Effect of pH on Pyrazole Binding to Liver Alcohol Dehydrogenase

zyxwvutsrqpo zyxwvutsrqpon Eur. J. Biochem. 114, 549-554 (1981) c; FEBS 1981 Effect of pH on Pyrazole Binding to Liver Alcohol Dehydrogenase Pia A N D E R S O N , Jan KVASSMAN, Anders LINDSTROM, Bertil OLDEN, and Costa PETTERSSON Avdelningen for Biokemi, Kemicentrum, Lunds Universitet, Lund (Received October 27, 1980) zyxwv 1. Kinetic and equilibrium data have been determined at different pH between 4 and 10 for binding of the inhibitor pyrazole to liver alcohol dehydrogenase and to the binary complexes formed between enzyme and NADH or NAD'. 2. Pyrazole binding to free enzyme requires the protonated form of an ionizing group with a pK, of 9.2, agreeing with the pKa value reported for the water molecule bound at the catalytic zinc ion of the enzyme subunit. The rate of association of the inhibitor to the enzyme . NAD' complex exhibits a similar pKa-7.6-dependence attributable to ionization of zinc-bound water in the latter binary complex. These observations lend support to the idea that pyrazole combines to the catalytic zinc ion on complex formation with the enzyme, zinc-bound water most likely being displaced by the inhibitor. 3. The rate of dissociation of the inhibitor from the ternary enzyme . NAD' . pyrazole complex is proportional to the hydrogen ion concentration over the examined pH range (4-8). This effect of pH, which is proposed to reflect ionization of the enzyme-bound inhibitor with a pK, value below 4 (indirectly estimated to 2.4), accounts for the exceptional stability of the ternary complex at neutral and alkaline pH. It is concluded that pyrazole, by analogy to water and alcohol ligands, undergoes a drastic pK, perturbation on binding to the catalytic zinc ion in the enzyme . NAD' complex. Pyrazole is an effective inhibitor of the liver alcohol dehydrogenase reaction, strictly competitive with the alcohol substrate [l]. Theorell et al. demonstrated that pyrazole inhibition is attributable to the formation of an exceptionally firm enzyme . NAD' . pyrazole complex showing a characteristic difference absorption band at about 295 nm [2]. Due to the stability and chromophoric properties of this ternary complex, pyrazole has turned out to be a most useful ligand from a methodological point of view. Transient-state kinetic methods based on the absorbance changes associated with pyrazole binding have been applied to monitor formation of the enzyme . NAD' complex during catalysis and to study the process of ternary-complex formation with non-chromophoric substrate and coenzyme analogues [3- 51. The tight binding of pyrazole in the ternary complex can be utilized to render NAD' dissociation insignificant in active-site titrations of the enzyme [2] and in kinetic studies of coenzyme binding [6]. The high affinity of the enzyme . NAD' complex for pyrazole makes it possible also to examine the kinetics of enzymic aldehyde reduction under single-turnover conditions by the 'catalytic suicide' method of McFarland and Bernhard [7]. Despite the extensive use of pyrazole for such methodological purposes in experiments carried out at different pH, detailed kinetic information about pyrazole binding to the enzyme . NAD' complex is available only for the reaction at pH 7 [ 3 ] . The present investigation was undertaken to characterize the kinetics of the binding process over the wide pH range (4- 10) usually considered in mechanistic studies of the enzyme. As will be illustrated in the following paper, such information is of primary interest for the design and evaluation of transient-state kinetic experiments in which pyrazole is used as a reporter ligand. Secondarily, examination of the kinetics of complex formation with pyrazole might provide inferences regarding the mechanism of substrate binding during liver alcohol dehydrogenase catalysis. Enzyme. Liver alcohol dehydrogenase or alcohol : NAD' oxidoreductase (EC 1.1.1.1). Reactions exhibiting half-times below 20 s were monitored photometrically by stop-flow techniques. The rapid-reaction EXPERIMENTAL PROCEDURE Materiuls Crystalline horse liver alcohol dehydrogenase from Boehringer (Mannheim) was further purified as described by McFarland and Bernhard [7]. When required, unbuffered enzyme solutions were prepared using 33 mM sodium sulphate for elution of the enzyme in the column chromatographic purification step. Enzyme concentrations were determined fluorimetrically by titration with NADH in the presence of isobutyramide [8] and are reported throughout as active-site concentrations. The coenzymes NAD' and NADH (Sigma Chemical Corp., grade 111 quality) were purified as described by Gurr et al. [9]. Pyrazole (Fluka AG) was recrystallized twice from water prior to use. Other reagents were of highest available quality and used without further purification. All experiments reported in the present investigation were performed at 25°C in 0.1 M ionic strength phosphate (pH 6- 10) or succinate (pH 4- 6) buffer solutions, prepared using water twice distilled from a quartz apparatus. zyxwvu zyxwvutsrqp Spectrometric Equipment and Dutu Processing equipment described previously [lo] was used, except that the Beckman DU monochromator was replaced by a Durrum D-I 30 monochromator. Reactions with half-times above 20 s were examined by standard spectrometric techniques using a Beckman Acta CIII spectrophotometer. Fluorescence measurements were made with a Perkin-Elmer MPF-2A spectrofluorimeter. Equilibrium constants for pyrazole binding to free enzyme and to the enzyme NADH complex were determined fluorimetrically as described by Sigman et al. [ I l l . Rates and amplitudes of the exponential transients observed in the kinetic binding studies were estimated by non-linear regression analysis as detailed previously [lo]. zyxwvutsrq zyxw zyxwvutsrq zyxwvutsr zyxwvutsr [Pyraroie] (mM) Pyruzolr Associutiorz Rutes Liver alcohol dehydrogenase (about 10 pM) was preincubated with 4 m M NAD' in phosphate buffer of appropriate pH between 6 and 10. The equilibrium solution thus obtained (which can be assumed to contain at least 967: of the enzyme as the binary enzyme . NAD' complex [12]) was mixed in the stop-flow apparatus with an equal volume of the same buffer solution containing varied concentrations of pyrazole (0.04- 10 mM). Apparent first-order rate constants for the exponential transient governing formation of the ternary enzyme . N . 4 D c . pyrazole complex were determined from the absorbance changes observed at 300 nm [3]. Fig. 1. Dependence on Iigund concentration of rlic role o f ' p y r ~ i z ibinding ~l~~ to the binary complex formed hetween liver alcoliol deliydrogenuse and N A D ' . Apparent first-order rate constants ( r ) determined at 25 C in 0.1 M ionic strength phosphate buffer from the absorbance changes observed at 300 nm o n reacting the enzyme (about 5 pM) with varied concentrations of pyrazole in the presence of 2 mM NAD' O t c zyxwv zyxwvutsrq P j w z o l e Dissociulion Rutc7.s Liver alcohol dehydrogenase (about 10 pM) was preincubated with 4 niM N A D and 60 pM pyrazole in phosphate buffer of appropriate pH between 6 and 8. Displacement of pyrazole from the ternary complex formed was then effected by the addition of ;I large excess of trifluoroethanol (SO m M after mixing) in the m n e buffer solution. Apparent first-order rate constants for the displacement reaction were determined from the absorbance changes observed at 300 nm, measurements being made by stop-flow techniques (pH 6 - 7 ) or conventional spectrophotometric techniques (pH 7 - 8). The transient-state kinetics of the reverse reaction (displacement of trifluoroethanol from the enzyme . NAD' . alcohol complex by the addition of pyrazole) were examined over the pIH range 4-6 by the pH-jump method described previously IS]. RESULTS Pyrurole As.soc.iutioii Rates Eqn (1) shows the reaction scheme for pyrazole (P) binding to the binary complex (EO) formed between liver alcohol dehydrogenase and NAD'. EO +P k- 1% 1 I I 4 I I I 6 8 10 PH Fig. 2. Effect of p H on the on-velocity constunifor pyrazole binding GO the binary complex formed between liver alrohol deliydrogenuse and NAD' , Estimates of the on-velocity constant ( k , ) calculated from data such as. and including, those in Fig. 1 (0)and Fig.5 (0). The curve drawn was calculated from Eqn (3) with k r = 350 sK' m M - ' and pK7 = 7.6 zyxw by stop-flow techniques [3] at different pH between 6 and 10. As indicated by the typical results in Fig. 1. I' was found to be linearly dependent on the pyrazole concentration at each fixed pH. Evaluation of the binding data in terms of Eqn (2) showed that the magnitude of k - l was too small (less than 5 s-') to be reliably determined from these experiments. The estimates of k l obtained at different pH are given in Fig.:! and conform to the relationship zyxwvutsrqponmlkjihgfedcbaZYXW zyxwvutsr zyxwvuts iI 4 EOP. (1) I When pyrazole is present in large excess to enzyme, the kinetics of the binding process in Eqn (1) are governed by a single exponential transient. The corresponding apparent first-order rate consiant ( r ) is given [13] by I' = -3 + k,[P]. (2) Estimates of I' for the binding of varied concentrations of pyrazole to the enzyme . NAD' complex were determined (3) with kf = 0.35 (+ 0.05) s-l pM-' and pK7 = 7.6 (I0.2). It may be concluded from this result that pyrazole association to the enzyme . N A D + complex requires the protonated form of an ionizing group exhibiting a pKa of 7.6 in the binary complex, agreeing with the pK, value reported for the water molecule bound at the catalytic zinc ion of the enzyme subunit [I. 14.151. zyxwvuts 55 1 2l 4 6 PH zyxwvutsrqp zyxwvutsrq 0' 0 I I I I zyxwvutsr zyxwvut zyxw 8 25 50 75 100 [Pyrarole](mM) Fig. 3. Eflect O f p H on the off-velocity constant,for pyrazole binding to the binary complex .formed between liver alcohol dehydrogenase and NAD' . Estimates of the off-velocity constant (k-1) calculated from data such as. and including, those in Fig. 5 (0).Values of k -1 for the pH range 6 - 8 ( 0 ) refer to the apparent first-order rate constant for the absorbance changes observed at 300nm on mixing the enzyme (10 pM), preincubated with 4 mM NAD' and 60 pM pyrazole, with 100 mM trifluoroethanol at 25°C in 0.1 M ionic strength phosphate buffer. The line drawn has a slope of - 1 Fig.4. Displacement of trijluoroethanol with pyrazole from the ternary complex formed with liver alcohol dehydrogenase and NAD' . Apparent first-order rate constants (r) calculated from the absorbance changes recorded at 300 nm on mixing enzyme (12 pM), preincubated at pH 6 with NAD ' (4 mM) and trifluoroethanol (160 mM) in 33 mM sodium sulphate, with a 0.1 M ionic strength succinate buffer solution containing varied concentrations of pyrazole. The pH of the reaction solution after mixing was 4.0. The curve drawn represents a least-squares fit of Eqn ( 5 ) with the assumption that kz = 60 s - ' mM-' and k - 2 = 330 sK1 Pyruzok Dissociation Rates centrations of pyrazole. Fig. 4 shows the variation with pyrazole concentration of the apparent first-order rate constant r for the displacement reaction at pH 4. As has been previously reported [S], r increases non-linearly with increasing pyrazole concentration towards a maximum value (about 330 s - ' ) representing the magnitude at pH 4 of the trifluoroethanol dissociation rate constant k - 2 . The present extension of the measurements to much lower pyrazole concentrations makes it possible to arrive also at an extrapolated value (about 50 s - ' ) of r at zero pyrazole concentration; according to Eqn (5), the latter value of r should provide an approximate measure of the pyrazole dissociation rate constant k-1. More precise estimates of k - I were determined by non-linear regression analysis. A fit of Eqn ( 5 ) to the data in Fig. 4 with theassumptions [ S l t h a t k ~= 60s-1 mM-'andk-2 = 330s-' gave k - I = 43 (k 6) s-' and kl = 0.40 (+ 0.07) s-' pM-'. Estimates of the pyrazole dissociation rate constant thus obtained over the pH range 4 - 6 are included in Fig. 3. They show that the proportionality between k - 1 and hydrogen ion concentration persists down to pH 4. The corresponding estimates of kl were found to be independent of pH and to differ insignificantly from thelimiting valuekf = 0.35 s-l pM-' calculated from the data in Fig. 2. It can be inferred from these results that the relationships for pyrazole binding established by the measurements above pH 6 are valid also over the pH range 4 - 6. The present estimates of k-1 (0.040 s - ' ) and kl (0.28 s-' pM-') at pH 7 correspond to an equilibrium constant of 0.14 pM for pyrazole dissociation from the ternary complex. These binding data agree with those reported by Shore et al. [3] for the reaction at pH 7 (kl = 0.284 s-l pM-' and k - l / k l = 0.142 pM). zy zyxwvutsrq zyxw Eqn (4) shows the reaction scheme for a competitive [I. 51 binding of pyrazole (P) and trifluoroethanol (T) to the enzyme . NAD' complex (EO) EOT-EO 1 > lPl EOP. (4) When ligand concentrations greatly exceed the total concentration of enzyme, the kinetics of the reaction system in Eqn (4) will be governed by two exponential transients [12]. Only the slower one of these transients would be expected to be detectable by stop-flow techniques, the corresponding apparent first-order rate constant ( r ) being given [I21 by (5) Determinations of the rate constant k-1 for pyrazole dissociation from the enzyme . NAD+ . pyrazole complex were performed by two different methods based on Eqn ( 5 ) . In one series of experiments pyrazole was preincubated with the enzyme . NAD' complex and displaced from the ternary complex formed by the addition of a large excess of trifluoroethanol. According to the binding data reported previously [5] and in the preceding section, ligand concentrations in these experiments were such ([PI = 30 pM and [TI = 50 mM) that Eqn (5) should reduce effectively to r = k-1 [32], i.e. the observed rate of the displacement process can be assumed to provide a direct measure of the pyrazole dissociation rate. The estimates of k - 1 thus obtained at different pH between 6 and 8 are given in Fig.3 and show that the magnitude of k-1 is approximately proportional to the hydrogen ion concentration over the examined pH range. Above pH 8, the pyrazole dissociation rate became too low to be reliably determined. In a second series of experiments trifluoroethanol was preincubated with the enzyme . NAD' complex at pH 6. Displacement of the alcohol from the ternary complex formed was then effected at lower pH by the addition of varied con- zyxwv Pyrazole Binding in the Absence of' N A D i Equilibrium constants (KE,p) for pyrazole dissociation from the binary enzyme . pyrazole complex were determined fluorimetrically using auramine 0 as a reporter ligand [ I l l . zy zyxwvutsrqponmlkj zyxwvutsrqp zyxwv zyxwvutsrqponmlkjihgf zyxwvuts /OH- 80 I- E "AD+ El i cn 40 E /L- "AD' zyxwvutsrqp zyxwvutsrqponm 6 I I I I 7 8 9 10 PH Fig. 5. Effect of p H on p.vru:olr binding to liver alcohol dehydrogenase in the ubsencr of' coc'n:j'mes. Equilibrium constants ( K E , ~for ) pyrazole dissociation from the binarycomplex formed with free enzymedetermined fluorimetrically at 25 C in 0.1 M ionic strength phosphate buffer using aurainine 0 as a reporter ligand. The curve drawn was calculated from Eqn (6) with K& = H mM and pKg = 9.2 Scheme 1. Cenerulizedmechanisnifor the binding of un ionizing ligundto the binary complex formed between liver alcohol dehydrogennse nnd NAD' The ionizing ligand (LH) is assumed to combine to the catalytic zinc ion of the enzyme subunit (E) with displacement of zinc-bound waler zyxwvutsr remains unaffected by p H between pH 6 and 10 can be interpreted in the same way. These results lend strong support to the idea that pyrazole combines to the catalytic zinc ion on complex formation with the enzyme [I , 2 ] , zinc-bound water most likely being displaced in the binding process. The data in Fig.3 establish that the rate constant for with K ~ J= 8 m M and pK9 = 9.2. It may be concluded that pyrazole dissociation from the enzyme NAD' . pyrazole pyrazole binding to free enzyme is dependent on a proton complex is approximately proportional to the hydrogen ion dissociation equilibrium with a pK, value agreeing with that concentration between pH 4 and 8. It can be inferred from this attributed to ionization of zinc-bound water in free enzyme observation that the ternary complex participates in an ionization step with a pKa value below 4, pyrazole desorption [ I , 15,161. The equilibrium constant ( K E ~p), for pyrazole dissociation occurring exclusively from the protonated form of the from the enzyme NADH . pyrazole complex was similarly complex. This ionization step appears to be characteristic for determined at different p H between 6 and 10, which gave the complex formed with pyrazole as the ligand. There is no KeR,P= 4 (21) mM independently of pH. N o attempt was similar effect of p H on the corresponding dissociation reaction made to obtain estimates of KER,Pover the pH range with ligands such as decanoate [ 5 ] and 2,2'-bipyridine [15], (10-12) where ionization of zinc-bound water in the en- which lack ionizing functional groups. Alcohol dissociation zyme . NADH complex becomes kinetically significant [17]. from the corresponding ternary complex, however, has been Displacement of pyrazole from the enzyme . N A D H . py- found to be controlled by ligand-dependent ionization steps razole complex by the addition of varied concentrations of with pKa values within the range 4- 7 [5,14]. Kvassman and N,N-dimethylaminocinnamaldehydeshowed n o rate limi- Pettersson [14] interpreted the latter effects of p H in terms of tation detectable by stop-flow techniques. This observation an alcohol/alcoholate ion equilibration of the enzyme-bound indicates that the rate constant for pyrazole dissociation from ligand, as indicated by the generalized binding mechanism in Scheme 1. They pointed out that alcoholate ion formation the latter ternary complex exceeds 500 s-'. must be drastically facilitated by coordination of the ligand to the catalytic zinc ion and by electrostatic interaction with the positive charge on the nicotinamide ring of the coenzyme. DISCUSSION Similar interactions have been proposed to account for the Mechanism of Pyra; ole Birdrig pK, of zinc-bound water in the enzyme . NAD' complex The catalytic zinc ion in the liver alcohol dehydrogenase [1,12] and can be anticipated to be at hand and facilitate subunit is bound by three protein hgands and coordinates a deprotonation of any ionizing ligand when bound to the water molecule as a fourth inner-sphere ligand [I]. Ionization catalytic zinc ion in the enzyme . NAD' complex. For these of this water molecule: which exhibits a pK, value of 9.2 in reasons, and since the kinetics of pyrazole binding to the free enzyme, 7.6 in the enzyme . NAD+ complex, and 11.2 in latter complex are consistent with Scheme 1 (with pK6 < 4), the enzyme . NADH complex [I, 14- 171, has been found to it seems reasonable to attribute the observed effect of pH on prevent complex formation between enzyme and external the pyrazole dissociation rate to ionization of the ligand at ligands which combine to the catalytic zinc ion [5,14- 171. the ternary-complex level. This would explain why pyrazole This is what might be expected if the external ligand is bound desorption requires protonation of the ternary complex. The by a substitution mechanism with displacement of zinc-bound coulombic force field provided by the catalytic zinc ion and water [16]. The present investigation provides evidence that enzyme-bound NAD+ will render the rate of dissociation of pyrazole binding to free enzyme (Fig. 5), as well as pyrazole the negatively charged ligand negligibly small compared to association to the cnzyme . NAD' complex (Fig. 2), is de- the rate of desorption of the neutral pyrazole molecule. pendent on the ionization state of zinc-bound water with According to the data in Fig.5, pyrazole binding to free preferential binding to the unionized form. The observation enzyme shows no p H dependence attributable to ionization that the affinity of pyrazole for the enzyme . N A D H complex of the ligand at the binary-complex level. This means that the The estimates obtained at different p H between 6 and 10 are given in Fig. 5 and conform to the relationship ' z zyxwvu 553 ionizing group with a pKa value below 4 in the ternary complex must exhibit a pK, value above 10 in the binary enzyme . pyrazole complex. The present results, therefore, provide additional evidence that the pKa of an ionizing ligand may be drastically decreased on binding to the enzyme . NAD' complex and indicate that the presence of NAD' is a most important factor contributing to this pKa perturbation. The binding mechanism in Scheme 1 is consistent with the spectrometric and titrimetric data reported by Theorell et al. [2], who found that ternary-complex formation between enzyme, NAD', and pyrazole results in a release of close to one proton per enzyme subunit at pH 7. Theorell et al. concluded from their results that N-2 of pyrazole combines to the active-site zinc ion, while N-1 of pyrazole was proposed to be covalently bound with liberation of a proton to C-4 of the nicotinamide ring of NAD'. We agree that the protons released on pyrazole binding to the enzyme . NAD' complex most likely derive from the imino group of the ligand, but do not consider it necessary to assume that pyrazole and NAD' are covalently linked in the ternary complex formed. Non-covalent interactions analogous to those proposed to account for the pK, perturbations of zincbound water and alcohol ligands [I41 might be sufficient to explain the increased acidity of the pyrazole imino group in the ternary complex. Two main inferences can be drawn from this observation. Firstly, it seems obvious that the exceptional stability of the enzyme . NAD+ . pyrazole complex at high pH ( k - l / k l z 0.03 pM) is attributable mainly or entirely to the low magnitude of pK6, i.e. to ionization of the ligand resulting from the drastic pKa perturbation brought about by complex formation. The low pK, value for this ionization process accounts for the steady decrease in stability of the ternary complex with increasing hydrogen ion concentration over the examined pH range. The equilibrium dissociation constant k - l / k l is of less remarkable magnitude (about 100 pM) at pH 4, and might well approach a limiting value exceeding 1 mM at extremely low pH. This would require that the limiting value at low pH of the corresponding off-velocity constant (Fig. 3) exceeds 350 s-*. Strong evidence that such is actually the case is provided by the observation that pyrazole dissociation from the ternary complex formed with reduced coenzyme occurs at an immeasurably high rate (> 500 s - l ) . Secondly, it may be noted that an energetic binding contribution of maximally 11 kJ/mol would be considerably smaller than that provided by an ordinarily weak hydrogen bond (20- 30 kJ/mol). The possibility that covalent interactions with NAD' contribute to the binding of pyrazole in the neutral state, therefore, is rendered highly unlikely by the present results. It then seems reasonable (and is justified by data reported for substrate binding to the enzyme [14]) to assume that interactions of similar nature and strength contribute to the binding of the neutral pyrazole molecule in the ternary complexes formed with NAD' and NADH. In other words, the value (4mM) observed for K E R ,below ~ pH 10 can be taken to provide an approximate measure of the actual magnitude of K&,. According to Eqn (S), this estimate of K & p corresponds to pK6 = 2.4 and hence to a NAD'-induced perturbation of the pKa of enzyme-bound pyrazole by more than 7 pK. A pKa perturbation of that order of magnitude would not be unreasonably large to derive mainly from electrostatic effects of the positively charged nicotinamide ring of enzyme-bound NAD'. Irrespective of the detailed nature of the interactions contributing to stabilization of the ionized state of the ligand in the ternary complex, the present investigation seems to establish that the mechanism of pyrazole binding to the enzyme . NAD+ complex is closely analogous to that of the binding of alcohol ligands. This provides additional justification for the methodological use of pyrazole as a chromophoric substrate analogue, at least in applications where the low dissociation rate of pyrazole at alkaline pH poses no problems. In particular, the kinetic data now reported will make it possible to use pyrazole as a wellcharacterized reporter ligand in transient-state kinetic studies of the binding of non-chromophoric alcohols. The value of the present results in the latter respect will be illustrated in the following paper. zyxwv zyxwvutsrqp Stability of the Enzyme . NAD' . Pyrazole Complex The data in Fig. 5 indicate that the affinity of free enzyme for pyrazole is of the same order of magnitude as the affinity for other inhibitory nitrogenous bases [l]. Pyrazole binding to the enzyme . NAD' complex, however, appears to be exceptionally tight, the corresponding apparent equilibrium dissociation constant ( k - , / k l ) approaching a value of 0.03 pM at high pH (Fig.2 and 3). According to Scheme 1, the pH dependence of k - l / k l can be expressed as where K & p denotes the true equilibrium constant for dissociation of the neutral pyrazole molecule from the ternary complex. Eqn (7) shows that the tight binding of pyrazole at high pH could be due either to a high value of K6 (stabilization through ionization of the bound ligand) or to a low value of Go,p(exceptionally firm binding of the neutral ligand) or both. The present kinetic studies have not been carried out at sufficiently low pH to provide any direct estimates of the constants K6 and K & p . Assuming that pK7 = 7.6 (Fig.2), however, it follows from Eqn (7) and the present estimates of k - l / k l (0.14 pM at pH 7) that the two constants are interrelated approximately through K&l,p= 1.1 K6. (8) Since the results in Fig.3 establish that pK6 < 4, it may be concluded from Eqn (8) that go,, > 0.1 mM. This means that binding of the neutral pyrazole molecule to the enzyme . NAD' complex is stabilized by less than 1.9 pK (corresponding to 11 kJ/mol) compared to pyrazole binding to free enzyme (KE,p= 8 mM) and to the enzyme . NADH complex (KER,P= 4 mM). This investigation was supported by grants from the Swedish Natural Science Research Council. REFERENCES 1. B r i n d h , C.-I., Jornvall, H., Eklund, H. & Furugren, B. (1975) in The Enzymes (Boyer, P. D., ed.) 3rd edn, vol. IIA, pp. 103-190, Academic Press, London. 2. Theorell, H. & Yonetani, T. (1963) Biochem. Z . 338, 537-553. zyxwvutsrqponmlk zyxwvutsrqpon zyxwvuts zyxw zyxwvutsrq zyxwvutsr 3 . Shore, J . D. &Gilleland. M . J . (1970)J. B i d . Chem. 245,3422-3425. 4. KvdsSman. J. & Pcrterason. G. (1976) Eur. J . Biochem. 6Y, 279-287. 5. Kvassman. J . & Petterson, G . (1980) Eur. J . Biochern. 103, 557-564. 6. DeTraglia. M . C.. Schmidt. J., Dunn, M. F. & McFarland, J. T. (1977) J . Biol. C'hem. 252. 3493 -3.500. 7. McFarland, J. 'I-. & Bernhard. S. A . (1972) Biochemistry, 11, 14861493. 8. Theorell, H. & McKinley-McKee, J . S. (1961) Actu Chem. ScanJ. 15, 1797 - 1810. 9. Gurr, P. A , . Bronsliill, P. M., IHancs, C. S. & Wong, J. T. F. (1972) CUM.J . B i o ~ ~ h 50, e f ~1376. 1384. 10. 11. 12. 13. 14. IS. 16. 17. Kvassman, J. & Pettersson, G . (1978) Eur. J . Bioclzem. 87. 417-427. Sigman, D. S. & Glazer, A. N. (1972) J . B i d . Chem. 247. 334-341. Kvassman, J. & Pettersson. G. (1979) Eur, J . Biochem. IOU, 115- 123. Hammes, G. G. & Schimmel, P. R. (1970) in The Enzyrnes (Boyer, P. D., ed.) 3rd edn. vol. 2, pp. 67- 114, Academic Press, London. Kvassman, J. & Petterson, G. (1980) Eur. J. Biochem. 103, 565-575. Evans, S. A. & Shore, J . D. (1980) J . Biol. Cliern. 255. 1509-1514. Andersson, P., Kvassman, J.. Lindstrom. A,, Oldtn, B. & Peltemon, G. (1980) Eur. J . Biochem. 108, 303-312. Andersson, P., Kvassman, J.. Lindstrom, A , , Olden, B. & Pettersson, G . (1981) Eur. J . Biochem. 113, 425-433. P. Andersson, J . Kvassman. A . Lindstrom, B. Olden, and G. Pettersson, Avdelningen for Biokemi, Kemicentrum, Lunds Universitet. Box 740, S-220 07 Lund. Sweden