Department of Chemistry and Biochemistry, University of Maryland, College Park, MD 20742, USA.
J Biomol NMR. 2011 May;50(1):19-33. doi: 10.1007/s10858-011-9474-8. Epub 2011 Feb 9.
We performed density functional calculations of backbone (15)N shielding tensors in the regions of beta-sheet and turns of protein G. The calculations were carried out for all twenty-four beta-sheet residues and eight beta-turn residues in the protein GB3 and the results were compared with the available experimental data from solid-state and solution NMR measurements. Together with the alpha-helix data, our calculations cover 39 out of the 55 residues (or 71%) in GB3. The applicability of several computational models developed previously (Cai et al. in J Biomol NMR 45:245-253, 2009) to compute (15)N shielding tensors of alpha-helical residues is assessed. We show that the proposed quantum chemical computational model is capable of predicting isotropic (15)N chemical shifts for an entire protein that are in good correlation with experimental data. However, the individual components of the predicted (15)N shielding tensor agree with experiment less well: the computed values show much larger spread than the experimental data, and there is a profound difference in the behavior of the tensor components for alpha-helix/turns and beta-sheet residues. We discuss possible reasons for this.
我们对蛋白 G 中β-折叠和转角区域的骨架(15)N 屏蔽张量进行了密度泛函计算。我们对蛋白 GB3 中的所有 24 个β-折叠残基和 8 个β-转角残基进行了计算,并将结果与来自固态和溶液 NMR 测量的可用实验数据进行了比较。与α-螺旋数据一起,我们的计算涵盖了 GB3 中 55 个残基中的 39 个(或 71%)。评估了先前开发的几种计算模型(Cai 等人,在 J Biomol NMR 45:245-253, 2009)在计算α-螺旋残基(15)N 屏蔽张量中的适用性。我们表明,所提出的量子化学计算模型能够预测与实验数据具有良好相关性的整个蛋白质的各向同性(15)N 化学位移。然而,预测(15)N 屏蔽张量的各个分量与实验数据的一致性较差:计算值的离散程度比实验数据大得多,并且对于α-螺旋/转角和β-折叠残基,张量分量的行为存在显著差异。我们讨论了这种差异的可能原因。