The SARS-CoV-2 main protease (M pro ) has a pivotal role in mediating viral genome replication and transcription of the coronavirus, making it a prime target for antiviral drug development. This study presents four high-resolution crystal structures of SARS-CoV-2 M pro in the apo state. Among these structures, we identify a previously unobserved, stable inactive conformation, characterized by a major displacement of the oxyanion loop (especially Phe140, Leu141 and Asn142), the rotation of some important residues (including His163 and His172) and the loss of many key interactions around the active site, which provides new insights into the conformational dynamics of the enzyme. We also report structures of M pro in complex with the covalent inhibitor pomotrelvir (PBI-0451) for the wild-type enzyme and three clinical variant-associated mutants (G15S, K90R and P132H). PBI-0451 adopts a virtually identical mode when bound to wild-type M pro and the K90R and G15S M pro mutants. However, the crystal structure of PBI-0451 bound to SARS-CoV-2 M pro P132H reveals a discernible shift of the Gln189-bearing loop and incomplete electron density for the Gln189 side chain, which may be responsible for the subtle modest reduction in PBI-0451 susceptibility. These findings expand the conformational landscape of M pro and offer structural guidance for the design of M pro -targeted antiviral drugs.
Single-particle cryo-electron microscopy (cryo-EM) can resolve multiple structurally distinct macromolecules (for example unintended host-derived proteins that remain during recombinant protein purification) within a heterogeneous specimen. Here, cryo-EM was initially performed to investigate an oligomeric form of recombinant human 2-oxoglutarate- and iron-dependent oxygenase domain-containing protein 1 (OGFOD1) purified from Escherichia coli . However, image classification and 3D reconstruction revealed OGFOD1-incompatible particle populations. Combined cryo-EM and mass spectrometry identified these particles as three endogenous E. coli proteins: the bifunctional enzyme ArnA, guanosine triphosphate cyclohydrolase I (GTP cyclohydrolase I) and glycerol dehydrogenase (GlyDH). Their structures were reconstructed at overall resolutions of 3.24, 3.00 and 3.02 Å, respectively, revealing expected hexameric, decameric and octameric protein assemblies. Structural comparisons with previously reported crystallographic and cryo-EM structures showed that the overall oligomeric architectures and active-site configurations were highly conserved; however, local conformational differences were observed in specific regions. Despite substantial differences in total molecular mass, the three protein assemblies had comparable predicted hydrodynamic radii, which plausibly explained their partial coelution during size-exclusion chromatography. These results show that rigid, symmetric endogenous complexes can dominate heterogeneous cryo-EM datasets and highlight the value of systematic analysis of off-target particle populations.
Large language models (LLMs) offer significant potential to accelerate scientific software development, yet their application to structural biology algorithms requires rigorous evaluation. This challenge is examined through the development of xcif , a new native parser for the crystallographic information file (CIF) format, designed to replace a stable but slow parser ( ucif ) in the Computational Crystallography Toolbox ( cctbx ). Using an expert-led, artificial intelligence (AI)-assisted workflow, implementation time was reduced from an estimated several months to approximately one week of active human effort. However, this efficiency relied entirely on domain knowledge to specify constraints, intercept plausible but flawed machine-generated code and design adversarial tests. The machine assistance did not replace programming expertise but rather shifted the human engineering burden from writing code to independent validation. Applying the new xcif parser across a corpus of approximately 550 000 archive files confirmed that it produces output identical to both the legacy implementation and GEMMI , a modern high-performance parser. xcif parses roughly an order of magnitude faster than the legacy cctbx parser and 1.3–1.5 times faster than GEMMI , while reducing peak memory use for larger files. This case study demonstrates that AI assistance provides a highly practical, cost-effective pathway for the renewal of mature structural biology infrastructure, provided that the development is strictly governed by expert oversight and objective verification.
Two corrections are made to the article by St John et al. (2026).
Cryo-electron tomography is an expanding technology for the study of macromolecules, viruses and cells. It is often applied to specimens that are too large or heterogeneous for methods based on 2D image averaging such as single-particle analysis, for example intracellular membranes or organelles. Current practice records a tilt series of projection images in rotation. Reconstruction is normally an ill-posed mathematical problem. Particularly for the under-determined case of sparse data, discrete tilt angles and a limited tilt range, characteristic artifacts appear in the reconstructed slices. Much of what appears as noise is in fact structural: the projection of contrast from different planes. Various schemes are employed to regularize the reconstruction, including machine-learning frameworks built on neural networks. To the extent that the noise is structural, based on the geometry of the tilt series, it might be suppressed by deconvolution with a suitable kernel. This was demonstrated and has been used regularly in cryo-STEM tomography of thick specimens where the under-sampling problem is particularly acute. Here, we present 3 dcon as an open-source extension of the entropy-regularized deconvolution algorithm that had been adopted from fluorescence microscopy. It takes advantage of modern computing hardware for convenient and fast processing. Deconvolution is entirely algorithmic, meaning that successful processing of the data does not depend on the data itself. As such, it should be robust in a wide variety of applications.
