The first few months of the SARS-CoV-2 pandemic illustrated, in many ways, the level of maturity and essential nature of modern structural biology. The outbreak was given official pandemic status on 11 Feb 2020 - six days after the release of the first crystal structure of the main protease; the first cryo-EM structure of the spike protein was released just two weeks later. These were the first of a flood of new structures - most, in a strong break with tradition, released well before the associated manuscripts. This, combined with the recent decision by the worldwide Protein Data Bank to allow re-versioning of submitted structures by the authors, allowed for an almost unprecedented scenario: while the experimentalists worked to get these critically important structures solved as quickly as possible, specialists in model building and refinement could check and (where necessary) improve their models, returning the results to the original authors often before their papers were ever published.

In this talk I will discuss some of my observations arising from inspecting and rebuilding some three dozen early SARS-CoV-2 and SARS-CoV-1 structures. In a great many respects, the remarkable improvement in the rate of modelling errors between SARS-CoV-1 and -2 structures shows just how far the field has come. However, the devil is in the details, and various classes of repeated errors in the modern structures point to the need for further improvement in model-building and validation methods.

The Protein Data Bank (PDB) [1] was established as the first open-access digital data resource in biology in 1971 with just seven X-ray crystallographic structures of proteins. Today, the single global archive houses more than 177,000 experimentally determined 3D structures of biological macromolecules that are made freely available to millions of users worldwide with no limitations on usage. This information facilitates basic and applied research and education across the sciences, impacting fundamental biology, biomedicine, biotechnology, bioengineering, and energy sciences. The PDB archive is managed jointly by the Worldwide Protein Data Bank [2-3] (wwPDB, wwpdb.org) which is committed to making PDB data Findable-Accessible-Interoperable–Reusable (FAIR) [4].

To ensure the highest quality structure data, the wwPDB OneDep system for structure deposition [5], validation [6], and biocuration [7] (deposit.wwpdb.org) provides enhanced validation reports. During 2019-2020, wwPDB implemented depositor-initiated coordinate versioning that enables the Depositor of Record (or Principal Investigator) to replace previously released x,y,z atomic coordinates without obsoleting the original PDB entry or changing the PDB ID. This feature was developed in response to feedback from PDB depositors, who were reluctant to update their structures because the newly issued PDB ID would differ from that reported in original structure publication. We thank all of many PDB depositors, who have proactively corrected their structures using the new versioning feature within OneDep. To date, more than 200 PDB structures have been updated with newly versioned atomic coordinates.

Since the early days of the COVID-19 pandemic, PDB data have informed our understanding of SARS-CoV-2 protein structure, function, and evolution, and facilitated structure-guided discovery and development of anti-coronaviral drugs, vaccines, and neutralizing monoclonal antibodies. More than 1,000 SARS-CoV-2 related protein structures are now freely available from the PDB, reflecting enormous efforts made by the structural biology community in fighting the pandemic. Occasionally, rapid PDB data deposition and publication of coronavirus structural studies driven by an understandable sense of urgency has resulted in public release of PDB structures containing minor errors. The wwPDB coordinate versioning feature described above has enabled rapid correction of SARS-CoV-2 related structures archived in the PDB.

The wwPDB strongly encourages PDB depositors to update structures as needed using OneDep. Doing so will improve the quality of data stored in the archive, while preserving original PDB IDs and maintaining connections to the scientific literature.

[1] Protein Data Bank. (1971). Crystallography: Protein Data Bank. Nature (London), New Biol. 233:223-223.

[2] Berman, H., Henrick, K., Nakamura, H. (2003). Announcing the worldwide protein data bank. Nat Struct Biol. 10:980.

[3] wwPDB consortium. (2019). Protein Data Bank: The single global archive for 3d macromolecular structure data. Nucleic Acids Res. 47:D520-D528.

[4] Wilkinson, MD, Dumontier, M, Aalbersberg, IJ, Appleton, G, Axton, M, Baak, A, Blomberg, N, Boiten, JW, da Silva Santos, LB, Bourne, PE, et al. (2016). The FAIR guiding principles for scientific data management and stewardship. Sci Data. 3:1-9

[5] Young, J. Y., Westbrook, J. D., Feng, Z., Sala, R., Peisach, E., Oldfield, T. J., Sen, S., Gutmanas, A., Armstrong, D. R., Berrisford, J. M., et al. (2017). OneDep: Unified wwpdb system for deposition, biocuration, and validation of macromolecular structures in the pdb archive. Structure. 25:536-545.

[6] Gore, S., Sanz Garcia, E., Hendrickx, P. M. S., Gutmanas, A., Westbrook, J. D., Yang, H., Feng, Z., Baskaran, K., Berrisford, J. M., Hudson, B. P., et al. (2017). Validation of structures in the protein data bank. Structure. 25:1916-1927.

[7] Young, J. Y., Westbrook, J. D., Feng, Z., Peisach, E., Persikova, I., Sala, R., Sen, S., Berrisford, J. M., Swaminathan, G. J., Oldfield, T. J., et al. (2018). Worldwide protein data bank biocuration supporting open access to high-quality 3d structural biology data. Database. 2018:bay002.

RCSB PDB is funded by the National Science Foundation (DBI-1832184), the US Department of Energy (DE-SC0019749), and the National Cancer Institute, National Institute of Allergy and Infectious Diseases, and National Institute of General Medical Sciences of the National Institutes of Health under grant R01GM133198.

COVID-19, caused by SARS-CoV-2, is a global health and economic catastrophe. The viral main protease (M^pro) is indispensable for SARS-CoV-2 replication and thus is an important target for small-molecule antivirals. Computer-assisted and structure-guided drug design strategies rely on atomic scale understanding of the target biomacromolecule traditionally derived from X-ray crystallographic data collected at cryogenic temperatures. Conventional protein X-ray crystallography is limited by possible cryo-artifacts and its inability to locate the functional hydrogen atoms crucial for understanding chemistry occurring in enzyme active sites. Neutrons are an ideal probe to observe the protonation states of ionizable amino acids at near-physiological temperature, directly determining their electric charges – crucial information for drug design. Our X-ray crystal structures of M^pro collected at near-physiological temperatures brought rapid insights into the reactivity of the catalytic cysteine, malleability of the active site, and binding modes with clinical protease inhibitors. The neutron crystal structures of ligand-free and inhibitor-bound M^pro were determined allowing the direct observation of protonation states of all residues in a coronavirus protein for the first time. At rest, the catalytic Cys-His dyad exists in the reactive zwitterionic state, with both Cys145 and His41 charged, instead of the anticipated neutral state. Covalent inhibitor binding results in modulation of the protonation states, retaining the overall electric charge of the M^pro active site cavity. In addition, high-throughput virtual screening in conjunction with in vitro assays identified a lead non-covalent compound with micromolar affinity, which is being used to design novel M^pro inhibitors. Our research is providing real-time data for atomistic design and discovery of M^pro inhibitors to combat the COVID-19 pandemic and prepare for future threats from pathogenic coronaviruses.

The COVID-19 pandemic, instigated by the SARS-CoV-2 coronavirus, continues to plague the globe. The SARS-CoV-2 main protease, or M^pro, is a promising target for development of novel antiviral therapeutics. Previous X-ray crystal structures of M^pro were obtained at cryogenic temperature or room temperature only. Here we report a series of high-resolution crystal structures of unliganded M^pro across multiple temperatures from cryogenic to physiological, and another at high humidity. We interrogate these datasets with parsimonious multiconformer models, multi-copy ensemble models, and isomorphous difference density maps. Our analysis reveals a temperature-dependent conformational landscape for M^pro, including a mobile water interleaved between the catalytic dyad, mercurial conformational heterogeneity in a key substrate-binding loop, and a far-reaching intramolecular network bridging the active site and dimer interface. Our results may inspire new strategies for antiviral drug development to counter-punch COVID-19 and combat future coronavirus pandemics.

Technological developments and growing interests in the study of cellular assemblies have led cryo-EM as a powerful technique to solve the three-dimensional structures of macromolecular complexes. More than 450 structures of molecular complexes from SARS-CoV-2 have been solved using cryo-EM and the structures have been crucial to understand molecular details behind the viral infection and development of drug molecules and vaccines. Given that majority of the cryo-EM reconstructions are solved at resolutions worse than 2.5Å and often the local resolution within the map varies considerably, atomic model validation is crucial to identify errors and less reliable areas of the model.

Here, we present the database CoVal, which is a repository of amino acid replacement mutations identified in the SARS-CoV-2 genome sequences, mapped onto protein structures from cryo-EM and X-ray crystallography. We provide information on the demographic distribution of these mutations, along with details on co-occuring mutations. CoVal gives easy access to mutation sites mapped to known structures with multiple metrics on the quality of the structure and agreement with experimental data. We also provide validation scores for the local quality of mutation site(s) and their structural neighbors. The database is freely accessible at: https://coval.ccpem.ac.uk. We also discuss tools for atomic model validation in the CCP-EM software suite [1] and results from the validation analysis of cryo-EM structures from SARS-CoV-2.