Structural Biology Research Articles (Page 1)

Relaxin-3 is a two-chain neuropeptide of the insulin/relaxin superfamily and the cognate ligand for the G protein-coupled receptor RXFP3. Since its discovery, the relaxin-3/RXFP3 signaling system has emerged as a key regulator of feeding behavior, stress responses, arousal, addiction, and cognitive function. Recent structural studies, including the first cryo-electron microscopy structures of RXFP3 bound to relaxin-3 and small molecules, have provided significant insights into ligand-receptor interactions. Together with mutagenesis and pharmacological studies, these advances have facilitated the design of diverse RXFP3 ligands, ranging from simplified and/or stapled single-chain analogs of relaxin-3 to grafted scaffolds and small-molecule modulators. Such tools have been instrumental for probing relaxin-3 biology in vivo and highlight the system's therapeutic potential for treating anxiety, depression, obesity, binge eating, and alcohol use disorder. However, challenges remain, particularly regarding blood-brain barrier penetration, receptor subtype selectivity, pharmacokinetic optimization, and safe long-term modulation. This review summarizes current knowledge of relaxin-3 structure, receptor interactions, and pharmacology and highlights how advances in peptide chemistry, structural biology, and small-molecule design are enabling the rational development of RXFP3-targeted therapeutics.

Serial crystallography (SX) has revolutionized structural biology by enabling high-resolution structure determination for important classes of proteins, including the study of relevant biomolecular reaction mechanisms. However, one of the ongoing challenges in this field remains the efficient use of precious macromolecule samples whose availability is often limited. Reducing sample consumption is thus critical in maximizing the potential of SX conducted at powerful X-ray sources such as synchrotrons and X-ray free-electron lasers (XFEL) to expand to a broader range of significant biological samples, gaining insights into unraveled biological reaction mechanisms. This review focuses on three primary sample delivery systems: fixed-targets, liquid injection, and hybrid methods, each with distinct advantages and limitations concerning sample consumption. The progress and challenges associated with these methods, highlighting advancements in reducing sample consumption and thus enabling the study of more diverse biological samples, are summarized. We compare the currently reported sample delivery methods in view of the minimum amount of sample required to obtain a full data set and discuss how the current approaches compare to this theoretical minimum. With this overview, we aim to provide a critical and comprehensive assessment of the current methods and experimental realizations for sample delivery in SX with proteins.

Cryogenic-electron tomography (cryo-ET) has provided an unprecedented glimpse into the nanoscale architecture of cells by combining cryogenic preservation of biological structures with electron tomography. Micropatterning of extracellular matrix proteins is increasingly used as a method to prepare adherent cell types for cryo-ET as it promotes optimal positioning of cells and subcellular regions of interest for vitrification, cryo-focused ion beam (cryo-FIB) milling, and data acquisition. Here we demonstrate a micropatterning workflow for capturing minimally adherent cell types, human T cells and Jurkat cells, for cryo-FIB and cryo-ET. Our affinity capture system facilitated the nanoscale imaging of Jurkat cells, revealing extracellular filamentous structures. It improved workflow efficiency by consistently producing grids with a sufficient number of well-positioned cells for an entire cryo-FIB session. Affinity capture can be extended to facilitate high-resolution imaging of other adherent and non-adherent cell types with cryo-ET.

In recent years, humanoid robots have made substantial advances in motion control and multimodal interaction. However, full-size humanoid robots face significant technical challenges due to their inherent geometric and physical properties, leading to large inertia of humanoid robots and substantial driving forces. These characteristics result in issues such as limited biomimetic capabilities, low control efficiency, and complex system integration, thereby restricting practical applications of full-size humanoid robots in real-world settings. To address these limitations, this paper incorporates a biomimetic design approach that draws inspiration from biological structures and movement mechanisms to enhance the robot’s human-like movements and overall efficiency. The platform introduced in this paper, Loong, is designed to overcome these challenges, offering a practically viable solution for full-size humanoid robots. The research team has innovatively used highly biomimetic joint designs to enhance Loong’s capacity for human-like movements and developed a multi-level control architecture along with a multi-master high-speed real-time communication mechanism that significantly improves its control efficiency. In addition, Loong incorporates a modular system integration strategy, which offers substantial advantages in mass production and maintenance, which improves its adaptability and practical utility for diverse operational environments. The biomimetic approach not only enhances Loong’s functionality but also enables it to perform better in complex and dynamic environments. To validate Loong’s design performance, extensive experimental tests were performed, which demonstrated the robot’s ability to traverse complex terrains such as 13 cm steps and 20° slopes and its competence in object manipulation and transportation. These innovations provide a new design paradigm for the development of full-size humanoid robots while laying a more compatible foundation for the development of hardware platforms for medium- and small-sized humanoid robots. This work makes a significant contribution to the practical deployment of humanoid robots.

Abstract Wearable ultrasound patches have the potential to transform medical diagnostics by eliminating spatiotemporal constraints. A key challenge lies in developing coupling layers that combine sufficient rigidity for probe support with softness and adhesion to eliminate air gaps and maintain stable skin contact. Inspired by biological gradient structures, this rigid‐soft conflict was addressed by engineering a gradiently crosslinked polydimethylsiloxane (PDMS) coupling layer. The high‐modulus upper layer provided mechanical support for the probe, while the low‐modulus, adhesive bottom layer ensured conformal contact with skin. This gradient structure was achieved by pouring PDMS prepolymers onto a glass slide functionalized with N‐(2‐aminoethyl)‐3‐aminopropyltrimethoxysilane (AEAPS), which anchored platinum (Pt) catalyst near the surface. A diffusion‐driven Pt gradient resulted in a corresponding crosslinking density gradient upon immediate curing. Incorporating 3 vol.% tungsten and 25 vol.% aluminum oxide further optimized acoustic impedance (1.53 MRayl) to match skin and maintained high adhesion strengths (7.33 and 6.55 N m −1 ). This gradient PDMS layer enabled long‐term, stable coupling and high‐quality imaging of radial artery blood flow, advancing the field of wearable ultrasound diagnostics.

Molecular toxicity prediction plays a crucial role in drug screening and environmental health risk assessment. Traditional toxicity prediction models primarily rely on molecular fingerprints and other structural features, while neglecting the complex biological mechanisms underlying compound toxicity, resulting in limited predictive accuracy, poor interpretability, and reduced generalizability. To address this challenge, this study proposes a novel molecular toxicity prediction framework that integrates knowledge graphs with Graph Neural Networks (GNNs). Specifically, we constructed a heterogeneous toxicological knowledge graph (ToxKG) based on ComptoxAI. ToxKG incorporates data from authoritative databases such as PubChem, Reactome, and ChEMBL, and covers multiple entities and relationships including chemicals, genes, signaling pathways, and bioassays. We then systematically evaluated six representative GNN models (GCN, GAT, R-GCN, HRAN, HGT, and GPS) on the Tox21 dataset. Experimental results demonstrate that heterogeneous graph models enriched with ToxKG information significantly outperform traditional models relying solely on structural features across multiple metrics including AUC, F1-score, ACC, and balanced accuracy (BAC). Notably, the GPS model achieved the highest AUC value (0.956) for key receptor tasks such as NR-AR, highlighting the critical role of biological mechanism information and heterogeneous graph structures in toxicity prediction. This study provides a promising pathway toward the development of interpretable and efficient intelligent models for toxicological risk assessment.

Structure-based drug discovery relies on three-dimensional protein structures to provide the atomic blueprints for small-molecule design, indicating where to place each atom to maximize favorable interactions. The advent of cryo-cooling crystals in crystallography greatly accelerated the ease and accessibility of structural data, making it a mainstay of most drug discovery efforts. However, despite its successes, including producing numerous clinically successful molecules, cryo-cooled samples only tell part of the structural story: they may leave out dynamic details or introduce artifacts that may lead drug discovery campaigns astray. In this Perspective, we highlight recent studies characterizing temperature-sensitive structural phenomena observed by crystallography. We showcase how leveraging information on rare, hidden conformational states informs ligand discovery via molecular docking. This demonstrates the value of performing structural studies at elevated temperatures, closer to where biology occurs, to 'unfreeze' structural ensembles for drug discovery and design.

Newcastle disease virus (NDV) is a representative paramyxovirus that usually causes severe infections and substantial economic losses to the global poultry industry. Over the years, NDV has attracted widespread attention as a promising oncolytic virotherapy agent and vector vaccine against many pathogens and an important prototype for elucidating the replication and pathogenesis of other paramyxoviruses. The F and HN glycoproteins are two kinds of glycosylated transmembrane proteins located on the virion envelope that play multiple roles in the virulence, infection, replication, and pathogenicity of NDV. In view of the ability to induce neutralizing and protective antibodies and the similarity in the structural features of the F and HN glycoproteins of NDV and other paramyxoviruses, investigating their structures and functions is beneficial for understanding the viral lifecycle and pathogenesis and developing more effective broad-spectrum antibodies or antiviral drugs against viral infection. This systematic review aims to summarize the structural features and membrane fusion mechanism of the F and HN glycoproteins and their relationships with viral virulence, pathogenic phenotype and thermostability, coupled with the crucial roles of F/HN-host protein/compound interactions in the infection, replication, and pathogenicity of NDV. Additionally, this review also highlights the importance of technologies such as protein‒protein interactome analysis, single-particle cryo-electron microscopy, genome-wide CRISPR/Cas9 library screening, and computational structural biology for providing novel viewpoints on the lifecycle and pathogenesis of NDV and related paramyxoviruses and valuable reference information for the future development of efficient treatment strategies targeting viral glycoproteins.

The contemporary urban environment is undergoing rapid transformation due to global climate challenges and the increasing demand for improved energy efficiency, reduced material consumption, and enhanced user comfort. Architecture and construction — fields deeply intertwined with resource management and environmental quality — must adopt innovative and sustainable design strategies. One of the most promising approaches is bionics (biomimetics), an interdisciplinary field that draws on principles from biological systems to inform design practices. Nature has long served as a source of architectural inspiration — from the symbolic and organic forms of Art Nouveau to contemporary projects enabled by digital technologies that allow for precise modeling of biological structures. However, modern biomimetics extends beyond aesthetics. It encompasses function, structure, and adaptive processes, integrating insights from biology, engineering, materials science, computer science, and architecture to develop nature-inspired systems that are more efficient, durable, and sustainable. Key technologies in this domain include parametric and generative design, environmental simulation, digital fabrication, and smart materials, all of which enable the creation of structures that respond dynamically to external stimuli. Additionally, the growing trend of biophilic design — often combined with biomimetics — contributes to the development of spaces that are not only energy-efficient but also promote the psychophysical well-being of users. This article aims to present the current state of knowledge regarding the application of bionic solutions in architecture and construction. Selected case studies are discussed to illustrate how biomimetic strategies have achieved high environmental, aesthetic, and functional performance. The article presents examples of rod structures based on minimal forms, as well as structures with topology inspired by nature and parametrically generated. It also explores specific architectural implementations and identifies challenges and future directions for this rapidly evolving field.

Chronic and neuropathic pain remain significant clinical challenges owing to limited efficacy and safety concerns associated with conventional analgesics, including opioids and NSAIDs. Voltage-gated sodium channels, particularly Nav1.7 and Nav1.8, have emerged as promising non-opioid targets for pain modulation, given their selective expression in peripheral nociceptors and critical roles in pain signal transmission. Recent advances in structural biology and pharmacology have enabled the development of highly selective inhibitors targeting these channels. This review explores sodium channel inhibitors currently in clinical development, with a focus on suzetrigine (VX-548), the first US Food and Drug Administration (FDA)-approved Nav1.8 inhibitor for acute pain, as well as other investigational agents such as ralfinamide, OLP-1002, LTGO-33 and HBW-004285. Despite setbacks in early candidates owing to selectivity and tolerability issues, ongoing trials demonstrate renewed optimism for a new class of analgesics that may overcome the limitations of traditional pain therapies. We discuss key pharmacological challenges observed in earlier trials including functional redundancy, species differences, and on-target side effects, and outline how emerging strategies, such as structural biology-guided design, combination therapies, and precision medicine, are paving the way for safer, more effective, nonaddictive pain treatments.

CASP16 is the most recent in a series of community experiments to rigorously assess the state of the art in areas of computational structural biology. The field has advanced enormously in recent years: in early CASPs, the assessments centered around whether the methods were at all useful. Now they mostly focus on how near we are to not needing experiments. In most areas, deep learning methods dominate, particularly AlphaFold variants and associated technology. In this round, there is no significant change in overall agreement between calculated monomer protein structures and their experimental counterparts, not because of method deficiencies but because, for most proteins, agreement is likely as high as can be obtained given experimental uncertainty. For protein complexes, huge gains in accuracy were made in the previous CASP, but there still appears to be room for further improvement. In contrast to these encouraging results, for RNA structures, the deep learning methods are notably unsuccessful at present and are not superior to traditional approaches. Both approaches still produce very poor results in the absence of structural homology. For macromolecular ensembles, the small CASP target set limits conclusions, but generally, in the absence of structural templates, results tend to be poor and detailed structures of alternative conformations are usually of relatively low accuracy. For organic ligand-protein structures and affinities (important for aspects of drug design), deep learning methods are substantially more successful than traditional ones on the relatively easy CASP target set, though the results often fall short of experimental accuracy. In the less glamorous but essential area of methods for estimating the accuracy, previous results found reliable accuracy estimates at the amino acid level. The present CASP results show that the best methods are also largely effective in selecting models of protein complexes with high interface accuracy. Will upcoming method improvements overcome the remaining barriers to reaching experimental accuracy in all categories? We will have to wait until the next CASP to find out, but there are two promising trends. One is the combination of traditional physics-inspired methods and deep learning, and the other is the expected increase in training data, especially for ligand-protein complexes.

Single-molecule localization microscopy generates point clouds corresponding to fluorophore localizations. Spatial cluster identification and analysis of these point clouds are crucial for extracting insights about molecular organization. However, this task becomes challenging in the presence of localization noise, high point density, or complex biological structures. Here, we introduce MIRO (Multifunctional Integration through Relational Optimization), an algorithm that uses recurrent graph neural networks to transform the point clouds in order to improve clustering efficiency when applying conventional clustering techniques. We show that MIRO supports simultaneous processing of clusters of different shapes and at multiple scales, demonstrating improved performance across varied datasets. Our comprehensive evaluation demonstrates MIRO’s transformative potential for single-molecule localization applications, showcasing its capability to revolutionize cluster analysis and provide accurate, reliable details of molecular architecture. In addition, MIRO’s robust clustering capabilities hold promise for applications in various fields such as neuroscience, for the analysis of neural connectivity patterns, and environmental science, for studying spatial distributions of ecological data.

Recent update on the discovery of indoleamine-2,3-dioxygenase 1 inhibitors targeting cancer immunotherapy.

Peptide nanodiscs: Versatile platforms for membrane protein functional reconstitution and structural studies: A review.

Does Sequence Clustering Confound AlphaFold2?

Seeing is believing-Plasmodium falciparum translation in action.

C-C motif glycoprotein ligand 5 (CCL5) and its GPCR CCR5: Macromolecular game-changers in cancer biology.

The macrophage migration inhibitory factor (MIF) family of cytokines comprised of the MIF and D-dopachrome tautomerase (or MIF-2) paralogs share identical tertiary and quaternary structures that contribute to their overlapping enzymatic and signaling activities. Recent investigations of MIF and MIF-2 have shown them to possess N-to-C-terminal allosteric crosstalk, but despite the similarity of this "allosteric pathway," its regulation of MIF and MIF-2 is not identical. Thus, structure alone does not preserve the precise allosteric mechanism, and additional residues that modulate MIF and MIF-2 function must be characterized. Cysteines have been identified as allosteric switches for the same biochemical functions of MIF, and small molecules targeting its N-terminal enzymatic site have affected the structure of three proximal cysteines. Ebselen is a compound that forms covalent selenylsulfide bonds with MIF cysteines and is hypothesized to destabilize and dissociate the MIF trimer into monomers. Ebselen-bound MIF also displays little-to-no catalysis or biological signaling. However, it is unclear whether Ebselen similarly affects the MIF-2 paralog, despite MIF-2 containing two related cysteines (MIF contains three). We used mutagenesis, nuclear magnetic resonance, molecular dynamics simulations, in vitro and in vivo biochemistry to investigate the mechanism of Ebselen as a modulator of MIF-2 via its cysteines. Our findings suggest that Ebselen partially disrupts the MIF-2 homotrimer, though the overall population of such a structure is <35%, even on the timescale of many hours. Ebselen does attenuate the biological functions of MIF-2, and solution structural biology captures the conformational transitions preceding the destabilized MIF-2 trimer.

Protein structure prediction has been revolutionized and generalized with the advent of cutting-edge AI methods such as AlphaFold, but reliance on computationally intensive multiple sequence alignments (MSA) remains a major limitation. We introduce DeepFold-PLM, a novel framework that integrates advanced protein language models with vector embedding databases to enhance ultra-fast MSA construction, remote homology detection, and protein structure prediction. DeepFold-PLM utilizes high-dimensional embeddings and contrastive learning, significantly accelerate MSA generation, achieving 47 times faster than standard methods, while maintaining prediction accuracy comparable to AlphaFold. In addition, it enhances structure prediction by extending modeling capabilities to multimeric protein complexes, provides a scalable PyTorch-based implementation for efficient large-scale prediction. Our method also effectively increases sequence diversity (Neff = 8.65 versus 4.83 with JackHMMER) enriching coevolutionary information critical for accurate structure prediction. DeepFold-PLM thus represents a versatile and practical resource that enables high-throughput applications in computational structural biology. Source codes and user-friendly Python API of all modules of DeepFold-PLM publicly available at https://github.com/DeepFoldProtein/DeepFold-PLM.

Viscous DNA and RNA: Quantum damped dynamical random systems.