脂質體萃取方法開發與鞘脂質於中風之研究 第一部分：利用環戊基甲基醚開發液液萃取法分析脂質體 第二部分：探討血中神經鞘脂質濃度與缺血性中風預後的相關性

Abstract

脂質體學是一門分析生物體內所有脂質種類的學問。作為代謝體學子科學，脂質體學也能呈現生物體的表現型與病理生理狀態。由於脂質的高度複雜性，脂質體的研究仍有許多挑戰。要達到良好的脂質體分析結果，除了要仰賴質譜分析，樣品的萃取方法也會影響結果解釋。儘管目前現有的脂質萃取方法針對大部分的脂質種類都有好的回收率，但是對於能涵蓋脂質的廣泛性仍有改善空間。另外，由於脂質的重要生理角色，利用脂質體學研究方法找尋具有潛力的生物標記也是具有潛力之策略。本論文分成兩部分並以脂質體學的角度來分別針對脂質萃取與鑑別脂質生物標記兩個主題做討論。 論文的第一部分，我們針對脂質體學發展了一個新穎的萃取方法。此萃取法是以1:1:0.725的環戊基甲基醚 (cyclopentyl methyl ether, CPME)–乙醇–水之體積比例所形成的雙層萃取系統。本萃取方法是利用上層的有機層來萃取大部分種類的脂質，並且同時發現在溶血磷脂 (lysophospholipids)、二酸甘油酯 (diacylglycerol)、與鞘氨醇磷酸鹽 (sphingosine-1-phosphate, S1P)，比起甲基三級丁基醚萃取法有較好的萃取率。另外，不能被溶解的殘餘物在離心後會被沉澱在萃取微管的底部，這使得整個實驗步驟變得更簡單，並且適用於高通量分析。由於能夠涵蓋更廣泛的脂質萃取，特別是較為親水種類的脂質，加上其溶劑組成皆具備有綠色化學的特性，環戊基甲基醚萃取法會是一個在例行脂質體分析上有優勢的萃取方法。 第二個部分，我們將脂質體學應用在中風研究，並找尋重要的生物標記來幫助診斷。中風，特別是缺血性中風，雖然已有多種治療方案，但仍是造成全球人口終生失能的主要病因。神經鞘脂，為脂質體中的其中一群，在生物體扮演非常重要的角色。我們先前研究中已經證明，在大腦中動脈阻塞的動物模型中，神經鞘脂會對急性缺血性腦損傷的反應產生顯著濃度變化，暗示著神經鞘脂的代謝途徑也許可作為治療缺血性中風的標的。為了要更進一步瞭解神經鞘脂在臨床上的角色，我們以急性缺血性中風 (acute ischemic stroke, AIS)病患作為研究對象，觀察其血液中神經鞘脂的濃度變化，並且檢示這些變化是否與AIS病患的預後有關。我們從臺大醫院收集30個非中風受測者血清樣品做控制組，也收集了87個AIS病患在中風後48小時與48到72小時的血清樣品。使用超高壓液相層析結合搭配電噴灑游離法的串聯式質譜儀分析血清中不同碳鏈長度的神經醯胺 (ceramide) 與鞘氨醇磷酸鹽濃度。結果顯示與控制組相比，較長長碳鏈神經醯胺與鞘氨醇磷酸鹽在AIS病人的血清中濃度較低，而長碳鏈神經醯胺則是比較高。重要的是，在調整其他潛在的臨床干擾因子後，從48到72小時的血清樣品中可觀察到高濃度的Cer(d18:1/18:0)、Cer(d18:1/20:0)和Cer(d18:1/22:0)，與病人預後不好(定義為經歷AIS三個月後之改良式雷氏量表評分≥2)有重要的關聯，暗示著這些神經醯胺可作為之後評估AIS病患的潛在預後標的。 近年來動脈血管內血栓移除術 (endovascular thrombectomy, EVT)已成為急性大血管阻塞的AIS病患重要的替代治療法，因此我們進一步觀察經歷EVT的AIS病患其血中神經醯胺與鞘氨醇磷酸鹽的濃度變化。在此研究中，我們收集75位經歷EVT手術的AIS病患以及19位控制組血漿，並且利用超高壓液相層析結合搭配電噴灑游離法的串聯式質譜儀分析。為了要檢驗這群病人的血中神經醯胺濃度是否與臨床預後有相關性，我們分別在三個時間點收集病患的血液樣品：手術前 (T1)、手術後當下 (T2)、與手術後24小時 (T3)。研究結果顯示，與控制組相比，三個時間點的血中長碳鏈神經醯胺 Cer(d18:1/16:0) 濃度、T1與T3的血中 Cer(d18:1/18:0) 濃度、T1的血中 Cer(d18:1/20:0) 濃度、與T1的血中較長長碳鏈神經醯胺 Cer(d18:1/24:1) 濃度在缺血性中風病患都較高，但三個時間點的血中鞘氨醇磷酸鹽濃度則較低。根據多變項分析的結果顯示，在調整其他潛在的臨床干擾因子後，血中較高濃度的Cer(d18:1/16:0)與Cer(d18:1/18:0)在三個時間點，Cer(d18:1/20:0)在T1與T2，與Cer(d18:1/24:0)在T2都仍與功能性預後不好有顯著關聯。這個結果暗示長碳鏈神經醯胺，特別是Cer(d18:1/16:0)與Cer(d18:1/18:0)，可以作為經歷EVT手術的AIS病患之早期預後診斷標的。 在本論文中，我們提出了一個適用於脂質體學研究的新的脂質萃取方法，並且針對缺血性中風找到了一個有潛力的預後生物標記。我們期望這個新穎的萃取方法可以適用在各種脂質體學的研究上，而被發現的標記則可以有益於中風病人的臨床照護。

Lipidomics is the systemic analysis of total lipid in biological systems. As the subdiscipline of metabolomics, lipidomics could exhibit the phenotype and pathophysiological status of an organism. Due to their great structural diversity, lipid analysis remains a challenge. Sample preparation techniques, as well as mass spectrometric analysis, are important to obtain comprehensive lipidomic results, influencing the biological interpretation. Although good recoveries of different lipid subclasses could be obtained for most current lipid extraction methods, the lipid coverage of these methods still needs to be improved. In addition, due to their versatile biological functions, lipidomics is considered a promising strategy for identifying biomarkers for diseases. In this dissertation, I will divide it into two parts, including the development of a lipid extraction method and the identification of lipid biomarkers for stroke. First, we developed a novel extraction method for lipidomics studies. This two-phase system contains cyclopentyl methyl ether (CPME), ethanol, and water, giving the final volume ratio of 1:1:0.725 (v/v/v). The mixture of CPME–ethanol–water used the upper phase to extract most of the major lipid classes and had higher recoveries of lysophospholipids, diacylglycerols, and sphingosine-1-phosphates, than the methyl tert-butyl ether (MTBE) extraction method. In addition, non-soluble debris is precipitated at the bottom of the extraction tube after centrifugation, making the whole experimental procedure more simple and suitable for high-throughput analysis. With expanded coverage of total lipid, including polar lipid classes, and the green solvent properties, the CPME method is represented as an attractive method for routine lipidomics analysis. In the second part, we applied a lipidomic approach to the analysis of stroke, and found potential biomarkers to help diagnose. Stroke, especially ischemic stroke, is the leading cause of lifelong disability worldwide, despite multiple available treatment options applied. Sphingolipids, one of the lipid subclasses, play important roles in biological systems. Our previous study demonstrated that sphingolipids significantly changed in response to acute ischemic brain injury in an animal model of middle cerebral artery occlusion (MCAO), suggesting that sphingolipid metabolism may serve as a potential therapeutic target for ischemic stroke. To further understand their potential roles in clinical, we investigated the levels of sphingolipids in acute ischemic stroke (AIS) patients and examined whether these lipids were associated with AIS outcome. We collected 87 plasma samples from patients with AIS at <48 and 48-72 h post stroke and 30 nonstroke controls at National Taiwan University Hospital (NTUH). The plasma concentrations of ceramides with different chain lengths and S1P were measured by using ultra-high-pressure liquid chromatography-electrospray ionization tandem mass spectrometry (UHPLC-ESI-MS/MS). The results showed that compared with nonstroke controls, very-long-chain ceramides and sphingosine-1-phosphate (S1P) were significantly decreased in AIS patients, while long-chain ceramides were significantly increased. Importantly, higher levels of Cer(d18:1/18:0), Cer(d18:1/20:0), and Cer(d18:1/22:0) at 48-72 h were significantly associated with poor functional outcomes, defined as a modified Rankin Scale (mRS) score ≥2 at 3 months after AIS, after adjusting for potential clinical confounders, suggesting that these ceramide species could be potential prognostic markers for patients with AIS. In addition, we further profiled ceramides and S1P in AIS patients subjected to endovascular thrombectomy (EVT) treatment, which has become the alternative therapy for AIS patients with acute large vessel occlusion. We investigated the same analytes in 75 AIS patients undergoing EVT and 19 age- and sex-matched controls by using UHPLC-ESI-MS/MS. To examine the association between ceramide levels and clinical outcomes among AIS patients undergoing EVT, plasma samples were collected from AIS patients who underwent EVT before (T1), immediately after (T2), and 24 hours after (T3) the procedures. The result showed that the plasma levels of long-chain ceramides Cer(d18:1/16:0) at all three time points, Cer(d18:1/18:0) at T1 and T3, and Cer(d18:1/20:0) at T1, and very-long-chain ceramide Cer(d18:1/24:1) at T1, in AIS patients were significantly higher than those in the controls, while the plasma levels of S1P in AIS patients were significantly lower at all three time points. Multivariable analysis showed that higher levels of Cer(d18:1/16:0) and Cer(d18:1/18:0) at three points, Cer(d18:1/20:0) at T1 and T2, and Cer(d18:1/24:0) at T2 remained significantly associated with poor functional outcomes after adjustment for potential confounding factors. These data suggest that the levels of plasma long-chain ceramides, especially Cer(d18:1/16:0) and Cer(d18:1/18:0), could be early prognostic indicators for AIS patients undergoing EVT. In summary, we proposed a new lipid extraction method for lipidomic studies and identified potential prognostic markers for ischemic stroke. We anticipate the novel extraction method could be applied to various lipidomic studies, and the identified markers could benefit clinical care for stroke patients.