【作者】 苑福升；

【导师】 杨小玉；

【作者基本信息】 吉林大学 ， 外科学， 2012， 博士

【摘要】 目的：本研究旨在通过建立代谢组学实验平台，对照分析动物成骨肉瘤（Osteosarcoma;OS）实验模型与临床OS、良性骨肿瘤患者的血清/尿液特征代谢产物，分别应用偏最小方差判别分析法（Partial least square-discriminate analysis;PLS-DA）等方法建立OS代谢谱数学模型，采用相应的可视化技术，通过对动物模型及OS患者的血液及尿液中的代谢产物参比研究，建立分辨率高，重复性好的OS血清、尿液代谢标志物组，并试图从机体的动态代谢途径寻找OS诊断标志物。方法：本研究分3部分进行1.建立模型动物OS的代谢谱。通过实验建立MG-63（人成骨肉瘤细胞株）C3H鼠模型和健康对照组，采用氯甲酸乙酯（Ethyl Chloroformate;ECF）衍生的气质联用（Gas chromatography/mass spectrometry;GC/MS）技术平台，分别观察模型鼠荷瘤后第3、5、7周的内源性小分子代谢谱相对于健康对照组尿液的变化。用主成分分析（Principal component analysis, PCA）的方法，判断荷瘤鼠同一时间点的尿液样本代谢谱轮廓和趋势；通过分析影响PLS-DA模型中的权重较大的变量(Variable importance inprojection；VIP)，鉴定差异代谢产物中主成分。2.建立临床骨肿瘤（Benign /Osteosarcoma）的代谢谱。将13名OS患者，18名良性肿瘤患者和20名健康对照人群分成3组，通过ECF衍生GC/MS技术平台对三组尿液样本代谢进行分析；同步基于三甲基硅烷（Trimethylsilyl；TMS）衍生的气相色谱飞行时间质谱联用分析（Gas chromatography/Time-of-flight mass spectrometry;GC/TOF-MS）分析的方法，考察三组血清样本。以此分别得到三组血清/尿液样本的代谢物谱变化。继而通过PCA的方法和PLS-DA，首先对比骨肿瘤组与健康对照组，进一步区分良性骨肿瘤和OS组，并判断模型中的VIP，鉴定出代谢产物中比较明显的成分(Principal components;PC)。3. OS代谢标记物与代谢通路相关的综合分析。将实验鼠和临床研究的结果予以对比，得出交集的代谢产物，在Metabolic Pathway以及KEGG(Kyoto Encyclopedia ofGenes and Genomes)数据库进行代谢通路的查询。通过综合分析，获得OS在发生发展过程中内源性小分子比较明显的代谢变化，推测OS代谢诊断标记物、主要代谢通路以及影响其代谢通路的关键酶。结果：1.实验中发现伴随着OS的形成与进展，PCA和PLS-DA等多维统计方法结果显示小鼠在实验第3周开始出现偏移、第5周已经初步显现分离趋势，第7周已经完全可以区分。利用PLS-DA模型实验中找出的与OS形成相关的37个差异代谢物，并利用NIST谱库等数据库，鉴定了其中10个代谢物。其中马尿酸，尿素，L-鸟氨酸在荷瘤组中含量较健康对照组低，而腐胺，亚精胺，异柠檬酸等含量较健康对照组高。2.利用非监督的PCA方法，得到健康对照组与骨肿瘤患者（Benign /Osteosarcoma）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（Orthogonal Partialleast square-discriminant analysis OPLS-DA）的方法能清晰地观察到OS、良性骨肿瘤患者与正常对照之间的代谢轮廓的差异，得到与OS临床病理相关的代谢通路的变化。血清代谢组学分析中：核糖醇、D-果糖、2 -酮戊二酸、乳酸、丁醛、丁酸、甘氨酸、a-氨基己二酸、L-鸟氨酸、甲基组氨酸、L-缬氨酸、十二烷酸等代谢物在OS患者血清中明显低于良性骨肿瘤患者；而瓜氨酸、L-酪氨酸、尿素、半乳糖、2-脱氧-D-半乳糖、硬脂酸、α-羟基丁酸等代谢物在OS患者血清中含量明显高于良性骨肿瘤。在尿液代谢组学分析中：甘氨酸、亮氨酸、赖氨酸、高香草酸、乳酸、柠檬酸等代谢物在OS患者尿液中含量明显低于良性骨肿瘤患者；马尿酸、谷氨酸、组胺、胱氨酸、腐胺等代谢物在OS患者尿液中含量明显高于良性骨肿瘤患者。3.我们发现了OS代谢产物的主要共同通路异常为肿瘤能量代谢异常和鸟氨酸循环-多胺代谢的异常。在OS尿液代谢组学研究中，柠檬酸、马尿酸盐、腐胺、亚精胺，鸟氨酸可能成为代谢诊断的标志物，关键酶推测是鸟氨酸脱羧酶（Ornithimedecarboxylase,ODC)，谷胱甘肽-S-转移酶（Gultathione S transferases;GST）的同工酶和GSTM1、GSTT1。OS血清代谢组学研究结果显示，苹果酸、十二烷酸、羊毛硫氨酸、有可能成为代谢诊断的标志物，推测的关键酶是鸟氨酸循环中负责调控/转运N-乙酰谷氨酸合酶(N-acetylglutamatesynthetase；NAGAS)和鸟氨酸转移酶（Ornithinecarbamyl transferase;OCT）。4.代谢差异结果表明：OS不仅影响了糖类、脂类和蛋自质三大物质的代谢，还对能量代谢、鸟氨酸循环-多胺代谢等多个生理系统产生显著影响，并且可能直接导致代谢通路损伤。结论：1.基于ECF和TMS衍生的GC/MS的OS代谢组学实验平台，进行OS动物模型和临床OS患者的血/尿液的代谢组学研究，能够初步得到良性，恶性和健康对照人群组别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OS的早期诊断上有很大的潜力。2.本研究中基于GC/TOF-MS的血清代谢组学方法同样比较稳定，可以清楚的区分OS、良性肿瘤和正常人的血清代谢轮廓。3.本实验证明：在模型动物实验和OS临床患者的尿液/血清的代谢组学研究中，证实能量代谢异常和鸟氨酸循环-多胺代谢等代谢通路碍在OS发生、发展病理过程中具有重要作用，与OS形成、恶化和转移具有重要相关性，有待于后续验证。本论文创新性研究成果：1.首次通过GC/MS和GC/TOF-MS实验平台对OS代谢组学进行研究，不但可以得到健康组与OS组代谢轮廓分离趋势，并可以进一步鉴别区分良性骨肿瘤与OS代谢模式。2.探索性地将动物实验与临床研究获得的结果进行整合分析，血清/尿液样本代谢组学分析结果交互验证，可以得到相对稳定的OS代谢标记物，可能启发基于色谱/质谱技术的代谢组学方法在OS的早期诊断的新思路。3.发现人与动物OS共同代谢通路为能量代谢，鸟氨酸循环-多胺代谢障碍为主导，得出重要结论：能量代谢障碍、鸟氨酸循环-多胺代谢等代谢通路障碍可能是OS的发生、发展、转移病理过程中重要的代谢模式，为OS代谢通路纵深研究提供了重要科学证据。更多还原

【Abstract】 Objective：This study is designed to learn through the metabonomics test platform, analyze animalmodels with healthy control group urine specimens through the experiments, and clinicallyresearch into patients with osteosarcoma, benign bone tumors and healthy control serum /urine specimens. It was established a spectral mathematical model of osteosarcomarespectively by the application of the Partial least square-discriminate analysis (PLS-DA)method and other methods. Relevant analysis techniques and statistical methods were usedto analysis and comparison of animal experiment and clinical study of samples of the endproduct of metabolism. High resolution, reproducible osteosarcoma serum and urinemetabolism markers matter group were established to try to find osteosarcoma diagnosticmarkers by the body dynamic metabolic pathways .Methods：This study was divided into three parts.1. Experiments of animal models of osteosarcoma metabolic spectrum.The MG-63 (humanosteosarcoma cell line) of C3H mouse model and the healthy control group wereestablished , the use of ECF derivative by GC / MS technology platforms were observedin a rat model of tumor-bearing after 3, 5 and 7 weeks of endogenous small moleculemetabolic spectrum relative to the changes in the urine of healthy control group. usingthe principal of the component analysis (PCA) method determine the tumor-bearing miceat the same time point urine samples metabolic spectrum profile and trends.Through theanalysis of variable importance in projection (VIP) of PLS-DA model, we could identifythe differences in metabolites of the main ingredients.2. Experiments of the metabolic spectrum of clinical bone tumors (Benign /osteosarcoma).13 patients with osteosarcoma, 18 patients with benign bone tumor and 20healthy controls were divided into three groups. ECF derivative by GC / MS technologyplatform to analyze three groups of urine samples of metabolism;Synchronization basedon the use Trimethylsilyl (TMS) derivatives by gas chromatography time-of-flight massspectrometry (GC / TOF-MS) analysis method to study three groups of serum samples.This metabolite spectral changes of serum / urine samples, respectively. Then, by themethod of PCA and PLS-DA/OPLS-DA, we compared the bone tumor group and thehealthy control group and further distinguished between benign bone tumors and osteosarcoma group. Through the analysis of VIP, we could identify the differences inmetabolites of the main ingredients.3. Comprehensive analysis of osteosarcoma metabolism markers and metabolic pathway.The results of animal experiment and clinical study were combined analysis, to come tothe intersection of metabolites in the Metabolic Pathway and the KEGG (KyotoEncyclopedia of Genes and Genomes) database query of the metabolic pathway.Throughcomprehensive analysis, we can get into osteosarcoma in a more obvious metabolicchanges during the development of endogenous small molecule, presumably into the theosteosarcoma metabolic biomarkers, major metabolic pathways, as well as its metabolicpathway enzymes.Results:1. The result of Animal experiment showed that model in three weeks of the experimentbegan to offset the first five weeks have been initially apparent separation, the first sevenweeks can distinguish.The formation of 37-related differences in metabolite identified inthe PLS-DA model experiments with osteosarcoma, and using the NIST library and otherdatabases identified 10 metabolites.Hippuric acid, Urea, L-ornithine content in the tumorgroup than the healthy control group is lower;The content of putrescine, spermidine,isocitrate is higher.2. By the unsupervised PCA method, we got the metabolic spectrum separation trend of thehealthy controls and patients with bone tumors (Benign / osteosarcoma)，We furtherused the orthogonal the partial least square-discriminant analysis (OPLS-DA) method toobserve clearly in the metabolic profile differences between the osteosarcoma, benignbone tumors and normal controls, and get the change of the metabolic pathways relatedto the clinical pathology of osteosarcoma.Serum metabonomics analysis: Ribitol、D-Fructose、2-Ketoglutaric acid、lactate、Butanal、Butanoic acid、Glycine、à-Aminoadipicacid、L-Ornithine、methylhistidine、L-Valine、Dodecanoic acid in the serum of patientswith osteosarcoma was significantly lower than in patients with benign bone tumors;Citrulline、L-Tyrosine、Urea、Galactopyranose、2-Deoxy-galactopyranose、Octadecanoicacid、α-Hydroxybutyric acid and other metabolites in the serum of patients withosteosarcoma were significantly higher than that of benign bone tumors. Urinemetabonomics analysis: glycine、leucine、lysine、homovanillate、lactate、citrate and othermetabolites in the urine of patients with osteosarcoma were significantly lower than in patients with benign bone tumors;hippurate、glutamate、histamine、Cystine、putresinein the urine of patients with osteosarcoma were significantly higher than in patients withbenign bone tumors3. We found abnormal osteosarcoma metabolites common pathway for tumor energymetabolism and ornithine cycle - polyamine metabolism abnormalities. Osteosarcomaurine metabonomics studies, citric acid, hippurate, putrescine, spermidine, ornithine maybecome the markers of the metabolic diagnosis, speculated that a key enzyme ornithinedecarboxylase (Ornithime decarboxylase ODC), S-transferases Gultathione (GST)isozyme of GSTM1, GSTT1.Into osteosarcoma serum metabonomics study results showthat malic acid, dodecanoic acid, Lanthionine may become the markers of the metabolicdiagnosis, speculated that a key enzyme responsible for the regulation of ornithine cycle /transit N-acetyl glutamate synthetase (NAGAS) and ornithine carbamyl transferase(OCT)。4. Metabolic differences: osteosarcoma not only affects the metabolism of carbohydrates,lipids and proteins in the three substances, and also have a significant impact on energymetabolism and ornithine cycle - polyamine metabolism and other physiologicalsystems,and may be a direct result of the metabolic pathways.Conclusion:1. Osteogenic sarcoma metabonomics experimental platform, based on ECF and TMSderivative by GC / MS study of the osteosarcoma animal models and clinicalmetabonomics into the blood / urine of patients with osteosarcoma, we can initially beenhealthy. malignant and healthy control population group separation of metabolicprofiling trends.This indicates that there is great potential in the early diagnosis ofosteosarcoma, based on chromatography / mass spectrometry method of urinemetabonomics.2. In this study, GC / TOF-MS-based serum Metabonomics also relatively stable, we canclearly divided into osteosarcoma, benign and normal serum metabolic profile.3. Model animal experiments and osteosarcoma clinical patient urine / serum metabonomicsconfirmed that the metabolic pathway for energy metabolism and ornithine cycle - andpolyamine metabolism obstruction occurred in osteosarcoma, the development ofpathological processes play an important role, and osteosarcoma formation, degradationand transfer to an important need to be follow-up verification.Contributions and innovations： 1. The first time we osteosarcoma metabonomics learn by GC / MS and GC / TOF-MSexperiment platform, not only can the healthy group and into the trend of metabolicprofiling of osteosarcoma group separation and can be further distinguishingbetween benign bone tumors with osteosarcoma metabolism mode.2. We exploratory results obtained in animal experiments and clinical studies tointegrate analysis, serum / urine samples metabonomics analysis results of crossvalidation can be relatively stable osteosarcoma metabolic markers. This mayinspire new ideas in the early diagnosis of osteosarcoma chromatography / massspectrometry-based metabonomics.3. We found that human and animals into osteosarcoma, a common metabolic pathwayfor energy metabolism and ornithine cycle - polyamine metabolic disorder led toimportant conclusions: the metabolic pathways of energy metabolism, ornithinecycle - and polyamine metabolism disorder may be osteogenic sarcoma,development, transfer pathological process of metabolism, for osteosarcomametabolic pathway in-depth studies provide important scientific evidence.更多还原