1. Introduction
Engineering of proteins to improve their thermostability is an area of intensive research. Robust enzymes are required to withstand industrial process conditions for developing biobricks for synthetic biology. Rational design approaches to engineering thermostability require structural information, but even with advanced computational methods it is challenging to predict with sufficient precision to anticipate the results of a given mutation. Directed evolution is an alternative when structures are unavailable but it requires extensive experiments for the screening of mutant libraries. Novel enzymes has been fueled by the diversification of enzymes for billions of years evolution[1]. Therefore, new design method based on the co-evolution of amino acid residues is proposed [2,3].
D-amino acids are key intermediates for pharmaceuticals and agrochemicals [4]. L-amino acid dehydrogenases dependent on NAD(P)H have been wildly applied for the effectively synthesis of L-amino acids. By contrast, wide application of D-amino acid dehydrogenase (DAADH) has not been achieved due to few DAADHs have been explored from natural resources. Furthermore, compared with L-amino acid dehydrogenases, DAADHs are unstable and low reaction rates[5]. Thermostability is the bottleneck for DAADHs. Therefore, creating a stable and high active DAADH and demonstrated that it is applicable in industry is a challenge[6]. Designed mutations based on the structures analysis ofmeso -DAPDH and DAADH to enhance enzyme activity[7]. Cheng et al.[8] studied the site-saturation mutagenesis of DAPDH fromSymbiobacterium thermophilum   (StDAPDH) and obtain a double-site mutant W121L-H227I, which showed improved activities towards various sterically bulky 2-keto acids.
The development of in sillco design of enzyme reduce the experimental costs and time[9].These interfacial interactions are generally regulated by the key residues on the surface of motif of enzyme. Zhu et al.[10] studied the modification of subunit interface of Leifsonia  alcohol dehydrogenase (LnADH) and yielded the mutant T100R-S148I with thermostability with and catalytic efficiency enhancement. Boucher et al. [11] investigated the structural basis for the hyperthermostability of a FN3-like protein domain from Thermoanaerobacter tengcongensis by molecular dynamics simulations. Three arginine residues (R23, R25, and R72), which stabilize the protein by salt bridges were identified and mutated into alanine, which reduced melting temperature from 97.5 to 22 °C.
How to quickly define the key interficial amino acid residues which greatly effect stabilitiy of enzyme is crticial for protein engineering. Co-evolution analysis can narrow down the selection range of mutated sites, and enhance the successful rate based on the natural evolved selection. Conservation and co-evolution analysis indicated the interfacial interaction of domains and subunits are key factors behind enzyme’s stability [12,13]. A limited number of co-evolved residues located at the oligomerization interfaces, which are expected to highlight most key sites involved in their thermostability[14]. Wang et al.[15] redesigned the interactions of the subunit interfaces of glycerol dehydratase from Klebsiella pneumoniae. Based on evolutionary analysis, three co-evolved residue pairs were selected for mutation to enhance subunits interfacial interactions. The design of salt bridges are considered to be an easy and efficient ways to assembly motifs, domains and subunits. Kleiner et al.[16] found that the interdimeric interface of methionine S-adenosyltransferase fromUreaplasma urealiticum is subject to rapid evolutionary changes. The indel loop affects the release pathway for meso -DAP and pyruvic acid and meso -DAP[17] . Insights from co-evolution perspective can define the key residues the interaction hotspots across the interfacial of subunits[18]. Therefore, the novel strategy combined with co-evolution and interfacial interaction network analysis is developed.
D-amino acid dehydrogenase (DAADH) from Ureibacillus thermosphaericus was selected for thermostability enhancement study (Scheme 1). Redesign of interfacial interactions of domains and subunits by integrating interfacial interaction network analysis with co-evolution analysis (Fig. 1). Several novel NADPH-dependent DAADHs were designed by multi-point mutations for interfacial interaction enhancement. Experiments were carried out to verify the design and reveal the mechanism of enhancement of thermostability.
Scheme.1 Reductive amination of phenylpyruvate catalyzed by DAADH