The variation of kinetic parameters was examined for bovine brain hexokinase with glucose and MgATP as substrates. The -log(V(,1)) and -log(V(,1)/K(,m)) profiles for both substrates were examined and seen to decrease below pH 6.5. All profiles asymptotically approached slopes of -1, indicating that the loss of activity in each instance was due to the protonation of a single group on the enzyme. Analysis of the data indicated two ionizable groups are involved in the reaction. One functions in the binding of ATP and in catalysis while the other participates in the binding of glucose;The -log(V(,1)) profiles showed a "hump" attributed to a loss of activity in the pH region 7.5-5.5. This inhibition was found to be caused by aluminum present in commercial preparations of ATP and probably inhibits as an AlATP complex. In an effort to remove this ion, we developed an extraction procedure that eliminates more than 99% of the contaminant with no loss of nucleotide. This method involves repeated extractions of the nucleotide solution with 8-hydroxyquinoline in chloroform, followed by one extraction at higher pH, that removes the final 10% of the aluminum contaminant that seems to be sequestered in a slowly dissociating complex and is unavailable to the 8-hydroxyquinoline at low pH;Two mechanisms have been suggested to account for the regulation of brain hexokinase by glucose 6-phosphate. One mechanism places glucose-6-P at an allosteric site, remote from the active site, while the second describes glucose-6-P as a simple product inhibitor of the enzyme, binding at the (gamma)-phosphate subsite within the ATP locus of the active site. To distinguish between these possibilities, we have undertaken a study of the back reaction of hexokinase I. Our data indicate that glucose-6-P displays classical Michaelis-Menten kinetics with brain hexokinase. This finding is consistent only with the high affinity glucose-6-P site on the enzyme being the catalytic site. The dissociation constant, estimated from the initial-rate experiments, is approximately 25 (mu)M, a value that agrees well with the inhibition constant for glucose-6-P in the forward direction.