Methods for the determination of the structures and dynamics of proteins by solid-state NMR spectroscopy

Protein molecules perform a vast array of functions in living organisms and the characterisation of their structures and dynamics is a key step towards a full understanding of many biological processes. Magic angle spinning (MAS) solid-state NMR (SSNMR) spectroscopy has emerged as a uniquely powerful technique for the extraction of such information at atomic resolution, with mounting successes founded on continual developments in methodology and technology. In this thesis, a number of new approaches for probing the structures and dynamics of proteins are presented, towards the aim of overcoming current challenges regarding sensitivity, spectral resolution and a shortage of quantitative experimental observables. A streamlined method for simultaneously obtaining long-distance homonuclear (13C-13C) and heteronuclear (15N-13C) contacts is introduced that relies on the third spinassisted recoupling (TSAR) mechanism. The experiment, dubbed "time-shared TSAR" (TSTSAR), effectively doubles the information content of spectra and reduces the required experimental time to that needed for just one of the equivalent PAR or PAINCP experiments. An approach for the quantitative study of large proteins and complexes is presented, relying on a combination of proton detection at "ultrafast" (≥55 kHz) MAS frequencies, sample deuteration and optional paramagnetic doping. This is successfully employed for the characterisation of a >300 kDa precipitated complex of the protein GB1 with full length human immunoglobulin (IgG), with only a few nanomoles of sample. Recent advances in MAS technology have enabled spinning frequencies of 100 kHz and above to be obtained. Using the dipeptide β-Asp-Ala, it is found that under such conditions, protons lines are narrowed to an extent similar to that achievable using contemporary homonuclear decoupling methods, leading to a time-efficient method for obtaining resolved spectra of small, natural-abundance molecules. Similar experiments with a GB1-IgG complex sample confirm the technology’s applicability to non-model biological systems, despite the tiny rotor volume of 0.7 μL (≤3 nanomoles of complex). 15N R1ρ relaxation rates are measured for the same complex and compared with identical measurements in crystalline GB1, allowing for a direct comparison between the slow (ns-ms) dynamics of the protein in different molecular environments. Motions on this time scale are found to be more prevalent in the complex, possibly evidence of an overall collective molecular motion. An approach for the measurement of aliphatic 13C relaxation rates in fully protonated samples is presented, based on a combination of ultrafast MAS rates and alternately labelled samples. Sample spinning at ≥80 kHz enables resolved 13Cα-1H correlations, forming a base for 13Cα relaxation experiments that are subsequently performed on crystalline [1,3-13C,15N]GB1 and analysed using a simple model-free (SMF) treatment. It is noted that without further data, this analysis is likely inadequate for an accurate description of the dynamics of the protein. The measurement of 13C’ R1ρ relaxation rates at ultrafast MAS rates is introduced as a probe of backbone protein dynamics in fully protonated samples. 13C and 15N R1 and R1ρ relaxation rates are measured in crystalline [U-13C,15N]GB1 and analysed using the SMF formalism. An examination of simulated spectral densities rationalises the apparent inconsistencies that arise from this and reveals that motions in GB1 occur on at least two time scales. A combined 15N/13C extended model-free (EMF) analysis is conducted for peptide plane motions in GB1, whereupon it is found that the addition of 13C data helps to remove fitting artefacts present in a 15N-only analysis