Investigation of Protein-DNA recognition using Molecular Dynamics and advanced sampling simulations

蛋白质-DNA相互作用对于DNA的复制和转录以及对于所有基因调控和DNA修复过程具有根本重要性。该项目的目的是使用计算机模拟来实现对间接DNA识别的更精确的理解，这是由序列依赖的DNA变形性决定的。此外，我们还希望了解蛋白质-DNA结合的过程，通过“滑动”沿着DNA的分子细节，从而分析DNA变形性的作用。在第一个项目资助阶段，我们成功地开发了一个序列依赖性DNA灵活性的近定量模型。该多模态伊辛模型包括DNA的可能的半稳定构象状态的分布和这些半稳定状态中的每一个中的构象的分布。该方法允许给定的DNA序列和给定的变形结构来计算变形的相关自由能。除此之外，还可以在原子分辨率下模拟结合在DNA小沟中的配体的滑动过程，并使用先进的采样方法研究蛋白质-DNA系统的滑动过程。在第二个资助阶段，我们计划改进和应用多模态伊辛模型系统地计算已知蛋白质-DNA复合物的间接DNA识别。此外，蛋白质-DNA结合的变构效应将通过蛋白质与邻近DNA位点的DNA结合来研究，这近年来越来越成为实验研究的焦点。在这里，我们期望具有显著范围的效果，特别是对于在其末端受到有限流动性的DNA片段，因为这里的整体曲率和整体扭曲是恒定的。也就是说，曲率或扭曲（或其他螺旋变量）的局部变化必然导致其他地方曲率或扭曲的耦合（反向）变化，这可以影响第二蛋白质结合。具有受限迁移率的DNA区段经常出现在包装的DNA中（例如在真核生物中），并且可以实现变构效应。在第二个项目阶段，我们还将使用先进的采样方法来研究配体在DNA凹槽中滑动过程的动力学，以及它是如何被DNA变形调制的。

Protein-DNA interactions are of fundamental importance for the replication and transcription of DNA as well as for all gene regulation and DNA repair processes. The aim of the project is to use computer simulations to achieve a more precise understanding of indirect DNA recognition, which is determined by the sequence-dependent DNA deformability. In addition, we also want to understand the process of protein-DNA binding by “sliding” along the DNA in molecular detail and thereby analyze the role of DNA deformability. In the first project funding phase, we managed to develop a near quantitative model of sequence-dependent DNA flexibility. This multimodal Ising model includes both the distribution of possible semi-stable conformational states of DNA and the distribution of conformations in each of these semi-stable states. The method allows for a given DNA sequence and given deformed structure to calculate the associated free energy of the deformation. In addition to this development, it was also possible to simulate at atomic resolution the sliding process for a ligand that binds in the DNA minor groove and to investigate the sliding process for a protein-DNA system using advanced sampling methods. In the second funding phase, we plan to improve and apply the multimodal Ising model to systematically calculate indirect DNA recognition for known protein-DNA complexes. Furthermore, allosteric effects of protein-DNA binding will be investigated through protein binding to DNA near a neighboring DNA site with a bound protein, which have increasingly been the focus of experimental studies in recent years. Here we expect effects with a significant range, especially for DNA segments that are subject to limited mobility at their ends, since the overall curvature and the overall twist are constant here. That is, a local change in curvature or twist (or other helical variables) necessarily causes a coupled (reverse) change in curvature or twist elsewhere, which can influence a second protein binding. DNA segments with restricted mobility occur frequently in packaged DNA (e.g. in eukaryotes) and can enable allosteric effects. In the second project phase we will also use advanced sampling methods to investigate the kinetics of the sliding processes of Ligands in DNA grooves and how it is modulated by DNA deformation.