节点文献

基于代谢组学技术的糖尿病特征代谢物及维医证型研究

Metabonomics Study of Diabetes Mellitus and Traditional Uygur Medicine Syndrome Based on Different Analytical Methods

【作者】 马晓丽

【导师】 毛新民;

【作者基本信息】 新疆医科大学 , 药理学, 2013, 博士

【摘要】 目的:本文采用1H-NMR、UPLC/TOF-MS作为代谢组学研究的检测手段,以糖尿病患者和正常人的尿液为研究对象,分析了糖尿病患者尿液的特征代谢谱;进一步采用HPLC和GC法的代谢靶标技术定量测定尿液及血清中氨基酸、脂肪酸类代谢物。采用模式识别方法对与糖尿病相关的代谢物质进行全面的研究,从西医诊断到维医辨证分型等不同层面阐述与糖尿病代谢相关的差异物质,为全面理解糖尿病在发病机制上的差异和联系提供了一个新视角和实验基础。方法:1.利用基于NMR的代谢组学方法研究维吾尔族2型糖尿病尿液代谢特征,建立稳定的NMR尿液一维氢谱,分析2型糖尿病人及正常人尿液特征代谢物。根据维医分型将糖尿病患者分为不同体液证型,进一步利用建立的方法寻找不同证型间的差异标志物。2.采用超高效液相色谱-四级杆串联飞行时间质谱(UPLC Q-TOFMS)技术分析糖尿病及正常人尿液的代谢谱,结合多元数据分析寻找糖尿病和正常人的差异物质。根据精确分子量、按照差异代谢物的种类在谱库中进行检索,对其进行系统的鉴定并与NMR结果进行比较。3.建立高效液相色谱(HPLC)分析生物样品中游离氨基酸的代谢靶标分析技术,以维吾尔族糖尿病及正常人血清/尿液为研究对象,测定氨基酸代谢谱,结合多元统计分析评估氨基酸代谢谱对糖尿病发病的影响。进一步将该方法应用于糖尿病小鼠模型中,充分研究糖尿病小鼠的氨基酸代谢变化。4.建立气相色谱(GC)技术检测血清中游离脂肪酸的技术平台,优化了GC色谱的最优条件。用该法分析2型糖尿病患者和正常健康者血清的7种游离脂肪酸,结合多元统计分析寻找脂肪酸代谢谱的差异,对脂肪酸代谢异常进行系统评价。结果:1.建立了1H-NMR的尿液代谢组学方法,糖尿病患者和正常人经代谢组学方法得到很好的区分,并找到了相对应的多个糖尿病标志物。DM组的大部分代谢物浓度较正常人低。结合两组样品的NMR谱图的不同轮廓,与正常人相比,DM患者尿液中的葡萄糖、马尿酸、甜菜碱、牛磺酸等升高;亮氨酸、异亮氨基酸、肌酸酐、柠檬酸、二甲胺等含量较正常人低。根据维吾尔医分型将患者分为不同异常体液证型,进一步利用NMR技术对不同证型糖尿病及健康人尿液进行代谢组学研究,发现三种异常体液糖尿病患者尿液差异物质相似,异常粘液质患者尿液中的代谢差异与其他两类都有显著差异,成为各个异常体液中代谢变化最为显著的体液类型。2.建立了UPLC-Q/TOF-MS的尿液代谢组学研究方法,糖尿病患者尿液中的糖、脂、氨基酸等代谢通路都发生了紊乱,在糖类代谢物中,钠葡萄糖醛酸含量明显高于正常对照组,在脂肪酸类物质中,与正常对照组相比,20-羧基花生四烯酸,棕榈酸在糖尿病患者尿液中比在正常人中高。氨基酸类代谢物中,酪氨酸、N-β-丙氨酰-L组氨酸、色氨酸含量较正常人低,戊二酸、烟酰甘氨酸、天冬氨酸、苯丙氨酸等含量升高。这些代谢物均是与体内糖、脂代谢相关的调节物质,进一步说明了糖尿病患者发生了相关代谢紊乱。1H-NMR技术所得到的标记物和UPLC-Q/TOF得到的标记物大部分不同,但是分组聚类结果类似。3.以反相色谱柱和梯度洗脱对生物样品氨基酸进行HPLC分离。通过色谱条件的优化及对方法学考察,说明选择的衍生化方法重复性好、稳定性高。另外,通过正交偏最小二乘判别分析(OPLS-DA)对HPLC得到的氨基酸数据进行分析。研究发现,HPLC分析得到的血清及尿液氨基酸代谢谱能明显表征出糖尿病患者和正常人的代谢表型差异,有效的区分糖尿病患者和健康对照组的氨基酸代谢轮廓。在糖尿病血清氨基酸代谢靶标分析中,糖尿病患者血清氨基酸代谢谱中对分组贡献最大的是天冬氨酸、谷氨酸、苏氨酸、丝氨酸,而在单维统计分析发现天冬氨酸和苯丙氨酸代谢异常最为明显,较正常人明显升高,而较正常人显著下降的氨基酸有丙氨酸、甘氨酸和赖氨酸。在尿液的代谢谱比较时发现多元分析中得到丝氨酸、苏氨酸、牛磺酸对组间贡献较大,而单维统计分析发现尿液中牛磺酸、精氨酸、丝氨酸和组氨酸较正常人显著下降,谷氨酸显著升高。两种方法处理得到的结果可以互为补充。在糖尿病动物模型氨基酸谱的研究中发现所有氨基酸在单维统计分析中均无差异,而在多元统计分析中,通过PLS-DA负载图可见,精氨酸、苯丙氨酸、赖氨酸、牛磺酸对组间贡献较大,这些氨基酸可能承载着小鼠成模后药理过程中代谢途径变化的重要信息。本研究说明HPLC靶标分析技术可以作为高效和简单的技术平台用于糖尿病的代谢轮廓研究,是一种潜在的早期诊断手段。4.建立了以硫酸/甲醇为甲酯化试剂测定血清中FFA的GC分析方法,同时优化了GC色谱的最优条件。利用多元统计分析结合模式识别方式,结合PCA和OPLS-DA对数据进行模式识别对两组人群进行了有效区分,发现肉豆蔻酸在糖尿病患者血清中升高,而油酸在糖尿病患者中特异性下调,提示这两种物质可能分别为潜在特异性标志物,为区分糖尿病和体内脂肪酸失衡的关系提供了重要的实验依据。结论:1H-NMR技术和UPLC-Q/TOF技术作为常用的代谢组学分析工具,在标记物发现中有一定的互补性。通过对正常人和糖尿病患者尿液的代谢谱研究,发现代谢组学具有很好的疾病诊断作用,进一步将其应用到维医不同证型糖尿病的特异代谢物寻找中,获得了区分不同证型糖尿病的特征代谢物;运用HPLC及GC的代谢靶标分析技术对血清/尿液中的氨基酸、脂肪酸进行定量分析,结合模式识别发现具有良好的聚类效果。代谢组学作为现代生命科学的一个新的分支,对研究糖尿病的潜在标志物研究具有重要价值。

【Abstract】 Objective:Metabonomics is a new method of studying dynamic change of endogenousmetabolite in biosystem and has been increasingly used as an effective tool for diseasediagnosis to elucidate significant changes in metabolism, also an ideal tool for studyingtraditional medicine theories.In our study, With the help of the recent advent oftechnologies for quantitative and comprehensive metabolites analyses, which provides amore sensitive and specific diagnosis than single biomarker approaches, and revealdistinct differences in metabolism between diseased individuals and healthy ones in anovel systems-level.Methods:1.In the present study, NMR was used in the urine metabolomics study, verifyingthe high performance of the system in separating endogenous metabolites and its abilityas a nearly diagnostic method for uygur medicine.2.Through studying the metabolic variations in urine of type2diabetes mellituspatients and healthy volunteers with UPLC-TOF-MS, more disease information could begain to help finding the common and specific biomarkers among the diseases, which gavenew eyes on the understanding of the mechanism of the diseases.Here, T2DM andmetabonomics were investigated by comparing small molecules in urine samples fromT2DM patients and healthy controls, the potential biomarkers of DM were found. 3.A robust and sensitive liquid chromatography spectrometry (HPLC) method wasdeveloped and optimized for quantification of various amino acids in serum and urine.Then the data was imported through the analysis by principal component analysis (PCA),and the least squares discriminant analysis (PLS-DA). The analytical serum amino acidsplatform can applied to investigate the effect of diabetes on amino acids in serum, urineand diabetes rats, Then, results shows the classification accuracy were good.4.To investigate the serum free fatty acid variation in diabetic patients and heathyvolunteers. the human serum were methylesterifiled by H2SO4/CH3OH, extracted byn-hexane, FID detector,7kinds of fatty acids can be separated completely with a goodlinear and chemometric resolution method were used to discriminated diabetes patientsand controls.Results:1.The application of NMR detection has been investigated for the production ofglobal metabolite profiles from human urine, Chemomatrics was performed to detect theperturbation metabolites as many as possible. Furthermore, the potential biomarkers werescreened out, which might be the target components in the future pathogenesis research,as well as predicted model was builded up using PLS-DA. The metabonomics basedNMR and OPLS can prove the difference in differentiated traditional Uygur Medicinesyndrome with diabetes, Eatabilished the approach with NMR for metabonomic study,some specific urine endogenous metabolites changes in the metabolic composition ofurine samples from diabetes, there are Elevated glucose,hippuric acid, betaine, taurine,etc.; leucine, isoleucine, creatinine, citric acid, dimethylamine, and other content is lowerthan normal control.2.A UPLC Q-TOF MS system was employed to distinguish the urine global profilesof38type2diabetes mellitus patients and32heatlty volunteers. PCA and PLS-DA wasused for group differentiation and potential biomarkers selection. As shown in the scoresplot, the distinct clustering among diabetes mellitus patients and health controls wasobserved. In carbohydrate metabolites, sodium glucuronic acid content was significantlyhigher than the normal control group, compared with the normal control group, thesubstance of the fatty acids were higher than control group,20-carboxy-arachidonic acid,palmitic acid in the urine of patients with diabetes were higher than in normal subjects.Metabolites of amino acids such as tyrosine, N-beta-alanyl-L histidine, tryptophancontent is lower than normal, glutaric acid, nicotinoyloxy glycine, aspartic acid, phenylalanine, etc. content were increased than normal people. These metabolites areassociated with in glucose and lipid regulating substances.This study indicatedthepotential of metabolomic strategies for explanation of the mechanism of diseases anddetection of specifical biomarkers.3.The HPLC based serum and urine amino acid metabolomics approach was alsoused to investigate the pathophysiological variation among type2diabetes mellituspatients compared to health control. The derivatization condition was optimized. Patternrecognition was carried out with PCA and PLS-DA.Then, the serum of type2diabetesrats were analysis by this method.In diabetes, the amino acid metabolism includingaspartate,glutamic, serine, threonine were found as the most obvious changed aminoacid.although aspartic acid and phenylalanine were significantly higher than normal by ttest, a significant decline in the amino acid are alanine, glycine and lysine. Metaboliteprofiles in the urine was analysised in a multivariate analysis and the result shows thatserine, threonine, taurine were most contributed to inter-group, one-dimensionalstatistical analysis found that the urine taurine, arginine, serine and histidine significantlylower than normal, and glutamic acid were increased significantly. Are two ways to dealwith the results obtained can be complementary.4.The GC based serum metabolomics approach was also used to investigate thepathophysiological variation among type2diabetes mellitus patients compared to healthcontrol. The derivatization condition was optimized. Pattern recognition was carried outwith PCA and OPLS-DA were used to analysis All fatty acids were suceessfullyidentified and quantified, and stable plasma FFA metabolie Profiling of both the controlsand TZDM patients. The results found elevated Myristicacid in the serum of withdiabetes patients, while oleic acid in diabetic patients specifically down regulatedcompared with healthy people, which suggested that these two substances were potentialspecific markers to distinguish between diabetes and control and fatty acid metabolic canprovide important experimental basis for diagonosis.Conclusion:The NMR separation of urine samples was performed and UPLC-TOF-MSmetabolomics study can clearly classify the diabetes groups and control group. Themetabolites gotton from NMR were different from those from UPLC/MS, but thedifferent metabolites maily including amino acid, glucose, and organic acid. Comparedwith healthy controls, the glycolysis metabolism, and amino acid metabolism of diabetes mellitus patients were disordered to some degree.It indicated that both common andspecific metabolic pathways of diabetes mellitus were disordered. The relationshipbetween the two methods could b efurther studied according to the results of the presenttest.The HPLC and GC based serum and urine metabolomics approach was also used toinvestigate the pathophysiological variation among type2diabetes mellitus patients. Thederivatization condition was optimized. Pattern recognition was carried out with PCA andPLS-DA to get the similar results. This study would provide important significance fordiagnosis and evaluation of the therapeutic efficacy for diabctie mellitus.

  • 【分类号】R587.1;R29
  • 【被引频次】23
  • 【下载频次】3319
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络