The structure of the tailed phage is composed of an icosahedral (or elongated icosahedral) head and a spiral symmetrical tail, which are connected by a portal located at a unique vertex of the icosahedron. A series of image-processing methods and tools have been developed to address the asymmetric structures of phages. However, the structural determination in small proteins within the head and flexible proteins of tailed phages remains a significant impediment, further hindering our deep understanding of the structural biology field. In this study, we developed a data-processing strategy for tailed phage and demonstrated its efficacy with three cryo-EM datasets, including podophage T7, siphophage T1, and myophage Mu. The proposed strategy combines conventional icosahedral reconstruction with local refinement and reconstruction and consists of four key modules: icosahedral reconstruction, selection of the unique vertex of the icosahedron, local asymmetric reconstruction and refinement, and local defocus refinement. The strategy has been successfully applied to determine the asymmetric structure of a range of tailed phages, with a particular focus on resolving the small proteins (core proteins and scaffolding proteins) within the head and flexible proteins on the tail. In addition, the local defocus refinement of our strategy approaches the approximate resolution limit of the icosahedral capsid. The proposed strategy is a viable scheme for determining the asymmetric structures of tailed phages, especially in podophages.
{"title":"A data-processing strategy of asymmetric reconstruction for tailed phages by Cryo-electron Microscopy","authors":"Wenyuan Chen , Jing Zheng , Junquan Zhou , Lingpeng Cheng , Hongrong Liu","doi":"10.1016/j.jsb.2025.108262","DOIUrl":"10.1016/j.jsb.2025.108262","url":null,"abstract":"<div><div>The structure of the tailed phage is composed of an icosahedral (or elongated icosahedral) head and a spiral symmetrical tail, which are connected by a portal located at a unique vertex of the icosahedron. A series of image-processing methods and tools have been developed to address the asymmetric structures of phages. However, the structural determination in small proteins within the head and flexible proteins of tailed phages remains a significant impediment, further hindering our deep understanding of the structural biology field. In this study, we developed a data-processing strategy for tailed phage and demonstrated its efficacy with three cryo-EM datasets, including podophage T7, siphophage T1, and myophage Mu. The proposed strategy combines conventional icosahedral reconstruction with local refinement and reconstruction and consists of four key modules: icosahedral reconstruction, selection of the unique vertex of the icosahedron, local asymmetric reconstruction and refinement, and local defocus refinement. The strategy has been successfully applied to determine the asymmetric structure of a range of tailed phages, with a particular focus on resolving the small proteins (core proteins and scaffolding proteins) within the head and flexible proteins on the tail. In addition, the local defocus refinement of our strategy approaches the approximate resolution limit of the icosahedral capsid. The proposed strategy is a viable scheme for determining the asymmetric structures of tailed phages, especially in podophages.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108262"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145495681","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-10-10DOI: 10.1016/j.jsb.2025.108252
Kazuaki Hoshi
Toll-like receptor 9 (TLR9) recognizes pathogenic DNA molecules containing unmethylated cytosine-phosphate-guanine motifs (CpG DNA) and initiates signaling cascades essential for enhancing immune responses. TLR9 is a type I transmembrane receptor comprising an N-terminal leucine-rich repeat (LRR) domain, a transmembrane domain, and a C-terminal Toll/interleukin-1 receptor (TIR) domain. Most studies have focused on the interaction between the LRR domain and its DNA ligands. However, the TIR domain is crucial for interacting with adapter proteins such as myeloid differentiation factor 88 (MyD88). The aim of this study was to predict changes in the orientation of the TIR domain in human TLR9 (hTLR9) and its complexes with agonistic or antagonistic DNA molecules using the AlphaFold server. AlphaFold predicted the overall structure of hTLR9 with high confidence scores, including part of the TIR domain. Interestingly, binding of agonistic and antagonistic DNA molecules to the N-terminal LRR domain induced a structural change in the orientation of the TIR domain compared to the unbound TLR9 structure. The TIR domain in the predicted hTLR9 model displayed a secondary structure similar to that of the previously reported human TLR1 crystal structure. The predicted model suggests that ligand binding to the N-terminal LRR domain causes a change in the orientation of the TIR domain of hTLR9, likely due to bending of the transmembrane region.
{"title":"Prediction of a structural change in the orientation of the cytoplasmic signaling unit of human Toll-like receptor 9 upon binding of agonistic and antagonistic DNA molecules","authors":"Kazuaki Hoshi","doi":"10.1016/j.jsb.2025.108252","DOIUrl":"10.1016/j.jsb.2025.108252","url":null,"abstract":"<div><div>Toll-like receptor 9 (TLR9) recognizes pathogenic DNA molecules containing unmethylated cytosine-phosphate-guanine motifs (CpG DNA) and initiates signaling cascades essential for enhancing immune responses. TLR9 is a type I transmembrane receptor comprising an N-terminal leucine-rich repeat (LRR) domain, a transmembrane domain, and a C-terminal Toll/interleukin-1 receptor (TIR) domain. Most studies have focused on the interaction between the LRR domain and its DNA ligands. However, the TIR domain is crucial for interacting with adapter proteins such as myeloid differentiation factor 88 (MyD88). The aim of this study was to predict changes in the orientation of the TIR domain in human TLR9 (hTLR9) and its complexes with agonistic or antagonistic DNA molecules using the AlphaFold server. AlphaFold predicted the overall structure of hTLR9 with high confidence scores, including part of the TIR domain. Interestingly, binding of agonistic and antagonistic DNA molecules to the N-terminal LRR domain induced a structural change in the orientation of the TIR domain compared to the unbound TLR9 structure. The TIR domain in the predicted hTLR9 model displayed a secondary structure similar to that of the previously reported human TLR1 crystal structure. The predicted model suggests that ligand binding to the N-terminal LRR domain causes a change in the orientation of the TIR domain of hTLR9, likely due to bending of the transmembrane region.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108252"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145280609","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Insoluble aggregated tau protein in the form of paired helical filaments is a causative agent of the neurofibrillary tangles observed in Alzheimer’s disease (AD). The hexapeptide 275VQIINK280 located in the microtubule-binding domain of tau plays a crucial role in the abnormal aggregation process. Therefore, targeting the VQIINK sequence with a tau aggregation inhibitor may be a promising therapeutic approach for AD. A previous study demonstrated that the Fab domain of the tau antibody (Fab2r3) inhibits tau aggregation by binding to the VQIINK sequence. By determining the three-dimensional structures of the Fab2r3-VQIINK peptide complex and apo Fab2r3, we elucidated the recognition mechanism between Fab2r3 and the VQIINK peptide. However, the basis for the selectivity of Fab2r3 for VQIINK was not completely clear. Therefore, the objective of this report is to investigate the selective binding mechanism of Fab2r3 against VQIINK peptide. Through isothermal titration calorimetry, we show that Ile-4 in the VQIINK peptide is crucial for the selectivity of Fab2r3. X-ray structural analysis of three complexes of Fab2r3 with Ile-4 mutated peptides (VQIVYK, VQILNK, and VQIFNK) suggested that the rigid conformation of a hydrophobic pocket in Fab2r3 plays a vital role in ligand selectivity. These findings may explain the effectiveness of Fab2r3 as a tau aggregation inhibitor.
{"title":"Precise ligand-selective mechanism at the fab domain of a tau-recognizing antibody","authors":"Tomohiro Tsuchida , Takahiro Tsuchiya , Katsuhiko Minoura , Yasuko In , Katsushiro Miyamoto , Taizo Taniguchi , Toshimasa Ishida , Koji Tomoo","doi":"10.1016/j.jsb.2025.108250","DOIUrl":"10.1016/j.jsb.2025.108250","url":null,"abstract":"<div><div>Insoluble aggregated tau protein in the form of paired helical filaments is a causative agent of the neurofibrillary tangles observed in Alzheimer’s disease (AD). The hexapeptide <sup>275</sup>VQIINK<sup>280</sup> located in the microtubule-binding domain of tau plays a crucial role in the abnormal aggregation process. Therefore, targeting the VQIINK sequence with a tau aggregation inhibitor may be a promising therapeutic approach for AD. A previous study demonstrated that the Fab domain of the tau antibody (Fab2r3) inhibits tau aggregation by binding to the VQIINK sequence. By determining the three-dimensional structures of the Fab2r3-VQIINK peptide complex and apo Fab2r3, we elucidated the recognition mechanism between Fab2r3 and the VQIINK peptide. However, the basis for the selectivity of Fab2r3 for VQIINK was not completely clear. Therefore, the objective of this report is to investigate the selective binding mechanism of Fab2r3 against VQIINK peptide. Through isothermal titration calorimetry, we show that Ile-4 in the VQIINK peptide is crucial for the selectivity of Fab2r3. X-ray structural analysis of three complexes of Fab2r3 with Ile-4 mutated peptides (VQIVYK, VQILNK, and VQIFNK) suggested that the rigid conformation of a hydrophobic pocket in Fab2r3 plays a vital role in ligand selectivity. These findings may explain the effectiveness of Fab2r3 as a tau aggregation inhibitor.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108250"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145125003","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In drug development, the efficacy of an antibody depends on how the antibody interacts with the target antigen. The strength of these interactions, measured through “binding affinity”, gives an indication of how successful an antibody is in neutralizing an antigen. Due to the high computational complexity of traditional techniques for binding affinity quantification, deep learning is recently employed for the task at hand. Despite the commendable improvements in deep learning-based binding affinity prediction, such approaches are highly dependent on the quality of the antibody–antigen structures and they tend to overlook the importance of capturing the evolutionary details of proteins upon mutation. Further, most of the existing datasets for the task only include antibody–antigen pairs related to one antigen variant and, thus, are not suitable for developing comprehensive data-driven approaches. To circumvent the said complexities, we first curate the largest and most generalized (i.e., including a wide array of antigen variants) datasets for antibody–antigen binding affinity prediction, consisting of more than sequence pairs, structure pairs and the corresponding continuous binding affinity values. Subsequently, we propose a novel deep geometric neural network comprising a structure-based model, which is to account atomistic-scale structural features, and a sequence-based model, which is to attribute sequential and evolutionary information, while sharing the learned information from each model through cross-attention blocks. Further, within each parallel model, we mimic the interaction space of antibodies and antigens through a set of multi-scale hierarchical attention blocks and the final latent vectors of each model are obtained by considering antibody and antigen representative vectors and the interaction vector. The proposed framework exhibited a 10% improvement in mean absolute error compared to the state-of-the-art models while showing a strong correlation () between the predictions and target values. Additionally, we extensively discuss the model optimization strategies, weight space analysis, and interpretability in a post-hoc fashion. We release our datasets and code publicly to support the development of antibody–antigen binding affinity prediction frameworks for the benefit of science and society.
{"title":"Deep geometric framework to predict antibody–antigen binding affinity","authors":"Nuwan Bandara , Dasun Premathilaka , Sachini Chandanayake , Sahan Hettiarachchi , Vithurshan Varenthirarajah , Aravinda Munasinghe , Kaushalya Madhawa , Subodha Charles","doi":"10.1016/j.jsb.2025.108257","DOIUrl":"10.1016/j.jsb.2025.108257","url":null,"abstract":"<div><div>In drug development, the efficacy of an antibody depends on how the antibody interacts with the target antigen. The strength of these interactions, measured through “binding affinity”, gives an indication of how successful an antibody is in neutralizing an antigen. Due to the high computational complexity of traditional techniques for binding affinity quantification, deep learning is recently employed for the task at hand. Despite the commendable improvements in deep learning-based binding affinity prediction, such approaches are highly dependent on the quality of the antibody–antigen structures and they tend to overlook the importance of capturing the evolutionary details of proteins upon mutation. Further, most of the existing datasets for the task only include antibody–antigen pairs related to one antigen variant and, thus, are not suitable for developing comprehensive data-driven approaches. To circumvent the said complexities, we first curate the largest and most generalized (i.e., including a wide array of antigen variants) datasets for antibody–antigen binding affinity prediction, consisting of more than <span><math><mrow><mn>100</mn><mi>K</mi></mrow></math></span> sequence pairs, <span><math><mrow><mn>8</mn><mi>K</mi></mrow></math></span> structure pairs and the corresponding continuous binding affinity values. Subsequently, we propose a novel deep geometric neural network comprising a structure-based model, which is to account atomistic-scale structural features, and a sequence-based model, which is to attribute sequential and evolutionary information, while sharing the learned information from each model through cross-attention blocks. Further, within each parallel model, we mimic the interaction space of antibodies and antigens through a set of multi-scale hierarchical attention blocks and the final latent vectors of each model are obtained by considering antibody and antigen representative vectors and the interaction vector. The proposed framework exhibited a 10% improvement in mean absolute error compared to the state-of-the-art models while showing a strong correlation (<span><math><mrow><mo>></mo><mn>0</mn><mo>.</mo><mn>87</mn></mrow></math></span>) between the predictions and target values. Additionally, we extensively discuss the model optimization strategies, weight space analysis, and interpretability in a post-hoc fashion. We release our datasets and code publicly to support the development of antibody–antigen binding affinity prediction frameworks for the benefit of science and society.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108257"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145370349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-10-16DOI: 10.1016/j.jsb.2025.108254
Agata Leszczuk , Nataliia Kutyrieva-Nowak , Sebastian Rueda , Amit Basu
Arabinogalactan proteins (AGPs) are cell wall-plasma membrane proteins with a high level of glycosylation. The selective and high-affinity binding between AGP and the Yariv reagent has been widely used to carry out functional studies on AGPs by disrupting AGP functions using a non-genetic tool. The current work aimed to determine the molecular features of cell walls during Arabidopsis thaliana seed germination under conditions where AGP functions are blocked. To achieve this, we used molecular & imaging methods with molecular probes and for the first time − a new tool for AGP detection − a fluorescent analogue of the Yariv reagent. The most significant changes included a decrease in the content of AGPs, due to the addition of the Yariv reagent, and subsequent changes only in the content of AGPs upon transfer from the Yariv reagent to fresh Yariv-free medium. Additionally, as a result of the presence of the Yariv reagent, changes in the molecular masses of the analysed cell wall components were observed: lack of AGPs with small molecular mass and disappearance of homogalacturonan with high molecular mass. This work provided the first example of AGP labelling using antibodies and AzYariv-Cy5, and highlights the utility of AzYariv-Cy5 as a broad-spectrum tool for AGP studies.
{"title":"Changes in Arabidopsis thaliana seedling cell wall assembly induced by treatment with Yariv reagent – Molecular features & visualization with immunocytochemistry and a fluorescent Yariv reagent","authors":"Agata Leszczuk , Nataliia Kutyrieva-Nowak , Sebastian Rueda , Amit Basu","doi":"10.1016/j.jsb.2025.108254","DOIUrl":"10.1016/j.jsb.2025.108254","url":null,"abstract":"<div><div>Arabinogalactan proteins (AGPs) are cell wall-plasma membrane proteins with a high level of glycosylation. The selective and high-affinity binding between AGP and the Yariv reagent has been widely used to carry out functional studies on AGPs by disrupting AGP functions using a non-genetic tool. The current work aimed to determine the molecular features of cell walls during <em>Arabidopsis thaliana</em> seed germination under conditions where AGP functions are blocked. To achieve this, we used molecular & imaging methods with molecular probes and for the first time − a new tool for AGP detection − a fluorescent analogue of the Yariv reagent. The<!--> <!-->most significant changes included a decrease in the content of AGPs, due to the addition of the Yariv reagent, and subsequent changes only in the content of AGPs upon transfer from the Yariv reagent to fresh Yariv-free medium. Additionally, as a result of the presence of the Yariv reagent, changes in the molecular masses of the analysed cell wall components were observed: lack of AGPs with small molecular mass and disappearance of homogalacturonan with high molecular mass. This work provided the first example of AGP labelling using antibodies and AzYariv-Cy5, and highlights the utility of AzYariv-Cy5 as a broad-spectrum tool for AGP studies.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108254"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145318232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-11-29DOI: 10.1016/j.jsb.2025.108267
Mart G.F. Last , Maartje van Klaveren , Lennert Janssen , Nickels Jensen , Isabelle Jansen , Stefan Jakobs , Lenard M. Voortman , Thomas H. Sharp
Correlating super-resolution fluorescence light microscopy with cryo-electron tomography (SRcryoCLEM) is a feasible way of targeting specific proteins of interest for high-resolution cryo-electron tomography (cryoET) imaging within cells. Among different approaches for performing super-resolution fluorescence microscopy on cryogenically preserved samples, cryo-single molecule localization microscopy (cryoSMLM) offers one of the highest imaging resolutions. Thus far, applications of cryoSMLM in SRcryoCLEM have been limited to targeting a single protein structure at a time, as the available palette of cryo-compatible reversibly photoswitchable fluorescent proteins, required for cryoSMLM imaging, is severely limited. Here, we present rsTagRFP and rsEGFP2 as a compatible pair of red and green fluorescent labels that enables dual-colour cryoSMLM, and thus dual-target SRcryoCLEM, in mammalian cells. We demonstrate the simultaneous targeting and identification of two separate structures, MAP2-decorated microtubules and vimentin intermediate filaments, with 30 nm accuracy and within the same cell.
{"title":"Dual-colour super-resolution cryoCLEM in mammalian cells using the fluorescent proteins rsTagRFP and rsEGFP2","authors":"Mart G.F. Last , Maartje van Klaveren , Lennert Janssen , Nickels Jensen , Isabelle Jansen , Stefan Jakobs , Lenard M. Voortman , Thomas H. Sharp","doi":"10.1016/j.jsb.2025.108267","DOIUrl":"10.1016/j.jsb.2025.108267","url":null,"abstract":"<div><div>Correlating super-resolution fluorescence light microscopy with cryo-electron tomography (SRcryoCLEM) is a feasible way of targeting specific proteins of interest for high-resolution cryo-electron tomography (cryoET) imaging within cells. Among different approaches for performing super-resolution fluorescence microscopy on cryogenically preserved samples, cryo-single molecule localization microscopy (cryoSMLM) offers one of the highest imaging resolutions. Thus far, applications of cryoSMLM in SRcryoCLEM have been limited to targeting a single protein structure at a time, as the available palette of cryo-compatible reversibly photoswitchable fluorescent proteins, required for cryoSMLM imaging, is severely limited. Here, we present rsTagRFP and rsEGFP2 as a compatible pair of red and green fluorescent labels that enables dual-colour cryoSMLM, and thus dual-target SRcryoCLEM, in mammalian cells. We demonstrate the simultaneous targeting and identification of two separate structures, MAP2-decorated microtubules and vimentin intermediate filaments, with 30 <!--> <!-->nm accuracy and within the same cell.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108267"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145654589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-10-30DOI: 10.1016/j.jsb.2025.108258
Alain Morales-Martínez , Edgar Garduño , José María Carazo , Carlos Oscar S. Sorzano , José Luis Vilas
Cryo-electron tomography (cryo-ET) is a microscopy technique that enables the acquisition of 3D images of biological samples. Research in cell biology has shown that cellular processes are carried out by groups of macromolecules that interact in a crowded environment. In such an environment, where multiple biological macromolecules coexist and intertwine, semantic segmentation becomes even more challenging but crucial to understanding the structure and function of macromolecular complexes. However, manual semantic segmentation can be time-consuming, highly subjective, and prone to variability, which poses significant obstacles in studies dealing with large volumes of data. In contrast, automated algorithms such as Convolutional Neural Networks (CNNs) can process large-scale datasets with minimal human resources, thereby reducing the subjectivity associated with manual segmentation. In this work, we propose a convolutional neural network architecture that combines the features of U-Net, DeepLab, SegNet, Gated-SCNN, LSTM (Long Short-Term Memory), RNN (Recurrent Neural Network), and GAN (Generative Adversarial Network) architectures. This hybrid architecture effectively learns to identify different types of membranes and can replicate the behavior of a skilled human annotator. This system demonstrates a strong ability to segment various cellular membranes and vesicle structures.
{"title":"Membrane and vesicle structure detection in cryo-electron tomography based on deep learning","authors":"Alain Morales-Martínez , Edgar Garduño , José María Carazo , Carlos Oscar S. Sorzano , José Luis Vilas","doi":"10.1016/j.jsb.2025.108258","DOIUrl":"10.1016/j.jsb.2025.108258","url":null,"abstract":"<div><div>Cryo-electron tomography (cryo-ET) is a microscopy technique that enables the acquisition of 3D images of biological samples. Research in cell biology has shown that cellular processes are carried out by groups of macromolecules that interact in a crowded environment. In such an environment, where multiple biological macromolecules coexist and intertwine, semantic segmentation becomes even more challenging but crucial to understanding the structure and function of macromolecular complexes. However, manual semantic segmentation can be time-consuming, highly subjective, and prone to variability, which poses significant obstacles in studies dealing with large volumes of data. In contrast, automated algorithms such as Convolutional Neural Networks (CNNs) can process large-scale datasets with minimal human resources, thereby reducing the subjectivity associated with manual segmentation. In this work, we propose a convolutional neural network architecture that combines the features of U-Net, DeepLab, SegNet, Gated-SCNN, LSTM (Long Short-Term Memory), RNN (Recurrent Neural Network), and GAN (Generative Adversarial Network) architectures. This hybrid architecture effectively learns to identify different types of membranes and can replicate the behavior of a skilled human annotator. This system demonstrates a strong ability to segment various cellular membranes and vesicle structures.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108258"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145424724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-11-07DOI: 10.1016/j.jsb.2025.108260
Chuanyun Luo, Mark Paetzel
Type I signal peptidase (SPase I) is an essential membrane-bound enzyme that removes amino-terminal signal peptides from secretory proteins. Owing to its critical role in bacterial viability and its periplasmic accessibility, SPase I has emerged as an attractive target for antibiotic development. Arylomycins, a class of macrocyclic lipohexapeptide natural products, inhibit SPase I by binding to its active site. Previous studies have identified a key resistance determinant—a proline residue at the base of the substrate-binding groove (Pro84 in Escherichia coli SPase I)—which reduces arylomycin affinity. Here, we present the crystal structure of the E. coli SPase I P84A mutant in complex with arylomycin A2, revealing that the introduced alanine enables an additional hydrogen bond between the enzyme backbone and the arylomycin N-terminal carbonyl, thus enhancing the affinity for arylomycins. Furthermore, a newly developed preprotein-binding assay utilizing a non-cleavable version of ProOmpA Nuclease A demonstrates that substituting SPase I Pro84 with serine or leucine disrupts substrate recognition, underscoring the delicate balance between inhibitor resistance and substrate processing. These findings reveal that residue Pro84 participates in the interaction between preprotein signal peptides and the E. coli SPase I substrate-binding groove, offering a foundation for designing next-generation arylomycin analogs with improved antibacterial potency.
I型信号肽酶（SPase I）是一种必需的膜结合酶，可从分泌蛋白中去除氨基末端信号肽。由于其在细菌活力和质周可及性中的关键作用，SPase I已成为抗生素开发的一个有吸引力的靶点。芳霉素是一类大环脂六肽的天然产物，通过结合酶I的活性位点抑制酶I。先前的研究已经确定了一个关键的抗性决定因素——在底物结合槽底部的脯氨酸残基（大肠杆菌pase I中的Pro84）——它降低了阿霉素的亲和力。在这里，我们展示了theE的晶体结构。coliSPase I P84A与arylomycin A2复合物发生突变，表明引入的丙氨酸使酶主链与arylomycin n端羰基之间形成额外的氢键，从而增强了对arylomycin的亲和力。此外，一项利用不可切割版本的ProOmpA核酸酶a的新开发的蛋白前结合试验表明，用丝氨酸或亮氨酸取代SPase I Pro84会破坏底物识别，强调抑制剂抗性和底物加工之间的微妙平衡。这些发现表明，残基Pro84参与了蛋白前信号肽与大肠杆菌SPase I底物结合槽的相互作用，为设计具有更高抑菌效力的下一代阿霉素类似物奠定了基础。
{"title":"Crystal structure of Escherichia coli type I signal peptidase P84A in complex with lipopeptide antibiotic arylomycin A2","authors":"Chuanyun Luo, Mark Paetzel","doi":"10.1016/j.jsb.2025.108260","DOIUrl":"10.1016/j.jsb.2025.108260","url":null,"abstract":"<div><div>Type I signal peptidase (SPase I) is an essential membrane-bound enzyme that removes amino-terminal signal peptides from secretory proteins. Owing to its critical role in bacterial viability and its periplasmic accessibility, SPase I has emerged as an attractive target for antibiotic development. Arylomycins, a class of macrocyclic lipohexapeptide natural products, inhibit SPase I by binding to its active site. Previous studies have identified a key resistance determinant—a proline residue at the base of the substrate-binding groove (Pro84 in<!--> <em>Escherichia coli</em> <!-->SPase I)—which reduces arylomycin affinity. Here, we present the crystal structure of the<!--> <em>E. coli</em> <!-->SPase I P84A mutant in complex with arylomycin A<sub>2</sub>, revealing that the introduced alanine enables an additional hydrogen bond between the enzyme backbone and the arylomycin N-terminal carbonyl, thus enhancing the affinity for arylomycins. Furthermore, a newly developed preprotein-binding assay utilizing a non-cleavable version of ProOmpA Nuclease A demonstrates that substituting SPase I Pro84 with serine or leucine disrupts substrate recognition, underscoring the delicate balance between inhibitor resistance and substrate processing. These findings reveal that residue Pro84 participates in the interaction between preprotein signal peptides and the <em>E. coli</em> SPase I substrate-binding groove, offering a foundation for designing next-generation arylomycin analogs with improved antibacterial potency.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108260"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145482362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-11-11DOI: 10.1016/j.jsb.2025.108261
Bo Zheng , Yibei Yu , Maonian Wu , Shaojun Zhu , Tao Wu , Cheng Qian
Cryogenic electron tomography is an important technique that enables the three-dimensional visualisation of microscopic samples. In cryogenic electron tomography, a series of two-dimensional projection images is acquired from different tilt angles of the sample and computationally reconstructed into a tomogram. The tilt range of the specimen stage is typically limited to a certain angular range. Beyond this range, the sample may become too thick for electrons to penetrate, and mechanical components such as the support grid or holder may obstruct the beam, resulting in a loss of image quality. This angular limitation leads to missing information in the reconstructed tomograms, known as the missing wedge problem. Moreover, the use of low-dose electron imaging and other experimental constraints introduces considerable noise, thereby reducing the signal-to-noise ratio of the reconstructed tomogram. In order to solve the problems of missing wedges and low signal-to-noise ratio of tomograms, the Fillnet tomogram restoration framework was designed in this study. The training pair generation module and the FFT_Unet model are specially designed in this framework to improve the accurate acquisition of three-dimensional features in tomograms. Different loss functions are also designed to improve the model’s attention to the special features of the samples.
{"title":"Fillnet: A cryogenic electron tomography restoration framework integrating FFT_Unet architecture and weight optimisation strategy","authors":"Bo Zheng , Yibei Yu , Maonian Wu , Shaojun Zhu , Tao Wu , Cheng Qian","doi":"10.1016/j.jsb.2025.108261","DOIUrl":"10.1016/j.jsb.2025.108261","url":null,"abstract":"<div><div>Cryogenic electron tomography is an important technique that enables the three-dimensional visualisation of microscopic samples. In cryogenic electron tomography, a series of two-dimensional projection images is acquired from different tilt angles of the sample and computationally reconstructed into a tomogram. The tilt range of the specimen stage is typically limited to a certain angular range. Beyond this range, the sample may become too thick for electrons to penetrate, and mechanical components such as the support grid or holder may obstruct the beam, resulting in a loss of image quality. This angular limitation leads to missing information in the reconstructed tomograms, known as the missing wedge problem. Moreover, the use of low-dose electron imaging and other experimental constraints introduces considerable noise, thereby reducing the signal-to-noise ratio of the reconstructed tomogram. In order to solve the problems of missing wedges and low signal-to-noise ratio of tomograms, the Fillnet tomogram restoration framework was designed in this study. The training pair generation module and the FFT_Unet model are specially designed in this framework to improve the accurate acquisition of three-dimensional features in tomograms. Different loss functions are also designed to improve the model’s attention to the special features of the samples.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 4","pages":"Article 108261"},"PeriodicalIF":2.7,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145513224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-01Epub Date: 2025-06-26DOI: 10.1016/j.jsb.2025.108230
Sebastian Dorawa , Katarzyna Biniek-Antosiak , Magdalena Bejger , Anna-Karina Kaczorowska , Karol Ciuchcinski , Agnieszka Godlewska , Magdalena Płotka , Gudmundur O. Hreggvidsson , Lukasz Dziewit , Tadeusz Kaczorowski , Wojciech Rypniewski
We presents the discovery and molecular characterization of a novel lytic enzyme from the extremophilic Thermus thermophilus MAT72 phage vB_Tt72. The protein of 346-aa (MW = 39,705) functions as phage vB_Tt72 endolysin and shows low sequence identity (<37 %) to members of M23 family of peptidoglycan hydrolases, except for two uncharacterized endopeptidases of T. thermophilus phages: φYS40 (87 %) and φTMA (88 %). The enzyme exhibits lytic activity mainly against bacteria of the genus Thermus and, to a lesser extent, against other Gram-negative and Gram-positive bacteria. The protein is monomeric in solution and is highly thermostable (Tm = 98.3 °C). It retains ∼ 50 % of its lytic activity after 90 min of incubation at 99 °C. Crystallographic analysis, at 2.2 Å resolution, revealed a fold characteristic of M23 metallopeptidases, accounting for 40 % of the structure. The remaining parts of the molecule are folded in a manner that was previously undescribed. The M23 fold contains a Zn2+ ion coordinated by a conserved His-Asp-His triad, and two conserved His residues essential for catalysis. The active site is occupied by a phosphate or a sulfate anion, while the substrate-binding groove contains a ligand, which is a fragment of E. coli peptidoglycan. The common sequence-based criteria failed to identify the protein as (hyper)thermophilic. It is likely that the protein’s thermal stability is owed to peculiar features of its three-dimensional structure. Instead of trimmed surface loops, observed in many thermostable proteins, the catalytic domain contains two long loops that interlace and form an α-helical bundle with its own hydrophobic core.
{"title":"Crystal structure, enzymatic and thermodynamic properties of the Thermus thermophilus phage Tt72 lytic endopeptidase with unique structural signatures of thermal adaptation","authors":"Sebastian Dorawa , Katarzyna Biniek-Antosiak , Magdalena Bejger , Anna-Karina Kaczorowska , Karol Ciuchcinski , Agnieszka Godlewska , Magdalena Płotka , Gudmundur O. Hreggvidsson , Lukasz Dziewit , Tadeusz Kaczorowski , Wojciech Rypniewski","doi":"10.1016/j.jsb.2025.108230","DOIUrl":"10.1016/j.jsb.2025.108230","url":null,"abstract":"<div><div>We presents the discovery and molecular characterization of a novel lytic enzyme from the extremophilic <em>Thermus thermophilus</em> MAT72 phage vB_Tt72. The protein of 346-aa (MW = 39,705) functions as phage vB_Tt72 endolysin and shows low sequence identity (<37 %) to members of M23 family of peptidoglycan hydrolases, except for two uncharacterized endopeptidases of <em>T. thermophilus</em> phages: φYS40 (87 %) and φTMA (88 %). The enzyme exhibits lytic activity mainly against bacteria of the genus <em>Thermus</em> and, to a lesser extent, against other Gram-negative and Gram-positive bacteria. The protein is monomeric in solution and is highly thermostable (T<sub>m</sub> = 98.3 °C). It retains ∼ 50 % of its lytic activity after 90 min of incubation at 99 °C. Crystallographic analysis, at 2.2 Å resolution, revealed a fold characteristic of M23 metallopeptidases, accounting for 40 % of the structure. The remaining parts of the molecule are folded in a manner that was previously undescribed. The M23 fold contains a Zn<sup>2+</sup> ion coordinated by a conserved His-Asp-His triad, and two conserved His residues essential for catalysis. The active site is occupied by a phosphate or a sulfate anion, while the substrate-binding groove contains a ligand, which is a fragment of <em>E. coli</em> peptidoglycan. The common sequence-based criteria failed to identify the protein as (hyper)thermophilic. It is likely that the protein’s thermal stability is owed to peculiar features of its three-dimensional structure. Instead of trimmed surface loops, observed in many thermostable proteins, the catalytic domain contains two long loops that interlace and form an α-helical bundle with its own hydrophobic core.</div></div>","PeriodicalId":17074,"journal":{"name":"Journal of structural biology","volume":"217 3","pages":"Article 108230"},"PeriodicalIF":3.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144522821","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号