In-silico driven engineering of enantioselectivity of a penicillin G acylase towards active pharmaceutical ingredients

Abstract Penicillin G acylase is one of the most employed enzymes in the pharmaceutical industry due to its role in the biotransformation of semi-synthetic β-lactam antibiotics. Recently, the enantioselectivity of the penicillin G acylase markedly broadened its application potential. In this study, we have evaluated effects of in-silico replacements of acyl-binding subsite residue Phe24β of the enzyme from Achromobacter sp. CCM 4824 to seven markedly smaller amino acids on its enantioselectivity towards industrially relevant compounds. Models of the two most promising mutants bearing substitutions Pheβ24Ala and Pheβ24Cys were investigated using molecular docking calculations. The Cys substitution revealed much better enantioselectivity traits with a set of seven substrates. To verify the relevance of in-silico predictions, we constructed a PGA A +  Phe24βCys mutant and determined its enantioselectivity in biocatalytic reactions. Since we experimentally confirmed all these predictions, we expanded our in-silico analysis to another set of seven compounds: the prediction suggested increased enantioselectivity for N -phenylacetyl- p- F-α-phenylalanine. The ( R )-enantiomer of this substrate is used as a building block in synthesis of important anti-cancer agent Abarelix. The enantioselectivity of PGA A +  Phe24βCys mutant towards this substrate was improved by 75% reaching E -value of about 70. Our results suggest the rapid identification of interesting replacements altering enantioselectivity using in-silico approach as the way for further expanding biotechnological application of penicillin G acylase.