建立一多層級╱多物理之理論計算平台以對蛋白質機器進行有預測效力的模擬

Abstract

在⽣生物系統中，蛋⽩白質機器通常以模組化、︑､多區塊的結構形式進⾏行功能．即便在不斷有熱 擾的影響之下，這些由蛋⽩白質分⼦子組成的機器還是能準確地完成⾮非常複雜的功能，如基因表現 和檢測與修復細胞內巨分⼦子的結構等． ⽽而蛋⽩白質各區塊的結構及序列如何控制模組間互動的構 形動態以達到⽣生物功能尚處於未知階段．除受限於實驗⽅方法對蛋⽩白質結構動態的測量通常僅能 對單⼀一模組進⾏行了解外， 理論計算也尚未能對由許多蛋⽩白質區塊所組成的複雜分⼦子結構進⾏行有 效模擬．其主因在於傳統上以全原⼦子模型為主的分⼦子模擬，常使描述多模組蛋⽩白質機器所需的 計算量超出現有電腦資源的負荷．在沒有計算理論的輔助下，開展新的實驗⽅方法以測量蛋⽩白質 機器之構形動態及其與⽣生物功能間的關係也受到限制． 有鑒於此，本計劃將開展⼀一套介觀層級計算⽅方法，混熱擾流⼒力與分⼦子動⼒力(hybrid FHD/MD)，以有效模擬多模組蛋⽩白質機器的構形動態．Ｈybrid FHD/MD 的主要創新在於⽤用粒 ⼦子的分⼦子⼒力學描述蛋⽩白質構形，⽽而場量的場量⼒力學描述溶液相的流體⾏行為，溶質與溶液作⽤用及 電動現象．此外，我們也將建⽴立ㄧ套多層級模擬理論，使 hybrid FHD/MD 能與全原⼦子模型及 其較精細的物理描述進⾏行耦合，達到準確建模的⽬目的．因此，全原⼦子模型在介觀層級層級的表 象將以hybrid FHD/MD 形式表達，並能以此較有計算效率的模型對蛋⽩白質機器的構形動態做 有預測⼒力的模擬．Ｈybrid FHD/MD 與其相應的多層級模擬理論將應⽤用于模擬纖維素酶與⾮非粒 腺體胜鏈合成酶等分⼦子機器的模組間互動．由於本計畫推動的理論計算架構包含了與實驗測量 緊密結合的獨特內涵。︒｡為此，我們將與時變單分⼦子福斯特共振能量轉移 (time-dependent single-molecule Förster resonance energy transfer) 專家，美國普林斯頓⼤大學楊皓教授合作，以驗 證理論模擬之預測並解出纖維素酶與⾮非粒腺體胜鏈合成酶模組間構形動態的機理，以及探討模 組間構形動態是否能經由與⼩小分⼦子作⽤用以調控蛋⽩白質機器的功能． 我們的⾧長遠⽬目標是能以電腦 計算加速⽣生物和醫學研究以減少藥物發展之耗時與試錯。

Protein machines in biology rely on modular structures and multiple components to carry out sophisticated processes such as regulation of gene expression, transcribing and translating genomic data, and harvesting and utilization of energy. However, what is the structure and sequence design allowing different biomolecular units to coordinate their activities in the presence of stochastic noises remains elusive. Not only the conformational dynamics of protein machines are difficult to measure, there is lack of effective methods in simulation and theory for modeling the collective behaviors of protein domains. In this work, we propose to breakthrough this limitation by the complementary development of a simulation method and a computational theory. On one hand, we will establish a hybrid fluctuating hydrodynamics and molecular dynamics (hybrid FHD/MD) approach for effective simulation of the conformational dynamics of protein complexes. On the other, a multiscale computational framework will be devised to bridge molecular and field mechanics for allowing the mixed representation in the hybrid FHD/MD model to interface and communicate with simulations at a finer-scale, typical with all-atom force fields. The development of the hybrid FHD/MD simulation method and the conjugate framework of multiscale coarse gaining aims two main objectives. The first is to make accessible the dynamics and statistics at the mesoscopic scale that are impractical to obtain with atomic models. The second is to enable the hybrid model to be used as a mesoscopic-scale theory to bridge the phenomenological and atomic scales of biology. This platform is specially designed to simulate the functional dynamics of protein machines as well as to trace the molecular origin of emergent properties, such as the free-energy landscapes and diffusion dynamics of the conformational changes in protein machines. The application systems of our development of simulation and theoretical methods will focus on the family 7 cellobiohydrolase I of Trichoderma reesei and the nonribosomal peptide synthetases of Bacillus subtilis. For both protein machines, we aim to elucidate the mechanism of their inter-domain coupling via applying the simulation method of hybridizing hydrodynamics and molecular dynamics and the theory of bridging molecular and field mechanics. Validation of the mechanism and theoretical predictions of their conformational dynamics will be conducted by comparing with the experimental data obtained by single-molecule Förster-type resonance energy transfer methods, biochemical assays, and the related molecular biology tools through collaboration. It is also important to emphasize that the proposed methodology development and research agenda have direct relevance to medicinal and biotechnological applications. Furthermore, the focused topic concerns the behaviors of a complex system at the mesoscopic scale. Therefore, the impact of the proposed framework of multiscale modeling and simulation may go beyond chemical and life sciences to other mesoscopic-scale problems in complex systems.