Cargando…

Rationalization of stereospecific binding of propranolol to cytochrome P450 2D6 by free energy calculations

ABSTRACT: Cytochrome P450 2D6 is a major drug-metabolising enzyme with a wide substrate range. A single-point mutation introduced in this enzyme induces stereoselective binding of R and S-propranolol whereas the wild type has no preference. The system has previously been studied both experimentally...

Descripción completa

Detalles Bibliográficos
Autores principales: Nagy, Gabor, Oostenbrink, Chris
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Springer-Verlag 2012
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3509327/
https://www.ncbi.nlm.nih.gov/pubmed/23086294
http://dx.doi.org/10.1007/s00249-012-0865-x
Descripción
Sumario:ABSTRACT: Cytochrome P450 2D6 is a major drug-metabolising enzyme with a wide substrate range. A single-point mutation introduced in this enzyme induces stereoselective binding of R and S-propranolol whereas the wild type has no preference. The system has previously been studied both experimentally and computationally (de Graaf et al. in Eur Biophys J 36:589–599, 2007a). The in silico study reported hysteresis and significant deviations from closure of thermodynamic cycles, probably because of lack of sampling. Here, we focus on the effect of prolonged simulation time and enhanced sampling methods, such as Hamiltonian replica exchange, to reduce these problems and to improve the precision of free energy calculations. Finally we rationalize the results at a molecular level and compare data with experimental findings and previously estimated free energies. GRAPHICAL ABSTRACT: Propranolol (PPD) binds stereospecifically to cytochrome P450 2D6, if the mutation F483A is introduced, by reducing binding affinity to R-PPD. This observation has previously been studied by molecular dynamics, by use of several different thermodynamic cycles. Previous computational results suffered from internal inconsistencies, probably because of insufficient sampling. Here we use prolonged simulation time and Hamiltonian replica exchange to overcome the inconsistencies, and give a possible rationalization of the phenomenon at a molecular level. [Image: see text]