Enhancing The Catalytic Activities Of Acylaminoacyl Peptidase And Galactosyltransferase By Directed Evolution

Enzymes are very useful tools for the synthesis of complex chiral molecules and hold a tremendous promise for the green and economical chemistry. However, while the number of biocatalysts has undergone an explosive growth over the past several years, there are disproportionately few examples of commercial scale applications of biocatalysts in industrial manufacturing and process development. This is mainly due to the fact that enzymes isolated directly from nature are evolved for a purpose other than being the catalysts of industrial processes, hence are usually not optimal for a certain application. Directed evolution, as a powerful tool for protein engineering, has made great achievement in enhancing various properties of enzymes since it has emerged in 1990s. By conjunction of consecutive rounds of diversification of target gene and selection for the desired catalytic activity, directed evolution mimics the key steps of natural evolution: mutagenesis, recombination and natural selection. It great accelerates the speed of molecular evolution and has the ability to obtain a protein with desired properties in months or even weeks, which usually takes millions of year in nature. More importantly, directed evolution is also a powerful tool to broaden our knowledge of protein chemistry. By analyzing the structure and property change of the mutants, directed evolution could provide valuable information of the underlying mechanism for enzyme catalysis, substrate specificity and stability.In this work, two important enzymes, acylaminoacyl peptidase from Aeropyrum pernix K1 andβ-1,3-galactosyltransferase CgtB from Campylobacter jejuni, were subjected to directed evolution. These two enzymes are extremely valuable for both research and industry applications. Different strategies of directed evolution were employed accroding to their different features. Several mutants with altered catalytic activity were obtained and fully characterized. Further kinetics and molecular modeling studies allow us have a deeper insight into the mechanism of mutation effect and the possible structural basis for the activity change.ApAPH from thermophilic archaeon Aeropyrum pernix K1 belongs to the prolyl oligopeptidase family (POP family), which represents a relative new class of serine protease. ApAPH catalyzes the removal of an N-acylated amino acid from the blocked peptides with different acyl groups at the N-terminus, and it also hydrolyzes p-nitrophenyl alkanoate esters of various alkyl chain lengths. The recombinant protein (apAPH) shows an optimal temperature at 90oC and is extremely stable. Its half life at 90oC is 18.6 hours. Recently, we have determined the crystal structure of apAPH and surprisingly found that its catalytic domain is very simliar to that of the hormone-sensitive lipase family. In addition, there is the canonical catalytic motif of lipase family (GXSXG) in the active site of apAPH, suggesting that apAPH has the potential to be evolved to an esterase.It has been shown that highly conserved residue that forms crucial structural elements of a catalytic center may be used to account for the evolutionary history of enzymes. Arg526, which is adjacent to the catalytic Asp524, is completely conserved among all APHs. In fact, there is an obviously Arg bias at this position in the whole POP family. The special location and the highly conservation suggests that the 526 residue may play a crucial role in substrate recognition and/or transition state stabilization. Here, we have chosen Arg526 for saturation mutagenesis and made a semi-rational design library of apAPH. Screening of the mutation library by high-throughput screening method resulted in mutants with up-to-6-fold increased esterase activities and decreased peptidase activities, converting apAPH from an enzyme with dual functions to an esterase with high specificity. Kinetic characterization of mutants and wild type showed that the activity changes were achieved by an increased kcat for the esterase activity and an increased Km for peptidase activity. Molecular dynamics simulation suggested that Arg526 plays crucial role in stabilizing the enzyme-substrate complex for the peptidase activity. On the other hand, bulky hydrophobic side chains at 526 position is more favorable for esterase activity since the hydrophobic side chain results in a more stable enzyme-substrate complex. This study provides the first direct evidence that the POP family and lipase family are evolutional related; and Arg526 may be an evolutional marker.Structure-based alignment indicated that the spatial position of the Glu88-Arg526 salt bridge is highly conserved among the POP family, implys that this interaction plays important role in this family. However, the effects of this inter-domain interactions on catalytic efficiency, substrate specificity, and stability are not clear yet. In order to study the functional and structural roles of this inter-domain salt bridge, a series of mutants on these two residues were created and characterized. The results had shown that E88A mutation on theβ-propeller domain almost abolished the catalysis for both of the ester and peptide substrate, indicating that Glu88 mutation is not only helps Arg526 to bind the substrate, but more importantly, it also plays crucial role in stabilizing the active site conformation. In contrast, E88A/R526V exhibited similar catalytic capability to R526V, suggesting that the Glu88 regulates the activity via an inter-domain salt bridge with the Arg526. Analysis of the pH dependence of the variants'reaction kinetics showed that the mutations make great influence on the electrostatic environment of the active site. In addition, chloride ions strongly modulated the activity of E88A mutant and the electrostatic environment of the active site. These phenomena imply that the activity change caused by E88A is closely related to the electrostatic environment near the active site. To further understand the experimental results, molecular dynamics simulation studies were undertaken with selected apAPH mutants. The results showed that the E88A mutant breaks the inter-domain salt bridge and results in the rotation of the guanidine side chain of Arg526, thus interfering with the charge-relay system of the catalytic triad. However, when sodium chloride was added to the system, the chloride anion restricts the guanidine group of Arg526 by electrostatic attraction and neutralizes the positive charge, therefore rescue the charge-relay system from disorder.In addition, we have found that this inter-domain salt bridge also plays important roles in stabilized the protein. Compared with the wild type apAPH, the half life of E88A, R526V and E88A/R526V mutants at 90℃have decreased 2-, 3- and 5-fold, respectively. And the DSC results showed that the melting temperatures of the mutants decreased 5-10℃. In addition, the Cm value of these mutants also decreased 0.4-0.6M. The observed alterations in the unfolding process and reductions in chemical stability of the mutants indicate that the evolutionary pressures promoting the conservation of the Glu88-Arg526 are due to the requirements for structural stability and efficient folding of apAPH. Therefore, this inter-domain salt bridge plays two major roles in the function of the enzyme: (1) neutralizing the positive charge of Arg526, thereby increasing the enzymatic activity and anchoring the substrate, and (2) forming the Glu88-Arg526 salt bridge, thereby stabilizing the protein. Our results provide a rare example of the regulation of catalytic activity and thermodynamic stability by a single residue on the non-catalytic domain. These new insights should be useful for the rational design of multi-domain enzymes.Glycosyltransferases (GT) are ubiquitous enzymes that catalyze the transfer reaction of a monosaccharide unit from an activated sugar phosphate (known as the "glycosyl donor") to an acceptor molecule. GTs are important biological catalysts in cellular systems which generating complex cell surface glycans including monosaccharide glycoside, oligosaccharide, or polysaccharide. GTs are now playing a key role for in vitro synthesis of oligosaccharides and has a number of important applications in pharmacal industry. However, assaying for GT activity is extremely challenging as no obvious change in fluorescence or absorbance is associated with glycosidic bond formation. In practical, the GT activity is usually determined by indirect ways, which make it difficult to be applied in a high-throughput manner. So far, the high-throughput screening method for GT activity is very rare, and screened library size is usually no more than 1000. To our knowledge, there is no ultra high-throughput screening method (>106) for GT has been not reported yet. In this thesis, we usedβ-1,3-galactosyltransferase CgtB from Campylobacter jejuni as a model enzyme to develop the first ultra high-throughput screening system for GT. CgtB transfers the galactose residue from UDP-Gal to GalNAc to form aβ-1,3 linkage. CgtB is the only enzyme that could be used in both of the synthesis of two important glycans: gangliosides GM1 and T-antigen (core 1 O-glycan). However, the native activity of CgtB is very low compared with other glycosyltransferases, which limit its utility in synthetic chemistry. In this work, we have successfully increased the activity of CgtB by directed evolution using this screening method.Two fluorescent substrates GalNAc-Bodipy and GalNAc-Coumarin were synthesized by chemical ways. Both of these substrates could transport both into and out of the cell freely by the lactose permease located on the cell membrane. When there is CgtB activity inside the cell, the enzyme transfers the galactose from UDP-Gal to the fluorescent acceptors to form a disaccharide. While the monosaccharide substrate could be removed from the cell, the fluorescent disaccharide is not recognized by lactose permease and thus trapped inside the cell. The trapped fluorescence can be detected by flow cytometry with ultra high-throughput (up to 107 per hour). By screening a random library of CgtB with 2~5 mutations per gene, 11 mutants with enhanced activity were isolate from 2×107 colonies. Sequencing results showed that 3 hot spots (K151E, L227G, E234D) occurred more than once in the mutants, suggesting that they play important roles in the increases of the activities. The mutant E14 has 5 mutation site(sN26K, C40S, K50R, K151E, L227G)and shows a kcat/Km value 8.73-fold higher than the wild type CgtB. And then, one round of DNA shuffling was performed to remove the negative mutation and optimize the combination of positive mutation sites. Screening of the shuffling library resulted in the best mutant S42, which has a kcat/Km value 16.12-fold higher than the wild type. Kinetic analysis showed that the activity increase was mainly due to the decrease of Km, which has decreased 22.34- and 82.02-fold in E14 and S42 compared to that of the wild type CgtB, respectively.We further constructed the single mutants of the"hot spots"(K151E, L227G, E234D) to understand their effects on the activity change. All of these mutants showed increased activities, but the kinetic results suggested that they have different effect on the kinetic parameters. K151E increase the kcat value twice, but L227G, E234D strongly decreased the Km value. The activities of the double mutation K151E/E234D were higher than that of both of the single mutant K151E and E234D but lower than the S42 mutant which contains all of the three mutation sites (K151E, L227G and E234D), suggesting that the effects of this mutation are additive.In this work, we have developed a novel screening method for the detection of the activity of glycosyltransferases. To our knowledge, this is the only method to date to detect the activity of this important enzyme at an ultra high-through rate. This approach can also be applied for screening the substrate specificity, stability and other properties of glycosyltransferases after simple modification, providing a powerful tool for the directed evolution of the enzymes of this important family.