Ultrahigh-Throughput Directed Evolution of a Metal-Free α/β-Hydrolase with a Cys-His-Asp Triad into an Efficient Phosphotriesterase

Abstract

Finding new mechanistic solutions for biocatalytic challenges is key in the evolutionary adaptation of enzymes, as well as in devising new catalysts. The recent release of man-made substances into the environment provides a dynamic testing ground for observing biocatalytic innovation at play. Phosphate triesters, used as pesticides, have only recently been introduced into the environment, where they have no natural counterpart. Enzymes have rapidly evolved to hydrolyze phosphate triesters in response to this challenge, converging onto the same mechanistic solution, which requires bivalent cations as a cofactor for catalysis. In contrast, the previously identified metagenomic promiscuous hydrolase P91, a homologue of acetylcholinesterase, achieves slow phosphotriester hydrolysis mediated by a metal-independent Cys-His-Asp triad. Here, we probe the evolvability of this new catalytic motif by subjecting P91 to directed evolution. By combining a focused library approach with the ultrahigh throughput of droplet microfluidics, we increase P91's activity by a factor of ≈360 (to a k_cat/K_M of ≈7 × 10⁵ M^-1 s^-1) in only two rounds of evolution, rivaling the catalytic efficiencies of naturally evolved, metal-dependent phosphotriesterases. Unlike its homologue acetylcholinesterase, P91 does not suffer suicide inhibition; instead, fast dephosphorylation rates make the formation of the covalent adduct rather than its hydrolysis rate-limiting. This step is improved by directed evolution, with intermediate formation accelerated by 2 orders of magnitude. Combining focused, combinatorial libraries with the ultrahigh throughput of droplet microfluidics can be leveraged to identify and enhance mechanistic strategies that have not reached high efficiency in nature, resulting in alternative reagents with novel catalytic machineries.

Figure 1

Library strategy for the directed evolution of P91. The active site of P91 was first mutationally explored (all spheres) by the screening of single-site saturation libraries for phosphotriesterase activity in multititer plates (see Figure S2). A subset of these residues was then combined in round 1 (green spheres) and round 2 (yellow spheres) into combinatorial multiple-site saturation mutagenesis libraries, which were screened in microfluidic droplets. In round 1, droplets were incubated off-chip, whereas in round 2, droplets were incubated in a delay line on-chip to maintain selection stringency by shortening the reaction time. Blowout: Catalytic triad of P91, consisting of Cys118, His199, and Asp167. Note the two conformations of Cys118 in the structure, an inward-pointing protected and an outward-pointing active conformation (for details on cysteine conformations, see Note S1.18 in the Supporting Information).

Figure 2

Microfluidic droplet screening assay. The microfluidic screening assay consists of three steps. (1) Encapsulation of bacterial cells together with a fluorogenic substrate and lysis agent into picoliter-sized aqueous droplets that are separated with fluorinated oil. The droplets serve as miniaturized reaction vessels that link phenotype (catalytic activity indicated by fluorescence) to genotype (gene sequence encoded on a plasmid). (2) Droplet incubation can be carried out in a delay line on-chip (for the range of minutes) or off-chip (for hours to weeks). (3) The droplets can then be sorted according to their fluorescence with an excitation laser that is focused on the droplet flow along a Y-shaped junction. When surpassing a preset fluorescence threshold, a single droplet can be electrophoretically deviated by the electrodes away from the waste channel into the hit collection channel. Additional chip details are given in Figures S3 and S5.

Figure 3

Directed evolution of P91 brings about the mutant P91-R2 that reaches the catalytic efficiencies (k_cat/K_M) of many engineered and naturally evolved metal-dependent phosphotriesterases. (a) Michaelis–Menten plots for P91-WT (blue) and the evolved variant P91-R2 (red) for the hydrolysis of fluorescein di(diethylphosphate) (FDDEP, 1), measured in 50 mM N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES)–NaOH, 150 mM NaCl, 1 mM tris(2-carboxyethyl)phosphine (TCEP), pH 8.0 at 25 °C. v₀, initial reaction velocity; [E], initial enzyme concentration. Enzyme concentration was 0.2 μM for P91-WT and 0.2 nM for P91-R2. Error bars represent the standard error of three measurements of separate enzyme purifications. (b) Comparison of catalytic efficiencies of promiscuous (hashed), engineered (filled), and naturally evolved (lined) phosphotriesterases from different protein superfamilies: ABH, α/β-hydrolases; BP, β-propellers; MBL, metallo-β-lactamases; AH, amidohydrolases. P91-WT is shown in blue, and the evolved variant P91-R2 is in red. For enzymes that were evolved by directed evolution, the number of rounds is indicated in the bar. Annotation for the bar labels a–l and the respective references are detailed in Table S2. Substrates of the enzymes shown differ in their leaving groups (p-nitrophenol, fluorescein, and umbelliferone) but are all diethyl-substituted phosphotriesters. These are among the most common organophosphate insecticides in agricultural use and, due to their high accessibility compared to highly regulated warfare agents, are used in most published studies dealing with kinetic enzyme characterizations and directed evolution studies. When evolved variants showed higher activity toward a different organophosphate substrate used in the respective study, this is additionally noted in Table S2.

Figure 4

Largest rate improvements by evolution are on intermediate formation. (a) Overview of the reaction scheme postulated in analogy to esterase catalysis by hydrolases with a catalytic triad. P91 (enzyme E) hydrolyzes phosphotriesters (substrate S) via a mechanism involving the formation (phosphorylation, k₂) and breakdown (dephosphorylation, k₃) of a covalent intermediate (E–I). (b) Examples of traces of kinetic bursts with nucleophile-exchanged P91 variants (Cys118Ser). Reaction time course of P91-WT Cys118Ser (blue, enzyme concentration 10 μM) and P91-R2 Cys118Ser (red, enzyme concentration 1 μM) with 100 μM FDDEP in 50 mM HEPES–NaOH, 150 mM NaCl, pH 8.0 at 25 °C. Note the different time scales for the different variants, allowing the use of a spectrophotometric microplate reader for the wild-type-derived enzyme while requiring the use of a stopped-flow instrument for the evolved variant (with each instrument using different relative fluorescence units, RFU). (c) Individual rate constants of nucleophile-exchanged P91 variants (Cys118Ser) were determined from burst kinetics. The phosphorylation rate constant (k₂) was measured with FDDEP (1), and the dephosphorylation rate constant (k₃) was measured with paraoxon-ethyl (2) (see Note S1.15 in the Supporting Information). While k₂ changed by almost 3 orders of magnitude over evolution, k₃ remained approximately within the same order of magnitude.

Figure 5

Comparison of reaction-type specificity and leaving group preference for P91-WT and P91-R2 indicates that the main difference is intermediate formation. The relative change in catalytic efficiency (k_cat/K_M) from wild type (P91-WT) to evolved P91 (P91-R2) was measured for carboxyesterase and phosphotriesterase activity with two different leaving groups, p-nitrophenol (yellow) and fluorescein (green). Relative changes in catalytic efficiencies were calculated as (k_cat/K_M)_P91-R2/(k_cat/K_M)_P91-WT. The k_cat/K_M values underlying this plot are listed in Table S1.

Figure 6

Brønsted analysis shows that the evolved variant P91-R2 accelerates intermediate formation by improved leaving group stabilization. The Brønsted plots show the linear free-energy relationship between the rate constant of hydrolysis of paraoxon derivatives 2 and 5–10 (a) and the pK_a values of their leaving group (also Figure S1 and Table S4) for P91-WT and P91-R2. Filled dots: k_cat in s^–1. Open circles: k_cat/K_M in M^–1 s^–1. (b) Brønsted plot for P91-WT: k_cat/K_M: β_LG = −1.07, R² = 0.93; k_cat: β_LG = −0.95, R² = 0.94. (c) Brønsted plot for P91-R2: k_cat/K_M: β_LG = −0.55, R² = 0.80; k_cat: β_LG = −0.68, R² = 0.94. As the slope of the linear fits (β_LG) is very similar for both kinetic parameters (k_cat and k_cat/K_M), intermediate formation (k₂) must be the rate-limiting step. The lower β_LG of P91-R2 (≈−0.6) compared to P91-WT (≈−1) indicates that the evolved variant has adapted to offset the charge which accumulates on the leaving group during the transition state. Details on the underlying kinetic measurements can be found in Figure S15 and Tables S5 and S6.

Figure 7

Rationalization of the effects of the identified mutations. (a) Positions identified in the preliminary mutational scanning shown in the structure of P91-WT (blue). The catalytic triad is shown in magenta, and positions chosen for mutation in the library are shown in green. Hydrogen bonds and putative new polar interactions are highlighted as dotted lines. (b) Mutations in P91-R2 (red) are shown in a structure created by AlphaFold2/ColabFold.^, P91 hydrolyses phosphotriesters via a mechanism involving the (c) formation (phosphorylation at rate constant k₂) and (d) breakdown (dephosphorylation with rate constant k₃) of a covalent intermediate. Both transition states have a trigonal–bipyramidal geometry around the phosphorus center during the formation of the covalent adduct and the breakdown of the covalent intermediate. His199 and Asp167 form a charge relay system, with His199 acting as a general acid/base catalyst. Ala38 contributes with its backbone amide to stabilization of the oxyanion and forms an oxyanion hole. The phosphate moiety is shown in blue, and the leaving group is shown in red. Leaving group charge offset and deprotonation of the water molecule are achieved with the participation of as yet unidentified residues of the enzyme, depicted as “HA/A^–” in green.