Discovery and synthetic exploitation of novel epoxide hydrolases from hot springs metagenomes

The presented work has been carried out within the framework of the HotZyme Project, an EU FP7 Collaborative programme that aims to exploit a metagenomic approach in diverse hot environments to identify new thermostable hydrolases with improved performances and/or novel functionalities for different industrial processes. One of the main targets of the Hotzyme project is the discovery of new epoxide hydrolases, which represent an attractive option for the synthesis of chiral epoxides and 1,2 diols, valuable building blocks for the synthesis of several pharmaceutical compounds. Most of the epoxide hydrolases described so far belong to the ?/?-hydrolase fold superfamily and hydrolyze epoxides by the action of a catalytic triad. Recently 2 atypical epoxide hydrolases have been characterized; specifically the limonene-1,2-epoxide hydrolases (LEH) from Rhodococcus erythropolis (Re-LEH) and the LEH from Mycobacterium tuberculosis (Mt-LEH). These two enzymes catalyze the hydrolysis of limonene-1,2-epoxide to give the vicinal diol limonene-1,2-diol and differ from the classic epoxide hydrolases with respect to smaller size, the fold and the catalytic mechanism. This relatively unexplored class of epoxide hydrolase has gained our attention. We performed a bioinformatics search of the metagenomic assemblies available from the Hotzyme partners using as queries the sequences of the Re-LEH and Mt-LEH. Two ORFs showing good similarity with the Mt-LEH were found in the metagenome of samples collected from hot springs at 46°C and 55 °C at neutral pH in Russia and in China, respectively. These new LEHs were successfully cloned and expressed in a soluble form in Escherichia coli. In addition the two purified enzymes have been crystallized and their structures have been solved to high resolution using the Mt-LEH enzyme as a model.