Mechanism of D-fructose isomerization by arthrobacter D-xylose isomerase

Abstract

The mechanism of D-fructose isomerization by Arthrobacter D-xylose isomerase suggested from X-ray-crystallographic studies was tested by detailed kinetic analysis of the enzyme with various metal ions at different pH values and temperatures. At D-fructose concentrations used in commercial processes Mg2+ is the best activator with an apparent dissociation constant of 63 μM; Co2+ and Mn2+ bind more strongly (apparent K(d) 20 μM and 10 μM respectively) but give less activity (45% and 8% respectively). Ca2+ is a strict competitive inhibitor versus Mg2+ (K(i) 3 μM) or Co2+ (K(i) 105 μM). The kinetics show a compulsory order of binding; Co2+ binds first to Site 2 and then to Site 1; then D-fructose binds at Site 1. At normal concentrations Mg2+ binds at Site 1, then D-fructose and then Mg2+ at Site 2. At very high Mg2+ concentrations (> 10 mM) the order is Mg2+ at Site 1, Mg2+ at Site 2, then D-fructose. The turnover rate (k(cat.)) is controlled by ionization of a residue with apparent pK(a) at 30°C of 6.0 ± 0.07 (Mg2+) or 5.3 ± 0.08 (Co2+) and ΔH = 23.5 kJ/mol. This appears to be His-219, which is co-ordinated to M[2]; protonation destroys isomerization by displacing M[2]: Co2+ binds more strongly at Site 2 than Mg2+, so competes more strongly against H+. The inhibition constant (K(i)) for the two competitive inhibitors 5-thio-α-D-glucopyranose and D-sorbitol is invariant with pH, but K(m(app.)) in the Mg[1]-enzyme is controlled by ionization of a group with pK(a) 6.8 ± 0.07 and ΔH = 27 kJ/mol, which appears to be His-53. This shows that K(m(app.)) is a complex constant that includes the rate of the ring-opening step catalysed by His-53, which explains the pH-dependence. In the Mg[1]Mg[2]-enzyme or Co[1]Co[2]-enzyme, the pK(a) is lower (6.2 ± 0.1 or 5.6 ± 0.08) because of the extra adjacent cation. Hence the results fit the previously proposed pathway, but show that the mechanisms differ for Mg2+ and Co2+ and that the rate-limiting step is isomerization and not ring-opening as previously postulated.