Estudi computacional de la selectivitat d’un enzim hidrolasa d’epòxids

Abstract

Enzymes are the most efficient catalysts found in nature. However, natural enzymes are not sufficiently efficient to perform chemical reactions that are important for the pharmaceutical and fine chemistry industries. Therefore, there is a need to understand how they are able to perform its function in order to improve their performance. Enzymes are biodegradable, efficient, selective, and sustainable biomolecules that can become an alternative to other kinds of catalysts from the environmental point of view. In this project, the epoxide hydrolase (EH) enzyme that is able to selectively hydrolyze racemic mixtures of epoxides will be studied in detail. Using conventional methods it is difficult to observe the ring opening of the epoxide starting from racemic mixtures. A number of pharmaceutical compounds, such as alprenolol and propanolol beta-blockers, require the use of epoxide as precursors. The use of EHs in industry is promising because they allow obtaining the final product with high purity. Here, the EH from Bacillus megaterium is studied because this particular EH shows a moderate enantioselectivity towards voluminous aromatic epoxides. However, this particular enzyme presents product inhibition which limits its application. To this end, the aim of this study is to understand the conformational changes that control the efficiency of the enzyme by means of molecular dynamics simulations in three different states: apo, in the presence of substrate (using both enantiomers) and in the presence of product. Molecular dynamics simulations show that all substrates are able to access the active site of the enzyme and to adopt the proper orientation for the reaction. On the other side, all products explore the active site region with less frequency; however, when they arrive at the active site they stay there to inhibit the enzyme. Finally, the formation of a tunnel that connects the substrate entrance region with the product release region has been identified in the simulations