Biocatalysis@TUDelft

Tagatose-1,6-bisphosphate aldolases (TagAs) have been understudied, with few available publications reporting poor stereoselectivity. This leads to their classification as the less promising aldolases. The aim of this concept is to revisit these findings and uncover inconsistencies to better understand their true potential. It is believed that, with further exploration, TagAs can prove as valuable as other aldolases in catalyzing stereoselective reactions.

Tagatose-1,6-bisphosphate aldolases (TagA) have been understudied. The limited publications on the subject report a poor stereoselectivity, which quickly leads to their classification among the less interesting aldolases. This conclusion, which appears inadequate, discourages further investigation by other researchers. The objective here is to re-examine all these studies in detail to finally uncover inconsistencies and unexplained observations regarding the behavior of these enzymes. Ultimately, there is a significant lack of research that can deepen the understanding of TagA, which will likely reveal its potential to be as valuable as its counterparts: fructose-1,6-bisphosphate, rhamnulose-1-phosphate, and fuculose-1-phosphate aldolases.

The noncanonical amino acid, cyanophenylalanine, is genetically inserted close to iron–sulfur clusters of either spinach ferredoxin or iron–iron hydrogenase, and functions as an infrared spectroscopic reporter for changes in cluster redox state.

The noncanonical amino acid, para-cyanophenylalanine (CNF), when incorporated into metalloproteins, functions as an infrared spectroscopic probe for the redox state of iron-sulfur clusters, offering a strategy for determining electron occupancy in the electron transport chains of complex metalloenzymes. A redshift of ≈1–2 cm⁻¹ in the nitrile (NC) stretching frequency is observed, following reduction of spinach ferredoxin modified to contain CNF close to its [2Fe–2S] center, and this shift is reversed on re-oxidation. We extend this to CNF positioned near to the proximal [4Fe–4S] cluster of the [FeFe] hydrogenase from Desulfovibrio desulfuricans. In combination with a distal [4Fe–4S] cluster and the [4Fe–4S] cluster of the active site ‘H-cluster’ ([4Fe–4S]_H), the proximal cluster forms an electron relay connecting the active site to the surface of the protein. Again, a reversible shift in wavenumber for CNF is observed, following cluster reduction in either apo-protein (containing the iron-sulfur clusters but lacking the active site) or holo-protein with intact active site, demonstrating the general applicability of this approach to studying complex metalloenzymes.

Enzymes achieve catalysis by dynamically sampling diverse conformational states. Beyond this plasticity, individual enzyme molecules occupy metastable substates, forming an ensemble of functional substates within a population. Since shifts in functional substate dynamics drive phenotypic variation, their evolutionary trajectories are central to the emergence of new functions. However, the challenge of measuring functional substates has hindered our understanding of their role in enzyme evolution and the optimization of conformational substates. Here, we address this gap by investigating how functional and conformational substates were modulated during enzyme functional transitions, using single-molecule kinetic assays and molecular dynamics simulations. We analyzed wild-type phosphotriesterase (PTE) and 18 evolved variants that transitioned from the native PTE to promiscuous arylesterase (AE) activity. Our findings reveal that evolutionary transitions reshape functional and conformational substate landscapes: PTE-specialized variants exhibit broader substate distributions, whereas AE-specialized variants display more uniform substates. These results provide the first direct evidence that enzyme evolution is accompanied by coordinated shifts in functional and conformational substate equilibria, optimizing both for the enzymes catalytic efficiency. This work highlights the power of single-molecule techniques in uncovering how heterogeneous enzyme populations navigate substate transitions and, ultimately, how these transitions shape enzyme evolvability.

Science, Volume 388, Issue 6743, Page 142-142, April 2025.

De novo protein design has been used to create functional protein assemblies, but its reliance on standard amino acids limits the integration of synthetic supramolecular strategies for precise control of molecular assemblies. Herein, we present an artificial protein assembly that integrates synthetic supramolecular design with de novo protein engineering. Focusing on unit and connection rigidity, we designed and created the Bi-Porphyrin Acquisition Designer protein (BiPAD), which captures two highly designable synthetic porphyrins, via state-of-the-art generative protein design to fuse α/β-folded porphyrin-binding motifs. BiPADs captured rigid porphyrins, each bearing an additional metal coordination site, resulting in a metal-responsive cyclic assembly with the intended structure. Furthermore, high-speed atomic force microscopy revealed dynamic structural changes in the BiPAD assembly. This work expands the designability of artificial protein assemblies, paving the way for the synergistic design of functional systems through the integration of protein engineering and synthetic chemistry.

De novo enzyme design starts from ideal active site descriptions consisting of constellations of catalytic residue functional groups around reaction transition state(s), and seeks to generate protein structures that can accurately hold the site in place. Highly active enzymes have been designed starting from such descriptions using the generative AI method RFdiffusion [1-3], but there are two current methodological limitations. First, the geometry of the active site can only be specified at the residue level, so for each catalytic residue functional group placed around the reaction transition state, the possible locations of the residue backbone must be enumerated by building side chain rotamers back from the functional group. Second, the location of the catalytic residues along the sequence must be specified in advance, which considerably limits the space of solutions which can be sampled. Here we describe a new deep generative method, Rosetta Fold diffusion 2 (RFdiffusion2), that solves both problems, enabling enzymes to be designed from sequence agnostic descriptions of functional group locations without inverse rotamer generation. We first evaluate RFdiffusion2 on an in silico enzyme design benchmark of 41 diverse active sites and find that it is able to successfully build proteins scaffolding all 41 sites, compared to 16/41 with prior state-of-the-art deep learning methods. Next, we design enzymes around three diverse catalytic sites and characterize the designs experimentally; in each case we identify active catalysts in testing less than 96 sequences. RFdiffusion2 demonstrates the potential of atomic resolution generative models for the design of de novo enzymes directly from their reaction mechanisms.

Proceedings of the National Academy of Sciences, Volume 122, Issue 14, April 2025.
SignificanceA fundamental concept in chemistry is that the properties of elements are periodic. Magnesium is a ubiquitous alkaline Earth (group II element) metal cofactor in enzymes that catalyzes many biochemical reactions, including DNA processing. We ...

Biocatalysis is a branch of catalysis that has allowed the development of a diverse range of sophisticated synthetic methodologies that target geometrically demanding structures such as pharmaceuticals, natural products, and their analogues. The routes are often more efficient due to enzymes innately high degree of chemo-, regio- and stereoselectivity, while making the overall process more sustainable when compared to their chemical equivalent, as enzymes are naturally biodegradable and operate under physiological conditions. Herein, we demonstrate the power of this catalytic approach via the development of a chemoenzymatic one-pot process that allowed access to the fungal metabolites tenuipesones A and B, along with their enantiomeric counterparts, in good yields (60-72 %) and excellent enantioselectivity (>99 % ee). The stereochemical outcome of the products was controlled through careful selection of the biocatalysts, enabling control of the configuration of the key C7 chiral center. This novel chemoenzymatic cascade process allowed access to all four possible stereoisomers, with their absolute stereochemistry being evaluated to confirm the true configuration of the natural isolates.

The role of amino-acid residues distant from an enzymes active site in facilitating the complete catalytic cycle--including substrate binding, chemical transformation, and product release-- remains poorly understood. Here, we investigate how distal mutations promote the catalytic cycle by engineering mutants of three de novo Kemp eliminases containing either active-site or distal mutations identified through directed evolution. Kinetic analyses, X-ray crystallography, and molecular dynamics simulations reveal that while active-site mutations create preorganized catalytic sites for efficient chemical transformation, distal mutations enhance catalysis by facilitating substrate binding and product release through tuning structural dynamics to widen the active-site entrance and reorganize surface loops. These distinct contributions work synergistically to improve overall activity, demonstrating that a well-organized active site, though necessary, is not sufficient for optimal catalysis. Our findings reveal critical roles that distal residues play in shaping the catalytic cycle to enhance efficiency, yielding valuable insights for enzyme design.

As one of the most potent and selective protein phosphatase inhibitors, fostriecin shows a broad range of anticancer activity. In light of fostriecin’s antiproliferative properties, a phase I clinical trial was conducted on the natural product, but was soon halted due to issues with compound stability and purity. Numerous efforts in the past two decades have yielded 17 successful syntheses that proceed in 19 to 34 steps. Herein, we develop a modular chemoenzymatic approach that provides fostriecin and its analogs in a collective manner in 9 steps (longest linear sequence). The synthesis features a convergent assembly of three key fragments and a late-stage chemoenzymatic derivatization of an advanced intermediate that (i) installs two of the key pharmacophores and (ii) allows ready diversification of the hydrophobic tail. A key feature in this derivatization is the optimization of an enzymatic C–H oxidation step through the concurrent use of decoy molecule strategy and rational enzyme engineering. Cumulatively, our strategy capitalizes on the exquisite chemoselectivity of enzymatic transformations while ensuring synthetic modularity and versatility for analog generation. This work will facilitate future investigation into the biological activities and medicinal chemistry of the natural product family.

The anti - and syn -selective epoxidation of various α,β-unsaturated aldehydes is promoted by an engineered variant of 2-deoxy-D-ribose-5-phosphate aldolase (DERA), giving rise to various α,β-epoxy-aldehydes with excellent enantiopurity (enantiomeric ratio up to 99 : 1).

Abstract

The enzyme 2-deoxy-D-ribose-5-phosphate aldolase (DERA) naturally catalyzes the reversible aldol addition between acetaldehyde and D-glyceraldehyde-3-phosphate to yield 2-deoxy-D-ribose-5-phosphate. Herein we describe the redesign of DERA into a proficient non-natural peroxygenase that promotes the asymmetric epoxidation of various α,β-unsaturated aldehydes. This repurposed aldolase, named DERA-EP, is able to utilize H₂O₂ to accomplish both anti- and syn-selective epoxidations of various α,β-unsaturated aldehydes to give the corresponding epoxides with moderate to high diastereoselectivity (diastereomeric ratio up to 99 : 1) and excellent enantioselectivity (enantiomeric ratio up to 99 : 1). Crystallographic analysis of DERA-EP in a substrate-free and substrate-bound state provides a structural context for the evolved activity, a clear explanation for the high enantioselectivity, and compelling evidence for catalysis via enzyme-bound iminium ion intermediates. The unprecedented anti-selectivity of DERA-EP with multiple α,β-unsaturated aldehydes is complementary to the syn-selectivity of previously reported enzyme-, metal- and organo-catalysts, making DERA-EP an attractive new asset to the toolbox of epoxidation catalysts.

Photoenzymatic catalysis facilitates stereoselective new-to-nature chemistry under mild conditions. In addition to the rational design of artificial photoenzymes, naturally occurring redox enzymes have been repurposed for this approach. Most prominently, flavin-containing cofactors can promote photoredox catalysis in the chiral protein environment, with several examples of enantioselective C–C bond forming reactions reported in recent years. Here, we add another class of natural enzymes, which utilize the pyrroloquinoline quinone (PQQ) cofactor, to the toolbox of photobiocatalysis. Although structurally distinct from flavin, PQQ exhibits mechanistic similarities, as it also absorbs visible light and is capable of single-electron transfer. First, we established the trimethyl ester PQQMe3 as a stand-alone photoredox catalyst in pure organic solvent. Upon excitation, PQQMe3 enables the redox-neutral radical cyclization of an N-(bromoalkyl)-substituted indole. We then tested a panel of PQQ-dependent sugar and alcohol dehydrogenases for photoenzymatic catalysis in aqueous buffer, focusing on a redox-neutral radical reaction to form oxindoles. Under optimized reaction conditions, we obtained 69% yield and an 82:18 enantiomeric ratio. Our work thus demonstrates that PQQ enzymes are capable of stereoselective photoredox catalysis. Future enzyme engineering efforts based on computational modelling and directed evolution will fully unlock their synthetic potential.

Nature Communications, Published online: 20 March 2025; doi:10.1038/s41467-025-58038-4

Enzyme kinetic parameter prediction is a challenge in enzyme discovery and engineering. Here, the authors train a robust deep learning model CataPro to predict enzyme kinetic parameters and validate its practicality through wet-lab experiments.

Thioketals are an important class of compounds that enable the selective preparation and protection of carbonyl compounds in chemical synthesis. Despite their synthetic utility, selective cleavage of thioketals often requires the use of harsh conditions and stoichiometric reagents that are largely bioincompatible. Herein, we describe a biocatalytic strategy for the selective cleavage of thioketals using enzymatic bromide recycling by vanadium-dependent haloperoxidase (VHPO) enzymes. This process involves thioketal cleavage through repetitive VHPO-mediated formation of hypohalous acid with a catalytic quantity of halide salt and hydrogen peroxide as the terminal oxidant. This method is demonstrated on a broad range of 1,3-thioketals in high yield and excellent chemoselectivity. The protocol has been demonstrated on gram-scale, run with lysate and whole cells, and has been extended to cleavage of 1,3-acetals, 1,4-thioketals, and 1,3-oxathiolanes.

A single-step biocatalytic route to complex indolizidine and quinolizidine alkaloids is described that relies on transaminase-triggered double intramolecular aza-Michael methodology. In one case, a retro-double intramolecular aza-Michael reaction enables dynamic kinetic resolution.

Abstract

Biocatalysis is now a well-established branch of catalysis and the growing toolbox of natural, evolved and designer enzymes is enabling chemistry previously deemed inaccessible. However, most enzyme methodologies have been developed for functional group interconversions, such as the conversion of a ketone into an amine or alcohol, and do not result in the generation of significant 3D molecular complexity. The application of enzyme-triggered reaction cascade methodologies has the potential to transform simple substrates into complex sp³-rich molecules in one step. Herein, we describe a single-step biocatalytic route to high-value, complex indolizidine, and quinolizidine alkaloids, which relies on a transaminase-triggered double intramolecular aza-Michael reaction. This approach allows access to architecturally complex, natural-product-like N-heterocycles and reveals intriguing examples of diastereoselectivity in these enzyme-triggered reactions. Significantly, we demonstrate an elegant example of a biocatalytic cascade where the transaminase plays a dual role in generating complex N-heterocycles and where a retro-double intramolecular aza-Michael reaction mediates a dynamic kinetic resolution and enables the isolation of sp³-rich indolizidine diastereoisomers containing five stereocenters, as single isomers.

The biosynthesis of saturated bioisosteres of ortho-disubstituted benzenes was reported by the artificial photoenzyme harboring a genetic incorporated photosensor under the light with several examples after the directed evolution. Our work provides a biocatalytic strategy for the synthesis of saturated bioisosteres of ortho-disubstituted benzenes and expands the utility of artificial photoenzyme for abiological reactions.

Abstract

Saturated bioisosteres of ortho-substituted benzenes are of significant interest due to their enhanced pharmacokinetic properties, such as improved metabolic stability and reduced toxicity, making them valuable in drug design and development. However, efficient synthesis of them remains a challenge in organic chemistry. Herein, we report the biocatalytic synthesis of saturated bioisosteres of ortho-substituted benzenes using engineered artificial photoenzymes. The artificial photoenzyme, incorporating genetically encoded unnatural amino acids with benzophenone photosensitizer residue, facilitate the formation of chiral saturated bioisosteres with moderate enantiomeric excess via the energy transfer process. Our results demonstrate the versatility of artificial photoenzymes in mediating new-to-nature reactions that are difficult to achieve with conventional chemical or enzymatic methods.

The nitrogen-nitrogen (N-N) bond motif comprises an important class of compounds for drug discovery. Synthetic methods are primarily based on the modification of N-N or N=N precursors, whereas selective methods for direct N-N coupling offer advantages in terms of atom economy and sustainability. In this context, enzymes such as piperazate synthases (PZSs), which naturally catalyze the N-N cyclization of L-N5-hydroxyornithine to the cyclic hydrazine L-piperazate, may allow an expansion of the current narrow range of chemical approaches for N-N coupling. In this study, we demonstrate that PZSs are able to catalyze the conversion of various N-hydroxylated diamines, which are different from the natural substrate. The N-hydroxylated diamines were obtained in situ using N-hydroxylating monooxygenases (NMOs), allowing subsequent cyclization by PZS, ultimately forming the N-N bond to yield various N-N bond-containing heterocycles. Using bioinformatic tools, we identified novel NMO and PZS homologs that exhibit distinct activity and stereoselectivity profiles. The screened panel yielded 17 hydroxylated diamines and new promiscuous NMOs, thereby expanding the substrate range of NMOs resulting in the formation of previously poorly accessible N-hydroxylated products as substrates for PZS. The subsequently investigated PZSs led to a series of 5- and 6-membered N-N bond-containing heterocycles, and the most promiscuous catalysts were used to scale up and optimize the synthesis, yielding the desired N-N bond-containing heterocycles with up to 45% isolated yield. Overall, our data provides essential insights into the substrate promiscuity and activity of NMOs and PZSs, further enhancing the potential of these biocatalysts for an expanded range of N-N coupling reactions.

An efficient artificial copper-dependent Michaelase featuring a metal-binding unnatural amino acid, e. g. bipyridyl alanine (BpyA), was optimized through directed evolution and applied in catalytic asymmetric additions of 2-acetyl azaarenes to nitroalkenes. Various chiral γ-nitro butyric acid derivatives were obtained in good yields with high enantioselectivities. Moreover, the reaction was performed at preparative scale and the product was used in follow-up derivatization reactions towards various high-value-added pharmaceutically relevant compounds.

Abstract

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) comprising a genetically encoded copper-binding ligand, i. e. (2,2-bipyridin-5-yl)alanine (BpyA), was developed. For the first time, such an ArM containing a non-canonical metal-binding amino acid was successfully optimized through directed evolution. The evolved Cu_Michaelase was applied in the copper-catalyzed asymmetric addition of 2-acetyl azaarenes to nitroalkenes, yielding various γ-nitro butyric acid derivatives, which are precursors for a range of high-value-added pharmaceutically relevant compounds, with good yields and high enantioselectivities (up to >99 % yield and 99 % ee). Additionally, the evolved variant could be further used in a preparative-scale synthesis, providing chiral products for diverse derivatizations. X-ray crystal structure analysis confirmed the binding of Cu(II) ions to the BpyA residues and showed that, in principle, there is sufficient space for the 2-acetyl azaarene substrate to coordinate. Kinetic studies showed that the increased catalytic efficiency of the evolved enzyme is due to improvements in apparent K _M for both substrates and a notable threefold increase in apparent k _cat for 2-acetyl pyridine. This work illustrates the potential of artificial metalloenzymes exploiting non-canonical metal-binding ligands for new-to-nature biocatalysis.

Nature, Published online: 12 February 2025; doi:10.1038/s41586-024-08553-z

A metalloenzyme capable of oxidatively cleaving cellulose, found in a microbial community specialized in lignocellulose degradation, could enable sustainable biofuel production.

Biocatalysis contributes significantly to the development of more sustainable synthetic pathways by using mild reaction conditions and water as a solvent. However, many relevant classes of compounds, including privileged groups in drug design, are not yet accessible via enzymatic pathways. In this context, the development of an enzymatic route to indazoles remains an unmet challenge. Here, we present the first example of nitroreductase-triggered indazole formation, in which 2-nitrobenzylamine derivatives are converted to reactive nitrosobenzylamine intermediates that spontaneously cyclize and aromatize to indazoles. Two nitroreductases, NfsA and BaNTR1, were found to accept a series of 2-nitrobenzylamine derivatives with excellent conversions (up to >99 %). In the case of N-substituted nitrosobenzylamines, 2H-indazoles were formed, whereas other derivatives led to 1H-indazoles. The synthetic value of the nitroreductase-triggered indazole formation was further demonstrated by successful coupling with an imine reductase (IRED15) in a sequential cascade reaction. With this cascade, N-methyl-2Hindazole was accessible from cheap 2-nitrobenzaldehyde and methylamine, resulting in 62 % isolated yield.

Promiscuity guided evolution of the decarboxylative aldolase, UstD, resulted in an efficient biocatalyst for synthesis of tertiary γ-hydroxy non-canonical amino acids. Simultaneous collection of variant activity and promiscuity was enabled by competition screening during directed evolution. Changes in promiscuity effectively identified distal residues that influence catalysis, a longstanding challenge in protein engineering.

Abstract

Many applications of enzymes benefit from activity on structurally diverse substrates. Here, we sought to engineer the decarboxylative aldolase UstD to perform a challenging C−C bond forming reaction with ketone electrophiles. The parent enzyme had only low levels of activity, portending multiple rounds of directed evolution and a possibility that mutations may inadvertently increase the specificity of the enzyme for a single model screening substrate. We show how to intentionally guide UstD towards generality through multi-generational directed evolution using substrate-multiplexed screening (SUMS). Mutations outside of the active site that impact catalytic function were immediately revealed by shifts in promiscuity, even when the overall activity was lower. By re-targeting these distal residues that couple to the active site with saturation mutagenesis, broadly activating mutations were readily identified. When analyzing active site mutants, SUMS identified both specialist enzymes that would have more limited utility as well as generalist enzymes with complementary activity on diverse substrates. These new UstD enzymes catalyze convergent synthesis of non-canonical amino acids bearing tertiary alcohol side chains. This methodology is easy to implement and enables the rapid and effective evolution of enzymes to catalyze desirable new functions.