Abstract
In industry the use of enzymes as catalysts has been hampered by several limitations, such as narrow substrate scope, poor stereo selectivity, or limited stability under process conditions. During the last decade, directed evolution has emerged as a powerful technology to address all these problems. Our group has recently proposed a new and efficient method in directed evolution of enzymes, namely Iterative Saturation Mutagenesis (ISM)1. Several successful results based on ISM approach2 generated some basic questions remained to be answered, such as how efficient is ISM, what are important factors determining the success of ISM, are there few or many solutions to the same problem, etc. Finding answers to these questions are the main goal of this thesis. This thesis is divided into two main projects, exploration of evolutionary pathways of Aspergillus niger epoxide hydrolase (ANEH) in improving its thermostability and enantioselectivity.
In exploration of ANEH thermostability evolutionary pathways, libraries based on B-FIT (B-factor iterative test) were generated and resulted in moderate improvement in T5060 values. Several additional libraries based on other strategies (e.g. salt bridge introduction, helix capping, N- or C-terminal stabilization, or loop stabilization) generated no positive results. Unlike the frequently applied strategy, “pick the best and continue”, three different types of mutants as parental sequences were employed. Five pathways, using the most-improved-mutant (2 pathways), moderate-improved-mutant (2 pathways), and the least improved mutant (1 pathway) obtained after initial round of saturation mutagenesis, were generated. Results from ISM indicated that the mutants became progressively more resistant to thermal denaturation in each round; however the accomplished improvements were varied among these five pathways. Surprisingly, the final best thermo mutant originated from a disfavored pathway (with the least-improved-mutant as parental sequence), which under the traditional directed evolution procedure usually would not be chosen. The other pathways, which were generated using the most- or moderate-improved parent, however, end up giving mutants with similar moderate thermostability improvements.
Upon exploration of ANEH enantioselectivity evolutionary pathways, libraries targeting the same residues as previous CAST (combinatorial active site test)3 were generated. The fitness pathway landscape generated revealed that there are two different types of trajectories exist connecting the wild type and the best mutant. The first type demonstrates a smooth continuous decrease in the free energy, whereas the second type of pathway is the one characterized by the existence of so called “local minima”. 16 out of 24 pathways show the characteristic of first type of trajectory (green pathway) whereas the remaining 8 pathways belong to the second one (red pathway). This experimental result exhibits the high probability in ISM approach to find enzyme with improved properties (67% probability is encountered). Twelve pathways (10 out of 16 green pathways and 2 out of 8 red pathways) lead to a highly enantioselective mutants, having E-values above 80. Many of these 12 pathways originated from library D, which actually attained the lowest improvement in initial ISM round. On the contrary only one pathway stemmed from library sites which gain the highest improvement in first round of ISM, library B. Previous CAST study generated a highly enantioselective mutant, LW202, having E-values of 115. The highest improvement achieved in this study, was obtained via pathway originating from library F, pathway FBED, leading to mutant GUY-228 with an E-value of 158. This pathway also appeared to be the shortest mutational trajectory in the sense that only three ISM rounds were necessary to elevate the selectivity factor. Three additional pathways resulted in mutants better than LW202, namely pathway DBEF, DEFB, and EDBF. The productivity of the approach was also proven by the high number of improved mutants encountered in exploring all of those libraries, a total of 68 hits (31 mutants which have E-values above 80 and 37 mutants with E-values 50-80) were obtained
This thesis clearly revealed that ISM is a highly efficient approach in directed evolution as such high probability is encountered in finding fitter enzyme. Together with previous results concerning the usage of a reduced amino acid alphabet, oversampling, and library quality control 2 , ISM is rapidly advancing to be the best, fastest, and efficient method in directed evolution.
References
[1] M. T. Reetz, J. D. Carballeira, Nature Protocols, 2007, 891.
[2] M. T. Reetz, Angewandte Chemie, in press.
[3] M. T. Reetz, L. W. Wang, M. Bocola, Angewandte Chemie-International Edition, 2006, 45.
