Computational Biophysics and Bioinformatics, Department of Physics and Astronomy, Clemson University, Clemson, SC 29634, USA.
Curr Pharm Des. 2013;19(23):4182-90. doi: 10.2174/1381612811319230004.
In this review we discuss the role of protonation states in receptor-ligand interactions, providing experimental evidences and computational predictions that complex formation may involve titratable groups with unusual pKa's and that protonation states frequently change from unbound to bound states. These protonation changes result in proton uptake/release, which in turn causes the pH-dependence of the binding. Indeed, experimental data strongly suggest that almost any binding is pH-dependent and to be correctly modeled, the protonation states must be properly assigned prior to and after the binding. One may accurately predict the protonation states when provided with the structures of the unbound proteins and their complex; however, the modeling becomes much more complicated if the bound state has to be predicted in a docking protocol or if the structures of either bound or unbound receptor-ligand are not available. The major challenges that arise in these situations are the coupling between binding and protonation states, and the conformational changes induced by the binding and ionization states of titratable groups. In addition, any assessment of the protonation state, either before or after binding, must refer to the pH of binding, which is frequently unknown. Thus, even if the pKa's of ionizable groups can be correctly assigned for both unbound and bound state, without knowing the experimental pH one cannot assign the corresponding protonation states, and consequently one cannot calculate the resulting proton uptake/release. It is pointed out, that while experimental pH may not be the physiological pH and binding may involve proton uptake/release, there is a tendency that the native receptor-ligand complexes have evolved toward specific either subcellular or tissue characteristic pH at which the proton uptake/release is either minimal or absent.
在这篇综述中,我们讨论了质子化状态在受体-配体相互作用中的作用,提供了实验证据和计算预测,表明复合物的形成可能涉及具有不寻常 pKa 的可滴定基团,并且质子化状态经常从非结合状态变为结合状态。这些质子化变化导致质子的摄取/释放,这反过来又导致结合的 pH 依赖性。事实上,实验数据强烈表明,几乎任何结合都是 pH 依赖性的,为了正确建模,必须在结合前后正确分配质子化状态。如果提供了未结合蛋白质及其复合物的结构,则可以准确预测质子化状态;但是,如果必须在对接方案中预测结合状态,或者如果未结合或结合的受体-配体的结构不可用,则建模变得更加复杂。在这些情况下出现的主要挑战是结合和质子化状态之间的耦合,以及结合和可滴定基团的离解状态引起的构象变化。此外,在结合前后对质子化状态的任何评估都必须参考结合 pH,而结合 pH 通常是未知的。因此,即使可以正确分配未结合和结合状态下可电离基团的 pKa,如果不知道实验 pH,就不能分配相应的质子化状态,因此也不能计算出由此产生的质子摄取/释放。需要指出的是,虽然实验 pH 可能不是生理 pH,结合可能涉及质子摄取/释放,但天然受体-配体复合物已经朝着特定的亚细胞或组织特征 pH 进化,在该 pH 下质子摄取/释放最小或不存在。