Behind substrate specificity switch in lactate dehydrogenase: Active-site loop dynamics and epistatic enhancement

Substrate specificity in lactate dehydrogenases (LDHs) controls metabolic flux and remains a central challenge in enzyme engineering, particularly when closely related substrates differ by only a single carboxylate group. Although Q102 (equivalent to Gln86 in Geobacillus stearothermophilus L-LDH) is recognized as a canonical substrate-specificity determinant, how this electrostatic switch cooperates with steric effects, conformational dynamics, and scaffold context to govern substrate recognition remains unclear. Here, we combined steady-state kinetics with molecular dynamics simulations to examine specificity switching from pyruvate to oxaloacetate in G. stearothermophilus L-LDH (GsLDH). The Q86R substitution enhanced oxaloacetate catalytic efficiency 34-fold (k_cat/K_m = 340 s^-1 mM^-1) through favorable electrostatic stabilization and increased substrate retention, but substantially reduced pyruvate activity. In contrast, I227V acted primarily through steric accommodation, broadening substrate acceptance by retaining substantial pyruvate activity (377 s^-1 mM^-1) while markedly increasing oxaloacetate activity (322 s^-1 mM^-1). The combined Q86R/I227V variant displayed positive epistasis and achieved the highest oxaloacetate efficiency among engineered LDH variants (738 s^-1 mM^-1), approaching that of native G. stearothermophilus malate dehydrogenase (GsMDH) (771 s^-1 mM^-1), indicating complementary contributions of electrostatic stabilization and steric optimization. Loop-state and productive-geometry analyses further showed that loop closure alone is permissive but insufficient for catalysis; productive turnover requires coupling between loop conformation, substrate retention, and hydride/proton-transfer geometry. Principal component analysis further revealed that the engineered mutations redistribute the conformational ensembles toward substrate-compatible productive states. Reciprocal mutations in GsMDH failed to restore efficient pyruvate catalysis, demonstrating that specificity switching is asymmetric and constrained by scaffold context. Collectively, these findings support an electrostatic-steric ensemble-selection model for LDH/MDH specificity and provide a mechanistic framework for engineering substrate specificity in Rossmann-fold dehydrogenases.