Mining unique cysteine synthetases and computational study on thoroughly eliminating feedback inhibition through tunnel engineering

L-cysteine is an essential component in pharmaceutical and agricultural industries, and synthetic biology has made strides in developing new metabolic pathways for its production, particularly in archaea with unique O-phosphoserine sulfhydrylases (OPSS) as key enzymes. In this study, we employed database mining to identify a highly catalytic activity OPSS from Acetobacterium sp. (AsOPSS). However, it was observed that the enzymatic activity of AsOPSS suffered significant feedback inhibition from the product L-cysteine, exhibiting an IC₅₀ value of merely 1.2 mM. A semi-rational design combined with tunnel analysis strategy was conducted to engineer AsOPSS. The best variant, AsOPSS^A218R was achieved, totally eliminating product inhibition without sacrificing catalytic efficiency. Molecular docking and molecular dynamic simulations indicated that the binding conformation of AsOPSS^A218R with L-cys was altered, leading to a reduced affinity between L-cysteine and the active pocket. Tunnel analysis revealed that the AsOPSS^A218R variant reshaped the landscape of the tunnel, resulting in the construction of a new tunnel. Furthermore, random acceleration molecular dynamics simulation and umbrella sampling simulation demonstrated that the novel tunnel improved the suitability for product release and effectively separated the interference between the product release and substrate binding processes. Finally, more than 45 mM of L-cysteine was produced in vitro within 2 h using the AsOPSS^A218R variant. Our findings emphasize the potential for relieving feedback inhibition by artificially generating new product release channels, while also laying an enzymatic foundation for efficient L-cysteine production.