Biomolecular Binding Mechanisms at High Temporal and Structural Resolution

Biomolecular binding is pervasive in biology and governs all biological process, including DNA replication and immune recognition. Determining binding mechanisms at the atomistic-level is a significant challenge due to experimental and computational limitations. All-atom molecular dynamics simulations in conjunction with weighted ensemble (WE)-enhanced sampling has shown to be capable of generating binding trajectories in which mechanisms can be curated. In this thesis, molecular dynamics simulations in coordination with WE were implemented to determine the binding mechanism of three systems: (1) Hoechst 33258 (H33258) binding to DNA, (2) daunomycin binding to DNA, and (3) a T cell receptor (TCR) binding to a peptide/major histocompatibility complex (pMHC).

We curated a binding mechanism using 2562 WE-simulated trajectories of H33258 binding to a DNA duplex. Specifically, the initial contacts between H33258 and DNA are driven electrostatically by H33258's positively charged group and the DNA's negatively charged backbone. Following the initial contacts, H33258 forms a hinge-like intermediate with one end in the minor groove of DNA. After hinge state formation, the other end of H33258 swings into the minor groove in a concerted motion and the spine of hydration along the minor groove is dehydrated.

We simulated 469 trajectories of daunomycin binding to a DNA duplex. Throughout the binding process, the DNA strand underwent structural changes, including DNA base pair rise, bending, and minor groove width changes. Post-intercalation, most binding trajectories needed an additional one to five nanoseconds of rearrangement to achieve the bound configuration.

75 binding trajectories of a TCR binding to a pMHC were compiled into a binding mechanism. The initial phase of the binding process consists of contacts being formed, some forming hydrogen bonds. Hydrogen bond lifetimes vary; some persist throughout binding and others are transient. The final phase of the binding process involves water exiting the interface as final contacts form.

生物分子间的结合在生物学中无处不在，并调控着所有生物过程，包括DNA复制和免疫识别。由于实验和计算的限制，在原子水平上确定结合机制是一项重大挑战。结合轨迹的生成，通过原子分子动力学模拟与加权集合（WE）增强采样相结合，已被证明能够揭示结合机制。在本论文中，通过分子动力学模拟与WE的协同作用，对三个系统的结合机制进行了研究：(1) Hoechst 33258（H33258）与DNA的结合，(2) daunomycin与DNA的结合，以及(3) T细胞受体（TCR）与肽/主要组织相容性复合物（pMHC）的结合。我们利用2562个WE模拟的H33258与DNA双链的结合轨迹，对结合机制进行了梳理。具体而言，H33258与DNA的初始接触由H33258的正电荷基团和DNA的负电荷骨架静电驱动。在初始接触之后，H33258形成了一种铰链状中间体，一端位于DNA的次级沟中。在铰链状态形成后，H33258的另一端协同运动进入次级沟，并导致沿次级沟的水合脊脱水。我们对daunomycin与DNA双链的结合进行了469个轨迹的模拟。在整个结合过程中，DNA链经历了结构变化，包括碱基对升高、弯曲和次级沟宽度变化。在嵌入后，大多数结合轨迹需要额外的1至5纳秒的重新排列才能达到结合构型。我们收集了75个TCR与pMHC结合的轨迹，以梳理结合机制。结合过程的初始阶段包括接触的形成，其中一些形成了氢键。氢键的寿命各异；有些在结合过程中持续存在，而有些则是短暂的。结合过程的最终阶段涉及水从界面逸出，最终接触形成。