De novo design of luciferases using deep learning.

Yeh AH, Norn C, Kipnis Y, Tischer D, Pellock SJ, Evans D, Ma P, Lee GR, Zhang JZ, Anishchenko I, Coventry B, Cao L, Dauparas J, Halabiya S, DeWitt M, Carter L, Houk KN, Baker D. · 2023

Read De novo design of luciferases using deep learning. by Yeh AH, Norn C, Kipnis Y, Tischer D, Pellock SJ, Evans D, Ma P, Lee GR, Zhang JZ, Anishchenko I, Coventry B, Cao L, Dauparas J, Halabiya S, DeWitt M, Carter L, Houk KN, Baker D. (2023) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc9946828.txt.

Introduction

Yeh and co-authors design new luciferase enzymes from scratch with deep learning. Classic enzyme design placed active sites into natural protein scaffolds. Few good scaffolds and complex sequence rules limited that route. The team used a family-wide hallucination method instead.

The method generates many idealized protein shapes with varied pocket forms. Matching sequences are designed to encode those shapes. The team built luciferases for two synthetic luciferin fuels. Their active sites set an arginine group beside the reaction site with tight shape fit. The best enzyme was small, at 13.9 kDa. It stayed stable above 95 °C and showed high selectivity. The work shows deep learning can create working enzymes to order.

Catalog record

CREATORYeh AH, Norn C, Kipnis Y, Tischer D, Pellock SJ, Evans D, Ma P, Lee GR, Zhang JZ, Anishchenko I, Coventry B, Cao L, Dauparas J, Halabiya S, DeWitt M, Carter L, Houk KN, Baker D.
YEAR2023
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC9946828

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

De novo design of luciferases using deep learning By Andy Hsien-Wei Yeh, Christoffer Norn, Yakov Kipnis, Doug Tischer, Samuel J. Pellock, Declan Evans, Pengchen Ma, Gyu Rie Lee, Jason Z. Zhang, Ivan Anishchenko, Brian Coventry, Longxing Cao Published in: Nature DOI: 10.1038/s41586-023-05696-3 ABSTRACT 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 scaffolds1,2, 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 diphenylterazine3 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 (kcat/Km = 106 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. MAIN Bioluminescent light produced by the enzymatic oxidation of a luciferin substrate by luciferases is widely used for bioassays and imaging in biomedical research. Because no excitation light source is needed, luminescent photons are produced in the dark; this results in higher sensitivity than fluorescence imaging in live animal models and in biological samples in which autofluorescence or phototoxicity is a concern4,5. However, the development of luciferases as molecular probes has lagged behind that of well-developed fluorescent protein toolkits for a number of reasons: (i) very few native luciferases have been identified6,7; (ii) many of those that have been identified require multiple disulfide bonds to stabilize the structure and are therefore prone to misfolding in mammalian cells8; (iii) most native luciferases do not recognize synthetic luciferins with more desirable photophysical properties9; and (iv) multiplexed imaging to follow multiple processes in parallel using mutually orthogonal luciferase–luciferin pairs has been limited by the low substrate specificity of native luciferases10,11. We sought to use de novo protein design to create luciferases that are small, highly stable, well-expressed in cells, specific for one substrate and need no cofactors to function. We chose a synthetic luciferin, diphenylterazine (DTZ), as the target substrate because of its high quantum yield, red-shifted emission3, favourable in vivo pharmacokinetics12,13 and lack of required cofactors for light emission. Previous computational enzyme design efforts have primarily repurposed native protein scaffolds in the Protein Data Bank (PDB)1,2, but there are few native structures with binding pockets appropriate for DTZ, and the effects of sequence changes on native proteins can be unpredictable (designed helical bundles have also been used as enzyme scaffolds14–16, but these are limited in number and most do not have pockets that are suitable for DTZ binding). To circumvent these limitations, we set out to generate large numbers of small and stable protein scaffolds with pockets of the appropriate size and shape for DTZ, and with clear sequence–structure relationships to facilitate subsequent active-site incorporation. To identify protein folds that are capable of hosting such pockets, we first docked DTZ into 4,000 native small-molecule-binding proteins. We found that many nuclear transport factor 2 (NTF2)-like folds have binding pockets with appropriate shape complementarity and size for DTZ placement (pink dashes in Fig. 1e), and hence selected the NTF2-like superfamily as the target topology.Fig. 1Generation of idealized scaffolds and computational design of de novo luciferases.a, Family-wide hallucination. Sequences encoding proteins with the desired topology are optimized by Markov chain Monte Carlo (MCMC) sampling with a multicomponent loss function. Structurally conserved regions (peach) are evaluated on the basis of consistency with input residue–residue distance and orientation distributions obtained from 85 experimental structures of NTF2-like proteins, whereas variable non-ideal regions (teal) are evaluated on the basis of the confidence of predicted inter-residue geometries calculated as the KL divergence between network predictions and the background distribution. The sequence-space MCMC sampling incorporates both sequence changes and insertions and deletions (see Supplementary Methods) to guide the hallucinated sequence towards encoding structures with the desired folds. Hydrogen-bonding networks are incorporated into the designed structures to increase structural specificity. b–d, The design of luciferase active sites. b, Generation of luciferase substrate (DTZ) conformers. c, Generation of a Rotamer Interaction Field (RIF) to stabilize anionic DTZ and form hydrophobic packing interactions. d, Docking of the RIF into the hallucinated scaffolds, and optimization of substrate–scaffold interactions using position-specific score matrices (PSSM)-biased sequence design. e, Selection of the NTF2 topology. The RIF was docked into 4,000 native small-molecule-binding proteins, excluding proteins that bind the luciferin substrate using more than five loop residues. Most of the top hits were from the NTF2-like protein superfamily (pink dashes). Using the family-wide hallucination scaffold generation protocol, we generated 1,615 scaffolds and found that these yielded better predicted RIF binding energies than the native proteins. f,g, Our DL-optimized scaffolds sample more within the space of the native structures (f) and have stronger sequence-to-structure relationships (more confident Alphafold2 structure predictions) (g) than native or previous non-deep-learning energy-optimized scaffolds. FAMILY-WIDE HALLUCINATION Native NTF2 structures have a range of pocket sizes and shapes but also contain features that are not ideal, such as long loops that compromise stability. To create large numbers of ideal NTF2-like structures, we developed a deep-learning-based ‘family-wide hallucination’ approach that integrates unconstrained de novo design17,18 and Rosetta sequence-design approaches19 to enable the generation of an essentially unlimited number of proteins that have a desired fold (Fig. 1a). The family-wide hallucination approach used the de novo sequence and structure discovery capability of unconstrained protein hallucination17,18 for loop and variable regions, and structure-guided sequence optimization for core regions. We used the trRosetta structure prediction neural network20, which is effective in identifying experimentally successful de-novo-designed proteins and hallucinating new globular proteins of diverse topologies. Starting from the sequences of 2,000 naturally occurring NTF2s, we carried out Monte Carlo searches in sequence space, at each step making a sequence change and predicting the structure using trRosetta. As the loss function guiding search, we used the confidence of the neural network in the predicted structure (as in our previous free hallucination study) supplemented with a topology-specific loss function over core residue pair geometries (see Supplementary Methods); in the loop regions, we also allowed the number of residues to vary, which resulted in short near ideal loops. To further encode structural specificity, we incorporated buried, long-range hydrogen-bonding networks. The resulting 1,615 family-wide hallucinated NTF2 scaffolds provided more shape-complementary binding pockets for DTZ than did native small-molecule-binding proteins (Fig. 1e). This method samples protein backbones that are closer to native NTF2-like proteins (Fig. 1f) and that have better scaffold quality metrics than those produced in a previous non-deep-learning energy-based approach21 (Fig. 1g). DE NOVO DESIGN OF LUCIFERASES FOR

Read the complete text