Internal protein dynamics on ps to μs timescales as studied by multi-frequency 15N solid-state NMR relaxation

Abstract

A comprehensive analysis of the dynamics of the SH3 domain of chicken alpha-spectrin is presented, based upon (15)N T1 and on- and off-resonance T1ρ relaxation times obtained on deuterated samples with a partial back-exchange of labile protons under a variety of the experimental conditions, taking explicitly into account the dipolar order parameters calculated from (15)N-(1)H dipole-dipole couplings. It is demonstrated that such a multi-frequency approach enables access to motional correlation times spanning about 6 orders of magnitude. We asses the validity of different motional models based upon orientation autocorrelation functions with a different number of motional components. We find that for many residues a "two components" model is not sufficient for a good description of the data and more complicated fitting models must be considered. We show that slow motions with correlation times on the order of 1-10 μs can be determined reliably in spite of rather low apparent amplitudes (below 1 %), and demonstrate that the distribution of the protein backbone mobility along the time scale axis is pronouncedly non-uniform and non-monotonic: two domains of fast (τ < 10(-10) s) and intermediate (10(-9) s < τ < 10(-7) s) motions are separated by a gap of one order of magnitude in time with almost no motions. For slower motions (τ > 10(-6) s) we observe a sharp ~1 order of magnitude decrease of the apparent motional amplitudes. Such a distribution obviously reflects different nature of backbone motions on different time scales, where the slow end may be attributed to weakly populated "excited states." Surprisingly, our data reveal no clearly evident correlations between secondary structure of the protein and motional parameters. We also could not notice any unambiguous correlations between motions in different time scales along the protein backbone emphasizing the importance of the inter-residue interactions and the cooperative nature of protein dynamics.

Studying dynamics by magic-angle spinning solid-state NMR spectroscopy: Principles and applications to biomolecules

“…The 15 N-15 N proton-driven spin diffusion rate constant has been estimated in a protonated protein at 10 kHz MAS to approximately 0.025 s −1 . At ambient temperature, backbone 15 N sites have typical R 1 rate constants of the order of 0.01 to 0.5 s −1 [96][97][98][99][100]; rate constants for side chain NH 2 groups (Asn, Gln) and Trp NHε have been reported of the order of 0.15 to 1 s −1 [101]; NH 3 + groups may have significantly higher rate constants. As the spin-diffusion rate constants are only slightly smaller than backbone R 1 rate constants, one may thus expect that spin diffusion has a noticeable effect on the measured rate constants.…”

Optimized “detectors” for dynamics analysis in solid-state NMR

“…This was inspired by the success of the Lipari-Szabo modelfree approach 4 or extensions of that approach [5][6][7] in solution-state NMR, achieved by setting the correlation time of the overall tumbling in solids to infinity. While the validity of the LipariSzabo model-free approach in solution-state NMR was justified in detail (similar correlation functions are derived by Halle and Wennerström, 8 with the validity of this and the model-free approach discussed in detail 9 ), studies in solid-state NMR have investigated primarily the effects that the model selection [10][11][12] and the data included 13 have on the analysis of the experimental data. Although there are indications that modeling the correlation function with a sum of exponential functions is not always reliable, these studies do not directly consider modeling behavior in the case that the real motion is too complex to be fully characterized by the experimental data.…”

“…In this example the detector responses are given by r 1 (q ,S) = aR 1,400 (q ,S) + bR 1,850 (q ,S) r 2 (q ,S) = cR 1,400 (q ,S) + dR 1,850 (q ,S) . (10) We may similarly define a detector sensitivity, which is obtained using the same linear combination, but now for the sensitivities of the two rate constants.…”

“…This linear combination, defined in Eq. (10), is basically a coordinate transformation from the black coordinate system in Fig. 3(a) to the coordinate system defined by the red and blue axes (Fig 3a and b) which are vectors tangential to the allowed region (blue area).…”

Dynamics in the solid-state: perspectives for the investigation of amyloid aggregates, membrane proteins and soluble protein complexes

“…Alternatively, site-specific incoherent T 2 type relaxation rates can be obtained from T 1q measurements (Akasaka et al 1983). This approach has been implemented recently for GB1 (Lewandowski et al 2011b), ubiquitin (Tollinger et al 2012), and the a-spectrin SH3 domain (Krushelnitsky et al 2010;Zinkevich et al 2013).…”