Protein Science最新文献

{"title":"Soluble protein analog selection engine (SPASE): An automated AI-powered server to improve protein engineering workflows.","authors":"Sacha T Larda, Alex Paré, Nicolas Doucet","doi":"10.1002/pro.70733","DOIUrl":"10.1002/pro.70733","url":null,"abstract":"<p><p>The design of proteins with desired biophysical properties, such as high solubility and low aggregation propensity, is crucial for various biotechnological and biomedical applications. While deep learning-based methods like ProteinMPNN have shown remarkable success in protein sequence design, their direct output may not always exhibit optimal solubility and aggregation properties. Here, we present Soluble Protein Analog Selection Engine (SPASE), a novel automated webserver that addresses this challenge by integrating ProteinMPNN with state-of-the-art tools for protein solubility prediction (Protein-Sol) and aggregation prediction (Aggrescan3D). SPASE automatically generates a diverse pool of protein variants using soluble ProteinMPNN, predicts the solubility of each analog, models their three-dimensional structures with ESMFold, and scores these variants based on their predicted solubility, aggregation propensity, and folding confidence. Computational benchmarking indicates that SPASE enriches for protein analogs with higher predicted solubility and lower predicted aggregation propensity than the average output of soluble ProteinMPNN. We discuss the advantages and limitations of the workflow, including challenges associated with protein novelty, solubility prediction, and aggregation assessment. These considerations highlight the value of integrated platforms for prioritizing protein designs across multiple predicted biophysical properties. Together, these results position the SPASE server as a practical and accessible computational platform for prioritizing protein engineering candidates for downstream experimental evaluation. SPASE is publicly available at https://proteinengineering.ca/, which serves as its stable public access portal.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70733"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13446025/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148679795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An ensemble learning framework for protein stability prediction with enhanced recognition of stabilizing mutations.","authors":"Yang Liu, Jian Zhang, Minghui Li","doi":"10.1002/pro.70728","DOIUrl":"10.1002/pro.70728","url":null,"abstract":"<p><p>Accurately predicting mutation-induced protein stability changes remains a central challenge in structural bioinformatics. Existing methods exhibit a strong bias toward destabilizing mutations, leading to limited performance for stabilizing mutations and constraining their utility in protein engineering. Here, we address this limitation through two complementary strategies: balanced dataset construction and integrative modeling. To mitigate the severe class imbalance in current stability datasets, we constructed undersampling-based balanced datasets and further evaluated reverse-mutation augmentation as a comparative strategy. Building on the rapid development of high-performing predictors, we hypothesized that integrating their outputs could exploit complementary strengths and improve predictive accuracy. Accordingly, we developed three modeling frameworks, including models based on handcrafted features, models using embedding representations extracted from ProteinMPNN, and ensemble models integrating a diverse set of state-of-the-art predictors. Across multiple independent test sets, ensemble models consistently outperformed individual approaches, with particularly pronounced gains in identifying stabilizing mutations. These findings demonstrate that combining undersampling-based balanced data construction with systematic predictor integration provides an effective and practical strategy for achieving more balanced and accurate protein stability prediction, and offers a useful framework for identifying stabilizing mutations in protein engineering and related applications. StaMutAble is freely available at: https://github.com/minghuilab/StaMutAble.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70728"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13439985/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148673978","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Structure and conformational dynamics of the Pseudomonas CbrA transceptor.","authors":"Melanie A Orlando, Tejas Shah, Matthew W Faber, Samik Bose, Benjamin J Orlando","doi":"10.1002/pro.70775","DOIUrl":"10.1002/pro.70775","url":null,"abstract":"<p><p>The CbrA protein is a central regulator of carbon metabolism, biofilm formation, and virulence in Pseudomonas species, but the molecular mechanisms by which CbrA links nutrient sensing to downstream signaling has remained unclear. CbrA is a rare \"transceptor\" that combines membrane transporter and histidine kinase domains into a single functional polypeptide. The structural basis for histidine recognition and membrane transport, as well as signaling through intracellular histidine kinase domains has remained elusive. Here we determined a cryo-EM structure of CbrA which provides key molecular details of the SLC5-STAC domains in this unusual system. Unexpectedly, the small peptide CbrX encoded upstream of CbrA formed a stable complex with the SLC5 transporter domain, but was not essential for growth of Pseudomonas putida on histidine as a sole carbon source. The cryo-EM structure reveals how histidine binds within the transporter, and molecular dynamics simulations provide insight into proton gradient driven conformational changes that enable histidine transport. These findings define the molecular architecture of key CbrA functional domains, and lay the foundation for developing a comprehensive understanding of coupling between membrane transport and downstream signaling pathways that guide essential physiological traits in Pseudomonas.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70775"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13488458/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148797417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Conformational asymmetry of human immunodeficiency virus-1 protease flaps increases along the darunavir drug resistance pathway.","authors":"Michel Guerrero, Roman Zadorozhnyi, Guowu Lin, Caitlin M Quinn, Guillermo Calero, Nese Kurt-Yilmaz, Celia A Schiffer, Tatyana Polenova, Angela M Gronenborn, Rieko Ishima","doi":"10.1002/pro.70736","DOIUrl":"10.1002/pro.70736","url":null,"abstract":"<p><p>Under selective drug pressure, the structure and dynamics of HIV-1 protease (PR) evolve to confer resistance to the drug. Here, we report on the conformational changes in PR flaps for drug-resistance mutations using ¹⁹F nuclear magnetic resonance (NMR) spectroscopy. We prepared wild-type (wt) PR and its three mutant variants containing 4, 10, or 11 mutations, namely 4Mut, 10Mut, and 11Mut, that were previously found in a viral passage study against darunavir (DRV). The proteins were uniformly labeled with ¹⁵N and with ¹⁹F amino-acid specifically in the 5-indole positions of two conserved tryptophan residues, W6 and W42, which are located near the dimer interface and at the edge of the flap region of PR, respectively. In solution, ¹⁹F NMR spectra of these proteins in the inhibitor-free forms showed a downfield shift of the W42 resonance as the number of mutations increased. In the DRV-bound form, the W42 resonance again shifted with increasing number of mutations and, importantly, split into two peaks, reflecting different local electronic environments of the two flaps of the dimer. Solid-state ¹⁹F magic angle spinning (MAS) NMR spectra of wt-PR and 10Mut also exhibited two W42 resonances in DRV-bound crystalline forms. These observations indicate that drug-resistance mutations increase the asymmetry between the two flaps in the PR-inhibitor interaction in the inhibitor-bound forms.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70736"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13439748/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148674019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Structural and functional insights into Pseudomonas aeruginosa ApaH, a diadenosine tetraphosphatase crucial for bacterial virulence.","authors":"Gianluca Pistoia, Matteo Cervoni, Flavia Catalano, Francesca Troilo, Francesca Guidi, Eleonora Comparini, Giuseppina Mignogna, Carlo Travaglini-Allocatelli, Alessandro Giuffrè, Antonio Coluccia, Francesco Imperi, Adele Di Matteo, Giorgio Giardina","doi":"10.1002/pro.70781","DOIUrl":"10.1002/pro.70781","url":null,"abstract":"<p><p>Infections by Pseudomonas aeruginosa are a major cause of severe morbidity and mortality in immunocompromised patients and people with cystic fibrosis, largely due to the pathogen's ability to resist antibiotic treatment and to deploy multiple virulence strategies that promote persistence in the host. We have recently uncovered that the signaling molecule diadenosine tetraphosphate (Ap4A) acts as a crucial regulator of P. aeruginosa virulence. Specifically, deletion of the Ap4A-degrading enzyme, diadenosine tetraphosphatase (ApaH), dramatically reduces the expression of key virulence factors. The structural and molecular properties of P. aeruginosa ApaH (PaApaH) remain uncharacterized. Here, we present an integrated biochemical, structural, computational, and phenotypic characterization of PaApaH. We define the molecular determinants of its manganese-dependent Ap4A hydrolytic mechanism and establish a direct functional link between PaApaH catalytic activity and virulence phenotypes in vivo. Together with the absence of ApaH homologs in eukaryotes, these findings place PaApaH as an attractive and selective target for antivirulence therapeutic strategies against P. aeruginosa infections.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70781"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13505596/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148813939","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multiplex tissue resource for high-resolution spatial protein profiling in the Human Protein Atlas.","authors":"Borbala Katona, Rutger Schutten, Filippa Bertilsson, Feria Hikmet, Per Adelsköld, Mattias Forsberg, Kalle von Feilitzen, Loren Méar, Mathias Uhlén, Cecilia Lindskog","doi":"10.1002/pro.70764","DOIUrl":"10.1002/pro.70764","url":null,"abstract":"<p><p>Spatially resolved protein expression is essential for understanding tissue organization, cellular specialization, and protein function. The open-access Human Protein Atlas database (www.proteinatlas.org) has generated an extensive antibody-based tissue resource for a majority of the human protein-coding genes using conventional immunohistochemistry, enabling body-wide annotation of protein expression across normal human tissues and major cell types. However, single-marker staining often lacks the cellular and subcellular context required to resolve rare cell populations, closely related cell states, or proteins with limited functional characterization. To address this, we established a multiplex tissue resource within the Human Protein Atlas based on a large-scale multiplex immunohistochemistry workflow. The iterative workflow combines optimized antibody panels targeting established markers of cell identity, tissue organization, cellular state, and subcellular structure with candidate proteins of interest. This allows protein expression to be interpreted directly within intact tissue architecture based on expression overlap between candidate proteins and panel markers. In version 25 of the Human Protein Atlas, 1106 proteins have been analyzed using eight multiplex antibody panels across nine tissue settings, including testis, motile ciliated epithelia, salivary gland, endocrine pancreas, and kidney. These panels resolve biological contexts such as stages of spermatogenesis, Sertoli cell and ciliary subcellular compartments, salivary gland acinar and ductal structures, pancreatic endocrine cell types, and nephron segments. Here, we present the design and implementation of the multiplex tissue resource and demonstrate its utility for refining spatial protein annotation across diverse human tissue systems. By providing high-resolution spatial context for protein expression in human tissues, this publicly available resource strengthens functional protein annotation and offers a framework for generating new hypotheses about protein roles in normal tissue biology.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70764"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13488366/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148797411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sequence-encoded determinants drive distinct early aggregation pathways and different structural outcomes in PNT1 fibrils across Henipavirus members.","authors":"Harshita Sawdekar, Julien Mignon, Frank Gondelaud, Anuja Walimbe, Samrat Mukhopadhyay, Joseph Chamieh, Sonia Longhi","doi":"10.1002/pro.70758","DOIUrl":"10.1002/pro.70758","url":null,"abstract":"<p><p>Protein aggregation is increasingly recognized as a biologically relevant process in viral proteins, yet the molecular determinants governing such phenomena remain poorly understood. Intrinsically disordered proteins (IDPs) are ubiquitous in viral proteomes, where conformational plasticity not only enables functional diversity but also permits aberrant self-assembly. Members of the Paramyxoviridae family, including henipaviruses such as the Nipah and Hendra viruses (two biosafety level-4 pathogens), encode the P, V, and W proteins that share a common intrinsically disordered N-terminal domain (NTD). V and W are virulence factors that undergo fibrillation. Here, we explored the aggregation landscape of seven homologous PNT1 subdomains from the NTD that harbors a conserved cryptic amyloidogenic region (CAR). By integrating Taylor dispersion analysis (TDA), Raman spectroscopy, and all-atom molecular dynamics (MD) simulations, we unraveled both early assembly kinetics and structural outcomes. Despite the conserved CAR motif, the PNT1 homologs exhibit strikingly divergent aggregation behaviors. TDA revealed distinct oligomerization pathways with variations in nucleation, growth kinetics, and monomer recruitment, indicating virus-specific pathways. Raman spectroscopy unveiled substantial variations in β-sheet content, side-chain packing, and aromatic residue environments, while MD simulations showed that conformational heterogeneity in the CAR flanking regions modulates hydrogen-bond networks and structural stability. These findings indicate that the sequence context beyond a conserved amyloidogenic core governs the aggregation landscape of Henipavirus P, V, and W proteins. Evolutionary diversification modulates conformational heterogeneity and assembly pathways, giving rise to distinct structural outcomes. This work sheds light on the molecular grammar underlying viral protein fibrillation with potential implications for viral pathogenicity.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70758"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13459193/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148707483","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The C117D oxidation mimic reveals the monomeric structure of SARS-CoV-2 main protease.","authors":"Sam Andress, Matthew J McLeod, Todd Holyoak","doi":"10.1002/pro.70753","DOIUrl":"10.1002/pro.70753","url":null,"abstract":"<p><p>The SARS-CoV-2 main protease (M^pro) is essential for viral replication and functions as a homodimer, with dimerization being critical for catalytic activity. M^pro contains an unusually high number of cysteine residues. Among these, C117 and the catalytic nucleophile C145 can form a reversible disulfide bond under oxidative conditions. To investigate the structural and functional consequences of irreversible oxidation of these residues, we generated oxidation mimics by substituting these residues with aspartate (C117D and C145D), to mimic the sulfinic acid oxidation state. Kinetic assays revealed that both variants are catalytically inactive, with C117D exhibiting at least 100-fold lower activity than wild-type (WT) M^pro. Small-angle X-ray scattering (SAXS) and differential scanning fluorimetry (DSF) demonstrated that C117D adopts a monomeric, destabilized state in solution, whereas C145D retains a dimeric conformation similar to WT. Crystallographic analysis of C117D revealed a dramatic rearrangement of domain III, involving a ~40° rotation relative to domains I and II, and disorder in the N- and C-terminal regions, disrupting the canonical dimerization interface. Local structural changes propagated from the C117D site to the active site, including an unwound oxyanion loop that provides structural evidence for the observed inactivity. Rescue of the stable, dimeric state for C117D was achieved through formation of the covalent C117D-GC376 complex. These findings establish a high-resolution structure of monomeric full-length SARS-CoV-2 M^pro and underscore a critical role of C117 in maintaining dimerization and enzymatic function. Furthermore, the unique monomeric domain II-III interface present in the monomeric form may offer opportunities for allosteric inhibitor design targeting M^pro dimerization.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70753"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13454373/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148702181","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ContrastQA: A label-guided graph contrastive learning-based approach for protein complex structure quality assessment.","authors":"Lei Zhang, Rui Ding, Xiao Chen, Jie Hou, Dong Si, Yang Wang, Keying Lin, Renzhi Cao","doi":"10.1002/pro.70734","DOIUrl":"10.1002/pro.70734","url":null,"abstract":"<p><p>Despite recent progress, the Estimation of Model Accuracy (EMA) for protein complexes remains less advanced compared to that for protein monomers. A key challenge lies in effectively integrating both interface-specific and global structural information to accurately assess the quality of protein complexes. Here, we introduce ContrastQA, the first EMA framework for protein complexes that incorporates the proposed label-guided graph contrastive learning based on interface quality. By integrating a geometric graph neural network to model global structural features, ContrastQA effectively captures both local (interface-level) and global (structure-level) information for accurate model quality estimation. ContrastQA achieved ranking losses of 0.123 and 0.116 on the TMscore and GDT-TS metrics on the CASP16 dataset, which are 0.015 (10.9%) and 0.012 (8.7%) lower than the second-best EMA method with ranking losses of 0.138 and 0.128. Our study demonstrates the strong effectiveness of the label-guided graph contrastive learning module, particularly in selecting high-quality models. These findings suggest that our graph contrastive learning framework serves as a valuable pre-training strategy for learning protein structure representations.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70734"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13469176/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148721732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting membrane protein localization by deep learning on structure and chemistry.","authors":"Bivek Pokhrel, Christian Munley, Miguel Pedraza, Edward Lyman","doi":"10.1002/pro.70747","DOIUrl":"10.1002/pro.70747","url":null,"abstract":"<p><p>It has been known since at least the 1980's that the structure and chemistry of membranes and membrane proteins are matched. Exploiting this fact, a graph neural network model of proteins was trained on experimentally determined membrane protein structures to predict the native membrane environment of transmembrane domains from their structure. The algorithm, \"GPSforTMDs,\" learns to generalize about membrane protein structure, obtains overall performance that is competitive with sequence-based methods, and obtains exceptional performance for some categories of membrane environment, even when training examples are few. Other categories it finds more challenging, in some cases for clear reasons (for example, compatibility of TMDs with membranes along the secretory pathway), and in other cases that are mysterious (mistaking archaeal TMDs for bacterial, and vice versa). The results motivate the need for high quality databases reporting TMD localization, and suggest that peering inside the algorithm will reveal new \"rules\" for membrane proteins. The code and associated database is available at https://github.com/bivekpok/GPSforTMDs.</p>","PeriodicalId":20761,"journal":{"name":"Protein Science","volume":"35 9","pages":"e70747"},"PeriodicalIF":5.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13446024/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148679792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}