R.B. Leveson-Gower

The vanadium-dependent haloperoxidase (VHPO) class of enzymes are discovered as an effective biocatalyst platform for nitrogen-halogen (N−X) bond formation. VHPOs perform selective halogenation on a range of substituted benzamidine hydrochlorides to produce the corresponding N’-halobenzimidamides and this technology is applied to 1,2,4-oxadiazole synthesis.

Abstract

Nitrogen-containing compounds are valuable synthetic intermediates and targets in nearly every chemical industry. While methods for nitrogen-carbon and nitrogen-heteroatom bond formation have primarily relied on nucleophilic nitrogen atom reactivity, molecules containing nitrogen-halogen bonds allow for electrophilic or radical reactivity modes at the nitrogen center. Despite the growing synthetic utility of nitrogen-halogen bond-containing compounds, selective catalytic strategies for their synthesis are largely underexplored. We recently discovered that the vanadium-dependent haloperoxidase (VHPO) class of enzymes are a suitable biocatalyst platform for nitrogen-halogen bond formation. Herein, we show that VHPOs perform selective halogenation of a range of substituted benzamidine hydrochlorides to produce the corresponding N’-halobenzimidamides. This biocatalytic platform is applied to the synthesis of 1,2,4-oxadiazoles from the corresponding N-acylbenzamidines in high yield and with excellent chemoselectivity. Finally, the synthetic applicability of this biotechnology is demonstrated in an extension to nitrogen-nitrogen bond formation and the chemoenzymatic synthesis of the Duchenne muscular dystrophy drug, ataluren.

Nature Chemistry, Published online: 28 August 2024; doi:10.1038/s41557-024-01608-8

Catalytic asymmetric radical dearomatization has remained a daunting task due to the challenges in exerting stereocontrol over highly reactive radical intermediates. Now, using metalloredox biocatalysis, new-to-nature radical dearomatases P450rad1–P450rad5 have been engineered to facilitate asymmetric dearomatization of a broad spectrum of aromatic substrates, including indoles, pyrroles and phenols.

The high-yielding fungal peroxygenase, MroUPO-TN, catalyzes the regioselective remote-site functionalization of hydrocarbons, providing useful and new synthetic building blocks in an economical and sustainable process. Bromocyclooctane affords 4-bromocyclooctanone in 80% yield, whereas 1-haloalkanes are transformed into the corresponding ω-1 haloketones. Deuterium labeling and ¹⁸O-labeling experiments show that the selectivity for 4-halocyclooctanones derives from selective, remote site oxygenation.

Abstract

We describe the discovery of an unspecific peroxygenase (UPO) variant that catalyzes the remote-site functionalization of halogenated and unsaturated hydrocarbons with high catalytic site-specificity. UPOs are fungal heme-thiolate biocatalysts with wide-ranging oxidative activities, including C─H bond oxygenation, usually with limited regioselectivity. We describe here a wild-type MroUPO, newly isolated in high yield from a previously uncharacterized strain of Marasmius rotula. This variant, MroUPO-TN, catalyzes the selective oxygenation of a range of haloalkanes, cyclic haloalkanes and cyclic olefins to generate useful remote-site haloketones. The regioselectivity for eight-membered rings reaches 99% with significant enantiomeric excess. Mechanistic studies performed with deuterated substrates and ¹⁸O-labeling experiments have revealed a synergy between intrinsic substrate properties and the highly aliphatic, heme active site. The observed selectivity offers routes to new and useful, bifunctional synthons and pharmacophores, thus providing practical ways to employ these natural and environmentally benign biocatalysts.

Biocatalysts are prized for their selectivity, tunability, and their compatibility with environmentally-friendly reaction conditions. Introduction of unnatural cofactors opens the door to new reactive enzymatic intermediates, and in turn, the possibility for new biochemical reactions. In the present study, we employed a de novo biosynthesized, non-natural cofactor, cobalt protoporphyrin IX,1 to generate a mono-nuclear cobalt hydride intermediate in the active site of a common P450 scaffold. We show that this cobalt hydride intermediate engages in metal-hydrogen atom transfer (M-HAT) reactivity, a well-studied and highly utilized reactivity pattern in synthetic chemistry,2 but which is not known to operate in nature. Because the required cofactor is fully biosynthesized and incorporated into proteins in vivo, the catalysts are highly amendable to directed evolution. We leveraged the ability to quickly access these new artificial metalloenzymes with a colorimetric screen and evolved new variants for M-HAT-mediated deallylation of phenols. We showed how common silanes have a propensity to hydrolysis that can be overcome with directed evolution by accelerating metal-hydride formation from a bulky, water stable silane. During this evolution, we discovered that variants were catalyzing HAT to the colorimetric probe itself, resulting in a unique reductive dearomatization reaction. This radical process occurs efficiently under aerobic conditions and reactions of this type have not been observed previously. These discoveries demonstrate how the tunability of biocatalytic systems can enable innovations in synthetic chemistry. We anticipate that further engineering and study of M-HAT biocatalysts will prompt new questions about hydrogen atom transfer reactivity and enable the adoption of biocatalysts for numerous synthetically useful transformations.

The unspecific peroxygenase rAaeUPO-PaDa-I-H catalyses the oxidation of Z- and E-allylic alcohols with complementary selectivity, giving epoxide and aldehyde/acid products, respectively. Both reactions were performed on a preparative scale with yields of up to 80 %, and the epoxidations proceed with excellent enantioselectivity (99 % ee).

Abstract

Unspecific peroxygenases (UPOs) catalyze the selective oxygenation of organic substrates using only hydrogen peroxide as the external oxidant. The PaDa−I variant of the UPO from Agrocybe aegerita catalyses the oxidation of Z- and E-allylic alcohols with complementary selectivity, giving epoxide and carboxylic acid/aldehyde products respectively. Both reactions can be performed on preparative scale with isolated yields up to 80 %, and the epoxidations proceed with excellent enantioselectivity (>99 % ee). The divergent reactions can also be used to transform E/Z mixtures of allylic alcohols, enabling both product series to be isolated from a single reaction. The utility of the epoxidation method is exemplified in the total synthesis of both enantiomers of the insect pheromone disparlure, including a highly enantioselective gram-scale transformation. These reactions provide further evidence for the potential of UPOs as catalysts for the scalable preparation of important oxygenated intermediates.

Bipyridyl-l-alanine (BpyAla) is a highly sought, metal-chelating non-canonical amino acid. However, its high cost has hindered many applications. Here, we develop a chemoenzymatic approach to efficiently construct BpyAla. This strategy is general to many types of metal-chelating amino acid, and we show that a newly available BpyAla analog can be incorporated into proteins using existing amber suppression technology.

Abstract

Metal-chelating noncanonical amino acids (ncAAs) are uniquely functional building blocks for proteins, peptide catalysts, and small molecule sensors. However, catalytic asymmetric approaches to synthesizing these molecules are hindered by their functional group variability and intrinsic propensity to ligate metals. In particular, bipyridyl-l-alanine (BpyAla) is a highly sought ncAA, but its complex, inefficient syntheses have limited utility. Here, we develop a chemoenzymatic approach to efficiently construct BpyAla. Three enzymes that can be produced in high titer together react to convert Gly and an aldehyde into the corresponding β-hydroxy ncAA, which is subsequently deoxygenated. We explore approaches to synthesizing biaryl aldehydes and show how the three-enzymatic cascade can access a range of α-amino acids with bulky side chains, including a variety of metal-chelating amino acids. We show that newly accessible BpyAla analogues are compatible with existing amber suppression technology, which will enable future merging of traditional synthetic and biosynthetic approaches to tuning metal reactivity.

C-H functionalization provides unparalleled benefits for the late-stage modification of complex molecules. In recent years, photocatalysis has progressed significantly due to its mild conditions, sustainability and selectivities. Current mechanisms for photocatalytic C-H functionalization primarily involve direct single-electron oxidation and direct hydrogen atom transfer (d-HAT) by radicals, both of which are invalid for the functionalization of highly electron-deficient C(sp3)-H bonds. Here, we developed a cooperative photoenzymatic system consisting of a flavin-dependent ene-reductase (ER) and an exogenous photocatalyst fluorescein (FI) to achieve enantiodivergent functionalization of electron-deficient C(sp3)-H bonds. Mechanistic studies revealed a novel pathway for radical intermediate formation via excited-state FI*-induced single-electron oxidation of carbanions under alkaline conditions. The overall catalytic efficiency is enhanced by the electron transfer (ET) between FMNox and FI-*, while the stereoselectivity is controlled by ERs through enantioselective hydrogen atom transfer (HAT). This suggests that thermodynamically unfavorable single-electron oxidation of highly electron-deficient species can occur through ion-to-radical conversion. We anticipate that our novel mode of photoenzymatic catalysis will inspire new pathways for generating free radical intermediates and foster innovative strategies for achieving photoenzymatic new-to-nature reactions. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=75 SRC="FIGDIR/small/627421v1_ufig1.gif" ALT="Figure 1"> View larger version (16K): org.highwire.dtl.DTLVardef@d74f1eorg.highwire.dtl.DTLVardef@18b5cdaorg.highwire.dtl.DTLVardef@7d0fe1org.highwire.dtl.DTLVardef@9d9dda_HPS_FORMAT_FIGEXP M_FIG C_FIG

The Nobel Prize for Chemistry 2024 was jointly awarded to David Baker for computational protein design and to Demis Hassabis and John Jumper for protein structure prediction. This highlight showcases the impact of the Nobel prize laureates’ contributions and summarizes the history, state of the art, applications and future directions of these methods.

Abstract

The ability to predict and design protein structures has led to numerous applications in medicine, diagnostics and sustainable chemical manufacture. In addition, the wealth of predicted protein structures has advanced our understanding of how life's molecules function and interact. Honouring the work that has fundamentally changed the way scientists research and engineer proteins, the Nobel Prize in Chemistry in 2024 was awarded to David Baker for computational protein design and jointly to Demis Hassabis and John Jumper, who developed AlphaFold for machine-learning-based protein structure prediction. Here, we highlight notable contributions to the development of these computational tools and their importance for the design of functional proteins that are applied in organic synthesis. Notably, both technologies have the potential to impact drug discovery as any therapeutic protein target can now be modelled, allowing the de novo design of peptide binders and the identification of small molecule ligands through in silico docking of large compound libraries. Looking ahead, we highlight future research directions in protein engineering, medicinal chemistry and material design that are enabled by this transformative shift in protein science.

Science Advances, Volume 10, Issue 48, November 2024.

The HG3-to-HG3.17 evolutionary pathway enhances Kemp eliminase activity by refining active-site interactions and enhancing conformational dynamics. Molecular dynamics simulations reveal that a water-mediated network of non-covalent interactions supports a catalytically competent conformation. Gln50 contributes to oxyanion stabilization, while Trp44 regulates Gln50 flexibility, influencing active-site preorganization and catalytic efficiency for the most evolved variant HG3.17.

Abstract

The base-promoted Kemp elimination reaction has been used as a model system for enzyme design. Among the multiple computationally designed and evolved Kemp eliminases generated along the years, the HG3-to-HG3.17 evolutionary trajectory is particularly interesting due to the high catalytic efficiency of HG3.17 and the debated role of glutamine 50 (Gln50) as potential oxyanion stabilizer. This study aims to elucidate the structural and dynamic changes along the evolutionary pathway from HG3 to HG3.17 that contribute to improved catalytic efficiency. In particular, we evaluate key variants along the HG3 evolutionary trajectory via molecular dynamics simulations coupled to non-covalent interactions and water analysis. Our computational study indicates that HG3.17 can adopt a catalytically competent conformation promoted by a water-mediated network of non-covalent interactions, in which aspartate 127 (Asp127) is properly positioned for proton abstraction and Gln50 and to some extent mutation cysteine 84 (Cys84) contribute to oxyanion stabilization. We find that HG3.17 exhibits a rather high flexibility of Gln50, which is regulated by the conformation adopted by the active site residue tryptophan 44 (Trp44). This interplay between Gln50 and Trp44 positioning induced by distal active site mutations affects the water-mediated network of non-covalent interactions, Gln50 preorganization, and water content of the active site pocket.

Imagine a computer capable of solving currently unsolvable problems. Quantum computing leverages the principles of quantum mechanics to tackle complex challenges that would take classical computers centuries to complete. In this Commentary, we explore the current state of quantum computing development and how it will revolutionise the way we discover and design biocatalysts for practical use.

Nature, Published online: 20 November 2024; doi:10.1038/s41586-024-08179-1

Photocatalysis at 40–60 °C is shown to be able to defluorinate perfluoroalkyl substances, known as ‘forever chemicals’, allowing the recycling of fluorine in polyfluoroalkyl and perfluoroalkyl substances as inorganic fluoride salt.

Artificial metalloenzymes (ArMs) are an attractive approach to achieving “new to nature” biocatalytic transformations. In this work, a novel copper dependent artificial Michaelase (Cu_Michaelase) was created that comprises a genetically encoded copper-binding ligand, i.e. (2,2-bipyridin-5-yl)alanine (BpyA). For the first time, such an ArM containing a non-canonical metal-binding amino acid was successfully optimized through directed evolution. The ArM was applied in the copper-catalyzed asymmetric Michael addition of 2-acetyl azaarenes to nitroalkenes, yielding various γ-nitro butyric acid derivatives, which are precursors for a range high-value-added pharmaceutically relevant compounds, with good yields and high enantioselectivities (up to >99% yield and 99% ee). The evolved ArM could even be used in a preparative scale synthesis and the products were further derivatized. X-ray crystal structure analysis confirms the binding of the Cu(II) ions to the BpyA residues and shows that, in principle, there is sufficient space for the 2-acetyl azaarene substrates to coordinate. Kinetic studies showed that the increased catalytic efficiency of the evolved enzyme is due to improvements in apparent KM for both substrates and a notable threefold increase in apparent kcat for the 2-acetyl azaarene. This work illustrates the potential of artificial metalloenzymes exploiting non-canonical metal-binding ligands for new-to-nature biocatalysis.

This study presents the biochemical and structural characterization of CyanoPOX, a heme-containing peroxidase from Cyanobacterium sp. TDX16. Expressed in E. coli, the enzyme exhibits properties similar to bovine lactoperoxidase despite low sequence identity. CyanoPOX demonstrates optimal activity in slightly acidic conditions and efficiently oxidizes various substrates. Structural analysis reveals a non-covalently bound b-type heme cofactor and a conserved active site pocket.

Abstract

Peroxidases belong to a group of enzymes that are widely found in animals, plants and microorganisms. These enzymes are effective biocatalysts for a wide range of oxidations on various substrates. This work presents a biochemical and structural characterization of a novel heme-containing peroxidase from Cyanobacterium sp. TDX16, CyanoPOX. This cyanobacterial enzyme was successfully overexpressed in Escherichia coli as a soluble, heme-containing monomeric enzyme. Although CyanoPOX shares relatively low sequence identity (37 %) with bovine lactoperoxidase, it displays comparable biochemical properties. CyanoPOX is most stable and active in slightly acidic conditions (pH 6–6.5) and moderately thermostable (melting temperature around 48 °C). Several compounds that are typical substrates for mammalian lactoperoxidases were tested to establish the catalytic potential of CyanoPOX. Potassium iodide showed the highest catalytic efficiency (126 mM⁻¹ s⁻¹), while various aromatic compounds were also readily converted. Structural elucidation of CyanoPOX confirmed the presence of a non-covalently bound b-type heme cofactor that is situated in the central core of the protein. Except for a highly similar overall structure, CyanoPOX also has a conserved active site pocket when compared with mammalian lactoperoxidases. Due to its catalytic properties and high expression in a bacterial host, this newly discovered peroxidase shows promise for applications.

Fungal unspecific peroxygenases (UPOs) are known for their biocatalytic potential but extensive engineering campaigns are challenging. In this work, we used an easy-to-express bacterial tyrosine hydroxylase as a template to broaden its peroxygenase activities by computational structure-assisted redesign. An engineered variant was generated capable of indigo production and enantioselective sulfoxidation demonstrating that alternative peroxygenases can be explored for biocatalytic applications.

Abstract

Fungal unspecific peroxygenases (UPOs, EC 1.11.2.1) are known for their unique biocatalytic properties and have been extensively studied since their discovery more than two decades ago. UPOs have found a wide range of applications in various fields, such as bioremediation, drug metabolism, and synthetic organic chemistry. Nonetheless, the production of UPOs seems limited to fungal expression systems which makes enzyme engineering challenging. It would be attractive to have a peroxygenase available that is easily expressed in a bacterial host. Here, we report on the expression and engineering of a bacterial peroxygenase, tyrosine hydroxylase (TyrH, EC 1.11.2.6). With only two rounds of computer-aided engineering, we have identified distal and active-site mutations that improve enzymatic activities and substrate scope, respectively. The combination of these mutations in a single variant led to an effective peroxygenase, capable of indigo production and enantioselective sulfoxidation and which can be overexpressed in Escherichia coli.

BackgroundBiotechnological applications are steadily growing and have become an important tool to reinvent the synthesis of chemicals and pharmaceuticals for lower dependence on fossil resources. In order to sustain this progression, new feedstocks for biotechnological hosts have to be explored. One-carbon (C1-)compounds, including formate, derived from CO2 or organic waste are accessible in large quantities with renewable energy, making them promising candidates. Previous studies showed that introduction of the formate assimilation machinery from Methylorubrum extorquens into Escherichia coli allows assimilation of formate into the established biotechnological host. Applying this established route for formate assimilation, we here investigated utilisation of formate for the production of value-added building blocks in E. coli using S-adenosylmethionine (SAM)-dependent methyltransferases. ResultsWe first analysed methylation activity in E. coli BL21 with a two-vector system to produce three different methyltransferases together with the formate assimilation machinery. Feeding isotopically labelled formate, products with 51 - 81% 13C-labelling could be obtained by maintaining in vivo methylation activity. Focussing on improvement of in vivo methylation, we analysed two further E. coli strains with an engineered C1-metabolism and, following condition optimisation, achieved a doubled methylation activity with a share of more than 70% formate-derived methyl groups. ConclusionsThis study demonstrates the efficient transformation of formate into methyl groups in E. coli. Our findings support that feeding formate can improve the availability of usable C1-compounds and, as a result, increase in vivo methylation activity in engineered E. coli.

Nature Catalysis, Published online: 03 December 2024; doi:10.1038/s41929-024-01258-6

The capability and importance of computational methods in organic chemistry is steadily increasing. This Review provides an overview of computational methods for the design of asymmetric catalysts, with the aim of avoiding specialist computational language to make the field more accessible to experimental chemists.

Enzymes are the quintessential green catalysts, but realizing their full potential for biotechnology typically requires improvement of their biomolecular properties. Catalysis enhancement, however, is often accompanied by impaired stability. Here, we show how the interplay between activity and stability in enzyme optimization can be efficiently addressed by coupling two recently proposed methodologies for guiding directed evolution. We first identify catalytic hotspots from chemical shift perturbations induced by transition-state-analogue binding and then use computational/phylogenetic design (FuncLib) to predict stabilizing combinations of mutations at sets of such hotspots. We test this approach on a previously designed de novo Kemp eliminase, which is already highly optimized in terms of both activity and stability. Most tested variants displayed substantially increased denaturation temperatures and purification yields. Notably, our most efficient engineered variant shows a ~3-fold enhancement in activity (kcat 1700 s-1, kcat/KM 4.3·105 M-1s-1) from an already heavily optimized starting variant, resulting in the most proficient proton-abstraction Kemp eliminase designed to date, with a catalytic efficiency on a par with naturally occurring enzymes. Molecular simulations pinpoint the origin of this catalytic enhancement as being due to the progressive elimination of a catalytically inefficient substrate conformation that is present in the original design. Remarkably, interaction network analysis identifies a significant fraction of catalytic hot-spots, thus providing a computational tool which we show to be useful even for natural-enzyme engineering. Overall, our work showcases the power of dynamically guided enzyme engineering as a design principle for obtaining novel biocatalysts with tailored physicochemical properties, towards even anthropogenic reactions.

Mushroom is a rich source of enzymes and therefore is a promising biocatalyst for organic synthesis. This review highlights key advances in mushroom-mediated redox reactions using whole cells, purified enzymes or recombinant enzymes, including the reduction of carbonyl compounds and carboxylic acids, oxidation of C−H bonds, epoxidation of olefins and oxidative cleavage of alkenes.

Abstract

The application of biocatalysts in organic synthesis has grown significantly in recent years, and both academia and industry are continuously searching for novel biocatalysts capable of performing challenging chemical reactions. Mushrooms are a rich source of ligninolytic and secondary metabolite biosynthetic enzymes, and therefore were considered promising biocatalysts for organic synthesis. This review focuses on the broad utilization potential of mushroom-based biocatalysts and highlights key advances in mushroom-mediated redox reactions. It mainly includes the reduction of ketones and carboxylic acids, hydroxylation of aromatic and aliphatic compounds, epoxidation of olefins, oxidative cleavage of alkenes, and other uncommon reactions catalyzed by the whole cells or purified enzymes of mushroom origin. Overall, a comprehensive overview of the applications of mushrooms as biocatalysts in organic synthesis is provided, which puts this versatile microorganism in the spotlight of further research.

A new class of artificial metalloenzymes containing genetically encoded noble-metal-binding sites featuring a non-canonical thiophenol-based amino acid, which serves as an excellent soft ligand for binding various noble metals, was developed. The corresponding gold(I) enzyme was characterised, confirming gold binding to the thiophenol, and successfully applied in catalytic hydroamination reactions.

Abstract

Incorporating noble metals in artificial metalloenzymes (ArMs) is challenging due to the lack of suitable soft coordinating ligands among natural amino acids. We present a new class of ArMs featuring a genetically encoded noble-metal-binding site based on a non-canonical thiophenol-based amino acid, 4-mercaptophenylalanine (pSHF), incorporated in the transcriptional regulator LmrR through stop codon suppression. We demonstrate that pSHF is an excellent ligand for noble metals in their low oxidation states. The corresponding gold(I) enzyme was characterised by mass spectrometry, UV/Vis spectroscopy and X-ray crystallography and successfully catalysed hydroamination reactions of 2-ethynyl anilines with turnover numbers over 50. Interestingly, two equivalents of gold(I) per protein dimer proved to be required for activity. Up to 98 % regioselectivity in the hydroamination of an ethynylphenylurea substrate was observed, yielding the corresponding phenyl-dihydroquinazolinone product, consistent with a π-activation mechanism by single gold centres. The ArM was optimized by site saturation mutagenesis using an on-bead screening protocol. This resulted in a single mutant that showed higher activity for one class of substrates. This work brings the power of noble-metal catalysis into the realm of enzyme engineering and establishes thiophenols as alternative ligands for noble metals, providing new opportunities in coordination chemistry and catalysis.

Trace iridium modified NiFe phosphide (Ir-NiFeP_x/CC) was in-situ grown on carbon cloth by a simple wet chemical method, followed by a phosphorization post treatment at a relatively low temperature. The best-performing electrocatalyst, Ir₂-NiFeP_x/CC, shows excellent OER activity, with an overpotential of 190 mV for alkaline OER at 10 mA cm⁻² and an ideal long-term stability over 90 h.

Abstract

Cost-effective electrocatalysts is a key constituent to establish the balance of cost and catalytic efficiency for oxygen evolution reaction (OER) via water electrolysis in the area of energy conversion and storage. NiFe phosphide decorated with trace amount of iridium (Ir) species in-situ grown on carbon cloth was prepared by a facile wet chemistry approach followed by a phosphorization post-treatment at a relative low temperature. The optimal electrocatalyst, Ir₂-NiFeP_x/CC, exhibits excellent OER activity, with an low overpotential of 190 mV at 10 mA cm⁻² for alkaline OER, and a desirable long-term durability over 90 h. The outstanding OER performance stems from the structural evolution via phosphorization process, Ir decoration with more high-valence stated Ir⁴⁺ species, and tight connection between individual components of the electrode, which gives rise to the strong activity to the active sites and faster reaction kinetics in the alkaline OER process. Mover, the Ir loading was as low as approximately ~1.7 wt % (0.29 mg cm⁻²), showing promissing propective in cost-effective OER.

Science Advances, Volume 10, Issue 40, October 2024.

The evolution of a promiscuous enzyme for its various activities often results in catalytically specialized variants. This is an important natural mechanism to ensure the proper functioning of natural metabolic networks. It also acts as both a curse and blessing for enzyme engineers, where enzymes that have undergone directed evolution may exhibit exquisite selectivity at the expense of a diminished overall catalytic repertoire. We previously performed two independent directed evolution campaigns on a promiscuous artificial enzyme that leverages the unique properties of a non-canonical amino acid (ncAA) para- aminophenylalanine (pAF) as catalytic residue, resulting in two evolved variants which are both catalytically specialized. Here, we combine mutagenesis, crystallography and computation to reveal the molecular basis of the specialization phenomenon. In one evolved variant, an unexpected change in quaternary structure biases substrate dynamics to promote enantioselective catalysis, whilst the other demonstrates synergistic cooperation between natural side chains and the pAF residue to form semi-synthetic catalytic machinery. Our analysis provides valuable insights for the future engineering of effective artificial enzymes which employ either the widely used LmrR scaffold or pAF catalytic residue.

Enzymes that proceed through multistep reaction mechanisms often utilize complex, polar active sites positioned with sub-angstrom precision to mediate distinct chemical steps, which makes their de novo construction extremely challenging. We sought to overcome this challenge using the classic catalytic triad and oxyanion hole of serine hydrolases as a model system. We used RFdiffusion1 to generate proteins housing catalytic sites of increasing complexity and varying geometry, and a newly developed ensemble generation method called ChemNet to assess active site geometry and preorganization at each step of the reaction. Experimental characterization revealed novel serine hydrolases that catalyze ester hydrolysis with catalytic efficiencies (kcat/Km) up to 3.8 x 103 M-1 s-1, closely match the design models (C RMSDs < 1 [A]), and have folds distinct from natural serine hydrolases. In silico selection of designs based on active site preorganization across the reaction coordinate considerably increased success rates, enabling identification of new catalysts in screens of as few as 20 designs. Our de novo buildup approach provides insight into the geometric determinants of catalysis that complements what can be obtained from structural and mutational studies of native enzymes (in which catalytic group geometry and active site makeup cannot be so systematically varied), and provides a roadmap for the design of industrially relevant serine hydrolases and, more generally, for designing complex enzymes that catalyze multi-step transformations.