趙木榮CHAO, MU-RONG胡鈞HU, JYUNJYUNHU職業安全衛生學系碩士班2026-10-052030-07-1520252025-05-29https://handle.ncl.edu.tw/11296/rgakp7https://scholars.csmu.edu.tw/handle/123456789/20540順鉑 (Cisplatin) 為臨床常用的鉑類抗癌藥物，透過與DNA中的核苷形成鍵結物 (Adducts)，進而阻礙DNA複製與轉錄，引發細胞凋亡。然而，其與DNA交互作用產生的鍵結物結構仍未完全揭示。本研究結合高解析質譜 (HRMS) 與FeatureHunter軟體，開發一套系統性篩選與鑑定順鉑誘導DNA鍵結物的非目標性數據處理策略，並以去氧核醣核苷、去氧核醣核苷酸和三種寡核苷酸作為模型，並探討其鍵結特性。並進一步驗證於小牛胸腺DNA (CT-DNA) 樣本中，以驗證方法於真實DNA樣本之可行性。本實驗藉由FeatureHunter軟體搜尋順鉑所誘導之DNA鍵結物，並探尋它們在CID與HCD碎裂模式下所造成的各種斷片特徵，以利解析分子結構特性。首先針對各式去氧核醣核苷、去氧核醣核苷酸和寡核苷酸的鍵結物模型樣本進行解析，由訊號排名前二十強的adducts的斷片特性顯示：順鉑所誘導之DNA鍵結物具有鉑同位素質量差與離子源氨基碎裂 (In-source NH₃) 的碎裂特徵，且可透過兩個去氧核醣加一個磷酸基團和氨基 (2dR+P+NH3)、一個去氧核醣和一個氨基 (dR+NH3) 的中性丟失特徵及電荷數判定鏈內、鏈間交聯或單鍵結物。接著應用FeatureHunter搭配上述中性丟失與同位素質量差特性於CT-DNA分析數據解析，在符合順鉑DNA鍵結物特性且強度排名前二十的離子中可鑑定出6個已知鍵結物與14個新的未知鍵結物。進一步透過OPLS-DA與Volcano plot圖進行統計分析，針對不同鍵結類型 (單鍵結、鏈內交聯與鏈間交聯物) 中具代表性的候選離子，挑選出4個具高辨識潛力的離子，作為未來評估順鉑化療之潛在生物標誌物。本研究成功展示順鉑與不同DNA結構之反應差異，並建立結合非目標質譜與統計分析的高效數據處理流程，提供全面探索順鉑-DNA鍵結物的嶄新工具與策略。Cisplatin is a widely used platinum-based anticancer drug in clinical settings. It exerts its cytotoxic effects primarily through the formation of DNA adducts by binding to nucleosides, thereby disrupting DNA replication and transcription and ultimately inducing apoptosis. However, the structural details of the cisplatin-DNA adducts remain incompletely elucidated. In this study, we integrated high-resolution mass spectrometry (HRMS) with the FeatureHunter software to develop a systematic, non-targeted data-processing strategy for screening and identifying cisplatin-induced DNA adducts. Deoxynucleosides, deoxynucleotides, and three types of oligomers were used as model compounds to investigate binding characteristics. The method was further validated using calf thymus DNA (CT-DNA) samples to assess its applicability in real biological matrices.Using FeatureHunter, we searched for cisplatin-induced DNA adducts and analyzed their fragmentation characteristics under both collision-induced dissociation (CID) and higher-energy collisional dissociation (HCD) modes to deduce structural features. In the initial analysis of model adducts formed with various deoxynucleosides, deoxynucleotides, and oligomers, the top twenty most intense signals revealed fragmentation features typical of cisplatin-DNA adducts, including platinum isotopic delta mass and in-source ammonia loss (NH₃). Characteristic neutral losses such as 2-deoxyribose with phosphate group and ammonia (2dR+P+NH₃) and deoxyribose with ammonia (dR+NH₃), along with charge state information, were used to distinguish monoadducts, intrastrand and interstrand crosslinks.Subsequently, the FeatureHunter platform, combined with these neutral loss and isotopic features, was applied to CT-DNA data. Among the top twenty most intense ions exhibiting cisplatin-adduct characteristics, six known and fourteen novel unknown adducts were identified. Further statistical analyses using OPLS-DA and volcano plots enabled the selection of four candidate ions with high discriminative potential across different binding types (monoadducts, intrastrand, and interstrand crosslinks), which may serve as potential biomarkers for evaluating cisplatin chemotherapy response. This study successfully demonstrates the differential reactivity of cisplatin with various DNA structures and establishes an efficient data-processing workflow combining non-targeted mass spectrometry and statistical analysis, offering a novel tool and strategy for comprehensive cisplatin-DNA adduct exploration.順鉑DNA鍵結體學天然同位素混成式四極柱/線性離子阱/軌道阱高解析質譜儀FeatureHunterDDA-MS2CisplatinDNA adductomicsNatural isotopeQ-LIT-Orbitrap MSData-dependent acquisition-MS2thesis