Enhancing the Catalytic Activity of Thermococcus kodakarensis RuBisCo via Computer-Aided Rational Design

Ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCo) is a prime target for enhancing the photosynthetic efficiency. Here, we employed bmDCA, a machine-learning implementation, to navigate the sequence space of Thermococcus kodakarensis RuBisCo (Tk-RuBisCo). After only two design-test cycles involving ∼30 variants, we isolated the double mutant SP9-Y168L-A361P, which exhibits a 1.5-fold increase in activity over the previous best mutant (SP9) under the same conditions. This gain far exceeds that achieved by conventional approaches. Notably, most bmDCA-designed double mutants display strictly additive effects, indicating that the incorporation of second-order coevolutionary information facilitates the identification of cooperative pairs. ¹H NMR results demonstrated that the Y168L-A361P mutation enhances the carboxylation rate to a much greater extent than does the oxygenation rate. Molecular-dynamics simulations revealed that Y168L and A361P─although neither contacts the substrate directly─augment binding of the carboxylation intermediate EI II through an extended interaction network that strengthens hydrogen bonds to the substrate carboxylate.