Final Report Summary - SOLID_NMR_DYNAMICS (Development of solid state NMR methods at 100 kHz magic angle spinning frequency for the study of internal protein dynamics and the application to membrane proteins)
The knowledge about protein dynamics reveals essential information about the conformational space sampled by the protein and the underlying energy landscape, thus, providing insights into the molecular mechanisms of protein function, e.g. molecular recognition processes. Recent developments in Magic Angle Spinning (MAS) technology allow nowadays MAS frequencies of more than 100 kHz and proton detection in solid-state NMR, even in the absence of sample deuteration. The possibility of proton detection in the solid state enables also the measurement of 15N relaxation rate constants, which characterize the internal dynamics of the protein backbone. 15N relaxation measurements were established during the last decades in solution NMR as a powerful tool to investigate internal backbone dynamics. 15N R1 and 15N R1ρ or CPMG-based R2 measurements can characterize internal dynamics on the fast ps-ns time scale. Relaxation-dispersion R1ρ or CPMG experiments can characterize internal dynamics in a time window from about 50 μs to 10 ms. However, there is a gap between the few ns and the 50 μs time range which is not accessible to solution-state NMR relaxation measurements. In fact, the overall tumbling of the protein, (with a correlation time of about 4 ns in the case ubiquitin in solution) limits the measurement of slower motions. In contrast, solid-state NMR relaxation measurements are not restricted by the overall tumbling limitation and, in particular 15N R1ρ relaxation experiments, can provide information on the ns-μs time window, thus, complementing solution-state NMR measurements.
The determination of relaxation rate constants in the solid state is in general more difficult than in solution state, as coherent interactions, e.g. due to dipolar coupling evolution, add to the apparent magnetization decay rate and need to be disentangled from the relaxation due to internal dynamics. We have measured various 15N and 1H relaxation constants in [2H,15N,13C]-labeled microcrystalline ubiquitin where only the backbone amide group is fully protonated, and we have empirically investigated the contribution of coherent effects.
We have also determined 15N R1ρ relaxation rate constants at different MAS frequencies and we have observed a MAS-frequency dependence of 15N R1ρ relaxation rate constants, with rate constants at 60 kHz MAS being significantly larger than at 110 kHz (Figure 1). We have excluded coherent contributions, e.g. 1H-15N dipolar coupling evolution, by repeating our measurements on [2H,15N,13C]-labeled microcrystalline ubiquitin where only 20% of the amides are protonated, i.e. a highly diluted proton network. Since very similar results were obtained for both samples, coherent contributions due to 1H-1H dipolar coupling could be excluded. A temperature effect due to air friction with the rotor at different MAS frequencies (and thus different radial velocities) was also ruled out, thus, leaving internal dynamics as only possible explanation of our observation of MAS dependence of 15N R1ρ rate constants.
The MAS dependence of 15N R1ρ rate constants originates from the fact that in the expression of 15N R1ρ in the solid state under MAS the J(0) spectral density of motion, or more precisely J(ω1) where ω1 is the spin-lock frequency, is replaced by spectral densities that depend on the sum and difference of spin-lock frequency, ω1, and MAS frequency (R. Kurbanov, T. Zinkevich, A. Krushelnitsky, J. Chem. Phys., 2011).
The above described 15N R1ρ relaxation experiments on ubiquitin with variable MAS frequency under fast MAS conditions (60-110 kHz) were evaluated using a model-free analysis and enabled us to detect low amplitude dynamics across the entire protein on a time scale of about 1 µs (Figure 2). On the same µs time scale we found larger amplitude motions for several sites, including the first loop region connecting β-strand 1 and 2, as well as the N-terminus of the α-helix (Glu 24, Asn 25) and Asp 52. These residues were known to be involved in a hydrogen-bond reordering process, however on a two to three orders of magnitude slower time scale. According to our analysis, the ns-µs backbone protein backbone dynamics occur in a narrow time scale range at around 1µs, but are of lower amplitude than concluded in previous solid-state NMR studies. Our study contributes to the debate about the presence and amplitude of protein backbone dynamics in the ns-µs time range, by revealing experimentally, with residue-specific resolution, the low µs time scale for protein backbone dynamics in microcrystalline ubiquitin. Amplitudes of motion in microcrystalline ubiquitin (for a time window <100 µs) appear however reduced compared to previous RDC-based studies in solution-state NMR (which are sensitive to <10 ms motions). While our study focused on very general questions about the amplitude and time scales of dynamics present in a model protein, the developed methodology can be expanded to more complicated system, e.g. membrane proteins in a lipid environment. Understanding of the internal and conformational dynamics in a membrane protein, especially those which are drug targets, can provide important mechanistic insight useful to the development of novel drugs targeted to those membrane proteins.
The determination of relaxation rate constants in the solid state is in general more difficult than in solution state, as coherent interactions, e.g. due to dipolar coupling evolution, add to the apparent magnetization decay rate and need to be disentangled from the relaxation due to internal dynamics. We have measured various 15N and 1H relaxation constants in [2H,15N,13C]-labeled microcrystalline ubiquitin where only the backbone amide group is fully protonated, and we have empirically investigated the contribution of coherent effects.
We have also determined 15N R1ρ relaxation rate constants at different MAS frequencies and we have observed a MAS-frequency dependence of 15N R1ρ relaxation rate constants, with rate constants at 60 kHz MAS being significantly larger than at 110 kHz (Figure 1). We have excluded coherent contributions, e.g. 1H-15N dipolar coupling evolution, by repeating our measurements on [2H,15N,13C]-labeled microcrystalline ubiquitin where only 20% of the amides are protonated, i.e. a highly diluted proton network. Since very similar results were obtained for both samples, coherent contributions due to 1H-1H dipolar coupling could be excluded. A temperature effect due to air friction with the rotor at different MAS frequencies (and thus different radial velocities) was also ruled out, thus, leaving internal dynamics as only possible explanation of our observation of MAS dependence of 15N R1ρ rate constants.
The MAS dependence of 15N R1ρ rate constants originates from the fact that in the expression of 15N R1ρ in the solid state under MAS the J(0) spectral density of motion, or more precisely J(ω1) where ω1 is the spin-lock frequency, is replaced by spectral densities that depend on the sum and difference of spin-lock frequency, ω1, and MAS frequency (R. Kurbanov, T. Zinkevich, A. Krushelnitsky, J. Chem. Phys., 2011).
The above described 15N R1ρ relaxation experiments on ubiquitin with variable MAS frequency under fast MAS conditions (60-110 kHz) were evaluated using a model-free analysis and enabled us to detect low amplitude dynamics across the entire protein on a time scale of about 1 µs (Figure 2). On the same µs time scale we found larger amplitude motions for several sites, including the first loop region connecting β-strand 1 and 2, as well as the N-terminus of the α-helix (Glu 24, Asn 25) and Asp 52. These residues were known to be involved in a hydrogen-bond reordering process, however on a two to three orders of magnitude slower time scale. According to our analysis, the ns-µs backbone protein backbone dynamics occur in a narrow time scale range at around 1µs, but are of lower amplitude than concluded in previous solid-state NMR studies. Our study contributes to the debate about the presence and amplitude of protein backbone dynamics in the ns-µs time range, by revealing experimentally, with residue-specific resolution, the low µs time scale for protein backbone dynamics in microcrystalline ubiquitin. Amplitudes of motion in microcrystalline ubiquitin (for a time window <100 µs) appear however reduced compared to previous RDC-based studies in solution-state NMR (which are sensitive to <10 ms motions). While our study focused on very general questions about the amplitude and time scales of dynamics present in a model protein, the developed methodology can be expanded to more complicated system, e.g. membrane proteins in a lipid environment. Understanding of the internal and conformational dynamics in a membrane protein, especially those which are drug targets, can provide important mechanistic insight useful to the development of novel drugs targeted to those membrane proteins.
