1. Introduction
Natural enzymes usually only evolve relevant catalytic performance according to their own needs. When enzymes are used in industry, methods are needed to be explored to tailor their activity towards industrially relevant substrates, and these enzymes should also be optimized towards industrial reaction conditions.1 To improve the production in industry which involves biocatalysts, increasing activity of enzymes for specific substrates is the key.2Besides, given that high temperatures in industrial processes including reaction, purification, packaging etc. provide benefits such as increased substrate solubility, improved diffusivity, decreased viscosity of the medium, and a lower risk of microbial contamination, thermostability of enzymes is another important property.3, 4
Enzymes obtained from natural recruitment and protein engineering have greatly contributed in various sets of applications. Over recent decades, the newly developed methods in the protein engineering, including directed evolution, semi-rational design, and de novodesign, have enabled to obtain numerous better enzymes for the industrial application.5-7 The catalytic activity and thermostability of many enzymes has been improved by directed evolution. By screening the triple mutant C168T/Q192H/Y7L with error-prone PCR and site-saturation, the thermostability and enzyme activity of GH11 xylanase from Aspergillus fumigatus RT-1 were improved.8 Lin et al. identified a N255D mutant by random mutagenesis with 14-fold higher activity than the wild type Horseradish Peroxidase.9 Yin et al. also constructed a mutation library with error-prone PCR. In this library, three five-linked mutants Bgl1D2, Bgl1D6 and Bgl1D20 stands out to have 2.3-2.6 times higher hydrolytic activity, while only Bgl1D2 becomes more stable. It has seven times higher thermostability, whereas Bgl1D6, Bgl1D20 shows no significant change in thermostability compared with the wild type.10
Despite of the success of the above trials, directed evolution is unfavorable when one considers the experimental resources it requires for mutants in a large library. Semi-rational design of proteins deals with this problem by introducing bioinformatics to rationally reduce the size of the mutant’s library. At present, the catalytic activity and thermostability of enzyme are mainly improved by semi-rational design. Generally, the non-conservative amino acids around the active site may be related to the catalytic performance of the enzyme, and hence the catalytic activity of enzymes can be improved by changing the non-conservative amino acids.11 Moreover, based on the correlation between thermostability of proteins and factors such as hydrophobicity, packing density,12, 13 number of disulfide bonds,14 strength of electrostatic interactions,15, 16 length of surface loops,13 conformational rigidity,17, 18 amino acid coupling patterns,19 and local structural entrop,20 bioinformatics software is generally developed to design proteins with good thermostability using proline theory,21 B-fitter,18Rosetta,22 molecular dynamics simulations23 and disulfide by design et al24.
Notably, among these semi-rational methods, mutants which have been designed to show improved thermostability, all display lower enzyme activity,25, 26 and vice versa.11 It is reasonable when we noticed the contradiction in the adjustment of protein structures between the two design strategies. Specifically, high catalytic activity was often obtained by reducing their surface hydrophobicity and hence increasing the flexibility of the structure,27-29 while enzymes with good stability were designed by making the structure more rigid, which involves enhancing their surface hydrophobicity.30-31 Thus, it is impossible to simply combine these strategies to design enzymes with both properties improved. Given that the high catalytic activity and good thermostability are both related to reduced costs, and vice versa, when one considers to optimize the total costs, which is often the case in the industry, new effective semi-rational design method is needed to be developed to simultaneously improve the catalytic activity and thermostability of enzymes.
To fill this need, here we proposed a double-screening strategy to obtain mutants with both properties improved based on compactional analysis and prediction of enzyme properties. Firstly, given that the non-conservative amino acids around the active site is related to the catalytic performance of the enzyme, mutation on the non-conserved residues in the catalytic region could bring potential higher activity. These mutants could be further screened virtually to select mutants with favorable heat stability. In this way, a relatively small library of mutants of potential simultaneous higher activity and thermostability could be further constructed and tested experimentally. Thus, not only does this strategy overcome the disadvantages of directed evolution, i.e. too large a library to do experiments, but also consider the shortcoming of the current semi-rational methods, i.e. limit the design too much on the local structure without considering the global effect. We tested this strategy on an extensively investigated enzyme in our lab, α-L-rhamnosidase.
α-L-Rhamnosidase is a glycoside hydrolase, which can effectively hydrolyze the rhamnose group at the end of most glycosides. It is widely used in the debittering of citrus juices,32, 33improving the aroma components of beverages.34, 35 To date, only 29 α-L-rhamnosidases have been biochemically characterized, and six of them, namely Bs RhaB (PDB entry 2OKX), Bt 1001 (PDB entry 3CIH), Sa Rha78A (PDB entry 3W5M), Ko Rha (PDB entry 4XHC), At Rha (PDB entry 6GSZ), and Dt Rha (PDB entry 6I60) have been illustrated in crystal structures in GH78 family.36-40 In previous studies, we cloned and expressed α-L-rhamnosidase (Rha1) from Aspergillus nigerJMU-TS528,41 which belongs to the GH78 family in the CAZy database. Simultaneously, we applied semi-conservative site-directed mutagenesis on the catalytic domain to increase the enzyme activity of Rha1.42 We also used two different methods to improve thermostability of Rha1, PoPMuSiC algorithm and lysine-arginine mutation on the surface of rRha1.43, 44 In this study, we applied the strategy mentioned above to improve the catalysis efficiency and thermostability simultaneously. Specifically, we computationally predicted the mutation in the catalytic region of Rha1 with the potential to improve the catalysis efficiency and thermostability by the aid of molecular docking and conservation degree and energy variation analysis. The predicted mutants were then expressed and validated in the enzymatic activity and thermostability. This is the first semi-rational design to improve catalysis efficiency and thermostability simultaneously of enzymes, which could be helpful to effective design α-L-rhamnosidases and other important enzymes in the industry.