用核磁共振波谱技术结合多维变量分析研究人体肺癌组织萃取液的代谢组特征  

Metabonomic Characteristics of Human Lung Cancer Using High Resolution 1H NMR Spectroscopy and Multivariate Data Analysis

作　　者：陈文学[1,2] 祖育昆[3] 卢韶华[4] 陈芬儿[1] 蓝文贤[5] 白春学[4] 邓风[2] 

机构地区：[1]复旦大学化学系复旦-DSM联合实验室,上海200433 [2]中国科学院武汉物理与数学研究所波谱与原子分子物理国家重点实验室,武汉430071 [3]华中科技大学同济医学院附属同济医院,武汉430030 [4]复旦大学附属中山医院,上海200233 [5]中国科学院上海有机化学研究所,上海200032

出　　处：《中国科学：生命科学》2010年第12期1145-1153,共9页Scientia Sinica(Vitae)

基　　金：国家自然科学基金(批准号:20872018);中国博士后科研基金(批准号:20090450065);波谱与原子分子物理国家重点实验室开放基金(批准号:T152805)资助项目

摘　　要：肺癌严重威胁着人类的健康,其发病率和死亡率均居世界恶性肿瘤首位.临床上,肺癌的诊断主要依靠组织病理切片.清楚地了解肿瘤的生化代谢特征能为肺癌的准确诊断提供重要帮助.当前,生物组织萃取液的高分辨核磁共振波谱作为一个极好的调查组织生化代谢的工具已得到广泛应用,特别是它与多维变量分析方法相结合已成为研究生物组织代谢特征的重要平台.本研究通过HRNMR波谱技术结合多维变量分析方法(主要包括主成分分析和偏最小二乘法-判别分析方法)得到了34个肺癌病人在3个不同组织位点的101例肺组织萃取液的代谢组特征.PCA研究显示,肺癌组织在不同的位点具有不同的代谢组学特征.与临近的非侵入肺组织相比,肺癌组织中乳酸的含量显著升高,而肌醇、谷氨酰胺和缬氨酸的含量显著下降.本研究表明,用正交偏最小二乘法-判别分析方法可很好地区分临近的非侵入组织和肺癌组织,该模型预测肺癌的准确性为100%.Lung cancer is a serious and leading disease imperiling people’s health with the highest morbidity and mortality among all malignant tumors. Clinical diagnosis of lung cancer mainly relies on histopathological evaluation of tissue specimen. Extensive knowledge of the metabolic biochemistry of tumors can potentially provide important information for accurate diagnosis of lung cancer. High resolution 1H NMR spectroscopy of biological tissues extracts has emerged and been widely acknowledged as an excellent tool in investigating tissue metabolism. Particularly, the combination of HRMAS NMR technique and multivariate data analysis （MVDA） has become an important metabonomics platform for studying the intact biological tissues. In this paper, the metabonomic characteristics of 101 lung tissue extracts from 34 patients with lung cancer have been obtained using the HR 1H NMR spectroscopy and the MVDA methods including principal component analysis （PCA） and/or orthogonal partial least squares-discriminant analysis （OPLS-DA）. The present study suggests clear differences among the metabonomic characteristics of lung tissues at various sites. Compared with the adjacent non-involved tissues, the lung cancer tissues had significantly higher levels of lactate but significantly lower levels of myo-inositol, glutamine and valine. Furthermore, OPLS-DA not only showed a good difference between the adjacent non-involved tissues and lung cancer tissues, but also produced 100% of predication for lung cancer.

关 键 词：肺癌 核磁共振 多维变量分析 模式识别 预测 

分 类 号：R734.2[医药卫生—肿瘤]