Directed Evolution of New and Improved Enzyme Functions Using an Evolutionary Intermediate and Multidirectional Search Joanne L. Porter,* ,† Priscilla L. S. Boon, † Tracy P. Murray, † Thomas Huber, † Charles A. Collyer, ‡ and David L. Ollis* ,† † Research School of Chemistry, Australian National University, Canberra, Australian Capital Territory 2601, Australia ‡ School of Molecular Bioscience, University of Sydney, Sydney, New South Wales 2006, Australia * S Supporting Information ABSTRACT: The ease with which enzymes can be adapted from their native roles and engineered to function specifically for industrial or commercial applications is crucial to enabling enzyme technology to advance beyond its current state. Directed evolution is a powerful tool for engineering enzymes with improved physical and catalytic properties and can be used to evolve enzymes where lack of structural information may thwart the use of rational design. In this study, we take the versatile and diverse α/β hydrolase fold framework, in the form of dienelactone hydrolase, and evolve it over three unique sequential evolutions with a total of 14 rounds of screening to generate a series of enzyme variants. The native enzyme has a low level of promiscuous activity toward p-nitrophenyl acetate but almost undetectable activity toward larger p-nitrophenyl esters. Using p-nitrophenyl acetate as an evolutionary intermediate, we have generated variants with altered specificity and catalytic activity up to 3 orders of magnitude higher than the native enzyme toward the larger nonphysiological p-nitrophenyl ester substrates. Several variants also possess increased stability resulting from the multidimensional approach to screening. Crystal structure analysis and substrate docking show how the enzyme active site changes over the course of the evolutions as either a direct or an indirect result of mutations. A lthough enzymes are remarkable catalysts, their use in many practical applications is limited by the availability of large quantities of stable enzyme with appropriate catalytic properties. Directed evolution can be usefully employed to enhance enzyme stability and alter substrate specificity to engineer tailored biocatalysts. 1-4 Mutations are introduced at the genetic level yielding a library of variant enzymes that are screened for a desired trait. It is extremely difficult to evolve an enzyme to have a completely new catalytic function; rather an existing and weak promiscuous activity can be exploited and greatly enhanced. Here we report the multistep process by which the substrate specificity of the enzyme dienelactone hydrolase (DLH) has been altered so that it can act upon ester and lipase substrates with long alkyl chains (C4-C12), while turnover of these substrates by the native enzyme is virtually undetectable. DLH is one of the many hydrolytic enzymes that has an α/β hydrolase fold. While these enzymes do not necessarily share any significant sequence similarity, they do share remarkably similar tertiary structures and a preserved arrangement of catalytic residues, suggesting possible evolution from a common ancestor. 5,6 The catalytic residues include a highly conserved triad, consisting of a nucleophile (serine, cysteine, or aspartic acid), an acid, and a histidine. The most significant variations among these enzymes are in the residues responsible for substrate binding. As well as hydrolases, enzymes exhibiting this characteristic fold can include proteases, esterases, peroxidases, lipases, and dehalogenases. 7 It appears as if nature has taken an effective catalytic core and altered the substrate binding domain to cater to a variety of substrates. We have set out to do the same with DLH from Pseudomonas knackmussii one of the simplest of known α/β hydrolase fold enzymes. 8-10 DLH is 25.5 kDa monomeric protein and is the third of four enzymes that constitute the chlorocatechol branch of the β-ketoadipate pathway. 11-14 It consists of 236 amino acids that form eight strands of β-sheet in the core of the enzyme with one buried α-helix and six others on the solvent accessible surface. 9 The catalytic triad consists of a cysteine (C123), an aspartic acid (D171), and a histidine (H202). 15 Previous studies have shown that mutation of the nucleophile to a serine (C123S) gives an enzyme with altered isomerase activity toward the physiological substrate. 16,17 The DLH C123S variant was the starting point for the present study, referred to as native DLH in the remainder of this report, since unlike the native cysteine, serine is not prone to oxidation. It has low but detectable activity toward simple substrates like p- nitrophenyl acetate and α-naphthyl acetate but almost nonexistent activity toward p-nitrophenyl butyrate and longer chain p-nitrophenyl esters, too low to detect in crude lysates during library screening. We set out to generate DLH variants that have good activity toward long chain p-nitrophenyl esters and in doing so alter the Received: October 7, 2014 Accepted: November 24, 2014 Published: November 24, 2014 Articles pubs.acs.org/acschemicalbiology © 2014 American Chemical Society 611 dx.doi.org/10.1021/cb500809f | ACS Chem. Biol. 2015, 10, 611-621