Determination of the geometric structure of the metal site in a blue copper protein by paramagnetic NMR

Biomolecular NMR spectroscopy in the era of artificial intelligence

Biomolecular nuclear magnetic resonance (NMR) spectroscopy and artificial intelligence (AI) have a burgeoning synergy. Deep learning-based structural predictors have forever changed structural biology, yet these tools currently face limitations in accurately characterizing protein dynamics, allostery, and conformational heterogeneity. We begin by highlighting the unique abilities of biomolecular NMR spectroscopy to complement AI-based structural predictions toward addressing these knowledge gaps. We then highlight the direct integration of deep learning approaches into biomolecular NMR methods. AI-based tools can dramatically improve the acquisition and analysis of NMR spectra, enhancing the accuracy and reliability of NMR measurements, thus streamlining experimental processes. Additionally, deep learning enables the development of novel types of NMR experiments that were previously unattainable, expanding the scope and potential of biomolecular NMR spectroscopy. Ultimately, a combination of AI and NMR promises to further revolutionize structural biology on several levels, advance our understanding of complex biomolecular systems, and accelerate drug discovery efforts.

Reparameterization of Non-Bonded Parameters for Copper Ions in Plastocyanin: An Adaptive Force Matching Study

Molecular mechanics rely on existing experimental and theoretical inputs to confidently calculate the trajectories of molecular systems. These calculations, however, are often hindered by missing force field parameters. A notable subject of this problem is metal centers of proteins. This study parameterized, through an adaptive force matching (AFM) workflow, the copper cofactor of plastocyanin in its two oxidation states. New 12-6 Lennard-Jones (LJ) parameters and atomic partial charges were generated to complete the non-bonded description of the copper site. Our models show uniform distorted tetrahedral structures for reduced plastocyanin, Cu(I), and oxidized plastocyanin, Cu(II). These structures align with the QM/MM MD results and existing crystallography studies. TD-DFT calculations, meanwhile, showed that conformations with elongated axial Cu-SMet and shortened equatorial Cu-SCys bonds retain the experimental UV-Vis profile of blue copper (BC) proteins, thus signifying the importance of Cu-S interactions on BC proteins' unique spectroscopic properties.

Exploring the use of cobalt(II) dipolar shifts in refining the structure of a zinc finger peptide

The uses of dipolar shifts due to cobalt(II) substituted for zinc(II) in a consensus zinc finger peptide for refining the NMR-determined structure were examined. Substantial differences between the calculated and observed chemical shift differences between the cobalt(II) and zinc(II) complexes were observed when these dipolar shifts were not used as constraints in the structure refinement. However, inclusion of these constraints resulted in excellent agreement with minor adjustments in the structure and a slight improvement in the precision of the structure determination. Other calculations revealed that the dipolar shifts were not adequate to determine the overall folded structure by themselves, but were useful in increasing the accuracy and precision of a structure determined based only on nuclear Overhauser effects constraints involving only backbone atoms.

Towards autonomous analysis of chemical exchange saturation transfer experiments using deep neural networks

Macromolecules often exchange between functional states on timescales that can be accessed with NMR spectroscopy and many NMR tools have been developed to characterise the kinetics and thermodynamics of the exchange processes, as well as the structure of the conformers that are involved. However, analysis of the NMR data that report on exchanging macromolecules often hinges on complex least-squares fitting procedures as well as human experience and intuition, which, in some cases, limits the widespread use of the methods. The applications of deep neural networks (DNNs) and artificial intelligence have increased significantly in the sciences, and recently, specifically, within the field of biomolecular NMR, where DNNs are now available for tasks such as the reconstruction of sparsely sampled spectra, peak picking, and virtual decoupling. Here we present a DNN for the analysis of chemical exchange saturation transfer (CEST) data reporting on two- or three-site chemical exchange involving sparse state lifetimes of between approximately 3-60 ms, the range most frequently observed via experiment. The work presented here focuses on the ¹H CEST class of methods that are further complicated, in relation to applications to other nuclei, by anti-phase features. The developed DNNs accurately predict the chemical shifts of nuclei in the exchanging species directly from anti-phase ¹H^N CEST profiles, along with an uncertainty associated with the predictions. The performance of the DNN was quantitatively assessed using both synthetic and experimental anti-phase CEST profiles. The assessments show that the DNN accurately determines chemical shifts and their associated uncertainties. The DNNs developed here do not contain any parameters for the end-user to adjust and the method therefore allows for autonomous analysis of complex NMR data that report on conformational exchange.

Measuring transverse relaxation in highly paramagnetic systems

The enhancement of nuclear relaxation rates due to the interaction with a paramagnetic center (known as Paramagnetic Relaxation Enhancement) is a powerful source of structural and dynamics information, widely used in structural biology. However, many signals affected by the hyperfine interaction relax faster than the evolution periods of common NMR experiments and therefore they are broadened beyond detection. This gives rise to a so-called blind sphere around the paramagnetic center, which is a major limitation in the use of PREs. Reducing the blind sphere is extremely important in paramagnetic metalloproteins. The identification, characterization, and proper structural restraining of the first coordination sphere of the metal ion(s) and its immediate neighboring regions is key to understand their biological function. The novel HSQC scheme we propose here, that we termed R2-weighted, HSQC-AP, achieves this aim by detecting signals that escaped detection in a conventional HSQC experiment and provides fully reliable R2 values in the range of 1H R2 rates ca. 50-400 s-1. Independently on the type of paramagnetic center and on the size of the molecule, this experiment decreases the radius of the blind sphere and increases the number of detectable PREs. Here, we report the validation of this approach for the case of PioC, a small protein containing a high potential 4Fe-4S cluster in the reduced [Fe4S4]2+ form. The blind sphere was contracted to a minimal extent, enabling the measurement of R2 rates for the cluster coordinating residues.

Paramagnetic NMR in solution and the solid state

The field of paramagnetic NMR has expanded considerably in recent years. This review addresses both the theoretical description of paramagnetic NMR, and the way in which it is currently practised. We provide a review of the theory of the NMR parameters of systems in both solution and the solid state. Here we unify the different languages used by the NMR, EPR, quantum chemistry/DFT, and magnetism communities to provide a comprehensive and coherent theoretical description. We cover the theory of the paramagnetic shift and shift anisotropy in solution both in the traditional formalism in terms of the magnetic susceptibility tensor, and using a more modern formalism employing the relevant EPR parameters, such as are used in first-principles calculations. In addition we examine the theory first in the simple non-relativistic picture, and then in the presence of spin-orbit coupling. These ideas are then extended to a description of the paramagnetic shift in periodic solids, where it is necessary to include the bulk magnetic properties, such as magnetic ordering at low temperatures. The description of the paramagnetic shift is completed by describing the current understanding of such shifts due to lanthanide and actinide ions. We then examine the paramagnetic relaxation enhancement, using a simple model employing a phenomenological picture of the electronic relaxation, and again using a more complex state-of-the-art theory which incorporates electronic relaxation explicitly. An additional important consideration in the solid state is the impact of bulk magnetic susceptibility effects on the form of the spectrum, where we include some ideas from the field of classical electrodynamics. We then continue by describing in detail the solution and solid-state NMR methods that have been deployed in the study of paramagnetic systems in chemistry, biology, and the materials sciences. Finally we describe a number of case studies in paramagnetic NMR that have been specifically chosen to highlight how the theory in part one, and the methods in part two, can be used in practice. The systems chosen include small organometallic complexes in solution, solid battery electrode materials, metalloproteins in both solution and the solid state, systems containing lanthanide ions, and multi-component materials used in pharmaceutical controlled-release formulations that have been doped with paramagnetic species to measure the component domain sizes.

Exchangeable deuterons introduce artifacts in amide 15N CEST experiments used to study protein conformational exchange

Protein molecules sample different conformations in solution and characterizing these conformations is crucial to understanding protein function. ¹⁵N CEST experiments are now routinely used to study slow conformational exchange of protein molecules between a 'visible' major state and 'invisible' minor states. These experiments have also been adapted to measure the solvent exchange rates of amide protons by exploiting the one bond deuterium isotope effect on the amide ¹⁵N chemical shifts. However at moderately high temperatures (~ 50 °C) that are sometimes required to populate protein minor conformers to levels (~ 1%) that can be detected by CEST experiments solvent H/D exchange can lead to 'dips' in low B₁¹⁵N CEST profiles that can be wrongly assigned to the conformational exchange process being characterized. This is demonstrated in the case of ~ 18 kDa T4 Lysozyme (T4L) at 50 °C and the ~ 11 kDa E. coli hibernation promoting factor (HPF) at 52 °C. This problem is trivially solved by eliminating the exchangeable deuterons in the solvent by using either an external D₂O lock or by using a small amount (~ 1-3%) of a molecule like d6-DMSO that does not contain exchangeable deuterons to lock the spectrometer.

Multiple frequency saturation pulses reduce CEST acquisition time for quantifying conformational exchange in biomolecules

Exchange between conformational states is required for biomolecular catalysis, allostery, and folding. A variety of NMR experiments have been developed to quantify motional regimes ranging from nanoseconds to seconds. In this work, we describe an approach to speed up the acquisition of chemical exchange saturation transfer (CEST) experiments that are commonly used to probe millisecond to second conformational exchange in proteins and nucleic acids. The standard approach is to obtain CEST datasets through the acquisition of a series of 2D correlation spectra where each experiment utilizes a single saturation frequency to ¹H, ¹⁵N or ¹³C. These pseudo 3D datasets are time consuming to collect and are further lengthened by reduced signal to noise stemming from the long saturation pulse. In this article, we show how usage of a multiple frequency saturation pulse (i.e., MF-CEST) changes the nature of data collection from series to parallel, and thus decreases the total acquisition time by an integer factor corresponding to the number of frequencies in the pulse. We demonstrate the applicability of MF-CEST on a Src homology 2 (SH2) domain from phospholipase Cγ and the secondary active transport protein EmrE as model systems by collecting ¹³C methyl and ¹⁵N backbone datasets. MF-CEST can also be extended to additional sites within proteins and nucleic acids. The only notable drawback of MF-CEST as applied to backbone ¹⁵N experiments occurs when a large chemical shift difference between the major and minor populations is present (typically greater than ~ 8 ppm). In these cases, ambiguity may arise between the chemical shift of the minor population and the multiple frequency saturation pulse. Nevertheless, this drawback does not occur for methyl group MF-CEST experiments or in cases where somewhat smaller chemical shift differences occur are present.

Probing conformational dynamics in biomolecules via chemical exchange saturation transfer: a primer

Although Chemical Exchange Saturation Transfer (CEST) type NMR experiments have been used to study chemical exchange processes in molecules since the early 1960s, there has been renewed interest in the past several years in using this approach to study biomolecular conformational dynamics. The methodology is particularly powerful for the study of sparsely populated, transiently formed conformers that are recalcitrant to investigation using traditional biophysical tools, and it is complementary to relaxation dispersion and magnetization transfer experiments that have traditionally been used to study chemical exchange processes. Here we discuss the concepts behind the CEST experiment, focusing on practical aspects as well, we review some of the pulse sequences that have been developed to characterize protein and RNA conformational dynamics, and we discuss a number of examples where the CEST methodology has provided important insights into the role of dynamics in biomolecular function.

Towards a protocol for solution structure determination of copper(II) proteins: the case of Cu(II)Zn(II) superoxide dismutase

We have developed an optimized protocol to solve the solution structure of copper(II) proteins. After assignment, proton-proton NOEs are used for the shell where 1H spectra are conveniently observed. In a shell closer to the metal ion, 13C NMR spectra with band-selective homonuclear decoupling provide the assignment of all nuclei except for those of the metal ligands. A convenient method for the measurement of 13C longitudinal-relaxation rates (R1) of carbonyls and carboxylate moieties is proposed. 1H NOEs and 1H and 13C R1 data are sufficient to produce a good/reasonable solution structure, as demonstrated for a monomeric species of superoxide dismutase, a 153-residue protein.

Computational approaches to predict binding interactions between mammalian tyrosinases and (S)-(+)-decursin and its analogues as potent inhibitors

Using computational methods, we elucidated the binding modes of (S)-(+)-decursin with mouse and human tyrosinases, and thereby provided insight to new tyrosinase inhibitors.

A computational study of the effects of (13) C-(13) C scalar couplings on (13) C CEST NMR spectra: towards studies on a uniformly (13) C-labeled protein

Read the label: The NMR CEST experiment can be used to reconstruct spectra of sparsely populated, transiently formed protein conformers so long as they exchange with a highly populated ground state with rates of 20-300 s(-1) . Here we establish that accurate (13) C chemical shifts of side-chain carbon nuclei can be obtained from uniformly (13) C-labeled samples, without interference from the coupled (13) C spin network.

Jahn-Teller dynamics in a series of high-symmetry Co(II) chelates determine paramagnetic relaxation enhancements

NMR paramagnetic relaxation enhancements (PREs) of a series of structurally characterized, trigonal bis-trispyrazolylborate (Tp) chelates of high-spin Co(II), spanning 100-850 MHz in field, are reported. Prior knowledge of the metal-nucleus distances allows numerical extraction of position-dependent electron spin relaxation rates (τ(c)(-1)) from direct measurement of the individual PREs of the four symmetry distinct protons in Co(Tp)(2), using available closed-form expressions. The data for this electronically complex system where spin-orbit coupling defines the ground state electronic structure are analyzed in terms of the Solomon-Bloembergen-Morgan (SBM) relations, as well as available zero-field splitting limit theories. A simple angular correction is shown to be sufficient to reconcile the individual τ(c)(T) data for the four classes of protons. The data identify a previously unrecognized dynamic Jahn-Teller effect in these historically important complexes, with a barrier of ~230 cm(-1), pointing to a level of dynamics in trispyrazolylborate chemistry that has not been described before, and further show that it is the Jahn-Teller that is responsible for the PREs in fluid solution. A field-dependent component is also identified for the two protons nearest g(//), which is suggested to arise due to Zeeman mixing of excited state character into the ground level.

Accurate structure and dynamics of the metal-site of paramagnetic metalloproteins from NMR parameters using natural bond orbitals

A natural bond orbital (NBO) analysis of unpaired electron spin density in metalloproteins is presented, which allows a fast and robust calculation of paramagnetic NMR parameters. Approximately 90% of the unpaired electron spin density occupies metal–ligand NBOs, allowing the majority of the density to be modeled by only a few NBOs that reflect the chemical bonding environment. We show that the paramagnetic relaxation rate of protons can be calculated accurately using only the metal–ligand NBOs and that these rates are in good agreement with corresponding rates measured experimentally. This holds, in particular, for protons of ligand residues where the point-dipole approximation breaks down. To describe the paramagnetic relaxation of heavy nuclei, also the electron spin density in the local orbitals must be taken into account. Geometric distance restraints for ¹⁵N can be derived from the paramagnetic relaxation enhancement and the Fermi contact shift when local NBOs are included in the analysis. Thus, the NBO approach allows us to include experimental paramagnetic NMR parameters of ¹⁵N nuclei as restraints in a structure optimization protocol. We performed a molecular dynamics simulation and structure determination of oxidized rubredoxin using the experimentally obtained paramagnetic NMR parameters of ¹⁵N. The corresponding structures obtained are in good agreement with the crystal structure of rubredoxin. Thus, the NBO approach allows an accurate description of the geometric structure and the dynamics of metalloproteins, when NMR parameters are available of nuclei in the immediate vicinity of the metal-site.

Hyperfine-shifted (13)C and (15)N NMR signals from Clostridium pasteurianum rubredoxin: extensive assignments and quantum chemical verification

Stable isotope-labeling methods, coupled with novel techniques for detecting fast-relaxing NMR signals, now permit detailed investigations of paramagnetic centers of metalloproteins. We have utilized these advances to carry out comprehensive assignments of the hyperfine-shifted ¹³C and ¹⁵N signals of the rubredoxin from Clostridium pasteurianum (CpRd) in both its oxidized and reduced states. We used residue-specific labeling (by chemical synthesis) and residue-type-selective labeling (by biosynthesis) to assign signals detected by one-dimensional ¹⁵N NMR spectroscopy, to nitrogen atoms near the iron center. We refined and extended these ¹⁵N assignments to the adjacent carbonyl carbons by means of one-dimensional ¹³C[¹⁵N] decoupling difference experiments. We collected paramagnetic-optimized SuperWEFT ¹³C[¹³C] constant time COSY (SW-CT-COSY) data to complete the assignment of ¹³C signals of reduced CpRd. By following these ¹³C signals as the protein was gradually oxidized, we transferred these assignments to carbons in the oxidized state. We have compared these assignments with hyperfine chemical shifts calculated from available X-ray structures of CpRd in its oxidized and reduced forms. The results allow the evaluation of the X-ray structural models as representative of the solution structure of the protein, and they provide a framework for future investigation of the active site of this protein. The methods developed here should be applicable to other proteins that contain a paramagnetic center with high spin and slow electron exchange.

NMR hyperfine shifts in blue copper proteins: a quantum chemical investigation

We present the results of the first quantum chemical investigations of 1H NMR hyperfine shifts in the blue copper proteins (BCPs): amicyanin, azurin, pseudoazurin, plastocyanin, stellacyanin, and rusticyanin. We find that very large structural models that incorporate extensive hydrogen bond networks, as well as geometry optimization, are required to reproduce the experimental NMR hyperfine shift results, the best theory vs experiment predictions having R2 = 0.94, a slope = 1.01, and a SD = 40.5 ppm (or approximately 4.7% of the overall approximately 860 ppm shift range). We also find interesting correlations between the hyperfine shifts and the bond and ring critical point properties computed using atoms-in-molecules theory, in addition to finding that hyperfine shifts can be well-predicted by using an empirical model, based on the geometry-optimized structures, which in the future should be of use in structure refinement.