Liver Alcohol Dehydrogenase-Coenzyme Reaction Rates*

Specific rate constants for the association and dissocia- tion of liver alcohol dehydrogenase.coenzyme complexes have been measured within the pH range 6 to 10.5. NADH and NAD+ association rates were measured under pseudo-first order conditions by monitoring changes in enzyme and coenzyme fluorescence. Dissociation rates were measured by replacing bound NADH or NAD+ with tight binding ternary complex inhibitors (NAD+, pyrazole) and (NADH, iso-butyramide), respectively. Results indicate that the rate constant for NAD+ and NADH association show the same pH rate profile with pK,, = 9.5. Furthermore, the pK,, for o-phenanthroline association to liver alcohol dehydrogenase is 8.1. Since x-ray studies indicate that o-phenanthroline displaces the water bound to Zny+ at the active site (Branden, G. in The Enzymes P., the most associated with ionization water

Specific rate constants for the association and dissociation of liver alcohol dehydrogenase.coenzyme complexes have been measured within the pH range 6 to 10.5. NADH and NAD+ association rates were measured under pseudofirst order conditions by monitoring changes in enzyme and coenzyme fluorescence. Dissociation rates were measured by replacing bound NADH or NAD+ with tight binding ternary complex inhibitors (NAD+, pyrazole) and (NADH, isobutyramide), respectively. Results indicate that the rate constant for NAD+ and NADH association show the same pH rate profile with pK,, = 9.5. Furthermore, the pK,, for o-phenanthroline association to liver alcohol dehydrogenase is 8. 1. Since x-ray studies indicate that o-phenanthroline displaces the water bound to Zny+ at the active site (Branden, C.-I., Jornvall, H., Eklund, H., and Furugaen, G. (1975) in The Enzymes (Boyer, P., ed) p. 133, Academic Press, New York cl)), the p&, of 8.1 is most likely associated with ionization of this water molecule. Thus ionization of water bound to zinc apparently does not control the rate of nucleotide association as previously suggested (Taniguchi, S., Theorell, H., and Akeson, A. (1967) Acta Chem. Sand. 21,1903 (2)).
The dissociation rate constant for E-NAD+ shows a pK,, = 8.0. By contrast, the rate constant for E-NADH dissociation exhibits a complex behavior not interpretable as associated with ionization of a single group. Agreement between equilibrium-binding constants for nucleotide and kdisx,c,atiun/ kaSSOCiatiOn ratios is quite good. Furthermore, agreement with rate constants determined from steady state kinetic experiments is good except for the rate of NADH association. Mechanistic consequences of the nucleotide kinetic results are discussed.
We have also discovered that liver alcohol dehydrogenase rapidly loses Zn'+ at pH values >lO; this finding makes previous studies at these very basic pH values suspect. to be the rate-limiting step for aliphatic aldehyde reduction (3)(4)(5)(6). However, the direct measurement of coenzyme association and dissociation rates as a function of pH have not been reported.
Two major proposals for the catalytic action of LADH' have been suggested.
One mechanism (Scheme 2) involves the active site zinc ion as a Lewis acid catalyst in the activation of the aldehyde carbonyl for reduction by NADH via inner sphere coordination of the carbonyl oxygen (7-9) and the subsequent acid catalzyed breakup of the zinc alcoholate product via the involvement of a protonic amino acid side chain residue (10). The second mechanism (Scheme 3) involves acid catalysis of aldehyde reduction by a water molecule bound to the active site zinc ion (11).
The apparent enzyme pK,, values which regulate the association and dissociation of NAD+, as determined from the analysis of the steady state kinetic behavior of liver alcohol dehydrogenase are pK,, = 9.5 for association and pK,, = 8.0 for dissociation (6). The apparent pK,, values which regulate the affinity of the site for NAD+ and NADH (2) have been reported as pK,, = 8.5 for NAD+ and pK, = 10.5 for NADH. The pK,, values for the enzyme group controlling NAD+ binding is perturbed from pK,, = 8.5 to pK,, = 6.5 in the presence of NAD+ .  If a single  ionizable  group controls  coenzyme  binding,  then sizeable  pK,,  perturbations  must accompany  the association  and dissociation steps for coenzyme  binding  to account  for the  Rates -Association rates for NAD+ between pH 7.9 and 10.5 were measured by mixing liver alcohol dehydrogenase (4 pN) with a series of NAD+ solutions (10 to 60 PM) in a stopped flow experiment.
Measurement was made of the decrease in enzyme fluorescence on formation of E. NAD complex (A,, = 280 nm, A,,,, > 300 nm); data were treated as described below. NADH Dissociation Rates -NADH dissociation rates were determined over the pH range 6.0 to 10.5 by mixing liver alcohol dehydrogenase (6.5 pi) preincubated with NADH (21 FM) in one syringe, with NAD+ (2.3 mM) and pyrazole (9.4 mM) in a second syringe. The decrease in fluorescence of the E NADH complex (A,, = 330 nm, A,,~ > 340 nm) was followed in the stopped flow spectrometer. Experiments at several NAD+ and pyrazole concentrations (Table II) confirmed that the observed process was the first order E (NADH) dissociation.  If these conditions do not hold, then the apparent rate of complex formation will depend on the concentration of one or more of the ligands involved (e.g. NAD+ or NADH).' The data presented in Table I demonstrate that the rate of formation of the ternary E(NADH, IBA) complex on mixing the E(NAD+) complex with a solution of NADH and IBA is insensitive to lo-fold changes in concentrations of either NAD+ or NADH. Fig. 1 shows a typical stopped flow trace of the single exponential increase in fluorescence at -390 nm (A,,. = 330 nm) associated with formation of the E(NADH, IBA) complex. The data in Table I were obtained from similar traces by monitoring the decrease in optical density at 355 nm (the maximum in the difference spectrum between free and bound NADH). second order rate process), and since rates measured both by fluorescence and by absorbance are identical, the observed rate process must be limited by the specific first order dissociation ofE(NAD+). Fig. 2 shows the results of this same experiment over the pH range 6.2 to 10.0. Note that the solid line represents the best tit theoretical protonic dissociation curve, calculated for a rate process dependent on the ionization of a group with pK,, = 8.0. The precision of the fit at high pH is good; however, at low pH values the observed rates reflect a response to pH which is slightly greater than first order in lH+l.
Formation of the E(NAD+) binary complex results in a significant quenching of the enzyme tryptophan fluorescence. Under the pseudo-first order condition [NAD+l > [enzyme], quenching is a single exponential rate process when the tryptophan fluorescence is monitored at -340 nm (A,,. = 280 nm). The dependence of the observed pseudo-first order rate constant on NAD+ concentration at pH 9.1 is shown in Fig. 3. Since the approximate equation for the single step reversible binding of NAD+ to liver alcohol dehydrogenase is as follows (18) k apu = k,, (NAD') + km, for the mechanism k+, E+NAD+,--' k-1 E -NAD+ the second order rate constant for association of NAD+, k+l, is obtained as the slope of the plot in Fig. 3. In calculating association rate constants for NAD+ and NADH we have always assigned intercepts as the measured dissociation rate constants, k,. Linear regression analysis was employed to determine the "best tit" line to experimental data. Fig. 4 shows the pH dependence of the specific second order association constants for NAD+-liver alcohol dehydrogenase binary complex formation for the pH range 7.9 to 10.5. At more acidic pH values the binding constant of NAD+ is not sufficiently large to yield suitably large amounts of fluorescence quenching. Over the limited pH range amenable to investigation, the values of the association rates can be fit to a theoretical acid dissociation curve with an apparent pK,, of 9.5 (Fig. 4). (2) +I The data presented in Table II show that the observed rate of NADH dissociation, as measured by the decrease in fluorescence for enzyme-bound NADH, is independent of the concentrations of NAD+ and pyrazole when the concentrations used are sufficiently high to insure that the rate of E(NAD+, pyr) complex formation (the rate constant at saturating [NAD+l is 100 s-' at pH 7.0 and pH 8.75) is rapid relative to E(NADH) dissociation (15). The pH dependence of the specific ECNADH) dissociation rate over the pH range 6.0 to 10.5 is shown in Fig.  5. In contrast to the behavior of the E (NAD+) complex, the rate of dissociation of the E(NADH) complex varies only slightly as a function of pH, showing a small, nearly linear increase in rate constant with increasing pH. That this process is indeed dependent on pH can be seen in Fig. 7

Liver Alcohol
Dehydrogenase-Coenzyme Reaction Rates constants not available from our data at a given pH value. At all pH values agreement is within a factor of 3, and at most pH values agreement is even better. This is an indication that over the range in which we are able to compute these calculated values, binding is a single step process and that the equations used in our calculations are adequate to tit the binding mechanism. Table IV shows a comparison of our directly measured rate constants (identified by the constants in Scheme 1) with those derived from steady state rate experiments. Fig. 8 also shows graphically the steady state results and pH dependence for various rate parameters. Agreement is generally good, although the pK, for rate of NADH association is different from ours, and we are unable to compare our results for the rate of NAD+ association below pH 8.0 because   l of experimental difficulties. At pH values above 9, earlier researchers have reported that either the NADH binding stoichiometry or the enzyme affinity is drastically altered (20-22). Because the decreased fluorescence enhancement on NADH binding below pH 10 is similar to the reported behavior of apozinc LADH (231, we investigated the hypothesis that enzyme loses Zn'+ at basic pH. This would not be unexpected since competition with OH-at basic pH is often seen with complexes of Zn'+. Atomic absorption measurements comparing enzyme handled identically at pH 8.7 uersus pH 11.0 shows the enzyme species at pH 8.7 contains 2.0 2 0.3 g atom of zinc per 40,000 daltons, whereas at pH 11.0, the zinc content decreases to 1.1 * 0.2 g atom of zinc per 40,000 daltons. Thus, above pH 10, the binding and kinetic experiments may be erroneous due to the loss of zinc ion from the active site. Where we have reported rate experiments at pH > 10, these pH values were achieved by "jumping" the pH to basic values in the stopped flow spectrometer, i.e. enzyme was never allowed to stand at pH > 10. The absence of an enhanced fluorescence emission spectrum for bound NADH is consistent with the properties of a zinc-deficient enzyme (23).

Several interesting
conclusions can be drawn from this study of the kinetics of dinucleotide association and dissociation. Steady state kinetic studies (3)(4)(5)(6) have suggested that the LADH reaction both in the direction of aldehyde reduction and in the direction of alcohol oxidation is a compulsory ordered mechanism in which coenzyme dissociation is the rate-limiting process. Our data are quite consistent with this conclusion. The rate constants for dissociation of NAD+ (Fig. 2) correlate well with the turnover numbers for acetaldehyde reduction at all pH values (Table IV, Fig. 8). Likewise, our measured dissociation rate constants for NADH (Fig. 5) correlate well with turnover numbers for ethyl alcohol oxidation (Table IV, Fig. 8). For aromatic substrates which exhibit smaller turnover numbers than the acetaldehyde turnover number (e.g. acetaldehyde, K,,, = 9.72 s-l, while for benzaldehyde, k,,, = 4.0 ss', and for P-napthaldehyde, k,,, = 0.4 s-l at pH = 8.75) (6,14), NAD+ dissociation is not rate-limiting.
It is likely that alcohol dissociation becomes rate-limiting for these tightly binding aromatic substrates.
The good agreement between our measured dissociation rates for NAD+ and NADH and the k,,, values for acetaldehyde and ethanol turnover has interesting consequences relative to subunit interactions between protomers. It has been suggested that in oligomeric enzymes showing half-of-the-sites reactivity, liganding by substrates at the inactive subunit may cause a reduction in the energy of activation of the ratelimiting step (24). If this were the case with liver alcohol  with a more acidic pK,, than our data. The apparent pK,, from our NADH association rate constant data is 9.4 to 9.6 while the pK, for the steady state data is -8.7 to 8.9 (Table IV, Fig. 8 (1,25). In order to test the first hypothesis we have investigated the pH dependence of the rate of phenanthroline binding (Fig. 7). The pK, for this process is 8.1 2 0.1 with rates being slower at basic pH values. Since