Rational design of E. coli thioesterase I (TesA) for improved specificity towards lauric acid

Free fatty acids (FFA) are precursors to important oleochemicals such as alcohols, aldehydes, and amines. The current supply of FFA is limited by availability of oil crops such as canola, rapeseed, and coconut. Demand for oil crops have created tensions between food, fuel, and environmental stewardship. Therefore, microbial hosts are being developed as an alternative production route for FFA. Thioesterases are enzymes that hydrolyze the thiol bond in fatty acyl-ACP producing a FFA and its substrate specificity determines the chain length obtained. Moreover, since this step acts as both a product sink and a deregulation mechanism it is critical to have a thioesterase that is highly active to keep the inhibitory fatty acyl-ACP pool low, and specific towards the desired chain length to maximize its production. Here, we used a computational approach to engineer the highly active E. coli thioesterase I (TesA) to decrease its specificity to its best natural substrate myristoyl-ACP (14:0) and increase the specificity to lauroyl-ACP (12:0). TesA crystal structure was used to guide the algorithm to calculate the binding energy of the substrates upon configurational changes imposed to TesA. Mutants predicted were validated experimentally on E. coli and results were used to guide subsequent iterations. After three round of predictions, seven mutants with main specificity towards lauroyl-ACP were obtained. The best two mutants exhibited 200% and 171% increases in lauroyl-ACP specificity compared to wild type without affecting the enzyme activity. These successful results highlight the potential of computational approaches combined with experimental validation in protein engineering.