Explore the vast range of titles available.

De novo enzyme design has sought to introduce active sites and substrate-binding pockets that are predicted to catalyse a reaction of interest into geometrically compatible native scaffolds 1 , 2 , but has been limited by a lack of suitable protein structures and the complexity of native protein sequence–structure relationships. Here we describe a deep-learning-based ‘family-wide hallucination’ approach that generates large numbers of idealized protein structures containing diverse pocket shapes and designed sequences that encode them. We use these scaffolds to design artificial luciferases that selectively catalyse the oxidative chemiluminescence of the synthetic luciferin substrates diphenylterazine 3 and 2-deoxycoelenterazine. The designed active sites position an arginine guanidinium group adjacent to an anion that develops during the reaction in a binding pocket with high shape complementarity. For both luciferin substrates, we obtain designed luciferases with high selectivity; the most active of these is a small (13.9 kDa) and thermostable (with a melting temperature higher than 95 °C) enzyme that has a catalytic efficiency on diphenylterazine ( k cat / K m  = 10 6  M −1  s −1 ) comparable to that of native luciferases, but a much higher substrate specificity. The creation of highly active and specific biocatalysts from scratch with broad applications in biomedicine is a key milestone for computational enzyme design, and our approach should enable generation of a wide range of luciferases and other enzymes. A deep-learning-based strategy is used to design artificial luciferases that catalyse the oxidative chemiluminescence of diphenylterazine with high substrate specificity and catalytic efficiency.

There has been considerable recent progress in designing new proteins using deep-learning methods 1 – 9 . Despite this progress, a general deep-learning framework for protein design that enables solution of a wide range of design challenges, including de novo binder design and design of higher-order symmetric architectures, has yet to be described. Diffusion models 10 , 11 have had considerable success in image and language generative modelling but limited success when applied to protein modelling, probably due to the complexity of protein backbone geometry and sequence–structure relationships. Here we show that by fine-tuning the RoseTTAFold structure prediction network on protein structure denoising tasks, we obtain a generative model of protein backbones that achieves outstanding performance on unconditional and topology-constrained protein monomer design, protein binder design, symmetric oligomer design, enzyme active site scaffolding and symmetric motif scaffolding for therapeutic and metal-binding protein design. We demonstrate the power and generality of the method, called RoseTTAFold diffusion (RFdiffusion), by experimentally characterizing the structures and functions of hundreds of designed symmetric assemblies, metal-binding proteins and protein binders. The accuracy of RFdiffusion is confirmed by the cryogenic electron microscopy structure of a designed binder in complex with influenza haemagglutinin that is nearly identical to the design model. In a manner analogous to networks that produce images from user-specified inputs, RFdiffusion enables the design of diverse functional proteins from simple molecular specifications. Fine-tuning the RoseTTAFold structure prediction network on protein structure denoising tasks yields a generative model for protein design that achieves outstanding performance on a wide range of protein structure and function design challenges.

Irinotecan treats a range of solid tumors, but its effectiveness is severely limited by gastrointestinal (GI) tract toxicity caused by gut bacterial β-glucuronidase (GUS) enzymes. Targeted bacterial GUS inhibitors have been shown to partially alleviate irinotecan-induced GI tract damage and resultant diarrhea in mice. Here, we unravel the mechanistic basis for GI protection by gut microbial GUS inhibitors using in vivo models.We use in vitro, in fimo, and in vivo models to determine whether GUS inhibition alters the anticancer efficacy of irinotecan. We demonstrate that a single dose of irinotecan increases GI bacterial GUS activity in 1 d and reduces intestinal epithelial cell proliferation in 5 d, both blocked by a single dose of a GUS inhibitor. In a tumor xenograft model, GUS inhibition prevents intestinal toxicity and maintains the antitumor efficacy of irinotecan. Remarkably, GUS inhibitor also effectively blocks the striking irinotecan-induced bloom of Enterobacteriaceae in immunedeficient mice. In a genetically engineered mouse model of cancer, GUS inhibition alleviates gut damage, improves survival, and does not alter gut microbial composition; however, by allowing dose intensification, it dramatically improves irinotecan’s effectiveness, reducing tumors to a fraction of that achieved by irinotecan alone, while simultaneously promoting epithelial regeneration. These results indicate that targeted gut microbial enzyme inhibitors can improve cancer chemotherapeutic outcomes by protecting the gut epithelium from microbial dysbiosis and proliferative crypt damage.

The gut microbiota harbor diverse β-glucuronidase (GUS) enzymes that liberate glucuronic acid (GlcA) sugars from small-molecule conjugates and complex carbohydrates. However, only the Enterobacteriaceae family of human gut-associated Proteobacteria maintain a GUS operon under the transcriptional control of a glucuronide repressor, GusR. Despite its potential importance in Escherichia, Salmonella, Klebsiella, Shigella, and Yersinia opportunistic pathogens, the structure of GusR has not been examined. Here, we explore the molecular basis for GusR-mediated regulation of GUS expression in response to small-molecule glucuronides. Presented are 2.1-Å-resolution crystal structures of GusRs from Escherichia coli and Salmonella enterica in complexes with a glucuronide ligand. The GusR-specific DNA operator site in the regulatory region of the E. coli GUS operon is identified, and structure-guided GusR mutants pinpoint the residues essential for DNA binding and glucuronide recognition. Interestingly, the endobiotic estradiol-17-glucuronide and the xenobiotic indomethacin-acyl-glucuronide are found to exhibit markedly differential binding to these GusR orthologs. Using structure-guided mutations, we are able to transfer E. coli GusR’s preferential DNA and glucuronide binding affinity to S. enterica GusR. Structures of putative GusR orthologs from GUS-encoding Firmicutes species also reveal functionally unique features of the Enterobacteriaceae GusRs. Finally, dominant-negative GusR variants are validated in cell-based studies. These data provide a molecular framework toward understanding the control of glucuronide utilization by opportunistic pathogens in the human gut.

Bacterial β-glucuronidase (GUS) enzymes cause drug toxicity by reversing Phase II glucuronidation in the gastrointestinal tract. While many human gut microbial GUS enzymes have been examined with model glucuronide substrates like p -nitrophenol-β-D-glucuronide ( p NPG), the GUS orthologs that are most efficient at processing drug-glucuronides remain unclear. Here we present the crystal structures of GUS enzymes from human gut commensals Lactobacillus rhamnosus , Ruminococcus gnavus , and Faecalibacterium prausnitzii that possess an active site loop (Loop 1; L1) analogous to that found in E. coli GUS, which processes drug substrates. We also resolve the structure of the No Loop GUS from Bacteroides dorei . We then compare the p NPG and diclofenac glucuronide processing abilities of a panel of twelve structurally diverse GUS proteins, and find that the new L1 GUS enzymes presented here process small glucuronide substrates inefficiently compared to previously characterized L1 GUS enzymes like E. coli GUS. We further demonstrate that our GUS inhibitors, which are effective against some L1 enzymes, are not potent towards all. Our findings pinpoint active site structural features necessary for the processing of drug-glucuronide substrates and the inhibition of such processing.

There has been considerable recent progress in protein structure prediction using deep neural networks to predict inter-residue distances from amino acid sequences 1 – 3 . Here we investigate whether the information captured by such networks is sufficiently rich to generate new folded proteins with sequences unrelated to those of the naturally occurring proteins used in training the models. We generate random amino acid sequences, and input them into the trRosetta structure prediction network to predict starting residue–residue distance maps, which, as expected, are quite featureless. We then carry out Monte Carlo sampling in amino acid sequence space, optimizing the contrast (Kullback–Leibler divergence) between the inter-residue distance distributions predicted by the network and background distributions averaged over all proteins. Optimization from different random starting points resulted in novel proteins spanning a wide range of sequences and predicted structures. We obtained synthetic genes encoding 129 of the network-‘hallucinated’ sequences, and expressed and purified the proteins in Escherichia coli ; 27 of the proteins yielded monodisperse species with circular dichroism spectra consistent with the hallucinated structures. We determined the three-dimensional structures of three of the hallucinated proteins, two by X-ray crystallography and one by NMR, and these closely matched the hallucinated models. Thus, deep networks trained to predict native protein structures from their sequences can be inverted to design new proteins, and such networks and methods should contribute alongside traditional physics-based models to the de novo design of proteins with new functions. The trRosetta neural network was used to iteratively optimise model proteins from random 100-amino-acid sequences, resulting in ‘hallucinated’ proteins, which when expressed in bacteria closely resembled the model structures.

The de novo design of small-molecule–binding proteins holds great promise as a potential tool to develop sensors on-demand for arbitrary small molecules. Here we combine deep learning and physics-based methods to generate a family of proteins with diverse and designable pocket geometries, which we employ to computationally design binders for six small-molecule targets. Biophysical characterization of the designed binders reveals nanomolar to low micromolar binding affinities and atomic-level design accuracy. Additionally, we use a cortisol binder to design a chemically induced dimerization (CID) system that enables the construction of a biosensor for cortisol detection. The approach described here demonstrates the potential of the NTF2 fold and deep learning-based protein design in sensor development, paving the way for future platforms to design binders and sensors for small molecules across analytical, environmental, and biomedical applications.

Modeling the conformational heterogeneity of protein-small molecule interactions is important for understanding natural systems and evaluating designed systems, but remains an outstanding challenge. We reasoned that while residue level descriptions of biomolecules are efficient for de novo structure prediction, for probing heterogeneity of interactions with small molecules in the folded state an entirely atomic level description could have advantages in speed and generality. We developed a graph neural network called PLACER (Protein-Ligand Atomistic Conformational Ensemble Resolver) trained to recapitulate correct atomic positions from partially corrupted input structures from the Cambridge Structural Database and the Protein Data Bank; the nodes of the graph are the atoms in the system. PLACER accurately generates structures of diverse organic small molecules given knowledge of their atom composition and bonding, and given a description of the larger protein context, builds up structures of small molecules and protein side chains for protein-small molecule docking. Because PLACER is rapid and stochastic, ensembles of predictions can be readily generated to map conformational heterogeneity. In enzyme design efforts described here and elsewhere, we find that using PLACER to assess the accuracy and pre-organization of the designed active sites results in higher success rates and higher activities; we obtain a preorganized retroaldolase with a / of 11000 M min , considerably higher than any pre-deep learning design for this reaction. We anticipate that PLACER will be widely useful for rapidly generating conformational ensembles of small molecule and small molecule-protein systems, and for designing higher activity preorganized enzymes.

Despite advances in de novo enzyme design, success has been largely limited to low energy barrier model reactions. Amide bonds such as those linking amino acids along the peptide backbone are stable for hundreds of years in neutral aqueous solution because of the high energy barrier to hydrolysis . Here we describe the use of a new deep learning method, RFD2-MI , to de novo design enzymes which utilize an activated cysteine nucleophile to hydrolyze the polypeptide backbone in a sequence-dependent manner, achieving rate enhancements over the background reaction ( / ) of up to 3 × 10 . The generated designs have folds very different from the proteases in nature (TM score < 0.50), and crystal structures are very close to the design models (Cα RMSDs < 1.2 Å), highlighting the accuracy of the design methodology. Our approach has broad utility for advancing the design of novel proteases for both biotechnical and medical applications.

Although significant progress has been made in creating metalloenzymes that hydrolyze activated esters , the energetically demanding cleavage of amide bonds has remained a major challenge for enzyme design: amide bonds are significantly more stable than ester bonds, the amine leaving groups in proteins are not activated, and peptide substrates are flexible making them difficult to bind precisely. Here, we report the design of zinc proteases from minimal catalytic motifs using a fine-tuned version of RoseTTAFold Diffusion 2, called RoseTTAFold Diffusion 2 for Molecular Interfaces , optimized for both enzyme and protein-protein interaction design. In a single one-shot design round of 135 designs, 36% of the designs had activity and cleaved precisely at the intended site. The most active design accelerated peptide bond hydrolysis more than 10 -fold over the uncatalyzed reaction . These results demonstrated that enzyme design has advanced well beyond model reactions with activated substrates, and open the door to design of proficient metallohydrolases for medicine and bioremediation.