Effects of mutations of the active site arginine residues in 4-oxalocrotonate tautomerase on the pK(a) values of active site residues and on the pH dependence of catalysis

The unusually low pK(a) value of the general base catalyst Pro-1 (pK(a) = 6 .4) in 4-oxalocrotonate tautomerase (4-OT) has been ascribed to both a low dielectric constant at the active site and the proximity of the cationic re sidues Arg-11 and Arg-39 [Stivers, J. T., Abeygunawardana, C., Mildvan, A. S., Hajipour. G., and Whitman, C. P. (1996) Biochemistry 35, 814-823]. In a ddition, the pH-rate profiles in that study showed an unidentified protonat ed group essential for catalysis with a pK(a) of 9.0. To address these issu es. the pK(a) values of the active site Pro-1 and lower limit pK(a) values of arginine residues were determined by direct N-15 NMR pH titrations. The pK(a) values of Pro-1 and of the essential acid group were determined indep endently from pH-rate profiles of the kinetic parameters of 4-OT in arginin e mutants of 4-OT and compared with those of wild type. The chemical shifts of all of the Arg N epsilon resonances in wild-type 4-OT and in the R11A a nd R39Q mutants were found to be independent of pH over the range 4.9-9.7, indicating that no arginine is responsible for the kinetically determined p K(a) of 9.0 for an acidic group in free 4-OT. With the R11A mutant, where k (cat)/K-m was reduced by a factor of 10(2.9), the pK(a) of Pro-1 was not si gnificantly altered from that of the wild-type enzyme (pK(a) = 6.4 +/- 0.2) as revealed by both direct N-15 NMR titration (pK(a) = 6.3 +/- 0.1) and th e pH dependence of k(cat)/K-m (pK(a) = 6.4 +/- 0.2). The pH-rate profiles o f both k(cat)/K-m and k(cat) for the reaction of the R11A mutant with the d icarboxylate substrate, 2-hydroxymuconate, showed humps, i.e., sharply defi ned maxima followed by nonzero plateaus, The humps disappeared in the react ion with the monocarboxylate substrate, 2-hydroxy-2,4-pentadienoate, indica ting that, unlike the wild-type enzyme which reacts only with the dianionic form of the dicarboxylic substrate, the R11A mutant reacts with both the 6 -COOH and 6-COO- forms, with the 6-COOH form being 12-fold more active. Thi s reversal in the preferred ionization state of the 6-carboxyl group of the substrate that occurs upon mutation of Arg-11 to Ala provides strong evide nce that Arg-11 interacts with the 6-carboxylate of the substrate. In the R 39Q mutant, where k(cat)/K-m was reduced by a factor of 10(3), the kinetica lly determined pK(a) value for Pro-1 was 4.6 +/- 0.2, while the ionization of Pro-1 showed negative cooperativity with an apparent pK(a) of 7.1 +/- 0. 1 determined by 1D N-15 NMR. From the Hill coefficient of 0.54, it can be s hown that the apparent pK(a) value of 7.1 could result most simply from the averaging of two limiting pK(a) values of 4.6 and 8.2. Mutation of Arg-39, by altering the structure of the beta-hairpin which covers the active site , could result in an increase in the solvent exposure of Pro-1, raising its upper limit pK(a) value to 8.2. In the R39A mutant, the kinetically determ ined pK(a) of Pro-1 was also low, 5.0 +/- 0.2, indicating that in both the R39Q and R39A mutants, only the sites with low pK(a) values were kineticall y operative. With the fully active R61A mutant, the kinetically determined pK(a) of Pro-1 (pK(a) = 6.5 +/- 0.2) agreed with that of wild-type 4-OT. It is concluded that the unusually low pK(a) of Pro-1 shows little contribu tion from electrostatic effects of the nearby cationic Arg-11, Arg-39, and Arg-61 residues but results primarily from a site of low local dielectric c onstant.