bims-strubi 2021-10-03 papers

bims-strubi

Biomed News

on Advances in structural biology

Issue of 2021–10–03
seventeen papers selected by
Alessandro Grinzato, European Synchrotron Radiation Facility

Structure determination of GPCRs: cryo-EM compared with X-ray crystallography.
Using machine learning to study protein-protein interactions: From the uromodulin polymer to egg zona pellucida filaments.
From Cells to Atoms: Cryo-EM as an Essential Tool to Investigate Pathogen Biology, Host-Pathogen Interaction and Drug Discovery.
AlphaFold 2: Why It Works and Its Implications for Understanding the Relationships of Protein Sequence, Structure, and Function.
The Atlas of Bacterial & Archaeal Cell Structure: an Interactive Open-Access Microbiology Textbook.
ProLIF: a library to encode molecular interactions as fingerprints.
Structure of filamentous viruses.
Divide-and-link peptide docking: a fragment-based peptide docking protocol.
Kinematic Reconstruction of Cyclic Peptides and Protein Backbones from Partial Data.
CovPDB: a high-resolution coverage of the covalent protein-ligand interactome.
Micropatterning Transmission Electron Microscopy Grids to Direct Cell Positioning within Whole-Cell Cryo-Electron Tomography Workflows.
Weight average approaches for predicting dynamical properties of biomolecules.
Comprehensive Survey of Recent Drug Discovery Using Deep Learning.
gmx_MMPBSA: A New Tool to Perform End-State Free Energy Calculations with GROMACS.
MDM-TASK-web: MD-TASK and MODE-TASK web server for analyzing protein dynamics.
Making it Rain: Cloud-Based Molecular Simulations for Everyone.
Pain or gain?

Biochem Soc Trans. 2021 Sep 28. pii: BST20210431. [Epub ahead of print]

Structure determination of GPCRs: cryo-EM compared with X-ray crystallography.

Javier García-Nafría, Christopher G Tate.

  G protein-coupled receptors (GPCRs) are the largest single family of cell surface receptors encoded by the human genome and they play pivotal roles in co-ordinating cellular systems throughout the human body, making them ideal drug targets. Structural biology has played a key role in defining how receptors are activated and signal through G proteins and β-arrestins. The application of structure-based drug design (SBDD) is now yielding novel compounds targeting GPCRs. There is thus significant interest from both academia and the pharmaceutical industry in the structural biology of GPCRs as currently only about one quarter of human non-odorant receptors have had their structure determined. Initially, all the structures were determined by X-ray crystallography, but recent advances in electron cryo-microscopy (cryo-EM) now make GPCRs tractable targets for single-particle cryo-EM with comparable resolution to X-ray crystallography. So far this year, 78% of the 99 GPCR structures deposited in the PDB (Jan-Jul 2021) were determined by cryo-EM. Cryo-EM has also opened up new possibilities in GPCR structural biology, such as determining structures of GPCRs embedded in a lipid nanodisc and multiple GPCR conformations from a single preparation. However, X-ray crystallography still has a number of advantages, particularly in the speed of determining many structures of the same receptor bound to different ligands, an essential prerequisite for effective SBDD. We will discuss the relative merits of cryo-EM and X-ray crystallography for the structure determination of GPCRs and the future potential of both techniques.

Keywords:  G-protein-coupled receptors; cryo-electron microscopy; crystallography

DOI:  https://doi.org/10.1042/BST20210431
Mol Reprod Dev. 2021 Sep 29.

Using machine learning to study protein-protein interactions: From the uromodulin polymer to egg zona pellucida filaments.

Luca Jovine.

  Neural network-based models for protein structure prediction have recently reached near-experimental accuracy and are fast becoming a powerful tool in the arsenal of biologists. As suggested by initial studies using RoseTTAFold or the ColabFold implementation of AlphaFold2, a particularly interesting future development will be the optimization of these computational methods to also routinely yield high-confidence predictions of protein-protein interactions. Here I use AlphaFold2 and ColabFold to investigate the activation and polymerization of uromodulin (UMOD)/Tamm-Horsfall protein, a zona pellucida (ZP) module-containing protein whose precursor and filamentous structures have been previously determined experimentally by X-ray crystallography and cryo-EM, respectively. Despite having no knowledge of the UMOD polymer structure (coordinates for which were neither used for model training nor as template), AlphaFold2/ColabFold are able to recapitulate a crucial conformational change underlying UMOD polymerization, as well as the general organization of protein subunits within the resulting filament. This surprising result is achieved by simply deleting from the input sequence a stretch of residues that correspond to a polymerization-inhibiting C-terminal propeptide. By mimicking in silico the activating effect of propeptide dissociation triggered by site-specific proteolysis of the protein precursor, this example has implications for the assembly of egg coat proteins and the many other molecules that also contain a ZP module. Most importantly, it shows the potential of exploiting machine learning not only to accurately predict the structures of individual proteins or complexes, but also to carry out computational experiments replicating specific molecular events.

Keywords:  artificial intelligence; protein polymerization; protein-protein interactions; uromodulin; zona pellucida

DOI:  https://doi.org/10.1002/mrd.23538
Mol Microbiol. 2021 Sep 30.

From Cells to Atoms: Cryo-EM as an Essential Tool to Investigate Pathogen Biology, Host-Pathogen Interaction and Drug Discovery.

Doulin C Shepherd, Somavally Dalvi, Debnath Ghosal.

Electron cryo-microscopy (cryo-EM) has lately emerged as a powerful method in structural biology and cell biology. While cryo-EM single-particle analysis (SPA) is now routinely delivering structures of purified proteins and protein complexes at near-atomic resolution, the use of electron cryo-tomography (cryo-ET), together with subtomogram averaging, is allowing visualization of macromolecular complexes in their native cellular environment, at unprecedented resolution. The unique ability of cryo-EM to provide information at many spatial resolution scales from ångströms to microns makes it an invaluable tool that bridges the classic 'resolution-gap' between structural biology and cell biology domains. Like in many other fields of biology, in recent years, cryo-EM has revolutionized our understanding of pathogen biology, host-pathogen interaction and has made significant strides towards structure-based drug discovery. In a very recent example, during this ongoing coronavirus disease (COVID-19) pandemic, the structure of the stabilized severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein was deciphered by SPA. This led to the development of multiple vaccines. Alongside, cryo-ET provided key insights into the structure of the native virion, mechanism of its entry, replication, and budding; demonstrating the unrivaled power of cryo-EM in investigating pathogen biology, host-pathogen interaction and drug discovery. In this review, we showcase a few examples of how different imaging modalities within cryo-EM have enabled the study of microbiology and host-pathogen interaction.

DOI: https://doi.org/10.1111/mmi.14820
J Chem Inf Model. 2021 Sep 29.

AlphaFold 2: Why It Works and Its Implications for Understanding the Relationships of Protein Sequence, Structure, and Function.

Jeffrey Skolnick, Mu Gao, Hongyi Zhou, Suresh Singh.

AlphaFold 2 (AF2) was the star of CASP14, the last biannual structure prediction experiment. Using novel deep learning, AF2 predicted the structures of many difficult protein targets at or near experimental resolution. Here, we present our perspective of why AF2 works and show that it is a very sophisticated fold recognition algorithm that exploits the completeness of the library of single domain PDB structures. It has also learned local side chain packing rearrangements that enable it to refine proteins to high resolution. The benefits and limitations of its ability to predict the structures of many more proteins at or close to atomic detail are discussed.

DOI: https://doi.org/10.1021/acs.jcim.1c01114
J Microbiol Biol Educ. 2021 ;pii: e00128-21. [Epub ahead of print]22(2):

The Atlas of Bacterial & Archaeal Cell Structure: an Interactive Open-Access Microbiology Textbook.

Catherine M Oikonomou, Grant J Jensen.

  Here, we describe a new open-access digital textbook for microbiology, The Atlas of Bacterial & Archaeal Cell Structure (available at cellstructureatlas.org). The book addresses a fundamental gap in existing textbooks, namely, what bacterial and archaeal cells look like and how the macromolecular structures they contain give rise to their diverse and complex functions. The interactive, multimedia resource features real data from more than 150 cells belonging to approximately 70 different species, imaged by cutting-edge cryogenic electron microscopy (cryo-EM). Complementary animations show the cellular machinery in action. Only a basic familiarity with fundamental biology concepts is required to understand the material, which targets a wide range of students in courses from general biology for nonmajors to specialized graduate-level microbiology. The content can be digested in several hours, making it well suited to be assigned as a supplemental resource for a course covering either more diverse topics in cell biology or a more specialized topic such as medical microbiology. By making this resource freely available online, we hope it will serve students in diverse educational settings, including self-directed learners.

Keywords:  archaea; bacteria; cryo-EM; cryo-electron tomography; educational resource; microbiology; online; open access; textbook

DOI:  https://doi.org/10.1128/jmbe.00128-21
J Cheminform. 2021 Sep 25. 13(1): 72

ProLIF: a library to encode molecular interactions as fingerprints.

Cédric Bouysset, Sébastien Fiorucci.

  Interaction fingerprints are vector representations that summarize the three-dimensional nature of interactions in molecular complexes, typically formed between a protein and a ligand. This kind of encoding has found many applications in drug-discovery projects, from structure-based virtual-screening to machine-learning. Here, we present ProLIF, a Python library designed to generate interaction fingerprints for molecular complexes extracted from molecular dynamics trajectories, experimental structures, and docking simulations. It can handle complexes formed of any combination of ligand, protein, DNA, or RNA molecules. The available interaction types can be fully reparametrized or extended by user-defined ones. Several tutorials that cover typical use-case scenarios are available, and the documentation is accompanied with code snippets showcasing the integration with other data-analysis libraries for a more seamless user-experience. The library can be freely installed from our GitHub repository ( https://github.com/chemosim-lab/ProLIF ).

Keywords:  Docking; Interaction fingerprint; Molecular dynamics; Python; Structural biology; Virtual screening

DOI:  https://doi.org/10.1186/s13321-021-00548-6
Curr Opin Virol. 2021 Sep 27. pii: S1879-6257(21)00104-8. [Epub ahead of print]51 25-33

Structure of filamentous viruses.

Giuseppe Zanotti, Alessandro Grinzato.

Despite filamentous viruses represent an important portion of the universe of viruses, their 3D structures available are quite limited, particularly if compared to the large number of structures of icosahedral viruses present in the Protein Data Bank. As a matter of fact, flexible filamentous viruses cannot be grown as single crystals and past structural studies have mostly been limited to X-ray fiber diffraction or to the determination of the structure of isolated viral proteins. Only very recently, several structures of filamentous viruses have become available, owing to the recent development of cryo-electron microscopy. This technique has given a strong impulse to the field and has allowed the building of reliable molecular models of entire viruses, in some cases at a nearly atomic resolution level. In this paper we briefly describe the architecture of filamentous viruses that infect bacteria, archaea, plants and humans. It is easy to foresee that more new structures of filamentous viruses will become available soon and they will allow a better understanding of the rules underlying the structural organization of these organisms so relevant for the life on our planet.

DOI: https://doi.org/10.1016/j.coviro.2021.09.006
Phys Chem Chem Phys. 2021 Oct 01.

Divide-and-link peptide docking: a fragment-based peptide docking protocol.

Lu Sun, Tingting Fu, Dan Zhao, Hongjun Fan, Shijun Zhong.

Protein-peptide interactions are crucial for various important cellular regulations, and are also a basis for understanding protein-protein interactions, protein folding and peptide drug design. Due to the limited structural data obtained using experimental methods, it is necessary to predict protein-peptide interaction modes using computational methods. In the present work, we designed a fragment-based docking protocol, Divide-and-Link Peptide Docking (DLPepDock), to predict protein-peptide binding modes. This protocol contains the following steps: dividing the peptide into fragments and separately docking the fragments using a third-party small molecular docking tool, linking the docked fragmental poses to form the whole peptide conformations via fragmental coordinate transformation using our in-house program, removing unreasonable poses according to several geometrical filters, extracting representative conformations after clustering for further minimization using the steepest descent and conjugation gradient methods based on a full-atom molecular force field and finally scoring using the MM/PBSA binding energy calculation implemented in Amber. When tested on the LEADS-PEP benchmark data set of 26 diverse complexes with peptides of 6-12 residues, FlexPepDock ab initio and AutoDock CrankPep achieved superior results. DLPepDock performed better than the other 15 docking protocols implemented in nine docking programs (HPepDock, DockThor, rDock, Glide, LeDock, AutoDock, AutoDock Vina, Surflex, and GOLD). The Linux scripts to call the third-party tools and run all the calculations.

DOI: https://doi.org/10.1039/d1cp02098f
J Chem Inf Model. 2021 Sep 27.

Kinematic Reconstruction of Cyclic Peptides and Protein Backbones from Partial Data.

Mosavverul Hassan, Evangelos A Coutsias.

We present an algorithm, QBKR (Quaternary Backbone Kinematic Reconstruction), a fast analytical method for an all-atom backbone reconstruction of proteins and linear or cyclic peptide chains from Cα coordinate traces. Unlike previous analytical methods for deriving all-atom representations from coarse-grained models that rely on canonical geometry with planar peptides in the trans conformation, our de novo kinematic model incorporates noncanonical, cis-trans, geometry naturally. Perturbations to this geometry can be effected with ease in our formulation, for example, to account for a continuous change from cis to trans geometry. A simple optimization of a spring-based objective function is employed for Cα-Cα distance variations that extend beyond the cis-trans limit. The kinematic construction produces a linked chain of peptide units, Cα-C-N-Cα, hinged at the Cα atoms spanning all possible planar and nonplanar peptide conformations. We have combined our method with a ring closure algorithm for the case of ring peptides and missing loops in a protein structure. Here, the reconstruction proceeding from both the N and C termini of the protein backbone (or in both directions from a starting position for rings) requires freedom in the position of one Cα atom (a capstone) to achieve a successful loop or ring closure. A salient feature of our reconstruction method is the ability to enrich conformational ensembles to produce alternative feasible conformations in which H-bond forming C-O or N-H pairs in the backbone can reverse orientations, thus addressing a well-known shortcoming in Cα-based RMSD structure comparison, wherein very close structures may lead to significantly different overall H-bond behavior. We apply the fixed Cα-based design to the reverse reconstruction from noisy Cryo-EM data, a posteriori to the optimization. Our method can be applied to speed up the process of an all-atom description from voluminous experimental data or subpar electron density maps.

DOI: https://doi.org/10.1021/acs.jcim.1c00453
Nucleic Acids Res. 2021 Sep 28. pii: gkab868. [Epub ahead of print]

CovPDB: a high-resolution coverage of the covalent protein-ligand interactome.

Mingjie Gao, Aurélien F A Moumbock, Ammar Qaseem, Qianqing Xu, Stefan Günther.

In recent years, the drug discovery paradigm has shifted toward compounds that covalently modify disease-associated target proteins, because they tend to possess high potency, selectivity, and duration of action. The rational design of novel targeted covalent inhibitors (TCIs) typically starts from resolved macromolecular structures of target proteins in their apo or holo forms. However, the existing TCI databases contain only a paucity of covalent protein-ligand (cP-L) complexes. Herein, we report CovPDB, the first database solely dedicated to high-resolution cocrystal structures of biologically relevant cP-L complexes, curated from the Protein Data Bank. For these curated complexes, the chemical structures and warheads of pre-reactive electrophilic ligands as well as the covalent bonding mechanisms to their target proteins were expertly manually annotated. Totally, CovPDB contains 733 proteins and 1,501 ligands, relating to 2,294 cP-L complexes, 93 reactive warheads, 14 targetable residues, and 21 covalent mechanisms. Users are provided with an intuitive and interactive web interface that allows multiple search and browsing options to explore the covalent interactome at a molecular level in order to develop novel TCIs. CovPDB is freely accessible at http://www.pharmbioinf.uni-freiburg.de/covpdb/ and its contents are available for download as flat files of various formats.

DOI: https://doi.org/10.1093/nar/gkab868
J Vis Exp. 2021 Sep 13.

Micropatterning Transmission Electron Microscopy Grids to Direct Cell Positioning within Whole-Cell Cryo-Electron Tomography Workflows.

Bryan S Sibert, Joseph Y Kim, Jie E Yang, Elizabeth R Wright.

Whole-cell cryo-electron tomography (cryo-ET) is a powerful technology that is used to produce nanometer-level resolution structures of macromolecules present in the cellular context and preserved in a near-native frozen-hydrated state. However, there are challenges associated with culturing and/or adhering cells onto TEM grids in a manner that is suitable for tomography while retaining the cells in their physiological state. Here, a detailed step-by-step protocol is presented on the use of micropatterning to direct and promote eukaryotic cell growth on TEM grids. During micropatterning, cell growth is directed by depositing extra-cellular matrix (ECM) proteins within specified patterns and positions on the foil of the TEM grid while the other areas remain coated with an anti-fouling layer. Flexibility in the choice of surface coating and pattern design makes micropatterning broadly applicable for a wide range of cell types. Micropatterning is useful for studies of structures within individual cells as well as more complex experimental systems such as host-pathogen interactions or differentiated multi-cellular communities. Micropatterning may also be integrated into many downstream whole-cell cryo-ET workflows, including correlative light and electron microscopy (cryo-CLEM) and focused-ion beam milling (cryo-FIB).

DOI: https://doi.org/10.3791/62992
Curr Opin Struct Biol. 2021 Sep 27. pii: S0959-440X(21)00130-5. [Epub ahead of print]72 88-94

Weight average approaches for predicting dynamical properties of biomolecules.

Kiyoshi Yagi, Suyong Re, Takaharu Mori, Yuji Sugita.

Recent advances in atomistic molecular dynamics (MD) simulations of biomolecules allow us to explore their conformational spaces widely, observing large-scale conformational fluctuations or transitions between distinct structures. To reproduce or refine experimental data using MD simulations, structure ensembles, which are characterized by multiple structures and their statistical weights on the rugged free-energy landscapes, are often used. Here, we summarize weight average approaches for various experimental measurements. Weight average approaches are now applied to hybrid quantum mechanics/molecular mechanics MD simulations to predict fast vibrational motions in a protein with a high accuracy for better understanding of molecular functions from atomic structures.

DOI: https://doi.org/10.1016/j.sbi.2021.08.008
Int J Mol Sci. 2021 Sep 15. pii: 9983. [Epub ahead of print]22(18):

Comprehensive Survey of Recent Drug Discovery Using Deep Learning.

Jintae Kim, Sera Park, Dongbo Min, Wankyu Kim.

  Drug discovery based on artificial intelligence has been in the spotlight recently as it significantly reduces the time and cost required for developing novel drugs. With the advancement of deep learning (DL) technology and the growth of drug-related data, numerous deep-learning-based methodologies are emerging at all steps of drug development processes. In particular, pharmaceutical chemists have faced significant issues with regard to selecting and designing potential drugs for a target of interest to enter preclinical testing. The two major challenges are prediction of interactions between drugs and druggable targets and generation of novel molecular structures suitable for a target of interest. Therefore, we reviewed recent deep-learning applications in drug-target interaction (DTI) prediction and de novo drug design. In addition, we introduce a comprehensive summary of a variety of drug and protein representations, DL models, and commonly used benchmark datasets or tools for model training and testing. Finally, we present the remaining challenges for the promising future of DL-based DTI prediction and de novo drug design.

Keywords:  artificial intelligence-based drug discovery; benchmark tool; de novo drug design; deep learning; drug–target interaction; molecular representation; virtual screening

DOI:  https://doi.org/10.3390/ijms22189983
J Chem Theory Comput. 2021 Sep 29.

gmx_MMPBSA: A New Tool to Perform End-State Free Energy Calculations with GROMACS.

Mario S Valdés-Tresanco, Mario E Valdés-Tresanco, Pedro A Valiente, Ernesto Moreno.

Molecular mechanics/Poisson-Boltzmann (Generalized-Born) surface area is one of the most popular methods to estimate binding free energies. This method has been proven to balance accuracy and computational efficiency, especially when dealing with large systems. As a result of its popularity, several programs have been developed for performing MM/PB(GB)SA calculations within the GROMACS community. These programs, however, present several limitations. Here we present gmx_MMPBSA, a new tool to perform end-state free energy calculations from GROMACS molecular dynamics trajectories. gmx_MMPBSA provides the user with several options, including binding free energy calculations with different solvation models (PB, GB, or 3D-RISM), stability calculations, computational alanine scanning, entropy corrections, and binding free energy decomposition. Noteworthy, several promising methodologies to calculate relative binding free energies such as alanine scanning with variable dielectric constant and interaction entropy have also been implemented in gmx_MMPBSA. Two additional tools-gmx_MMPBSA_test and gmx_MMPBSA_ana-have been integrated within gmx_MMPBSA to improve its usability. Multiple illustrating examples can be accessed through gmx_MMPBSA_test, while gmx_MMPBSA_ana provides fast, easy, and efficient access to different graphics plotted from gmx_MMPBSA output files. The latest version (v1.4.3, 26/05/2021) is available free of charge (documentation, test files, and tutorials included) at https://github.com/Valdes-Tresanco-MS/gmx_MMPBSA.

DOI: https://doi.org/10.1021/acs.jctc.1c00645
Comput Struct Biotechnol J. 2021 ;19 5059-5071

MDM-TASK-web: MD-TASK and MODE-TASK web server for analyzing protein dynamics.

Olivier Sheik Amamuddy, Michael Glenister, Thulani Tshabalala, Özlem Tastan Bishop.

  The web server, MDM-TASK-web, combines the MD-TASK and MODE-TASK software suites, which are aimed at the coarse-grained analysis of static and all-atom MD-simulated proteins, using a variety of non-conventional approaches, such as dynamic residue network analysis, perturbation-response scanning, dynamic cross-correlation, essential dynamics and normal mode analysis. Altogether, these tools allow for the exploration of protein dynamics at various levels of detail, spanning single residue perturbations and weighted contact network representations, to global residue centrality measurements and the investigation of global protein motion. Typically, following molecular dynamic simulations designed to investigate intrinsic and extrinsic protein perturbations (for instance induced by allosteric and orthosteric ligands, protein binding, temperature, pH and mutations), this selection of tools can be used to further describe protein dynamics. This may lead to the discovery of key residues involved in biological processes, such as drug resistance. The server simplifies the set-up required for running these tools and visualizing their results. Several scripts from the tool suites were updated and new ones were also added and integrated with 2D/3D visualization via the web interface. An embedded work-flow, integrated documentation and visualization tools shorten the number of steps to follow, starting from calculations to result visualization. The Django-powered web server (available at https://mdmtaskweb.rubi.ru.ac.za/) is compatible with all major web browsers. All scripts implemented in the web platform are freely available at https://github.com/RUBi-ZA/MD-TASK/tree/mdm-task-web and https://github.com/RUBi-ZA/MODE-TASK/tree/mdm-task-web.

Keywords:  MD-TASK; MODE-TASK; Molecular dynamics analysis; Normal mode analysis; Residue network analysis

DOI:  https://doi.org/10.1016/j.csbj.2021.08.043
J Chem Inf Model. 2021 Oct 01.

Making it Rain: Cloud-Based Molecular Simulations for Everyone.

Pablo R Arantes, Marcelo D Polêto, Conrado Pedebos, Rodrigo Ligabue-Braun.

We present a user-friendly front-end for running molecular dynamics (MD) simulations using the OpenMM toolkit on the Google Colab framework. Our goals are (1) to highlight the usage of a cloud-computing scheme for educational purposes for a hands-on approach when learning MD simulations and (2) to exemplify how low-income research groups can perform MD simulations in the microsecond time scale. We hope this work facilitates teaching and learning of molecular simulation throughout the community.

DOI: https://doi.org/10.1021/acs.jcim.1c00998
Elife. 2021 09 29. pii: e73378. [Epub ahead of print]10

Pain or gain?

Valeria Kalienkova.

  The 3D structures of a membrane protein called TMEM120A suggest that it may act as an enzyme in fat metabolism rather than as an ion channel that senses mechanical pain.

Keywords:  TMEM120; coenzyme A; cryo-EM structure; human; ion channel; mechanosensation; membrane protein; molecular biophysics; mouse; neuroscience; structural biology

DOI:  https://doi.org/10.7554/eLife.73378