今天推送的文章发表在Green Chemistry的“Engineering the substrate preference of glucose oxidase for the enzymatic oxidation of xylose”，通讯作者为山东大学微生物技术国家重点实验室的刘国栋副教授。
Glucose oxidase (GOx, EC 1.1.3.4) is widely used in glucose monitoring, food and feed addition, and biofuel cell construction. D-xylose is the main component of hemicellulose and is the second most abundant sugar in plant biomass. Efficient conversion of d-xylose is the key to the utilization of lignocellulosic feedstock. Unlike GOx, which is widely used for the detection and biotransformation of d-glucose, enzymes that use oxygen as an electron acceptor to efficiently catalyze the oxidation of d-xylose to d-xylose-1,5-lactone have not been reported. Aspergillus niger GOx (AnGOx) is 270-fold less active against d-xylose than against d-glucose. It is common to oxidize d-xylose with excessive amounts of AnGOx, which is expensive for industrial processes. Many sugar dehydrogenases also have the ability to oxidize d-xylose at the C1 position and have been used in the development of d-xylose biosensors or in the production of d-xylate. However, the dependence of dehydrogenases on electron acceptors other than oxygen requires the use of whole cells as catalysts or the development of electron acceptor-regeneration systems, which increases the complexity and cost of their applications.
In this study, the authors engineered the substrate specificity of AnGOx using a semi-rational design approach to increase its activity against d-xylose by 5.7-fold. Molecular dynamics (MD) simulations were performed to explore the underlying mechanism of the mutant's enhanced affinity for d-xylose. In addition, the authors utilized the AnGOx double mutant to produce both sodium lignate and sodium gluconate from corn cob hydrolysate. This study deepens the understanding of GOx substrate specificity and provides an alternative method for the production of lignate from lignocellulosic biomass.
1
Screening for GOx mutants with enhanced activity against d-xylose
Since the three-dimensional structure of AnGOx, as determined by X-ray diffraction, does not contain substrates/products or their analogues. Previous studies have used MD simulations to predict dense interactions between d-glucose and amino acid residues in the AnGOx active site pocket. These interactions were later supported by superimposing the structure of AnGOx onto Aspergillus flavus glucose dehydrogenase (AfGDH) complexed with the oxidation product LGC (PDB:4YNU). Considering that AfGDH exhibits relatively high activity against d-xylose (about 20% of d-glucose activity), the authors engineered AnGOx based on its active site structure compared to AfGDH. Site-directed mutagenesis was performed on 10 residues (T110A, N217S, R176S, F215Y, Q347S, F351Y, D416H, F414L, D424E, and T331N) around the substrate binding pocket, respectively, and mutants were expressed in Saccharomyces cerevisiae. Activity measurements showed that none of the mutants showed increased activity against D-xylose. These results suggest that the underlying mechanism of substrate-specific differences between AnGOx and AfGDH may involve changes in multiple residues.
The authors then focused on residues that interact with the C6 hydroxymethyl group of d-glucose in the active site (Figure 1A). Depending on the structure superimposed on the AfGDH-LGC complex, the T110 of AnGOx can interact with the C6–OH group of d-glucose via hydrogen bonding (Figure 1B). Therefore, T110 was chosen to use the degenerate NNK codon for saturation mutagenesis. Saccharomyces cerevisiae transformants expressing AnGOx mutants were cultured with d-fructose as a carbon source to avoid any interference from d-glucose oxidation, and culture supernatants were collected for the d-xylose oxidase activity assay (Figure 1C). Transformants expressing T110V and T110I mutants have significantly increased activity against d-xylose compared to transformants expressing wild-type (WT)AnGOx. Since the expression levels of the target protein may vary between different Saccharomyces cerevisiae transformants, this may interfere with the comparison of enzyme activity. With the exception of T110, F414 is very close to the hydroxyl group at the C6 position. Therefore, a second round of saturation mutagenesis was performed on the F414 residue of the T110V mutant. Among the 96 transformants, 6 transformants had significantly higher activity against d-xylose than those expressing the parental mutant T110V. All 6 transformants are F414L mutations, making this position the same as that of AfGDH (L401).
2
Characterization of substrate-specific altered AnGOx mutants
The low expression levels of AnGOx and its mutants in Saccharomyces cerevisiae are not sufficient for its characterization and application. The authors used Pichia ferment for expression and purification instead. The oxidative activity of T110I, T110V, and T110V/F414L against d-xylose was 0.6-fold, 1.2-fold, and 5.7-fold that of WT, respectively (Table 1). In contrast, the d-xylose oxidase activity of the F414L single mutant is similar to that of WT. Therefore, the effect of the F414L mutation on the oxidative activity of d-xylose depends on the T110V mutation. The activity of T110I, T110V, and T110V/F414L on the native substrate d-glucose was reduced by 96.2%, 87.0%, and 94.4%, respectively, relative to WT, suggesting a selectivity-activity trade-off in engineering. For T110V/F214L, the specific activity of d-xylose is 1.4-fold that of glucose, while the activity of d-fructose, d-mannose, d-galactose, and l-arabinose is almost undetectable (Figure 2A). Thus, the mutant maintained stereoselectivity for the monosaccharides tested.
The kinetic parameters showed (Table 2) that the Km value of WT for d-xylose was 1.5 times that of T110V and 4.8 times that of T110V/F414L, indicating that the mutant had a higher substrate affinity for d-xylose. However, the turnover number (KCAT) of both mutants decreased significantly. It is speculated that the increased activity of T110V and T110V/F414L against d-xylose is mainly caused by their increased affinity for substrates. In addition, the kcat/km of D-glucose oxidation was reduced by both mutants compared to the wild type, which was attributed to reduced substrate affinity and reduced turnover. These results suggest that mutations expand the substrate range of AnGOx, but at the cost of reduced product yields.
3
Molecular dynamics simulations
The only difference between the structures of D-xylose and D-glucose is the lack of an outer cyclic hydroxymethyl group in D-xylose. To understand the underlying mechanism of the increased activity of d-xylose oxidase, the authors used AnGOx, T110V, and T110V/F414L complexes with d-xylose to perform all-atom MD simulations. RMSD showed that all systems except WT reached equilibrium during the 100ns simulation, indicating that these mutations promoted the stable binding of d-xylose to the enzyme. Changes in RMSF values show that mutations do not affect the flexibility of residues in the substrate binding pocket. The authors then studied the distance between the O1 atom of d-xylose and the two key histidine catalytic residues (H516 and H559). In T110V, the O1 of d-xylose is closer to H516 and in T110V/F414L, the distance from the two histidine residues is lower (Figure 3A–C). In the conformation of T110V/F414L, 91.6% O1 xylose–H516 is < 7.0Å and 86.2% O1xylene–H559 is < 9.0Å. These frequencies were significantly higher than WT (24.7% and 39.0%, respectively).
To learn more about the interaction between d-xylose and the active site, 10 residues within a distance of LGC 6Å were selected in the modeled AnGOx-LGC complex (Figure 3D) and their interaction energies with d-xylose during the simulation were calculated. Electrostatic and hydrophobic interactions with H516 and H559 are enhanced in both mutants (Figures 3E and F). In addition, the electrostatic interaction with adjacent residues of H516, such as N514 and R512, is enhanced. Conversely, both electrostatic interactions with Y68 and hydrophobic interactions with F/L414, Y68, G108, and T/V110, which dominate d-xylose binding to the active site, were reduced in the mutant (Figure 3F). The conformations produced during MD simulation also indicate different binding patterns of d-xylose in the active site. In WT, d-xylose pulls d-xylose away from the catalytic residues H516 and H559 deep in the pocket through various interactions between hydroxyl groups on its pyranose ring with Y68 and T110 and hydrophobic interaction with F414 and site B located at the opening of the substrate binding pocket along with FAD (Figure 3G). Thus, although the substrate entry channel is wide enough to accommodate d-xylose, T110 and F414, which are important for the stability of d-glucose, become detrimental to the correct localization of the substrate at a smaller size. Since the interaction of T110 and F414 is disrupted in the mutant, there is a greater chance that d-xylose will interact with site a, which allows its C1 hydroxyl group to enter the catalytic histidine residue and FAD to initiate the reaction (Figures 3H and I). However, stabilizing d-xylose only by residues at site A increases the freedom of movement of the substrate and may lead to low catalytic efficiency of the mutant (Table 2).
4
T110V/F414L co-converts d-glucose and d-xylose to algonuconic acid
When used as an additive for cement and concrete, xysulfonate (lignanate) exhibits similar properties to gluconate and has huge market potential. In addition, D-xylonic acid is an intermediate molecule in the synthesis of several derivatives, such as 1,2,4-butanetriol and 1,4-butanediol, and is considered a value-added platform for the production of biochemicals. Therefore, the authors tested the ability of T110V/F414L mutants with improved d-xylose oxidative activity to co-produce xylose and gluconate from d-xylose and d-glucose. Reactions are carried out in a bioreactor with controlled pH and continuous aeration. Catalase is added to the system to eliminate the toxic product H2O2. LC-MS analysis revealed the catalytic preparation of lignate and gluconate using a mixture of pure D-xylose and D-glucose as substrates.
Subsequently, corn cob hydrolysates containing D-xylose and D-glucose (lignocellulose derivatives) were used as substrates for the enzymatic production of aldose (Figure 5A). Unlike reactions that use pure D-xylose and D-glucose as substrates, oxidase and catalase need to be added again to maintain a continuous reaction, suggesting the presence of enzyme-inhibiting compounds in the hydrolysate. After 48 h of reaction, the conversion rates of D-xylose and D-glucose were 76.8% and 100%, respectively (Figure 5B). The results show that engineered AnGOx with an expanded substrate range has the potential to be used for the enzymatic production of algonuconic acid from lignocellulosic materials. However, due to the low turnover number, mutants may lose their advantage at high substrate concentrations (Table 2).