Study On N-Glycomics In Malignancies Of Digestive System

Every year, roughly 7.2 to 7.5 million people worldwide die from cancer, 55% of which occur in developing countries. The survival of cancer is highly dependent on the pathological stage upon diagnosis, early detection of cancer greatly increases five-year survival rate. However, current tumor markers have poor sensitivity and specificity or they are not elevated early enough in the cancer process to be useful for screening. Newer markers are needed to overcome these problems and allow diagnosis of cancer at an earlier stage. Glycomics is the comprehensive study of all glycans expressed in biological systems. Glycans are an important class of post-translational modifications. Glycosylation is therefore highly sensitive to the biochemical environment and has been implicated in cancer. We propose that these cancer-associated glycan variants be exploited for biomarker development aimed at diagnosing early-stage cancer. DNA sequencer-assisted fluorophore - assisted carbohydrate electrophoresis (DSA-FACE) was proved in the previous study to be a rapid, highly sensitive and high throughput method for N-glycome profiling, making it is available to apply N-glycomics to clinical diagnosis and prediction of cancer.This study aims to construct and assess the diagnostic value of N-glycan based diagnostic model in large sample clinical case control study for gastrointestinal cancer including hepatocellular carcinoma, colorectal cancer and gastric cancer.Part 1: N-glycomic study for early diagnosis and its mechanism in HCC The part study aims to identify and validate the diagnostic values of N-glycan based diagnostic model in hepatocellular carcinoma (HCC) and to uncover the underlying molecular mechanism. Multiparameters diagnostic models were constructed based on N-glycan alterations with logistic step-wise regression and the receiver operating characteristic curve analysis. A total of 698 subjects including 477 HBV-related HCC, 62 liver cirrhosis, 80 HBV carriers and 79 healthy controls were recruited as prospective groups. N-glycome of serum glycoprotein was profiled by DSA-FACE. Western blotting and RT-PCR were used to analyze the expression of GDP-fuc-Tr involved in core-fucosylation modification. The result showed that the sensitivity of two diagnostic model were both over 80% in HCC group. The level of total core fucose residues was increased significantly in HCC; however, the mRNA expression of GDP-fuc-Tr was no significant difference between tumor tissues and adjacent noncancerous tissue. The diagnostic models improve the efficacy in HCC diagnosis, but the mechanism of glycosylation has long remained obscure.Part 2: Identification and assessment of new biomarkers for colorectal cancer with serum N-glycan profilingThis study aims to identify and validate the diagnostic values of N-glycan markers in colorectal cancer (CRC) and to uncover the underlying molecular mechanism. A total of 347 subjects including CRC, adenoma and healthy controls were divided into training (n=287) and retrospective validation group (n=60). N-glycan profiling of serum was analyzed by DNA sequencer-assisted flurophore-assisted carbohydrate electrophoresis (DSA-FACE). N-glycan markers were identified and two diagnostic models were constructed and validated based on DSA-FACE with logistic step-wise regression and the receiver operating characteristic curve analysis. Their diagnostic performances were further assessed in another independent prospective (n=43) and follow-up (n=20) cohort. Lectin blot and RT-PCR were used to analyze the total core-fucosylated residues and molecular expression involved in core-fucosylation modification. Two diagnostic models nominated as CRCglycoA and CRCglycoB were constructed to differentiate CRC from normal and adenoma respectively. The area under the ROC (AUC) of both CRCglycoA and CRCglycoB were higher than that of CEA (CRCglycoA: 0.92 vs. 0.81; CRCglycoB: 0.81 vs. 0.73). The sensitivity of CRCglycoA was improved 21.7% to 25% and accuracy improved 11.63% to 18% in training, retrospective and prospective groups compared to that of CEA. The sensitivity of CRCglycoB was improved 20-28.23%. Both the N-glycan markers and diagnostic models were reversed after curative surgery. The level of total core fucose residues and fucosyltransferase (Fut8) were decreased significantly in CRC. The N-glycan markers based diagnostic models are new valuable non-invasive alternatives for CRC identification. Decreased Fut8 might be responsible for the decreased level of total core-fucosylated modification in both tissue and serum of CRC.Part 3: serum N-glycan profiling of patients with gastric cancerTo compare the serum N-glycan profiling among gastric caner, ulcer and healthy control try to searching for new biomarkers for gastric cancer. A total of 184 subjects including gastric caner, adenoma and healthy controls were recruited. N-glycan profiling of serum was analyzed by DNA sequencer-assisted flurophore-assisted carbohydrate electrophoresis (DSA-FACE). The diagnostic model was constructed based on N-glycan alterations with logistic step-wise regression and the receiver operating characteristic curve analysis. The result showed that peaks5, 9 in gastric cancer group were higher than those in the control groups; Peaks 2,3,4,6 and 7 were lower in gastric cancer than those in control groups (p<0.05). The area under the ROC (AUC) of GCglyco and logp5/4 were 0.937 and 0.897 respectively. Compared with routine clinical tumor markers, The Accuracy of GCglyco was improved 24%. Logp5/4 was a gradually increasing as tumor pathological grading. The diagnostic models are promising and innovative in gastric cancer diagnosis and progression monitoring.The improvements of diagnostic efficacies confirm again are more powerful and promising to be a noninvasive serum maker in early cancer diagnosis. Future goal is to investigate the relevant regulatory molecules concerning the N-glycan markers, so as to make further improvement in cancer study from both the mechanism and diagnostic efficacy points of view.