AI Prediction of Structural Stability of Nanoproteins Based on Structures and Residue Properties by Mean Pooled Dual Graph Convolutional Network.

The structural stability of proteins is an important topic in various fields such as biotechnology, pharmaceuticals, and enzymology. Specifically, understanding the structural stability of protein is crucial for protein design. Artificial design, while pursuing high thermodynamic stability and rigidity of proteins, inevitably sacrifices biological functions closely related to protein flexibility. The thermodynamic stability of proteins is not always optimal when they are highest to perfectly perform their biological functions. Extensive theoretical and experimental screening is often required to obtain stable protein structures. Thus, it becomes critically important to develop a stability prediction model based on the balance between protein stability and bioactivity. To design protein drugs with better functionality in a broader structural space, a novel protein structural stability predictor called PSSP has been developed in this study. PSSP is a mean pooled dual graph convolutional network (GCN) model based on sequence characteristics and secondary structure, distance matrix, graph, and residue properties of a nanoprotein to provide rapid prediction and judgment. This model exhibits excellent robustness in predicting the structural stability of nanoproteins. Comparing with previous artificial intelligence algorithms, the results indicate this model can provide a rapid and accurate assessment of the structural stability of artificially designed proteins, which shows the great promises for promoting the robust development of protein design.

Folds from fold: Exploring topological isoforms of a single-domain protein.

Expanding the protein fold space beyond linear chains is of fundamental significance, yet remains largely unexplored. Herein, we report the creation of seven topological isoforms (i.e., linear, cyclic, knot, lasso, pseudorotaxane, and catenane) from a single protein fold precursor by rewiring the connectivity of secondary structure elements of the SpyTag-SpyCatcher complex and mutating the reactive residue on SpyTag to abolish the isopeptide bonding. These topological isoforms can be directly expressed in cells. Their topologies were confirmed by combined techniques of proteolytic digestion, fluorescence correlation spectroscopy (FCS), size-exclusion chromatography (SEC), and topological transformation. To study the effects of topology on their structures and properties, their biophysical properties were characterized by differential scanning calorimetry (DSC), heteronuclear single quantum coherence nuclear magnetic resonance spectroscopy (HSQC-NMR), and circular dichroism (CD) spectroscopy. Molecular dynamics (MD) simulations were further performed to reveal the atomic details of structural changes upon unfolding. Both experimental and simulation results suggest that they share a similar, well-folded hydrophobic core but exhibit distinct folding/unfolding dynamic behaviors. These results shed light onto the folding landscape of topological isoforms derived from the same protein fold. As a model system, this work improves our understanding of protein structure and dynamics beyond linear chains and suggests that protein folds are highly amenable to topological variation.

Insights into the binding recognition and computational design of IL-2 muteins with enhanced predicted binding affinity to the IL-2 receptor α.

Interleukin-2 (IL-2) is an immune system regulator that has received approval for cancer treatment. However, high-dose IL-2 therapy has seen restricted use due to its low efficacy and on-target toxicity. To enhance the effectiveness of IL-2 therapy, it is essential to engineer IL-2 molecules to enhance their specificity toward target cell populations. In this study, molecular dynamics (MD) simulations and Rosetta software were utilized to design novel high-affinity IL-2Rα-binding IL-2 muteins. MD simulations were used to identify the target residues of IL-2 for design, and Rosetta software were then employed to predict potential IL-2 muteins with higher binding affinity toward IL-2Rα. Rosetta generated two potential designed IL-2 muteins. The results of the MD validation and MM/GBSA analysis indicated that both designed IL-2 muteins exhibited greater predicted binding affinities toward IL-2Rα than that of the native proteins. RMSF analysis demonstrated that the structural fluctuations of free IL-2 and designed muteins were similar, indicating that the mutations did not alter the intramolecular force responsible for IL-2's stability and folding. These designed IL-2 muteins may have potential benefits for cancer immunotherapy.

Exploring "dark-matter" protein folds using deep learning.

De novo protein design explores uncharted sequence and structure space to generate novel proteins not sampled by evolution. A main challenge in de novo design involves crafting "designable" structural templates to guide the sequence searches toward adopting target structures. We present a convolutional variational autoencoder that learns patterns of protein structure, dubbed Genesis. We coupled Genesis with trRosetta to design sequences for a set of protein folds and found that Genesis is capable of reconstructing native-like distance and angle distributions for five native folds and three novel, the so-called "dark-matter" folds as a demonstration of generalizability. We used a high-throughput assay to characterize the stability of the designs through protease resistance, obtaining encouraging success rates for folded proteins. Genesis enables exploration of the protein fold space within minutes, unrestricted by protein topologies. Our approach addresses the backbone designability problem, showing that small neural networks can efficiently learn structural patterns in proteins. A record of this paper's transparent peer review process is included in the supplemental information.

Computational Stabilization of a Non-heme Iron Enzyme Enables Efficient Evolution of New Function.

Deep learning tools for enzyme design are rapidly emerging, and there is a critical need to evaluate their effectiveness in engineering workflows. Here we show that the deep learning-based tool ProteinMPNN can be used to redesign Fe(II)/αKG superfamily enzymes for greater stability, solubility, and expression while retaining both native activity and industrially-relevant non-native functions. This superfamily has diverse catalytic functions and could provide a rich new source of biocatalysts for synthesis and industrial processes. Through systematic comparisons of directed evolution trajectories for a non-native, remote C(sp3)-H hydroxylation reaction, we demonstrate that the stabilized redesign can be evolved more efficiently than the wild-type enzyme. After three rounds of directed evolution, we obtained a 6-fold activity increase from the wild-type parent and an 80-fold increase from the stabilized variant. To generate the initial stabilized variant, we identified multiple structural and sequence constraints to preserve catalytic function. We applied these criteria to produce stabilized, catalytically active variants of a second Fe(II)/αKG enzyme, suggesting that the approach is generalizable to additional members of the Fe(II)/αKG superfamily. ProteinMPNN is user-friendly and widely-accessible, and our results provide a framework for the routine implementation of deep learning-based protein stabilization tools in directed evolution workflows for novel biocatalysts.

A Nonlinear Peptide Topology for Synthetic Virions.

a nonlinear de novo peptide topology for the assembly of synthetic virions is reported. The topology is a backbone cyclized amino-acid sequence in which polar l- and hydrophobic d-amino acid residues of the same-type alternate. This arrangement introduces pseudo C4 symmetries of side chains within the same cyclopeptide ring, allowing for the lateral propagation of cyclopeptides into networks with a [3/6, 4]-fold rotational symmetry closing into virus-like shells. A combination of computational and experimental approaches was used to establish that the topology forms morphologically uniform, nonaggregating and nontoxic nanoscale shells. These effectively encapsulate genetic cargo and promote its intracellular delivery and a target genetic response. The design introduces a nanotechnology inspired solution for engineering virus-like systems thereby expanding traditional molecular biology approaches used to create artificial biology to chemical space.

How the technologies behind self-driving cars, social networks, ChatGPT, and DALL-E2 are changing structural biology.

The performance of deep Neural Networks (NNs) in the text (ChatGPT) and image (DALL-E2) domains has attracted worldwide attention. Convolutional NNs (CNNs), Large Language Models (LLMs), Denoising Diffusion Probabilistic Models (DDPMs)/Noise Conditional Score Networks (NCSNs), and Graph NNs (GNNs) have impacted computer vision, language editing and translation, automated conversation, image generation, and social network management. Proteins can be viewed as texts written with the alphabet of amino acids, as images, or as graphs of interacting residues. Each of these perspectives suggests the use of tools from a different area of deep learning for protein structural biology. Here, I review how CNNs, LLMs, DDPMs/NCSNs, and GNNs have led to major advances in protein structure prediction, inverse folding, protein design, and small molecule design. This review is primarily intended as a deep learning primer for practicing experimental structural biologists. However, extensive references to the deep learning literature should also make it relevant to readers who have a background in machine learning, physics or statistics, and an interest in protein structural biology.

Revolutionizing Molecular Design for Innovative Therapeutic Applications through Artificial Intelligence.

The field of computational protein engineering has been transformed by recent advancements in machine learning, artificial intelligence, and molecular modeling, enabling the design of proteins with unprecedented precision and functionality. Computational methods now play a crucial role in enhancing the stability, activity, and specificity of proteins for diverse applications in biotechnology and medicine. Techniques such as deep learning, reinforcement learning, and transfer learning have dramatically improved protein structure prediction, optimization of binding affinities, and enzyme design. These innovations have streamlined the process of protein engineering by allowing the rapid generation of targeted libraries, reducing experimental sampling, and enabling the rational design of proteins with tailored properties. Furthermore, the integration of computational approaches with high-throughput experimental techniques has facilitated the development of multifunctional proteins and novel therapeutics. However, challenges remain in bridging the gap between computational predictions and experimental validation and in addressing ethical concerns related to AI-driven protein design. This review provides a comprehensive overview of the current state and future directions of computational methods in protein engineering, emphasizing their transformative potential in creating next-generation biologics and advancing synthetic biology.

Target-templated Construction of Functional Proteomimetics Using Photo-foldamer Libraries.

Current methods for proteomimetic engineering rely on structure-based design. Here we describe a design strategy that allows the construction of proteomimetics against challenging targets without a priori characterization of the target surface. Our approach relies on (i) a 100-membered photoreactive foldamer library, the members of which act as local surface mimetics, and (ii) the subsequent affinity maturation of the primary hits using systems chemistry. Two surface-oriented proteinogenic side chains drove the interactions between the short helical foldamer fragments and the proteins. Diazirine-based photo-crosslinking was applied to sensitively detected and localize binding even to shallow and dynamic patches on representatively difficult targets. Photo-foldamers identified functionally relevant protein interfaces, allosteric and previously unexplored targetable regions on the surface of STAT3 and an oncogenic K-Ras variant. Target-templated dynamic linking of foldamer hits resulted in two orders of magnitude affinity improvement in a single step. The dimeric K-Ras ligand mimicked protein-like catalytic functions. The photo-foldamer approach thus enables the highly efficient mapping of protein-protein interaction sites and provides a viable starting point for proteomimetic ligand development without a priori structural hypotheses.

Integrating Computational Design and Experimental Approaches for Next-Generation Biologics.

Therapeutic protein engineering has revolutionized medicine by enabling the development of highly specific and potent treatments for a wide range of diseases. This review examines recent advances in computational and experimental approaches for engineering improved protein therapeutics. Key areas of focus include antibody engineering, enzyme replacement therapies, and cytokine-based drugs. Computational methods like structure-based design, machine learning integration, and protein language models have dramatically enhanced our ability to predict protein properties and guide engineering efforts. Experimental techniques such as directed evolution and rational design approaches continue to evolve, with high-throughput methods accelerating the discovery process. Applications of these methods have led to breakthroughs in affinity maturation, bispecific antibodies, enzyme stability enhancement, and the development of conditionally active cytokines. Emerging approaches like intracellular protein delivery, stimulus-responsive proteins, and de novo designed therapeutic proteins offer exciting new possibilities. However, challenges remain in predicting in vivo behavior, scalable manufacturing, immunogenicity mitigation, and targeted delivery. Addressing these challenges will require continued integration of computational and experimental methods, as well as a deeper understanding of protein behavior in complex physiological environments. As the field advances, we can anticipate increasingly sophisticated and effective protein therapeutics for treating human diseases.

Development of artificial photosystems based on designed proteins for mechanistic insights into photosynthesis.

This review aims to provide an overview of the progress in protein-based artificial photosystem design and their potential to uncover the underlying principles governing light-harvesting in photosynthesis. While significant advances have been made in this area, a gap persists in reviewing these advances. This review provides a perspective of the field, pinpointing knowledge gaps and unresolved challenges that warrant further inquiry. In particular, it delves into the key considerations when designing photosystems based on the chromophore and protein scaffold characteristics, presents the established strategies for artificial photosystems engineering with their advantages and disadvantages, and underscores the recent breakthroughs in understanding the molecular mechanisms governing light-harvesting, charge separation, and the role of the protein motions in the chromophore's excited state relaxation. By disseminating this knowledge, this article provides a foundational resource for defining the field of bio-hybrid photosystems and aims to inspire the continued exploration of artificial photosystems using protein design.

Biomimetic androgen receptor-based AIE biosensor for detecting bisphenol analogues: An integrating in silico topological analysis, molecular docking, and experimental validation study.

Bisphenol analogues are the typical class of endocrine disrupting chemicals (EDCs) that interfere with binding of endogenous hormones to androgen receptor (AR). With the expansion of industrial activities and the intensification of environmental pollution, an increasing array of bisphenol analogues is being released into the environment and food chain. This highlights the urgency to develop sensitive methods for the detection of bisphenol analogues. Here, we propose a biomimetic AR-based biosensor platform for detecting bisphenol analogues (BPF, TBBPA, and TBBPS) by binding with Aggregation-Induced Emission (AIE) probes. Following a comparison of the PROSS and ABACUS methods, biomimetic AR was designed using the ABACUS approach and subsequently expressed in vitro via the E. coli expression system. Through molecular docking and the observation of fluorescence changes upon binding with biomimetic AR, BS-46006 was selected as the AIE probe for the biosensor. The biomimetic AR-based biosensor showed sensitive detections of BPF, TBBPA, and TBBPS within a range of 0-50 mM. To further elucidate the multi-residue recognition mechanism, molecular orbitals, Electron Localization Function (ELF), and Localized Orbital Locator (LOL) were systematically calculated in this study. Lowest unoccupied molecular orbital and highest occupied molecular orbital indicated the energy gap of BPF, TBBPA, and TBBPS, which correspond to 0.12812, 0.19689, and 0.18711 eV, respectively. ELF and LOL offered clearer perspective through heat maps to visually represent the electron delocalization in BPF, TBBPA, and TBBPS. The matrix effect analysis suggested that the responses of bisphenol analogues in soil matrices could be effectively mitigated through sample pretreatment. The analysis of spiked soil samples showed the acceptable recoveries ranged from 91 % to 105 %. Additionally, the biomimetic AR-based AIE biosensor, which combines multi-residue detection with Tolerable Daily Intakes, shows great promise for the risk assessment of bisphenol analogues. This research may present a viable approach for the analysis of environmental pollutants.

Engineering conditional protein-protein interactions for dynamic cellular control.

Conditional protein-protein interactions enable dynamic regulation of cellular activity and are an attractive approach to probe native protein interactions, improve metabolic engineering of microbial factories, and develop smart therapeutics. Conditional protein-protein interactions have been engineered to respond to various chemical, light, and nucleic acid-based stimuli. These interactions have been applied to assemble protein fragments, build protein scaffolds, and spatially organize proteins in many microbial and higher-order hosts. To foster the development of novel conditional protein-protein interactions that respond to new inputs or can be utilized in alternative settings, we provide an overview of the process of designing new engineered protein interactions while showcasing many recently developed computational tools that may accelerate protein engineering in this space.

All-Atom Protein Sequence Design Based on Geometric Deep Learning.

Designing sequences for specific protein backbones is a key step in creating new functional proteins. Here, we introduce GeoSeqBuilder, a deep learning framework that integrates protein sequence generation with side chain conformation prediction to produce the complete all-atom structures for designed sequences. GeoSeqBuilder uses spatial geometric features from protein backbones and explicitly includes three-body interactions of neighboring residues. GeoSeqBuilder achieves native residue type recovery rate of 51.6%, comparable to ProteinMPNN and  other leading methods, while accurately predicting side chain conformations. We first used GeoSeqBuilder to design sequences for thioredoxin and a hallucinated three-helical bundle protein. All the 15 tested sequences expressed as soluble monomeric proteins with high thermal stability, and the 2 high-resolution crystal structures solved closely match the designed models. The generated protein sequences exhibit low similarity (minimum 23%) to the original sequences, with significantly altered hydrophobic cores. We further redesigned the hydrophobic core of glutathione peroxidase 4, and 3 of the 5 designs showed improved enzyme activity. Although further testing is needed, the high experimental success rate in our testing demonstrates that GeoSeqBuilder is a powerful tool for designing novel sequences for predefined protein structures with atomic details. GeoSeqBuilder is available at https://github.com/PKUliujl/GeoSeqBuilder.

Transplanted Murine Tumours SPECT Imaging with 99mTc Delivered with an Artificial Recombinant Protein.

99mTc is a well-known radionuclide that is widely used and readily available for SPECT/CT (Single-Photon Emission Computed Tomography) diagnosis. However, commercial isotope carriers are not specific enough to tumours, rapidly clear from the bloodstream, and are not safe. To overcome these limitations, we suggest immunologically compatible recombinant proteins containing a combination of metal binding sites as 99mTc chelators and several different tumour-specific ligands for early detection of tumours. E1b protein containing metal-binding centres and tumour-specific ligands targeting integrin αvß3 and nucleolin, as well as a short Cys-rich sequence, was artificially constructed. It was produced in E. coli, purified by metal-chelate chromatography, and used to obtain a complex with 99mTc. This was administered intravenously to healthy Balb/C mice at an activity dose of about 80 MBq per mouse, and the biodistribution was studied by SPECT/CT for 24 h. Free sodium 99mTc-pertechnetate at the same dose was used as a reference. The selectivity of 99mTc-E1b and the kinetics of isotope retention in tumours were then investigated in experiments in C57Bl/6 and Balb/C mice with subcutaneously transplanted lung carcinoma (LLC) or mammary adenocarcinoma (Ca755, EMT6, or 4T1). The radionuclide distribution ratio in tumour and adjacent normal tissue (T/N) steadily increased over 24 h, reaching 15.7 ± 4.2 for EMT6, 16.5 ± 3.8 for Ca755, 6.7 ± 4.2 for LLC, and 7.5 ± 3.1 for 4T1.

Metals at the Helm: Revolutionizing Protein Assembly and Applications.

Protein assembly is an essential process in biological systems, where proteins self-assemble into complex structures with diverse functions. Inspired by the exquisite control over protein assembly in nature, scientists have been exploring ways to design and assemble protein structures with precise control over their topologies and functions. One promising approach for achieving this goal is through metal coordination, which utilizes metal-binding motifs to mediate protein-protein interactions and assemble protein complexes with controlled stoichiometry and geometry. Metal coordination provides a modular and tunable approach for protein assembly and de novo structure design, where the metal ion acts as a molecular glue that holds the protein subunits together in a specific orientation. Metal-coordinated protein assemblies have shown great potential for developing functional metalloproteinase, novel biomaterials and integrated drug delivery systems. In this review, an overview of the recent advances in protein assemblies benefited from metal coordination is provided, focusing on various protein arrangements in different dimensions including protein oligomers, protein nanocage and higher-order protein architectures. Moreover, the key metal-binding motifs and strategies used to assemble protein structures with precise control over their properties are highlighted. The potential applications of metal-mediated protein assemblies in biotechnology and biomedicine are also discussed.

Protein stability models fail to capture epistatic interactions of double point mutations.

There is strong interest in accurate methods for predicting changes in protein stability resulting from amino acid mutations to the protein sequence. Recombinant proteins must often be stabilized to be used as therapeutics or reagents, and destabilizing mutations are implicated in a variety of diseases. Due to increased data availability and improved modeling techniques, recent studies have shown advancements in predicting changes in protein stability when a single point mutation is made. Less focus has been directed toward predicting changes in protein stability when there are two or more mutations, despite the significance of mutation clusters for disease pathways and protein design studies. Here, we analyze the largest available dataset of double point mutation stability and benchmark several widely used protein stability models on this and other datasets. We identify a blind spot in how predictors are typically evaluated on multiple mutations, finding that, contrary to assumptions in the field, current stability models are unable to consistently capture epistatic interactions between double mutations. We observe one notable deviation from this trend, which is that epistasis-aware models provide marginally better predictions on stabilizing double point mutations. We develop an extension of the ThermoMPNN framework for double mutant modeling as well as a novel data augmentation scheme which mitigates some of the limitations in available datasets. Collectively, our findings indicate that current protein stability models fail to capture the nuanced epistatic interactions between concurrent mutations due to several factors, including training dataset limitations and insufficient model sensitivity.

Inventing Novel Protein Folds.

The vastness of unexplored protein fold universe remains a significant question. Through systematic de novo design of proteins with novel αß-folds, we demonstrated that nature has only explored a tiny portion of the possible folds. Numerous possible protein folds are still untouched by nature. This review outlines this study and discusses the prospects for design of functional proteins with novel folds.

Synthetic directed evolution for targeted engineering of plant traits.

Improving crop traits requires genetic diversity, which allows breeders to select advantageous alleles of key genes. In species or loci that lack sufficient genetic diversity, synthetic directed evolution (SDE) can supplement natural variation, thus expanding the possibilities for trait engineering. In this review, we explore recent advances and applications of SDE for crop improvement, highlighting potential targets (coding sequences and cis-regulatory elements) and computational tools to enhance crop resilience and performance across diverse environments. Recent advancements in SDE approaches have streamlined the generation of variants and the selection processes; by leveraging these advanced technologies and principles, we can minimize concerns about host fitness and unintended effects, thus opening promising avenues for effectively enhancing crop traits.

PB-GPT: An innovative GPT-based model for protein backbone generation.

With advanced computational methods, it is now feasible to modify or design proteins for specific functions, a process with significant implications for disease treatment and other medical applications. Protein structures and functions are intrinsically linked to their backbones, making the design of these backbones a pivotal aspect of protein engineering. In this study, we focus on the task of unconditionally generating protein backbones. By means of codebook quantization and compression dictionaries, we convert protein backbone structures into a distinctive coded language and propose a GPT-based protein backbone generation model, PB-GPT. To validate the generalization performance of the model, we trained and evaluated the model on both public datasets and small protein datasets. The results demonstrate that our model has the capability to unconditionally generate elaborate, highly realistic protein backbones with structural patterns resembling those of natural proteins, thus showcasing the significant potential of large language models in protein structure design.