癌症仍然是一種死亡率很高的重要疾病。臨床上檢體採樣及侵入性的考量,使用液態活檢來檢測、診斷和監測癌症有其優勢。液態活檢同時具有非侵入性,使得在所有臨床階段的樣本採集變得更加容易和可能。而蛋白質體學分析被認為在開發生物標記方面有巨大潛力,並能為生物學研究的進展提供了寶貴的見解。因此,結合蛋白質體學分析和液態活檢在癌症研究方面有助於科學家取得重大進展。本研究整合液態活檢和蛋白質體分析,分別在兩個關鍵領域有所發現。(一) 訓練人員使用 LC-MS/MS 檢測甲狀腺球蛋白的方法,在本次研究中顯現了使用LC-MS/MS作為監測甲狀腺癌的方法的前景。(二) 分析自體抗體圖譜,用以開發生物標記並深入生物機制的研究。研究結果顯示,新的LC-MS/MS的方法有改進LC-MS/MS對於樣品分析的精準度,但需要更進一步的驗證。另一方面在自體抗體圖譜分析研究中我們發現雌激素透過 PRKCZ 刺激的信號傳導與半胱天冬酶介導的細胞骨架蛋白裂解這兩個生物路徑與ICIs 的反應存在相關性。同時也觀察到 CASP6 和雌激素相關基因具有作為預後生物標記的潛力。 透過本次的研究,針對LC-MS/MS 檢測甲狀腺球蛋白,我們從臨床檢驗的角度出發完成了對於這項方法學的人員訓練的檢驗。而關於 ICIs 反應的生物標記的開發,我們鑑定出 10 種中心自體抗體,雖然結果顯示這些中心抗體目前無法作為生物標記,但它們有潛力作為未來開發預測模型的依據。
Cancer continues to significantly impact global health due to the high mortality rates associated with the disease. Adopting liquid biopsy as a method for cancer detection, diagnosis, and monitoring is notably advantageous due to its non-invasive nature, rendering sample collection simpler and less strenuous. The proteomics analysis holds significant potential in discovering biomarkers and contributes valuable insights to the progress of biological research. As a result, the combination of proteomics analysis and liquid biopsy has the potential to make significant strides in addressing cancer-related challenges. This study integrates liquid biopsy and proteomic analysis in two key areas. The first involves personnel training in thyroglobulin measurement using LC-MS/MS, showing promise as a biomarker for monitoring thyroid cancer. The second area pertains to identifying autoantibody profiles for developing biomarkers and in-depth mechanistic research. LC-MS/MS training results show that the modified method is an improvement, but further validation in more samples is needed. In the analysis of the autoantibody profile, the study found that estrogen-stimulated signaling through PRKCZ and caspase-mediated cleavage of cytoskeletal proteins were related to ICIs response. It was also observed that CASP6 and estrogen-related genes hold potential as predictive biomarkers. Additionally, the identification of potential ICIs response biomarkers revealed ten hub autoantibodies. Although these hub antibodies do not currently meet the criteria for biomarkers, they exhibit promise for developing a prediction model.