Development of Solid State NMR Methods and Technology

National Institute Of Diabetes And Digestive And Kidney Diseases

Investigators

Linked publications, trials & patents

Abstract

Progress in FY2023 was in the following areas: (1) TIME-RESOLVED SOLID STATE NMR OF AMYLOID-BETA SELF-ASSEMBLY: In FY2023, we used our rapid mixing/rapid freezing technology, developed in recent years, to examine the earliest stages of self-assembly by the 40-residue amyloid-beta peptide (Ab40), following a rapid pH drop from pH 12 to pH 7. Solid state NMR spectra of 13C-labeled Ab40 samples, recorded at 25 K with signal enhancements from dynamic nuclear polarization, show a dramatic change in the Ab40 conformational distribution within 1 millisecond, from a random-coil-like distribution to a distribution that favors beta-strand conformations. After this large initial change, subsequent changes are subtle, up to times as long as one hour. To complement the time-resolved solid state NMR data, we performed time-resolved light scattering measurements under the same solvent conditions, by using a stopped-flow fluorescence instrument with the detection wavelength set equal to the excitation wavelength. The light-scattering data allow Ab40 oligomer sizes to be quantified. We find that Ab40 remains monomeric up to about 10 milliseconds, is approximately octameric after 0.5 seconds, and forms oligomers comprising 100 molecules after 30 minutes. Thus, although Ab40 is widely considered to be a random-coil, intrinsically disordered peptide, based on experiments that were performed previously by other labs under non-aggregating conditions, Ab40 monomers actually adopt beta-strand-rich conformations that resemble their conformations in fibrils as soon as the solvent conditions change to conditions that favor aggregation. A paper describing this work was published in Nature Communications in 2023. (2) TIME-RESOLVED SOLID STATE NMR OF RAPID PROTEIN FOLDING: Using the rapid inverse temperature jump method that we reported in 2022, we have carried out a study of the folding process of the 35-residue protein HP35, a widely-studied model system for rapid protein folding. Previous experimental and computational studies of HP35 folding indicated a folding rate around 105 per second (folding times in the 5-20 microsecond range). In our time-resolved solid state NMR experiments, we heat HP35 solutions to 95 C for 30 ms, cool them to 30 C within 0.6 ms, allow the solutions to remain at 30 C for a variable evolution time tau, then freeze the solutions within about 100 microseconds. Low-temperature solid state NMR spectra of the frozen solutions show that HP35 is nearly folded at the smallest possible values of tau, consistent with previous studies. However, HP35 is not completely folded. Changes in solid state NMR signals from amino acid sidechains occur as tau increases from near zero to 10 ms, with time constants in the 3-10 ms range. Thus, the time-resolved solid state NMR data show that a previously unobserved structural annealing process occurs on the millisecond time scale, after microsecond-tim-scale main folding process. A paper describing this work has been submitted for publication. (3) MICRON-SCALE MRI: In 2022, we published a paper in PNAS that reported 1.7 micron spatial resolution in 1H MRI of simple test samples. This is the current world's record for spatial resolution in inductively-detected (i.e., normal) MRI. The resolution is limited by the signal-to-noise ratio (SNR) for detection of 1H NMR signals. We enhance the SNR by performing the experiments at very low sample temperatures (5 K in the PNAS paper) and using dynamic nuclear polarization (DNP) to increase nuclear spin polarizations above their thermal-equilibrium values. The SNR can be further increased by reducing the effective noise temperature of the signal detection circuitry. In FY2023, we have investigated the use of cooled electronics, including a cold transmit/receive switching circuit and a cold preamplifier to reduce the noise temperature. A circuit design that seems to be effective has been developed and tested. We are currently incorporating this circuit into a liquid helium cryostat for future experiments. If the noise temperature can be reduced from 80 K (as in our previously published work) to less than 4 K, which appears to be possible, the overall SNR will be enhanced by a factor of 4.5. This should allow us to achieve 1.0 micron isotropic spatial resolution. Experiments on real biological samples can then be attempted.