SummaryUpon heterologous overexpression, many proteins misfold or aggregate, thus resulting in low functional yields. Human acetylcholinesterase (hAChE), an enzyme mediating synaptic transmission, is a typical case of a human protein that necessitates mammalian systems to obtain functional expression. We developed a computational strategy and designed an AChE variant bearing 51 mutations that improved core packing, surface polarity, and backbone rigidity. This variant expressed at ∼2,000-fold higher levels in E. coli compared to wild-type hAChE and exhibited 20°C higher thermostability with no change in enzymatic properties or in the active-site configuration as determined by crystallography. To demonstrate broad utility, we similarly designed four other human and bacterial proteins. Testing at most three designs per protein, we obtained enhanced stability and/or higher yields of soluble and active protein in E. coli. Our algorithm requires only a 3D structure and several dozen sequences of naturally occurring homologs, and is available at http://pross.weizmann.ac.il.

Summary Substantial improvements in enzyme activity demand multiple mutations at spatially proximal positions in the active site. Such mutations, however, often exhibit unpredictable epistatic (non-additive) effects on activity. Here, we describe FuncLib - an automated method for designing multipoint mutations at enzyme active sites using phylogenetic analysis and Rosetta design calculations. We applied FuncLib to two unrelated enzymes, a phosphotriesterase and an acetyl-CoA synthetase. All designs were active and most showed activity profiles that significantly differed from the wild type and from one another. Several dozen designs with only 3-6 active-site mutations exhibited 10-4,000-fold higher efficiencies with a range of alternative substrates, including the hydrolysis of the toxic organophosphate nerve agents soman and cyclosarin and the synthesis of butyryl-CoA. FuncLib is implemented as a web-server (http://FuncLib.weizmann.ac.il); it circumvents iterative, high-throughput screens and opens the way to design highly efficient and diverse catalytic repertoires.

Organophosphate nerve agents are extremely lethal compounds. Rapid in vivo organophosphate clearance requires bioscavenging enzymes with catalytic efficiencies of >10(7) (M(-1) min(-1)). Although serum paraoxonase (PON1) is a leading candidate for such a treatment, it hydrolyzes the toxic S(p) isomers of G-agents with very slow rates. We improved PON1's catalytic efficiency by combining random and targeted mutagenesis with high-throughput screening using fluorogenic analogs in emulsion compartments. We thereby enhanced PON1's activity toward the coumarin analog of S(p)-cyclosarin by ∼10(5)-fold. We also developed a direct screen for protection of acetylcholinesterase from inactivation by nerve agents and used it to isolate variants that degrade the toxic isomer of the coumarin analog and cyclosarin itself with k(cat)/K(M) ∼ 10(7) M(-1) min(-1). We then demonstrated the in vivo prophylactic activity of an evolved variant. These evolved variants and the newly developed screens provide the basis for engineering PON1 for prophylaxis against other G-type agents.

The ability to redesign enzymes to catalyze non-cognate chemical transformations would have wide-ranging applications. We developed a computational method for repurposing the reactivity of active site functional groups of metalloenzymes to catalyze new reactions. Using this method, we engineered a zinc-containing murine adenosine deaminase to catalyze the hydrolysis of a model organophosphate with a catalytic efficiency kcat/Km ~104 M−1s−1 after directed evolution. In the high-resolution crystal structure of the enzyme, all but one of the designed residues adopt the designed conformation. The designed enzyme efficiently catalyzed the hydrolysis of the RP-isomer of a coumarinyl analog of the nerve agent cyclosarin, and showed striking substrate selectivity for coumarinyl leaving-groups. Computational redesign of native enzyme active sites complements directed evolution methods and offers a general approach for exploring their untapped catalytic potential for new reactivities.

As a result of an author oversight in the originally published version of this article, the Rosetta script file design.xml provided in Data S1 was missing the res_type_constraint value, causing code crash. The value appears in the script filterscan.xml that is used just before design.xml. We updated the file design.xml, which now includes the res_type_constraint value reported in the main text (0.4). In addition, a file called design_text had no use and was omitted in this update. The authors apologize for the error and any inconvenience that may have resulted.

VX and its Russian (RVX) and Chinese (CVX) analogues rapidly inactivate acetylcholinesterase and are the most toxic stockpile nerve agents. These organophosphates have a thiol leaving group with a choline-like moiety and are hydrolyzed very slowly by natural enzymes. We used an integrated computational and experimental approach to increase Brevundimonas diminuta phosphotriesterase's (PTE) detoxification rate of V-agents by 5000-fold. Computational models were built of the complex between PTE and V-agents. On the basis of these models, the active site was redesigned to be complementary in shape to VX and RVX and to include favorable electrostatic interactions with their choline-like leaving group. Small libraries based on designed sequences were constructed. The libraries were screened by a direct assay for V-agent detoxification, as our initial studies showed that colorimetric surrogates fail to report the detoxification rates of the actual agents. The experimental results were fed back to improve the computational models. Overall, five rounds of iterating between experiment and model refinement led to variants that hydrolyze the toxic SP isomers of all three V-agents with kcat/KM values of up to 5 × 10(6) M(-1) min(-1) and also efficiently detoxify G-agents. These new catalysts provide the basis for broad spectrum nerve agent detoxification.

To understand the role of glycosylation in the circulation of cholinesterases, we compared the mean residence time of five tissue-derived and two recombinant cholinesterases (injected intravenously in mice) with their oligosaccharide profiles. Monosaccharide composition analysis revealed differences in the total carbohydrate, galactose, and sialic acid contents. The molar ratio of sialic acid to galactose residues on tetrameric human serum butyrylcholinesterase, recombinant human butyrylcholinesterase, and recombinant mouse acetylcholinesterase was found to be approximately 1.0. For Torpedo californica acetylcholinesterase, monomeric and tetrameric fetal bovine serum acetylcholinesterase, and equine serum butyrylcholinesterase, this ratio was approximately 0.5. However, the circulatory stability of cholinesterases could not be correlated with the sialic acid-to-galactose ratio. Fractionation of the total pool of oligosaccharides obtained after neuraminidase digestion revealed one major oligosaccharide for human serum butyrylcholinesterase and three or four major oligosaccharides in other cholinesterases. The glycans of tetrameric forms of plasma cholinesterases (human serum butyrylcholinesterase, fetal bovine serum acetylcholinesterase, and equine serum butyrylcholinesterase) clearly demonstrated a reduced heterogeneity and higher maturity compared with glycans of monomeric fetal bovine serum acetylcholinesterase, dimeric tissue-derived T. californica acetylcholinesterase, and recombinant cholinesterases. T. californica acetylcholinesterase, recombinant cholinesterases, and monomeric fetal bovine serum acetylcholinesterase showed a distinctive shorter mean residence time (44-304 min) compared with tetrameric forms of plasma cholinesterases (1902-3206 min). Differences in the pharmacokinetic parameters of cholinesterases seem to be due to the combined effect of the molecular weight and charge- and size-based heterogeneity in glycans.

No abstract