Shen J
Hoechst Marion Roussel, Inc., Cincinnati, Ohio 45215-6300, USA.
J Med Chem. 1997 Aug 29;40(18):2953-8. doi: 10.1021/jm9606958.
A tight-binding thermolysin inhibitor, Cbz-Phe-psi[PO2NH]-Leu-Ala (ZFpLA, Ki = 0.068 nM), and its analogs, ZRp(O)LA (R = Ala, Leu or Phe) have been studied using the finite difference solution to the linearized Poisson-Boltzmann equation (FDPB) and solvation entropy correction (SEC). The binding energy difference between conformationally different thermolysin inhibitors ZFpLA and ZGpLL is estimated using three approaches. Two of approaches use the X-ray structures of ZFpLA-thermolysin and ZGpLL-thermolysin structures. The third one uses both X-ray structures to calculate binding energy differences from ZFpLA and ZGpLL to a hypothetical intermediate MepLA. All the results are qualitatively correct with one closely reproducing the experimental value. The enhancement of the ZFpLA binding is attributed largely to the solvation entropy or "hydrophobic force". The binding mode of the ZGpLR N-terminal moiety appears to be electrostatically unfavorable. Reducing the polarity of that moiety is predicted to enhance binding affinity. The binding trends due to the hydrophobic variation of ZRp(O)LA are calculated within 1 kcal/mol of the experimental values. Increasing lipophilicity of a ligand favors the binding due to the difference of surface area change between the free state and the bound state. The analysis of energetic components shows that these trends are not specific for the binding of phosphorus-containing inhibitors but are generally true for protein-ligand interactions. The electrostatic calculation does not support the involvement of the second protonation of ZFpLA in binding. Therefore, reexamining the second protonation of ZFpLA or seeking further experimental support seems appropriate. The structural sensitivity of the FDPB calculation was assessed by using ligand and receptor structures from different X-ray studies of thermolysin. The small deviations (< 0.3 A) in the receptor structures do not cause significant changes in electrostatic binding energy if there is no structural change in modified regions.
一种紧密结合的嗜热菌蛋白酶抑制剂Cbz-Phe-ψ[PO₂NH]-Leu-Ala(ZFpLA,Ki = 0.068 nM)及其类似物ZRp(O)LA(R = Ala、Leu或Phe)已通过线性化泊松-玻尔兹曼方程的有限差分解(FDPB)和溶剂化熵校正(SEC)进行了研究。使用三种方法估计构象不同的嗜热菌蛋白酶抑制剂ZFpLA和ZGpLL之间的结合能差异。其中两种方法使用ZFpLA-嗜热菌蛋白酶和ZGpLL-嗜热菌蛋白酶结构的X射线晶体结构。第三种方法使用这两种X射线晶体结构来计算从ZFpLA和ZGpLL到假设中间体MepLA的结合能差异。所有结果在定性上都是正确的,其中一个结果与实验值非常接近。ZFpLA结合的增强主要归因于溶剂化熵或“疏水作用力”。ZGpLR N端部分的结合模式在静电方面似乎不利。预计降低该部分的极性会增强结合亲和力。ZRp(O)LA疏水变化引起的结合趋势在实验值的1 kcal/mol范围内计算得出。由于游离态和结合态之间表面积变化的差异,配体亲脂性的增加有利于结合。能量成分分析表明,这些趋势并非含磷抑制剂结合所特有的,而是蛋白质-配体相互作用的普遍规律。静电计算不支持ZFpLA的第二次质子化参与结合。因此,重新审视ZFpLA的第二次质子化或寻求进一步的实验支持似乎是合适的。通过使用来自嗜热菌蛋白酶不同X射线研究的配体和受体结构,评估了FDPB计算的结构敏感性。如果修饰区域没有结构变化,受体结构中的小偏差(< 0.3 Å)不会导致静电结合能发生显著变化。