基於奈米探針質譜法分析消化道系統癌症之血清澱粉樣蛋白A異構體模式

Abstract

儘管消化道系統癌症(Gastro-intestinal cancers, GI cancers)的發病率很高，目前仍然沒有合適的生物標記作為診斷，為了解決當前未滿足的臨床需求，我們使用了血清澱粉樣蛋白A，一種在許多癌症類型中濃度會上升的急性期蛋白。血清澱粉樣蛋白A(Serum amyloid A, SAA)是一種急性反應期蛋白。我們團隊先前發現了由24種等位基因異構體組成的異構模式，並使用支援向量機（Support vector machine, SVM）將其整合到分類模型中，展示了其在區分胃癌患者、胃炎和無病患者中的潛在應用性。 在這項工作中，我們建立了一個基於奈米探針的自動純化技術結合基質輔助雷射脫附/游離飛行時間質譜儀（Matrix-assisted laser desorption/ionization mass spectrometry, MALDI-TOF MS）的方法，並用於血清澱粉樣蛋白A異構模式分析。為了評估其在胃癌診斷中的效用，根據臨床實驗室標準協會（Clinical Laboratory Standard Institute, CLSI）和美國食品和藥物管理局（United States Food and Drugs Administration ,US FDA）指南對該方法進行了評估。在論文的第一部分中，藉由市售的血清澱粉樣蛋白A標準品和人體血清庫評估了線性，偵測和定量極限，精密度和干擾物測試的分析項目。我們發現在平方根變換後，從15.6 ng-500 ng（血清澱粉樣蛋白A/內標準品比值為0.0065-54.77）有良好的線性（r2 = 0.9917）、偵測和定量極限為8.32 ng（血清澱粉樣蛋白A/內標準品比值為0.0025），並且該方法在三種不同濃度下的20日重複性和裝置內精密度結果中也是可被接受的（測試χ2 = 54.09, 41.46, 與 54.45 < 臨界χ2 = 55.8)。同時，我們也評估了常見的干擾物（如白蛋白，結合膽紅素和血紅蛋白）的潛在影響，結果表明它們並無在本方法中的自動純化部分造成干擾。在本文的第二部分，我們試圖通過研究血清澱粉樣蛋白A在細胞系模型中的表達來描述血清澱粉樣蛋白A在癌症中的機制。在正常情況和IL-18誘導狀態下，僅在GES-1裂解液（正常細胞株）中偵測到血清澱粉樣蛋白A，而在胃細胞株中均未發現血清澱粉樣蛋白A。但因為樣品中的血清澱粉樣蛋白A含量極低，目前無法進行進一步的使用LC-MS/MS分析以確認血清澱粉樣蛋白A的存在及異構模式。 綜上所述，在詳細的分析項目評估中我們展示了方法中的線性，分析靈敏度，精確度和抗干擾性。其結果表明了該測定足夠穩健，並有可能成為在臨床中進行篩檢的常規工具。

Despite the high incidence of gastrointestinal (GI) cancers, an accurate non-invasive screening tool is lacking. To address this unmet clinical need, we employ serum amyloid A, an acute phase protein that is upregulated in many cancers types. Its variant pattern, composed of 24 allelic variants, was previously discovered by our group and integrated into a classification model using support vector machines (SVM), which demonstrated its potential application for differentiation of patients with gastric cancer from gastritis and disease-free individuals. In this work, we established an automated nanoparticle-based enrichment and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) method for SAA variant pattern analysis. To evaluate its utility in the diagnosis of gastric cancer, the method was evaluated according to the Clinical Laboratory Standards Institute (CLSI) and United States Food and Drugs Administration (US FDA) guidelines. In the first part of the thesis, the analytical merits in terms of linearity, limit of detection and quantification (LoD and LoQ), precision, and interference were evaluated by commercially available SAA and human serum pool. We found good linearity (r2=0.9917) from 15.6 ng-500 ng (0.0065-54.77 in SAA/ISD) after square root transformation, and LoD and LoQ of 8.32 ng (0.0025 in SAA/ISD ratio). The method has acceptable 20-day repeatability and within-device precision at 3 different concentration levels (test χ2 = 54.09, 41.46, and 54.45 < critical χ2 = 55.8). The potential impact of common interferents, such as albumin, conjugated bilirubin and hemoglobin, were evaluated. The results show that they do not interfere with the automated enrichment of this method. In the second part of the thesis, we attempted to delineate the mechanism of SAA in cancer by investigating its expression in cell line models. SAA was only detected in GES-1 lysate (normal cell line) and was not found in gastric cell lines, at both normal and IL-18-induced states. The low amount of SAA in the sample did not allow further LC-MS/MS analysis to confirm the presence of SAA, let alone the variant pattern of SAA. In summary, a detailed evaluation of the analytical merits demonstrated the linearity, analytical sensitivity, precision, and resistance to interference of our assay. Our results demonstrate the robustness of this assay and high potential to be a routine tool in the clinic practice for screening.