West N J, Smith L J
New Chemistry Laboratory, University of Oxford, South Parks Road, Oxford, OX1 3QT, UK.
J Mol Biol. 1998 Jul 31;280(5):867-77. doi: 10.1006/jmbi.1998.1911.
The behaviour of amino acid side-chains in proteins in solution has been characterised by analysing NMR 3JHalphaH beta coupling constants and crystallographic chi1 torsion angles. Side-chains both in the core of native folded proteins and in situations where there is an absence of close packing including the random coil state have been considered. An analysis of experimental 3JHalphaH beta coupling constant data for ten proteins shows that in the core of native proteins a very close similarity is observed between the chi1 conformations adopted in solution and in crystals. There is clear evidence, however, for significant motional averaging about the chi1 torsion angles in solution. Using a model of a Gaussian distribution about the average torsion angles the extent of these fluctuations has been quantified; the standard deviation for the motion is 26 degrees, the fluctuations about chi1 in the protein core being similar in size to those found for main-chain phi torsion angles in solution. From the distribution of chi1 torsion angles in a data base of protein crystal structures, torsion angle populations and coupling constants have been predicted for a random coil polypeptide. Significant variations in the chi1 distributions for different amino acids give differences in the predicted coupling constants; for 3JHalphaH beta, for example, values of 5.1 and 5.7 Hz are predicted for serine compared with 4.9 and 9.9 Hz for leucine. Experimental data for short unstructured peptides show an excellent agreement with the predictions, indicating that the overall chi1 distributions in protein crystals reflect the local preferences of the amino acids. Predictions from the protein data base therefore provide an important framework for interpreting experimental data for non-native protein conformations and for residues on the surface of folded proteins.
通过分析核磁共振3JHαHβ耦合常数和晶体学χ1扭转角,对溶液中蛋白质氨基酸侧链的行为进行了表征。研究考虑了天然折叠蛋白质核心中的侧链以及不存在紧密堆积的情况(包括无规卷曲状态)。对十种蛋白质的实验3JHαHβ耦合常数数据进行分析表明,在天然蛋白质核心中,溶液和晶体中采用的χ1构象之间存在非常紧密的相似性。然而,有明确证据表明溶液中围绕χ1扭转角存在显著的运动平均化。使用围绕平均扭转角的高斯分布模型,对这些波动的程度进行了量化;运动的标准偏差为26度,蛋白质核心中围绕χ1的波动大小与溶液中主链φ扭转角的波动相似。根据蛋白质晶体结构数据库中χ1扭转角的分布,预测了无规卷曲多肽的扭转角群体和耦合常数。不同氨基酸的χ1分布存在显著差异,导致预测的耦合常数不同;例如,对于3JHαHβ,丝氨酸的预测值为5.1和5.7 Hz,而亮氨酸的预测值为4.9和9.9 Hz。短无结构肽的实验数据与预测结果非常吻合,表明蛋白质晶体中的整体χ1分布反映了氨基酸的局部偏好。因此,来自蛋白质数据库的预测为解释非天然蛋白质构象以及折叠蛋白质表面残基的实验数据提供了重要框架。