Increasing the accuracy of exchange parameters reporting on slow dynamics by performing CEST experiments with 'high' B1 fields.

Over the last decade chemical exchange saturation transfer (CEST) NMR methods have emerged as powerful tools to characterize biomolecular conformational dynamics occurring between a visible major state and 'invisible' minor states. The ability of the CEST experiment to detect these minor states, and provide precise exchange parameters, hinges on using appropriate B1 field strengths during the saturation period. Typically, a pair of B1 fields with ω1 (=2πB1) values around the exchange rate kex are chosen. Here we show that the transverse relaxation rate of the minor state resonance (R2,B) also plays a crucial role in determining the B1 fields that lead to the most informative datasets. Using [Formula: see text]  ≥ kex, to guide the choice of B1, instead of kex, leads to data wherefrom substantially more accurate exchange parameters can be derived. The need for higher B1 fields, guided by K, is demonstrated by studying the conformational exchange in two mutants of the 71 residue FF domain with kex âˆ¼ 11 s-1 and âˆ¼ 72 s-1, respectively. In both cases analysis of CEST datasets recorded using B1 field values guided by kex lead to imprecise exchange parameters, whereas using B1 values guided by K resulted in precise site-specific exchange parameters. The conclusions presented here will be valuable while using CEST to study slow processes at sites with large intrinsic relaxation rates, including carbonyl sites in small to medium sized proteins, amide 15N sites in large proteins and when the minor state dips are broadened due to exchange among the minor states.

Solution-state methyl NMR spectroscopy of large non-deuterated proteins enabled by deep neural networks.

Methyl-TROSY nuclear magnetic resonance (NMR) spectroscopy is a powerful technique for characterising large biomolecules in solution. However, preparing samples for these experiments is demanding and entails deuteration, limiting its use. Here we demonstrate that NMR spectra recorded on protonated, uniformly 13C labelled samples can be processed using deep neural networks to yield spectra that are of similar quality to typical deuterated methyl-TROSY spectra, potentially providing information for proteins that cannot be produced in bacterial systems. We validate the methodology experimentally on three proteins with molecular weights in the range 42-360 kDa. We further demonstrate the applicability of our methodology to 3D NOESY spectra of Escherichia coli Malate Synthase G (81 kDa), where observed NOE cross-peaks are in good agreement with the available structure. The method represents an advance in the field of using deep learning to analyse complex magnetic resonance data and could have an impact on the study of large biomolecules in years to come.

Picosecond Dynamics of a Small Molecule in Its Bound State with an Intrinsically Disordered Protein.

Intrinsically disordered proteins (IDPs) are highly dynamic biomolecules that rapidly interconvert among many structural conformations. These dynamic biomolecules are involved in cancers, neurodegeneration, cardiovascular illnesses, and viral infections. Despite their enormous therapeutic potential, IDPs have generally been considered undruggable because of their lack of classical long-lived binding pockets for small molecules. Currently, only a few instances are known where small molecules have been observed to interact with IDPs, and this situation is further exacerbated by the limited sensitivity of experimental techniques to detect such binding events. Here, using experimental nuclear magnetic resonance (NMR) spectroscopy 19F transverse spin-relaxation measurements, we discovered that a small molecule, 5-fluoroindole, interacts with the disordered domains of non-structural protein 5A from hepatitis C virus with a Kd of 260 ± 110 µM. Our analysis also allowed us to determine the rotational correlation times (τc) for the free and bound states of 5-fluoroindole. In the free state, we observed a rotational correlation time of 27.0 ± 1.3 ps, whereas in the bound state, τc only increased to 46 ± 10 ps. Our findings imply that it is possible for small molecules to engage with IDPs in exceptionally dynamic ways, in sharp contrast to the rigid binding modes typically exhibited when small molecules bind to well-defined binding pockets within structured proteins.

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.

Intrinsic structural dynamics dictate enzymatic activity and inhibition.

Enzymes are known to sample various conformations, many of which are critical for their biological function. However, structural characterizations of enzymes predominantly focus on the most populated conformation. As a result, single-point mutations often produce structures that are similar or essentially identical to those of the wild-type enzyme despite large changes in enzymatic activity. Here, we show for mutants of a histone deacetylase enzyme (HDAC8) that reduced enzymatic activities, reduced inhibitor affinities, and reduced residence times are all captured by the rate constants between intrinsically sampled conformations that, in turn, can be obtained independently by solution NMR spectroscopy. Thus, for the HDAC8 enzyme, the dynamic sampling of conformations dictates both enzymatic activity and inhibitor potency. Our analysis also dissects the functional role of the conformations sampled, where specific conformations distinct from those in available structures are responsible for substrate and inhibitor binding, catalysis, and product dissociation. Precise structures alone often do not adequately explain the effect of missense mutations on enzymatic activity and drug potency. Our findings not only assign functional roles to several conformational states of HDAC8 but they also underscore the paramount role of dynamics, which will have general implications for characterizing missense mutations and designing inhibitors.

Less is more: A simple methyl-TROSY based pulse scheme offers improved sensitivity in applications to high molecular weight complexes.

The HMQC pulse sequence and variants thereof have been exploited in studies of high molecular weight protein complexes, taking advantage of the fact that fast and slow relaxing magnetization components are sequestered along two distinct magnetization transfer pathways. Despite the simplicity of the HMQC scheme an even shorter version can be designed, based on elimination of the terminal refocusing period, as a further means of increasing signal. Here we present such an experiment, and show that significant sensitivity gains, in some cases by factors of two or more, are realized in studies of proteins varying in molecular masses from 100 kDa to 1 MDa.

Optimizing frequency sampling in CEST experiments.

For the past decade chemical exchange saturation transfer (CEST) experiments have been successfully applied to study exchange processes in biomolecules involving sparsely populated, transiently formed conformers. Initial implementations focused on extensive sampling of the CEST frequency domain, requiring significant measurement times. Here we show that the lengthy sampling schemes often used are not optimal and that reduced frequency sampling schedules can be developed without a priori knowledge of the exchange parameters, that only depend on the chosen B1 field, and, to a lesser extent, on the intrinsic transverse relaxation rates of ground state spins. The reduced sampling approach described here can be used synergistically with other methods for reducing measurement times such as those that excite multiple frequencies in the CEST dimension simultaneously, or make use of non-uniform sampling of indirectly detected time domains, to further decrease measurement times. The proposed approach is validated by analysis of simulated and experimental datasets.

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 1HN 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.

Structural mass spectrometry decodes domain interaction and dynamics of the full-length Human Histone Deacetylase 2.

Human Histone Deacetylase 2 (HDAC2) belongs to a conserved enzyme superfamily that regulates deacetylation inside cells. HDAC2 is a drug target as it is known to be upregulated in cancers and neurodegenerative disorders. It consists of globular deacetylase and C-terminus intrinsically-disordered domains [1-3]. To date, there is no full-length structure of HDAC2 available due to the high intrinsic flexibility of its C-terminal domain. The intrinsically-disordered domain, however, is known to be important for the enzymatic function of HDAC2 [1, 4]. Here we combine several structural Mass Spectrometry (MS) methodologies such as denaturing, native, ion mobility and chemical crosslinking, alongside biochemical assays and molecular modelling to study the structure and dynamics of the full-length HDAC2 for the first time. We show that MS can easily dissect heterogeneity inherent within the protein sample and at the same time probe the structural arrangement of the different conformers present. Activity assays combined with data from MS and molecular modelling suggest how the structural dynamics of the C-terminal domain, and its interactions with the catalytic domain, regulate the activity of this enzyme.

Post-translational insertion of boron in proteins to probe and modulate function.

Boron is absent in proteins, yet is a micronutrient. It possesses unique bonding that could expand biological function including modes of Lewis acidity not available to typical elements of life. Here we show that post-translational Cß-Bγ bond formation provides mild, direct, site-selective access to the minimally sized residue boronoalanine (Bal) in proteins. Precise anchoring of boron within complex biomolecular systems allows dative bond-mediated, site-dependent protein Lewis acid-base-pairing (LABP) by Bal. Dynamic protein-LABP creates tunable inter- and intramolecular ligand-host interactions, while reactive protein-LABP reveals reactively accessible sites through migratory boron-to-oxygen Cß-Oγ covalent bond formation. These modes of dative bonding can also generate de novo function, such as control of thermo- and proteolytic stability in a target protein, or observation of transient structural features via chemical exchange. These results indicate that controlled insertion of boron facilitates stability modulation, structure determination, de novo binding activities and redox-responsive 'mutation'.

Virtual Homonuclear Decoupling in Direct Detection Nuclear Magnetic Resonance Experiments Using Deep Neural Networks.

Nuclear magnetic resonance (NMR) experiments are frequently complicated by the presence of homonuclear scalar couplings. For the growing body of biomolecular 13C-detected NMR methods, one-bond 13C-13C couplings significantly reduce sensitivity and resolution. The solution to this problem has typically been to perform virtual decoupling by recording multiple spectra and taking linear combinations. Here, we propose an alternative method of virtual decoupling using deep neural networks, which only requires a single spectrum and gives a significant boost in resolution while reducing the minimum effective phase cycles of the experiments by at least a factor of 2. We successfully apply this methodology to virtually decouple in-phase CON (13CO-15N) protein NMR spectra, 13C-13C correlation spectra of protein side chains, and 13Cα-detected protein 13Cα-13CO spectra where two large homonuclear couplings are present. The deep neural network approach effectively decouples spectra with a high degree of flexibility, including in cases where existing methods fail, and facilitates the use of simpler pulse sequences.

Aromatic side-chain flips orchestrate the conformational sampling of functional loops in human histone deacetylase 8.

Human histone deacetylase 8 (HDAC8) is a key hydrolase in gene regulation and an important drug-target. High-resolution structures of HDAC8 in complex with substrates or inhibitors are available, which have provided insights into the bound state of HDAC8 and its function. Here, using long all-atom unbiased molecular dynamics simulations and Markov state modelling, we show a strong correlation between the conformation of aromatic side chains near the active site and opening and closing of the surrounding functional loops of HDAC8. We also investigated two mutants known to allosterically downregulate the enzymatic activity of HDAC8. Based on experimental data, we hypothesise that I19S-HDAC8 is unable to release the product, whereas both product release and substrate binding are impaired in the S39E-HDAC8 mutant. The presented results deliver detailed insights into the functional dynamics of HDAC8 and provide a mechanism for the substantial downregulation caused by allosteric mutations, including a disease causing one.

Methodological advancements for characterising protein side chains by NMR spectroscopy.

The surface of proteins is covered by side chains of polar amino acids that are imperative for modulating protein functionality through the formation of noncovalent intermolecular interactions. However, despite their tremendous importance, the unique structures of protein side chains require tailored approaches for investigation by nuclear magnetic resonance spectroscopy and so have traditionally been understudied compared with the protein backbone. Here, we review substantial recent methodological advancements within nuclear magnetic resonance spectroscopy to address this issue. Specifically, we consider advancements that provide new insight into methyl-bearing side chains, show the potential of using non-natural amino acids and reveal the actions of charged side chains. Combined, the new methods promise unprecedented characterisations of side chains that will further elucidate protein function.

FID-Net: A versatile deep neural network architecture for NMR spectral reconstruction and virtual decoupling.

In recent years, the transformative potential of deep neural networks (DNNs) for analysing and interpreting NMR data has clearly been recognised. However, most applications of DNNs in NMR to date either struggle to outperform existing methodologies or are limited in scope to a narrow range of data that closely resemble the data that the network was trained on. These limitations have prevented a widescale uptake of DNNs in NMR. Addressing this, we introduce FID-Net, a deep neural network architecture inspired by WaveNet, for performing analyses on time domain NMR data. We first demonstrate the effectiveness of this architecture in reconstructing non-uniformly sampled (NUS) biomolecular NMR spectra. It is shown that a single network is able to reconstruct a diverse range of 2D NUS spectra that have been obtained with arbitrary sampling schedules, with a range of sweep widths, and a variety of other acquisition parameters. The performance of the trained FID-Net in this case exceeds or matches existing methods currently used for the reconstruction of NUS NMR spectra. Secondly, we present a network based on the FID-Net architecture that can efficiently virtually decouple 13Cα-13Cß couplings in HNCA protein NMR spectra in a single shot analysis, while at the same time leaving glycine residues unmodulated. The ability for these DNNs to work effectively in a wide range of scenarios, without retraining, paves the way for their widespread usage in analysing NMR data.

The anatomy of unfolding of Yfh1 is revealed by site-specific fold stability analysis measured by 2D NMR spectroscopy.

Most techniques allow detection of protein unfolding either by following the behaviour of single reporters or as an averaged all-or-none process. We recently added 2D NMR spectroscopy to the well-established techniques able to obtain information on the process of unfolding using resonances of residues in the hydrophobic core of a protein. Here, we questioned whether an analysis of the individual stability curves from each resonance could provide additional site-specific information. We used the Yfh1 protein that has the unique feature to undergo both cold and heat denaturation at temperatures above water freezing at low ionic strength. We show that stability curves inconsistent with the average NMR curve from hydrophobic core residues mainly comprise exposed outliers that do nevertheless provide precious information. By monitoring both cold and heat denaturation of individual residues we gain knowledge on the process of cold denaturation and convincingly demonstrate that the two unfolding processes are intrinsically different.

NUScon: a community-driven platform for quantitative evaluation of nonuniform sampling in NMR.

Although the concepts of nonuniform sampling (NUS​​​​​​​) and non-Fourier spectral reconstruction in multidimensional NMR began to emerge 4 decades ago , it is only relatively recently that NUS has become more commonplace. Advantages of NUS include the ability to tailor experiments to reduce data collection time and to improve spectral quality, whether through detection of closely spaced peaks (i.e., "resolution") or peaks of weak intensity (i.e., "sensitivity"). Wider adoption of these methods is the result of improvements in computational performance, a growing abundance and flexibility of software, support from NMR spectrometer vendors, and the increased data sampling demands imposed by higher magnetic fields. However, the identification of best practices still remains a significant and unmet challenge. Unlike the discrete Fourier transform, non-Fourier methods used to reconstruct spectra from NUS data are nonlinear, depend on the complexity and nature of the signals, and lack quantitative or formal theory describing their performance. Seemingly subtle algorithmic differences may lead to significant variabilities in spectral qualities and artifacts. A community-based critical assessment of NUS challenge problems has been initiated, called the "Nonuniform Sampling Contest" (NUScon), with the objective of determining best practices for processing and analyzing NUS experiments. We address this objective by constructing challenges from NMR experiments that we inject with synthetic signals, and we process these challenges using workflows submitted by the community. In the initial rounds of NUScon our aim is to establish objective criteria for evaluating the quality of spectral reconstructions. We present here a software package for performing the quantitative analyses, and we present the results from the first two rounds of NUScon. We discuss the challenges that remain and present a roadmap for continued community-driven development with the ultimate aim of providing best practices in this rapidly evolving field. The NUScon software package and all data from evaluating the challenge problems are hosted on the NMRbox platform.

A distal regulatory region of a class I human histone deacetylase.

Histone deacetylases (HDACs) are key enzymes in epigenetics and important drug targets in cancer biology. Whilst it has been established that HDACs regulate many cellular processes, far less is known about the regulation of these enzymes themselves. Here, we show that HDAC8 is allosterically regulated by shifts in populations between exchanging states. An inactive state is identified, which is stabilised by a range of mutations and resembles a sparsely-populated state in equilibrium with active HDAC8. Computational models show that the inactive and active states differ by small changes in a regulatory region that extends up to 28 Å from the active site. The regulatory allosteric region identified here in HDAC8 corresponds to regions in other class I HDACs known to bind regulators, thus suggesting a general mechanism. The presented results pave the way for the development of allosteric HDAC inhibitors and regulators to improve the therapy for several disease states.

Multiquantum Chemical Exchange Saturation Transfer NMR to Quantify Symmetrical Exchange: Application to Rotational Dynamics of the Guanidinium Group in Arginine Side Chains.

Chemical exchange saturation transfer (CEST) NMR experiments have emerged as a powerful tool for characterizing dynamics in proteins. We show here that the CEST approach can be extended to systems with symmetrical exchange, where the NMR signals of all exchanging species are severely broadened. To achieve this, multiquantum CEST (MQ-CEST) is introduced, where the CEST pulse is applied to a longitudinal multispin order density element and the CEST profiles are encoded onto nonbroadened nuclei. The MQ-CEST approach is demonstrated on the restricted rotation of guanidinium groups in arginine residues within proteins. These groups and their dynamics are essential for many enzymes and for noncovalent interactions through the formation of hydrogen bonds, salt-bridges, and π-stacking interactions, and their rate of rotation is highly indicative of the extent of interactions formed. The MQ-CEST method is successfully applied to guanidinium groups in the 19 kDa L99A mutant of T4 lysozyme.

Mechanism of Crosstalk between the LSD1 Demethylase and HDAC1 Deacetylase in the CoREST Complex.

The transcriptional corepressor complex CoREST is one of seven histone deacetylase complexes that regulate the genome through controlling chromatin acetylation. The CoREST complex is unique in containing both histone demethylase and deacetylase enzymes, LSD1 and HDAC1, held together by the RCOR1 scaffold protein. To date, it has been assumed that the enzymes function independently within the complex. Now, we report the assembly of the ternary complex. Using both structural and functional studies, we show that the activity of the two enzymes is closely coupled and that the complex can exist in at least two distinct states with different kinetics. Electron microscopy of the complex reveals a bi-lobed structure with LSD1 and HDAC1 enzymes at opposite ends of the complex. The structure of CoREST in complex with a nucleosome reveals a mode of chromatin engagement that contrasts with previous models.