Validated determination of NRG1 Ig-like domain structure by mass spectrometry coupled with computational modeling

Radical Protein Footprinting in Mammalian Whole Blood

Hydroxyl Radical Protein Footprinting (HRPF) is a powerful method to probe the solvent-accessible surface area of proteins. It is mostly used to study the higher-order structure of proteins, as well as protein-protein and protein-carbohydrate interactions. Hydroxyl radicals are generated by the photolysis of hydrogen peroxide and these radicals modify the surface amino acids. Bottom-up proteomics is then applied and peptide oxidation is calculated and correlated with solvent accessibility. It is mainly performed in vitro; however, it has been recently used in living systems, including live cells, live nematodes, and 3D cell cultures. Mammalian tissues are still out of reach as they absorb UV strongly, hindering radical generation. Here, we describe the first example of RPF in mammalian stabilized whole blood. Using photoactivation of persulfate with a commercially available FOX Photolysis System modified for sample handling and inline mixing, we demonstrate the first labeling of proteins in whole blood. We demonstrate that the RPF protocol does not alter the blood cell gross morphology outside of a moderate hypertonicity equivalent to sodium chloride exposure prior to labeling. We detail an improved quenching protocol to limit background labeling in persulfate RPF. We describe the labeling of the top ten most abundant proteins in the blood. We demonstrate the equivalence of ex vivo labeling in whole blood with labeling of the same structure in vitro using hemoglobin as a test system. Overall, these results now open the possibility of performing RPF-based structural proteomics in pre-clinical models and using readily available clinical samples.

Systematic Fe(II)-EDTA Method of Dose-Dependent Hydroxyl Radical Generation for Protein Oxidative Footprinting

Correlating the structure and dynamics of proteins with biological function is critical to understanding normal and dysfunctional cellular mechanisms. We describe a quantitative method of hydroxyl radical generation via Fe(II)-ethylenediaminetetraacetic acid (EDTA)-catalyzed Fenton chemistry that provides ready access to protein oxidative footprinting using equipment commonly found in research and process control laboratories. Robust and reproducible dose-dependent oxidation of protein samples is observed and quantitated by mass spectrometry with as fine a single residue resolution. An oxidation analysis of lysozyme provides a readily accessible benchmark for our method. The efficacy of our oxidation method is demonstrated by mapping the interface of a RAS-monobody complex, the surface of the NIST mAb, and the interface between PRC2 complex components. These studies are executed using standard laboratory tools and a few pennies of reagents; the mass spectrometry analysis can be streamlined to map the protein structure with single amino acid residue resolution.

Dimethylthiourea as a Quencher in Hydroxyl Radical Protein Footprinting Experiments

Hydroxyl radical protein footprinting (HRPF) is a mass-spectrometry-based method for studying protein structures, interactions, conformations, and folding. This method is based on the irreversible labeling of solvent-exposed amino acid side chains by hydroxyl radicals. While catalase is commonly used as a quencher after the labeling of a protein by the hydroxyl radicals to efficiently remove the remaining hydrogen peroxide, it has some disadvantages. Catalase quenching adds a relatively high amount of protein to the sample, limiting the sensitivity of the method due to dynamic range issues and causing significant issues when dealing with more complex samples. We evaluated dimethylthiourea (DMTU) as a replacement for catalase in the quenching HRPF reactions. We observed that DMTU is highly effective at quenching HRPF oxidation. DMTU does not cause the background protein issues that catalase does, resulting in an increased number of protein identifications from complex mixtures. We recommend the replacement of catalase quenching with DMTU for all HRPF experiments.

Blind assessment of monomeric AlphaFold2 protein structure models with experimental NMR data

Recent advances in molecular modeling of protein structures are changing the field of structural biology. AlphaFold-2 (AF2), an AI system developed by DeepMind, Inc., utilizes attention-based deep learning to predict models of protein structures with high accuracy relative to structures determined by X-ray crystallography and cryo-electron microscopy (cryoEM). Comparing AF2 models to structures determined using solution NMR data, both high similarities and distinct differences have been observed. Since AF2 was trained on X-ray crystal and cryoEM structures, we assessed how accurately AF2 can model small, monomeric, solution protein NMR structures which (i) were not used in the AF2 training data set, and (ii) did not have homologous structures in the Protein Data Bank at the time of AF2 training. We identified nine open-source protein NMR data sets for such "blind" targets, including chemical shift, raw NMR FID data, NOESY peak lists, and (for 1 case) 15N-1H residual dipolar coupling data. For these nine small (70-108 residues) monomeric proteins, we generated AF2 prediction models and assessed how well these models fit to these experimental NMR data, using several well-established NMR structure validation tools. In most of these cases, the AF2 models fit the NMR data nearly as well, or sometimes better than, the corresponding NMR structure models previously deposited in the Protein Data Bank. These results provide benchmark NMR data for assessing new NMR data analysis and protein structure prediction methods. They also document the potential for using AF2 as a guiding tool in protein NMR data analysis, and more generally for hypothesis generation in structural biology research.

Blind Assessment of Monomeric AlphaFold2 Protein Structure Models with Experimental NMR Data

Recent advances in molecular modeling of protein structures are changing the field of structural biology. AlphaFold-2 (AF2), an AI system developed by DeepMind, Inc., utilizes attention-based deep learning to predict models of protein structures with high accuracy relative to structures determined by X-ray crystallography and cryo-electron microscopy (cryoEM). Comparing AF2 models to structures determined using solution NMR data, both high similarities and distinct differences have been observed. Since AF2 was trained on X-ray crystal and cryoEM structures, we assessed how accurately AF2 can model small, monomeric, solution protein NMR structures which (i) were not used in the AF2 training data set, and (ii) did not have homologous structures in the Protein Data Bank at the time of AF2 training. We identified nine open source protein NMR data sets for such "blind" targets, including chemical shift, raw NMR FID data, NOESY peak lists, and (for 1 case) 15 N- 1 H residual dipolar coupling data. For these nine small (70 - 108 residues) monomeric proteins, we generated AF2 prediction models and assessed how well these models fit to these experimental NMR data, using several well-established NMR structure validation tools. In most of these cases, the AF2 models fit the NMR data nearly as well, or sometimes better than, the corresponding NMR structure models previously deposited in the Protein Data Bank. These results provide benchmark NMR data for assessing new NMR data analysis and protein structure prediction methods. They also document the potential for using AF2 as a guiding tool in protein NMR data analysis, and more generally for hypothesis generation in structural biology research.

Highlights

AF2 models assessed against NMR data for 9 monomeric proteins not used in training.AF2 models fit NMR data almost as well as the experimentally-determined structures. RPF-DP, PSVS , and PDBStat software provide structure quality and RDC assessment. RPF-DP analysis using AF2 models suggests multiple conformational states.

Protein complex prediction using Rosetta, AlphaFold, and mass spectrometry covalent labeling

Covalent labeling (CL) in combination with mass spectrometry can be used as an analytical tool to study and determine structural properties of protein-protein complexes. However, data from these experiments is sparse and does not unambiguously elucidate protein structure. Thus, computational algorithms are needed to deduce structure from the CL data. In this work, we present a hybrid method that combines models of protein complex subunits generated with AlphaFold with differential CL data via a CL-guided protein-protein docking in Rosetta. In a benchmark set, the RMSD (root-mean-square deviation) of the best-scoring models was below 3.6 Å for 5/5 complexes with inclusion of CL data, whereas the same quality was only achieved for 1/5 complexes without CL data. This study suggests that our integrated approach can successfully use data obtained from CL experiments to distinguish between nativelike and non-nativelike models.

Application of Multiple Length Cross-linkers to the Characterization of Gaseous Protein Structure

The speed, sensitivity, and tolerance of heterogeneity, as well as the kinetic trapping of solution-like states during electrospray, make native mass spectrometry an attractive method to study protein structure. Increases in the resolution of ion mobility measurements and in mass resolving power and range are leading to the increase of the information content of intact protein measurements and an expanded role of mass spectrometry in structural biology. Herein, a suite of different length noncovalent (sulfonate to positively charged side chain) cross-linkers was introduced via gas-phase ion/ion chemistry and used to determine distance restraints of kinetically trapped gas-phase structures of native-like cytochrome c ions. Electron capture dissociation allowed for the identification of cross-linked sites. Different length linkers resulted in distinct pairs of side chains being linked, supporting the ability of gas-phase cross-linking to be structurally specific. The gas-phase lengths of the cross-linkers were determined by conformational searches and density functional theory, allowing for the interpretation of the cross-links as distance restraints. These distance restraints were used to model gas-phase structures with molecular dynamics simulations, revealing a mixture of structures with similar overall shape/size but distinct features, thereby illustrating the kinetic trapping of multiple native-like solution structures in the gas phase.