Istomin Andrei Y, Jacobs Donald J, Livesay Dennis R
Department of Physics and Optical Science, University of North Carolina at Charlotte 28223, USA.
Protein Sci. 2007 Nov;16(11):2564-9. doi: 10.1110/ps.073124507.
The time it takes for proteins to fold into their native states varies over several orders of magnitude depending on their native-state topology, size, and amino acid composition. In a number of previous studies, it was found that there is strong correlation between logarithmic folding rates and contact order for proteins that fold with two-state kinetics, while such correlation is absent for three-state proteins. Conversely, strong correlations between folding rates and chain length occur within three-state proteins, but not in two-state proteins. Here, we demonstrate that chain lengths and folding rates of two-state proteins are not correlated with each other only when all-alpha, all-beta, and mixed-class proteins are considered together, which is typically the case. However, when considering all-alpha and all-beta two-state proteins separately, there is significant linear correlation between folding rate and size. Moreover, the sets of data points for the all-alpha and all-beta classes define asymptotes of lower and upper limits on folding rates of mixed-class proteins. By analyzing correlation of other topological parameters with folding rates of two-state proteins, we find that only the long-range order exhibits correlation with folding rates that is uniform over all three classes. It is also the only descriptor to provide statistically significant correlations for each of the three structural classes. We give an interpretation of this observation in terms of Makarov and Plaxco's diffusion-based topomer-search model.
蛋白质折叠成其天然状态所需的时间因天然状态的拓扑结构、大小和氨基酸组成而异,相差好几个数量级。在之前的一些研究中发现,对于以两态动力学折叠的蛋白质,对数折叠速率与接触序之间存在很强的相关性,而对于三态蛋白质则不存在这种相关性。相反,折叠速率与链长之间的强相关性出现在三态蛋白质中,而在两态蛋白质中则不存在。在这里,我们证明,只有当将所有α螺旋、所有β折叠和混合类蛋白质一起考虑时(通常情况如此),两态蛋白质的链长和折叠速率才不相互关联。然而,当分别考虑所有α螺旋和所有β折叠的两态蛋白质时,折叠速率与大小之间存在显著的线性相关性。此外,所有α螺旋和所有β折叠类别的数据点集定义了混合类蛋白质折叠速率的下限和上限渐近线。通过分析其他拓扑参数与两态蛋白质折叠速率的相关性,我们发现只有长程有序与折叠速率呈现出在所有三个类别中都一致的相关性。它也是唯一能为三个结构类别中的每一个提供具有统计学意义相关性的描述符。我们根据马卡罗夫和普拉斯科基于扩散的拓扑异构体搜索模型对这一观察结果进行了解释。