2型糖尿病患者血浆中脂肪酸代谢谱的研究

【作者】 谭斌斌；

【导师】 梁逸曾；

【作者基本信息】 中南大学 ， 分析化学， 2010， 硕士

【摘要】 代谢组学是继基因组学、转录组学和蛋白质组学之后在学术界兴起的又一门新兴系统生物学分支。其在疾病诊断、药物的药效和毒性评价、药物作用机理研究及中药现代化等方面展现了广阔的前景。当前代谢组学研究中存在两个难点同时也是热点问题：一是代谢产物的定性、定量分析,二是代谢产物复杂数据分析体系中有用信息的提取和挖掘。本论文针对这两个问题,利用气相色谱-质谱(GC-MS)联用仪,结合化学计量学方法中的多元分辨和模式识别方法,对2型糖尿病(T2DM)及相关疾病患者血浆中的脂肪酸代谢谱进行了全面、细致的研究。本实验采用一种简单、稳定的氢氧化钠-甲醇／硫酸-甲醇甲酯化方法对人体血浆中的游离脂肪酸(FFA)进行衍生和提取,并运用GC-MS联用技术对衍生产物进行分析检测。借助于直观推导式演进特征投影法(HELP)和选择性离子分析法(SIA)对重叠色谱峰进行分辨和解析,对所有FFA成分进行了准确的定性、定量分析,建立了稳定的2型糖尿患者和健康人血浆中的游离脂肪酸代谢谱。本论文运用一种新的变量筛选方法——竞争性自适应的重加权采样法(CARS),结合偏最小二乘-线性判别分析法(PLS-LDA)对2型糖尿病患者和健康人进行判别分析,建立判别模型,取得了较好的预测效果,并筛选出三个潜在的生物标记物：油酸(OLA),α-亚麻酸(ALA)和二十碳五烯酸(EPA)。这对2型糖尿病的临床诊断和评价治疗效果具有非常重要的指导意义。本论文还分别对成人腹型肥胖患者、儿童腹型肥胖患者、成人腹型肥胖合并2型糖尿病患者及相关并发症患者血浆中的游离脂肪酸代谢谱进行了研究,发现不同疾病状态的患者之间存在脂肪酸代谢模式的差别。进一步考察脂肪酸代谢模式与临床指标之间的相关关系,结果显示：人体血浆中的脂肪酸代谢谱与临床参数之间存在良好的线性关系,并筛选出一些与疾病相关的重要脂肪酸成分和临床参数。为肥胖、2型糖尿病及相关并发症的发病机理研究和指导不同病人的临床治疗提供有用的信息。本研究充分展现了代谢组学研究技术结合化学计量学方法在2型糖尿病及相关疾病研究方面的巨大潜力和优势。这将为各种复杂的代谢性疾病研究提供新的研究策略和技术平台。更多还原

【Abstract】 Metabolomics is a rising branch of the systems biology methodology after genomics and proteomics. It has been widely applied into disease diagnosis, drug effect and toxicity evaluation, drug mechanism, modernization of Traditional Chinese Medicine (TCM) and other fields. There are two difficult and hot problems in metabolomics research, one is the qualitative and quantitative analysis of metabonome and the other is extracting and mining useful information from complex data system of all metabonome. Aiming at the two problems, a comprehensive and painstaking study on the plasma fatty acid metabolic profiling of type 2 diabetes mellitus (T2DM) and correlated disease patients was done with the help of gas chromatography-mass spectrometry (GC-MS) and chemometric methods in this thesis.NaOH-CH3OH/H2SO4-CH3OH esterification method, a simple and stable method, was used to derviate the free fatty acids (FFA) from plasma in this experiment. And then, the products of derivative was detected and analysized by GC-MS. Two chemometric resolution methods, heuristic evolving latent projections (HELP) and selective ion analysis (SIA), have been used to effectively handle overlapping peaks of GC-MS data. All fatty acids were successfully identified and quantified, and stable plasma FFA metabolic profiling of both health controls and T2DM patients were established.In this thesis, a newly proposed competitive adaptive reweighted sampling (CARS) method, which can be used to screen variables, coupled with partial least squares linear discriminant analysis (PLS-LDA) was applied for the discriminant analysis between T2DM patients from healthy controls. A discriminant model was established, which got a good result for prediction, and three fatty acids, which were oleic acid (OLA), a-linolenic acid (ALA), and eicosapentaenoic acid (EPA), were finally selected as potential biomarkers. This study would provide important guiding significance for clinical diagnosis and evaluation of the therapeutic efficacy for type 2 diabetic mellitus. In this study, it was found that there was difference on fatty acid metabolic patterns among different stages of disease, based on the fatty acid metabolic profiling about abdominal obesity patients, abdominal obesity children, and abdominal obesity with T2DM complicated with hyperlipidemia patients. The further research demonstrated that there existed linear relationship between fatty acid metabolic pattern and clinical parameters. Several important fatty acids and clinical parameters relating to disease have been selected, which could provide important information for the study of pathogenesis and clinical intervention on T2DM and correlated disease.This study has fully demonstrated the huge potential and vantage of the combination of metabolomics and chemometrics on the research of T2DM and correlated disease. It will provide a novel study strategy and technology platform for the research on complex metabolic disease.更多还原