Assessment of Force Field Accuracy Using Cryogenic Electron Microscopy Data of Hyper-thermostable Glutamate Dehydrogenase

Assessment of Force Field Accuracy Using Cryogenic Electron Microscopy Data of Hyper-thermostable Glutamate Dehydrogenase
复制标题

使用超热稳定谷氨酸脱氢酶的低温电子显微镜数据评估力场精度

DOI:
10.1021/acs.jpcb.0c04464
复制
发表时间:
2020
期刊:
The Journal of Physical Chemistry B
影响因子:
--
通讯作者:
Nakasako Masayoshi
Nakasako Masayoshi
中科院分区:
--
文献类型:
--
作者:
Oroguchi Tomotaka;Oide Mao;Wakabayashi Taiki;Nakasako Masayoshi

文献摘要

相似文献

在生物物理相关的微秒时间尺度内进行分子动力学 (MD) 模拟是研究生物分子过程的强大工具,但结果通常显示出力场依赖性。因此,利用溶液中生物分子的实验数据评估力场精度对于模拟研究至关重要。在这里,我们建议使用通过冷冻电子显微镜(cryoEM)获得的结构模型,它提供了模拟溶液环境的玻璃冰中的生物分子结构。通过将 AMBER(ff99SB-ILDN-NMR、ff14SB、ff15ipq 和 ff15FB)和 CHARMM(CHARMM22 和 CHARMM36m)力场的 MD 轨迹与热稳定六聚谷氨酸脱氢酶 (GDH) 的冷冻电镜数据进行比较来评估其准确性,其中包括分辨率约为 3 的冷冻电镜图。 Å 和亚基的结构模型反映了 GDH 中发生的域运动中的亚稳态构象。在评估中,我们验证了冷冻电镜数据中二级结构和亚基间相互作用的再现性和稳定性的力场。此外,我们评估了关于冷冻电磁数据预期的域运动中能量景观的再现性的力场。结果,在六个力场中,ff15FB和ff99SB-ILDN-NMR与实验表现出良好的一致性。本研究证明了高分辨率冷冻电镜图的优势,并提出了重现实验观察到的蛋白质结构的最佳力场。
Molecular dynamics (MD) simulations in biophysically relevant time scales of microseconds is a powerful tool for studying biomolecular processes, but results often display force field dependency. Therefore, assessment of force field accuracy using experimental data of biomolecules in solution is essential for simulation studies. Here, we propose the use of structural models obtained via cryo-electron microscopy (cryoEM), which provides biomolecular structures in vitreous ice mimicking the environment in solution. The accuracy of the AMBER (ff99SB-ILDN-NMR, ff14SB, ff15ipq, and ff15FB) and CHARMM (CHARMM22 and CHARMM36m) force fields was assessed by comparing their MD trajectories with the cryoEM data of thermostable hexameric glutamate dehydrogenase (GDH), which included a cryoEM map at a resolution of approximately 3 Å and structure models of subunits reflecting metastable conformations in domain motion occurring in GDH. In the assessment, we validated the force fields with respect to the reproducibility and stability of secondary structures and intersubunit interactions in the cryoEM data. Furthermore, we evaluated the force fields regarding the reproducibility of the energy landscape in the domain motion expected from the cryoEM data. As a result, among the six force fields, ff15FB and ff99SB-ILDN-NMR displayed good agreement with the experiment. The present study demonstrated the advantages of the high-resolution cryoEM map and suggested the optimal force field to reproduce experimentally observed protein structures.