- 學位論文
非標靶代謝體學的方法開發與應用
Methodology Development and Application of Untargeted Metabolomics
摘要
代謝體學(Metabolomics)是一門探討生物體中小分子代謝物之組成的學門，相較於蛋白質體學、基因體學，代謝物在生物體中的狀態能夠提供最接近表現型的資訊。由於其結構差異，代謝物在生物體中具有提供能量、提供大分子結構基礎、傳遞訊息、作為蛋白質受體、輔酶等等功能。了解代謝物的全貌與交互關係，是近十年來生物化學家致力的課題。代謝體學又分成標靶代謝體學(targeted metabolomics)與非標靶代謝體學(untargeted metabolomics)兩種，非標靶代謝體學利用串聯質譜(MS/MS)圖譜與線上資料庫比對，進行無偏頗的代謝物結構鑑定。我們使用液相層析-串聯質譜儀(Liquid Chromatography-Mass Spectrometry)方法進行胞外小體(exosome)與載脂蛋白E基因剔除大鼠(ApoE-/- rat)之非標靶代謝體分析，並發現癌症轉移相關的小分子生物標記物(biomarker)與可能作為抗心血管疾病藥物之代謝物。此外我們更開發了一套新的非數據依賴擷取方法：選擇邊緣擷取(Selected MARgins acquisiTion)，此方法使用特別設計過的前驅物隔離區間(isolation window)並藉此增進串聯質譜圖譜的反褶積(deconvolution)與代謝物鑑定結果。 腫瘤衍生胞外小體(tumor-derived exosome)在癌症發展與轉移中扮演重要角色，胞外小體也被發現富含生物標記物，大量微小分子核糖核酸(miRNA)和蛋白質生物標記物被發現，甚至有些生物標記物已經在臨床使用。相對的，胞外小體中的小分子代謝物組成或小分子生物標記物則較少被關注，我們應用蛋白質體研究中廣為使用的奈流超高效液相層析(nanoUPLC)-串聯質譜法，分析不同癌症細胞株分泌之胞外小體的代謝體差異，我們發現少數代謝物具有顯著性差異，具有作為癌症轉移生物標記物之潛力。 就如同癌症細胞會利用分泌胞外小體進循環系統與其他器官溝通，腸道菌也會分泌代謝物進入宿主的循環系統，進而影響了宿主的代謝。許多證據表明腸道菌與宿主的代謝與疾病進程相關，在脂肪代謝、心血管疾病的影響上更是極其顯著。我們利用非標靶代謝體學，比較無特定病原(specific-pathogen-free)和無菌(germ free)之ApoE-/-大鼠的血清和肝臟中代謝物的差異，了解腸道菌帶來的代謝體變化，並發現許多腸道菌衍生代謝物與心血管疾病相關。 最後我們開發了一種新的非數據依賴擷取方法，在此方法中我們根據LC-MS的波峰分布挑選一組非等距之前驅物隔離區間，並藉此得到更清楚、更容易進行反褶積(deconvolution)的二級質譜圖譜，從而得到更完整也更可信之代謝物鑑定。 在此研究中，我們將非標靶代謝體應用在不同的生物系統上，包含開發癌症轉移相關之生物標記物與發現心血管疾病相關之代謝物。此外，我們開發了一個新的非數據依賴擷取方法，利用設計過的前驅物隔離區間增進串聯質譜數據的反褶積。
並列摘要
Metabolomics is the study of investigating the small molecules composition of biological samples. Compared to proteomics or genomics, the information metabolomics provides is more related to phenotype. Due to the high diversity of metabolite structures, metabolites act as many different roles in organisms, including energy preserver, building blocks of biomolecules, hormone, neurotransmitter, coenzyme, etc. Biochemists have been devoted to understanding the whole metabolome and the interaction between metabolites and enzymes in the past decade. Unlike targeted metabolomics, untargeted metabolomics utilizes MS/MS comparison with online databases to identified regulated metabolites in an unbiased fashion. We used liquid chromatography-tandem mass spectrometry (LC-MS/MS) based method to study the metabolome in tumor-derived exosomes and ApoE knockout rats to find small molecule biomarkers for organ-tropic metastasis and potential drugs for cardiovascular disease. We further developed a data-independent method, Selected MARgins acquisiTion (SMART), for untargeted metabolomic featuring designed isolation windows. SMART improved MS/MS spectrum deconvolution and resulted in better metabolite identification. Recently, tumor-derived exosomes received tremendous attention because of its crucial role in cancer progression and metastasis. Tumor-derived exosomes were a rich source of biomarkers, some protein and microRNA biomarkers were discovered and even translated into clinical use. However, few studies focused on metabolite composition or biomarkers in exosomes. We performed nanoUPLC-MS/MS-based untargeted metabolomics to analyze the differences between exosomes secreted from cancerous cell lines. We’ve found some metabolites which were significantly changed and could be potential biomarkers for tumor metastasis. As tumors conduct cross-organ communication by secreting exosomes to the circulation system, gut microbiota also tends to secrete metabolites to the circulation system and affect the host’s metabolism. Evidence showed that gut microbiota is associated with several metabolic pathways and disease progression, including the lipid metabolism pathway and cardiovascular disease. To survey the effect of gut microbiota on cardiovascular disease, we applied untargeted metabolomics to study the differences between germ-free and specific-pathogen-free ApoE-/- rats. We have discovered some microbiota-derived metabolites related to cardiovascular disease. Finally, we developed a novel data-independent acquisition (DIA) method named SMART. When using the SMART method, we generated a set of unequal width isolation windows according to the LC-MS peak distribution. Consequently, we could simplify MS/MS spectra and improve peaks deconvolution, and resulted in a more comprehensive and convincing metabolites identification. In this study, we applied untargeted metabolomics on different biological systems to discover tumor-metastasis-associated biomarkers and cardiovascular-disease-related metabolites. Furthermore, we developed a new DIA method, SMART, featuring designed isolation windows and improve MS/MS data deconvolution.