請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/77924
標題: | 利用個人化膜蛋白質體學與多重反應監測質譜法探勘胃癌診斷生物指標 Mining biomarkers for gastric cancer diagnosis by personalized membrane proteomics and multiple reaction monitoring mass spectrometry analysis |
作者: | Tai-Du Lin 林泰都 |
指導教授: | 翁啟惠(Chi-Huey Wong) |
關鍵字: | 胃癌,生物標記蛋白,蛋白質體學,重反應監測質譜法, Gastric cancer,Biomarker,Membrane proteome,multiple reaction monitoring mass spectrometry analysis, |
出版年 : | 2017 |
學位: | 博士 |
摘要: | Gastric cancer (GC) is one of the leading causes of cancer-related deaths worldwide. At present, however, the reported biomarkers, such as CEA, CA19-9, and CA72-4, have low sensitivity and specificity for diagnosis of gastric cancer (GC). Development of molecular biomarker for early diagnosis of GC is urgently needed. The membrane proteins hold promises for cancer detection because most FDA-approved cancer biomarkers are secreted membrane proteins. In attempt to identify GC biomarker with better diagnosis ability and reliability, in this study, we established a label-free discovery-through-verification proteomics pipeline to discover and verify biomarker candidates.
In the discovery phase, membrane proteomics analysis combining an efficient gel-assisted digestion protocol and a label-free quantification method were performed for 24 pairs of tumor and adjacent normal tissues from GC patients, including 8 and 16 from early stage (I and II stages) and late stages, respectively. Under 1.7% false discovery rate (FDR) and 95% confidence interval, the analysis quantified 1746 proteins, including the previously reported GC biomarkers: CEA, CA15-3, and CA125. We further filtered 35 potential biomarker candidates based on the following criteria: (1) Experimental evidence of presence in serum or informatic evidence showing secretion ability; (2) High frequency of overexpression (tumor/normal ratio ≧2) among 50% of total patients or stage-1 patients were candidate biomarkers for high detection sensitivity or early diagnosis of GC, respectably. By Western blotting (WB) analysis in normal and GC cell lines, only five proteins, BRI3BP, CLDN3, EPCAM, MME, and PLSCR1 have good quality of antibody to show positive staining. CLDN3, EPCAM, and PLSCR1 show significant overexpression in cancer cells. However, many of them do not have high-quality antibody or immunoassay available for quantification. In the second part of verification stage, the 35 biomarkers were verified by optimized MRM-MS approach by the following steps: (1) Best transitions for each candidate were selected from our in-house membrane proteome spectral libraries, SRMAtlas or PeptideAtlas; (2) selected transitions were tested on mixed gastric cancer cell lines and tissues to evaluate their reliability. A total of 815 transition from 163 peptides were selected for the 35 candidate proteins; 32 candidates show overexpression in at least 50% of patients or at least two stage-I patients except COX6BA and MME. The MRM-MS quantification result revealed that 8 candidates including ABHD12, CLDN3, DHCR7, EPCAM, GPRC5A, PLSCR1, SE1L1, and TMCO1 show >2-fold overexpression in tumor of more than 75% patients. To further evaluate the clinical relevance of these candidates, 3 candidates, EPCAM, CLDN3, and PLSCR1, with available antibodies were examined by tissue microarray (TMA) in 97 GC patients. These 3 candidates exhibited excellent discrimination between GC and normal mucosa (AUC range from 0,818 to 0.892). In addition, combined use of these 3 candidates as the biomarker panel has the best AUC (0.964). The results further showed that overexpression of EPCAM and CLDN3 not only occur in tumor part but also in the precancerous lesion, intestinal metaplasia (IM). In contrary, PLSCR1 is only significantly overexpressed in tumor tissue. These results suggested the promise of this biomarker panel for prediction and early diagnosis GC. Furthermore, the functional roles of the 8 candidates in GC tumorigenesis were studied by the protein-protein interaction (PPI) network constructed based on our membrane proteomic dataset and the TCGA Stomach Adenocarcinoma transcriptome datasets. These biomarker candidates and their PPI neighbors are majorly enriched in VEGF, MAPK and PI3K/AKT signaling pathway which had been reported to be involved in GC and PI3K, AKT, VEGFR, and EGFR have been targeted for treatment of GC. The results demonstrated the power of tissue membrane proteomics for the discovery of valuable biomarker candidates for prediction and early diagnosis of GC. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/77924 |
DOI: | 10.6342/NTU201702865 |
全文授權: | 有償授權 |
顯示於系所單位: | 生化科學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-106-D99b46008-1.pdf 目前未授權公開取用 | 4.83 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。