Predicting how mutations affect protein stability and protein-protein binding affinity is crucial for protein engineering and drug development. Although several computational tools have been developed for these tasks, they often require specialized expertise and are difficult to integrate into unified workflows. Here, we present PythiaStudio (https://pythiastudio.wulab.xyz), a comprehensive web platform that integrates our recently developed Pythia, Pythia-PPI, and Pythia-Pocket models with complementary protein analysis tools. The platform enables users to predict mutational effects on protein stability and protein-protein binding affinity, ligand binding pocket, through an intuitive interface. Additional features include fitness and structure prediction. PythiaStudio provides interactive visualization tools, including mutation heatmaps, sortable result tables, and structure viewers. Importantly, the platform offers an integrated engineering workflow that combines stability and fitness predictions to guide rational protein design. We demonstrate the utility of this workflow through multiple cases, including different glycoside hydrolases and amidases. In these cases, the two-step computational redesign strategy successfully improved both thermostability and catalytic activity. PythiaStudio democratizes access to state-of-the-art deep learning-based protein engineering methods, enabling researchers without computational expertise to perform sophisticated protein engineering.

Heme oxidase-like diiron oxidases (HDOs) constitute an expanding superfamily with intricate and diverse catalytic functions. Among them, AetD has been identified as a nitrile-forming enzyme, although its catalytic mechanism remains poorly understood. In this study, AetDs unexpected reaction pathway promiscuity is demonstrated, showcasing its ability to convert (bromine-substituted) l-tryptophan into not only indole-3-carbonitrile but also a range of 3 '-C functionalized indole derivatives, including amine, alcohol, aldehyde, and carboxylic acid. Through the analysis of product distributions, interconversions, isotope-labeling experiments, and QM/MM simulations, it is revealed that the highly reactive diiron (III, IV) X species, featuring an Fe2(IV)=O moiety, is capable of initiating hydrogen atom transfer from distinct sites on the substrate. These findings highlight how the flexible diiron structure of AetD enables diverse functional group transformations, establishing this enzyme as a powerful catalytic tool for complex and challenging biotransformations.

The active hydroxyl group of cellulose plays a crucial role in regulating the microstructure of cellulose-derived hard carbon, which ultimately affects its sodium storage capacity. Through small-angle X-ray scattering (SAXS) and X-ray atomic pair distribution function (PDF) analysis, we proved that modification of cellulose by esterification crosslinking can introduce more closed pores into the carbonized hard carbon, which is beneficial for promoting sodium ion storage. Our results demonstrate that by optimizing the conditions used for esterification cross-linking modification, the sodium storage capacity of cellulose-derived hard carbon could be increased from 254 to 348 mAh g−1, with an increase in plateau capacity from 140 to 230 mAh g−1. This study makes a significant contribution towards establishing industrial applications for cellulose-derived hard carbon.

Polyethylene terephthalate (PET) is a widely used thermoplastic material that contributes significantly to global plastic pollution. To address the pressing need to recycle fossil-derived PET and the critical importance of PET biodepolymerization and recycling technologies as promising green solutions, researchers worldwide are actively developing novel PET-degrading biocatalysts and efficient processes. The advancement of PET depolymerases and PET-degrading microorganisms is regarded as a key aspect of this study. Current studies primarily focus on the exploration, development, and enhancement of effective enzymes and strains, involving their isolation and screening from nature, along with protein design and engineering as informed by the elucidation of enzymatic mechanisms. Significant efforts have been dedicated toward process optimization to enhance the PET hydrolysis reaction. However, translating these laboratory findings into real-world applications remains a significant challenge that is contingent on the applicability of the developed biocatalysts and processes under industrial conditions. This review summarizes the cutting-edge foundational outcomes in the field of PET biodepolymerization and discusses the current challenges and potential solutions to advance PET depolymerization and, ultimately, PET recycling.

Recycling thermoset polyurethanes is hindered by their cross-linked structures and chemically stable urethane bonds. Although chemo-enzymatic approaches offer promise, known urethanases remain inefficient under industrial glycolysis conditions. Here, we present GRASE [graph neural network (GNN)-based recommendation of active and stable enzymes], a GNN-based framework that integrates self-supervised and supervised learning to identify efficient, glycolysis-compatible urethanases. Among these, AbPURase exhibited two orders of magnitude greater activity than previously known enzymes in 6 molar diethylene glycol, enabling near-complete depolymerization of commercial polyurethane at kilogram scale within 8 hours. Structural analysis revealed that a tightly packed hydrophobic core and proline-stabilized lid loop may confer AbPURase's stability and efficiency in harsh solvents. This work highlights how deep learning accelerates the discovery of biocatalysts with industrial potential and addresses a critical barrier in polyurethane recycling.

Protein C-terminal functionalization is pivotal for the development of therapeutic peptides and in chemical protein synthesis. Conventional traceless functionalization strategies rely on the use of inteins, which are often prone to premature hydrolysis. Herein, we repurposed a computationally redesigned peptide amidase variant (termed PAM27) into a glycine-specific carboxypeptidase, enabling traceless protein C-terminal functionalization via cleavage of a glycine residue and conjugation with amines. Twenty computationally designed stabilizing mutations were incorporated for enhancing enzyme stability, while seven substitutions aimed at reducing steric hindrance expanded the acyl acceptor channel, enabling the transit of glycine. PAM27 was found to exhibit stringent specificity for P1 ' glycine and broad tolerance for diverse P1 residues as well as nucleophilic agents, facilitating its integration with peptiligase-catalyzed ligation and bio-orthogonal chemical conjugation. PAM27 serves as a promising platform for developing versatile enzymes capable of catalyzing traceless C-terminal modification of various protein or peptide targets with diverse functional moieties.

Biotechnological strategies for plastic depolymerization and recycling have emerged as transformative approaches to combat the global plastic pollution crisis, aligning with the principles of a sustainable and circular economy. Despite advances in engineering PET hydrolases, the degradation process is frequently compromised by product inhibition and the heterogeneity of final products, thereby obstructing subsequent PET recondensation and impeding the synthesis of high-value derivatives. In this work, we utilized previously devised computational strategies to redesign a thermostable DuraMHETase, achieving an apparent melting temperature of 72 degrees C in complex with MHET and a 6-fold higher in total turnover number (TTN) toward MHET than the wild-type enzyme at 60 degrees C. The fused enzyme system composed of DuraMHETase and TurboPETase demonstrated higher efficiency than other PET hydrolases and the separated dual enzyme systems. Furthermore, we identified both exo-and endo-PETase activities in DuraMHETase, whereas the endo-activity was previously unobserved at ambient temperatures. These results expand the functional scope of MHETase beyond mere intermediate hydrolysis, and may provide guidance for the development of more synergistic approaches to plastic biodepolymerization and recycling. (c) 2025, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved.

CRISPR/Cas12a, a promising gene editing technology, faces limitations due to its requirement for a thymine (T)-rich protospacer adjacent motif (PAM). Despite the development of Cas12a variants with expanded PAM profiles, many genomic loci, especially those with guanine-cytosine (GC)-rich PAMs, have remained inaccessible. This study develops a small RNA toxin-aided strategy to evolve ErCas12a for targeting GC-rich PAMs, resulting in the creation of enhanced ErCas12a (enErCas12a). EnErCas12a demonstrates the ability to recognize GC-rich PAMs and target five times more PAM sequences than the wild-type ErCas12a. Furthermore, enErCas12a achieves efficient gene editing in both bacterial and mammalian cells at various sites with non-canonical PAMs, including GC-rich PAMs such as GCCC, CGCC, and GGCC, which are inaccessible to previous Cas12a variants. Moreover, enErCas12a effectively targets PAM sequences with a GC content exceeding 75% in mammalian cells, providing a valuable alternative to the existing Cas12a toolkit. Importantly, enErCas12a maintains high specificity at targets with canonical PAMs, while also demonstrating enhanced specificity at targets with non-canonical PAMs. Collectively, this work establishes enErCas12a as a promising tool for gene editing in both eukaryotes and prokaryotes.

Predicting free energy changes (ΔΔG) is essential for enhancing our understanding of protein evolution and plays a pivotal role in protein engineering and pharmaceutical development. While traditional methods offer valuable insights, they are often constrained by computational speed and reliance on biased training datasets. These constraints become particularly evident when aiming for accurate ΔΔG predictions across a diverse array of protein sequences. Herein, we introduce Pythia, a self-supervised graph neural network specifically designed for zero-shot ΔΔG predictions. Our comparative benchmarks demonstrate that Pythia outperforms other self-supervised pretraining models and force field-based approaches while also exhibiting competitive performance with fully supervised models. Notably, Pythia shows strong correlations and achieves a remarkable increase in computational speed of up to 105-fold. We further validated Pythia’s performance in predicting the thermostabilizing mutations of limonene epoxide hydrolase, leading to higher experimental success rates. This exceptional efficiency has enabled us to explore 26 million high-quality protein structures, marking a significant advancement in our ability to navigate the protein sequence space and enhance our understanding of the relationships between protein genotype and phenotype. In addition, we established a web server at https://pythia.wulab.xyz to allow users to easily perform such predictions.

The growing demand for biocatalysts in biomass processing highlights the necessity of enhancing the thermostability of glycoside hydrolases. However, improving both thermostability and activity is often hindered by trade-offs between backbone rigidity and the flexibility of substrate-binding regions. In this study, Bacillus subtilis cellulase and β-glucanase were engineered using a two-step process incorporating the computational tools Pythia and ESM-2, which were found complementary in improving stability and activity. The engineered cellulase and β-glucanase exhibited increases in their apparent melting temperatures (5.8 °C and 8.4 °C), accompanied by up to a 1.5-fold increase in initial activities. At 50 °C, while the wild-type cellulase lost 60% of its activity after 24 h and wild-type β-glucanase lost activity completely in 2 h, the engineered cellulase-M5 retained its initial activity, and β-glucanase-M7 displayed a 2.2-fold increase in its half-life. Structural analysis indicated that Pythia-identified mutations likely enhanced backbone robustness through refined polar and hydrophobic interactions, while beneficial mutations from ESM-2 appeared to affect polysaccharide-binding regions. This two-step computational redesign offers a promising approach for optimizing both thermostability and activity in glycoside hydrolases and other enzyme families with extensive sequence diversity.

The utilization of polyethylene terephthalate (PET) has caused significant and prolonged ecological repercussions. Enzymatic degradation is an environmentally friendly approach to addressing PET contamination. Hydrolysis of mono(2-hydroxyethyl) terephthalate (MHET), a competitively inhibited intermediate in PET degradation, is catalyzed by MHET degrading enzymes. Herein, we employed bioinformatic methods that combined with sequence and structural information to discover an MHET hydrolase, BurkMHETase. Enzymatic characterization showed that the enzyme was relatively stable at pH 7.5-10.0 and 30-45 ℃. The kinetic parameters kcat and Km on MHET were (24.2±0.5)/s and (1.8±0.2) μmol/L, respectively, which were similar to that of the well-known IsMHETase with higher substrate affinity. BurkMHETase coupled with PET degradation enzymes improved the degradation of PET films. Structural analysis and mutation experiments indicated that BurkMHETase may have evolved specific structural features to hydrolyze MHET. For MHET degrading enzymes, aromatic amino acids at position 495 and the synergistic interactions between active sites or distal amino acids appear to be required for MHET hydrolytic activity. Therefore, BurkMHETase may have substantial potential in a dual-enzyme PET degradation system while the bioinformatic methods can be used to broaden the scope of applicable MHETase enzymes.

Bacterial small molecule metabolites such as adenosine-diphosphate- d - glycero -β- d - manno -heptose (ADP-heptose) and their derivatives act as effective innate immune agonists in mammals. We show that functional nucleotide-diphosphate-heptose biosynthetic enzymes (HBEs) are distributed widely in bacteria, archaea, eukaryotes, and viruses. We identified a conserved STT R5 motif as a hallmark of heptose nucleotidyltransferases that can synthesize not only ADP-heptose but also cytidine-diphosphate (CDP)– and uridine-diphosphate (UDP)–heptose. Both CDP- and UDP-heptoses are agonists that trigger stronger alpha-protein kinase 1 (ALPK1)–dependent immune responses than ADP-heptose in human and mouse cells and mice. We also produced ADP-heptose in archaea and verified its innate immune agonist functions. Hence, the β- d - manno -heptoses are cross-kingdom, small-molecule, pathogen-associated molecular patterns that activate the ALPK1-dependent innate immune signaling cascade.

Poly(3-hydroxybutyrate- co -3-hydroxyvalerate) (PHBV) is a type of polyhydroxyalkanoates (PHA) that exhibits numerous outstanding properties and is naturally synthesized and elaborately regulated in various microorganisms. However, the regulatory mechanism involving the specific regulator PhaR in Haloferax mediterranei , a major PHBV production model among Haloarchaea, is not well understood. In our previous study, we showed that deletion of the phosphoenolpyruvate (PEP) synthetase-like ( pps -like) gene activates the cryptic phaC genes in H. mediterranei , resulting in enhanced PHBV accumulation. In this study, we demonstrated the specific function of the PPS-like protein as a negative regulator of phaR gene expression and PHBV synthesis. Chromatin immunoprecipitation (ChIP), in situ fluorescence reporting system, and in vitro electrophoretic mobility shift assay (EMSA) showed that the PPS-like protein can bind to the promoter region of phaRP . Computational modeling revealed a high structural similarity between the rifampin phosphotransferase (RPH) protein and the PPS-like protein, which has a conserved ATP-binding domain, a His domain, and a predicted DNA-binding domain. Key residues within this unique DNA-binding domain were subsequently validated through point mutation and functional evaluations. Based on these findings, we concluded that PPS-like protein, which we now renamed as PspR, has evolved into a repressor capable of regulating the key regulator PhaR, and thereby modulating PHBV synthesis. This regulatory network (PspR-PhaR) for PHA biosynthesis is likely widespread among haloarchaea, providing a novel approach to manipulate haloarchaea as a production platform for high-yielding PHA. Key points • The repressive mechanism of a novel inhibitor PspR in the PHBV biosynthesis was demonstrated • PspR is widespread among the PHA accumulating haloarchaea • It is the first report of functional conversion from an enzyme to a trans-acting regulator in haloarchaea

We have developed the GReedy Accumulated strategy for Protein Engineering (GRAPE) to improve enzyme stability across various applications, combining advanced computational methods with a unique clustering and greedy accumulation approach to efficiently explore epistatic effects with minimal experimental effort. To make this strategy accessible to nonexperts, we introduced GRAPE-WEB, an automated, user-friendly web server that allows the design, inspection, and combination of stabilizing mutations without requiring extensive bioinformatics knowledge. GRAPE-WEB's robust performance and accessibility provide a comprehensive and adaptable approach to protein thermostability design, suitable for both newcomers and experienced practitioners in the field. The web server is accessible at https://grape.wulab.xyz.

Biotechnological plastic recycling has emerged as a suitable option for addressing the pollution crisis. A major breakthrough in the biodegradation of poly(ethylene terephthalate) (PET) is achieved by using a LCC variant, which permits 90% conversion at an industrial level. Despite the achievements, its applications have been hampered by the remaining 10% of nonbiodegradable PET. Herein, we address current challenges by employing a computational strategy to engineer a hydrolase from the bacterium HR29. The redesigned variant, TurboPETase, outperforms other well-known PET hydrolases. Nearly complete depolymerization is accomplished in 8 h at a solids loading of 200 g kg −1 . Kinetic and structural analysis suggest that the improved performance may be attributed to a more flexible PET-binding groove that facilitates the targeting of more specific attack sites. Collectively, our results constitute a significant advance in understanding and engineering of industrially applicable polyester hydrolases, and provide guidance for further efforts on other polymer types.

Abstract Biotechnological plastic depolymerization and recycling have emerged as suitable options for addressing the plastic-waste pollution crisis in a circular plastic economy. Enzymatic degradation of poly(ethylene terephthalate) (PET), as the most typical representative, has evolved over the past two decades, with a major breakthrough achieved by using the LCC variant that permitted 90% conversion of PET on an industrial scale. Despite the achievements, the last 10% residual PET becomes nonbiodegradable due to physical aging, which has hampered its application in real industrial scenarios. In the present study, we addressed current challenges by employing a computational strategy that incorporates a protein language model and force-field-based algorithms to engineer a hydrolase from the bacterium HR29. The redesigned variant, TurboPETase, outperformed all the PET hydrolases reported thus far with regard to industrial application, enabling nearly 100% depolymerization of untreated PET containers, pretreated postconsumer PET bottles and their lower-grade products. The full degradation of pretreated PET at high industrially relevant scales (up to 300 g L-1) can be accomplished in as little as 10 h, with a maximum production rate of 77.3 gTPAeq L 1 h-1, demonstrating great potential for enzymatic PET recycling. Kinetic parameters derived from the inverse Michaelis‒Menten model and structural analysis suggest that the improved depolymerization performance may be attributed to a more flexible PET-binding groove that facilitates the targeting of more specific attack sites. Collectively, our results constitute a significant advance in the understanding and engineering of effective industrially applicable polyester hydrolases and provide guidance for further efforts on other mass-produced polymer types in this intriguing research field.

Predicting free energy changes (ΔΔG) is of paramount significance in advancing our comprehension of protein evolution and holds profound implications for protein engineering and pharmaceutical development. Traditional methods, however, often suffer from limitations such as sluggish computational speed or heavy reliance on biased training datasets. These challenges are magnified when aiming for accurate ΔΔG prediction across the vast universe of protein sequences. In this study, we present Pythia, a self-supervised graph neural network tailored for zero-shot ΔΔG predictions. In comparative benchmarks with other self-supervised pre-training models and force field-based methods, Pythia outshines its contenders with superior correlations while operating with the fewest parameters, and exhibits a remarkable acceleration in computational speed, up to 10 5 -fold. The efficacy of Pythia is corroborated through its application in predicting thermostable mutations of limonene epoxide hydrolase (LEH) with significant higher experimental success rates. This efficiency propels the exploration of 26 million high-quality protein structures. Such a grand-scale application signifies a leap forward in our capacity to traverse the protein sequence space and potentially enrich our insights into the intricacies of protein genotype-phenotype relationships. We provided a web app at https://pythia.wulab.xyz for users to conveniently execute predictions. Keywords: self-supervised learning, protein mutation prediction, protein thermostability

从合成生物学的角度来认知生命，其本质是可数据化与可设计性。生命体中绝大多数的催化功能是由酶来实现的，因此催化元件是合成生物学中最核心的元件之一。序列决定构象，而构象则决定功能。基于空间结构的催化元件序列数字化设计，是合成生物学研究的重要热点和前沿领域。它既可为开发合成生物学功能器件，特别是全新化学催化器件提供大量原型分子，同时也为发展模块化、工程化调控元件提供设计模板和指导规律。文章针对近年来出现的生物元件，尤其是催化元件的前沿进展进行简要介绍。

The rising demand for regioselective protein modifications in chemical biology and pharmaceutical manufacturing has fueled efforts to develop diverse techniques that functionalize native amino acid residues. Although many powerful strategies have provided elegant solutions for functionalizing N termini and side chains, sequence-unconstrained versatile C terminal functionalization remains a challenge. Here, we report an engineered peptide amidase (PAM) for C-terminal traceless functionalization in aqueous solution with a broad spectrum of both nucleophiles and protein sequences, excellent yields (up to 98%), and good compatibility with the click reaction. Computational analysis suggested an expanded nucleophile pocket induced by the introduction of an unusual glycine-rich motif, which may enrich the structural diversity in protein design. We anticipate that the successfully engineered PAM holds great potential in the applications of protein chemistry and proteomics, and highlights the employment of serine hydrolases in catalyzing acyl shift reactions that compete with hydrolysis under aqueous conditions.