Quantitatively predictive biology of aquaporins 1, 5, and GlpF

点击翻译按钮获取中文摘要

Project Summary The state-of-the-art high performance computing enables researchers to simulate the motions of millions of atoms interacting with one another. Now it is feasible to produce quantitative predictions of biological functions of a protein that are “deterministic” out of the atomistic interactions and motions that are stochastic in nature. In this project, the researchers propose to study the functions of two human aquaporins and look for ways to modulate/inhibit them. They will build the aquaporins and their biological environments from atoms up, simulate their stochastic dynamics, and elucidate their deterministic functional behaviors under various controllable conditions. Specifically, they aim to find inhibitors of two water channels (AQP1 and AQP5) and one glycerol channel (GlpF) by accurately quantifying the binding affinities of dozens of candidate inhibitors. The PI developed a new method, the hybrid steered molecular dynamics (hSMD) method, for the purpose of this project and related research. Using hSMD, the researchers will be free from the problem of systematic error amplifications inherent in the current methods of the literature. 5% errors in the input will translate into 5% errors in the final results for binding affinities. They will be able to take full advantage of the high resolution protein structures and the mature CHARMM force field parameters. They will harness the massively parallel computing power of the day to improve the currently investigated candidate inhibitors and to find new inhibitors in a quantitatively predictive manner. Upon completion of the project, two types of aquaporin inhibitors, the extracellular channel entry blockers and the deep channel cloggers, will be ready for clinical trials as drugs for treatment of hypertension, refractory edema, and elevated airway mucus secretion during anesthesia.