The analogous equations for the CSA relaxation mechanism are presented in the SI. These equations, as well as previous theoretical analyses of R1ρ relaxation in rotating solids [23] and [24], demonstrate that the sampling frequencies in the R1ρ experiments are the combinations of ω1 and ωR instead

of ω1 only: the dominant contribution to R1ρ PFT�� cell line comes from the spectral density functions J(ω1 ± ωR) and J(ω1 ± 2ωR). The numerically simulated R1ρ vs ω1 dependence [25] show that at ω1 < ωR, R1ρ increases with increasing ω1, which can be explained only by J(ω1 − ωR) term. Thus, R1ρ depends not only on the spin-lock field, but on the MAS frequency as well. The MAS dependence of R1ρ is the key point of the present work, as it is highly informative for slow molecular dynamics. Fig. 1 presents analytical simulations of R1ρ for different correlation times of motion. It is evident that R1ρ in a rotating solid follows conventional wisdom (i.e., behaves like “normal” static R1ρ or R2) only if the correlation time is much shorter than (ω1 ± ωR)−1 and (ω1 ± 2ωR)−1. CDK activity Otherwise, we observe a non-trivial dependence on ωR. Recently, Lewandowski et al. have measured the

R1ρ(ωR) dependence integrated over all residues of a solid protein [16], which was found to feature a shown strong, sharply increasing ωR dependence at low ωR and a plateau at high ωR (∼60 kHz). Since their experiment was conducted on a protonated protein, such dependence was correctly explained by the coherent contribution which dominates at low and is negligible at high ωR. However, in a deuterated

protein, the coherent contribution is expected to be obviously much smaller at low ωR or, as demonstrated by our data (see below), even completely negligible. Fig. 2 shows the average 15N R1ρ(ωR) measured in deuterated SH3 domain with 20% back-exchanged labile protons. The relaxation decays were measured for the whole integral intensity of 1D proton-detected spectrum (see Figs. S1 and S2 of the SI). The observed positive dependence unambiguously and without any Tolmetin mathematical treatment allows for two important conclusions. First, the coherent contribution to R1ρ in a deuterated protein is much smaller than incoherent relaxation even at low ωR values. Otherwise, a negative R1ρ(ωR) dependence would be expected [16]. Even if one assumes that the coherent contribution is non-negligible at 4 kHz MAS (which is rather unlikely for the reason described at the end of this paragraph), it is absolutely negligible at slightly faster MAS rates due to its very strong MAS dependence [16]. Note that the coherent contribution at slow MAS in a fully protonated protein is about 10,000 s−1 [16], whereas in the deuterated protein R1ρ has a value of about 10 s−1 ( Fig. 2).