基于气相色谱-质谱联用的结直肠癌代谢组模式特征性研究

Study on specific metabonomic profiling of serum from colorectal cancer patients by gas chromatography-mass spectrometry

目的研究结直肠癌患者血清的代谢表型改变，寻找与结直肠癌发病过程密切相关的小分子代谢物．试图建立应用于结直肠癌诊断的特征性代谢模式。方法利用代谢组学中气相色谱．质谱联用仪（GC—MS）方法对比结直肠癌患者（肠癌组）31例和正常成人（对照组）8例血清的代谢谱变化．同时利用美国标准与技术研究院质谱数据库对所有共有的内源性代谢物做鉴定，应用数学模型PLS—DA进行模式特征性分类．同时对肠癌组与对照组的区分率利用监督性的聚类分析进行分析。结果共鉴定出34个内源性代谢物（包括糖类、氨基酸、脂肪酸及一些中间代谢小分子）。与对照组代谢物相对含量相比．肠癌组血清中L—valine、L—threonine、1-deoxyglucose、Glyeine和Ribitol的含量显著下降．其中3-hydroxybutyric acid的含量则显著升高（P〈0．01，P〈0．05）。同时PLS—DA和聚类分析均能把肠癌组和对照组很好地区分开．利用上述6种结直肠癌相关代谢物进行监督性的聚类分析．显示结直肠癌的诊断率达93．5％。结论GC—MS作为代谢组学的一种研究方法在肿瘤生物学领域中可行．利用GC—MS所发现的结直肠癌患者血清特征性代谢谱的变化及相应的代谢分子标志物．将有利于结直肠癌的临床研究。

Objective To study the specific metabonomic profiling of serum from colorectal cancer patients to find out the low molecule metabolites associated intimately with colorectal cancer, and to establish specific metabolic model for the diagnosis of colorectal cancer. Methods The metabonomic profiles of the serum samples from colorectal cancer （CRC） patients （n=31） and healthy adults （n=8） were investigated by gas chromatography-mass spectrometry （GC-MS） technique combined with a commercial mass spectral library for the peak clustering based on metabolites. Results Thirty-four endogenous metabolites including some amino acids, carbohydrates, fatty acids and other intermediate metabolites were identified. By t test statistics with P〈0.05, P〈0.01 respectively, L-valine, L-threonine, 1-deoxyglucose, glycine and ribitol levels were decreased significantly, but 3-hydroxybutyric acid level was increased significantly in the CRC patient group as compared with healthy adult group. PLS-DA based on these metabolites discriminated two groups for each other. Hierarchical clustering based on above 6 significant differential metabolites revealed that the prediction accuracy of colorectal cancer group was 93.5%. Conclusion GC-MS technique is an alternative tool for the metabonomic study and would be certainly beneficial to the pathological research, early diagnosis and therapy evaluation of CRC.