Department of Physics, University of Texas at San Antonio, San Antonio, TX 78249, USA.
Department of Physics, University of Texas at San Antonio, San Antonio, TX 78249, USA.
Biochim Biophys Acta Biomembr. 2017 Aug;1859(8):1310-1316. doi: 10.1016/j.bbamem.2017.04.022. Epub 2017 Apr 25.
Measuring or computing the single-channel permeability of aquaporins/aquaglyceroporins (AQPs) has long been a challenge. The measured values scatter over an order of magnitude but the corresponding Arrhenius activation energies converge in the current literature. Osmotic flux through an AQP was simulated as water current forced through the channel by kilobar hydraulic pressure or theoretically approximated as single-file diffusion. In this paper, we report large scale simulations of osmotic current under sub M gradient through three AQPs (water channels AQP4 and AQP5 and glycerol-water channel GlpF) using the mature particle mesh Ewald technique (PME) for which the established force fields have been optimized with known accuracy. These simulations were implemented with hybrid periodic boundary conditions devised to avoid the artifactitious mixing across the membrane in a regular PME simulation. The computed single-channel permeabilities at 5°C and 25°C are in agreement with recently refined experiments on GlpF. The Arrhenius activation energies extracted from our simulations for all the three AQPs agree with the in vitro measurements. The single-file diffusion approximations from our large-scale simulations are consistent with the current literature on smaller systems. From these unambiguous agreements among the in vitro and in silico studies, we observe the quantitative accuracy of the all-atom force fields of the current literature for water-channel biology. We also observe that AQP4, that is particularly rich in the central nervous system, is more efficient in water conduction and more temperature-sensitive than other water-only channels (excluding glycerol channels that also conduct water when not inhibited by glycerol).
长期以来，测量或计算水通道蛋白/甘油通道蛋白（AQP）的单通道渗透率一直是一个挑战。测量值的分布范围跨越一个数量级，但相应的阿仑尼乌斯活化能在当前文献中收敛。通过千巴水压迫使水通过通道模拟 AQP 的渗透通量，或者从理论上近似为单分子扩散。在本文中，我们报告了使用成熟的粒子网格 Ewald 技术（PME）对三种 AQP（水通道 AQP4 和 AQP5 以及甘油水通道 GlpF）在亚毫摩尔梯度下的渗透电流进行大规模模拟的结果，其中建立的力场已通过已知精度进行了优化。这些模拟采用混合周期性边界条件实现，旨在避免在常规 PME 模拟中在膜内出现人为混合。在 5°C 和 25°C 下计算出的单通道渗透率与最近对 GlpF 的改进实验结果一致。从我们的模拟中提取的所有三种 AQP 的 Arrhenius 活化能与体外测量结果一致。从我们的大规模模拟中得出的单分子扩散近似值与较小系统的当前文献一致。从这些体外和计算机研究之间的明确一致性中，我们观察到当前文献中用于水通道生物学的全原子力场的定量准确性。我们还观察到，在中枢神经系统中特别丰富的 AQP4 在导水方面比其他仅含水分的通道（不被甘油抑制时也能导水的甘油通道除外）更有效，并且对温度更敏感。
