Thayer Kelly M, Stetson Sean, Caballero Fernando, Chiu Christopher, Han In Sub Mark
College of Integrative Sciences, Wesleyan University, Middletown, CT 06457 USA.
Department of Chemistry, Wesleyan University, Middletown, CT 06457 USA.
Biophys Rev. 2024 Jul 11;16(4):479-496. doi: 10.1007/s12551-024-01207-4. eCollection 2024 Aug.
The tumor suppressor protein p53, a transcription factor playing a key role in cancer prevention, interacts with DNA as its primary means of determining cell fate in the event of DNA damage. When it becomes mutated, it opens damaged cells to the possibility of reproducing unchecked, which can lead to formation of cancerous tumors. Despite its critical role, therapies at the molecular level to restore p53 native function remain elusive, due to its complex nature. Nevertheless, considerable information has been amassed, and new means of investigating the problem have become available.
We consider structural, biophysical, and bioinformatic insights and their implications for the role of direct and indirect readout and how they contribute to binding site recognition, particularly those of low consensus. We then pivot to consider advances in computational approaches to drug discovery.
We have conducted a review of recent literature pertinent to the p53 protein.
Considerable literature corroborates the idea that p53 is a complex allosteric protein that discriminates its binding sites not only via consensus sequence through direct H-bond contacts, but also a complex combination of factors involving the flexibility of the binding site. New computational methods have emerged capable of capturing such information, which can then be utilized as input to machine learning algorithms towards the goal of more intelligent and efficient de novo allosteric drug design.
Recent improvements in machine learning coupled with graph theory and sector analysis hold promise for advances to more intelligently design allosteric effectors that may be able to restore native p53-DNA binding activity to mutant proteins.
The ideas brought to light by this review constitute a significant advance that can be applied to ongoing biophysical studies of drugs for p53, paving the way for the continued development of new methodologies for allosteric drugs. Our discoveries hold promise to provide molecular therapeutics which restore p53 native activity, thereby offering new insights for cancer therapies.
Structural representation of the p53 DBD (PDBID 1TUP). DNA consensus sequence is shown in gray, and the protein is shown in blue. Red beads indicate hotspot residue mutations, green beads represent DNA interacting residues, and yellow beads represent both.
肿瘤抑制蛋白p53是一种在癌症预防中起关键作用的转录因子,它与DNA相互作用,作为在DNA损伤情况下决定细胞命运的主要方式。当它发生突变时,会使受损细胞有不受控制地繁殖的可能性,进而可能导致癌性肿瘤的形成。尽管其作用至关重要,但由于其性质复杂,恢复p53天然功能的分子水平疗法仍然难以捉摸。然而,已经积累了大量信息,并且出现了研究该问题的新方法。
我们考虑结构、生物物理和生物信息学方面的见解及其对直接和间接读出作用的影响,以及它们如何有助于结合位点识别,特别是那些低一致性的结合位点。然后我们转而考虑药物发现计算方法的进展。
我们对与p53蛋白相关的近期文献进行了综述。
大量文献证实,p53是一种复杂的变构蛋白,它不仅通过直接氢键接触的共有序列来区分其结合位点,还通过涉及结合位点灵活性的复杂因素组合来区分。新的计算方法已经出现,能够捕捉此类信息,然后可将其用作机器学习算法的输入,以实现更智能、高效的从头变构药物设计目标。
机器学习与图论和扇区分析的最新进展有望推动更智能地设计变构效应物,这些效应物或许能够恢复突变蛋白的天然p53-DNA结合活性。
本综述揭示的观点构成了一项重大进展,可应用于正在进行的p53药物生物物理研究,为变构药物新方法的持续发展铺平道路。我们的发现有望提供恢复p53天然活性的分子疗法,从而为癌症治疗提供新的见解。
p53 DNA结合结构域(PDBID 1TUP)的结构表示。DNA共有序列以灰色显示,蛋白质以蓝色显示。红色珠子表示热点残基突变,绿色珠子代表与DNA相互作用的残基,黄色珠子代表两者。