陳玉如陳頌方Chen, Yu-JuChen, Sung-Fang陳冠甫Chen, Guan-Fu2023-12-082028-06-012023-12-082023https://etds.lib.ntnu.edu.tw/thesis/detail/23ab252cd9378e952cfddcd3eae1ec8a/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/120963以質譜法為基礎的蛋白質體學是一個步驟繁雜，且須搭配高品質液相串聯質譜儀的技術，使得在臨床蛋白質體學上研究超微量樣品時依舊是個挑戰。已知在微量到超微量樣品的前處理流程中，低樣品損失決定分析的靈敏度和蛋白質體覆蓋率。然而，以目前方法學處理超微量樣品上常伴隨著無可避免地顯著樣品損失，也造成在深度分析超微量樣品上是個難以解決的課題。為了解決在超微量樣品製備上的瓶頸，吾人志在開發精簡的樣品前處理流程並結合等壓同位素標記法以利訊號增幅在小於兩百顆細胞的超微量實驗。此研究的第一部分，透過人類非小細胞肺癌細胞取得胜肽後，再經由序列稀釋的胜肽來模擬及評估等壓同位素標記法應用於超微量樣品的可能性。結果顯示0.2奈克的胜肽在搭配40奈克的胜肽增幅技術下，可以鑑定到6177條的同位素修飾胜肽和986個無定量缺失值的蛋白。此後，吾人透過此實驗模型來優化超微量樣品的等壓同位素標記流程，結果顯示相較於過往數據94%增幅在同位素修飾胜肽其鑑定數從6177上升至11954，且有70%增幅在無定量缺失值的蛋白其鑑定數從986上升至1684。此研究的第二部分，開發精簡的樣品前處理流程並且結合優化的等壓同位素標記法以利於製備數以百計的人類非小細胞肺癌細胞樣品。吾人精簡化原先的液相水解流程，包括實驗反應的體積由20微升縮小至5微升，並省略同位素標記反應前的緩衝溶液置換流程，這些改動有效的增加在製備少量細胞樣本的樣品回收率。符合預期，以手動處理200顆人類非小細胞肺癌細胞樣品上獲得68%增幅在同位素修飾胜肽其鑑定數由3170上升至5334，和20%增幅在無定量缺失值蛋白其鑑定數由720上升至868。此外，皮爾森相關係數分析也透露此精簡化流程在處理相同細胞數的實驗下具有高度再現性，且在三重複的實驗中具有小於20%的變異係數。此研究的第四部份，吾人將優化的流程應用於小鼠外周血單核細胞（PBMC）和兩種與小鼠相關的細胞系巨噬細胞系（RAW 264.7）和Lewis肺癌細胞系（LLC1）以評估方法的性能。在200個細胞中，數據可靠地根據其蛋白質的豐度將細胞進行分群，並在生物信息學分析中揭示了它們的蛋白質功能差異。此研究的第五部分，將優化的流程與微型化平台-微流控晶片相結合，以發展高度精簡的前處理流程，晶片的整合使細胞分選、計數、成像和樣本處理得以在單個封閉系統下執行，有助於達到根據等壓同位素標記的單細胞蛋白質體學分析。通過在90分鐘的液相層析並結合200倍的胜肽訊號增幅技術，各個非小細胞肺癌的單細胞樣本可以鑑定到8123個同位素修飾胜肽及1230個蛋白鑑定數。此外，在三重複分析中，大於90%的共同鑑定蛋白是無定量缺失值的。最後，定量結果顯示參考樣本（2個細胞）較單個細胞樣本之間的平均同位素修飾胜肽豐度增加了1.6倍，意味著此方法具有客觀定性及定量各單細胞樣本的能力。Mass spectrometry-based proteomics required a multiple-step workflow followed by high-performance LC-MS/MS analysis, posing challenges for clinicalproteomics of limited amounts of samples. For micro-to-nanoscale (micro-to-nanogram sample) proteomics, a sample preparation protocol with minimum sample loss determines the profiling sensitivity and proteome coverage. However, significant sample loss in the current methodologies of nanoscale sample preparation is inevitable and problematic.To address the challenges for nanoscale sample processing, we aim to develop a streamlined sample preparation method incorporating the TMT isotopic labeling for signal amplification in the minute samples (≤ 200 cells). First, the diluted peptides from the PC9 cell line are used to simulate the labeling reaction toward the nanoscale samples. It reveals that 6177 labeled peptides and 986 proteins with no missing value are identified in the 0.2 ng dilute-based PC9 sample by 40 ng peptide to boost the signal. Later, applying the optimized labeling workflow on the same model, it acquires a 94% increase for labeled peptide identification from 6177 to 11954, and a 70% increase for proteins with no missing value from 986 to 1684. Second, streamlined sample preparation is developed and integrated with the optimized labeling workflow for processing hundreds of PC9 cell samples. In the modified in-solution digestion workflow, reducing the working volume from 20 µL to 5 µL and bypassing the buffer reconstitution step before the TMT labeling, effectively improving the sample recovery in the low-cell experiment. As expected, the modified in-solution digestion workflow allows a 68% increase in the identification of the labeled peptides (3170 to 5334 peptides), and a 20% increase from 720 to 868 in the proteins with no missing value from 200 PC9 cells. Furthermore, the Pearson correlation reveals high reproducibility between samples of the same cell amount and has less than 20% coefficient of variation in the triplicate analysis. Lastly, the optimized assay was applied to mouse peripheral blood mononuclear cells (PBMC) and two mouse-related cell lines (macrophage cell line (RAW 264.7) and Lewis lung carcinoma cell line (LLC1)) to evaluate the method performance. By 200 cells, the data robustly classify cells based on their protein abundance and reveal their protein functional difference in the bioinformatic analysis.In the second part, we integrated the optimized protocol with a miniaturized platform, microfluidic chip, for highly streamlined TMT-based single-cell proteomics sample preparation. On-chip handling enables cell sorting, counting, imaging, and sample processing on a single enclosed device. Integrating the microfluidic chip with a 200-fold carrier peptide sample in a 90-min LC gradient shows 8123 peptides from 1230 proteins in individual PC9 single cells. Moreover, higher than 90% of proteins were commonly identified without missing values in triplicate analysis. Lastly, the quantification results show a 1.6-fold in the averaged TMT-peptide abundance between the reference channels (2 cells) and single-cell channels.等壓同位素標記法訊號放大技術微流控晶片單細胞蛋白質體學TMT isotopic labelingSignal amplificationMicrofluidic chipSingle-cell Proteomicsetd