结核性脑膜炎患者血清及脑脊液代谢组学研究

发布时间：2018-05-18 00:25

本文选题：结核性脑膜炎 + 超高效液相色谱　；参考：《天津医科大学》2017年硕士论文

【摘要】：目的:探寻结核性脑膜炎(Tuberculous meningitis,TBM)患者脑脊液和血清中的特异性代谢物质,通过分析研究这些代谢物质在患者体内数量和性质的改变,以及其在疾病衍变过程中的代谢机制,寻找有代表性的代谢离子,辅助临床对结核性脑膜炎的早期诊断,病情监测,以改善患者预后。方法:采用超高效液相色谱与质谱联用(UPLC-MS)代谢组学技术平台共分析脑脊液标本47例,血清标本107例。样本来自取2014年11月-2016年12月在天津市海河医院脑系科和结核科住院患者的脑脊液标本共47例,其中TBM的患者21例,非结核性脑膜炎患者26例,另选取健康体检者30例。最终研究对象共有102例,其中男57例,女45例;将其分成:结核性脑膜炎组患者21例,男/女为12/9,平均年龄(48.01±12.35)岁。对照组包括非结核性脑膜炎组(N-TBM组)26例,肺结核组(TB组)25例,健康对照组(N组)30例。其中N-TBM组男/女为14/12,平均年龄(45.56±19.43)岁,包括细菌性脑膜炎7例;病毒性脑膜炎15例,其它原因引起的脑膜炎4例;TB组25例,其中男15例,女10例,平均年龄(46.56±15.43)岁;健康对照组共30例,其中男16例,女14例,平均年龄(43.56±21.43)岁。脑脊液和血清样品检测均采用UPLC-MS。初步数据处理采用Mzmine2.0软件;初步处理后的数据录入SIMCA-P+12.0.1.0软件的模式识别系统构建主成分分析(PCA)和正交偏最小二乘(OPLS-DA)分析模型,进行特异代谢物的过滤筛选,再利用二分类OPLS-DA模型中的变量投影重要性值(VIP)和VIP图,载荷图(S)对代谢标志物进一步筛选;初筛后的代谢物质采用SPSS 21.0进行独立样本非参数检验,鉴定特异性代谢标志物;通过ROC曲线面积,评价特征性代谢物诊断结核性脑膜炎的诊断效能。结果:分别以TBM和N-TBM两组的脑脊液和血清代谢物数据结果构建OPLS-DA模型,发现可以很好地区分结核性脑膜炎组和非结核性脑膜炎组,该模型预测准确度达到100%。脑脊液寻找到23个特征代谢物,发现了5个代谢物具有很好诊断效能;两组血清筛出了31个特征离子,发现了6个代谢物具有很好诊断效能。脑脊液中的特征代谢离子包括番茄红素甲酸脂(Methyl(7Z,9Z,9'Z)-6'-apo-ycaroten-6'-oate)、二氢顺丁烯二酰亚胺β-D葡萄糖苷(Dihydromaleimide beta-Dglucoside)、去水山梨糖醇单油酸酯(Sorbitan oleate)、人参环氧炔醇亚油酸脂(Panaxydol linoleate)。血清中的可能特征代谢标志物包括硬脂酸酰胺(Octadecanamide)、溶血磷脂酰胆碱(LysoPC(18:0))、磷脂酸20:4(5Z,8Z,11Z,14Z)e/2:0)、植烷酸、二十二烷酸6种代谢物有望成为结核脑膜炎的血清特异性诊断标志物。结论:UPLC-MS的代谢组学研究技术结合正交偏最小二乘判别分析(OPLS-DA)模型的多元统计学分析方法,能够发现与结核性脑膜炎的特异性强且有一定诊断效能的代谢产物,通过研究分析这些代谢产物在结核性脑膜炎的疾病发生发展过程中的产生机制,有助于对结核性脑膜炎的早期诊断,为临床的靶向治疗等进一步研究提供理论依据和研究方向。
[Abstract]:Objective: To explore the specific metabolic substances in the cerebrospinal fluid and serum of patients with Tuberculous meningitis (TBM). Through the analysis and study of the changes in the number and nature of these metabolic substances in the patients, and the metabolic mechanisms in the process of disease evolution, a representative metabolic ion is sought to assist the clinical treatment of tuberculous brain. Early diagnosis and monitoring of the disease in order to improve the prognosis of the patients. Methods: a total of 47 specimens of cerebrospinal fluid (CSF) specimens and 107 serum samples were analyzed by the UPLC-MS (super high performance liquid chromatography and mass spectrometry) technique platform. The samples were taken from the cerebrospinal fluid of the patients in the Department of brain and tuberculosis of Tianjin Haihe Hospital in December November 2014. A total of 47 cases, including 21 cases of TBM, 26 cases of non tuberculous meningitis and 30 cases of health examination, were selected in 102 cases, including 57 male and 45 women, which were divided into 21 cases of tuberculous meningitis group, male / female 12/9, average age (48.01 + 12.35) years. The control group included non tuberculous meningitis group (group N-TBM) 26. There were 25 cases in the pulmonary tuberculosis group (group TB) and 30 cases in the healthy control group (group N). Among them, the male / female group in group N-TBM was 14/12, the average age was (45.56 + 19.43) years, including 7 cases of bacterial meningitis, 15 cases of viral meningitis, 4 cases of other causes of meningitis, and 25 cases in group TB, including 15 men, 10 women, average age (46.56 + 15.43) years, and healthy control groups, among them, among them a total of 30 cases in healthy control group, among them, healthy control group altogether 30 cases, among them, healthy control group altogether 30 cases, among them, healthy control group. There were 16 males and 14 females with an average age of (43.56 + 21.43) years. The detection of cerebrospinal fluid and serum samples were performed using Mzmine2.0 software by UPLC-MS. preliminary data processing. The primary processed data entered the SIMCA-P+12.0.1.0 software pattern recognition system to construct principal component analysis (PCA) and positive partial least squares (OPLS-DA) analysis model for specific metabolism. Filtering and screening of the material, then using the value of the variable projection value (VIP) and VIP diagram in the two classification OPLS-DA model, and the load map (S) to further screen the metabolic markers; the metabolic substances after the initial screening used SPSS 21 for the independent sample nonparametric test, identified the specific metabolites, and evaluated the characteristic metabolite diagnosis through the ROC curve area. The diagnostic effectiveness of tuberculous meningitis. Results: the OPLS-DA model was constructed with the results of TBM and N-TBM two groups of cerebrospinal fluid and serum metabolite data. It was found that the tuberculous meningitis group and non tuberculous meningitis group could be well divided. The prediction accuracy reached to the 100%. cerebrospinal fluid to find 23 Characteristic Metabolites and found 5 metabolites. The two groups of serum screened 31 characteristic ions and found that 6 metabolites have good diagnostic efficiency. The characteristic metabolic ions in the cerebrospinal fluid include lycopene formate (Methyl (7Z, 9Z, 9'Z) -6'-apo-ycaroten-6'-oate), and two HP Ding Xier imide beta -D glucoside (Dihydromaleimide beta-Dglucoside). Sorbitol mono oleate (Sorbitan oleate) and ginseng epoxy alkynol linoleic acid (Panaxydol linoleate). The possible characteristic metabolic markers in the serum include stearamide (Octadecanamide), lysophosphatidyl choline (LysoPC (18:0)), phosphatidic acid 20:4 (5Z, 8Z, 11Z, 14Z), and 6 metabolites of alkanoic acid and twenty-two alkanoic acid. Conclusion: the multivariate statistical analysis of UPLC-MS's metabonomics combined with orthogonal partial least squares discriminant analysis (OPLS-DA) model can detect the specific and diagnostic metabolites of tuberculous meningitis, and analyze these metabolites by analyzing these metabolites. The mechanism of the occurrence and development of tuberculous meningitis is helpful to the early diagnosis of tuberculous meningitis and to provide theoretical basis and research direction for further study of clinical target therapy.
【学位授予单位】：天津医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R529.3

【参考文献】