Rational pH-engineering of the thermostable xylanase based on computational model

Locating sites of amino acids related to enzyme properties is still a challenge for rational engineering. Based on the strategy that sites of amino acids can be located by dipeptides, a computational model was made for pH-related dipeptides of xylanase. According to the dipeptide model, pH of a thermostable xylanase B from Thermotoga maritima was rationally designed by locating pH-related amino acids in its sequence and structure. In agreement with expectation, the optimum pH (pH_opt) of the xylanase was improved by five amino acids substitutions: E70Q, E74Q, E77Q, G85Q and T87Q. In parallel assay reactions, at 90 °C, its pH_opt increases to 5.5 from 5.1, and its whole pH profile also shifts 0.5 units towards alkaline area; at 80 °C, the relative activity decreases very little over a wide pH range from 5.25 to 6.0. This result demonstrated that the bioinformatics model is useful for pH rational design and engineering of xylanase, a model molecular of a large family of ～10% proteins with (β/α)₈-barrel structure.