【作者】 贺敏；

【导师】 梁逸曾；

【作者基本信息】 中南大学 ， 分析化学， 2013， 博士

【摘要】 目前,合成类新药的发现呈放缓趋势,同时一个新的高效低毒药物商品化要花费巨大的财力和时间。天然药物及其制剂已经在全球大部分国家被广泛应用了几千年,有着悠久的历史,并且有着显著的疗效,因而成为了各国学者研究的热点。但是,天然药物的成份复杂,是一个“黑色分析体系”,如何更有效地分离、鉴定天然药物有效成份,并进一步进行药理毒理筛选及机制评估是我们应该解决的问题。本文以止咳平喘类天然药物为研究对象,详细讨论了使用化学计量学方法用于天然药物复杂体系解析和毒理研究。一,将化学计量学分辨方法应用于天然药物LC-DAD、GC-MS和UPLC-MS数据的纯色谱曲线、纯光谱和纯质谱的获取。这些方法包括平滑、手动线性扣除、自适应迭代加权惩罚最小二乘、直观推导式演进特征投影法、交互移动窗口因子分析和选择性离子分析等,这些方法在处理复杂成份分析时具有较高的应用价值。同时,三种化学计量学解析方法的共同点和不同点通过一些实验数据得以阐述。二,程序升温保留指数被应用于挥发油成份的进一步鉴定；一个由质谱特征和等效链长建立的特殊保留指数数据库用于脂肪酸的定性；一个定量结构-色谱保留指数相关模型被建立用来预测搜索结果中没有NIST保留指数的化合物,并通过文献验证。三,准确的质量测定通过两种方法获取,一种通过LC-QTOF-MS等高分辨仪器获取,在准确的质量测定后,在建立的天然药物单体数据库中进行搜索定性,可以锁定主要成份,对于同分异构体,根据参考文献和化合物结构判断流出顺序；第二种是数学处理的方法,我们使用Origin软件进行分子离子峰或关键碎片同位素结构解析和高斯拟合,同时使用外部校正方法进行校正,这种方法能够区别NIST MS库大量搜索结果中具有不同分子量的化合物。四,一种通过化合物-蛋白质相互作用的方法被应用去评估升压机制。我们的方法假设不同升压机制的化合物应该结合到不同的靶标蛋白,因此不同的机制可以通过化合物-蛋白质相互作用予以区分。首先,与血压升高有关的天然药物成份和靶标蛋白被收集,使用一个随机森林模型计算化合物-蛋白相互作用概率,然后根据参考文献和其它方法判断可信度。从一个热图、化合物-蛋白相互作用网络图,可以清晰地观察到天然药物成份和不同的靶标蛋白的相互作用关系。最终,使用主成份分析对这些预测概率进行处理,这些升压靶标能够划分为三个大的区域。本文探索了使用化合物-蛋白质相互作用进行药理(毒理)机制分类的可行性,这个方案同时也适用于未知化合物的药理(毒理)机制推导。图38幅,表17个,参考文献341篇。更多还原

【Abstract】 Currently, the discovery of new synthetic drugs has shown a trend of slowing down, and the new drugs will spend a lot of money and time before the commercialization. Natural medicines and their preparations have been widely used for thousands of years in most countries around the world, which has a significant effect. Therefore, natural medicines have become a hot research topic. However, natural medicine is a "black analysis system", we should solve the separation and identification problem of the active ingredients came from natural medicines, even toxicology screening and evaluation. In this study, a detailed discussion was done based on antitussive natural medicine, which involves the use of chemometric methods used for complex analytical system and toxicological research.1. Chemometric resolution methods were used in the natural medicine data came from high performance liquid chromatography-diode array detector (HPLC-DAD), gas chromatography-mass spectrometry (GC-MS), ultra performance liquid chromatography-mass spectrometry (UPLC-MS), and pure chromatographic curve, pure UV spectra and pure mass spectrometry were obtained. These chemometrics methods include smoothing and filtering, ordinary manual linear deduction, adaptive iteratively reweighted penalized least squares (airPLS), heuristic evolving latent projections (HELP) and alternative moving window factor analysis (AMWFA), selective ion analysis (SIA) and so on. These methods possess practical value in laboratory when facing complicated components analysis. Simultaneously, the common and different features among HELP, SIA and AMWFA were compared by using some experimental data.2. Temperature-programmed retention indices (PTRIs) were applied in the further identification of chemical composition from the essential oils; the equivalent chain length (ECL), fraction chain length (FCL), an established special retention indices library integrated with mass spectrometry were also applied to further identify the composition of fatty acids including total fatty acids, esterified fatty acids, free fatty acids; In addition, a quantitative structure-retention relationship (QSRR) model with good predictive ability was established and the in-silico RI was applied in qualitative identification combined with NIST MS library search results. Candidate compounds were found to have a moderate matching between the predicted RI values against the experimentally determined values, and incorrect formulas were excluded.3. Accurate mass determination was obtained through two methods. The first method is high resolution instruments, such as liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QTOF-MS). After accurate mass determination, an established monomer search database was used for qualitative, the main ingredients could be locked, and the results showed that this method is effective and feasible. For some isomer, we could determine the peak elution order using some references as well as the chemical structure analysis; the second method is mathematical process for the data came from low resolution instruments, we used a simple external calibration method to obtain the accurate mass of molecular ion or key fragment, which included overlapped isotopes structures resolution and Gaussian fitting using Origin software. The calibration method was able to distinguish different molecular weights among a large number of known NIST MS library search results.4. A method was applied to evaluate pressor mechanisms through compound-protein interactions. Our method assumed that the compounds with different pressor mechanisms should bind to different target proteins, and thereby these mechanisms could be differentiated using compound-protein interactions. Phytochemical components and tested target proteins related to blood pressure (BP) elevation were collected. Then, in silico compound-protein interactions prediction probabilities were calculated using a random forest model, which have been implemented in a web server, and the credibility was judged using related literature and other methods. Further, a heat map was constructed, it clearly showed different prediction probabilities accompanied with hierarchical clustering analysis results. Followed by a compound-protein interaction network was depicted according to the results, we can see the connectivity layout of phytochemical components with different target proteins within the BP elevation network, which guided the hypothesis generation of poly-pharmacology. Lastly, principal components analysis (PCA) was carried out upon the prediction probabilities, and pressor targets could be divided into three large classes. This work explored the possibility for pharmacological or toxicological mechanism classification using compound-protein interactions. Such approaches could also be used to deduce pharmacological or toxicological mechanisms for uncharacterized compounds.更多还原