Chalikian T V, Totrov M, Abagyan R, Breslauer K J
Department of Chemistry Rutgers, State University of New Jersey Piscataway 08855, USA.
J Mol Biol. 1996 Jul 26;260(4):588-603. doi: 10.1006/jmbi.1996.0423.
We report the first thermodynamic characterization of protein hydration that does not depend on model compound data but rather is based exclusively on macroscopic (volumetric) and microscopic (X-ray) measurements on protein molecules themselves. By combining these macroscopic and microscopic characterizations, we describe a quantitative model that allows one for the first time to predict the partial specific volumes, v(zero), and the partial specific adiabatic compressibilities, ks(zero), of globular proteins from the crystallographic coordinates of the constituent atoms, without using data derived from studies on low-molecular-mass model compounds. Specifically, we have used acoustic and densimetric techniques to determine v(zero) and ks(zero) for 15 globular proteins over a temperature range from 18 to 55 degrees C. For the subset of the 12 proteins with known three-dimensional structures, we calculated the molecular volumes as well as the solvent-accessible surface areas of the constituent charged, polar and nonpolar atomic groups. By combining these measured and calculated properties and applying linear regression analysis, we determined, as a function of temperature, the average hydration contributions to v(zero) and ks(zero) of 1 A2 of the charged, polar, and nonpolar solvent-accessible protein surfaces. We compared these results with those derived from studies on low-molecular-mass compounds to assess the validity of existing models of protein hydration based on small molecule data. This comparison revealed the following features: the hydration contributions to v(zero) and ks(zero) of charged protein surface groups are similar to those of charged groups in small organic molecules. By contrast, the hydration contributions to v(zero) and ks(zero) of polar protein surface groups are qualitatively different from those of polar groups in low-molecular-mass compounds. We suggest that this disparity may reflect the presence of networks of water molecules adjacent to polar protein surface areas, with these networks involving waters from second and third coordination spheres. For nonpolar protein surface groups, we find the ability of low-molecular-mass compounds to model successfully protein properties depends on the temperature domain being examined. Specifically, at room temperatures and below, the hydration contribution to ks(zero) of protein nonpolar surface atomic groups is close to that of nonpolar groups in small organic molecules. By contrast, at higher temperatures, the hydration contribution to ks(zero) of protein nonpolar surface groups becomes more negative than that of nonpolar groups in small organic molecules. We suggest that this behaviour may reflect nonpolar groups on protein surfaces being hydrated independently at low temperatures, while at higher temperatures some of the solvating waters become influenced by neighboring polar groups. We discuss the implications of our aggregate results in terms of various approaches currently being used to describe the hydration properties of globular proteins, particularly focusing on the limitations of existing additive models based on small molecule data.
我们报告了蛋白质水合作用的首个热力学表征,该表征不依赖于模型化合物数据,而是完全基于对蛋白质分子本身的宏观(体积)和微观(X射线)测量。通过结合这些宏观和微观表征,我们描述了一个定量模型,该模型首次使人们能够根据组成原子的晶体学坐标预测球状蛋白质的偏比容v(0)和偏比绝热压缩系数ks(0),而无需使用来自低分子量模型化合物研究的数据。具体而言,我们使用声学和密度测定技术在18至55摄氏度的温度范围内测定了15种球状蛋白质的v(0)和ks(0)。对于12种具有已知三维结构的蛋白质子集,我们计算了组成带电、极性和非极性原子基团的分子体积以及溶剂可及表面积。通过结合这些测量和计算得到的性质并应用线性回归分析,我们确定了作为温度函数的带电、极性和非极性溶剂可及蛋白质表面每1 Ų对v(0)和ks(0)的平均水合贡献。我们将这些结果与来自低分子量化合物研究的结果进行比较,以评估基于小分子数据的现有蛋白质水合模型的有效性。这种比较揭示了以下特征:蛋白质带电表面基团对v(0)和ks(0)的水合贡献与小有机分子中带电基团的类似。相比之下,蛋白质极性表面基团对v(0)和ks(0)的水合贡献与低分子量化合物中极性基团的在性质上有所不同。我们认为这种差异可能反映了极性蛋白质表面区域附近存在水分子网络,这些网络涉及来自第二和第三配位球的水。对于非极性蛋白质表面基团,我们发现低分子量化合物成功模拟蛋白质性质的能力取决于所研究的温度范围。具体而言,在室温及以下,蛋白质非极性表面原子基团对ks(0)的水合贡献接近小有机分子中非极性基团的贡献。相比之下,在较高温度下,蛋白质非极性表面基团对ks(0)的水合贡献比小有机分子中非极性基团的更负。我们认为这种行为可能反映了蛋白质表面的非极性基团在低温下独立水合,而在较高温度下一些溶剂化水受到相邻极性基团的影响。我们根据目前用于描述球状蛋白质水合性质的各种方法讨论了我们综合结果的含义,特别关注基于小分子数据的现有加和模型的局限性。