Engineering Candida tenuis Xylose reductase for improved utilization of NADH: antagonistic effects of multiple side chain replacements and performance of site-directed mutants under simulated in vivo conditions

Abstract

Six single- and multiple-site variants of Candida tenuis xylose reductase that were engineered to have side chain replacements in the coenzyme 2'-phosphate binding pocket were tested for NADPH versus NADH selectivity (R(sel)) in the presence of physiological reactant concentrations. The experimental R(sel) values agreed well with predictions from a kinetic mechanism describing mixed alternative coenzyme utilization. The Lys-274-->Arg and Arg-280-->His substitutions, which individually improved wild-type R(sel) 50- and 20-fold, respectively, had opposing structural effects when they were combined in a double mutant.

FIG. 1.

Coenzyme binding to wild-type C. tenuis XR. The interactions with adenosine ribose moieties in NADP⁺ (left panel) and NAD⁺ (right panel) are shown.

FIG. 2.

Experimental assay for determination of the coenzyme selectivities of C. tenuis XR and mutants of this enzyme. Values for the K274R N276D double mutant are shown. The arrows indicate the time of addition of xylose reductase (arrow a), the time that the reaction was stopped by heating (arrow b), the time of addition of formate (arrow c), and the time of addition of C. bodidinii formate dehydrogenase (arrow d).

FIG. 3.

Experimental selectivities can be correctly predicted from a kinetic model for mixed alternative coenzyme utilization. The experimental R′_sel values were obtained using equation 1. The standard deviations of R′_sel are 20% for the wild type, 9% for the K274R mutant, and <5% for all other mutants. R′_sel values were calculated using equation 2.