Outcomes of the EMDataResource cryo-EM Ligand Modeling Challenge.

The EMDataResource Ligand Model Challenge aimed to assess the reliability and reproducibility of modeling ligands bound to protein and protein-nucleic acid complexes in cryogenic electron microscopy (cryo-EM) maps determined at near-atomic (1.9-2.5 Å) resolution. Three published maps were selected as targets: Escherichia coli beta-galactosidase with inhibitor, SARS-CoV-2 virus RNA-dependent RNA polymerase with covalently bound nucleotide analog and SARS-CoV-2 virus ion channel ORF3a with bound lipid. Sixty-one models were submitted from 17 independent research groups, each with supporting workflow details. The quality of submitted ligand models and surrounding atoms were analyzed by visual inspection and quantification of local map quality, model-to-map fit, geometry, energetics and contact scores. A composite rather than a single score was needed to assess macromolecule+ligand model quality. These observations lead us to recommend best practices for assessing cryo-EM structures of liganded macromolecules reported at near-atomic resolution.

DoubleHelix: nucleic acid sequence identification, assignment and validation tool for cryo-EM and crystal structure models.

Sequence assignment is a key step of the model building process in both cryogenic electron microscopy (cryo-EM) and macromolecular crystallography (MX). If the assignment fails, it can result in difficult to identify errors affecting the interpretation of a model. There are many model validation strategies that help experimentalists in this step of protein model building, but they are virtually non-existent for nucleic acids. Here, I present doubleHelix-a comprehensive method for assignment, identification, and validation of nucleic acid sequences in structures determined using cryo-EM and MX. The method combines a neural network classifier of nucleobase identities and a sequence-independent secondary structure assignment approach. I show that the presented method can successfully assist sequence-assignment step in nucleic-acid model building at lower resolutions, where visual map interpretation is very difficult. Moreover, I present examples of sequence assignment errors detected using doubleHelix in cryo-EM and MX structures of ribosomes deposited in the Protein Data Bank, which escaped the scrutiny of available model-validation approaches. The doubleHelix program source code is available under BSD-3 license at https://gitlab.com/gchojnowski/doublehelix.

Cryo-EM model validation recommendations based on outcomes of the 2019 EMDataResource challenge.

This paper describes outcomes of the 2019 Cryo-EM Model Challenge. The goals were to (1) assess the quality of models that can be produced from cryogenic electron microscopy (cryo-EM) maps using current modeling software, (2) evaluate reproducibility of modeling results from different software developers and users and (3) compare performance of current metrics used for model evaluation, particularly Fit-to-Map metrics, with focus on near-atomic resolution. Our findings demonstrate the relatively high accuracy and reproducibility of cryo-EM models derived by 13 participating teams from four benchmark maps, including three forming a resolution series (1.8 to 3.1 Å). The results permit specific recommendations to be made about validating near-atomic cryo-EM structures both in the context of individual experiments and structure data archives such as the Protein Data Bank. We recommend the adoption of multiple scoring parameters to provide full and objective annotation and assessment of the model, reflective of the observed cryo-EM map density.

AlphaPulldown-a python package for protein-protein interaction screens using AlphaFold-Multimer.

SUMMARY: The artificial intelligence-based structure prediction program AlphaFold-Multimer enabled structural modelling of protein complexes with unprecedented accuracy. Increasingly, AlphaFold-Multimer is also used to discover new protein-protein interactions (PPIs). Here, we present AlphaPulldown, a Python package that streamlines PPI screens and high-throughput modelling of higher-order oligomers using AlphaFold-Multimer. It provides a convenient command-line interface, a variety of confidence scores and a graphical analysis tool. AVAILABILITY AND IMPLEMENTATION: AlphaPulldown is freely available at https://www.embl-hamburg.de/AlphaPulldown. SUPPLEMENTARY INFORMATION: Supplementary note is available at Bioinformatics online.

RNA 3D structure modeling by fragment assembly with small-angle X-ray scattering restraints.

SUMMARY: Structure determination is a key step in the functional characterization of many non-coding RNA molecules. High-resolution RNA 3D structure determination efforts, however, are not keeping up with the pace of discovery of new non-coding RNA sequences. This increases the importance of computational approaches and low-resolution experimental data, such as from the small-angle X-ray scattering experiments. We present RNA Masonry, a computer program and a web service for a fully automated modeling of RNA 3D structures. It assemblies RNA fragments into geometrically plausible models that meet user-provided secondary structure constraints, restraints on tertiary contacts, and small-angle X-ray scattering data. We illustrate the method description with detailed benchmarks and its application to structural studies of viral RNAs with SAXS restraints. AVAILABILITY AND IMPLEMENTATION: The program web server is available at http://iimcb.genesilico.pl/rnamasonry. The source code is available at https://gitlab.com/gchojnowski/rnamasonry.

Asymmetric horseshoe-like assembly of peroxisomal yeast oxalyl-CoA synthetase.

Oxalyl-CoA synthetase from Saccharomyces cerevisiae is one of the most abundant peroxisomal proteins in yeast and hence has become a model to study peroxisomal translocation. It contains a C-terminal Peroxisome Targeting Signal 1, which however is partly dispensable, suggesting additional receptor bindings sites. To unravel any additional features that may contribute to its capacity to be recognized as peroxisomal target, we determined its assembly and overall architecture by an integrated structural biology approach, including X-ray crystallography, single particle cryo-electron microscopy and small angle X-ray scattering. Surprisingly, it assembles into mixture of concentration-dependent dimers, tetramers and hexamers by dimer self-association. Hexameric particles form an unprecedented asymmetric horseshoe-like arrangement, which considerably differs from symmetric hexameric assembly found in many other protein structures. A single mutation within the self-association interface is sufficient to abolish any higher-level oligomerization, resulting in a homogenous dimeric assembly. The small C-terminal domain of yeast Oxalyl-CoA synthetase is connected by a partly flexible hinge with the large N-terminal domain, which provides the sole basis for oligomeric assembly. Our data provide a basis to mechanistically study peroxisomal translocation of this target.

Structural studies of RNA-protein complexes: A hybrid approach involving hydrodynamics, scattering, and computational methods.

The diverse functional cellular roles played by ribonucleic acids (RNA) have emphasized the need to develop rapid and accurate methodologies to elucidate the relationship between the structure and function of RNA. Structural biology tools such as X-ray crystallography and Nuclear Magnetic Resonance are highly useful methods to obtain atomic-level resolution models of macromolecules. However, both methods have sample, time, and technical limitations that prevent their application to a number of macromolecules of interest. An emerging alternative to high-resolution structural techniques is to employ a hybrid approach that combines low-resolution shape information about macromolecules and their complexes from experimental hydrodynamic (e.g. analytical ultracentrifugation) and solution scattering measurements (e.g., solution X-ray or neutron scattering), with computational modeling to obtain atomic-level models. While promising, scattering methods rely on aggregation-free, monodispersed preparations and therefore the careful development of a quality control pipeline is fundamental to an unbiased and reliable structural determination. This review article describes hydrodynamic techniques that are highly valuable for homogeneity studies, scattering techniques useful to study the low-resolution shape, and strategies for computational modeling to obtain high-resolution 3D structural models of RNAs, proteins, and RNA-protein complexes.

RNA-Puzzles Round II: assessment of RNA structure prediction programs applied to three large RNA structures.

This paper is a report of a second round of RNA-Puzzles, a collective and blind experiment in three-dimensional (3D) RNA structure prediction. Three puzzles, Puzzles 5, 6, and 10, represented sequences of three large RNA structures with limited or no homology with previously solved RNA molecules. A lariat-capping ribozyme, as well as riboswitches complexed to adenosylcobalamin and tRNA, were predicted by seven groups using RNAComposer, ModeRNA/SimRNA, Vfold, Rosetta, DMD, MC-Fold, 3dRNA, and AMBER refinement. Some groups derived models using data from state-of-the-art chemical-mapping methods (SHAPE, DMS, CMCT, and mutate-and-map). The comparisons between the predictions and the three subsequently released crystallographic structures, solved at diffraction resolutions of 2.5-3.2 Å, were carried out automatically using various sets of quality indicators. The comparisons clearly demonstrate the state of present-day de novo prediction abilities as well as the limitations of these state-of-the-art methods. All of the best prediction models have similar topologies to the native structures, which suggests that computational methods for RNA structure prediction can already provide useful structural information for biological problems. However, the prediction accuracy for non-Watson-Crick interactions, key to proper folding of RNAs, is low and some predicted models had high Clash Scores. These two difficulties point to some of the continuing bottlenecks in RNA structure prediction. All submitted models are available for download at http://ahsoka.u-strasbg.fr/rnapuzzles/.

Sequence-specific cleavage of dsRNA by Mini-III RNase.

Ribonucleases (RNases) play a critical role in RNA processing and degradation by hydrolyzing phosphodiester bonds (exo- or endonucleolytically). Many RNases that cut RNA internally exhibit substrate specificity, but their target sites are usually limited to one or a few specific nucleotides in single-stranded RNA and often in a context of a particular three-dimensional structure of the substrate. Thus far, no RNase counterparts of restriction enzymes have been identified which could cleave double-stranded RNA (dsRNA) in a sequence-specific manner. Here, we present evidence for a sequence-dependent cleavage of long dsRNA by RNase Mini-III from Bacillus subtilis (BsMiniIII). Analysis of the sites cleaved by this enzyme in limited digest of bacteriophage Φ6 dsRNA led to the identification of a consensus target sequence. We defined nucleotide residues within the preferred cleavage site that affected the efficiency of the cleavage and were essential for the discrimination of cleavable versus non-cleavable dsRNA sequences. We have also determined that the loop α5b-α6, a distinctive structural element in Mini-III RNases, is crucial for the specific cleavage, but not for dsRNA binding. Our results suggest that BsMiniIII may serve as a prototype of a sequence-specific dsRNase that could possibly be used for targeted cleavage of dsRNA.

RNA Bricks--a database of RNA 3D motifs and their interactions.

The RNA Bricks database (http://iimcb.genesilico.pl/rnabricks), stores information about recurrent RNA 3D motifs and their interactions, found in experimentally determined RNA structures and in RNA-protein complexes. In contrast to other similar tools (RNA 3D Motif Atlas, RNA Frabase, Rloom) RNA motifs, i.e. 'RNA bricks' are presented in the molecular environment, in which they were determined, including RNA, protein, metal ions, water molecules and ligands. All nucleotide residues in RNA bricks are annotated with structural quality scores that describe real-space correlation coefficients with the electron density data (if available), backbone geometry and possible steric conflicts, which can be used to identify poorly modeled residues. The database is also equipped with an algorithm for 3D motif search and comparison. The algorithm compares spatial positions of backbone atoms of the user-provided query structure and of stored RNA motifs, without relying on sequence or secondary structure information. This enables the identification of local structural similarities among evolutionarily related and unrelated RNA molecules. Besides, the search utility enables searching 'RNA bricks' according to sequence similarity, and makes it possible to identify motifs with modified ribonucleotide residues at specific positions.

ClaRNA: a classifier of contacts in RNA 3D structures based on a comparative analysis of various classification schemes.

The understanding of folding and function of RNA molecules depends on the identification and classification of interactions between ribonucleotide residues. We developed a new method named ClaRNA for computational classification of contacts in RNA 3D structures. Unique features of the program are the ability to identify imperfect contacts and to process coarse-grained models. Each doublet of spatially close ribonucleotide residues in a query structure is compared to clusters of reference doublets obtained by analysis of a large number of experimentally determined RNA structures, and assigned a score that describes its similarity to one or more known types of contacts, including pairing, stacking, base-phosphate and base-ribose interactions. The accuracy of ClaRNA is 0.997 for canonical base pairs, 0.983 for non-canonical pairs and 0.961 for stacking interactions. The generalized squared correlation coefficient (GC2) for ClaRNA is 0.969 for canonical base pairs, 0.638 for non-canonical pairs and 0.824 for stacking interactions. The classifier can be easily extended to include new types of spatial relationships between pairs or larger assemblies of nucleotide residues. ClaRNA is freely available via a web server that includes an extensive set of tools for processing and visualizing structural information about RNA molecules.

Characterization of the termini of the West Nile virus genome and their interactions with the small isoform of the 2' 5'-oligoadenylate synthetase family.

2' 5'-Oligoadenylate synthetases (OAS) are interferon-stimulated proteins that act in the innate immune response to viral infection. Upon binding viral double-stranded RNA, OAS enzymes produce 2'-5'-linked oligoadenylates that stimulate RNase L and ultimately slow viral propagation. Truncations/mutations in the smallest human OAS isoform, OAS1, results in susceptibility to West Nile virus (WNV). We have previously demonstrated in vitro the interaction between OAS1 and the 5'-terminal region of the WNV RNA genome. Here we report that the 3'-terminal region is also able to mediate specific interaction with and activation of OAS1. Binding and kinetic experiments identified a specific stem loop within the 3'-terminal region that is sufficient for activation of the enzyme. The solution conformation of the 3'-terminal region was determined by small angle X-ray scattering, and computational models suggest a conformationally restrained structure comprised of a helix and short stem loop. Structural investigation of the 3'-terminal region in complex with OAS1 is also presented. Finally, we show that genome cyclization by base pairing between the 5'- and 3'-terminal regions, a required step for replication, is not sufficient to protect WNV from OAS1 recognition in vitro. These data provide a physical framework for understanding recognition of the highly structured terminal regions of a flaviviral genome by an innate immune enzyme.

Brickworx builds recurrent RNA and DNA structural motifs into medium- and low-resolution electron-density maps.

Brickworx is a computer program that builds crystal structure models of nucleic acid molecules using recurrent motifs including double-stranded helices. In a first step, the program searches for electron-density peaks that may correspond to phosphate groups; it may also take into account phosphate-group positions provided by the user. Subsequently, comparing the three-dimensional patterns of the P atoms with a database of nucleic acid fragments, it finds the matching positions of the double-stranded helical motifs (A-RNA or B-DNA) in the unit cell. If the target structure is RNA, the helical fragments are further extended with recurrent RNA motifs from a fragment library that contains single-stranded segments. Finally, the matched motifs are merged and refined in real space to find the most likely conformations, including a fit of the sequence to the electron-density map. The Brickworx program is available for download and as a web server at http://iimcb.genesilico.pl/brickworx.

Computational modeling of protein-RNA complex structures.

Protein-RNA interactions play fundamental roles in many biological processes, such as regulation of gene expression, RNA splicing, and protein synthesis. The understanding of these processes improves as new structures of protein-RNA complexes are solved and the molecular details of interactions analyzed. However, experimental determination of protein-RNA complex structures by high-resolution methods is tedious and difficult. Therefore, studies on protein-RNA recognition and complex formation present major technical challenges for macromolecular structural biology. Alternatively, protein-RNA interactions can be predicted by computational methods. Although less accurate than experimental measurements, theoretical models of macromolecular structures can be sufficiently accurate to prompt functional hypotheses and guide e.g. identification of important amino acid or nucleotide residues. In this article we present an overview of strategies and methods for computational modeling of protein-RNA complexes, including software developed in our laboratory, and illustrate it with practical examples of structural predictions.

Solution conformation of adenovirus virus associated RNA-I and its interaction with PKR.

Adenovirus virus-associated RNA (VAI) provides protection against the host antiviral response in part by inhibiting the interferon-induced double stranded RNA-activated protein kinase (PKR). VAI consists of three base-paired regions; the apical stem responsible for the interaction with double-stranded RNA binding motifs (dsRBMs) of PKR, the central stem required for inhibition, and the terminal stem. The solution conformation of VAI and VAI lacking the terminal stem were determined using SAXS that suggested extended conformations that are in agreement with their secondary structures. Solution conformations of VAI lacking the terminal stem in complex with the dsRBMs of PKR indicated that the apical stem interacts with both dsRNA-binding motifs whereas the central stem does not. Hydrodynamic properties calculated from ab initio models were compared to experimentally determined parameters for model validation. Furthermore, SAXS envelopes were used as a constraint for the in silico modeling of tertiary structure for RNA and RNA-protein complex. Finally, full-length PKR was also studied, but concentration-dependent changes in hydrodynamic parameters prevented ab initio shape determination. Taken together, results provide an improved structural framework that further our understanding of the role VAI plays in evading host innate immune responses.

Computational modeling of RNA 3D structures, with the aid of experimental restraints.

In addition to mRNAs whose primary function is transmission of genetic information from DNA to proteins, numerous other classes of RNA molecules exist, which are involved in a variety of functions, such as catalyzing biochemical reactions or performing regulatory roles. In analogy to proteins, the function of RNAs depends on their structure and dynamics, which are largely determined by the ribonucleotide sequence. Experimental determination of high-resolution RNA structures is both laborious and difficult, and therefore, the majority of known RNAs remain structurally uncharacterized. To address this problem, computational structure prediction methods were developed that simulate either the physical process of RNA structure formation ("Greek science" approach) or utilize information derived from known structures of other RNA molecules ("Babylonian science" approach). All computational methods suffer from various limitations that make them generally unreliable for structure prediction of long RNA sequences. However, in many cases, the limitations of computational and experimental methods can be overcome by combining these two complementary approaches with each other. In this work, we review computational approaches for RNA structure prediction, with emphasis on implementations (particular programs) that can utilize restraints derived from experimental analyses. We also list experimental approaches, whose results can be relatively easily used by computational methods. Finally, we describe case studies where computational and experimental analyses were successfully combined to determine RNA structures that would remain out of reach for each of these approaches applied separately.

Master corepressor inactivation through multivalent SLiM-induced polymerization mediated by the oncogene suppressor RAI2.

While the elucidation of regulatory mechanisms of folded proteins is facilitated due to their amenability to high-resolution structural characterization, investigation of these mechanisms in disordered proteins is more challenging due to their structural heterogeneity, which can be captured by a variety of biophysical approaches. Here, we used the transcriptional master corepressor CtBP, which binds the putative metastasis suppressor RAI2 through repetitive SLiMs, as a model system. Using cryo-electron microscopy embedded in an integrative structural biology approach, we show that RAI2 unexpectedly induces CtBP polymerization through filaments of stacked tetrameric CtBP layers. These filaments lead to RAI2-mediated CtBP nuclear foci and relieve its corepressor function in RAI2-expressing cancer cells. The impact of RAI2-mediated CtBP loss-of-function is illustrated by the analysis of a diverse cohort of prostate cancer patients, which reveals a substantial decrease in RAI2 in advanced treatment-resistant cancer subtypes. As RAI2-like SLiM motifs are found in a wide range of organisms, including pathogenic viruses, our findings serve as a paradigm for diverse functional effects through multivalent interaction-mediated polymerization by disordered proteins in healthy and diseased conditions.

Outcomes of the EMDataResource Cryo-EM Ligand Modeling Challenge.

The EMDataResource Ligand Model Challenge aimed to assess the reliability and reproducibility of modeling ligands bound to protein and protein/nucleic-acid complexes in cryogenic electron microscopy (cryo-EM) maps determined at near-atomic (1.9-2.5 Å) resolution. Three published maps were selected as targets: E. coli beta-galactosidase with inhibitor, SARS-CoV-2 RNA-dependent RNA polymerase with covalently bound nucleotide analog, and SARS-CoV-2 ion channel ORF3a with bound lipid. Sixty-one models were submitted from 17 independent research groups, each with supporting workflow details. We found that (1) the quality of submitted ligand models and surrounding atoms varied, as judged by visual inspection and quantification of local map quality, model-to-map fit, geometry, energetics, and contact scores, and (2) a composite rather than a single score was needed to assess macromolecule+ligand model quality. These observations lead us to recommend best practices for assessing cryo-EM structures of liganded macromolecules reported at near-atomic resolution.

RIBER/DIBER: a software suite for crystal content analysis in the studies of protein-nucleic acid complexes.

Co-crystallization experiments of proteins with nucleic acids do not guarantee that both components are present in the crystal. We have previously developed DIBER to predict crystal content when protein and DNA are present in the crystallization mix. Here, we present RIBER, which should be used when protein and RNA are in the crystallization drop. The combined RIBER/DIBER suite builds on machine learning techniques to make reliable, quantitative predictions of crystal content for non-expert users and high-throughput crystallography.

Sequence-assignment validation in protein crystal structure models with checkMySequence.

Sequence-register shifts remain one of the most elusive errors in experimental macromolecular models. They may affect model interpretation and propagate to newly built models from older structures. In a recent publication, it was shown that register shifts in cryo-EM models of proteins can be detected using a systematic reassignment of short model fragments to the target sequence. Here, it is shown that the same approach can be used to detect register shifts in crystal structure models using standard, model-bias-corrected electron-density maps (2mFo - DFc). Five register-shift errors in models deposited in the PDB detected using this method are described in detail.