Structure and evolutionary trace-assisted screening of a residue swapping the substrate ambiguity and chiral specificity in an esterase

Abstract

Our understanding of enzymes with high substrate ambiguity remains limited because their large active sites allow substrate docking freedom to an extent that seems incompatible with stereospecificity. One possibility is that some of these enzymes evolved a set of evolutionarily fitted sequence positions that stringently allow switching substrate ambiguity and chiral specificity. To explore this hypothesis, we targeted for mutation a serine ester hydrolase (EH₃) that exhibits an impressive 71-substrate repertoire but is not stereospecific (e.e. 50%). We used structural actions and the computational evolutionary trace method to explore specificity-swapping sequence positions and hypothesized that position I244 was critical. Driven by evolutionary action analysis, this position was substituted to leucine, which together with isoleucine appears to be the amino acid most commonly present in the closest homologous sequences (max. identity, ca. 67.1%), and to phenylalanine, which appears in distant homologues. While the I244L mutation did not have any functional consequences, the I244F mutation allowed the esterase to maintain a remarkable 53-substrate range while gaining stereospecificity properties (e.e. 99.99%). These data support the possibility that some enzymes evolve sequence positions that control the substrate scope and stereospecificity. Such residues, which can be evolutionarily screened, may serve as starting points for further designing substrate-ambiguous, yet chiral-specific, enzymes that are greatly appreciated in biotechnology and synthetic chemistry.

Keywords: Crystal structure; EA, evolutionary action; ET, evolutionary trace; Eapp, apparent enantioselectivity; Esterase; Evolutionary trace; HEPES, 40 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; Ni-NTA, nickel-nitrilotriacetic acid; Promiscuity; Protein engineering; Specificity.

Graphical abstract

Fig. 1

Optimal parameters for the activity and stability of purified EH₃. (A) Temperature profile determined as follows: protein, 2 μg; [p-nitrophenyl propionate (pNPC₃)], 0.8 mM; pH, 50 mM Britton and Robinson buffer pH 8.0; T, 5–80 °C; reaction volume, 200 μl. (B) The thermal denaturation curve of EH₃ at pH 7.0 was measured by ellipticity changes at 220 nm and obtained at different temperatures. (C) The pH profile was determined as follows: protein, 2 μg; [pNPC₃], 0.8 mM; T, 30 °C; pH, 50 mM Britton and Robinson buffer from 4.0 to 10.0; reaction volume, 200 μl. Graphics were created with SigmaPlot version 14.0. The data are not fitted to any model.

Fig. 2

Non-chiral substrate specificity. The k_cat (s⁻¹) values of the EH₃, EH_3I244L and EH_3I244F variants were measured for 53 non-chiral carboxylic esters found to be hydrolyzed by any of the enzyme variants. The substrates, with the hydrophobicity (log P) and volume (ų) indicated (details in Table S2), are ranked based on hierarchical clustering according to substrate similarity profiles. For k_cat determination, calculated on a continuous pH indicator assay, the conditions were as follows: [enzyme], 0–270 µg ml⁻¹; [ester], 50 mM to ensure substrate saturation; reaction volume, 44 µl; T, 30 °C; and pH, 8.0. Abbreviations are as follows: BFPME: benzoic acid, 4-formyl-, phenylmethyl ester; BHPP: benzyl (R)-2-hydroxy-3-phenylpropionate. LogP values and molecular volume of each ester were calculated using ACD/ChemSketch 2015.2.5 and Molinspiration software, respectively. For raw data and details, see Table S2.

Fig. 3

Chiral substrate specificity. The k_cat (s⁻¹) values of the EH₃, EH_3I244L and EH_3I244F variants measured for 18 chiral carboxylic esters found to be hydrolyzed by any of the enzyme variants. Abbreviations are as follows: E(R)CHB, ethyl (R)-4-chloro-3-hydroxybutyrate; E(S)CHB, ethyl (S)-4-chloro-3-hydroxybutyrate. Figure preparation and experimental details are shown in Fig. 2. The structures of methyl-(R)-2-phenylpropanoate and methyl-(S)-2-phenylpropanoate used for soaking and investigation of chiral specificity are shown. LogP values and the molecular volume of each ester were calculated using ACD/ChemSketch 2015.2.5 and Molinspiration software, respectively. For raw data and details, see Table S2.

Fig. 4

Crystal structure of EH₃. (A) Molecular surface of the catalytic domain (wheat) with the α-helices making up the cap domain depicted as a cartoon (plum); for secondary structure numbering, see Fig. S1. The catalytic triad is shown as sticks (orange). The region comprising α1- α2 is highly flexible, and P47 acts as a hinge (green sticks). (B) Superimposition of the EH₃ subunit (plum) and its homologues, BFAE (slate, PDB ID: 1JKM) and rPPE_S159A/W187H (violet, PDB ID: 4OB6). The cap domain presents the largest differences that configure markedly divergent active sites. The folding characteristics of Est22 and Est25 are most similar to those of EH₃ and BFAE, respectively, and have been omitted for clarity. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Evolutionary trace ranks for the EH₃ protein. The analysis used 410 homologous sequences of EH₃ with sequence identity as low as 20%. The ET ranks are represented on the structure with a color scale (the most important residues are red, and the least important residues are green). While the catalytic residues were ranked within the top most important residues (S192 was 3%, D291 was 2%, and H321 was 1%), residue I244 was ranked in the top 12%, and it was the most important residue of loop 240–249 in contact with the catalytic residues. The figure was generated using the PDB structure 6SXP, PyMOL (version 1.8), ET (with the position-specific option), and the PyMOL ET viewer . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Active site of EH₃. (A) Methyl (2R)-2-phenylpropanoate and (B) methyl (2S)-2-phenylpropanoate bound at the catalytic site of EH_3S192A, showing the 2Fo-Fc electron density maps contoured at 0.9 and 0.8 σ in orange. (C) Active site channels of EH_3S192A, as calculated by CAVER , with bound methyl (2S)-2-phenylpropanoate and two glycerol molecules. The residues surrounding each cavity are shown. (D) Nearest environment and conserved binding mode of methyl (2R)-2-phenylpropanoate (slate) and methyl (2S)-2-phenylpropanoate (pale green) in the complexes; the closest distance from each substrate to the EH₃ residue is shown. The putative position of the modeled I244F mutant is shown as gray sticks. Panels C and D show the same color code as Fig. 4A. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)