【作者】 刘超；

【导师】 林世贤；

【作者基本信息】 浙江大学 ， 细胞生物学， 2023， 博士

【摘要】 蛋白质的翻译后修饰普遍存在于生物体中,参与体内的各种生命过程,调控蛋白质的结构、功能、稳定性、细胞定位和相互作用。目前,研究发现细胞中存在有400多种不同的蛋白质翻译后修饰类型,包括磷酸化、乙酰化、脂质化、甲基化、泛素化和糖基化等。由于蛋白质的动态性质以及获得均质蛋白技术的局限性,研究特定蛋白质翻译后修饰的生物学功能仍具有极大的挑战性。利用遗传密码扩展技术可以在细菌、酵母、哺乳动物细胞以及活体动物的蛋白质上位点特异性地引入携带天然修饰或天然修饰类似物的翻译后修饰非天然氨基酸,从而获得具有位点选择性修饰的均质化蛋白,实现特定翻译后修饰蛋白质的功能研究。本论文中,我们首先利用遗传密码扩展技术研究组蛋白的甲基化修饰,将富电子的色氨酸类衍生物定点引入到组蛋白读码结构域的关键色氨酸位点。通过结合亲和力检测实验,发现6-甲氧基取代色氨酸的引入能够将读码结构域与组蛋白H3K4me3的结合亲和力提高8倍左右。同时,这一现象也能帮助我们更好地理解阳离子-π非共价相互作用的化学识别原理,为进一步提高这种非共价相互作用的结合亲和力的色氨酸衍生物的设计铺平了道路。随后,通过构建设计多价串联的带6-甲氧基取代色氨酸的阅读器,我们开发了一个纳摩级别亲和力的超级读码结构域,实现对H3K4me3信号的高效检测和成像。另外,我们开发了一套计算机辅助虚拟筛选的策略,通过筛选得到了一系列长度可调、可以模拟天然脂化修饰的类似物,并将这些脂质化类似物通过遗传编码引入到大肠杆菌和哺乳动物细胞中的蛋白质的特定位点。我们发现这些脂质化模拟物可以增强目的蛋白与人血清白蛋白的结合亲和力,从而延长疾病治疗蛋白在小鼠体内的半衰期。同时,将这些长度可调的脂质类似物整合到多种功能信号蛋白的特定位点,实现天然的脂质化修饰的模拟,并成功锚定在膜上用于膜的定位与信号转导。因此,我们的策略能在生命系统中精准合成脂化修饰的蛋白质,并能够在不同生理和病理条件下对数百种底物蛋白进行蛋白质脂化的功能获得性研究,为治疗性候选药物的开发提供一种思路。更多还原

【Abstract】 Proteins post-translational modifications(PTMs)are ubiquitous in all living organisms and can participate in various biological processes in vivo,regulating protein structure,function,stability,cell localization and interaction.At present,more than 400 different modification types have been identified,including phosphorylation,acetylation,lipidation,methylation,ubiquitination,and glycosylation.On account of the dynamic property of proteins and the technical limitations for acquiring homogeneous proteins,the investigation of the exact biological function of specific post-translational modifications remains extremely challenging.Unnatural amino acids carrying natural modifications or their analogues can be site-specifically incorporated into the proteins of bacteria,yeast,mammalian cells or whole animals aided by genetic code expansion strategy,so that homogeneous proteins with site-selective modifications can be obtained for the analysis of physical and chemical properties.In this thesis,we genetically incorporated several electron-rich Trp derivatives in a site-specific manner into the tryptophan position of the histone reader,and we observed that the binding affinity of the reader with 6-methoxy-Trp to histone H3K4me3 increased by more than eightfold.This finding helps us to better understand the chemical mechanism of cation-π interactions,and paved the way for the design of other Trp derivatives that could further improve the binding affinity of this noncovalent interactions.Moreover,in combination with the design of multivalent readers,we further improved the binding affinity of engineered PHD domains to H3K4me3 to single-digit binding affinity for efficient detection and imaging of H3K4me3 signals.In addition,we developed a strategy for computational-aided virtual screening of lipid mimics,which then could be site-specifically incorporated into protein in E.coli and mammalian cells by genetic code expansion strategy.Next,we demonstrated that these lipid mimics could be used to enhance the binding affinity of the target protein to human serum albumin,which could be used to extend the half-life of the disease therapeutic protein in living mice.Moreover,these length-tunable lipid analogues can be incorporated into specific sites of various functional proteins to mimic natural lipidation modifications and help them anchor to membranes for localization and signal transduction.In conclusion,our strategy has enabled the precise synthesis of lipid-modified proteins in all living systems and the ability to perform gain-of-function studies of protein lipidation on hundreds of substrate proteins under different physiological and pathological conditions,which lays the foundation for the development of therapeutic drug candidates.更多还原