Targeting MYC G-quadruplexes by virtual screening of their dynamic ensembles for drug development

富含鸟嘌呤的DNA序列可以形成四链G-四链体（G4），其存在于癌基因的启动子区域中。例如，G4形成下调了突出的MYC癌基因的转录，从而抑制了许多癌细胞中不可或缺的Myc蛋白。因此，药物诱导的MYC G4稳定已被提议作为未来的癌症治疗。大多数G4相互作用化合物堆叠在刚性G-核心上，并且该结合口袋由柔性环和侧翼序列的不同组合补充，对于每个G4序列几乎是唯一的。这种结合口袋的巨大可变性允许开发针对MYC G4的高选择性药物。原则上，计算对接可以通过虚拟筛选来识别有前途的小分子，基于它们特定的分子相互作用。然而，虚拟筛选的可靠性是有限的G4，因为传统的方法不能正确地捕捉其动态结合位点的构象景观。因此，感兴趣的化合物被忽略，并出现假阳性。在这个项目中，我想通过一种改进的MYC G4特异性虚拟筛选方法来克服这一限制，该方法可以确定并解释结合位点的动态性质。这将通过G4动态系综的首次测定来实现，G4动态系综代表柔性结合位点的相关构象。通过核磁共振光谱法测定残留偶极耦合（RDC）将以原子分辨率测量两种选定的MYC G4的动力学。将通过分子动力学计算对MYC G4的可能构象进行采样，并根据RDC数据选择动态系综的构象异构体。随后，将检查在虚拟筛选中使用G4动态系综，以找到新的，高选择性的配体。最后，在体外和体内的测试中确定的化合物将验证他们的选择从虚拟筛选，并确定其价值为未来的研究。

Guanine-rich DNA sequences can form the four-stranded G-quadruplexes (G4s), which are found in promoter regions of oncogenes. For example, the G4 formation down-regulates the transcription of the prominent MYC oncogene and thereby suppresses the Myc protein, which is indispensable in many cancer cells. Therefore, drug-induced stabilization of the MYC G4s has been proposed as future cancer therapy. Most G4-interactive compounds stack upon the rigid G-core and this binding pocket is complemented by different combination of flexible loop and flanking sequences, almost unique for each G4 sequence. This enormous variability of binding pockets allows the development of highly-selective drugs for MYC G4s. In principle, computational docking can identify promising small molecules through virtual screening, based on their specific molecular interactions. However, the reliability of virtual screening is limited for G4s since conventional approaches cannot correctly capture the conformational landscape of their dynamic binding sites. Thus, interesting compounds are overlooked and false-positives occur. In this project, I want to overcome this limitation with an improved MYC G4-specific virtual screening approach that determines and accounts for the dynamic nature of binding sites. This will be accomplished by the first determination of G4 dynamic ensembles, which represent the relevant conformations of the flexible binding sites. The determination of residual dipolar couplings (RDCs) by nuclear magnetic resonance spectroscopy will measure the dynamics of two selected MYC G4s with atomic resolution. Possible conformations of the MYC G4s will be sampled by molecular dynamics calculations and the conformers of the dynamic ensembles will be selected based on the RDC data. Subsequently, the use of the G4 dynamic ensembles in virtual screenings will be examined to find novel, highly-selective ligands. Finally, in vitro and in vivo tests of the identified compounds will validate their selection from virtual screenings and determine their value for future studies.