lDDT: a local superposition-free score for comparing protein structures and models using distance difference tests

Intra-Host Co-Existing Strains of SARS-CoV-2 Reference Genome Uncovered by Exhaustive Computational Search

The COVID-19 pandemic caused by SARS-CoV-2 has had a severe impact on people worldwide. The reference genome of the virus has been widely used as a template for designing mRNA vaccines to combat the disease. In this study, we present a computational method aimed at identifying co-existing intra-host strains of the virus from RNA-sequencing data of short reads that were used to assemble the original reference genome. Our method consisted of five key steps: extraction of relevant reads, error correction for the reads, identification of within-host diversity, phylogenetic study, and protein binding affinity analysis. Our study revealed that multiple strains of SARS-CoV-2 can coexist in both the viral sample used to produce the reference sequence and a wastewater sample from California. Additionally, our workflow demonstrated its capability to identify within-host diversity in foot-and-mouth disease virus (FMDV). Through our research, we were able to shed light on the binding affinity and phylogenetic relationships of these strains with the published SARS-CoV-2 reference genome, SARS-CoV, variants of concern (VOC) of SARS-CoV-2, and some closely related coronaviruses. These insights have important implications for future research efforts aimed at identifying within-host diversity, understanding the evolution and spread of these viruses, as well as the development of effective treatments and vaccines against them.

Do "Newly Born" orphan proteins resemble "Never Born" proteins? A study using three deep learning algorithms

"Newly Born" proteins, devoid of detectable homology to any other proteins, known as orphan proteins, occur in a single species or within a taxonomically restricted gene family. They are generated by the expression of novel open reading frames, and appear throughout evolution. We were curious if three recently developed programs for predicting protein structures, namely, AlphaFold2, RoseTTAFold, and ESMFold, might be of value for comparison of such "Newly Born" proteins to random polypeptides with amino acid content similar to that of native proteins, which have been called "Never Born" proteins. The programs were used to compare the structures of two sets of "Never Born" proteins that had been expressed-Group 1, which had been shown experimentally to possess substantial secondary structure, and Group 3, which had been shown to be intrinsically disordered. Overall, although the models generated were scored as being of low quality, they nevertheless revealed some general principles. Specifically, all four members of Group 1 were predicted to be compact by all three algorithms, in agreement with the experimental data, whereas the members of Group 3 were predicted to be very extended, as would be expected for intrinsically disordered proteins, again consistent with the experimental data. These predicted differences were shown to be statistically significant by comparing their accessible surface areas. The three programs were then used to predict the structures of three orphan proteins whose crystal structures had been solved, two of which display novel folds. Surprisingly, only for the protein which did not have a novel fold, and was taxonomically restricted, rather than being a true orphan, did all three algorithms predict very similar, high-quality structures, closely resembling the crystal structure. Finally, they were used to predict the structures of seven orphan proteins with well-identified biological functions, whose 3D structures are not known. Two proteins, which were predicted to be disordered based on their sequences, are predicted by all three structure algorithms to be extended structures. The other five were predicted to be compact structures with only two exceptions in the case of AlphaFold2. All three prediction algorithms make remarkably similar and high-quality predictions for one large protein, HCO_11565, from a nematode. It is conjectured that this is due to many homologs in the taxonomically restricted family of which it is a member, and to the fact that the Dali server revealed several nonrelated proteins with similar folds. An animated Interactive 3D Complement (I3DC) is available in Proteopedia at http://proteopedia.org/w/Journal:Proteins:3.

Integrated model of the vertebrate augmin complex

Accurate segregation of chromosomes is required to maintain genome integrity during cell division. This feat is accomplished by the microtubule-based spindle. To build a spindle rapidly and with high fidelity, cells take advantage of branching microtubule nucleation, which rapidly amplifies microtubules during cell division. Branching microtubule nucleation relies on the hetero-octameric augmin complex, but lack of structure information about augmin has hindered understanding how it promotes branching. In this work, we combine cryo-electron microscopy, protein structural prediction, and visualization of fused bulky tags via negative stain electron microscopy to identify the location and orientation of each subunit within the augmin structure. Evolutionary analysis shows that augmin's structure is highly conserved across eukaryotes, and that augmin contains a previously unidentified microtubule binding site. Thus, our findings provide insight into the mechanism of branching microtubule nucleation.

Effective Molecular Dynamics from Neural Network-Based Structure Prediction Models

Recent breakthroughs in neural network-based structure prediction methods, such as AlphaFold2 and RoseTTAFold, have dramatically improved the quality of computational protein structure prediction. These models also provide statistical confidence scores that can estimate uncertainties in the predicted structures, but it remains unclear to what extent these scores are related to the intrinsic conformational dynamics of proteins. Here, we compare AlphaFold2 prediction scores with explicit large-scale molecular dynamics simulations of 28 one- and two-domain proteins with varying degrees of flexibility. We demonstrate a strong correlation between the statistical prediction scores and the explicit motion derived from extensive atomistic molecular dynamics simulations and further derive an elastic network model based on the statistical scores of AlphFold2 (AF-ENM), which we benchmark in combination with coarse-grained molecular dynamics simulations. We show that our AF-ENM method reproduces the global protein dynamics with improved accuracy, providing a powerful way to derive effective molecular dynamics using neural network-based structure prediction models.

In Silico Exploration of Alternative Conformational States of VDAC

VDAC (Voltage-Dependent Anion-selective Channel) is the primary metabolite pore in the mitochondrial outer membrane (OM). Atomic structures of VDAC, consistent with its physiological "open" state, are β-barrels formed by 19 transmembrane (TM) β-strands and an N-terminal segment (NTERM) that folds inside the pore lumen. However, structures are lacking for VDAC's partially "closed" states. To provide clues about possible VDAC conformers, we used the RoseTTAFold neural network to predict structures for human and fungal VDAC sequences modified to mimic removal from the pore wall or lumen of "cryptic" domains, i.e., segments buried in atomic models yet accessible to antibodies in OM-bound VDAC. Predicted in vacuo structures for full-length VDAC sequences are 19-strand β-barrels similar to atomic models, but with weaker H-bonding between TM strands and reduced interactions between NTERM and the pore wall. Excision of combinations of "cryptic" subregions yields β-barrels with smaller diameters, wide gaps between N- and C-terminal β-strands, and in some cases disruption of the β-sheet (associated with strained backbone H-bond registration). Tandem repeats of modified VDAC sequences also were explored, as was domain swapping in monomer constructs. Implications of the results for possible alternative conformational states of VDAC are discussed.

Orthogonal analysis of variants in APOE gene using in-silico approaches reveals novel disrupting variants

Introduction: Alzheimer's disease (AD) is one of the most prominent medical conditions in the world. Understanding the genetic component of the disease can greatly advance our knowledge regarding its progression, treatment and prognosis. Single amino-acid variants (SAVs) in the APOE gene have been widely investigated as a risk factor for AD Studies, including genome-wide association studies, meta-analysis based studies, and in-vivo animal studies, were carried out to investigate the functional importance and pathogenesis potential of APOE SAVs. However, given the high cost of such large-scale or experimental studies, there are only a handful of variants being reported that have definite explanations. The recent development of in-silico analytical approaches, especially large-scale deep learning models, has opened new opportunities for us to probe the structural and functional importance of APOE variants extensively. Method: In this study, we are taking an ensemble approach that simultaneously uses large-scale protein sequence-based models, including Evolutionary Scale Model and AlphaFold, together with a few in-silico functional prediction web services to investigate the known and possibly disease-causing SAVs in APOE and evaluate their likelihood of being functional and structurally disruptive. Results: As a result, using an ensemble approach with little to no prior field-specific knowledge, we reported 5 SAVs in APOE gene to be potentially disruptive, one of which (C112R) was classificed by previous studies as a key risk factor for AD. Discussion: Our study provided a novel framework to analyze and prioritize the functional and structural importance of SAVs for future experimental and functional validation.

Deep learning for protein complex structure prediction

Recent developments in the structure prediction of protein complexes have resulted in accuracies rivalling experimental methods in many cases. The high accuracy is mainly observed in dimeric complexes and other problems such as protein disorder and predicting the structure of host-pathogen interactions remain. This review highlights the foundation for current accurate structure prediction of protein complexes and possible ways to address the remaining limitations.

Likelihood-based docking of models into cryo-EM maps

Optimized docking of models into cryo-EM maps requires exploiting an understanding of the signal expected in the data to minimize the calculation time while maintaining sufficient signal. The likelihood-based rotation function used in crystallography can be employed to establish plausible orientations in a docking search. A phased likelihood translation function yields scores for the placement and rigid-body refinement of oriented models. Optimized strategies for choices of the resolution of data from the cryo-EM maps to use in the calculations and the size of search volumes are based on expected log-likelihood-gain scores computed in advance of the search calculation. Tests demonstrate that the new procedure is fast, robust and effective at placing models into even challenging cryo-EM maps.

Pump-like channelrhodopsins: Not just bridging the gap between ion pumps and ion channels

Channelrhodopsins are microbial rhodopsins that work as light-gated ion channels. Their importance has become increasingly recognized due to their ability to control the membrane potential of specific cells in a light-dependent manner. This technology, termed optogenetics, has revolutionized neuroscience, and numerous channelrhodopsin variants have been isolated or engineered to expand the utility of optogenetics. Pump-like channelrhodopsins (PLCRs), one of the recently discovered channelrhodopsin subfamilies, have attracted broad attention due to their high sequence similarity to ion-pumping rhodopsins and their distinct properties, such as high light sensitivity and ion selectivity. In this review, we summarize the current understanding of the structure-function relationships of PLCRs and discuss the challenges and opportunities of channelrhodopsin research.

Ensemble-based, high-throughput virtual screening of potential inhibitor targeting putative farnesol dehydrogenase of Metisa plana (Lepidoptera: Psychidae)

Metisa plana (Lepidoptera: Psychidae) bagworm is a leaf-eater caterpillar ubiquitously found as a damaging pest in oil palm plantations, specifically in Malaysia. Various strategies have been implemented, including the usage of chemical insecticides. However, the main challenges include the development of insecticide resistance and its detrimental effects on the environment and non-target organisms. Therefore, a biorational insecticide is introduced by targeting the juvenile hormone (JH) biosynthetic pathway, which is mainly present in the insect and vital for the insect's growth, diapause, metamorphosis, and adult reproduction. This study aimed to investigate the potential inhibitor for the rate-limiting enzyme involved in the JH pathway known as farnesol dehydrogenase. A 255 amino acids sequence encoded for the putative M. plana farnesol dehydrogenase (MpFolDH) open reading frame had been identified and isolated. The three-dimensional structure of MpFolDH was predicted to have seven β- sheets with α-helices at both sides, showing typical characteristics for classical short-chain dehydrogenase and associated with oxidoreductase activity. Then, the ensemble-based virtual screening was conducted based on the ZINC20 database, in which 43 768 compounds that fulfilled pesticide-likeness criteria were screened by site-specific molecular docking. After a short molecular dynamics simulation (5 ns) was conducted towards 102 compounds, only the top 10 compounds based on their most favourable binding energy were selected for a more extended simulation (100 ns). Based on the protein-ligand stability, protein compactness, residues rigidity, binding interaction, binding energy throughout the 100 ns simulation, and physicochemical analysis, ZINC000408743205 was selected as a potential inhibitor for this enzyme. Amino acids decomposition analysis indicates Ile18, Ala95, Val198 and Val202 were the critical contributor residues for MpFolDH-inhibitors(s) complex.

Assessing structure and disorder prediction tools for de novo emerged proteins in the age of machine learning

Background: De novo protein coding genes emerge from scratch in the non-coding regions of the genome and have, per definition, no homology to other genes. Therefore, their encoded de novo proteins belong to the so-called "dark protein space". So far, only four de novo protein structures have been experimentally approximated. Low homology, presumed high disorder and limited structures result in low confidence structural predictions for de novo proteins in most cases. Here, we look at the most widely used structure and disorder predictors and assess their applicability for de novo emerged proteins. Since AlphaFold2 is based on the generation of multiple sequence alignments and was trained on solved structures of largely conserved and globular proteins, its performance on de novo proteins remains unknown. More recently, natural language models of proteins have been used for alignment-free structure predictions, potentially making them more suitable for de novo proteins than AlphaFold2. Methods: We applied different disorder predictors (IUPred3 short/long, flDPnn) and structure predictors, AlphaFold2 on the one hand and language-based models (Omegafold, ESMfold, RGN2) on the other hand, to four de novo proteins with experimental evidence on structure. We compared the resulting predictions between the different predictors as well as to the existing experimental evidence. Results: Results from IUPred, the most widely used disorder predictor, depend heavily on the choice of parameters and differ significantly from flDPnn which has been found to outperform most other predictors in a comparative assessment study recently. Similarly, different structure predictors yielded varying results and confidence scores for de novo proteins. Conclusions: We suggest that, while in some cases protein language model based approaches might be more accurate than AlphaFold2, the structure prediction of de novo emerged proteins remains a difficult task for any predictor, be it disorder or structure.

Helical allophycocyanin nanotubes absorb far-red light in a thermophilic cyanobacterium

To compete in certain low-light environments, some cyanobacteria express a paralog of the light-harvesting phycobiliprotein, allophycocyanin (AP), that strongly absorbs far-red light (FRL). Using cryo-electron microscopy and time-resolved absorption spectroscopy, we reveal the structure-function relationship of this FRL-absorbing AP complex (FRL-AP) that is expressed during acclimation to low light and that likely associates with chlorophyll a-containing photosystem I. FRL-AP assembles as helical nanotubes rather than typical toroids due to alterations of the domain geometry within each subunit. Spectroscopic characterization suggests that FRL-AP nanotubes are somewhat inefficient antenna; however, the enhanced ability to harvest FRL when visible light is severely attenuated represents a beneficial trade-off. The results expand the known diversity of light-harvesting proteins in nature and exemplify how biological plasticity is achieved by balancing resource accessibility with efficiency.

Unveiling the Impact of Gene Presence/Absence Variation in Driving Inter-Individual Sequence Diversity within the CRP-I Gene Family in Mytilus spp

Mussels (Mytilus spp.) tolerate infections much better than other species living in the same marine coastal environment thanks to a highly efficient innate immune system, which exploits a remarkable diversification of effector molecules involved in mucosal and humoral responses. Among these, antimicrobial peptides (AMPs) are subjected to massive gene presence/absence variation (PAV), endowing each individual with a potentially unique repertoire of defense molecules. The unavailability of a chromosome-scale assembly has so far prevented a comprehensive evaluation of the genomic arrangement of AMP-encoding loci, preventing an accurate ascertainment of the orthology/paralogy relationships among sequence variants. Here, we characterized the CRP-I gene cluster in the blue mussel Mytilus edulis, which includes about 50 paralogous genes and pseudogenes, mostly packed in a small genomic region within chromosome 5. We further reported the occurrence of widespread PAV within this family in the Mytilus species complex and provided evidence that CRP-I peptides likely adopt a knottin fold. We functionally characterized the synthetic peptide sCRP-I H1, assessing the presence of biological activities consistent with other knottins, revealing that mussel CRP-I peptides are unlikely to act as antimicrobial agents or protease inhibitors, even though they may be used as defense molecules against infections from eukaryotic parasites.

Helical allophycocyanin nanotubes absorb far-red light in a thermophilic cyanobacterium

To compete in certain low-light environments, some cyanobacteria express a paralog of the light-harvesting phycobiliprotein, allophycocyanin (AP), that strongly absorbs far-red light (FRL). Using cryo-electron microscopy and time-resolved absorption spectroscopy, we reveal the structure-function relationship of this FRL-absorbing AP complex (FRL-AP) that is expressed during acclimation to low light and that likely associates with chlorophyll a-containing photosystem I. FRL-AP assembles as helical nanotubes rather than typical toroids due to alterations of the domain geometry within each subunit. Spectroscopic characterization suggests that FRL-AP nanotubes are somewhat inefficient antenna; however, the enhanced ability to harvest FRL when visible light is severely attenuated represents a beneficial trade-off. The results expand the known diversity of light-harvesting proteins in nature and exemplify how biological plasticity is achieved by balancing resource accessibility with efficiency.

Cryo-EM structure of the fully assembled Elongator complex

Transfer RNA (tRNA) molecules are essential to decode messenger RNA codons during protein synthesis. All known tRNAs are heavily modified at multiple positions through post-transcriptional addition of chemical groups. Modifications in the tRNA anticodons are directly influencing ribosome decoding and dynamics during translation elongation and are crucial for maintaining proteome integrity. In eukaryotes, wobble uridines are modified by Elongator, a large and highly conserved macromolecular complex. Elongator consists of two subcomplexes, namely Elp123 containing the enzymatically active Elp3 subunit and the associated Elp456 hetero-hexamer. The structure of the fully assembled complex and the function of the Elp456 subcomplex have remained elusive. Here, we show the cryo-electron microscopy structure of yeast Elongator at an overall resolution of 4.3 Å. We validate the obtained structure by complementary mutational analyses in vitro and in vivo. In addition, we determined various structures of the murine Elongator complex, including the fully assembled mouse Elongator complex at 5.9 Å resolution. Our results confirm the structural conservation of Elongator and its intermediates among eukaryotes. Furthermore, we complement our analyses with the biochemical characterization of the assembled human Elongator. Our results provide the molecular basis for the assembly of Elongator and its tRNA modification activity in eukaryotes.

High-accuracy protein model quality assessment using attention graph neural networks

Great improvement has been brought to protein tertiary structure prediction through deep learning. It is important but very challenging to accurately rank and score decoy structures predicted by different models. CASP14 results show that existing quality assessment (QA) approaches lag behind the development of protein structure prediction methods, where almost all existing QA models degrade in accuracy when the target is a decoy of high quality. How to give an accurate assessment to high-accuracy decoys is particularly useful with the available of accurate structure prediction methods. Here we propose a fast and effective single-model QA method, QATEN, which can evaluate decoys only by their topological characteristics and atomic types. Our model uses graph neural networks and attention mechanisms to evaluate global and amino acid level scores, and uses specific loss functions to constrain the network to focus more on high-precision decoys and protein domains. On the CASP14 evaluation decoys, QATEN performs better than other QA models under all correlation coefficients when targeting average LDDT. QATEN shows promising performance when considering only high-accuracy decoys. Compared to the embedded evaluation modules of predicted ${C}_{\alpha^{-}} RMSD$ (pRMSD) in RosettaFold and predicted LDDT (pLDDT) in AlphaFold2, QATEN is complementary and capable of achieving better evaluation on some decoy structures generated by AlphaFold2 and RosettaFold. These results suggest that the new QATEN approach can be used as a reliable independent assessment algorithm for high-accuracy protein structure decoys.

SARS-CoV-2 receptor-binding domain deep mutational AlphaFold2 structures

Leveraging recent advances in computational modeling of proteins with AlphaFold2 (AF2) we provide a complete curated data set of all single mutations from each of the 7 main SARS-CoV-2 lineages spike protein receptor binding domain (RBD) resulting in 3819X7 = 26733 PDB structures. We visualize the generated structures and show that AF2 pLDDT values are correlated with state-of-the-art disorder approximations, implying some internal protein dynamics are also captured by the model. Joint increasing mutational coverage of both structural and phenotype data coupled with advances in machine learning can be leveraged to accelerate virology research, specifically future variant prediction. We hope this data release can offer assistance into further understanding of the local and global mutational landscape of SARS-CoV-2 as well as provide insight into the biological understanding that 3D structure acts as a bridge between protein genotype and phenotype.

Seasonal effects decouple SARS-CoV-2 haplotypes worldwide

Background: Variants of concern (VOCs) have been replacing each other during the still rampant COVID-19 pandemic. As a result, SARS-CoV-2 populations have evolved increasingly intricate constellations of mutations that often enhance transmissibility, disease severity, and other epidemiological characteristics. The origin and evolution of these constellations remain puzzling. Methods: Here we study the evolution of VOCs at the proteome level by analyzing about 12 million genomic sequences retrieved from GISAID on July 23, 2022. A total 183,276 mutations were identified and filtered with a relevancy heuristic. The prevalence of haplotypes and free-standing mutations was then tracked monthly in various latitude corridors of the world. Results: A chronology of 22 haplotypes defined three phases driven by protein flexibility-rigidity, environmental sensing, and immune escape. A network of haplotypes illustrated the recruitment and coalescence of mutations into major VOC constellations and seasonal effects of decoupling and loss. Protein interaction networks mediated by haplotypes predicted communications impacting the structure and function of proteins, showing the increasingly central role of molecular interactions involving the spike (S), nucleocapsid (N), and membrane (M) proteins. Haplotype markers either affected fusogenic regions while spreading along the sequence of the S-protein or clustered around binding domains. Modeling of protein structure with AlphaFold2 showed that VOC Omicron and one of its haplotypes were major contributors to the distortion of the M-protein endodomain, which behaves as a receptor of other structural proteins during virion assembly. Remarkably, VOC constellations acted cooperatively to balance the more extreme effects of individual haplotypes. Conclusions: Our study uncovers seasonal patterns of emergence and diversification occurring amid a highly dynamic evolutionary landscape of bursts and waves. The mapping of genetically-linked mutations to structures that sense environmental change with powerful ab initio modeling tools demonstrates the potential of deep-learning for COVID-19 predictive intelligence and therapeutic intervention.

SARS-CoV-2 Spike Protein Post-Translational Modification Landscape and Its Impact on Protein Structure and Function via Computational Prediction

To elucidate the role of post-translational modifications (PTMs) in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein's structure and virulence, we generated a high-resolution map of 87 PTMs using liquid chromatography with tandem mass spectrometry data on the extracted spike protein from SARS-CoV-2 virions and then reconstituted its structure heterogeneity caused by PTMs. Nonetheless, Alphafold2, a high-accuracy artificial intelligence tool to perform protein structure prediction, relies solely on primary amino acid sequence, whereas the impact of PTM, which often modulates critical protein structure and function, is much ignored. To overcome this challenge, we proposed the mutagenesis approach-an in silico, site-directed amino acid substitution to mimic the influence of PTMs on protein structure due to altered physicochemical properties in the post-translationally modified amino acids-and then reconstituted the spike protein's structure from the substituted sequences by Alphafold2. For the first time, the proposed method revealed predicted protein structures resulting from PTMs, a problem that Alphafold2 has yet to address. As an example, we performed computational analyses of the interaction of the post-translationally modified spike protein with its host factors such as angiotensin-converting enzyme 2 to illuminate binding affinity. Mechanistically, this study suggested the structural analysis of post-translationally modified protein via mutagenesis and deep learning. To summarize, the reconstructed spike protein structures showed that specific PTMs can be used to modulate host factor binding, guide antibody design, and pave the way for new therapeutic targets. The code and Supplementary Materials are freely available at https://github.com/LTZHKUSTGZ/SARS-CoV-2-spike-protein-PTM.

Human Glucosylceramide Synthase at Work as Provided by "In Silico" Molecular Docking, Molecular Dynamics, and Metadynamics

Glucosylceramide synthase (GCS) is an enzyme that catalyzes the first reaction of ceramide glycosylation in sphingolipid metabolism. It represents a primary target in the pharmacological treatment of some lysosomal storage diseases (LSDs), such as Gaucher and Niemann-Pick syndromes. In this study, starting from the model reported in the AlphaFold Protein Structure Database, the location and conformations of GCS substrates and cofactors have been provided by a step-by-step in silico procedure, by which the functional manganese ion and the substrates have been inserted in the GCS structure through combined molecular docking and full-atomistic molecular dynamics approaches, including metadynamics. A detailed analysis by structural dynamics of the complete model system, i.e., the enzyme anchored to the plasma membrane, containing the manganese ion and the two substrates, has been carried out to identify its complex conformational landscape by means of well-tempered metadynamics. A final structure was selected, in which both substrates were present in the active site of the enzyme at minimum distance, thus giving support to a SNi-type reaction mechanism for catalysis. Asp236, Glu235, and Asp144 are found to interact with the metal cofactor, which is able to trap the phosphates of UDP-glucose, while Gly210, Trp276, and Val208 cooperate to provide its correct orientation. Phe205, Cys207, Tyr237, and Leu284 form a pocket for the polar head of the ceramide, which is transiently placed in position to determine the catalytic event, when His193 interacts with the head of the ceramide, thus anchoring the substrate to the active site.

The structure of the high-affinity nickel-binding site in the Ni,Zn-HypA•UreE2 complex

The maturation pathway for the nickel-dependent enzyme urease utilizes the protein UreE as a metallochaperone to supply Ni(II) ions. In Helicobacter pylori urease maturation also requires HypA and HypB, accessory proteins that are commonly associated with hydrogenase maturation. Herein we report on the characterization of a protein complex formed between HypA and the UreE2 dimer. Nuclear magnetic resonance (NMR) coupled with molecular modelling show that the protein complex apo, Zn-HypA•UreE2, forms between the rigorously conserved Met-His-Glu (MHE motif) Ni-binding N-terminal sequence of HypA and the two conserved His102A and His102B located at the dimer interface of UreE2. This complex forms in the absence of Ni(II) and is supported by extensive protein contacts that include the use of the C-terminal sequences of UreE2 to form additional strands of β-sheet with the Ni-binding domain of HypA. The Ni-binding properties of apo, Zn-HypA•UreE2 and the component proteins were investigated by isothermal titration calorimetry using a global fitting strategy that included all of the relevant equilibria, and show that the Ni,Zn-HypA•UreE2 complex contains a single Ni(II)-binding site with a sub-nanomolar KD. The structural features of this novel Ni(II) site were elucidated using proteins produced with specifically deuterated amino acids, protein point mutations, and the analyses of X-ray absorption spectroscopy, hyperfine shifted NMR features, as well as molecular modeling coupled with quantum-mechanical calculations. The results show that the complex contains a six-coordinate, high-spin Ni(II) site with ligands provided by both component proteins.

TopEnzyme: a framework and database for structural coverage of the functional enzyme space

MOTIVATION

TopEnzyme is a database of structural enzyme models created with TopModel and is linked to the SWISS-MODEL repository and AlphaFold Protein Structure Database to provide an overview of structural coverage of the functional enzyme space for over 200 000 enzyme models. It allows the user to quickly obtain representative structural models for 60% of all known enzyme functions.

RESULTS

We assessed the models with TopScore and contributed 9039 good-quality and 1297 high-quality structures. Furthermore, we compared these models to AlphaFold2 models with TopScore and found that the TopScore differs only by 0.04 on average in favor of AlphaFold2. We tested TopModel and AlphaFold2 for targets not seen in the respective training databases and found that both methods create qualitatively similar structures. When no experimental structures are available, this database will facilitate quick access to structural models across the currently most extensive structural coverage of the functional enzyme space within Swiss-Prot.

AVAILABILITY AND IMPLEMENTATION

We provide a full web interface to the database at https://cpclab.uni-duesseldorf.de/topenzyme/.

An informatic workflow for the enhanced annotation of excretory/secretory proteins of Haemonchus contortus

Major advances in genomic and associated technologies have demanded reliable bioinformatic tools and workflows for the annotation of genes and their products via comparative analyses using well-curated reference data sets, accessible in public repositories. However, the accurate in silico annotation of molecules (proteins) encoded in organisms (e.g., multicellular parasites) which are evolutionarily distant from those for which these extensive reference data sets are available, including invertebrate model organisms (e.g., Caenorhabditis elegans - free-living nematode, and Drosophila melanogaster - the vinegar fly) and vertebrate species (e.g., Homo sapiens and Mus musculus), remains a major challenge. Here, we constructed an informatic workflow for the enhanced annotation of biologically-important, excretory/secretory (ES) proteins ("secretome") encoded in the genome of a parasitic roundworm, called Haemonchus contortus (commonly known as the barber's pole worm). We critically evaluated the performance of five distinct methods, refined some of them, and then combined the use of all five methods to comprehensively annotate ES proteins, according to gene ontology, biological pathways and/or metabolic (enzymatic) processes. Then, using optimised parameter settings, we applied this workflow to comprehensively annotate 2591 of all 3353 proteins (77.3%) in the secretome of H. contortus. This result is a substantial improvement (10-25%) over previous annotations using individual, "off-the-shelf" algorithms and default settings, indicating the ready applicability of the present, refined workflow to gene/protein sequence data sets from a wide range of organisms in the Tree-of-Life.

The accessory protein TagV is required for full Type VI secretion system activity in Serratia marcescens

The bacterial Type VI secretion system (T6SS) is a dynamic macromolecular structure that promotes inter- and intra-species competition through the delivery of toxic effector proteins into neighbouring cells. The T6SS contains 14 well-characterised core proteins necessary for effector delivery (TssA-M, PAAR). In this study, we have identified a novel accessory component required for optimal T6SS activity in the opportunistic pathogen Serratia marcescens, which we name TagV. Deletion of tagV, which encodes an outer membrane lipoprotein, caused a reduction in the T6SS-dependent antibacterial activity of S. marcescens Db10. Mutants of S. marcescens lacking the core component TssJ, a distinct outer membrane lipoprotein previously considered essential for T6SS firing, retained a modest T6SS activity that could be abolished through deletion of tagV. TagV did not interact with the T6SS membrane complex proteins TssL or TssM, but is proposed to bind to peptidoglycan, indicating that the mechanism by which TagV promotes T6SS firing differs from that of TssJ. Homologues of tagV were identified in several other bacterial genera, suggesting that the accessory function of TagV is not restricted to S. marcescens. Together, our findings support the existence of a second, TssJ-independent mechanism for T6SS firing that is dependent upon the activity of TagV proteins.

Distinct Energy-Coupling Factor Transporter Subunits Enable Flavin Acquisition and Extracytosolic Trafficking for Extracellular Electron Transfer in Listeria monocytogenes

A variety of electron transfer mechanisms link bacterial cytosolic electron pools with functionally diverse redox activities in the cell envelope and extracellular space. In Listeria monocytogenes, the ApbE-like enzyme FmnB catalyzes extracytosolic protein flavinylation, covalently linking a flavin cofactor to proteins that transfer electrons to extracellular acceptors. L. monocytogenes uses an energy-coupling factor (ECF) transporter complex that contains distinct substrate-binding, transmembrane, ATPase A, and ATPase A' subunits (RibU, EcfT, EcfA, and EcfA') to import environmental flavins, but the basis of extracytosolic flavin trafficking for FmnB flavinylation remains poorly defined. In this study, we show that the EetB and FmnA proteins are related to ECF transporter substrate-binding and transmembrane subunits, respectively, and are essential for exporting flavins from the cytosol for flavinylation. Comparisons of the flavin import versus export capabilities of L. monocytogenes strains lacking different ECF transporter subunits demonstrate a strict directionality of substrate-binding subunit transport but partial functional redundancy of transmembrane and ATPase subunits. Based on these results, we propose that ECF transporter complexes with different subunit compositions execute directional flavin import/export through a broadly conserved mechanism. Finally, we present genomic context analyses that show that related ECF exporter genes are distributed across members of the phylum Firmicutes and frequently colocalize with genes encoding flavinylated extracytosolic proteins. These findings clarify the basis of ECF transporter export and extracytosolic flavin cofactor trafficking in Firmicutes. IMPORTANCE Bacteria import vitamins and other essential compounds from their surroundings but also traffic related compounds from the cytosol to the cell envelope where they serve various functions. Studying the foodborne pathogen Listeria monocytogenes, we find that the modular use of subunits from a prominent class of bacterial transporters enables the import of environmental vitamin B2 cofactors and the extracytosolic trafficking of a vitamin B2-derived cofactor that facilitates redox reactions in the cell envelope. These studies clarify the basis of bidirectional small-molecule transport across the cytoplasmic membrane and the assembly of redox-active proteins within the cell envelope and extracellular space.

Variations in exons 11 and 12 of the multi-pest resistance wheat gene Lr34 are independently additive for leaf rust resistance

INTRODUCTION

Characterization of germplasm collections for the wheat leaf rust gene Lr34 previously defined five haplotypes in spring wheat. All resistant lines had a 3-bp TTC deletion (null) in exon 11, resulting in the absence of a phenylalanine residue in the ABC transporter, as well as a single nucleotide C (Tyrosine in Lr34+) to T (Histidine in Lr34-) transition in exon 12. A rare haplotype present in Odesskaja 13 and Koktunkulskaja 332, both of intermediate rust resistance, had the 3-bp deletion typical of Lr34+ in exon 11 but the T nucleotide of Lr34- in exon 12.

METHODS

To quantify the role of each mutation in leaf rust resistance, Odesskaja 13 and Koktunkulskaja 332 were crossed to Thatcher and its near-isogenic line Thatcher-Lr34 (RL6058). Single seed descent populations were generated and evaluated for rust resistance in six different rust nurseries.

RESULTS

The Odesskaja 13 progeny with the TTC/T haplotype were susceptible with an average severity rating of 62.3%, the null/T haplotype progeny averaged 39.7% and the null/C haplotype was highly resistant, averaging 13.3% severity. The numbers for the Koktunkulskaja 332 crosses were similar with 63.5%, 43.5% and 23.7% severity ratings, respectively. Differences between all classes in all crosses were statistically significant, indicating that both mutations are independently additive for leaf rust resistance. The three-dimensional structural models of LR34 were used to analyze the locations and putative interference of both amino acids with the transport channel. Koktunkulskaja 332 also segregated for marker csLV46 which is linked to Lr46. Rust severity in lines with Lr34+ and csLV46+ had significantly lower rust severity ratings than those without, indicating the additivity of the two loci.

DISCUSSION

This has implications for the deployment of Lr34 in wheat cultivars and for the basic understanding of this important wheat multi-pest durable resistance gene.

Comparison of the structure-function properties of wild-type human apoA-V and a C-terminal truncation associated with elevated plasma triglycerides

Background

Plasma triglycerides (TGs) are causally associated with coronary artery disease and acute pancreatitis. Apolipoprotein A-V (apoA-V, gene APOA5) is a liver-secreted protein that is carried on triglyceride-rich lipoproteins and promotes the enzymatic activity of lipoprotein lipase (LPL), thereby reducing TG levels. Little is known about apoA-V structure-function; naturally occurring human APOA5 variants can provide novel insights.

Methods

We used hydrogen-deuterium exchange mass spectrometry to determine the secondary structure of human apoA-V in lipid-free and lipid-associated conditions and identified a C-terminal hydrophobic face. Then, we used genomic data in the Penn Medicine Biobank to identify a rare variant, Q252X, predicted to specifically eliminate this region. We interrogated the function of apoA-V Q252X using recombinant protein in vitro and in vivo in apoa5 knockout mice.

Results

Human apoA-V Q252X carriers exhibited elevated plasma TG levels consistent with loss of function. Apoa5 knockout mice injected with AAV vectors expressing wildtype and variant APOA5-AAV recapitulated this phenotype. Part of the loss of function is due to reduced mRNA expression. Functionally, recombinant apoA-V Q252X was more readily soluble in aqueous solutions and more exchangeable with lipoproteins than WT apoA-V. Despite lacking the C-terminal hydrophobic region (a putative lipid binding domain) this protein also decreased plasma TG in vivo.

Conclusions

Deletion of apoA-V's C-terminus leads to reduced apoA-V bioavailability in vivo and higher TG levels. However, the C-terminus is not required for lipoprotein binding or enhancement of intravascular lipolytic activity. WT apoA-V is highly prone to aggregation, and this property is markedly reduced in recombinant apoA-V lacking the C-terminus.

Evaluation of AlphaFold structure-based protein stability prediction on missense variations in cancer

Identifying pathogenic missense variants in hereditary cancer is critical to the efforts of patient surveillance and risk-reduction strategies. For this purpose, many different gene panels consisting of different number and/or set of genes are available and we are particularly interested in a panel of 26 genes with a varying degree of hereditary cancer risk consisting of ABRAXAS1, ATM, BARD1, BLM, BRCA1, BRCA2, BRIP1, CDH1, CHEK2, EPCAM, MEN1, MLH1, MRE11, MSH2, MSH6, MUTYH, NBN, PALB2, PMS2, PTEN, RAD50, RAD51C, RAD51D, STK11, TP53, and XRCC2. In this study, we have compiled a collection of the missense variations reported in any of these 26 genes. More than a thousand missense variants were collected from ClinVar and the targeted screen of a breast cancer cohort of 355 patients which contributed to this set with 160 novel missense variations. We analyzed the impact of the missense variations on protein stability by five different predictors including both sequence- (SAAF2EC and MUpro) and structure-based (Maestro, mCSM, CUPSAT) predictors. For the structure-based tools, we have utilized the AlphaFold (AF2) protein structures which comprise the first structural analysis of this hereditary cancer proteins. Our results agreed with the recent benchmarks that computed the power of stability predictors in discriminating the pathogenic variants. Overall, we reported a low-to-medium-level performance for the stability predictors in discriminating pathogenic variants, except MUpro which had an AUROC of 0.534 (95% CI [0.499-0.570]). The AUROC values ranged between 0.614-0.719 for the total set and 0.596-0.682 for the set with high AF2 confidence regions. Furthermore, our findings revealed that the confidence score for a given variant in the AF2 structure could alone predict pathogenicity more robustly than any of the tested stability predictors with an AUROC of 0.852. Altogether, this study represents the first structural analysis of the 26 hereditary cancer genes underscoring 1) the thermodynamic stability predicted from AF2 structures as a moderate and 2) the confidence score of AF2 as a strong descriptor for variant pathogenicity.

Assessing protein homology models with docking reproducibility

Results of the recent Critical Assessment of Protein Structure (CASP) competitions demonstrate that protein backbones can be predicted with very high accuracy. In particular, the artificial intelligence methods of AlphaFold 2 from DeepMind were able to produce structures that were similar enough to experimental structures that many described the problem of protein prediction solved. However, for such structures to be used for drug docking studies requires precision in the placement of side chain atoms as well. Here we built a library of 1334 small molecules and examined how reproducibly they bound to the same site on a protein using QuickVina-W, a branch of the program Autodock that is optimized for blind searches. We discovered that the higher the backbone quality of the homology model the greater the similarity between the small molecule docking to the experimental and modeled structures. Furthermore, we found that specific subsets of this library were particularly useful for identifying small differences between the best of the best modeled structures. Specifically, when the number of rotatable bonds in the small molecule increased, differences in binding sites became more apparent.

Structure of the Saccharolobus solfataricus type III-D CRISPR effector

CRISPR-Cas is a prokaryotic adaptive immune system, classified into six different types, each characterised by a signature protein. Type III systems, classified based on the presence of a Cas10 subunit, are rather diverse multi-subunit assemblies with a range of enzymatic activities and downstream ancillary effectors. The broad array of current biotechnological CRISPR applications is mainly based on proteins classified as Type II, however recent developments established the feasibility and efficacy of multi-protein Type III CRISPR-Cas effector complexes as RNA-targeting tools in eukaryotes. The crenarchaeon Saccharolobus solfataricus has two type III system subtypes (III-B and III-D). Here, we report the cryo-EM structure of the Csm Type III-D complex from S. solfataricus (SsoCsm), which uses CRISPR RNA to bind target RNA molecules, activating the Cas10 subunit for antiviral defence. The structure reveals the complex organisation, subunit/subunit connectivity and protein/guide RNA interactions of the SsoCsm complex, one of the largest CRISPR effectors known.

Quality assessment of VHH models

Heavy Chain Only Antibodies are specific to Camelid species. Despite the lack of the light chain variable domain, their heavy chain variable domain (VH) domain, named VHH or nanobody, has promising potential applications in research and therapeutic fields. The structural study of VHH is therefore of great interest. Unfortunately, considering the huge amount of sequences that might be produced, only about one thousand of VHH experimental structures are publicly available in the Protein Data Bank, implying that structural model prediction of VHH is a necessary alternative to obtaining 3D information besides its sequence. The present study aims to assess and compare the quality of predictions from different modelling methodologies. Established comparative & homology modelling approaches to recent Deep Learning-based modelling strategies were applied, i.e. Modeller using single or multiple structural templates, ModWeb, SwissModel (with two evaluation schema), RoseTTAfold, AlphaFold 2 and NanoNet. The prediction accuracy was evaluated using RMSD, TM-score, GDT-TS, GDT-HA and Protein Blocks distance metrics. Besides the global structure assessment, we performed specific analyses of Frameworks and CDRs structures. We observed that AlphaFold 2 and especially NanoNet performed better than the other evaluated softwares. Importantly, we performed molecular dynamics simulations of an experimental structure and a NanoNet predicted model of a VHH in order to compare the global structural flexibility and local conformations using Protein Blocks. Despite rather similar structures, substantial differences in dynamical properties were observed, which underlies the complexity of the task of model evaluation.Communicated by Ramaswamy H. Sarma.

Deleterious, protein-altering variants in the transcriptional coregulator ZMYM3 in 27 individuals with a neurodevelopmental delay phenotype

Neurodevelopmental disorders (NDDs) result from highly penetrant variation in hundreds of different genes, some of which have not yet been identified. Using the MatchMaker Exchange, we assembled a cohort of 27 individuals with rare, protein-altering variation in the transcriptional coregulator ZMYM3, located on the X chromosome. Most (n = 24) individuals were males, 17 of which have a maternally inherited variant; six individuals (4 male, 2 female) harbor de novo variants. Overlapping features included developmental delay, intellectual disability, behavioral abnormalities, and a specific facial gestalt in a subset of males. Variants in almost all individuals (n = 26) are missense, including six that recurrently affect two residues. Four unrelated probands were identified with inherited variation affecting Arg441, a site at which variation has been previously seen in NDD-affected siblings, and two individuals have de novo variation resulting in p.Arg1294Cys (c.3880C>T). All variants affect evolutionarily conserved sites, and most are predicted to damage protein structure or function. ZMYM3 is relatively intolerant to variation in the general population, is widely expressed across human tissues, and encodes a component of the KDM1A-RCOR1 chromatin-modifying complex. ChIP-seq experiments on one variant, p.Arg1274Trp, indicate dramatically reduced genomic occupancy, supporting a hypomorphic effect. While we are unable to perform statistical evaluations to definitively support a causative role for variation in ZMYM3, the totality of the evidence, including 27 affected individuals, recurrent variation at two codons, overlapping phenotypic features, protein-modeling data, evolutionary constraint, and experimentally confirmed functional effects strongly support ZMYM3 as an NDD-associated gene.

AlphaFold2-multimer guided high-accuracy prediction of typical and atypical ATG8-binding motifs

Macroautophagy/autophagy is an intracellular degradation process central to cellular homeostasis and defense against pathogens in eukaryotic cells. Regulation of autophagy relies on hierarchical binding of autophagy cargo receptors and adaptors to ATG8/LC3 protein family members. Interactions with ATG8/LC3 are typically facilitated by a conserved, short linear sequence, referred to as the ATG8/LC3 interacting motif/region (AIM/LIR), present in autophagy adaptors and receptors as well as pathogen virulence factors targeting host autophagy machinery. Since the canonical AIM/LIR sequence can be found in many proteins, identifying functional AIM/LIR motifs has proven challenging. Here, we show that protein modelling using Alphafold-Multimer (AF2-multimer) identifies both canonical and atypical AIM/LIR motifs with a high level of accuracy. AF2-multimer can be modified to detect additional functional AIM/LIR motifs by using protein sequences with mutations in primary AIM/LIR residues. By combining protein modelling data from AF2-multimer with phylogenetic analysis of protein sequences and protein-protein interaction assays, we demonstrate that AF2-multimer predicts the physiologically relevant AIM motif in the ATG8-interacting protein 2 (ATI-2) as well as the previously uncharacterized noncanonical AIM motif in ATG3 from potato (Solanum tuberosum). AF2-multimer also identified the AIM/LIR motifs in pathogen-encoded virulence factors that target ATG8 members in their plant and human hosts, revealing that cross-kingdom ATG8-LIR/AIM associations can also be predicted by AF2-multimer. We conclude that the AF2-guided discovery of autophagy adaptors/receptors will substantially accelerate our understanding of the molecular basis of autophagy in all biological kingdoms.

Broad protective vaccination against systemic Escherichia coli with autotransporter antigens

Extraintestinal pathogenic Escherichia coli (ExPEC) is the leading cause of adult life-threatening sepsis and urinary tract infections (UTI). The emergence and spread of multidrug-resistant (MDR) ExPEC strains result in a considerable amount of treatment failure and hospitalization costs, and contribute to the spread of drug resistance amongst the human microbiome. Thus, an effective vaccine against ExPEC would reduce morbidity and mortality and possibly decrease carriage in healthy or diseased populations. A comparative genomic analysis demonstrated a gene encoding an invasin-like protein, termed sinH, annotated as an autotransporter protein, shows high prevalence in various invasive ExPEC phylogroups, especially those associated with systemic bacteremia and UTI. Here, we evaluated the protective efficacy and immunogenicity of a recombinant SinH-based vaccine consisting of either domain-3 or domains-1,2, and 3 of the putative extracellular region of surface-localized SinH. Immunization of a murine host with SinH-based antigens elicited significant protection against various strains of the pandemic ExPEC sequence type 131 (ST131) as well as multiple sequence types in two distinct models of infection (colonization and bacteremia). SinH immunization also provided significant protection against ExPEC colonization in the bladder in an acute UTI model. Immunized cohorts produced significantly higher levels of vaccine-specific serum IgG and urinary IgG and IgA, findings consistent with mucosal protection. Collectively, these results demonstrate that autotransporter antigens such as SinH may constitute promising ExPEC phylogroup-specific and sequence-type effective vaccine targets that reduce E. coli colonization and virulence.

PTCHD1 Binds Cholesterol but Not Sonic Hedgehog, Suggesting a Distinct Cellular Function

Deleterious mutations in the X-linked Patched domain-containing 1 (PTCHD1) gene may account for up to 1% of autism cases. Despite this, the PTCHD1 protein remains poorly understood. Structural similarities to Patched family proteins point to a role in sterol transport, but this hypothesis has not been verified experimentally. Additionally, PTCHD1 has been suggested to be involved in Hedgehog signalling, but thus far, the experimental results have been conflicting. To enable a variety of biochemical and structural experiments, we developed a method for expressing PTCHD1 in Spodoptera frugiperda cells, solubilising it in glycol-diosgenin, and purifying it to homogeneity. In vitro and in silico experiments show that PTCHD1 function is not interchangeable with Patched 1 (PTCH1) in canonical Hedgehog signalling, since it does not repress Smoothened in Ptch1^-/- mouse embryonic fibroblasts and does not bind Sonic Hedgehog. However, we found that PTCHD1 binds cholesterol similarly to PTCH1. Furthermore, we identified 13 PTCHD1-specific protein interactors through co-immunoprecipitation and demonstrated a link to cell stress responses and RNA stress granule formation. Thus, our results support the notion that despite structural similarities to other Patched family proteins, PTCHD1 may have a distinct cellular function.

Genomic landscape of the emerging XDR Salmonella Typhi for mining druggable targets clpP, hisH, folP and gpmI and screening of novel TCM inhibitors, molecular docking and simulation analyses

Typhoid fever is transmitted by ingestion of polluted water, contaminated food, and stool of typhoid-infected individuals, mostly in developing countries with poor hygienic environments. To find novel therapeutic targets and inhibitors, We employed a subtractive genomics strategy towards Salmonella Typhi and the complete genomes of eight strains were primarily subjected to the EDGAR tool to predict the core genome (n = 3207). Human non-homology (n = 2450) was followed by essential genes identification (n = 37). The STRING database predicted maximum protein-protein interactions, followed by cellular localization. The virulent/immunogenic ability of predicted genes were checked to differentiate drug and vaccine targets. Furthermore, the 3D models of the identified putative proteins encoded by the respective genes were constructed and subjected to druggability analyses where only "highly druggable" proteins were selected for molecular docking and simulation analyses. The putative targets ATP-dependent CLP protease proteolytic subunit, Imidazole glycerol phosphate synthase hisH, 7,8-dihydropteroate synthase folP and 2,3-bisphosphoglycerate-independent phosphoglycerate mutase gpmI were screened against a drug-like library (n = 12,000) and top hits were selected based on H-bonds, RMSD and energy scores. Finally, the ADMET properties for novel inhibitors ZINC19340748, ZINC09319798, ZINC00494142, ZINC32918650 were optimized followed by binding free energy (MM/PBSA) calculation for ligand-receptor complexes. The findings of this work are expected to aid in expediting the identification of novel protein targets and inhibitors in combating typhoid Salmonellosis, in addition to the already existing therapies.

How good are AlphaFold models for docking-based virtual screening?

A crucial component in structure-based drug discovery is the availability of high-quality three-dimensional structures of the protein target. Whenever experimental structures were not available, homology modeling has been, so far, the method of choice. Recently, AlphaFold (AF), an artificial-intelligence-based protein structure prediction method, has shown impressive results in terms of model accuracy. This outstanding success prompted us to evaluate how accurate AF models are from the perspective of docking-based drug discovery. We compared the high-throughput docking (HTD) performance of AF models with their corresponding experimental PDB structures using a benchmark set of 22 targets. The AF models showed consistently worse performance using four docking programs and two consensus techniques. Although AlphaFold shows a remarkable ability to predict protein architecture, this might not be enough to guarantee that AF models can be reliably used for HTD, and post-modeling refinement strategies might be key to increase the chances of success.

Improved model quality assessment using sequence and structural information by enhanced deep neural networks

Protein model quality assessment plays an important role in protein structure prediction, protein design and drug discovery. In this work, DeepUMQA2, a substantially improved version of DeepUMQA for protein model quality assessment, is proposed. First, sequence features containing protein co-evolution information and structural features reflecting family information are extracted to complement model-dependent features. Second, a novel backbone network based on triangular multiplication update and axial attention mechanism is designed to enhance information exchange between inter-residue pairs. On CASP13 and CASP14 datasets, the performance of DeepUMQA2 increases by 20.5 and 20.4% compared with DeepUMQA, respectively (measured by top 1 loss). Moreover, on the three-month CAMEO dataset (11 March to 04 June 2022), DeepUMQA2 outperforms DeepUMQA by 15.5% (measured by local AUC0,0.2) and ranks first among all competing server methods in CAMEO blind test. Experimental results show that DeepUMQA2 outperforms state-of-the-art model quality assessment methods, such as ProQ3D-LDDT, ModFOLD8, and DeepAccNet and DeepUMQA2 can select more suitable best models than state-of-the-art protein structure methods, such as AlphaFold2, RoseTTAFold and I-TASSER, provided themselves.

Design of a pan-betacoronavirus vaccine candidate through a phylogenetically informed approach

Coronaviruses are a diverse family of viruses that crossed over into humans at least seven times, precipitating mild to catastrophic outcomes. The severe acute respiratory syndrome coronavirus 2 pandemic renewed efforts to identify strains with zoonotic potential and to develop pan-coronavirus vaccines. The analysis of 2181 coronavirus genomes (from 102 host species) confirmed the limited sequence conservation across genera (alpha-, beta-, delta-, and gammacoronavirus) and proteins. A phylogenetically informed pan-coronavirus vaccine was not feasible because of high genetic heterogeneity across genera. We focused on betacoronaviruses and identified nonhuman-infecting receptor binding domain (RBD) sequences that were more genetically similar to human coronaviruses than expected given their phylogenetic divergence. These human-like RBDs defined three phylogenetic clusters. A vaccine candidate based on a representative sequence for each cluster covers the diversity estimated to protect against existing and future human-infecting betacoronaviruses. Our findings emphasize the potential value of conceptualizing prophylaxis against zoonoses in terms of genetic, rather than species, diversity.

Design of Diverse Asymmetric Pockets in De Novo Homo-oligomeric Proteins

A challenge for design of protein-small-molecule recognition is that incorporation of cavities with size, shape, and composition suitable for specific recognition can considerably destabilize protein monomers. This challenge can be overcome through binding pockets formed at homo-oligomeric interfaces between folded monomers. Interfaces surrounding the central homo-oligomer symmetry axes necessarily have the same symmetry and so may not be well suited to binding asymmetric molecules. To enable general recognition of arbitrary asymmetric substrates and small molecules, we developed an approach to designing asymmetric interfaces at off-axis sites on homo-oligomers, analogous to those found in native homo-oligomeric proteins such as glutamine synthetase. We symmetrically dock curved helical repeat proteins such that they form pockets at the asymmetric interface of the oligomer with sizes ranging from several angstroms, appropriate for binding a single ion, to up to more than 20 Å across. Of the 133 proteins tested, 84 had soluble expression in E. coli, 47 had correct oligomeric states in solution, 35 had small-angle X-ray scattering (SAXS) data largely consistent with design models, and 8 had negative-stain electron microscopy (nsEM) 2D class averages showing the structures coming together as designed. Both an X-ray crystal structure and a cryogenic electron microscopy (cryoEM) structure are close to the computational design models. The nature of these proteins as homo-oligomers allows them to be readily built into higher-order structures such as nanocages, and the asymmetric pockets of these structures open rich possibilities for small-molecule binder design free from the constraints associated with monomer destabilization.

Silencing of a Pectin Acetylesterase (PAE) Gene Highly Expressed in Tobacco Pistils Negatively Affects Pollen Tube Growth

Successful plant reproduction and fruit formation depend on adequate pollen and pistil development, and pollen-pistil interactions. In Nicotiana tabacum, pollen tubes grow through the intercellular spaces of pistil-specialized tissues, stigmatic secretory zone, and stylar transmitting tissue (STT). These intercellular spaces are supposed to be formed by the modulation of cell wall pectin esterification. Previously we have identified a gene preferentially expressed in pistils encoding a putative pectin acetylesterase (PAE), named NtPAE1. Here, we characterized the NtPAE1 gene and performed genome-wide and phylogenetic analyses of PAEs. We identified 30 PAE sequences in the N. tabacum genome, distributed in four clades. The expression of NtPAE1 was assessed by RT-qPCR and in situ hybridization. We confirmed NtPAE1 preferential expression in stigmas/styles and ovaries and demonstrated its high expression in the STT. Structural predictions and comparisons between NtPAE1 and functional enzymes validated its identity as a PAE. Transgenic plants were produced, overexpressing and silencing the NtPAE1 gene. Overexpressed plants displayed smaller flowers while silencing plants exhibited collapsed pollen grains, which hardly germinate. NtPAE1 silencing plants do not produce fruits, due to impaired pollen tube growth in their STTs. Thus, NtPAE1 is an essential enzyme regulating pectin modifications in flowers and, ultimately, in plant reproduction.

InterPro in 2022

The InterPro database (https://www.ebi.ac.uk/interpro/) provides an integrative classification of protein sequences into families, and identifies functionally important domains and conserved sites. Here, we report recent developments with InterPro (version 90.0) and its associated software, including updates to data content and to the website. These developments extend and enrich the information provided by InterPro, and provide a more user friendly access to the data. Additionally, we have worked on adding Pfam website features to the InterPro website, as the Pfam website will be retired in late 2022. We also show that InterPro's sequence coverage has kept pace with the growth of UniProtKB. Moreover, we report the development of a card game as a method of engaging the non-scientific community. Finally, we discuss the benefits and challenges brought by the use of artificial intelligence for protein structure prediction.

RCSB Protein Data Bank (RCSB.org): delivery of experimentally-determined PDB structures alongside one million computed structure models of proteins from artificial intelligence/machine learning

The Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB), founding member of the Worldwide Protein Data Bank (wwPDB), is the US data center for the open-access PDB archive. As wwPDB-designated Archive Keeper, RCSB PDB is also responsible for PDB data security. Annually, RCSB PDB serves >10 000 depositors of three-dimensional (3D) biostructures working on all permanently inhabited continents. RCSB PDB delivers data from its research-focused RCSB.org web portal to many millions of PDB data consumers based in virtually every United Nations-recognized country, territory, etc. This Database Issue contribution describes upgrades to the research-focused RCSB.org web portal that created a one-stop-shop for open access to ∼200 000 experimentally-determined PDB structures of biological macromolecules alongside >1 000 000 incorporated Computed Structure Models (CSMs) predicted using artificial intelligence/machine learning methods. RCSB.org is a 'living data resource.' Every PDB structure and CSM is integrated weekly with related functional annotations from external biodata resources, providing up-to-date information for the entire corpus of 3D biostructure data freely available from RCSB.org with no usage limitations. Within RCSB.org, PDB structures and the CSMs are clearly identified as to their provenance and reliability. Both are fully searchable, and can be analyzed and visualized using the full complement of RCSB.org web portal capabilities.

Using AlphaFold to predict the impact of single mutations on protein stability and function

AlphaFold changed the field of structural biology by achieving three-dimensional (3D) structure prediction from protein sequence at experimental quality. The astounding success even led to claims that the protein folding problem is "solved". However, protein folding problem is more than just structure prediction from sequence. Presently, it is unknown if the AlphaFold-triggered revolution could help to solve other problems related to protein folding. Here we assay the ability of AlphaFold to predict the impact of single mutations on protein stability (ΔΔG) and function. To study the question we extracted the pLDDT and <pLDDT> metrics from AlphaFold predictions before and after single mutation in a protein and correlated the predicted change with the experimentally known ΔΔG values. Additionally, we correlated the same AlphaFold pLDDT metrics with the impact of a single mutation on structure using a large scale dataset of single mutations in GFP with the experimentally assayed levels of fluorescence. We found a very weak or no correlation between AlphaFold output metrics and change of protein stability or fluorescence. Our results imply that AlphaFold may not be immediately applied to other problems or applications in protein folding.

1D2DSimScore: A novel method for comparing contacts in biomacromolecules and their complexes

The biologically relevant structures of proteins and nucleic acids and their complexes are dynamic. They include a combination of regions ranging from rigid structural segments to structural switches to regions that are almost always disordered, which interact with each other in various ways. Comparing conformational changes and variation in contacts between different conformational states is essential to understand the biological functions of proteins, nucleic acids, and their complexes. Here, we describe a new computational tool, 1D2DSimScore, for comparing contacts and contact interfaces in all kinds of macromolecules and macromolecular complexes, including proteins, nucleic acids, and other molecules. 1D2DSimScore can be used to compare structural features of macromolecular models between alternative structures obtained in a particular experiment or to score various predictions against a defined "ideal" reference structure. Comparisons at the level of contacts are particularly useful for flexible molecules, for which comparisons in 3D that require rigid-body superpositions are difficult, and in biological systems where the formation of specific inter-residue contacts is more relevant for the biological function than the maintenance of a specific global 3D structure. Similarity/dissimilarity scores calculated by 1D2DSimScore can be used to complement scores describing 3D structural similarity measures calculated by the existing tools.

Contact-Assisted Threading in Low-Homology Protein Modeling

The ability to successfully predict the three-dimensional structure of a protein from its amino acid sequence has made considerable progress in the recent past. The progress is propelled by the improved accuracy of deep learning-based inter-residue contact map predictors coupled with the rising growth of protein sequence databases. Contact map encodes interatomic interaction information that can be exploited for highly accurate prediction of protein structures via contact map threading even for the query proteins that are not amenable to direct homology modeling. As such, contact-assisted threading has garnered considerable research effort. In this chapter, we provide an overview of existing contact-assisted threading methods while highlighting the recent advances and discussing some of the current limitations and future prospects in the application of contact-assisted threading for improving the accuracy of low-homology protein modeling.

ManyFold: an efficient and flexible library for training and validating protein folding models

SUMMARY

ManyFold is a flexible library for protein structure prediction with deep learning that (i) supports models that use both multiple sequence alignments (MSAs) and protein language model (pLM) embedding as inputs, (ii) allows inference of existing models (AlphaFold and OpenFold), (iii) is fully trainable, allowing for both fine-tuning and the training of new models from scratch and (iv) is written in Jax to support efficient batched operation in distributed settings. A proof-of-concept pLM-based model, pLMFold, is trained from scratch to obtain reasonable results with reduced computational overheads in comparison to AlphaFold.

AVAILABILITY AND IMPLEMENTATION

The source code for ManyFold, the validation dataset and a small sample of training data are available at https://github.com/instadeepai/manyfold.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

3D-equivariant graph neural networks for protein model quality assessment

MOTIVATION

Quality assessment (QA) of predicted protein tertiary structure models plays an important role in ranking and using them. With the recent development of deep learning end-to-end protein structure prediction techniques for generating highly confident tertiary structures for most proteins, it is important to explore corresponding QA strategies to evaluate and select the structural models predicted by them since these models have better quality and different properties than the models predicted by traditional tertiary structure prediction methods.

RESULTS

We develop EnQA, a novel graph-based 3D-equivariant neural network method that is equivariant to rotation and translation of 3D objects to estimate the accuracy of protein structural models by leveraging the structural features acquired from the state-of-the-art tertiary structure prediction method-AlphaFold2. We train and test the method on both traditional model datasets (e.g. the datasets of the Critical Assessment of Techniques for Protein Structure Prediction) and a new dataset of high-quality structural models predicted only by AlphaFold2 for the proteins whose experimental structures were released recently. Our approach achieves state-of-the-art performance on protein structural models predicted by both traditional protein structure prediction methods and the latest end-to-end deep learning method-AlphaFold2. It performs even better than the model QA scores provided by AlphaFold2 itself. The results illustrate that the 3D-equivariant graph neural network is a promising approach to the evaluation of protein structural models. Integrating AlphaFold2 features with other complementary sequence and structural features is important for improving protein model QA.

AVAILABILITY AND IMPLEMENTATION

The source code is available at https://github.com/BioinfoMachineLearning/EnQA.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Quality Estimates for 3D Protein Models

Protein structure modeling is one of the most advanced and complex processes in computational biology. One of the major problems for the protein structure prediction field has been how to estimate the accuracy of the predicted 3D models, on both a local and global level, in the absence of known structures. We must be able to accurately measure the confidence that we have in the quality predicted 3D models of proteins for them to become widely adopted by the general bioscience community. To address this major issue, it was necessary to develop new model quality assessment (MQA) methods and integrate them into our pipelines for building 3D protein models. Our MQA method, called ModFOLD, has been ranked as one of the most accurate MQA tools in independent blind evaluations. This chapter discusses model quality assessment in the protein modeling field, demonstrating both its strengths and limitations. We also present some of the best methods according to independent benchmarking data, which has been gathered in recent years.