also ‘locked’ in place by a conformational change occurring in the enzyme during or after cofactor binding. Second, the apoenzyme, following removal of the deforming agent, has a welldefined structure. This second point appears to have a profound effect on the reconstitution of apodehydratase with pyridoxal-P. On renaturing, the cofactor binds to a single specific lysine of 19 possible lysine residues. This fact, and the multistep kinetics of the binding process, imply that the particular lysyl residue binding pyridoxal-P is in a structured environment which acts as a catalytic site to promote bond formation in a highly specific fashion. In this sense the incipient binding site may resemble the substrate binding site of a normal enzyme. The overall course of the absorbance changes at 436 nm has established that, both in the presence and absence of 2-mercaptoethanol, a pseudo first-order change is followed by a true first-order change. To avoid artifacts in the detailed analysis of the pseudo first-order reaction, the ‘short form’ of the Guggenheim method [9,10] was used to calculate the pseudo first-order rate constants for this reaction. The average coefficient of confidence for all linear fits to Guggenheim points was 0.95, ensuring that noise problems had not induced ‘wild’ values in the determination. An apparent first-order rate constant which is proportional to increasing concentrations of a reactant maintained in large excess normally arises from B e C. a second-order equilibrium of the form: A However, such a system does not show saturation

+

kinetics as was found for D-serine dehydratase (Fig. 5A). The two most straight-forward mechanisms which display such behavior are :

where A, B and C represent starting materials or intermediates, A* is an alternative or tautomeric form of A and P is the final product. It should be kept in mind that this system is treated separately from the final slow reaction with the smaller absorbance change. Mechanism (1) is relatively uncommon in nature [19], whereas mechanism (2) appears to be widely distributed. In this analysis, only mechanism (2) will be discussed since mechanism (1) can be eliminated on the basis of the temperature-jump results subsequently discussed. The lack of a lag phase in the semilog plots used to obtain apparent rate constants indicates k2 % k3, i.e. equilibrium conditions apply to the system. Were k3 9 k2, the primary reaction would display a lag phase. If, however, k3 z kz, the slope of the doublereciprocal plot in Fig. 5 B would curve upward with decreasing pyridoxal-P concentrations. Were k4 in mechanism (2) not equal to zero, the double-reciprocal plot would be nonlinear showing a decreasing slope at decreasing concentrations of B [19]. Note that the lack of a finite k4, as well as the final slow first-order reaction connected with the smaller absorbance change, means that the equilibrium concentration of free pyridoxal-P, apodehydratase or any of the possible intermediates in the equilibrium are extremely small once reconstitution is complete. Mechanism (2) with kt % k3 and k4 = 0 has thus been shown to be responsible for the changes causing the major transmittance decrease at 436 nm. Having established that equilibrium conditions prevail in the reaction system under study, the apparent first-order rate constants may be plotted according to Strickland et al. [19] in a manner resembling a Lineweaver-Burk plot (Fig. 5 B), providing the true first-order rate constant ‘k3’ and the equilibrium constant of the initial rapid reaction. When the figure is constructed using the calculated concentrations of free pyridoxal-P, the plots of apparent rate constant vs concentration can be superimposed regardless whether obtained in the presence or absence of 2mercaptoethanol. The same is true, of course, for the corresponding double-reciprocal plots. This finding is in agreement with the results of Diebler and Thorneley [26] and of O’Leary and Malik [2] who showed the active pyridoxal-P species to be the free cofactor in other systems. It is also significant that either in the presence or absence of 2-mercaptoethanol, the mechanism established by Eqn (2) and responsible

213

T. A . Reed and K. D. Schnackerz

for the larger absorbance change shows a rapid initial step preceding the observed absorbance change. A similar result was postulated by O’Leary and Malik [2] for the reconstitution of glutamate decarboxylase on the basis of the kinetics of developing enzyme activity. Relaxation curves induced by temperature-jump were obtained after reconstitution had been initiated by stopped-flow mixing, but prior to completion of the reaction. The system was in transient equilibrium, as established in the previous section. A relaxation curve was only observed in the presence of 2-mercaptoethanol. Since this relaxation curve is not related to the mercaptoethanol/water/pyridoxal-Psystem itself, it appears to correlate with the initial rapid absorbance change observed in stopped-flow experiments in the presence of 2-mercaptoethanol. The probable cause of this anomaly has been determined by computer simulation [ l l ] of the reacting system in the presence and absence of 2-mercaptoethanol. Due to the size of the rate constants involved and the concentrations of the reactants present, a change in the concentration of the hemimercaptal of pyridoxal-P ‘follows’ the concentration of free pyridoxal-P very closely. Thus, initial rapid binding of pyridoxal-P to the apoenzyme is mirrored by a shift in the concentration of the hemimercaptal even though this initial binding does not appear to alter significantly the absorption coefficient of the bound pyridoxal-P. This change in the hemimercaptal concentration serves as an indicator of pyridoxal-P binding and is observed by the stopped-flow spectrometer as the binding change. The rate of change of pyridoxal-P hemimercaptal in the system composed of pyridoxal-P-hydrate, free pyridoxal-P and pyridoxal-P hemimercaptal is so rapid that the change of free to bound pyridoxal-P is exactly followed. The adjustment of the pyridoxal-P hydrate concentration is very slow and the concentration of this species is constant during the time course of the initial rapid binding reaction. This does not disturb the clear visualization of the initial binding step in the presence of 2-mercaptoethanol. In the absence of 2-mercaptoethanol, no change is observed since the initial binding of the pyridoxal-P to apoenzyme does not cause a sufficiently large absorbance change in itself to be observed and the adjustment of the equilibrium between pyridoxal-P and water occurs so slowly as to be lost against the very large changes in absorbance induced by subsequent steps of the reconstitution process. The increase of reciprocal relaxation time as a function of free pyridoxal-P concentration (Fig. 6 A) is then significant in differentiating between mechanisms (1) and (2) since it appears to reflect the course of the initial reaction step. Reaction mechanism (1) would call for a decrease in the reciprocal relaxation time as a function of free pyridoxal-P [27] and can therefore be discarded from serious consideration.

A discrepancy is noted between the values of the equilibrium constant for the initial rapid reaction as determined by analysis of the pseudo first-order stopped-flow data (Fig. 5) and the data obtained from temperature-jump in conjunction with stoppedflow mixing. This is due to the different ionic strength of the reaction mixtures used in the two types of measurement. Detailed analysis of the effect of ionic strength on protein structure has been presented by Edsal and Wyman [28] and more recently reviewed by Record et al. [29]. Hess [30] has shown experimentally that ionic strength exerts a marked influence on the structure-function relationship of the purple membrane. This is confirmed by the stopped-flow reactions studied at high ionic strength which show a greatly decreased rate of reconstitution as revealed by the 436-nm transmittance plotted as a function of time. In summary, detailed analysis of the stopped-flow curves representing Schiff base formation during reconstitution of D-serine dehydratase has established a rapid initial binding step followed by a rapid increase in 436-nm absorbance corresponding to Schiff base formation in the reconstituting enzyme. This second phase has been shown to be a first-order reaction. This is followed by a final, slower first-order change which is independent of the rate of formation of the Schiff base. The rapid initial binding step has been confirmed by both direct observation of a very fast phase in the stopped-flow curves of systems containing 2-mercaptoethanol and temperature-jump experiments using stopped-flow mixing to permit the observation of relaxation times of transient intermediates. This direct observation has been made possibly by an indicator reaction involving the equilibrium between free pyridoxal-P and its hemimercaptal. The final slow change in 436-nm absorbance presumably represents a conformational change of the enzyme which results in the Schiff-base-bound pyridoxal-P being locked into the enzyme. These steps are shown in the following mechanism in which A represents apoenzyme, P represents free pyridoxal-P, A-P is an initial complex between these reactants, (AP) is the complex after Schiff base formation and H is the holoenzyme : A

+P

A-P $ (AP)

+

H

Although these reactions display the reconstitution and alterations in structure as a function of Schiff base formation, specific conformational changes in the protein moiety of the enzyme are coordinated with these Schiff base changes and the development of enzyme activity. The structural changes of the protein moiety as reflected by fluorescence kinetics and dichroic spectral changes are currently under investigation.

214

T. A. Reed and K. D. Schnackerz: Reconstitution of D-Serine Apodehydratase with Pyridoxal-P

This research was supported in part by a grant of the Deutsche Forschungsgemeinschaft (Schn 139/3). We are indebted to Mrs G. Wirths for preparing the enzyme. We are also grateful to Professor Benno Hess for giving us access to the IBM 360/44 Computer at the Max-Planck-Institut fur Ernahrungsphysiologie, Dortmund. The authors wish to thank Professor Ernst Helmreich for his stimulating discussions, particular those in reference to the problem of conformational change ‘locking’ the cofactor into the protein. We are indebted to L. Veil and H. Winkler, both at the Max-PlanckInstitur f u r Biophysikalische Chemie, Gottingen, for performing the temperature-jump experiments and the determination of the rate constants for the reaction between pyridoxal-P and 2-mercaptoethanol, respectively.

REFERENCES 1. Snell, E. E. & DiMari, S. J. (1970) in The Enzymes (Boyer, P. D., ed.) vol. 2, pp. 335-370, Academic Press, New York. 2. O’Leary, M. H. & Malik, J. M. (1972) J . B i d . Chem. 247, 7097-7105. 3. Churchich, J. E. & Farrelly, J. G. (1969) J . B i d . Chem. 244, 3685- 3690. 4. Banks, B. C . E. &Vernon, C . A . (1961) J . Chem. Soc. (Lond.) 1968- 1972. 5. Arrio-DuPont, M. (1972) Eur. J . Biochem. 30, 307-317. 6. Dowhan, W. & Snell, E. E. (1970) J . Bid. Chem. 245, 46294635. 7. Reed, T. A,, Potsch, R. & Schnackerz, K. D. (1972) Abstr. Cornmun. 8th Meet. Fed. Eur. Biochem. Soc., no 380. 8. Reed, T. A,, Schnackerz, K. D. & Veil, L. (1973) Abstr. 9th Int. Congr. Biochem., no 204. 9. Guggenheim, E. A. (1926) Philos. Mug. 2, 538-542. 10. Gutfreund, H. (1972) Enzymes: Physical Principles, pp. 118, Wiley-Interscience, New York. 11. Reference deleted. 12. Curtis, A. R. &Chance, E. M. (1974) CHEKund C H E K M A T : Two Chemical Reaction Programs, United Kingdom Atomic

Energy Authority, publication no 7345, H.M. Stationary Office, London. 13. Morozov, Yu. V., Bazhulina, A. P., Cherkashina, L. P. & Karpeisky, M. Y. (1967) Biofizika, 12, 397-404. 14. Ahrens, L. H., Maass, G., Schuster, P. & Winkler, H. (1969) FEBS Lett. 5, 327-330. 15. Schuster, P. & Winkler, H. (1970) Tetrahedron, 26,2249-2256. 16. Benesi, H. A. & Hildebrand, J. H. (1949) J . Am. Chem. Soc. 71, 2703 - 2707. 17. Eigen, M. & DeMayer, L. (1963) in Technique of Organic Chemistry (Friess, S . L., Lewis, E. S. & Weissberger, A,, eds) vol. 8, part 2, pp. 895 - 1054, Wiley, New York. 18. Huang, Y. Z. & Snell, E. E. (1972) J . Bid. Chem. 247, 73587364. 19. Strickland, S., Palmer, G . & Massey, V. (1975) J . Biol. Chem. 250,4048 -4052. 20. Newton, W. A,, Morino, Y. & Snell, E. E. (1965) J . Bid. Chem. 240, 1211-1218. 21. Chou, P. Y. & Fassman, G. D. (1974) Biochemistry, 13, 222245. 22. Ikai, A. & Tanford, C . (1973) Mol. Biol. 73, 145-163. 23. Ikai, A., Fish, W. W. & Tanford, C . (1973) J. Mol. Bid. 73, 165 - 184. 24. Hess, B., Wetlaufer, D. B., Bornmann, L., Reed, T. & Tamaki, N. (1975) Proc. FEBS Meet. 40, 49-62. 25. Reference deleted. 26. Diebler, H. & Thorneley, R. N. F. (1973) J . Am. Chem. Soc. 95,896 - 904. 27. Halford, S. E. (1972) Biochem. J . 126, 727-738. 28. Edsal, J. T. & Wyman, J. (1958) Biophysical Chemistry, vol. 1, Academic Press, New York. 29. Record, T. M., Andersen, C. D. & Lohman, T. M. (1978) Q. Rev. Biophys. 11, 103-178. 30. Hess, B. (1977) Ionic Control of Proton and Electron Transfer, Linderstr4m-Lang Conference on Biochemical Control Mechanisms of Respiration of the Molecular and Tissue Level, Espoc, Finland.

K. D. Schnackerz, Physiologisch-Chemisches Institut der Julius-Maximilians-Universitat Wurzburg, KoellikerstraDe 2, D-8700 Wurzburg, Federal Republic of Germany T. A. Reed, Max-Planck-Institut fur Ernahrungsphysiologie, Rheinlanddamm 201, D-4600 Dortmund, Federal Republic of Germany