Optimum levels of exchangeable protons in perdeuterated proteins for proton detection in MAS solid-state NMR spectroscopy.

ABSTRACT

We present a systematic study of the effect of the level of exchangeable protons on the observed amide proton linewidth obtained in perdeuterated proteins. Decreasing the amount of D(2)O employed in the crystallization buffer from 90 to 0%, we observe a fourfold increase in linewidth for both (1)H and (15)N resonances. At the same time, we find a gradual increase in the signal-to-noise ratio (SNR) for (1)H-(15)N correlations in dipolar coupling based experiments for H(2)O concentrations of up to 40%. Beyond 40%, a significant reduction in SNR is observed. Scalar-coupling based (1)H-(15)N correlation experiments yield a nearly constant SNR for samples prepared with < or =30% H(2)O. Samples in which more H(2)O is employed for crystallization show a significantly reduced NMR intensity. Calculation of the SNR by taking into account the reduction in (1)H T (1) in samples containing more protons (SNR per unit time), yields a maximum SNR for samples crystallized using 30 and 40% H(2)O for scalar and dipolar coupling based experiments, respectively. A sensitivity gain of 3.8 is obtained by increasing the H(2)O concentration from 10 to 40% in the CP based experiment, whereas the linewidth only becomes 1.5 times broader. In general, we find that CP is more favorable compared to INEPT based transfer when the number of possible (1)H,(1)H interactions increases. At low levels of deuteration (> or =60% H(2)O in the crystallization buffer), resonances from rigid residues are broadened beyond detection. All experiments are carried out at MAS frequency of 24 kHz employing perdeuterated samples of the chicken alpha-spectrin SH3 domain.