Homonuclear Super-Resolution NMR Spectroscopy.

Homonuclear 1H NMR (nuclear magnetic resonance) spectra such as the [1H,1H]-NOESY (Nuclear Overhauser Enhancement spectroscopy) are essential for biomolecular structural biology. However, the spectrum contains hundreds to thousands of cross peaks meaning that within approximately 100 ppm2 there is significant signal overlap. Spectral resolution is thus a limiting factor for data interpretation for dynamics and structure elucidation. Acquiring the spectra at higher magnetic fields such as at a 1.2 GHz 1H frequency helps to reduce spectral crowding, since resolution scales proportionally to the magnetic field strength. Here, we show that the linewidths' of cross peaks in [1H,1H]-NOESY and [1H,1H]-TOCSY spectra can be further reduced by a factor of 2-3 in each dimension by super-resolution spectroscopy. In the indirect dimension a composite exponential-cosine weighted number of scans along the time increments are recorded and digitally smoothened by a window function. Furthermore, measurement time saving by reduced-acquisition super-resolution (RASR) is introduced. Application to the 20 kDa protein KRAS shows that highly resolved NMR spectra suitable for automated analysis can be acquired within less than 3 hours. The method opens an avenue towards automated chemical shift assignment, dynamics and structure determination of unlabeled small and medium size proteins within 24 hours.

1H, 13C, and 15N resonance assignments and solution structure of the N-terminal divergent calponin homology (NN-CH) domain of human intraflagellar transport protein 54.

The intraflagellar transport (IFT) machinery plays a crucial role in the bidirectional trafficking of components necessary for ciliary signaling, such as the Hedgehog, Wnt/PCR, and cAMP/PKA systems. Defects in some components of the IFT machinery cause dysfunction, leading to a wide range of human diseases and developmental disorders termed ciliopathies, such as nephronophthisis. The IFT machinery comprises three sub-complexes: BBsome, IFT-A, and IFT-B. The IFT protein 54 (IFT54) is an important component of the IFT-B sub-complex. In anterograde movement, IFT54 binds to active kinesin-II, walking along the cilia microtubule axoneme and carrying the dynein-2 complex in an inactive state, which works for retrograde movement. Several mutations in IFT54 are known to cause Senior-Loken syndrome, a ciliopathy. IFT54 possesses a divergent Calponin Homology (CH) domain termed as NN-CH domain at its N-terminus. However, several aspects of the function of the NN-CH domain of IFT54 are still obscure. Here, we report the 1H, 15N, and 13C resonance assignments of the NN-CH domain of human IFT54 and its solution structure. The NN-CH domain of human IFT54 adopts essentially the α1-α2-α3-α4-α5 topology as that of mouse IFT54, whose structure was determined by X-ray crystallographic study. The structural information and assignments obtained in this study shed light on the molecular function of the NN-CH domain in IFT54.

The 100-protein NMR spectra dataset: A resource for biomolecular NMR data analysis.

Multidimensional NMR spectra are the basis for studying proteins by NMR spectroscopy and crucial for the development and evaluation of methods for biomolecular NMR data analysis. Nevertheless, in contrast to derived data such as chemical shift assignments in the BMRB and protein structures in the PDB databases, this primary data is in general not publicly archived. To change this unsatisfactory situation, we present a standardized set of solution NMR data comprising 1329 2-4-dimensional NMR spectra and associated reference (chemical shift assignments, structures) and derived (peak lists, restraints for structure calculation, etc.) annotations. With the 100-protein NMR spectra dataset that was originally compiled for the development of the ARTINA deep learning-based spectra analysis method, 100 protein structures can be reproduced from their original experimental data. The 100-protein NMR spectra dataset is expected to help the development of computational methods for NMR spectroscopy, in particular machine learning approaches, and enable consistent and objective comparisons of these methods.

Conformations of Macrocyclic Peptides Sampled by Nuclear Magnetic Resonance: Models for Cell-Permeability.

The biological activities and pharmacological properties of peptides and peptide mimetics are determined by their conformational states. Therefore, a detailed understanding of the conformational landscape is crucial for rational drug design. Nuclear magnetic resonance (NMR) is the only method for structure determination in solution. However, it remains challenging to determine the structures of peptides using NMR because of very weak nuclear Overhauser effects (NOEs), the semiquantitative nature of the rotating frame Overhauser effect (ROE), and the low number of NOEs/ROEs in N-methylated peptides. In this study, we introduce a new approach to investigating the structures of modified macrocyclic peptides. We utilize exact NOEs (eNOEs) in viscous solvent mixtures to replicate various cellular environments. eNOEs provide detailed structural information for highly dynamic modified peptides. Structures of high precision were obtained for cyclosporin A, with a backbone atom rmsd of 0.10 Å. Distinct conformational states in different environments were identified for omphalotin A (OmphA), a fungal nematotoxic and multiple backbone N-methylated macrocyclic peptides. A model for cell-permeation is presented for OmphA, based on its structures in polar, apolar, and mixed polarity solvents. During the transition from a polar to an apolar environment, OmphA undergoes a rearrangement of its H-bonding network, accompanied by a cis to trans isomerization of the ω torsion angle within a type VIa ß-turn. We hypothesize that the kinetics of these conformational transitions play a crucial role in determining the membrane-permeation capabilities of OmphA.

Time-optimized protein NMR assignment with an integrative deep learning approach using AlphaFold and chemical shift prediction.

Chemical shift assignment is vital for nuclear magnetic resonance (NMR)-based studies of protein structures, dynamics, and interactions, providing crucial atomic-level insight. However, obtaining chemical shift assignments is labor intensive and requires extensive measurement time. To address this limitation, we previously proposed ARTINA, a deep learning method for automatic assignment of two-dimensional (2D)-4D NMR spectra. Here, we present an integrative approach that combines ARTINA with AlphaFold and UCBShift, enabling chemical shift assignment with reduced experimental data, increased accuracy, and enhanced robustness for larger systems, as presented in a comprehensive study with more than 5000 automated assignment calculations on 89 proteins. We demonstrate that five 3D spectra yield more accurate assignments (92.59%) than pure ARTINA runs using all experimentally available NMR data (on average 10 3D spectra per protein, 91.37%), considerably reducing the required measurement time. We also showcase automated assignments of only 15N-labeled samples, and report improved assignment accuracy in larger synthetic systems of up to 500 residues.

Efficient determination of the accessible conformation space of multi-domain complexes based on EPR PELDOR data.

Many proteins can adopt multiple conformations which are important for their function. This is also true for proteins and domains that are covalently linked to each other. One important example is ubiquitin, which can form chains of different conformations depending on which of its lysine side chains is used to form an isopeptide bond with the C-terminus of another ubiquitin molecule. Similarly, ubiquitin gets covalently attached to active-site residues of E2 ubiquitin-conjugating enzymes. Due to weak interactions between ubiquitin and its interaction partners, these covalent complexes adopt multiple conformations. Understanding the function of these complexes requires the characterization of the entire accessible conformation space and its modulation by interaction partners. Long-range (1.8-10 nm) distance restraints obtained by EPR spectroscopy in the form of probability distributions are ideally suited for this task as not only the mean distance but also information about the conformation dynamics is encoded in the experimental data. Here we describe a computational method that we have developed based on well-established structure determination software using NMR restraints to calculate the accessible conformation space using PELDOR/DEER data.

An Analysis of Nucleotide-Amyloid Interactions Reveals Selective Binding to Codon-Sized RNA.

Interactions between RNA and proteins are the cornerstone of many important biological processes from transcription and translation to gene regulation, yet little is known about the ancient origin of said interactions. We hypothesized that peptide amyloids played a role in the origin of life and that their repetitive structure lends itself to building interfaces with other polymers through avidity. Here, we report that short RNA with a minimum length of three nucleotides binds in a sequence-dependent manner to peptide amyloids. The 3'-5' linked RNA backbone appears to be well-suited to support these interactions, with the phosphodiester backbone and nucleobases both contributing to the affinity. Sequence-specific RNA-peptide interactions of the kind identified here may provide a path to understanding one of the great mysteries rooted in the origin of life: the origin of the genetic code.

Facilitating the structural characterisation of non-canonical amino acids in biomolecular NMR.

Peptides and proteins containing non-canonical amino acids (ncAAs) are a large and important class of biopolymers. They include non-ribosomally synthesised peptides, post-translationally modified proteins, expressed or synthesised proteins containing unnatural amino acids, and peptides and proteins that are chemically modified. Here, we describe a general procedure for generating atomic descriptions required to incorporate ncAAs within popular NMR structure determination software such as CYANA, CNS, Xplor-NIH and ARIA. This procedure is made publicly available via the existing Automated Topology Builder (ATB) server (https://atb.uq.edu.au, last access: 17 February 2023) with all submitted ncAAs stored in a dedicated database. The described procedure also includes a general method for linking of side chains of amino acids from CYANA templates. To ensure compatibility with other systems, atom names comply with IUPAC guidelines. In addition to describing the workflow, 3D models of complex natural products generated by CYANA are presented, including vancomycin. In order to demonstrate the manner in which the templates for ncAAs generated by the ATB can be used in practice, we use a combination of CYANA and CNS to solve the structure of a synthetic peptide designed to disrupt Alzheimer-related protein-protein interactions. Automating the generation of structural templates for ncAAs will extend the utility of NMR spectroscopy to studies of more complex biomolecules, with applications in the rapidly growing fields of synthetic biology and chemical biology. The procedures we outline can also be used to standardise the creation of structural templates for any amino acid and thus have the potential to impact structural biology more generally.

Chemical shift transfer: an effective strategy for protein NMR assignment with ARTINA.

Chemical shift transfer (CST) is a well-established technique in NMR spectroscopy that utilizes the chemical shift assignment of one protein (source) to identify chemical shifts of another (target). Given similarity between source and target systems (e.g., using homologs), CST allows the chemical shifts of the target system to be assigned using a limited amount of experimental data. In this study, we propose a deep-learning based workflow, ARTINA-CST, that automates this procedure, allowing CST to be carried out within minutes or hours of computational time and strictly without any human supervision. We characterize the efficacy of our method using three distinct synthetic and experimental datasets, demonstrating its effectiveness and robustness even when substantial differences exist between the source and target proteins. With its potential applications spanning a wide range of NMR projects, including drug discovery and protein interaction studies, ARTINA-CST is anticipated to be a valuable method that facilitates research in the field.

Fibrils Emerging from Droplets: Molecular Guiding Principles behind Phase Transitions of a Short Peptide-Based Condensate Studied by Solid-State NMR.

Biochemical reactions occurring in highly crowded cellular environments require different means of control to ensure productivity and specificity. Compartmentalization of reagents by liquid-liquid phase separation is one of these means. However, extremely high local protein concentrations of up to 400 mg/ml can result in pathological aggregation into fibrillar amyloid structures, a phenomenon that has been linked to various neurodegenerative diseases. Despite its relevance, the process of liquid-to-solid transition inside condensates is still not well understood at the molecular level. We thus herein use small peptide derivatives that can undergo both liquid-liquid and subsequent liquid-to-solid phase transition as model systems to study both processes. Using solid-state nuclear magnetic resonance (NMR) and transmission electron microscopy (TEM), we compare the structure of condensed states of leucine, tryptophan and phenylalanine containing derivatives, distinguishing between liquid-like condensates, amorphous aggregates and fibrils, respectively. A structural model for the fibrils formed by the phenylalanine derivative was obtained by an NMR-based structure calculation. The fibrils are stabilised by hydrogen bonds and side-chain π-π interactions, which are likely much less pronounced or absent in the liquid and amorphous state. Such noncovalent interactions are equally important for the liquid-to-solid transition of proteins, particularly those related to neurodegenerative diseases.

NMRtist: an online platform for automated biomolecular NMR spectra analysis.

SUMMARY: We present NMRtist, an online platform that combines deep learning, large-scale optimization and cloud computing to automate protein NMR spectra analysis. Our website provides virtual storage for NMR spectra deposition together with a set of applications designed for automated peak picking, chemical shift assignment and protein structure determination. The system can be used by non-experts and allows protein assignments and structures to be determined within hours after the measurements, strictly without any human intervention. AVAILABILITY AND IMPLEMENTATION: NMRtist is freely available to non-commercial users at https://nmrtist.org. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Protein structure determination in human cells by in-cell NMR and a reporter system to optimize protein delivery or transexpression.

Most experimental methods for structural biology proceed in vitro and therefore the contribution of the intracellular environment on protein structure and dynamics is absent. Studying proteins at atomic resolution in living mammalian cells has been elusive due to the lack of methodologies. In-cell nuclear magnetic resonance spectroscopy (in-cell NMR) is an emerging technique with the power to do so. Here, we improved current methods of in-cell NMR by the development of a reporter system that allows monitoring the delivery of exogenous proteins into mammalian cells, a process that we called here "transexpression". The reporter system was used to develop an efficient protocol for in-cell NMR which enables spectral acquisition with higher quality for both disordered and folded proteins. With this method, the 3D atomic resolution structure of the model protein GB1 in human cells was determined with a backbone root-mean-square deviation (RMSD) of 1.1 Å.

Atomic resolution protein allostery from the multi-state structure of a PDZ domain.

Recent methodological advances in solution NMR allow the determination of multi-state protein structures and provide insights into structurally and dynamically correlated protein sites at atomic resolution. This is demonstrated in the present work for the well-studied PDZ2 domain of protein human tyrosine phosphatase 1E for which protein allostery had been predicted. Two-state protein structures were calculated for both the free form and in complex with the RA-GEF2 peptide using the exact nuclear Overhauser effect (eNOE) method. In the apo protein, an allosteric conformational selection step comprising almost 60% of the domain was detected with an "open" ligand welcoming state and a "closed" state that obstructs the binding site by changing the distance between the ß-sheet 2, α-helix 2, and sidechains of residues Lys38 and Lys72. The observed induced fit-type apo-holo structural rearrangements are in line with the previously published evolution-based analysis covering ~25% of the domain with only a partial overlap with the protein allostery of the open form. These presented structural studies highlight the presence of a dedicated highly optimized and complex dynamic interplay of the PDZ2 domain owed by the structure-dynamics landscape.

Rapid protein assignments and structures from raw NMR spectra with the deep learning technique ARTINA.

Nuclear Magnetic Resonance (NMR) spectroscopy is a major technique in structural biology with over 11,800 protein structures deposited in the Protein Data Bank. NMR can elucidate structures and dynamics of small and medium size proteins in solution, living cells, and solids, but has been limited by the tedious data analysis process. It typically requires weeks or months of manual work of a trained expert to turn NMR measurements into a protein structure. Automation of this process is an open problem, formulated in the field over 30 years ago. We present a solution to this challenge that enables the completely automated analysis of protein NMR data within hours after completing the measurements. Using only NMR spectra and the protein sequence as input, our machine learning-based method, ARTINA, delivers signal positions, resonance assignments, and structures strictly without human intervention. Tested on a 100-protein benchmark comprising 1329 multidimensional NMR spectra, ARTINA demonstrated its ability to solve structures with 1.44 Å median RMSD to the PDB reference and to identify 91.36% correct NMR resonance assignments. ARTINA can be used by non-experts, reducing the effort for a protein assignment or structure determination by NMR essentially to the preparation of the sample and the spectra measurements.

1H, 13C, and 15N resonance assignments and solution structures of the KH domain of human ribosome binding factor A, mtRbfA, involved in mitochondrial ribosome biogenesis.

Ribosome biogenesis is a complicated, multistage process coordinated by ribosome assembly factors. Ribosome binding factor A (RbfA) is a bacterial one, which possesses a single structural type-II KH domain. By this domain, RbfA binds to a 16S rRNA precursor in small ribosomal subunits to promote its 5'-end processing. The human RbfA homolog, mtRbfA, binds to 12S rRNAs in the mitoribosomal small subunits and promotes its critical maturation process, the dimethylation of two highly conserved consecutive adenines, which differs from that of RbfA. However, the structural basis of the mtRbfA-mediated maturation process is poorly understood. Herein, we report the 1H, 15N, and 13C resonance assignments of the KH domain of mtRbfA and its solution structure. The mtRbfA domain adopts essentially the same α1-ß1-ß2-α2(kinked)-ß3 topology as the type-II KH domain. Comparison with the RbfA counterpart showed structural differences in specific regions that function as a putative RNA-binding site. Particularly, the α2 helix of mtRbfA forms a single helix with a moderate kink at the Ser-Ala-Ala sequence, whereas the corresponding α2 helix of RbfA is interrupted by a distinct kink at the Ala-x-Gly sequence, characteristic of bacterial RbfA proteins, to adopt an α2-kink-α3 conformation. Additionally, the region linking α1 and ß1 differs considerably in the sequence and structure between RbfA and mtRbfA. These findings suggest some variations of the RNA-binding mode between them and provide a structural basis for mtRbfA function in mitoribosome biogenesis.

A B-factor for NOEs?

Nuclear Overhauser effects (NOEs) are influenced by motion. Here, we derive exact, analytical results for a model of isotropic, harmonic fluctuations of atom positions that corresponds to the one underlying crystallographic B-factors. The model includes steric repulsion and yields closed-form expressions for the expected value of general invertible functions of the distance between two atoms, with the special case r-6 for NOEs. We discuss the implications for the definition of an NOE-based B-factor in solution NMR.

Optimization and validation of multi-state NMR protein structures using structural correlations.

Recent advances in the field of protein structure determination using liquid-state NMR enable the elucidation of multi-state protein conformations that can provide insight into correlated and non-correlated protein dynamics at atomic resolution. So far, NMR-derived multi-state structures were typically evaluated by means of visual inspection of structure superpositions, target function values that quantify the violation of experimented restraints and root-mean-square deviations that quantify similarity between conformers. As an alternative or complementary approach, we present here the use of a recently introduced structural correlation measure, PDBcor, that quantifies the clustering of protein states as an additional measure for multi-state protein structure analysis. It can be used for various assays including the validation of experimental distance restraints, optimization of the number of protein states, estimation of protein state populations, identification of key distance restraints, NOE network analysis and semiquantitative analysis of the protein correlation network. We present applications for the final quality analysis stages of typical multi-state protein structure calculations.

Atomic-resolution chemical characterization of (2x)72-kDa tryptophan synthase via four- and five-dimensional 1H-detected solid-state NMR.

NMR chemical shifts provide detailed information on the chemical properties of molecules, thereby complementing structural data from techniques like X-ray crystallography and electron microscopy. Detailed analysis of protein NMR data, however, often hinges on comprehensive, site-specific assignment of backbone resonances, which becomes a bottleneck for molecular weights beyond 40 to 45 kDa. Here, we show that assignments for the (2x)72-kDa protein tryptophan synthase (665 amino acids per asymmetric unit) can be achieved via higher-dimensional, proton-detected, solid-state NMR using a single, 1-mg, uniformly labeled, microcrystalline sample. This framework grants access to atom-specific characterization of chemical properties and relaxation for the backbone and side chains, including those residues important for the catalytic turnover. Combined with first-principles calculations, the chemical shifts in the ß-subunit active site suggest a connection between active-site chemistry, the electrostatic environment, and catalytically important dynamics of the portal to the ß-subunit from solution.

1H, 13C and 15N resonance assignments and solution structures of the two RRM domains of Matrin-3.

Matrin-3 is a multifunctional protein that binds to both DNA and RNA. Its DNA-binding activity is linked to the formation of the nuclear matrix and transcriptional regulation, while its RNA-binding activity is linked to mRNA metabolism including splicing, transport, stabilization, and degradation. Correspondingly, Matrin-3 has two zinc finger domains for DNA binding and two consecutive RNA recognition motif (RRM) domains for RNA binding. Matrin-3 has been reported to cause amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) when its disordered region contains pathogenic mutations. Simultaneously, it has been shown that the RNA-binding activity of Matrin-3 mediated by its RRM domains, affects the formation of insoluble cytoplasmic granules, which are related to the pathogenic mechanism of ALS/FTD. Thus, the effect of the RRM domains on the phase separation of condensed protein/RNA mixtures has to be clarified for a comprehensive understanding of ALS/FTD. Here, we report the 1H, 15N, and 13C resonance assignments of the two RNA binding domains and their solution structures. The resonance assignments and the solution structures obtained in this work will contribute to the elucidation of the molecular basis of Matrin-3 in the pathogenic mechanism of ALS and/or FTD.

PDBcor: An automated correlation extraction calculator for multi-state protein structures.

Allostery and correlated motion are key elements linking protein dynamics with the mechanisms of action of proteins. Here, we present PDBCor, an automated and unbiased method for the detection and analysis of correlated motions from experimental multi-state protein structures. It uses torsion angle and distance statistics and does not require any structure superposition. Clustering of protein conformers allows us to extract correlations in the form of mutual information based on information theory. With PDBcor, we elucidated correlated motion in the WW domain of PIN1, the protein GB3, and the enzyme cyclophilin, in line with reported findings. Correlations extracted with PDBcor can be utilized in subsequent assays including nuclear magnetic resonance (NMR) multi-state structure optimization and validation. As a guide for the interpretation of PDBcor results, we provide a series of protein structure ensembles that exhibit different levels of correlation, including non-correlated, locally correlated, and globally correlated ensembles.