【作者】 虞科；

【导师】 程翼宇；

【作者基本信息】 浙江大学 ， 药物分析学， 2007， 博士

【摘要】 药物分析学受当前医药及生命科学急速发展的驱动,研究领域大大拓展,已从经典的药品质量检验逐渐覆盖了整个医药分析技术应用范畴,并从药物组成分析延伸至体内物质分析。与此同时,药物分析学与信息科学交叉综合加速,学科发展空间进一步扩大。在此背景下,药物分析学已出现以多学科集群创新为特征的发展态势,这将给学科发展带来重大挑战和难得机遇。据此,本论文围绕中药等复杂物质体系分析和生物体内代谢物组分析等研究对象,将色谱及联用分析技术与计算智能技术等相结合,创新发展了一系列基于药物分析信息学的医药色谱分析新方法,其主要学术贡献有:1、针对中药毛细管电泳(CE)分离条件凭经验难以选择问题,创建了基于遗传算法、粒子群优化算法和模拟退火技术等三种进化算法的CE分离条件优化计算方法,可在较少试验次数下获得最佳分离条件,从而解决了复杂中药CE分离条件的优化问题,并已成功应用于中药制剂一双丹颗粒中多种活性成分毛细管胶束电动色谱(MEKC)的同时测定。该方法可用于快速建立中药多成分的MEKC测定方法。2、CE是蛋白质、多肽、核酸及其他生物分子分离和分析的重要技术之一,CE淌度作为生物多肽分子的固有物化属性,已成为多肽乃至蛋白质分析中的重要参数。本文将多肽分子描述符与机器学习算法相结合,创建了生物多肽分子CE淌度的非线性预测方法,已成功应用于预测102个生物多肽分子的CE淌度,预测准确度达到90%以上。此法可实际指导多肽分子CE分离,大幅缩减分析试验的次数,具有预测精度高、稳定性好等优点。3、研究建立了测定血样中多种人参皂苷浓度的HPLC/MS分析方法,已应用于测定参麦注射液静脉给药后家兔血浆中5种人参皂苷的药时曲线。研究发现二醇型人参皂苷在血中衰减较为缓慢,而三醇型人参皂苷在血中衰减较快。这一结果提示:虽然二醇型和三醇型人参皂苷的化学结构较为相似,但是这两类化合物可能存在较大的药代动力学差异。4、代谢组学(Metabonomics)是近年发展起来的一门新兴学科,它利用高通量、高灵敏度与高精确度的仪器分析技术和信息技术,可对机体内源性代谢物组进行“全景式”分析,是药物分析学研究领域的学科前沿。核磁共振(NMR)和液质联用(LC/MS)技术是当前代谢组学研究的两大主流技术,但由于机体代谢物组的多样性和复杂性,目前尚未能实现代谢物组的全谱无偏分析,企待拓展技术手段。本文研究创建了生物体内代谢物组气质联用(GC/MS)计算分析方法。运用所建的代谢物组计算分析方法,取得了两点突破:(1)在世界上首次研究发现了乙型病毒(HBV)所致肝衰竭病人的血清代谢物组特征谱。通过多元统计分析和模式识别技术研究发现肝衰竭病人和正常人的血清代谢物组存在显著的模式差异,而且还发现了不同病情(程度)肝衰竭患者的血清代谢组特征谱。这一重要发现将为肝病的临床药物治疗学研究开辟了新思路。(2)研究了在对乙酰氨基酚和乌头碱不同作用时间和不同给药剂量作用下的小鼠/大鼠尿液代谢物组特征的动态变化模式,初步揭示了在外源性物质干预下受试动物尿液代谢物组的随动变化规律。更多还原

【Abstract】 Pharmaceutical analysis has being boosted by the rapid development of medical and life sciences,as well as the related technologies.The research fields of pharmaceutical analysis have been expanded from the classic quality monitoring of drug products to the development of analytical techniques for the medical and pharmaceutical study.The focus has also been transformed from drug composition into body fluids.Furthermore,the accelerated intersection with information science further enlarged the fields of pharmaceutical analysis.Therefore,the innovation on the basis of integration across multiple disciplines,changing the traditional point of view on pharmaceutical analysis,has been presented,which bring tremendous challenges and opportunities for pharmaceutical analysis researchers. In this dissertation,aiming to the analysis of complex systems such as traditional Chinese medicines and bio-fluids,a series of novel methods,integrated with modern chromatographic technologies,computational intelligence and metabonomics are developed.The main innovation and academic contributions of these works are presented as follows:1.An approach to the optimization of Capillary electrophoresis(CE) separation condition based on multivariate statistics and genetic algorithms was established,and the best condition would be obtained with a lesser number of tests,which solved the issue of the CE condition optimization for the separation of complex chemical system. This method had been successfully applied to optimizing the separation of the active components in the traditional Chinese medicine ’SHUANGDAN’ granule,and an analytical method for the simultaneous determination of the active components such as protocatechuic aldehyde,paeonol,danshensu,salvianolic acid B was established.2.CE has been one important technique for the separation and analysis of proteins,peptides,nucleic acids and other biological molecules,and CE mobility,as the inherent physical and chemical property for peptides,has been an important parameter in the research of peptides and proteins.In this thesis,nonlinear quantitative structure-mobility relationship models based on the molecular descriptors and machine learning algorithms were introduced for predicting the CE mobilities of peptides.The developed nonlinear models had high accuracy and stability,and had been successfully used to predict the CE mobility of a variety of peptides,resulting in the prediction accuracy higher than 90%.The developed model can not only significantly reduce the test number and practically guide the separation of peptides, but also play positive role in promoting the proteomics research.3.Pharmacokinetics of the drug products derived from traditional Chinese medicines is one of the important research areas in the current study on pharmaceutical analysis.The current attractive safety problems of injections made from traditional Chinese medicines are closely related with their pharmacokinetic characteristics.Ginseng is a common-used drug material in traditional Chinese medicines,and ginsenosides are the main active components.In this thesis,a sensitive, specific and reliable HPLC/MS method for the simultaneous determination of ginsenosides Rg1,Rf,Re,Rd and Rb1 in the rabbit plasma was developed.This method was used for the measurement of blood drug concentration-time profiles of five ginsenosides after intravenous administration of ’SHENMAI’ injection,a well-sold injection made from ginseng and ophiopogonis,in rabbits.The result showed that the elimination of diol-ginsenosides in the blood was much more slowly than that of glycerol-ginsenosides,indicating that although the chemical structure of these two types of ginsensides was quite similar,a big difference was presented within their phamacokinetics.4.Metabonomics is the branch of science concerned with the quantitative understandings of the metabolite complement of integrated living systems and its dynamic responses to the changes of both endogenous factors(such as physiology and development) and exogenous factors(such as environmental factors and xenobiotics). NMR and LC/MS were two mainly used techniques in metabonomics;however,there is still lack of a real technique which can fulfill the global analysis of the whole metabonome just with only one run.In this thesis,a novel GC/MS based computational analytical method was developed for the metabonomic study.GC/MS technique and bioinformatics techniques were integrated in this method,which was capable for the metabolites profiling of various biological matrices.The developed metabonomics method was applied to the characterizing the serum metabonome profile of HBV-caused liver failure patients.This discovery indicated that the characteristic profile of serum metabonome is effective to predict the severity of liver failure and its prognosis of drug treatment,which promises to promote the personalized drug therapy of liver disease patients.Furthermore,the developed metabonomics method was also used for the evaluation of toxicity of acetaminophen and aconitine with different exposure times and doses on the mice or rats.The time related changes of metabolites profiling of the animals treated with different level of toxins have been discovered,which would be helpful to develop the more advanced safety assessment systems.更多还原