Computer simulation studies of the catalytic mechanism of human aldose reductase

Citation
P. Varnai et A. Warshel, Computer simulation studies of the catalytic mechanism of human aldose reductase, J AM CHEM S, 122(16), 2000, pp. 3849-3860
Citations number
73
Categorie Soggetti
Chemistry & Analysis",Chemistry
Journal title
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY
ISSN journal
00027863 → ACNP
Volume
122
Issue
16
Year of publication
2000
Pages
3849 - 3860
Database
ISI
SICI code
0002-7863(20000426)122:16<3849:CSSOTC>2.0.ZU;2-1
Abstract
Aldose reductase, an NADPH dependent oxidoreductase, has received considera ble attention due to its possible link to diabetic and galactosemic complic ations. It is known that the catalytic reaction involves a hydride shift fr om NADPW and a proton transfer from a suitable proton donor to the carbonyl group of the substrate. However, the details of the process are still uncl ear. The present work explores the catalytic mechanism of the enzyme by usi ng the semi-microscopic protein dipoles Langevin dipoles (PDLD/S) and the e mpirical valence bond (EVB) methods. The pK(a) values of His-110 and Tyr-48 are evaluated to determine which of these two residues donates the proton in the reaction. It is found that the free energy of protonation of His-110 in its protein site is similar to 9 kcal/mol and hence the pK(a) of this r esidue is abnormally low. Consequently, His-110 is not protonated in the ac tive site of aldose reductase. On the other hand, it is found that the pK(a ) of Tyr-48 is lowered to similar to 8.5 in the active site due to the stab ilization by the unique local environment of the phenol group. We conclude that Tyr-48 acts as the proton donor in the reduction of aldehydes by aldos e reductase, while the neutral His-110 has a role in substrate binding duri ng the catalysis. To obtain a quantitative picture of the energetics of dif ferent feasible catalytic mechanisms in the protein we follow the EVE philo sophy and calibrate the potential surface of the catalytic reaction in a so lvent cage by using the relevant energetics from experiments. It is found t hat a mechanism where a proton transfer precedes the hydride transfer is un favorable in the solvent cage, relative to the alternative mechanism where the hydride transfer precedes protonation. Furthermore, our study of the re action in the actual protein environment indicates that an initial proton t ransfer step would require prohibitively high energy. Thus, the most probab le catalytic mechanism commences with the hydride shift, followed by a prot on transfer from Tyr-48. The calculations show that in water the activation barrier for the hydride shift is similar to 20 kcal/mol, which is far abov e the barrier of the subsequent proton transfer. The protein environment st abilizes the transition state of the hydride shift by similar to 3 kcal/mol and destabilizes the intermediate state by similar to 8 kcal/mol relative to the corresponding states in the water cage. This finding is consistent w ith the physiological role of the enzyme in detoxification where: it cataly zes the reduction of a wide range of carbonyl-containing substrates without particular specificity. It is argued that it may be difficult for an enzym e to both satisfy this demand and catalyze the reaction beyond the simple r ole of bringing the proton and hydride donor groups to the proximity of the substrate.