【作者】 邱云平；

【导师】 贾伟； 陈闽军；

【作者基本信息】 上海交通大学 ， 药理学， 2008， 博士

【摘要】 大肠癌(colon cancer)包括结肠癌和直肠癌,因此又称为结直肠癌(colorectal cancer),是一种常见的消化道恶性肿瘤,死亡率在世界范围内居各种肿瘤的第三位,在西方居第二位。在我国随着人们饮食结构的改变,高蛋白食物摄入量的增加,大肠癌的发病率和死亡率正在逐步上升。大肠癌的发生和发展受到基因和环境等多种因素的影响,其病理发生的机制比较复杂,从单一的基因突变和分子通道的变化难以全面理解大肠癌的发生机制,给大肠癌的治疗带来困难。另外,大肠癌的治疗效果与病理分期有很大的关系,晚期患者的术后5年存活率小于8%,而早期患者术后5年存活率可以达到90%以上。然而令人遗憾的是目前为止还没有一个快速、准确而又适合于大规模筛查的大肠癌早期诊断方法,价格昂贵而又对人体伤害较大的内窥镜检查仍然是大肠癌诊断的金标准,这严重限制了大肠癌的早期诊断。代谢组学(metabonomics/ metabolomics)是继基因组学、转录组学、蛋白质组学之后提出的一门新的学科,是系统生物学中的一个重要组成部分,近年来已经迅速成为生命科学研究的热点之一。生物机体在受到外界病源或环境改变的刺激时,会产生多层次、多器官和多组织的应答反应,这些应答反应最终将影响到终端代谢水平的变化,代谢组学正是研究这种机体在受到内因、外因的刺激(或是病灶)时,机体内的终端小分子代谢物(通常指分子量小于1000Da)的变化规律的科学,从整体上动态地阐述生理或病理状态下机体对外界刺激的应答和动态系统性的变化过程,为全面理解多因素影响下的大肠癌的发生机制研究将提供一个新的视角。同时,从整体上考察人体的生理状态,能得到更多的疾病相关信息,有利于大肠癌的早期诊断。另外,代谢组学技术分析的样本主要是尿液和血液,其采集过程对人体无伤害或是有极小的伤害,因此代谢组学的方法将有利于大肠癌的大规模筛查。本论文建立了基于色谱质谱联用技术的大肠癌尿样和血清的代谢组学研究方法,考察了大肠癌癌前病变模型大鼠以及中药干预下的大鼠尿样的代谢组学变化;并分析了临床大肠癌病人与正常对照人群尿液和血液的代谢谱的差异,主要内容和结果如下:1、采用氯甲酸乙酯(ECF)衍生,气相色谱质谱联用(GC/MS)分析的代谢组学方法,考察了大鼠在二甲肼(DMH)诱导下形成异常病灶腺窝(ACF)过程以及中药干预下内源性小分子代谢物谱的变化,找出了与大肠癌癌前病变发生密切相关的代谢通路并探讨了中药的作用靶点。用主成分分析(PCA)的方法,能清晰地将造模七周后的大鼠尿液样本与同一时间点的正常对照组样本区分开,而中药治疗组的动物的代谢谱与正常对照组比较接近,这一结果与组织病理学研究结果中的ACF数值相一致。通过研究影响最小方差判别分析法(PLS-DA)模型中的权重较大的变量,实验中发现伴随着ACF的形成,大鼠体内的能量代谢、色氨酸代谢、多胺代谢以及肠道菌群的结构等与实验大鼠大肠癌癌前病变高度相关。对中药黄连和吴茱萸生物碱提取物干预下的大鼠尿样代谢组的研究结果表明环氧化酶-2(COX-2)和细胞色素酶P450以及肠道菌群可能是中药干预的靶点。2、采用氯甲酸乙酯衍生和气相色谱质谱联用的代谢组学技术,分析了大肠癌病人和正常对照人群的尿液中代谢谱的差异,利用非监督的PCA方法,能得到正常人与大肠癌患者代谢谱的分离趋势。进一步利用监督的正交偏最小方差判别分析(OPLS-DA)的方法能清晰地观察到大肠癌患者与正常对照之间的代谢轮廓的差异,其中包括5例病理分期为Ⅰ期的病人也能与正常人完全分开,说明以GC/MS分析为基础的尿液代谢组学的研究对大肠癌的诊断甚至早期诊断有很大的潜力。利用OPLS-DA的方法,试验中观察到大肠癌患者不同病理分期(Ⅱ,Ⅲ患者)之间的区分趋势,说明基于尿液的代谢组学方法对于大肠癌的病理分期有一定的判别作用。通过寻找对区分大肠癌病人和正常对照之间的OPLS-DA模型贡献较大的变量,可以得到与大肠癌临床病理相关的代谢通路的变化,主要包括色氨酸的代谢、组氨酸的代谢和三羧酸循环以及肠道菌群的变化。3、在优化的条件下,采用超高效液相四级杆串联飞行时间质谱联用仪(UPLC/QTOFMS)技术的代谢组学方法分析了大肠癌患者与正常对照的代谢组学差异。试验中得到与基于GC/MS代谢组学分析类似的结果,也能清晰地区分大肠癌病人和正常对照尿样,并且对于不同病理分期的病人(Ⅱ,Ⅲ病人)的之间的区分要稍好于基于GC/MS代谢组学分析的结果。同时,本实验进一步观察到与大肠癌相关的代谢通路变化――酪氨酸代谢、同型半胱氨酸代谢等。4、建立了基于三甲基硅烷(TMS)衍生、气相色谱飞行时间质谱(GC/TOFMS)分析的血清代谢组学研究方法。该方法通过线性、重复性、稳定性和回收率等方法学的实验考察,结果表明该方法稳定可靠。利用基于GC/TOFMS的分析方法,分析了大肠癌患者与正常对照人群血清样本的代谢物变化。通过OPLS-DA模型分析,实验中能清晰地观察到大肠癌患者和正常人样本代谢谱之间的分离,并观察到处于Ⅱ、Ⅲ期的大肠癌患者代谢谱的分离趋势,区分效果与基于UPLC/QTOFMS的尿样分析的代谢组学研究结果类似。同时,对大肠癌和正常人OPLS-DA模型贡献最大的代谢物的研究表明,油胺代谢、有氧和无氧能量代谢、鸟氨酸循环以及一些其它的氨基酸代谢在大肠癌患者血清中的含量有着显著的变化。本论文的研究结果能够得出了以下几点结论:1、基于氯甲酸乙酯衍生的GC/MS分析方法适合于化学诱导剂诱导的大肠癌癌前病变动物实验、中药干预动物模型实验以及临床大肠癌病人的尿液代谢组学研究。2、基于GC/MS和UPLC/QTOFMS技术的代谢组学分析方法能够准确地将包括早期病人在内的大肠癌病人与正常人的尿样区分开,并且能观察到不同病理分期下的病人代谢轮廓的分离趋势,表明基于色谱质谱的尿样代谢组学方法在大肠癌的早期诊断甚至分期上有很大的潜力。3、本研究中建立的基于GC/TOFMS的血清代谢组学方法学稳定可靠,并且能很好地区分大肠癌病人与正常人的血清代谢谱(其中包括一些早期病人),并观察到Ⅱ、Ⅲ病人之间的差别。4、癌前病变的模型动物实验和大肠癌临床病人的尿样和血样的代谢组学研究结果表明与大肠癌密切相关的代谢通路变化包括油胺代谢、能量代谢、色氨酸的代谢、多胺的代谢以及肠道菌群结构的变化。更多还原

【Abstract】 Colon cancer, also known as colorectal cancer (CRC), is one of the most common malignant tumors in the digestive tract. CRC represents the third leading deaths among the cancers worldwide, and the second in the Western developed countries. In China, the number of new cases and deaths of CRC are rapidly increasing along with the elevation of living standards and the increased consumption of high protein diet. As a multi-factorial and polygenic condition, CRC has complex molecular mechanisms which are not well understood by biomedical scientists and pathologists, and therefore, necessitate systematic investigations. Meanwhile, the clinical efficacy on CRC treatment is highly correlated with the pathological stages of the disease. For example, the five-year survival rate of stage I patient is about 93%, and this number sharply decreases to only about 8% for the stage IV patients. Despite of the apparent benefit on early treatment, there are few proper measures to detect CRC at early stage. The current“golden diagnostic tool”, colonoscopy, is not suitable for most of the population due to the cost and inconvenience in clinic use. It is therefore of vital importance to develop alternative or complementary measures for early diagnosis of CRC.Metabonomics/ metabolomics is a newly developed approach, as an integral part of the systems biology encompassing a number of omics sciences such as genomics, transcriptomics and proteomics, and has become one of the hottest subjects worldwide. Metabonomics uses multivariate statistical technique to analyze highly complex data sets generated by high-throughput spectroscopy such as nuclear magnetic resonance (NMR) and mass spectrometry (MS) of biological samples to capture metabolic variations in response to genetic modifications and environmental stimuli. As metabolites in serum and urine contain the general functional information generated by the biochemical regulatory systems in the whole body, metabonomics reveals a systems and dynamic outcome of the development of a pathological state. Therefore, the high-flux metabolic information originated from variations of global metabonome reveals important clues to disease onset and development, and thus, can be used for diagnostic applications such as CRC detection. Meanwhile, the noninvasive or less invasive bio-fluid sample (urine and serum) collection makes such a methodology easy to be adopted in clinical diagnosis or in routine physical examination. In this study, we try to establish chromatography in hyphenation with mass spectrometry metabonomic technology to investigate metabolic variations in precancerous colon rats and CRC patients and to investigate the applicability of such metabonomics method in CRC diagnosis.Main methods and results:1. We used ethyl chloroformate derivatization and gas chromatography-mass spectrometry (GC/MS) based metabonomic analysis of urines from 1,2-dimethylhydrazine (DMH)-induced precancerous colon rats, herbal medicine treated rats and healthy controls. The time-dependent variations of metabolite profile showed a progressive deviation of the metabolism in the model group from the initial pattern over time and a systemic recovery of the metabolism in the treatment group, which is consistent with the histological results. Additionally, the in-depth study of these metabolite alterations also allowed the simultaneous identification of key sites and pathways, such as gut microflora, COX 2 and cytochrome P450, which were closely associated with the herbal medicine treatment.2. We used a GC/MS-based metabonomic approach to investigate the CRC-related pathophysiological variations of the urinary metabolite profiles from 51 CRC patients in comparison to those from 39 age-matched volunteers. A group of metabolites significantly differed in CRC patients from the healthy controls were identified including the decreased levels of succinate, butyrate, citrate, histidine and 2-hydroxyhippurate, and the increased levels of histamine, hippurate, 5-hydroxyindoleacetate and glycine. In addition, such a GC/MS-based metabonomic approach was able to clearly recognize five patients (stage I) from the healthy controls and appeared to reveal a different metabolic pattern of patients at different pathological features (stage II and stage III).3. UPLC/QTOFMS based urinary metabolic profiling method was basically established. By comparing different solvent and dilution fold, we selected 2 fold diluting with pure water of urine as the final pretreatment method. Through this method, the similar separation between urines from CRC patients and healthy controls with GC/MS analysis and a slight better separation between stage II and stage III patients was obtained.4. The GC/TOFMS based serum metabonomic analysis method was optimized. The method validations revealed a wide linearity range, good repeatability and acceptable recovery rate for the proposed method. Based on the established method in this study, sera from CRC patients and healthy controls were analyzed to detect metabolite variations associated with CRC morbidity. The similar separation results were obtained with urine metabolic profiles. After identification of metabolites significantly varied in the CRC patients, oleamide metabolism, anaerobic and aerobic energy metabolism, ornithine metabolism and some other amino acids metabolism were associated with CRC morbidity.Conclusions:1. Using ethyl chloroformate derivatization and GC/MS analysis based urinary metabonomic study, we can visualize time-dependent variations in the DMH-induced precancerous colon rat and the reversal effect of herbal medicine, which reveals great potential of metabonomics used in detection of insight pathological variations and in tracking of the effect of drug intervention.2. Based on GC/MS and UPLC/QTOFMS urine metabonomic analysis, we can precisely distinguish CRC patients from healthy controls including 5 patients of early stage (stage I) and can generally differentiate patients of different pathological stage (stage II and stage III), which reveals great potential of urine metabolic profile in CRC diagnosis.3. The established GC/TOFMS based serum metabonomic analysis method in this research was reliable and stable. Using this method, we can clearly visualize the metabonomic differences in the OPLS-DA model between CRC patients and healthy controls and can even tell apart from patients of different pathological stages (stage II and stage III). 4. From the results of urine metabolic profile of DMH induced precancerous colon rats and biofluids (urine and serum) from CRC patients, we found that oleamide metabolism, energy supply metabolism, tryptophan metabolism, polyamine metabolism, and the altered gut flora structure are closely related to CRC morbidity.更多还原