Biocatalysis@TUDelft

Nature Biotechnology, Published online: 14 October 2025; doi:10.1038/s41587-025-02840-4

Recent patents relating to directed evolution methods and products.

In this study, formatotrophic biotransformations in recombinant Cupriavidus necator catalyzed by Rieske oxygenases (ROs) toward various oxyfunctionalized olefins are demonstrated. The oxidation of formate regenerates NADH, the ROs’ cofactor, essentially fueling the reaction. The formatotrophically cultivated cells performed at the same efficiency as heterotrophically grown cells of the same strain. NDO: naphthalene dioxygenase, CDO: cumene dioxygenase. CBB cycle: Calvin-Benson-Bassham cycle.

The use of single carbon (C₁) molecules, such as carbon dioxide or formate, is crucial in the transition from a linear, petroleum-based economy to a circular bioeconomy. Formate can serve as both a carbon and energy source, further enhancing its attractiveness as a feedstock. Cupriavidus necator, a lithoautotrophic microbial chassis strain, provides an opportunity to leverage formate for the synthesis of valuable products. However, its ability to grow on formate and the subsequent coupling of that process to recombinantly produced redox enzymes for the efficient production of high-value-added products in a biotransformation has not yet been established. Here, we report the development of a formate-driven C. necator whole-cell chassis that recombinantly produces Rieske oxygenases (ROs) and elaborate on possible stress responses of the cells during formatotrophic cultivation. The whole-cell chassis efficiently catalyzes the oxyfunctionalization of olefins fueled by formate oxidation. For instance, styrene is dihydroxylated to (R)-1-phenylethane-1,2-diol in an excellent 95% yield and with good enantioselectivity (74% ee) under formatotrophic conditions. The product yield and optical purity obtained demonstrate the synthetic usefulness of formate-fueled whole-cell bio-transformations in C. necator.

Hydrogenases catalyze reversible H₂ production and are potential models for renewable energy catalysts. Here, the full redox landscape of a group 3 [NiFe]-hydrogenase from methanothermococcus thermolithotrophicus is elucidated, resembling group 1 enzymes. Structural and spectroscopic analyses reveal a catalytic-ready state with nickel seesaw coordination, enabling intermediate trapping and advancing mechanistic understanding of oxygen-sensitive [NiFe] enzymes.

The radical SAM enzyme NirJ is responsible for the removal of two propionate side chains from its substrate didecarboxy-siroheme during heme d ₁ biosynthesis. The cleavage products of the NirJ reaction were not known so far. In this study, it is shown that acrylate is formed as the reaction byproduct and that the resulting tetrapyrrole product carries methylene groups after the removal of the propionate groups.

Heme d ₁ is an iron-containing, modified tetrapyrrole that serves as an essential prosthetic group in cytochrome cd ₁ nitrite reductases. The biosynthesis of heme d ₁ from the precursor siroheme requires three or four enzymatic steps, including the removal of two propionate side chains, the latter being catalyzed by the radical SAM enzyme NirJ. Although the removal of the propionate side chains by NirJ has been shown previously, several aspects of NirJ catalysis remained elusive, including the type of its auxiliary iron–sulfur cluster as well as the identity of the cleavage byproduct and the actual product of the NirJ reaction. Here, we demonstrate by Mössbauer spectroscopy that NirJ contains a [4Fe-4S] cluster ligated by cysteine residues as its auxiliary cluster. We show that acrylate is released during the NirJ reaction as the cleavage byproduct, as observed by HPLC-UV and HPLC–MS analysis of enzyme activity assay mixtures after derivatization. Finally, we provide strong evidence from HPLC-UV/Vis and HPLC-MS analysis that the NirJ reaction product contains methylene groups at positions C3 and C8 of the tetrapyrrole macrocycle. Based on these results, we propose a revised version of the NirJ reaction mechanism, including a potential role of the auxiliary iron–sulfur cluster as an electron donor for radical quenching.

A reversible chemoenzymatic labeling strategy for the efficient analysis of protein O-Glc was developed, and numerous canonical and noncanonical O-Glc modifications were identified across human cell lines with mass spectrometry (MS) technology, suggesting protein O-Glc is a widespread post-translational protein modification.

Abstract

O-linked glucose (O-Glc), the β-linked modification of serine residues, is a rare form of protein glycosylation first identified on proteins containing epidermal growth factor (EGF)-like domains (canonical O-Glc). Several recent studies revealed that proteins lacking EGF-like domains could also undergo O-Glc modification (noncanonical O-Glc). However, the biosynthetic origin and biological function of protein O-glucosylation remain poorly understood and debated, owing to the lack of effective analytical tools. Here, a reversible chemoenzymatic labeling strategy for O-Glc analysis is described. By the strategy described, a large number of canonical and noncanonical O-Glc sites were identified in human cell lines, indicating that protein O-Glc is a widespread post-translational protein modification.

A combined synthetic and biochemical approach links the flavoenzyme EpxF to three key steps in epoxomicin biosynthesis. Crystallography, mutagenesis, and ¹³C-labeling complete the mechanistic picture of α′,β′-epoxyketone formation and illustrate the potential of flavin-mediated transformations.

Abstract

Epoxomicin is a highly potent natural proteasome inhibitor and the structural scaffold for the anticancer drug carfilzomib. The biosynthesis of its α′,β′-epoxyketone warhead involves the flavoenzyme EpxF, but a molecular understanding of the key catalytic reaction cascade remained elusive. Here, we disclose detailed mechanistic insights by characterizing all intermediates in the sequential steps of decarboxylation, desaturation, and epoxidation with synthetic flavins and the flavin-dependent oxidoreductase EpxF. A high-resolution crystal structure of EpxF revealed the architecture of the active site and enabled the identification of key catalytic residues. Exploratory docking based on this structure served as a qualitative tool to guide mutagenesis and rationalize substrate recognition. NMR studies with a ¹³C-labeled epoxomicin precursor and structure-based EpxF variants further supported the proposed mechanism. Our integrated approach revealed similarities between synthetic and natural flavin catalysts and offers avenues for developing sustainable biomimetic reactions.

Proceedings of the National Academy of Sciences, Volume 122, Issue 41, October 2025.

Proceedings of the National Academy of Sciences, Volume 122, Issue 41, October 2025.
SignificanceThe sequence of amino acids within a protein dictates its structure and function. Protein engineering campaigns seek to discover protein sequences with desired functions. Data-driven models of the sequence–function relationship can be used to ...

Proceedings of the National Academy of Sciences, Volume 122, Issue 41, October 2025.
SignificanceGenerative modeling has enabled novel approaches to the design of functional molecules. However, its practical utility has been impeded by its tendency to propose molecules that are difficult or impossible to synthesize. To address this issue, ...

Proceedings of the National Academy of Sciences, Volume 122, Issue 41, October 2025.
A distinguishing feature of the neural network models used in Physics and Chemistry is that they must obey basic underlying symmetries, such as symmetry to translations, rotations, and the exchange of identical particles. Over the course of the last ...

Proceedings of the National Academy of Sciences, Volume 122, Issue 41, October 2025.
SignificanceHydrogenases are large and efficient metalloenzymes that catalyze the conversion between protons and hydrogen, with an inorganic active site that coordinates cheap transition metals. They have long been considered to replace precious metals as ...

Proceedings of the National Academy of Sciences, Volume 122, Issue 41, October 2025.
SignificanceEnhancing enzyme catalytic efficiency and thermal stability is crucial for biocatalytic and industrial applications. While structure-based protein sequence design can improve thermal stability, its success in enzyme engineering is limited by ...

Maculalactones are bioactive γ-butyrolactone natural products with promising antifouling properties, yet their limited availability constrains further development. Here, we report the identification and characterization of the key biosynthetic enzymes MacE and MacF from the maculalactone (mac) biosynthetic gene cluster in Nodularia sp. NIES-3585. In vitro reconstitution confirmed that MacE catalyzes acyloin condensation between phenylpyruvic acids, while MacF mediates subsequent fusion of the MacE product with cinnamoyl-CoA derivatives. Structure-guided engineering improved enzyme solubility, enabling systematic substrate scope expansion. Through a one-pot enzymatic synthesis, we synthesized 13 maculalactone derivatives with yields reaching more than 96%, while a complementary combinatorial biosynthesis approach enabled the incorporation of sterically demanding indole moieties, thereby significantly expanding the maculalactone structural space. Bioactivity screening against A549 lung adenocarcinoma cells revealed moderate cytotoxicity (IC₅₀ = 20.82–56.64 μM), with methylation of phenol groups significantly enhancing activity. Several analogs exhibited notable antibacterial activity against clinically relevant Gram-positive pathogens, including Streptococcus pneumoniae. Our work provides a sustainable platform for producing structurally diverse maculalactones, laying the foundation for future biomedical applications.

Type II NADH dehydrogenases (NDH-2s) are accessory enzymes of the bacterial electron transport chain (ETC). While they functionally overlap with Complex I, their main role is not proton translocation but maintaining the intracellular NADH/NAD+ balance. Although often non-essential, NDH-2 become crucial in species lacking complex I, serving as the primary electron entry point into the ETC. Their virtual absence in mammals makes these enzymes attractive targets for antimicrobial drug development and mitochondrial functional restoration. NDH-2s catalyze electron transfer from NADH to quinones, yet two distinct catalytic mechanisms have been described for members of the family: a classical ping-pong mechanism and an atypical ternary mechanism involving the formation of a charge transfer complex (CTC). The molecular basis of these mechanisms remains unclear. Also, their occurrence among NDH-2s from different bacterial lineages in unknown. Here we combined molecular phylogenetics, ancestral sequence reconstruction, expression and biochemical characterization of ancestral and modern enzymes and, molecular dynamics simulations to explore the mechanistic versatility of NDH-2s across Bacteria. Our results show the atypical ternary mechanism is restricted to the Firmicutes (Bacillota) lineage and it is defined by the presence of a single substitution located at the bottom of the active site. This work provides an evolutionary framework for understanding NDH-2 mechanistic versatility. Besides, it establishes a basis for drug discovery targeting pathogenic strains and opens avenues to develop innovative strategies to complement dysfunctional mitochondria.

Enzyme photobiocatalysis uses light to drive high-energy transformations but is limited by the rarity of photoenzymes. Fatty acid photodecarboxylase (FAP), a recently discovered photoenzyme, enables fatty acid conversion to alkanes/alkenes via excitation of an FAD cofactor, though its poor photostability and photoinactivation has hindered industrial applications. Here, we combine protein engineering approaches with biocatalytic and biophysical techniques, as well as computational chemistry, to demonstrate that additional active site cysteine residues can suppress oxygen-mediated inactivation processes that are driven by the FAD triplet-excited state. We identify a number of positions close to the FAD for cysteine residues that lead to a significant enhancement in activity as a result of an increase in the number of catalytic turnovers and improved photostability. The additional cysteine residues quench the triplet excited state of the FAD cofactor via a proposed proton-coupled electron transfer mechanism, resulting in lower levels of harmful reactive oxygen species. Our study highlights promising routes to mitigate non-productive, photoinactivation pathways in FAP and informs the rational design of new flavin-based photoenzymes.

Fatty acid photodecarboxylase (FAP) converts fatty acids to hydrocarbons upon irradiation, rendering it one of the few known natural photoenzymes with potential for biofuel production. However, photoinactivation remains a major barrier for industrial applications. Here, we employ light footprinting ion mobility mass spectrometry (IMMS), together with complementary biophysical approaches, to investigate conformational changes and stability of FAP upon irradiation. Mass spectra of FAP reveal four major proteoforms: FAP is always bound to the FAD cofactor and up to two endogenous fatty acid substrates in the dark, whereas upon irradiation these non-covalent species are released to produce the apo form. Ion mobility data show an irradiation-induced conformational change in FAP concomitant with photoinactivation. Collision cross section distributions reveal the effect of fatty acid binding on the conformation and photoactivity of the protein: FAP adopts extended conformations under irradiation and a gradual decay in photoactivity. The unfolding of both apo-FAP and holo-FAP was examined using collisionally activating IMS (aIMS), which showed irradiation destabilizes the native conformations of holo-FAP. Multivariate analysis using IMMS data identified characteristic FAP analogues and FAD fragments distinguishing dark and irradiated states. The kinetics analysis of proteoform transitions were determined by light foot printing IMMS, revealing the events occurring during the productive and non-productive FAP photocycles at the proteoform and conformation levels. This study provides a new experimental approach for investigating photocatalytic and photoinactivation processes, and highlights the complex conformational dynamics required for efficient photocatalysis, which is crucial for understanding catalytic mechanism and guiding future protein engineering and design.

Directed evolution is a method for engineering biological systems or components, such as proteins, wherein desired traits are optimised through iterative rounds of mutagenesis and selection of fit variants. The process of protein directed evolution can be envisaged as navigation over high-dimensional landscapes with numerous local maxima. The performance of any strategy in navigating such a landscape is dependent on the ruggedness of that landscape. However, this information is generally unavailable at the outset of an experiment, and cannot currently be computed using analytical methods. Here we propose SLIDE, Sequence-free Landscape Inference for Directed Evolution, which consists of two parts. First, SLIDE provides an estimation for landscape ruggedness from a mutating population using only population-level phenotypic data and an estimation of mutation rate. Ruggedness information in itself is valuable in protein design, for instance in predicting evolutionary stability. Second, SLIDE offers a framework for using the estimated ruggedness metric to select high-performing parameters for directed evolution control. Using theoretical NK landscapes and four real-world protein fitness landscapes, we demonstrate improvement upon the performance of standard selection strategies, particularly on rugged landscapes, using a pipeline that could also be combined with emerging AI-based methods for driving direction evolution.

Halogenation enhances the stability and function of pharmaceuticals, biomaterials, and industrial compounds. However, chemical halogenation lacks stereoselectivity and requires the use of toxic or expensive chemicals. Although enzymatic halogenation can improve selectivity and reduce environmental impact, current halogenases are inefficient and insoluble, leading to low yields that limit their applications. Here, we develop RebHEvo4, a soluble and highly active tryptophan halogenase, containing 12 mutations that confer 37-fold and 44-fold increases in 7-chloro and 7-bromotryptophan production respectively, in vivo. To create RebHEvo4, we devised an aminoacyl tRNA synthetase based halogenase biosensor and conducted over 500 hours of phage-assisted continuous evolution (PACE). Use of RebHEvo4 in a bioreactor resulted in the production of 2.7 g/L of halogenated tryptophan. When coupled with a downstream enzyme, RebHEvo4 allowed 36-fold increased yields of halogenated tryptamines compared to the wild-type enzyme. Additionally, RebHEvo4 enabled efficient production of genetically encoded antimicrobial halogenated peptides. The efficient, site-specific halogenation by our evolved halogenase will accelerate sustainable biomanufacturing of halogenated drugs.