A theoretical study of glucosamine synthase

Continuing our theoretical studies of glucosamine synthase catalysis, we have carried out MNDO and ab initio calculations of the first stage of the reaction, which involves the attack of a cysteine thiol group from the enzyme active site on the side chain carboxyamide group of glutamine, producing ammonia and thioester. The reactants were modelled by methyl mercaptate and acetamide, respectively. For two considered mechanisms of the reaction the energy surfaces were evaluated. Mechanism I, proposed by Chmara et al. (1985) involves the nucleophilic attack of a deprotonated thiol group on the carbonyl carbon atom. Mechanism II, postulated in our previous work (Tempczyk et al. 1989), assumes the concerted binding of the mercaptate sulphur to the carbonyl carbon and the sulfhydryl hydrogen to the amide nitrogen with simultaneous breaking of the S-H bond. The energy surface of mechanism I shows no minimum on the approach of the mercaptide anion towards the carbonyl carbon, which is also consistent with ab initio calculations in a 4-31 G basis set. Therefore, mechanism I seems to be unlikely. The same analysis of mechanism II shows that it leads to the desired products: methyl thioacetate and ammonia. The presence of a sulfhydryl hydrogen causes apparent pyramidicity of the amido nitrogen and lengthening of the C-N bond in the transition state, making conditions for the release of the ammonia molecule. The MNDO calculated energy barrier of the reaction is 50.1 kcal/mol and the approximate 4-31 G ab initio barrier (at the MNDO geometries of the substrate complex and the transition state) is 63 kcal/mol. The biggest energy contribution to the barrier comes from the breaking of the S-H bond, which also causes a large charge separation in the transition state. The latter affect may result in the stabilisation of the transition state in a real enzymatic environment when compared to the gas phase, e.g. by the interaction of the reacting center with a pair of oppositely charged amino acid side chains such as lysinium and glutamate (aspartate), which are present in the enzyme studied. To estimate the magnitude of this effect, molecular mechanics calculations were carried out on the reaction center at the transition state in our proposed model of the enzymatic active site. The site was supplemented by ammonium and acetate ion, which were to mimic the lysinium and glutamate/aspartate side chains. A transition state stabilization energy of 20 kcal/mol was obtained and this lowers the energy barrier to about 30 kcal/mol. This value is within the thermal energy range of an average protein and indicates that our mechanism is a possible route of glucosamine synthase catalysis. Offprint requests to: E. Borowski