- Email: [email protected]
Drosophila lebanonensis alcohol dehydrogenase: pH dependence of the kinetic coefficients
Biochimica et Biophysica Acta 1431 (1999) 74^86
Drosophila lebanonensis alcohol dehydrogenase: pH dependence of the kinetic coe¤cients Michelle Kaaber Brendskag a , John S. McKinley-McKee b , Jan-Olof Winberg a
a;
*
Department of Biochemistry, Institute of Medical Biology, University of TromsÖ, 9037 TromsÖ, Norway b Biochemical Institute, University of Oslo, Oslo, Norway Received 22 October 1998; received in revised form 14 January 1999; accepted 28 January 1999
Abstract The alcohol dehydrogenase (ADH) from Drosophila lebanonensis shows 82% positional identity to the alcohol dehydrogenases from Drosophila melanogaster. These insect ADHs belong to the short-chain dehydrogenase/reductase family which lack metal ions in their active site. In this family, it appears that the function of zinc in medium chain dehydrogenases has been replaced by three amino acids, Ser138 , Tyr151 and Lys155 . The present work on D. lebanonensis ADH has been performed in order to obtain information about reaction mechanism, and possible differences in topology and electrostatic properties in the vicinity of the catalytic residues in ADHs from various species of Drosophila. Thus the pH dependence of various kinetic coefficients has been studied. Both in the oxidation of alcohols and in the reduction of aldehydes, the reaction mechanism of D. lebanonensis ADH in the pH 6^10 region was consistent with a compulsory ordered pathway, with the coenzymes as the outer substrates. Over the entire pH region, the rate limiting step for the oxidation of secondary alcohols such as propan-2-ol was the release of the coenzyme product from the enzyme-NADH complex. In the oxidation of ethanol at least two steps were rate limiting, the hydride transfer step and the dissociation of NADH from the binary enzyme-NADH product complex. In the reduction of acetaldehyde, the rate limiting step was the dissociation of NAD from the binary enzyme-NAD product complex. The pH dependences of the kon velocity curves for the two coenzymes were the opposite of each other, i.e. kon increased for NAD and decreased for NADH with increasing pH. The two curves appeared complex and the kon velocity for the two coenzymes seemed to be regulated by several groups in the free enzyme. The kon velocity for ethanol and the ethanol competitive inhibitor pyrazole increased with pH and was regulated through the ionization of a single group in the binary enzyme-NAD complex, with a pKa value of 7.1. The kon velocity for acetaldehyde was pH independent and showed that in the enzyme-NADH complex, the pKa value of the catalytic residue must be above 10. The koff velocity of NAD appeared to be partly regulated by the catalytic residue, and protonation resulted in an increased dissociation rate. The koff velocity for NADH and the hydride transfer step was pH independent. In D. lebanonensis ADH, the pKa value of the catalytic residue was 0.5 pH units lower than in the ADHS alleloenzyme from D. melanogaster. Thus it can be concluded that while most of the topology of the active site is mainly conserved in these two distantly related enzymes, the microenvironment and electrostatic properties around the catalytic residues differ. ß 1999 Elsevier Science B.V. All rights reserved.
Abbreviations: ADH, alcohol dehydrogenase (EC 1.1.1.1); ADHS , slow alleloenzyme from Drosophila melanogaster; E, enzyme; EO or ES1 , enzyme-NAD ; EOI, enzyme-NAD -pyrazole; ER or ESP1 , enzyme-NADH ; ES1 S2 , enzyme-NAD -alcohol; ESP1 SP2 , enzymeNADH-aldehyde ; I, inhibitor; S1 or O, NAD ; SP1 or R, NADH; S2 , alcohol; SP2 , aldehyde * Corresponding author. Fax: +47 (77) 645350; E-mail: [email protected] 0167-4838 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 9 ) 0 0 0 2 8 - X
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
75
Keywords: Drosophila; Alcohol dehydrogenase; Reaction mechanism; pH study; Catalytic site; Inhibition
1. Introduction The gene for alcohol dehydrogenase (Adh) has been cloned and sequenced from various Drosophila species, and the enzyme has been puri¢ed and biochemically characterized from some of these species (for review, see [1^3]). The ADH enzyme catalyzes the interconversion of alcohols and their corresponding aldehyde/ketone products (Eq. 1). Alcohol NAD Haldehyde=ketone NADH H 1 Drosophila ADH is a dimer of Mr 54 800, consisting of two identical subunits [1^3], and belongs to the short-chain dehydrogenase/reductase (SDR) family [4,5] which lack metal ions in their active site. In this family, Ser138 , Tyr151 and Lys155 in the amino acid sequence are conserved (Drosophila lebanonensis nomenclature) [3,6,7], and site-directed mutagenesis showed that all three residues in contrast to the two cysteine residues in the enzyme were essential for activity [8^11]. Binding of alcohol and alcohol competitive inhibitors to the Drosophila melanogaster ADHS alleloenzyme was dependent on a single residue in the active site, with a pKa value of about 7.6 in the binary enzyme-NAD (ES1 ) complex, and above 10 in the binary enzyme-NADH (ESP1 ) complex [12]. Previously we showed that certain steps in the mechanism of alcohol oxidation for Drosophila ADH are di¡erent from that known to apply for zinc mediated liver ADH catalysis [13]. To facilitate hydride transfer from the alcohol, the group with a pKa of 7.6 in the D. melanogaster ADHS alleloenzyme must be a residue with a full negative charge, as its function is to produce an alcoholate anion [13]. This indicated that Tyr151 could be the residue which interacts with the hydroxyl group in the alcohol and hence takes over the function of zinc in horse liver ADH [12,13]. X-Ray crystallography of ¢ve members of the SDR family [14^19] showed that Tyr151 and Lys155 reside in K-helix F, suggesting that an interaction between Lys155 and Tyr151 decreases the pKa of the tyrosine group. These studies also indicated that the highly conserved Ser138
residue was located in the catalytic region of the enzyme. Based on studies using alternate substrates, deadend and product inhibitors [3,12,20], it has been shown that Drosophila ADHs follow a compulsory ordered mechanism in the interconversion of alcohols and their corresponding aldehyde and ketone products (Scheme 1). The ADH from the distantly related species D. lebanonensis has only 82% positional identity to the ADH alleloenzymes from D. melanogaster [3,21^23]. Both these Drosophila species are well adapted to environments with a high alcohol content due to a large amount of ADH compared to other Drosophila species [24]. However, D. lebanonensis £ies also contains approximately ¢ve times as much ADH as D. melanogaster £ies [24]. Due to the large amount of ADH in D. lebanonensis £ies, future studies concerning general properties of the insect ADHs will no doubt be performed on the enzyme in this Drosophila species. The substrate specificity of D. lebanonensis ADH was similar, but not identical to that of the other Drosophila ADHs studied [3,25]. These studies focused on the similarities of the di¡erent Drosophila ADHs and stressed that the major topology of the active site region was conserved. However, it is also important to design experiments to determine what distinguishes these enzymes, and this has been done in the present work. Minor di¡erences in the topology and electrostatic properties of the alcohol binding region might result not only in changes in the pKa value of the amino acid that binds the hydroxyl group of an alcohol and the polar group of alcohol competitive inhibitors, but also in changes in the individual rate constants that build up the kinetic coe¤cients. This may result in altered parameters in£uencing for example inhibitory patterns and/or the magnitude of primary and secondary isotope e¡ects. It was also of interest to investigate whether D. lebanonensis ADH like the D. melanogaster ADHS alleloenzyme followed a strict compulsory ordered mechanism over the entire pH region studied [12], as a change in reaction mechanism will in£uence the interpretation of the pH dependence of the di¡erent kinetic coe¤cients.
BBAPRO 35865 30-3-99
76
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
2. Experimental 2.1. Reagents Grade III NAD , NADH, and [2 H8 ]propan-2-ol were from Sigma. Ethanol (96%) was from A/S Vinmonopolet and [2 H6 ]ethanol was from Norsk Hydro. Anhydrous acetaldehyde (puriss p.a.), and pyrazole were from Fluka. Propan-2-ol (99.7%) was from Merck. 2.2. Enzyme D. lebanonensis ADH was puri¢ed as described previously [25]. Freeze-dried samples were dissolved in 0.1 M phosphate bu¡er pH 7.0, and dialyzed against two changes of the same bu¡er at 4³C. Any denatured protein was removed by centrifugation for 20 min at 25 000Ug. 2.3. Rate assay To determine the amount of enzyme, i.e. the enzyme active site concentration in the assay cuvette, the spectrophotometric rate assay for D. lebanonensis ADH previously described was used [25]. The enzyme concentration is expressed as the amount of subunits in nM. This is twice the amount of enzyme molecules as the enzyme is a dimer. The assay solution consisted of 0.5 mM NAD and 100 mM ethanol in a total volume of 1 ml 0.1 M glycine-NaOH bu¡er pH 9.5. In this assay, 1 WM of enzyme active sites gave a rate of 0.82 vA340nm /min. 2.4. Kinetic measurements The oxidation of alcohols and the reduction of acetaldehyde was studied by steady state kinetics at 23.5³C, and the initial rates were measured by following the appearance of NADH £uorescence, using a sensitive ¢lter £uorometer as described previously by Winberg and McKinley-McKee [26]. To normalize the results due to any variations in the light intensity of the xenon lamp, the de£ection of a £uorescent Perspex standard (equivalent to a NADH concentration of about 7 WM) was measured at intervals during initial rate measurements. Since
NADH £uorescence is quenched by increasing NADH concentrations, a calibration curve of the £uorescence against NADH concentration was determined at 23.5³C, with slit widths of 2 mm for both exciting and emitted light as described previously [26]. All initial rate determinations were within the linear part of the standard curve. Bu¡ers used were 0.1 M pyrophosphate (pH 9^10), or 0.1 M phosphate (pH 6^8). The reported aldehyde concentrations represent the sum of the free and hydrated species present. In all experiments, the Lineweaver-Burk plots obtained were linear within the experimental error. The data ¢tted Eq. 2: e=v0 P0 P1 =S1 P2 =S2 P12 =S1 S2
2
where e is the concentration of enzyme active sites and S1 and S2 are coenzyme and substrate respectively. The unprimed symbols (P, S) are used for the NAD /alcohol reactions and the primed symbols (PP, SP) are used for the NADH/aldehyde reactions. The kinetic coe¤cients in Eq. 2 were obtained from primary and secondary plots [27]. Vmax /e or kcat is 1/P0 , the catalytic center activity (molecules of product produced per second per enzyme subunit active site). Km for coenzyme and substrate is P1 /P0 and P2 /P0 , while kcat /Km for coenzyme and substrate is 1/P1 and 1/P2 respectively. The concentrations of coenzyme and substrate used were within the following ranges: NAD 5^2000 WM, ethanol 1^500 mM, propan-2-ol 0.1^40 mM, [2 H8 ]propan-2-ol 0.75^ 80 mM, [2 H6 ]ethanol 20^100 mM for the alcohol/ NAD reactions and NADH 1^10 WM, acetaldehyde 0.06^2 mM, for the aldehyde/NADH reactions. Inhibition experiments were carried out with either a constant NAD concentration and varied ethanol concentrations, or constant ethanol concentration and varied NAD concentrations. 2.5. pH dependence of kinetic coe¤cients Scheme 1 shows a general compulsory ordered mechanism at a single pH value. This and other mechanisms can be described by the initial rate Eq. 2, where the kinetic coe¤cients are complex functions of the rate constants. Typical features for a compulsory ordered mechanism are that the rate
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
77
Scheme 1.
constants for the formation and dissociation of the forward binary enzyme-coenzyme complex are given by k1 = 1/P1 , k31 = P12 /P1 P2 and its dissociation constant by P12 /P2 , while primed values are for the reverse binary complex. The Haldane relation for Keq is PP12 /P12 . If protonation of an essential amino acid in the enzyme is of kinetic signi¢cance in the pH region studied, two charged forms E and EH of the free enzyme appear. The equilibrium between E and EH is characterized by a proton dissociation constant K1 and the fractional concentration of the two di¡erent enzyme forms is then E = 1/(1+[H ]/K1 ) and EH = 1/(1+K1 /[H ]). If both enzyme species react with the substrate at any pH value, the sum of the products of the relative concentrations of each charge type of the enzyme and the appropriate rate constants should yield the observed rate at that pH value. The experimentally obtained rate constant (k) at any pH value can then be expressed as Eq. 3: k
k k 1 H =K 1 1 K 1 =H
3
where k* and k** are either the association or the dissociation rates of the substrate with the charged enzyme forms E and EH. For a substrate-competitive inhibitor or a substrate which binds to both enzyme species, the pH dependence of the inhibitor constant or enzyme-substrate dissociation constant Ki is then expressed as Eq. 4: K 1 H =K 1 Ki i 1 H =K 2
4
where K1 is the proton dissociation constant of the equilibrium between E and EH, and K2 that of the corresponding equilibrium of the enzyme-inhibitor/ substrate complexes. The relation of the enzyme-inhibitor or enzyme-substrate dissociation constant (Ki *) of the unprotonated enzyme and the corresponding constant (Kj ) of the protonated enzyme is then K1 Kj = K2 Ki * [28].
Fig. 1. Primary (a,d) and secondary (b,c,e,f) plots at pH 10 and 23.5³C: variation of the speci¢c initial rate, e/v, with the reciprocal of the ethanol concentration (a) and NAD concentration (b), for several constant NAD and ethanol concentrations. (a) The NAD concentrations were: a, 0.5 mM; E, 0.02 mM; O, 0.01 mM; P, 0.0067 mM; 7, 0.005 mM. (d) The ethanol concentrations were: a, 100 mM; E, 8 mM; O, 4 mM; P, 2 mM; 7, 1 mM. Variation of the intercepts (b,e) and slopes (c,f) of the primary plots with the reciprocal of the NAD concentration (b,c) and ethanol concentration (e,f). P0 is determined from the intercepts in (b) and (e), P1 from the slope in (b) and intercept in (f), P2 from the intercept in (c) and slope in (e), and P12 from the slopes in (c) and (f).
BBAPRO 35865 30-3-99
78
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
3. Results 3.1. Oxidation of alcohols Initial rate data for the ethanol/NAD reaction at pH 10.0 are shown in the primary plots of Fig. 1a,d. The intercepts and slopes obtained are plotted in Fig. 1b,c,e,f. The slopes and intercepts of these secondary plots yield values for the four kinetic coe¤cients in Eq. 2. Similar experiments were performed with ethanol in the pH range 6.0^10.0. In all cases, both primary and secondary plots were linear within the experimental error over the range of substrate and coenzyme concentrations used. Table 1 lists the initial rate parameters at 23.5³C. The kinetic coe¤cients for the oxidation of the alternative substrates [2 H6 ]ethanol, propan-2-ol and [2 H8 ]propan-2-ol at pH 6.0 and 9.5 at 23.5³C are listed in Table 2. In the case of propan-2-ol, there H H was no kinetic isotope e¡ect on either PD 0 /P0 (kcat / D D H H D kcat ) or P1 /P1 ((kcat /Km ) /(kcat /Km ) for NAD ) at the two pH values tested. At pH 9.5, a kinetic isoH H tope e¡ect of 2.6 was observed on PD 2 /P2 ((kcat /Km ) / D (kcat /Km ) for alcohol), while this parameter was only 1.3 at pH 6.0. At both pH values, a kinetic H isotope e¡ect was also observed on PD 12 /P12 . With H ethanol, a primary kinetic isotope e¡ect (PD 0 /P0 ) of 3.6 appeared at pH 6.0, similar to previous observations at pH 9.5 [25]. No isotope e¡ect was observed H D H for PD 1 /P1 , while the e¡ect was 4.5 on P2 /P2 and 9.3 D H on P12 /P12 .
Fig. 2. Variation with pH of (a) 1/P2 for ethanol, (b) 1/KEO;I for pyrazole. (a) Variation with pH of 1/P2 for ethanol; the theoretical curve represents an acid dissociation constant of pKa 7.1. (b) Variation with pH of 1/KEO;I for pyrazole; the theoretical curve represents an acid dissociation constant of pKa 7.1. In both (a) and (b), the pKa value was determined using the ENZFITTER program (Elsevier-BIOSOFT).
6.0^10.0 are listed in Table 3. In all cases, both primary and secondary plots were linear within the experimental error over the range of substrate and coenzyme concentrations used. 3.3. Inhibition studies Pyrazole was a competitive inhibitor against varied ethanol and a ¢xed concentration of NAD (1 mM) in the pH range 6.0^10.0 at 23.5³C. The Kis and KEO;I values determined are listed in Table 4. Pyrazole showed uncompetitive inhibition against varied NAD and a ¢xed concentration of 100 mM of ethanol at pH 6.0 and 10.0. NADH was a NAD competitive inhibitor at pH 6.0 and 10.0, and the obtained Kis values were 0.9 and 2.5 WM, respectively.
3.2. Reduction of acetaldehyde The initial rate parameters describing the reduction of acetaldehyde by NADH in the pH range
Table 1 pH dependence of the kinetic coe¤cients for the oxidation of ethanol by NAD at 23.5³C pH
P0 (s)
P1 (WMs)
P2 (mMs)
P12 (mM2 s)
P1 /P0 (WM)
P2 /P0 (mM)
P12 /P2 (WM)
10.0 9.5a 9.0 8.0 7.0 6.0
0.41 0.42 0.49 0.40 0.47 0.50
3.0 4.5 5.4 8.2 24 49
2.3 2.5 2.3 2.8 5.0 23
0.038 0.059 0.16 0.43 2.7 21
7.3 11 11 21 51 98
5.6 6.0 4.7 7.0 11 46
17 23 70 156 551 913
The kinetic coe¤cients are those in the reciprocal rate equation (Eq. 2), where S1 and S2 are NAD and ethanol, respectively. a Results from [25].
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
79
Table 2 Kinetic coe¤cients for the oxidation of alternate alcohol substrates with NAD at 23.5³C Substrate
pH
P0 (s)
P1 (WMs)
P2 (mMs)
P12 (mM2 s)
P1 /P0 (WM)
P2 /P0 (mM)
P12 /P2 (WM)
Ethanol [2 H6 ]Ethanol Propan-2-ol [2 H8 ]Propan-2-ol Propan-2-ola [2 H8 ]Propan-2-ol
6.0 6.0 6.0 6.0 9.5 9.5
0.50 1.78 0.15 0.16 0.14 0.13
49 29 46 35 3.8 3.7
23 103 4.1 5.4 0.16 0.41
21 196 3.0 11 0.0027 0.027
98 16 307 219 28 28
46 58 27 34 1.1 3.2
913 1903 732 2037 17 66
The kinetic coe¤cients are those in the reciprocal rate equation (Eq. 2). The bu¡ers used are 0.1 M phosphate bu¡er at pH 6.0 and 0.1 M glycine-NaOH bu¡er at pH 9.5. a Results from [25].
3.4. pH dependence of selected kinetic coe¤cients
4. Discussion
Fig. 2a shows for the oxidation of ethanol the variation of 1/P2 with pH, and a pKa value of 7.1 was determined. Fig. 2b shows the variation of 1/KEO;I with pH for the ethanol competitive inhibitor pyrazole and a pKa value of 7.1 was obtained. The kinetic coe¤cient P0 for the oxidation of ethanol and propan-2-ol did not vary within the pH region studied (Tables 1 and 2), and neither did the PP2 coe¤cient for the reduction of acetaldehyde (Table 3). Fig. 3a shows that 1/P1 (k1 ) increases and 1/PP1 (kP1 ) decreases with increasing pH. The dissociation rate of NAD from the binary enzyme-NAD complex (k31 ) was estimated as the average of 1/PP0 and P12 /P1 P2 (Tables 1^3) and Fig. 3b shows that k31 varies with pH. Fig. 3c shows that KE;O decreases and KE;R increases with increasing pH. A plot of log(PP2 [H ]/P2 ) vs. pH shows a slope of 31 in the pH region 7^10 (Fig. 3d).
4.1. Reaction mechanism Previously it was shown that the D. melanogaster alleloenzyme ADHS followed a compulsory ordered reaction mechanism with the coenzymes as the outer substrates in the pH region 6.0^10.0 (Scheme 1). This conclusion was based on Dalziel relations of the kinetic coe¤cients, alternate substrates, dead-end and product inhibition studies [3,12,20]. The same mechanism was suggested for D. lebanonensis ADH at pH 9.5, using alternate substrates and dead-end inhibitors [25]. Pyrazole, which was an ethanol competitive inhibitor over the entire pH region with D. lebanonensis ADH, showed uncompetitive inhibition against varied NAD at both pH 6 and 10. This indicated that the D. lebanonensis enzyme followed a compulsory ordered mechanism over the entire pH region studied. In the case of a rapid random mechanism, pyrazole should have shown a non-competi-
Table 3 pH dependence of the kinetic coe¤cients for the reduction of acetaldehyde by NADH at 23.5³C pH
PP0 (s)
PP1 (WMs)
PP2 (mMs)
PP12 (mM2 s)
PP1 /PP0 (WM)
PP2 /PP0 (mM)
PP12 /PP2 (WM)
10.0 9.0 8.0 7.5 7.0 6.0
0.47 0.14 0.050 0.045 0.041 0.040
1.22 0.52 0.36 0.27 0.20 0.14
0.25 0.37 0.35 0.54 0.43 0.36
0.00162 0.00097 0.00093 0.00068 0.00105 0.00047
2.6 3.7 7.2 9.0 4.9 3.5
0.53 2.6 7.0 18 11 9.0
6.5 2.6 2.7 1.3 2.4 1.3
The kinetic coe¤cients are those in the reciprocal rate equation (Eq. 2), where SP1 and SP2 are NADH and acetaldehyde, respectively. Bu¡ers used are as described in Section 2.
BBAPRO 35865 30-3-99
80
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
tive pattern against varied NAD . In a compulsory ordered mechanism, the value of the P coe¤cient for the outer substrate should be the same independent of the nature of the inner substrate. Table 2 shows that this was the case for the P1 coe¤cient at both pH 6.0 and 9.5 with di¡erent alcohol substrates. Thus the results obtained with two independent methods were consistent with alcohols being oxidized in a compulsory ordered mechanism over the entire pH region tested. A compulsory ordered reaction mechanism requires that the maximum rate relations P1 P2 /P12 PP0 (1) and PP1 PP2 /PP12 P0 (2) should be less than unity if ternary complexes are kinetically signi¢cant, and equal to unity in the limiting case of a TheorellChance mechanism, i.e. the dissociation of the outer product from the binary enzyme-product complex is rate limiting. For D. lebanonensis ADH, the maximum rate relation (1) can be regarded as unity in the pH region 6^10 (Table 5), with the deviation from unity attributable to the experimental errors in the kinetic coe¤cients determined. Thus in the reduction of acetaldehyde to ethanol, the dissociation of S1 from the ES1 complex (k31 ) is rate limiting between pH 6 and 10. This is similar to the previous observations with the ADHS alleloenzyme from D. melanogaster [12]. Table 4 pH dependence of the inhibitor constants Kis and KEO;I for pyrazole at 23.5³C pH 10.0 9.0 8.0 7.0 6.0
23.5³C Kis (WM)
KEO;I (WM)
12 13 17 34 121
12 12 15 22 63
pH dependence of the competitive inhibition constants Kis and KEO;I for pyrazole against varied ethanol and a ¢xed concentration of NAD (O) of 1.0 mM. The bu¡ers used at the di¡erent pH values are as described in Section 2. The neutral form of pyrazole was calculated from a pKa value of 2.5 [37] and used in the determination of the Kis and KEO;I values. Kis is equal to [I]/((si /s0 )31), where [I] is the inhibitor concentration and si and s0 are respectively the inhibited and uninhibited slopes from the Lineweaver-Burk plots. KEO;I is equal to Kis /((P12 /P2 [O])+1), and represents the dissociation constant of I from the ternary enzyme-NAD -inhibitor complex. P12 and P2 are the kinetic coef¢cients from Table 1.
Fig. 3. Variation with pH of (a) the speci¢c rates for the formation of the binary enzyme-NADH (kP1 ) and enzyme-NAD (k1 ) complexes, (b) the speci¢c dissociation rate (k31 ) of the enzyme-NAD complex, (c) the equilibrium constants of the enzyme-NAD (KE;O ) and enzyme-NADH (KE;R ) complexes, (d) log(PP2 [H ]/P2 ), the equilibrium constant of the reaction of the enzyme-coenzyme complexes with substrates. Variation with pH of (a) the speci¢c rates for the formation of the enzyme-NADH complex (1/PP1 or kP1 : b) and the enzyme-NAD complex (1/P1 or k1 : a). (b) The speci¢c dissociation rate (k31 ) of the enzyme-NAD complex. The k31 constant is the average of 1/PP0 for acetaldehyde reduction and P12 /P1 P2 for alcohol oxidation (Tables 1^3). (c) The dissociation constants of the enzymeNAD (KE;O : b) and enzyme-NADH (KE;R : a) complexes. The dissociation constants are the average values (mean þ S.D.) obtained from Table 6. (d) Log(PP2 [H ]/P2 ), the equilibrium constant of the reaction of the enzyme-coenzyme complexes with their substrates.
In the case of propan-2-ol and other secondary alcohols, it was previously shown that they were oxidized in a Theorell-Chance mechanism with rate limiting dissociation of SP1 from the ESP1 complex at pH 9.5 [3,12,20,25]. The present work shows that this is also the case at pH 6. This is based on the fact that there was no primary isotope e¡ect with propan-2-ol H and [2 H8 ]propan-2-ol, i.e. the ratio PD 0 /P0 is unity (Table 2). In a case where PP1 PP2 /PP12 is from the acetaldehyde reduction and represents 1/kP31 , and
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
1/P0 is for propan-2-ol oxidation and equals kP31 , the maximum rate relation will be unity. This is also the case for D. lebanonensis ADH in the pH region 6.0^ 10.0 as the maximum rate relation (3) shows (Table 5). Thus propan-2-ol is oxidized in a TheorellChance type of mechanism over the entire pH region studied. Table 5 shows that the maximum rate relation (2) is less than unity, which indicates a compulsory ordered mechanism where the rate limiting step in the oxidation of ethanol involves steps prior to NADH release from the binary ESP1 (ER) product complex or a combination of the two product release steps and the hydride transfer step. With Drosophila ADHs, imidazole was previously shown to form an ES1 I (EOI) complex in the ¢rst phase of the reaction resulting in inhibition of reaction. In addition it can form an ESP1 I (ERI) complex in the second phase of the reaction where SP1 dissociates faster from the ternary ESP1 I than from the binary ESP1 , and hence results in a stimulatory reaction [25,26,30]. In a reaction sequence where both these complexes are formed, an inhibitory pattern called inhibition with stimulation will appear [25,26,30]. Imidazole was an ethanol competitive inhibitor [25], which showed that the contribution of kP31 (ESP1 dissociation) to the rate limiting step must be limited. Previous studies with various secondary alcohols at pH 9.5 gave an average value of 7.4 s31 for kP31 [25], which shows that the contribution of the ESP1 dissociation step to the P0 value for ethanol oxidation is approx. 32%. H The primary kinetic isotope e¡ect on PD 0 /P0 for ethaTable 5 Maximum rate relations at 23.5³C
81
nol and [2 H6 ]ethanol (Table 2) indicates that the interconversion of the ternary complexes, i.e. the hydride transfer step (k), largely contributes to the rate limiting step. If it is assumed that kP32 Fk and kP, then P0 equals 1/k+1/kP31 and k is calculated to be 3.4 s31 . Thus, the hydride transfer step contributes 68% of the P0 value. This is similar to the ADHS alleloenzyme from D. melanogaster, where kP31 and k were estimated to be 3.4 s31 (average value of kcat for secondary alcohols) and 2.0 s31 , respectively [12,29]. In the D. melanogaster enzyme, the ESP1 dissociation step and the hydride transfer step contributed 38% and 62%, respectively, to the P0 value for ethanol oxidation. The di¡erent magnitude and hence contribution of k and kP31 to the rate limiting step in the oxidation of ethanol in these two Drosophila ADHPs were seen in the di¡erent magnitudes of H the primary kinetic isotope e¡ects (PD 0 /P0 ); 2.3 in D. S melanogaster ADH [29] and 3.0^3.6 in D. lebanonensis ADH ([25]; present work). In addition, imidazole was a competitive inhibitor with stimulation against varied ethanol with the former enzyme [26], while it was a competitive inhibitor with the latter enzyme [25]. Thus it can be concluded that the two ADHs from the distantly related species D. melanogaster and D. lebanonensis follow the same reaction mechanism over the pH region 6^10, and that they di¡er only in the magnitude of various rate constants and kinetic coe¤cients. These di¡erences result not only in a di¡erent magnitude of the primary kinetic isotope e¡ect with ethanol, but also in a di¡erent inhibitory pattern of imidazole vs. ethanol as a varied substrate. 4.2. Dissociation and equilibrium constants
pH
P1 P2 /P12 PP0 (1)
PP1 PP2 /PP12 P0 (2)
PP1 PP2 /PP12 P0 (3)
10.0 9.0 8.0 7.0 6.0
0.4 0.6 1.1 1.1 1.3
0.5 0.4 0.2 0.2 0.2
1.3 1.4 1.0 0.6 0.8
The values for the ratio of the maximum speci¢c rate (1/P0 and 1/PP0 ) to the rate of dissociation of the enzyme-coenzyme product complex (PP12 /PP1 PP2 and P12 /P1 P2 ) calculated for a compulsory ordered mechanism. The primed kinetic coe¤cients are for the acetaldehyde/NADH reaction and the unprimed coe¤cients in relations (1) and (2) are from the ethanol/NAD reaction. In relation (3), the P0 value is for propan-2-ol oxidation, and as this value did not vary with pH it was set at 0.145 s.
4.2.1. KE;O and KE;R As previously shown for the ADHS alleloenzyme from D. melanogaster [12], the proposed reaction mechanism indicates that there should be relations in addition to P12 /P2 and PP12 /PP2 describing the dissociation constants KE;O and KE;R for the enzymecoenzyme complexes. As 1/P1 for both ethanol and propan-2-ol is equal to k1 and 1/PP0 for acetaldehyde is equal to k31 , P1 /PP0 should be equal to KE;O . For the acetaldehyde reaction 1/PP1 = kP1 and for propan2-ol 1/P0 = kP31 and hence PP1 /P0 equals KE;R . Table 6 shows that these additional relations for KE;O and
BBAPRO 35865 30-3-99
82
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
Table 6 Comparison of the kinetic coe¤cients giving the enzyme-coenzyme dissociation constants at 23.5³C pH
10.0 9.5 9.0 8.0 7.0 6.0
KE;O (WM)
KE;R (WM)
P12 /P2 Ethanol
P1 /PP0 Ethanol/acetaldehyde
PP12 /PP2 Acetaldehyde
PP1 /P0 Acetaldehyde/propan-2-ol
17 23 70 156 551 913
6.4 N.D. 39 164 585 1225
6.5 N.D. 2.6 2.7 2.4 1.3
8.4 N.D. 3.7 2.5 1.4 1.0
The enzyme-coenzyme dissociation constants KE;O and KE;R are based on an ordered reaction mechanism in both directions (Scheme 1) with rate-limiting enzyme-coenzyme dissociation for propan-2-ol oxidation and acetaldehyde reduction. The P0 value for propan-2ol oxidation did not vary with pH and was therefore set at 0.145 s. The KE;O (WM) for propan-2-ol at pH 9.5 and 6.0 was as follows: P12 /P2 : 17, 732; P1 /PP0 : N.D., 1100. N.D., not determined.
KE;R are consistent with the P12 /P2 and PP12 /PP2 values in the pH region 6^10.0. The KE;O values calculated from propan-2-ol oxidation correspond with those calculated from ethanol oxidation (Table 6). 4.2.2. Keq The Haldane relation PP12 [H ]/P12 is the overall equilibrium constant Keq for the reversible reactions studied [27]: K eq
Alcohol NAD aldehyde=ketone NADH H where Keq = [NADH][carbonyl compound][H ]/ [NAD ][alcohol] and is independent of the enzyme catalyzing the reaction. The average value of 18 pM obtained for the overall equilibrium constant for the Table 7 pH dependence of the Haldane relations for the overall reaction at 23.5³C pH
PP12 /P12
PP12 [H ]/P12 (pM)
PP0 PP1 PP2 [H ]/P0 P1 P2 (pM)
10.0 9.0 8.0 7.0 6.0
4.3U1032 6.1U1033 2.2U1033 3.9U1034 2.2U1035
4.3 6.1 22 39 22
14 17 19 20 12
In column 4 the primed kinetic coe¤cients are for the acetaldehyde/NADH reaction, P0 is for the propan-2-ol/NAD reaction, where P0 equals kP31 which does not vary with pH and is therefore set at 0.145 s for all pH values. The kinetic coe¤cients P1 and P2 are for the ethanol/NAD reaction.
ethanol/acetaldehyde reaction (Table 7) corresponds well to the values previously obtained for D. melanogaster ADHS [12], horse liver and yeast ADH [31^ 35]. If the reaction mechanism for the oxidation of ethanol and propan-2-ol and the reduction of acetaldehyde is correct, Keq can also be expressed as PP0 PP1 PP2 [H ]/P0 P1 P2 (Table 7). The primed kinetic coe¤cients are for the acetaldehyde reduction. P0 is 1/kP31 for the propan-2-ol reaction and P1 (1/k1 ) and P2 are for the ethanol reaction. An average Keq of 16 pM was obtained over the pH range 6^10. This ¢ts well with the average PP12 [H ]/P12 value of 18 pM. The consistency of the di¡erent equilibrium and dissociation constants from the di¡erent relations of the kinetic coe¤cients supports the proposed reaction mechanism for the oxidation of alcohols and the reduction of aldehydes. 4.3. Variation of rate and equilibrium constants with pH 4.3.1. kon : coenzymes There are no titratable groups of the two coenzymes in the region pH 6.0^10.0. Therefore, the variation of the kon velocities with pH for the two coenzymes re£ects titratable groups in the free enzyme. It is noticeable that the kon velocity curves for the two coenzymes are the opposite of each other, i.e. 1/P1 (k1 ) or kon for NAD increases and 1/PP1 (kP1 ) or kon for NADH decreases with increasing pH (Fig. 3a). Both curves appear complex and are likely to involve
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
several ionizing groups. At least one residue with a pKa value in the region 9^10 regulates binding of NAD to the free enzyme. In the case of the curve for kP1 , the protonation of at least two groups with pKa values in the region 6^8 and 9^10 regulates NADH binding to the free enzyme. These results show that the groups in the enzyme that ionize in the pH region studied, and regulate the kon velocity for the two coenzymes must be located in close proximity or in the region that binds the nicotinamide part of the coenzymes, as this is the only part of the coenzymes that di¡er. The form of the two kon velocity curves (Fig. 3a) cannot exclude that the same amino acids regulate the kon velocities for the two coenzymes. Previously, similar but not identical curves were obtained with the D. melanogaster ADHS alleloenzyme [12]. Several investigations of di¡erent Drosophila ADHs indicated that residue 191 was involved in coenzyme binding (for review see [3]), and that this group which is a lysine in the ADHS alleloenzyme could be responsible for the observed pH e¡ect on k1 [12]. As this group is a threonine in D. lebanonensis ADH [3,21,22], the similarity of the curves for the two enzymes suggests that the e¡ect cannot be due to residue 191. In the D. melanogaster enzyme, kP1 was estimated to vary with the ionization of a single residue with a pKa around 8.0 [12]. However, a reexamination of this curve (Fig. 4 in [12]) shows that the curve £attens out between pH 7.5 and 8.5 and hence has a similar form to the curve presented for the D. lebanonensis enzyme in the present work. In spite of the 20% di¡erence in amino acid sequence between these two enzymes, the same amino acid residues in the two enzymes may regulate the kon velocities for the coenzymes. The slight di¡erence in the pH pro¢le between the two enzymes might re£ect an altered microenvironment in the vicinity of the nicotinamide binding region. 4.3.2. koff : coenzymes The variation of the koff velocities with pH for the two coenzymes depends on titratable groups in the binary enzyme-coenzyme complexes. The koff velocity (kP31 ) for NADH from the binary ESP1 complex does not vary with pH, which indicates that there are no ionizable groups of catalytic importance within the pH region studied for this binary complex. This corresponds with the observations for D. mela-
83
nogaster ADHS [12], where the koff velocity was approximately half that of the D. lebanonensis enzyme. The koff velocity (k31 ) for NAD from the binary ES1 complex appears to vary with the ionization of at least two groups in the ES1 complex of D. lebanonensis ADH. Visual inspection of the curve in Fig. 3b suggests pKa values around 7 and 8.5. Protonation of the groups enhances the dissociation rate. The dissociation rate of NAD from the binary ES1 complex with D. melanogaster ADHS varied with the ionization of a single group, with a pKa value of 7.6 [12]. As the koff velocity for NAD was pH dependent and for NADH was pH independent, it must be assumed that the amino acids in question are located in the vicinity of the nicotinamide group of the coenzymes. In the ESP1 complex these amino acids must have pKa values above 10, and the positive charge of the nicotinamide group in NAD causes a large perturbation of the pKa value(s). What is the nature of these amino acids? Previously with the D. melanogaster enzyme, we suggested that the group was the same as that responsible for alcohol binding. This appears a likely explanation as the discussion on alcohol and pyrazole binding indicates and where the corresponding amino acid in D. lebanonensis ADH has a pKa value of 7.1. However, with D. lebanonensis there appears to be an additional amino acid with a pKa value within the pH region studied that a¡ects the dissociation rate of NAD from the enzyme. 4.3.3. Haldane relation PP2 [H +]/P2 In a compulsory ordered mechanism, the Haldane relations are PP2 /P2 = k2 kkP32 /kP2 kPk32 and PP2 [H ]/ P2 = [ER][aldehyde][H ]/[EO][alcohol] for the bound coenzymes and substrates [31]. This is analogous to the overall Haldane relation PP12 [H ]/P12 = Keq for the coenzymes and substrates. The relation PP2 [H ]/P2 can also be expressed as Keq KE;O /KE;R , and the variation of log(PP2 [H ]/P2 ) with pH re£ects the di¡erent e¡ects of pH variation on groups in the binary ES1 and ESP1 complexes. Between pH 7 and 10, the plot of log(PP2 [H ]/P2 ) versus pH showed a slope of approx. 31 (Fig. 3d). This indicates that at least one coenzyme-linked acid group in the enzyme has a lower pKa value in the enzyme complex with NAD (ES1 ) than with NADH (ESP1 ).
BBAPRO 35865 30-3-99
84
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
4.3.4. KEO;I , P2 and PP2 The kinetic coe¤cient PP2 for the reduction of acetaldehyde did not vary within the pH 6.0^10.0 region studied (Table 3). The PP2 coe¤cient is related to the kinetic rate constants in Scheme 1 as follows: P0 2
1 k0 32 k 1 1 k0 2 k32 k0
5
As PP2 does not vary with pH, kP2 and the two quotients kP32 /kP and k/k32 must be pH independent. In the oxidation of propan-2-ol, 1/P0 equals kP31 . This coe¤cient did not vary with pH (Table 2). In the oxidation of ethanol, P0 equals 1/k+1/kP31 +1/ kP32 +kP/kkP32 , and the main rate limiting steps could be attributed to 1/k+1/kP31 , i.e. the hydride transfer and the ESP1 dissociation steps. As P0 did not vary with pH, it can be concluded that also k is pH independent (Table 1). This indicates that this is also the case with k32 , i.e. the dissociation of ethanol from the ternary ES1 S2 complex. Thus, there are no ionizable groups of catalytic importance within the pH region 6.0^10.0 in the ES1 S2 , and the ESP1 complexes. The lack of pH variation for PP2 also shows that the two rate constants kP32 and kP which re£ect the ionization properties of the ESP1 SP2 complex either must show the same pH dependence or be pH independent. The present work with D. lebanonensis ADH shows that the variation of 1/P2 against pH for ethanol oxidation depends on the ionization of a single residue. This was also the case for the ADHS alleloenzyme from D. melanogaster [12]. Eq. 6 shows the relation of P2 to the kinetic rate constants in Scheme 1: P2
1 k32 k0 1 1 0 k2 k k 32
6
The pKa value obtained for the D. lebanonensis enzyme was 7.1. As discussed for the pH independence of PP2 , the absence of [H ] proportionality of the two quotients k32 /k and kP/kP32 indicates that the pH dependence of the alcohol binding to the binary enzyme-NAD complex is derived exclusively from the on-velocity constant k2 . For 1/KEO;I , the pH pro¢le for the alcohol competitive inhibitor pyrazole was almost identical to that for 1/P2 , and gave a pKa value of 7.1. Therefore, there is no drastic stickiness of the substrate and the pKa value obtained from the
1/P2 pro¢le re£ects the true pKa of the amino acid in the ES1 complex. Previously with the ADHS enzyme from D. melanogaster, we showed that analogously to the zinc containing ADHs, the role of Drosophila ADH is to decrease the pKa for the enzyme bound alcohol substrates. Therefore, an alcohol at physiological pH will be present predominantly as alcoholate ion in the productive ternary complex [13]. The negative charge on the substrate can be expected to drastically facilitate subsequent hydride transfer. The recently published three-dimensional structures of ¢ve shortchain dehydrogenases [14^18] indicated that the hydroxyl group of the substrate pointed between the two highly conserved residues Ser138 and Tyr151 (D. lebanonensis nomenclature). The present study does not determine which of these two residues is responsible for the variation of P2 and KEO;I with pH and hence acts as a base catalyst of the hydride transfer step, by increasing the negative charge on the substrate at the ternary complex level. However, future studies on this subject through either inactivation kinetics with speci¢c inactivation reagents against the two conserved amino acids and/or the determination of the temperature dependence of the pKa value of the catalytic residue should help to solve this question. The present study on D. lebanonensis ADH and our previous pH study on the alleloenzyme ADHS from D. melanogaster [12] facilitate such studies, as kinetic coe¤cients at di¡erent pH values now are known for two di¡erent Drosophila ADHs. Previous works on the substrate speci¢city of different Drosophila ADHs have focused on the similarities between these enzymes [3,36]. These studies indicated that in spite of considerable di¡erences in the amino acid sequence of the ADHs from the distantly related species D. melanogaster and D. lebanonensis, their active site topologies must have been largely conserved during the evolution. The present work on pH dependence of the kinetic coe¤cients also shows that the properties of the active site and hence kinetics have been largely conserved in these enzymes. However, it also points to striking di¡erences between the ADHs from these two distantly related species, some of which have been discussed above. It is also noticeable that the active site residue in the D. lebanonensis enzyme has a pKa which is 0.5 pH units
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
below the corresponding value in the D. melanogaster ADHS alleloenzyme. This indicates di¡erences in the electrostatic properties of the microenvironment around the catalytic site in these two Drosophila ADHs. These di¡erences may be the cause of the changed magnitude of the Ki constants for pyrazole binding to the enzyme-NAD complexes in these two enzymes, where the inhibitor binds 1.8^2.8 times stronger to the D. melanogaster enzyme over the entire pH region. The di¡erence in microenvironment may also explain why the D. lebanonensis ADH oxidizes some larger alcohols at a faster rate than the D. melanogaster enzyme [3,25].
[9]
[10]
[11]
[12]
Acknowledgements [13]
We are grateful to Dr. Elvira Juan, Department of Genetics, University of Barcelona, for a generous gift of D. lebanonensis ADH and also providing us with D. lebanonensis £ies.
[14]
[15]
References [1] G.K. Chambers, The Drosophila-alcohol dehydrogenase gene-enzyme system, Adv. Genet. 25 (1988) 39^107. [2] G.K. Chambers, Gene expression, adaption and evolution in higher organisms. Evidence from studies of Drosophila alcohol dehydrogenases, Comp. Biochem. Physiol. 99B (1991) 723^730. [3] J.O. Winberg, J.S. McKinley-McKee, Kinetic interpretation of active site topologies and residue exchanges in Drosophila alcohol dehydrogenases, Int. J. Biochem. 24 (1992) 169^181. [4] H. Jo«rnvall, B. Persson, J. Je¡ery, Alcohol and polyol dehydrogenases are both divided into two protein types. Structural properties cross-relate the di¡erent enzyme activities within each type, Proc. Natl. Acad. Sci. USA 78 (1981) 4226^4230. [5] H. Jo«rnvall, B. Persson, M. Krook, R. Kaiser, Alcohol dehydrogenases, Biochem. Soc. Trans. 18 (1990) 169^171. [6] H. Jo«rnvall, B. Persson, M. Krook, S. Atrian, R. Gonza©lezDuarte, J. Je¡rey, D. Ghosh, Short-chain dehydrogenases/ reductases (SDR), Biochemistry 34 (1995) 6003^6013. [7] G. Labesse, A. Vidal-Cros, J. Chomilier, M. Gaudry, J.-P. Mornon, Structural comparisons lead to the de¢nition of a new superfamily of NAD(P)(H)-accepting oxidoreductases: the single-domain reductases/epimerases/dehydrogenases (the `RED' family), Biochem. J. 304 (1994) 95^99. [8] Z. Chen, L. Lu, M. Shirley, W.R. Lee, S.H. Chang, Sitedirected mutagenesis of glycine-14 and two `critical' cysteinyl
[16]
[17]
85
residues in Drosophila alcohol dehydrogenase, Biochemistry 29 (1990) 1112^1118. Z. Chen, J.C. Jiang, Z.-G. Lin, W.R. Lee, M.E. Baker, S.H. Chang, Site-speci¢c mutagenesis of Drosophila alcohol dehydrogenase: evidence for involvement of tyrosine-152 and lysine-156 in catalysis, Biochemistry 32 (1993) 3342^ 3346. N. Cols, G. Marfany, S. Atrian, R. Gonza©lez-Duarte, E¡ect of site-directed mutagenesis on conserved positions of Drosophila alcohol dehydrogenase, FEBS Lett. 319 (1993) 90^ 94. N. Cols, S. Atrian, J. Benach, R. Ladenstein, R. Gonza©lezDuarte, Drosophila alcohol dehydrogenase: evaluation of Ser139 site-directed mutagenesis, FEBS Lett. 413 (1997) 191^193. J.O. Winberg, J.S. McKinley-McKee, The ADHS alleloenzyme of alcohol dehydrogenase from Drosophila melanogaster: variation of kinetic parameters with pH, Biochem. J. 255 (1988) 589^599. J.S. McKinley-McKee, J.O. Winberg, G. Pettersson, Mechanism of action of Drosophila melanogaster alcohol dehydrogenase, Biochem. Int. 25 (1991) 879^885. D. Ghosh, C.M. Weeks, P. Grochulski, W.L. Duax, M. Erman, R.L. Rimsay, J.C. Orr, Three-dimensional structure of holo 3K,20L-hydroxysteroid dehydrogenase: a member of a short-chain dehydrogenase family, Proc. Natl. Acad. Sci. USA 88 (1991) 10064^10068. D. Ghosh, Z. Wawrzak, C.M. Weeks, W.L. Duax, M. Erman, The re¢ned three-dimensional structure of 3K,20L-hydroxysteroid dehydrogenase and possible roles of the residues conserved in short-chain dehydrogenases, Structure 2 (1994) 629^640. K.I. Varughese, M.M. Skinner, J.M. Whiley, D.A. Matthews, N.H. Xuong, Crystal structure of rat liver dihydropteridine reductase, Proc. Natl. Acad. Sci. USA 89 (1992) 6080^6084. D. Ghosh, V.Z. Pletnev, D.W. Zhu, Z. Wawrzak, W.L. Duax, W. Pangborn, F. Labrie, S.-X. Lin, Structure of huî man estrogenic 17L-hydroxysteroid dehydrogenase at 2.20 A
resolution, Structure 3 (1995) 503^513. [18] N. Tanaka, T. Nonaka, M. Nakanishi, Y. Deyaskiki, A. Hara, Y. Mitsui, Crystal structure of the ternary complex î resolution: the of mouse lung carbonyl reductase at 1.8 A structural origin of coenzyme speci¢city in the short-chain dehydrogenase/reductase family, Structure 4 (1996) 33^45. [19] J. Benach, S. Atrian, R. Gonza©lez-Duarte, R. Ladenstein, The re¢ned crystal structure of Drosophila lebanonensis alcoî resolution, J. Mol. Biol. 282 hol dehydrogenase at 1.9 A (1998) 383^399. [20] J.O. Winberg, J.S. McKinley-McKee, Drosophila melanogaster alcohol dehydrogenase: product inhibition studies, Biochem. J. 301 (1994) 901^909. [21] R. Albalat, R. Gonza©lez-Duarte, Nucleotide sequence of the Adh gene of Drosophila lebanonensis, Nucleic Acids Res. 18 (1990) 6706. [22] E. Juan, M. Papaceit, A. Quintana, Nucleotide sequence of
BBAPRO 35865 30-3-99
86
[23]
[24]
[25]
[26]
[27]
[28]
[29]
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86 the Adh gene of Drosophila lebanonensis, Nucleic Acids Res. 18 (1990) 6420. A. Villarroya, E. Juan, B. Egestad, H. Jo«rnvall, The primary structure of alcohol dehydrogenase from Drosophila lebanonensis: extensive variation within insect `short-chain' alcohol dehydrogenase lacking zinc, Eur. J. Biochem. 180 (1989) 191^197. J. Fibla, L. Enjuanes, R. Gonza©lez-Duarte, Inter speci¢c of Drosophila alcohol dehydrogenase by an immunoenzymatic assay using monoclonal antibodies, Biochem. Biophys. Res. Commun. 160 (1989) 638^647. J.O. Winberg, R. Hovik, J.S. McKinley-McKee, E. Juan, R. Gonza©lez-Duarte, Biochemical properties of alcohol dehydrogenase from Drosophila lebanonensis, Biochem. J. 235 (1986) 481^490. J.O. Winberg, J.S. McKinley-McKee, Drosophila alcohol dehydrogenase: mechanism of aldehyde oxidation and dismutation, Biochem. J. 329 (1998) 561^570. K. Dalziel, Initial steady state velocities in the evaluation on enzyme-coenzyme-substrate reactions, Acta Chem. Scand. 11 (1957) 1706^1723. W.W. Cleland, Determining the chemical mechanisms of enzyme-catalyzed reactions by kinetic studies, Adv. Enzymol. Relat. Areas Mol. Biol. 45 (1977) 273^387. J.O. Winberg, R. Hovik, J.S. McKinley-McKee, The alcohol dehydrogenase alleloenzymes ADHS and ADHF from the fruit£y Drosophila melanogaster: an enzymatic rate assay to determine the active-site concentration, Biochem. Genet. 23 (1985) 205^216.
[30] R. Hovik, J.-O. Winberg, J.S. McKinley-McKee, Drosophila melanogaster alcohol dehydrogenase: substrate speci¢city of the AdhF alleloenzyme, Insect Biochem. 14 (1984) 345^351. [31] K. Dalziel, Kinetic studies of liver alcohol dehydrogenase and pH e¡ects with coenzyme preparations of high purity, J. Biol. Chem. 238 (1963) 2850^2858. [32] K. Dalziel, Kinetic studies of liver alcohol dehydrogenases, Biochem. J. 84 (1962) 244^254. [33] K. Ba«cklin, The equilibrium constants of the system ethanol, aldehyde, DPN , DPNH and H , Acta Chem. Scand. 12 (1958) 1279^1285. [34] H. Theorell, A.P. Nygaard, R. Bonnichsen, Studies on liver alcohol dehydrogenase. III. The in£uence of pH and some anions on the reaction velocity constants, Acta Chem. Scand. 9 (1955) 1148^1165. [35] C.J. Dickenson, F.M. Dickinson, A study of the pH- and temperature dependence of the reactions of yeast alcohol dehydrogenase with ethanol, acetaldehyde and butyraldehyde as substrates, Biochem. J. 147 (1975) 303^311. [36] K.Th. Eisses, S.L. Davies, G.K. Chambers, Substrate and inhibitor speci¢cities of the thermostable alcohol dehydrogenase alleloenzymes ADH71k and ADHFCh:D . of Drosophila melanogaster, Biochem. Genet. 32 (1994) 91^101. [37] D.D. Perrin, Dissociation Constants of Organic Bases in Aqueous Solution (Pure Appl. Chem. Suppl.), Butterworth, London, 1972.
BBAPRO 35865 30-3-99
Drosophila lebanonensis alcohol dehydrogenase: pH dependence of the kinetic coe¤cients Michelle Kaaber Brendskag a , John S. McKinley-McKee b , Jan-Olof Winberg a
a;
*
Department of Biochemistry, Institute of Medical Biology, University of TromsÖ, 9037 TromsÖ, Norway b Biochemical Institute, University of Oslo, Oslo, Norway Received 22 October 1998; received in revised form 14 January 1999; accepted 28 January 1999
Abstract The alcohol dehydrogenase (ADH) from Drosophila lebanonensis shows 82% positional identity to the alcohol dehydrogenases from Drosophila melanogaster. These insect ADHs belong to the short-chain dehydrogenase/reductase family which lack metal ions in their active site. In this family, it appears that the function of zinc in medium chain dehydrogenases has been replaced by three amino acids, Ser138 , Tyr151 and Lys155 . The present work on D. lebanonensis ADH has been performed in order to obtain information about reaction mechanism, and possible differences in topology and electrostatic properties in the vicinity of the catalytic residues in ADHs from various species of Drosophila. Thus the pH dependence of various kinetic coefficients has been studied. Both in the oxidation of alcohols and in the reduction of aldehydes, the reaction mechanism of D. lebanonensis ADH in the pH 6^10 region was consistent with a compulsory ordered pathway, with the coenzymes as the outer substrates. Over the entire pH region, the rate limiting step for the oxidation of secondary alcohols such as propan-2-ol was the release of the coenzyme product from the enzyme-NADH complex. In the oxidation of ethanol at least two steps were rate limiting, the hydride transfer step and the dissociation of NADH from the binary enzyme-NADH product complex. In the reduction of acetaldehyde, the rate limiting step was the dissociation of NAD from the binary enzyme-NAD product complex. The pH dependences of the kon velocity curves for the two coenzymes were the opposite of each other, i.e. kon increased for NAD and decreased for NADH with increasing pH. The two curves appeared complex and the kon velocity for the two coenzymes seemed to be regulated by several groups in the free enzyme. The kon velocity for ethanol and the ethanol competitive inhibitor pyrazole increased with pH and was regulated through the ionization of a single group in the binary enzyme-NAD complex, with a pKa value of 7.1. The kon velocity for acetaldehyde was pH independent and showed that in the enzyme-NADH complex, the pKa value of the catalytic residue must be above 10. The koff velocity of NAD appeared to be partly regulated by the catalytic residue, and protonation resulted in an increased dissociation rate. The koff velocity for NADH and the hydride transfer step was pH independent. In D. lebanonensis ADH, the pKa value of the catalytic residue was 0.5 pH units lower than in the ADHS alleloenzyme from D. melanogaster. Thus it can be concluded that while most of the topology of the active site is mainly conserved in these two distantly related enzymes, the microenvironment and electrostatic properties around the catalytic residues differ. ß 1999 Elsevier Science B.V. All rights reserved.
Abbreviations: ADH, alcohol dehydrogenase (EC 1.1.1.1); ADHS , slow alleloenzyme from Drosophila melanogaster; E, enzyme; EO or ES1 , enzyme-NAD ; EOI, enzyme-NAD -pyrazole; ER or ESP1 , enzyme-NADH ; ES1 S2 , enzyme-NAD -alcohol; ESP1 SP2 , enzymeNADH-aldehyde ; I, inhibitor; S1 or O, NAD ; SP1 or R, NADH; S2 , alcohol; SP2 , aldehyde * Corresponding author. Fax: +47 (77) 645350; E-mail: [email protected] 0167-4838 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 9 ) 0 0 0 2 8 - X
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
75
Keywords: Drosophila; Alcohol dehydrogenase; Reaction mechanism; pH study; Catalytic site; Inhibition
1. Introduction The gene for alcohol dehydrogenase (Adh) has been cloned and sequenced from various Drosophila species, and the enzyme has been puri¢ed and biochemically characterized from some of these species (for review, see [1^3]). The ADH enzyme catalyzes the interconversion of alcohols and their corresponding aldehyde/ketone products (Eq. 1). Alcohol NAD Haldehyde=ketone NADH H 1 Drosophila ADH is a dimer of Mr 54 800, consisting of two identical subunits [1^3], and belongs to the short-chain dehydrogenase/reductase (SDR) family [4,5] which lack metal ions in their active site. In this family, Ser138 , Tyr151 and Lys155 in the amino acid sequence are conserved (Drosophila lebanonensis nomenclature) [3,6,7], and site-directed mutagenesis showed that all three residues in contrast to the two cysteine residues in the enzyme were essential for activity [8^11]. Binding of alcohol and alcohol competitive inhibitors to the Drosophila melanogaster ADHS alleloenzyme was dependent on a single residue in the active site, with a pKa value of about 7.6 in the binary enzyme-NAD (ES1 ) complex, and above 10 in the binary enzyme-NADH (ESP1 ) complex [12]. Previously we showed that certain steps in the mechanism of alcohol oxidation for Drosophila ADH are di¡erent from that known to apply for zinc mediated liver ADH catalysis [13]. To facilitate hydride transfer from the alcohol, the group with a pKa of 7.6 in the D. melanogaster ADHS alleloenzyme must be a residue with a full negative charge, as its function is to produce an alcoholate anion [13]. This indicated that Tyr151 could be the residue which interacts with the hydroxyl group in the alcohol and hence takes over the function of zinc in horse liver ADH [12,13]. X-Ray crystallography of ¢ve members of the SDR family [14^19] showed that Tyr151 and Lys155 reside in K-helix F, suggesting that an interaction between Lys155 and Tyr151 decreases the pKa of the tyrosine group. These studies also indicated that the highly conserved Ser138
residue was located in the catalytic region of the enzyme. Based on studies using alternate substrates, deadend and product inhibitors [3,12,20], it has been shown that Drosophila ADHs follow a compulsory ordered mechanism in the interconversion of alcohols and their corresponding aldehyde and ketone products (Scheme 1). The ADH from the distantly related species D. lebanonensis has only 82% positional identity to the ADH alleloenzymes from D. melanogaster [3,21^23]. Both these Drosophila species are well adapted to environments with a high alcohol content due to a large amount of ADH compared to other Drosophila species [24]. However, D. lebanonensis £ies also contains approximately ¢ve times as much ADH as D. melanogaster £ies [24]. Due to the large amount of ADH in D. lebanonensis £ies, future studies concerning general properties of the insect ADHs will no doubt be performed on the enzyme in this Drosophila species. The substrate specificity of D. lebanonensis ADH was similar, but not identical to that of the other Drosophila ADHs studied [3,25]. These studies focused on the similarities of the di¡erent Drosophila ADHs and stressed that the major topology of the active site region was conserved. However, it is also important to design experiments to determine what distinguishes these enzymes, and this has been done in the present work. Minor di¡erences in the topology and electrostatic properties of the alcohol binding region might result not only in changes in the pKa value of the amino acid that binds the hydroxyl group of an alcohol and the polar group of alcohol competitive inhibitors, but also in changes in the individual rate constants that build up the kinetic coe¤cients. This may result in altered parameters in£uencing for example inhibitory patterns and/or the magnitude of primary and secondary isotope e¡ects. It was also of interest to investigate whether D. lebanonensis ADH like the D. melanogaster ADHS alleloenzyme followed a strict compulsory ordered mechanism over the entire pH region studied [12], as a change in reaction mechanism will in£uence the interpretation of the pH dependence of the di¡erent kinetic coe¤cients.
BBAPRO 35865 30-3-99
76
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
2. Experimental 2.1. Reagents Grade III NAD , NADH, and [2 H8 ]propan-2-ol were from Sigma. Ethanol (96%) was from A/S Vinmonopolet and [2 H6 ]ethanol was from Norsk Hydro. Anhydrous acetaldehyde (puriss p.a.), and pyrazole were from Fluka. Propan-2-ol (99.7%) was from Merck. 2.2. Enzyme D. lebanonensis ADH was puri¢ed as described previously [25]. Freeze-dried samples were dissolved in 0.1 M phosphate bu¡er pH 7.0, and dialyzed against two changes of the same bu¡er at 4³C. Any denatured protein was removed by centrifugation for 20 min at 25 000Ug. 2.3. Rate assay To determine the amount of enzyme, i.e. the enzyme active site concentration in the assay cuvette, the spectrophotometric rate assay for D. lebanonensis ADH previously described was used [25]. The enzyme concentration is expressed as the amount of subunits in nM. This is twice the amount of enzyme molecules as the enzyme is a dimer. The assay solution consisted of 0.5 mM NAD and 100 mM ethanol in a total volume of 1 ml 0.1 M glycine-NaOH bu¡er pH 9.5. In this assay, 1 WM of enzyme active sites gave a rate of 0.82 vA340nm /min. 2.4. Kinetic measurements The oxidation of alcohols and the reduction of acetaldehyde was studied by steady state kinetics at 23.5³C, and the initial rates were measured by following the appearance of NADH £uorescence, using a sensitive ¢lter £uorometer as described previously by Winberg and McKinley-McKee [26]. To normalize the results due to any variations in the light intensity of the xenon lamp, the de£ection of a £uorescent Perspex standard (equivalent to a NADH concentration of about 7 WM) was measured at intervals during initial rate measurements. Since
NADH £uorescence is quenched by increasing NADH concentrations, a calibration curve of the £uorescence against NADH concentration was determined at 23.5³C, with slit widths of 2 mm for both exciting and emitted light as described previously [26]. All initial rate determinations were within the linear part of the standard curve. Bu¡ers used were 0.1 M pyrophosphate (pH 9^10), or 0.1 M phosphate (pH 6^8). The reported aldehyde concentrations represent the sum of the free and hydrated species present. In all experiments, the Lineweaver-Burk plots obtained were linear within the experimental error. The data ¢tted Eq. 2: e=v0 P0 P1 =S1 P2 =S2 P12 =S1 S2
2
where e is the concentration of enzyme active sites and S1 and S2 are coenzyme and substrate respectively. The unprimed symbols (P, S) are used for the NAD /alcohol reactions and the primed symbols (PP, SP) are used for the NADH/aldehyde reactions. The kinetic coe¤cients in Eq. 2 were obtained from primary and secondary plots [27]. Vmax /e or kcat is 1/P0 , the catalytic center activity (molecules of product produced per second per enzyme subunit active site). Km for coenzyme and substrate is P1 /P0 and P2 /P0 , while kcat /Km for coenzyme and substrate is 1/P1 and 1/P2 respectively. The concentrations of coenzyme and substrate used were within the following ranges: NAD 5^2000 WM, ethanol 1^500 mM, propan-2-ol 0.1^40 mM, [2 H8 ]propan-2-ol 0.75^ 80 mM, [2 H6 ]ethanol 20^100 mM for the alcohol/ NAD reactions and NADH 1^10 WM, acetaldehyde 0.06^2 mM, for the aldehyde/NADH reactions. Inhibition experiments were carried out with either a constant NAD concentration and varied ethanol concentrations, or constant ethanol concentration and varied NAD concentrations. 2.5. pH dependence of kinetic coe¤cients Scheme 1 shows a general compulsory ordered mechanism at a single pH value. This and other mechanisms can be described by the initial rate Eq. 2, where the kinetic coe¤cients are complex functions of the rate constants. Typical features for a compulsory ordered mechanism are that the rate
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
77
Scheme 1.
constants for the formation and dissociation of the forward binary enzyme-coenzyme complex are given by k1 = 1/P1 , k31 = P12 /P1 P2 and its dissociation constant by P12 /P2 , while primed values are for the reverse binary complex. The Haldane relation for Keq is PP12 /P12 . If protonation of an essential amino acid in the enzyme is of kinetic signi¢cance in the pH region studied, two charged forms E and EH of the free enzyme appear. The equilibrium between E and EH is characterized by a proton dissociation constant K1 and the fractional concentration of the two di¡erent enzyme forms is then E = 1/(1+[H ]/K1 ) and EH = 1/(1+K1 /[H ]). If both enzyme species react with the substrate at any pH value, the sum of the products of the relative concentrations of each charge type of the enzyme and the appropriate rate constants should yield the observed rate at that pH value. The experimentally obtained rate constant (k) at any pH value can then be expressed as Eq. 3: k
k k 1 H =K 1 1 K 1 =H
3
where k* and k** are either the association or the dissociation rates of the substrate with the charged enzyme forms E and EH. For a substrate-competitive inhibitor or a substrate which binds to both enzyme species, the pH dependence of the inhibitor constant or enzyme-substrate dissociation constant Ki is then expressed as Eq. 4: K 1 H =K 1 Ki i 1 H =K 2
4
where K1 is the proton dissociation constant of the equilibrium between E and EH, and K2 that of the corresponding equilibrium of the enzyme-inhibitor/ substrate complexes. The relation of the enzyme-inhibitor or enzyme-substrate dissociation constant (Ki *) of the unprotonated enzyme and the corresponding constant (Kj ) of the protonated enzyme is then K1 Kj = K2 Ki * [28].
Fig. 1. Primary (a,d) and secondary (b,c,e,f) plots at pH 10 and 23.5³C: variation of the speci¢c initial rate, e/v, with the reciprocal of the ethanol concentration (a) and NAD concentration (b), for several constant NAD and ethanol concentrations. (a) The NAD concentrations were: a, 0.5 mM; E, 0.02 mM; O, 0.01 mM; P, 0.0067 mM; 7, 0.005 mM. (d) The ethanol concentrations were: a, 100 mM; E, 8 mM; O, 4 mM; P, 2 mM; 7, 1 mM. Variation of the intercepts (b,e) and slopes (c,f) of the primary plots with the reciprocal of the NAD concentration (b,c) and ethanol concentration (e,f). P0 is determined from the intercepts in (b) and (e), P1 from the slope in (b) and intercept in (f), P2 from the intercept in (c) and slope in (e), and P12 from the slopes in (c) and (f).
BBAPRO 35865 30-3-99
78
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
3. Results 3.1. Oxidation of alcohols Initial rate data for the ethanol/NAD reaction at pH 10.0 are shown in the primary plots of Fig. 1a,d. The intercepts and slopes obtained are plotted in Fig. 1b,c,e,f. The slopes and intercepts of these secondary plots yield values for the four kinetic coe¤cients in Eq. 2. Similar experiments were performed with ethanol in the pH range 6.0^10.0. In all cases, both primary and secondary plots were linear within the experimental error over the range of substrate and coenzyme concentrations used. Table 1 lists the initial rate parameters at 23.5³C. The kinetic coe¤cients for the oxidation of the alternative substrates [2 H6 ]ethanol, propan-2-ol and [2 H8 ]propan-2-ol at pH 6.0 and 9.5 at 23.5³C are listed in Table 2. In the case of propan-2-ol, there H H was no kinetic isotope e¡ect on either PD 0 /P0 (kcat / D D H H D kcat ) or P1 /P1 ((kcat /Km ) /(kcat /Km ) for NAD ) at the two pH values tested. At pH 9.5, a kinetic isoH H tope e¡ect of 2.6 was observed on PD 2 /P2 ((kcat /Km ) / D (kcat /Km ) for alcohol), while this parameter was only 1.3 at pH 6.0. At both pH values, a kinetic H isotope e¡ect was also observed on PD 12 /P12 . With H ethanol, a primary kinetic isotope e¡ect (PD 0 /P0 ) of 3.6 appeared at pH 6.0, similar to previous observations at pH 9.5 [25]. No isotope e¡ect was observed H D H for PD 1 /P1 , while the e¡ect was 4.5 on P2 /P2 and 9.3 D H on P12 /P12 .
Fig. 2. Variation with pH of (a) 1/P2 for ethanol, (b) 1/KEO;I for pyrazole. (a) Variation with pH of 1/P2 for ethanol; the theoretical curve represents an acid dissociation constant of pKa 7.1. (b) Variation with pH of 1/KEO;I for pyrazole; the theoretical curve represents an acid dissociation constant of pKa 7.1. In both (a) and (b), the pKa value was determined using the ENZFITTER program (Elsevier-BIOSOFT).
6.0^10.0 are listed in Table 3. In all cases, both primary and secondary plots were linear within the experimental error over the range of substrate and coenzyme concentrations used. 3.3. Inhibition studies Pyrazole was a competitive inhibitor against varied ethanol and a ¢xed concentration of NAD (1 mM) in the pH range 6.0^10.0 at 23.5³C. The Kis and KEO;I values determined are listed in Table 4. Pyrazole showed uncompetitive inhibition against varied NAD and a ¢xed concentration of 100 mM of ethanol at pH 6.0 and 10.0. NADH was a NAD competitive inhibitor at pH 6.0 and 10.0, and the obtained Kis values were 0.9 and 2.5 WM, respectively.
3.2. Reduction of acetaldehyde The initial rate parameters describing the reduction of acetaldehyde by NADH in the pH range
Table 1 pH dependence of the kinetic coe¤cients for the oxidation of ethanol by NAD at 23.5³C pH
P0 (s)
P1 (WMs)
P2 (mMs)
P12 (mM2 s)
P1 /P0 (WM)
P2 /P0 (mM)
P12 /P2 (WM)
10.0 9.5a 9.0 8.0 7.0 6.0
0.41 0.42 0.49 0.40 0.47 0.50
3.0 4.5 5.4 8.2 24 49
2.3 2.5 2.3 2.8 5.0 23
0.038 0.059 0.16 0.43 2.7 21
7.3 11 11 21 51 98
5.6 6.0 4.7 7.0 11 46
17 23 70 156 551 913
The kinetic coe¤cients are those in the reciprocal rate equation (Eq. 2), where S1 and S2 are NAD and ethanol, respectively. a Results from [25].
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
79
Table 2 Kinetic coe¤cients for the oxidation of alternate alcohol substrates with NAD at 23.5³C Substrate
pH
P0 (s)
P1 (WMs)
P2 (mMs)
P12 (mM2 s)
P1 /P0 (WM)
P2 /P0 (mM)
P12 /P2 (WM)
Ethanol [2 H6 ]Ethanol Propan-2-ol [2 H8 ]Propan-2-ol Propan-2-ola [2 H8 ]Propan-2-ol
6.0 6.0 6.0 6.0 9.5 9.5
0.50 1.78 0.15 0.16 0.14 0.13
49 29 46 35 3.8 3.7
23 103 4.1 5.4 0.16 0.41
21 196 3.0 11 0.0027 0.027
98 16 307 219 28 28
46 58 27 34 1.1 3.2
913 1903 732 2037 17 66
The kinetic coe¤cients are those in the reciprocal rate equation (Eq. 2). The bu¡ers used are 0.1 M phosphate bu¡er at pH 6.0 and 0.1 M glycine-NaOH bu¡er at pH 9.5. a Results from [25].
3.4. pH dependence of selected kinetic coe¤cients
4. Discussion
Fig. 2a shows for the oxidation of ethanol the variation of 1/P2 with pH, and a pKa value of 7.1 was determined. Fig. 2b shows the variation of 1/KEO;I with pH for the ethanol competitive inhibitor pyrazole and a pKa value of 7.1 was obtained. The kinetic coe¤cient P0 for the oxidation of ethanol and propan-2-ol did not vary within the pH region studied (Tables 1 and 2), and neither did the PP2 coe¤cient for the reduction of acetaldehyde (Table 3). Fig. 3a shows that 1/P1 (k1 ) increases and 1/PP1 (kP1 ) decreases with increasing pH. The dissociation rate of NAD from the binary enzyme-NAD complex (k31 ) was estimated as the average of 1/PP0 and P12 /P1 P2 (Tables 1^3) and Fig. 3b shows that k31 varies with pH. Fig. 3c shows that KE;O decreases and KE;R increases with increasing pH. A plot of log(PP2 [H ]/P2 ) vs. pH shows a slope of 31 in the pH region 7^10 (Fig. 3d).
4.1. Reaction mechanism Previously it was shown that the D. melanogaster alleloenzyme ADHS followed a compulsory ordered reaction mechanism with the coenzymes as the outer substrates in the pH region 6.0^10.0 (Scheme 1). This conclusion was based on Dalziel relations of the kinetic coe¤cients, alternate substrates, dead-end and product inhibition studies [3,12,20]. The same mechanism was suggested for D. lebanonensis ADH at pH 9.5, using alternate substrates and dead-end inhibitors [25]. Pyrazole, which was an ethanol competitive inhibitor over the entire pH region with D. lebanonensis ADH, showed uncompetitive inhibition against varied NAD at both pH 6 and 10. This indicated that the D. lebanonensis enzyme followed a compulsory ordered mechanism over the entire pH region studied. In the case of a rapid random mechanism, pyrazole should have shown a non-competi-
Table 3 pH dependence of the kinetic coe¤cients for the reduction of acetaldehyde by NADH at 23.5³C pH
PP0 (s)
PP1 (WMs)
PP2 (mMs)
PP12 (mM2 s)
PP1 /PP0 (WM)
PP2 /PP0 (mM)
PP12 /PP2 (WM)
10.0 9.0 8.0 7.5 7.0 6.0
0.47 0.14 0.050 0.045 0.041 0.040
1.22 0.52 0.36 0.27 0.20 0.14
0.25 0.37 0.35 0.54 0.43 0.36
0.00162 0.00097 0.00093 0.00068 0.00105 0.00047
2.6 3.7 7.2 9.0 4.9 3.5
0.53 2.6 7.0 18 11 9.0
6.5 2.6 2.7 1.3 2.4 1.3
The kinetic coe¤cients are those in the reciprocal rate equation (Eq. 2), where SP1 and SP2 are NADH and acetaldehyde, respectively. Bu¡ers used are as described in Section 2.
BBAPRO 35865 30-3-99
80
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
tive pattern against varied NAD . In a compulsory ordered mechanism, the value of the P coe¤cient for the outer substrate should be the same independent of the nature of the inner substrate. Table 2 shows that this was the case for the P1 coe¤cient at both pH 6.0 and 9.5 with di¡erent alcohol substrates. Thus the results obtained with two independent methods were consistent with alcohols being oxidized in a compulsory ordered mechanism over the entire pH region tested. A compulsory ordered reaction mechanism requires that the maximum rate relations P1 P2 /P12 PP0 (1) and PP1 PP2 /PP12 P0 (2) should be less than unity if ternary complexes are kinetically signi¢cant, and equal to unity in the limiting case of a TheorellChance mechanism, i.e. the dissociation of the outer product from the binary enzyme-product complex is rate limiting. For D. lebanonensis ADH, the maximum rate relation (1) can be regarded as unity in the pH region 6^10 (Table 5), with the deviation from unity attributable to the experimental errors in the kinetic coe¤cients determined. Thus in the reduction of acetaldehyde to ethanol, the dissociation of S1 from the ES1 complex (k31 ) is rate limiting between pH 6 and 10. This is similar to the previous observations with the ADHS alleloenzyme from D. melanogaster [12]. Table 4 pH dependence of the inhibitor constants Kis and KEO;I for pyrazole at 23.5³C pH 10.0 9.0 8.0 7.0 6.0
23.5³C Kis (WM)
KEO;I (WM)
12 13 17 34 121
12 12 15 22 63
pH dependence of the competitive inhibition constants Kis and KEO;I for pyrazole against varied ethanol and a ¢xed concentration of NAD (O) of 1.0 mM. The bu¡ers used at the di¡erent pH values are as described in Section 2. The neutral form of pyrazole was calculated from a pKa value of 2.5 [37] and used in the determination of the Kis and KEO;I values. Kis is equal to [I]/((si /s0 )31), where [I] is the inhibitor concentration and si and s0 are respectively the inhibited and uninhibited slopes from the Lineweaver-Burk plots. KEO;I is equal to Kis /((P12 /P2 [O])+1), and represents the dissociation constant of I from the ternary enzyme-NAD -inhibitor complex. P12 and P2 are the kinetic coef¢cients from Table 1.
Fig. 3. Variation with pH of (a) the speci¢c rates for the formation of the binary enzyme-NADH (kP1 ) and enzyme-NAD (k1 ) complexes, (b) the speci¢c dissociation rate (k31 ) of the enzyme-NAD complex, (c) the equilibrium constants of the enzyme-NAD (KE;O ) and enzyme-NADH (KE;R ) complexes, (d) log(PP2 [H ]/P2 ), the equilibrium constant of the reaction of the enzyme-coenzyme complexes with substrates. Variation with pH of (a) the speci¢c rates for the formation of the enzyme-NADH complex (1/PP1 or kP1 : b) and the enzyme-NAD complex (1/P1 or k1 : a). (b) The speci¢c dissociation rate (k31 ) of the enzyme-NAD complex. The k31 constant is the average of 1/PP0 for acetaldehyde reduction and P12 /P1 P2 for alcohol oxidation (Tables 1^3). (c) The dissociation constants of the enzymeNAD (KE;O : b) and enzyme-NADH (KE;R : a) complexes. The dissociation constants are the average values (mean þ S.D.) obtained from Table 6. (d) Log(PP2 [H ]/P2 ), the equilibrium constant of the reaction of the enzyme-coenzyme complexes with their substrates.
In the case of propan-2-ol and other secondary alcohols, it was previously shown that they were oxidized in a Theorell-Chance mechanism with rate limiting dissociation of SP1 from the ESP1 complex at pH 9.5 [3,12,20,25]. The present work shows that this is also the case at pH 6. This is based on the fact that there was no primary isotope e¡ect with propan-2-ol H and [2 H8 ]propan-2-ol, i.e. the ratio PD 0 /P0 is unity (Table 2). In a case where PP1 PP2 /PP12 is from the acetaldehyde reduction and represents 1/kP31 , and
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
1/P0 is for propan-2-ol oxidation and equals kP31 , the maximum rate relation will be unity. This is also the case for D. lebanonensis ADH in the pH region 6.0^ 10.0 as the maximum rate relation (3) shows (Table 5). Thus propan-2-ol is oxidized in a TheorellChance type of mechanism over the entire pH region studied. Table 5 shows that the maximum rate relation (2) is less than unity, which indicates a compulsory ordered mechanism where the rate limiting step in the oxidation of ethanol involves steps prior to NADH release from the binary ESP1 (ER) product complex or a combination of the two product release steps and the hydride transfer step. With Drosophila ADHs, imidazole was previously shown to form an ES1 I (EOI) complex in the ¢rst phase of the reaction resulting in inhibition of reaction. In addition it can form an ESP1 I (ERI) complex in the second phase of the reaction where SP1 dissociates faster from the ternary ESP1 I than from the binary ESP1 , and hence results in a stimulatory reaction [25,26,30]. In a reaction sequence where both these complexes are formed, an inhibitory pattern called inhibition with stimulation will appear [25,26,30]. Imidazole was an ethanol competitive inhibitor [25], which showed that the contribution of kP31 (ESP1 dissociation) to the rate limiting step must be limited. Previous studies with various secondary alcohols at pH 9.5 gave an average value of 7.4 s31 for kP31 [25], which shows that the contribution of the ESP1 dissociation step to the P0 value for ethanol oxidation is approx. 32%. H The primary kinetic isotope e¡ect on PD 0 /P0 for ethaTable 5 Maximum rate relations at 23.5³C
81
nol and [2 H6 ]ethanol (Table 2) indicates that the interconversion of the ternary complexes, i.e. the hydride transfer step (k), largely contributes to the rate limiting step. If it is assumed that kP32 Fk and kP, then P0 equals 1/k+1/kP31 and k is calculated to be 3.4 s31 . Thus, the hydride transfer step contributes 68% of the P0 value. This is similar to the ADHS alleloenzyme from D. melanogaster, where kP31 and k were estimated to be 3.4 s31 (average value of kcat for secondary alcohols) and 2.0 s31 , respectively [12,29]. In the D. melanogaster enzyme, the ESP1 dissociation step and the hydride transfer step contributed 38% and 62%, respectively, to the P0 value for ethanol oxidation. The di¡erent magnitude and hence contribution of k and kP31 to the rate limiting step in the oxidation of ethanol in these two Drosophila ADHPs were seen in the di¡erent magnitudes of H the primary kinetic isotope e¡ects (PD 0 /P0 ); 2.3 in D. S melanogaster ADH [29] and 3.0^3.6 in D. lebanonensis ADH ([25]; present work). In addition, imidazole was a competitive inhibitor with stimulation against varied ethanol with the former enzyme [26], while it was a competitive inhibitor with the latter enzyme [25]. Thus it can be concluded that the two ADHs from the distantly related species D. melanogaster and D. lebanonensis follow the same reaction mechanism over the pH region 6^10, and that they di¡er only in the magnitude of various rate constants and kinetic coe¤cients. These di¡erences result not only in a di¡erent magnitude of the primary kinetic isotope e¡ect with ethanol, but also in a di¡erent inhibitory pattern of imidazole vs. ethanol as a varied substrate. 4.2. Dissociation and equilibrium constants
pH
P1 P2 /P12 PP0 (1)
PP1 PP2 /PP12 P0 (2)
PP1 PP2 /PP12 P0 (3)
10.0 9.0 8.0 7.0 6.0
0.4 0.6 1.1 1.1 1.3
0.5 0.4 0.2 0.2 0.2
1.3 1.4 1.0 0.6 0.8
The values for the ratio of the maximum speci¢c rate (1/P0 and 1/PP0 ) to the rate of dissociation of the enzyme-coenzyme product complex (PP12 /PP1 PP2 and P12 /P1 P2 ) calculated for a compulsory ordered mechanism. The primed kinetic coe¤cients are for the acetaldehyde/NADH reaction and the unprimed coe¤cients in relations (1) and (2) are from the ethanol/NAD reaction. In relation (3), the P0 value is for propan-2-ol oxidation, and as this value did not vary with pH it was set at 0.145 s.
4.2.1. KE;O and KE;R As previously shown for the ADHS alleloenzyme from D. melanogaster [12], the proposed reaction mechanism indicates that there should be relations in addition to P12 /P2 and PP12 /PP2 describing the dissociation constants KE;O and KE;R for the enzymecoenzyme complexes. As 1/P1 for both ethanol and propan-2-ol is equal to k1 and 1/PP0 for acetaldehyde is equal to k31 , P1 /PP0 should be equal to KE;O . For the acetaldehyde reaction 1/PP1 = kP1 and for propan2-ol 1/P0 = kP31 and hence PP1 /P0 equals KE;R . Table 6 shows that these additional relations for KE;O and
BBAPRO 35865 30-3-99
82
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
Table 6 Comparison of the kinetic coe¤cients giving the enzyme-coenzyme dissociation constants at 23.5³C pH
10.0 9.5 9.0 8.0 7.0 6.0
KE;O (WM)
KE;R (WM)
P12 /P2 Ethanol
P1 /PP0 Ethanol/acetaldehyde
PP12 /PP2 Acetaldehyde
PP1 /P0 Acetaldehyde/propan-2-ol
17 23 70 156 551 913
6.4 N.D. 39 164 585 1225
6.5 N.D. 2.6 2.7 2.4 1.3
8.4 N.D. 3.7 2.5 1.4 1.0
The enzyme-coenzyme dissociation constants KE;O and KE;R are based on an ordered reaction mechanism in both directions (Scheme 1) with rate-limiting enzyme-coenzyme dissociation for propan-2-ol oxidation and acetaldehyde reduction. The P0 value for propan-2ol oxidation did not vary with pH and was therefore set at 0.145 s. The KE;O (WM) for propan-2-ol at pH 9.5 and 6.0 was as follows: P12 /P2 : 17, 732; P1 /PP0 : N.D., 1100. N.D., not determined.
KE;R are consistent with the P12 /P2 and PP12 /PP2 values in the pH region 6^10.0. The KE;O values calculated from propan-2-ol oxidation correspond with those calculated from ethanol oxidation (Table 6). 4.2.2. Keq The Haldane relation PP12 [H ]/P12 is the overall equilibrium constant Keq for the reversible reactions studied [27]: K eq
Alcohol NAD aldehyde=ketone NADH H where Keq = [NADH][carbonyl compound][H ]/ [NAD ][alcohol] and is independent of the enzyme catalyzing the reaction. The average value of 18 pM obtained for the overall equilibrium constant for the Table 7 pH dependence of the Haldane relations for the overall reaction at 23.5³C pH
PP12 /P12
PP12 [H ]/P12 (pM)
PP0 PP1 PP2 [H ]/P0 P1 P2 (pM)
10.0 9.0 8.0 7.0 6.0
4.3U1032 6.1U1033 2.2U1033 3.9U1034 2.2U1035
4.3 6.1 22 39 22
14 17 19 20 12
In column 4 the primed kinetic coe¤cients are for the acetaldehyde/NADH reaction, P0 is for the propan-2-ol/NAD reaction, where P0 equals kP31 which does not vary with pH and is therefore set at 0.145 s for all pH values. The kinetic coe¤cients P1 and P2 are for the ethanol/NAD reaction.
ethanol/acetaldehyde reaction (Table 7) corresponds well to the values previously obtained for D. melanogaster ADHS [12], horse liver and yeast ADH [31^ 35]. If the reaction mechanism for the oxidation of ethanol and propan-2-ol and the reduction of acetaldehyde is correct, Keq can also be expressed as PP0 PP1 PP2 [H ]/P0 P1 P2 (Table 7). The primed kinetic coe¤cients are for the acetaldehyde reduction. P0 is 1/kP31 for the propan-2-ol reaction and P1 (1/k1 ) and P2 are for the ethanol reaction. An average Keq of 16 pM was obtained over the pH range 6^10. This ¢ts well with the average PP12 [H ]/P12 value of 18 pM. The consistency of the di¡erent equilibrium and dissociation constants from the di¡erent relations of the kinetic coe¤cients supports the proposed reaction mechanism for the oxidation of alcohols and the reduction of aldehydes. 4.3. Variation of rate and equilibrium constants with pH 4.3.1. kon : coenzymes There are no titratable groups of the two coenzymes in the region pH 6.0^10.0. Therefore, the variation of the kon velocities with pH for the two coenzymes re£ects titratable groups in the free enzyme. It is noticeable that the kon velocity curves for the two coenzymes are the opposite of each other, i.e. 1/P1 (k1 ) or kon for NAD increases and 1/PP1 (kP1 ) or kon for NADH decreases with increasing pH (Fig. 3a). Both curves appear complex and are likely to involve
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
several ionizing groups. At least one residue with a pKa value in the region 9^10 regulates binding of NAD to the free enzyme. In the case of the curve for kP1 , the protonation of at least two groups with pKa values in the region 6^8 and 9^10 regulates NADH binding to the free enzyme. These results show that the groups in the enzyme that ionize in the pH region studied, and regulate the kon velocity for the two coenzymes must be located in close proximity or in the region that binds the nicotinamide part of the coenzymes, as this is the only part of the coenzymes that di¡er. The form of the two kon velocity curves (Fig. 3a) cannot exclude that the same amino acids regulate the kon velocities for the two coenzymes. Previously, similar but not identical curves were obtained with the D. melanogaster ADHS alleloenzyme [12]. Several investigations of di¡erent Drosophila ADHs indicated that residue 191 was involved in coenzyme binding (for review see [3]), and that this group which is a lysine in the ADHS alleloenzyme could be responsible for the observed pH e¡ect on k1 [12]. As this group is a threonine in D. lebanonensis ADH [3,21,22], the similarity of the curves for the two enzymes suggests that the e¡ect cannot be due to residue 191. In the D. melanogaster enzyme, kP1 was estimated to vary with the ionization of a single residue with a pKa around 8.0 [12]. However, a reexamination of this curve (Fig. 4 in [12]) shows that the curve £attens out between pH 7.5 and 8.5 and hence has a similar form to the curve presented for the D. lebanonensis enzyme in the present work. In spite of the 20% di¡erence in amino acid sequence between these two enzymes, the same amino acid residues in the two enzymes may regulate the kon velocities for the coenzymes. The slight di¡erence in the pH pro¢le between the two enzymes might re£ect an altered microenvironment in the vicinity of the nicotinamide binding region. 4.3.2. koff : coenzymes The variation of the koff velocities with pH for the two coenzymes depends on titratable groups in the binary enzyme-coenzyme complexes. The koff velocity (kP31 ) for NADH from the binary ESP1 complex does not vary with pH, which indicates that there are no ionizable groups of catalytic importance within the pH region studied for this binary complex. This corresponds with the observations for D. mela-
83
nogaster ADHS [12], where the koff velocity was approximately half that of the D. lebanonensis enzyme. The koff velocity (k31 ) for NAD from the binary ES1 complex appears to vary with the ionization of at least two groups in the ES1 complex of D. lebanonensis ADH. Visual inspection of the curve in Fig. 3b suggests pKa values around 7 and 8.5. Protonation of the groups enhances the dissociation rate. The dissociation rate of NAD from the binary ES1 complex with D. melanogaster ADHS varied with the ionization of a single group, with a pKa value of 7.6 [12]. As the koff velocity for NAD was pH dependent and for NADH was pH independent, it must be assumed that the amino acids in question are located in the vicinity of the nicotinamide group of the coenzymes. In the ESP1 complex these amino acids must have pKa values above 10, and the positive charge of the nicotinamide group in NAD causes a large perturbation of the pKa value(s). What is the nature of these amino acids? Previously with the D. melanogaster enzyme, we suggested that the group was the same as that responsible for alcohol binding. This appears a likely explanation as the discussion on alcohol and pyrazole binding indicates and where the corresponding amino acid in D. lebanonensis ADH has a pKa value of 7.1. However, with D. lebanonensis there appears to be an additional amino acid with a pKa value within the pH region studied that a¡ects the dissociation rate of NAD from the enzyme. 4.3.3. Haldane relation PP2 [H +]/P2 In a compulsory ordered mechanism, the Haldane relations are PP2 /P2 = k2 kkP32 /kP2 kPk32 and PP2 [H ]/ P2 = [ER][aldehyde][H ]/[EO][alcohol] for the bound coenzymes and substrates [31]. This is analogous to the overall Haldane relation PP12 [H ]/P12 = Keq for the coenzymes and substrates. The relation PP2 [H ]/P2 can also be expressed as Keq KE;O /KE;R , and the variation of log(PP2 [H ]/P2 ) with pH re£ects the di¡erent e¡ects of pH variation on groups in the binary ES1 and ESP1 complexes. Between pH 7 and 10, the plot of log(PP2 [H ]/P2 ) versus pH showed a slope of approx. 31 (Fig. 3d). This indicates that at least one coenzyme-linked acid group in the enzyme has a lower pKa value in the enzyme complex with NAD (ES1 ) than with NADH (ESP1 ).
BBAPRO 35865 30-3-99
84
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
4.3.4. KEO;I , P2 and PP2 The kinetic coe¤cient PP2 for the reduction of acetaldehyde did not vary within the pH 6.0^10.0 region studied (Table 3). The PP2 coe¤cient is related to the kinetic rate constants in Scheme 1 as follows: P0 2
1 k0 32 k 1 1 k0 2 k32 k0
5
As PP2 does not vary with pH, kP2 and the two quotients kP32 /kP and k/k32 must be pH independent. In the oxidation of propan-2-ol, 1/P0 equals kP31 . This coe¤cient did not vary with pH (Table 2). In the oxidation of ethanol, P0 equals 1/k+1/kP31 +1/ kP32 +kP/kkP32 , and the main rate limiting steps could be attributed to 1/k+1/kP31 , i.e. the hydride transfer and the ESP1 dissociation steps. As P0 did not vary with pH, it can be concluded that also k is pH independent (Table 1). This indicates that this is also the case with k32 , i.e. the dissociation of ethanol from the ternary ES1 S2 complex. Thus, there are no ionizable groups of catalytic importance within the pH region 6.0^10.0 in the ES1 S2 , and the ESP1 complexes. The lack of pH variation for PP2 also shows that the two rate constants kP32 and kP which re£ect the ionization properties of the ESP1 SP2 complex either must show the same pH dependence or be pH independent. The present work with D. lebanonensis ADH shows that the variation of 1/P2 against pH for ethanol oxidation depends on the ionization of a single residue. This was also the case for the ADHS alleloenzyme from D. melanogaster [12]. Eq. 6 shows the relation of P2 to the kinetic rate constants in Scheme 1: P2
1 k32 k0 1 1 0 k2 k k 32
6
The pKa value obtained for the D. lebanonensis enzyme was 7.1. As discussed for the pH independence of PP2 , the absence of [H ] proportionality of the two quotients k32 /k and kP/kP32 indicates that the pH dependence of the alcohol binding to the binary enzyme-NAD complex is derived exclusively from the on-velocity constant k2 . For 1/KEO;I , the pH pro¢le for the alcohol competitive inhibitor pyrazole was almost identical to that for 1/P2 , and gave a pKa value of 7.1. Therefore, there is no drastic stickiness of the substrate and the pKa value obtained from the
1/P2 pro¢le re£ects the true pKa of the amino acid in the ES1 complex. Previously with the ADHS enzyme from D. melanogaster, we showed that analogously to the zinc containing ADHs, the role of Drosophila ADH is to decrease the pKa for the enzyme bound alcohol substrates. Therefore, an alcohol at physiological pH will be present predominantly as alcoholate ion in the productive ternary complex [13]. The negative charge on the substrate can be expected to drastically facilitate subsequent hydride transfer. The recently published three-dimensional structures of ¢ve shortchain dehydrogenases [14^18] indicated that the hydroxyl group of the substrate pointed between the two highly conserved residues Ser138 and Tyr151 (D. lebanonensis nomenclature). The present study does not determine which of these two residues is responsible for the variation of P2 and KEO;I with pH and hence acts as a base catalyst of the hydride transfer step, by increasing the negative charge on the substrate at the ternary complex level. However, future studies on this subject through either inactivation kinetics with speci¢c inactivation reagents against the two conserved amino acids and/or the determination of the temperature dependence of the pKa value of the catalytic residue should help to solve this question. The present study on D. lebanonensis ADH and our previous pH study on the alleloenzyme ADHS from D. melanogaster [12] facilitate such studies, as kinetic coe¤cients at di¡erent pH values now are known for two di¡erent Drosophila ADHs. Previous works on the substrate speci¢city of different Drosophila ADHs have focused on the similarities between these enzymes [3,36]. These studies indicated that in spite of considerable di¡erences in the amino acid sequence of the ADHs from the distantly related species D. melanogaster and D. lebanonensis, their active site topologies must have been largely conserved during the evolution. The present work on pH dependence of the kinetic coe¤cients also shows that the properties of the active site and hence kinetics have been largely conserved in these enzymes. However, it also points to striking di¡erences between the ADHs from these two distantly related species, some of which have been discussed above. It is also noticeable that the active site residue in the D. lebanonensis enzyme has a pKa which is 0.5 pH units
BBAPRO 35865 30-3-99
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86
below the corresponding value in the D. melanogaster ADHS alleloenzyme. This indicates di¡erences in the electrostatic properties of the microenvironment around the catalytic site in these two Drosophila ADHs. These di¡erences may be the cause of the changed magnitude of the Ki constants for pyrazole binding to the enzyme-NAD complexes in these two enzymes, where the inhibitor binds 1.8^2.8 times stronger to the D. melanogaster enzyme over the entire pH region. The di¡erence in microenvironment may also explain why the D. lebanonensis ADH oxidizes some larger alcohols at a faster rate than the D. melanogaster enzyme [3,25].
[9]
[10]
[11]
[12]
Acknowledgements [13]
We are grateful to Dr. Elvira Juan, Department of Genetics, University of Barcelona, for a generous gift of D. lebanonensis ADH and also providing us with D. lebanonensis £ies.
[14]
[15]
References [1] G.K. Chambers, The Drosophila-alcohol dehydrogenase gene-enzyme system, Adv. Genet. 25 (1988) 39^107. [2] G.K. Chambers, Gene expression, adaption and evolution in higher organisms. Evidence from studies of Drosophila alcohol dehydrogenases, Comp. Biochem. Physiol. 99B (1991) 723^730. [3] J.O. Winberg, J.S. McKinley-McKee, Kinetic interpretation of active site topologies and residue exchanges in Drosophila alcohol dehydrogenases, Int. J. Biochem. 24 (1992) 169^181. [4] H. Jo«rnvall, B. Persson, J. Je¡ery, Alcohol and polyol dehydrogenases are both divided into two protein types. Structural properties cross-relate the di¡erent enzyme activities within each type, Proc. Natl. Acad. Sci. USA 78 (1981) 4226^4230. [5] H. Jo«rnvall, B. Persson, M. Krook, R. Kaiser, Alcohol dehydrogenases, Biochem. Soc. Trans. 18 (1990) 169^171. [6] H. Jo«rnvall, B. Persson, M. Krook, S. Atrian, R. Gonza©lezDuarte, J. Je¡rey, D. Ghosh, Short-chain dehydrogenases/ reductases (SDR), Biochemistry 34 (1995) 6003^6013. [7] G. Labesse, A. Vidal-Cros, J. Chomilier, M. Gaudry, J.-P. Mornon, Structural comparisons lead to the de¢nition of a new superfamily of NAD(P)(H)-accepting oxidoreductases: the single-domain reductases/epimerases/dehydrogenases (the `RED' family), Biochem. J. 304 (1994) 95^99. [8] Z. Chen, L. Lu, M. Shirley, W.R. Lee, S.H. Chang, Sitedirected mutagenesis of glycine-14 and two `critical' cysteinyl
[16]
[17]
85
residues in Drosophila alcohol dehydrogenase, Biochemistry 29 (1990) 1112^1118. Z. Chen, J.C. Jiang, Z.-G. Lin, W.R. Lee, M.E. Baker, S.H. Chang, Site-speci¢c mutagenesis of Drosophila alcohol dehydrogenase: evidence for involvement of tyrosine-152 and lysine-156 in catalysis, Biochemistry 32 (1993) 3342^ 3346. N. Cols, G. Marfany, S. Atrian, R. Gonza©lez-Duarte, E¡ect of site-directed mutagenesis on conserved positions of Drosophila alcohol dehydrogenase, FEBS Lett. 319 (1993) 90^ 94. N. Cols, S. Atrian, J. Benach, R. Ladenstein, R. Gonza©lezDuarte, Drosophila alcohol dehydrogenase: evaluation of Ser139 site-directed mutagenesis, FEBS Lett. 413 (1997) 191^193. J.O. Winberg, J.S. McKinley-McKee, The ADHS alleloenzyme of alcohol dehydrogenase from Drosophila melanogaster: variation of kinetic parameters with pH, Biochem. J. 255 (1988) 589^599. J.S. McKinley-McKee, J.O. Winberg, G. Pettersson, Mechanism of action of Drosophila melanogaster alcohol dehydrogenase, Biochem. Int. 25 (1991) 879^885. D. Ghosh, C.M. Weeks, P. Grochulski, W.L. Duax, M. Erman, R.L. Rimsay, J.C. Orr, Three-dimensional structure of holo 3K,20L-hydroxysteroid dehydrogenase: a member of a short-chain dehydrogenase family, Proc. Natl. Acad. Sci. USA 88 (1991) 10064^10068. D. Ghosh, Z. Wawrzak, C.M. Weeks, W.L. Duax, M. Erman, The re¢ned three-dimensional structure of 3K,20L-hydroxysteroid dehydrogenase and possible roles of the residues conserved in short-chain dehydrogenases, Structure 2 (1994) 629^640. K.I. Varughese, M.M. Skinner, J.M. Whiley, D.A. Matthews, N.H. Xuong, Crystal structure of rat liver dihydropteridine reductase, Proc. Natl. Acad. Sci. USA 89 (1992) 6080^6084. D. Ghosh, V.Z. Pletnev, D.W. Zhu, Z. Wawrzak, W.L. Duax, W. Pangborn, F. Labrie, S.-X. Lin, Structure of huî man estrogenic 17L-hydroxysteroid dehydrogenase at 2.20 A
resolution, Structure 3 (1995) 503^513. [18] N. Tanaka, T. Nonaka, M. Nakanishi, Y. Deyaskiki, A. Hara, Y. Mitsui, Crystal structure of the ternary complex î resolution: the of mouse lung carbonyl reductase at 1.8 A structural origin of coenzyme speci¢city in the short-chain dehydrogenase/reductase family, Structure 4 (1996) 33^45. [19] J. Benach, S. Atrian, R. Gonza©lez-Duarte, R. Ladenstein, The re¢ned crystal structure of Drosophila lebanonensis alcoî resolution, J. Mol. Biol. 282 hol dehydrogenase at 1.9 A (1998) 383^399. [20] J.O. Winberg, J.S. McKinley-McKee, Drosophila melanogaster alcohol dehydrogenase: product inhibition studies, Biochem. J. 301 (1994) 901^909. [21] R. Albalat, R. Gonza©lez-Duarte, Nucleotide sequence of the Adh gene of Drosophila lebanonensis, Nucleic Acids Res. 18 (1990) 6706. [22] E. Juan, M. Papaceit, A. Quintana, Nucleotide sequence of
BBAPRO 35865 30-3-99
86
[23]
[24]
[25]
[26]
[27]
[28]
[29]
M. Kaaber Brendskag et al. / Biochimica et Biophysica Acta 1431 (1999) 74^86 the Adh gene of Drosophila lebanonensis, Nucleic Acids Res. 18 (1990) 6420. A. Villarroya, E. Juan, B. Egestad, H. Jo«rnvall, The primary structure of alcohol dehydrogenase from Drosophila lebanonensis: extensive variation within insect `short-chain' alcohol dehydrogenase lacking zinc, Eur. J. Biochem. 180 (1989) 191^197. J. Fibla, L. Enjuanes, R. Gonza©lez-Duarte, Inter speci¢c of Drosophila alcohol dehydrogenase by an immunoenzymatic assay using monoclonal antibodies, Biochem. Biophys. Res. Commun. 160 (1989) 638^647. J.O. Winberg, R. Hovik, J.S. McKinley-McKee, E. Juan, R. Gonza©lez-Duarte, Biochemical properties of alcohol dehydrogenase from Drosophila lebanonensis, Biochem. J. 235 (1986) 481^490. J.O. Winberg, J.S. McKinley-McKee, Drosophila alcohol dehydrogenase: mechanism of aldehyde oxidation and dismutation, Biochem. J. 329 (1998) 561^570. K. Dalziel, Initial steady state velocities in the evaluation on enzyme-coenzyme-substrate reactions, Acta Chem. Scand. 11 (1957) 1706^1723. W.W. Cleland, Determining the chemical mechanisms of enzyme-catalyzed reactions by kinetic studies, Adv. Enzymol. Relat. Areas Mol. Biol. 45 (1977) 273^387. J.O. Winberg, R. Hovik, J.S. McKinley-McKee, The alcohol dehydrogenase alleloenzymes ADHS and ADHF from the fruit£y Drosophila melanogaster: an enzymatic rate assay to determine the active-site concentration, Biochem. Genet. 23 (1985) 205^216.
[30] R. Hovik, J.-O. Winberg, J.S. McKinley-McKee, Drosophila melanogaster alcohol dehydrogenase: substrate speci¢city of the AdhF alleloenzyme, Insect Biochem. 14 (1984) 345^351. [31] K. Dalziel, Kinetic studies of liver alcohol dehydrogenase and pH e¡ects with coenzyme preparations of high purity, J. Biol. Chem. 238 (1963) 2850^2858. [32] K. Dalziel, Kinetic studies of liver alcohol dehydrogenases, Biochem. J. 84 (1962) 244^254. [33] K. Ba«cklin, The equilibrium constants of the system ethanol, aldehyde, DPN , DPNH and H , Acta Chem. Scand. 12 (1958) 1279^1285. [34] H. Theorell, A.P. Nygaard, R. Bonnichsen, Studies on liver alcohol dehydrogenase. III. The in£uence of pH and some anions on the reaction velocity constants, Acta Chem. Scand. 9 (1955) 1148^1165. [35] C.J. Dickenson, F.M. Dickinson, A study of the pH- and temperature dependence of the reactions of yeast alcohol dehydrogenase with ethanol, acetaldehyde and butyraldehyde as substrates, Biochem. J. 147 (1975) 303^311. [36] K.Th. Eisses, S.L. Davies, G.K. Chambers, Substrate and inhibitor speci¢cities of the thermostable alcohol dehydrogenase alleloenzymes ADH71k and ADHFCh:D . of Drosophila melanogaster, Biochem. Genet. 32 (1994) 91^101. [37] D.D. Perrin, Dissociation Constants of Organic Bases in Aqueous Solution (Pure Appl. Chem. Suppl.), Butterworth, London, 1972.
BBAPRO 35865 30-3-99