NMR spectroscopy method development

The authors report signal enhancements in high-resolution liquid-state NMR by a new setup for dynamic nuclear polarization. Two-dimensional NMR techniques are used to transfer hyperpolarization within the nuclei of molecules such as metabolites or drugs.

Solution NMR spectroscopy provides rich structural information on biomolecules, however, its resolution becomes limited when molecular size increases, due to short-lived nuclear magnetic responses to electromagnetic radiation. Here, the authors sustain long-lived coherences for the aliphatic protons of glycine residues within protein lysozyme, yielding substantial through-space magnetization transfers, and mapping interacting atoms in the protein structure.

Dynamic nuclear polarization is a technique to enhance NMR signals, often achieved by continuous-wave irradiation of the samples. Here, the authors create large polarization gradients by using high-power chirped inversion pulses at 94 GHz at an average power of a few hundred mW with fast build-up times, reaching NMR signal enhancements of up to 340.

NMR relaxometry can provide information on paramagnetic as well as diamagnetic compounds in solution by analyzing nuclear-spin relaxation times, and, operated at ultra-low magnetic fields, slow processes can be studied. Here, the authors employ zero and ultra-low magnetic fields using an atomic magnetometer as a simple, portable, robust, inexpensive, and sensitive tool to analyze chemical solutions and biofluids.

Zero- and ultralow-field (ZULF) nuclear magnetic resonance (NMR) has emerged as a complementary technique to conventional high-field (≳1 T) NMR, however, the ability to distinguish heteronuclear coupling remains challenging due to the low abundance of certain nuclei. Here, the authors apply non-hydrogenative parahydrogen-induced hyperpolarization to identify naturally abundant molecules, and demonstrate the detection of pyridine derivatives at a concentration level of ~1 mM.

Protein liquid-liquid phase separation is an important phenomenon in biology. Here, the authors demonstrate an approach to characterize the evolution of protein phases in both time and space using a fluorinated probe molecule in NMR spectroscopy.

Zero to ultralow-field NMR provides chemical information in the absence of a high magnetic field but it is difficult to measure molecules with quadrupolar nuclei due to their fast relaxation. This study examines zero-field J-spectra from isotopologues of ammonium cations, with quadrupolar nuclei, revealing the presence of a primary isotope effect of −58 mHz.

An integrated approach combing solution and solid-state NMR, molecular dynamics simulations and neutron reflectometry is applied to characterize dynamic membrane bound forms of ADP-ribosylation factor 1 (Arf1).

The main limitations in NMR techniques are low sensitivity and the requirement for complex instrumentation. Here the authors show that a microfluidic chip with a single untuned planar spiral microcoil, combined with laser-light induced hyperpolarization, allows for multidimensional and heteronuclear Nuclear Magnetic Resonance spectroscopy on picomole quantities of material.

In situ illumination of liquid-state NMR samples allows to characterise light-dependent chemical and biological phenomena, but, in practice, the position of an NMR sample deep within the bore of a spectrometer magnet renders such illumination challenging. Here, the authors demonstrate the working principles of a sample illumination device, with an LED array positioned directly at the top of special sample tube, which is inserted into the NMR spectrometer.

Here, the authors show deep neural networks can be trained to transform crowded ¹H,¹³C-NMR spectra of large proteins into readily interpretable spectra, negating the requirement for uniform deuteration to analyse complex NMR spectra.

Here, the authors utilized an evolutionary algorithm and artificial intelligence to design new basic 2D biomolecular NMR experiments to accelerate the acquisition of large biomolecular spectra. The method enables recording the spectra of poorly soluble or unstable macromolecules and analyzing the kinetics of biomolecular aggregation and oligomerization. The authors laid the foundation for accelerating multidimensional NMR experiments at high and ultra-high magnetic fields.

The authors present BARASA, an approach to assign backbone triple resonance spectra of proteins that augments traditional approaches with a Bayesian statistical analysis of the observed chemical shifts. The algorithm employs a simulated annealing engine to establish a consensus set of resonance assignments and is tested against systems ranging in size to over 450 amino acids including examples of intrinsically disordered proteins.