The Characteristic Changes And Significance Of Abnormal Glycosylation Of N-glycans In Ovarian Carcinoma

Ovarian cancer is hormone dependent tumor,and it is the leading cause of death in gynecological tumors.Some 70% of women with ovarian cancer present at stage III or IV,the five-year survival rate is still only 20%-40% despite the increasing knowledge about surgery and chemotherapy.When caught at stage I,the five-year survival rates are as high as 90%.Considering the prognosis of advanced ovarian cancer,we need a noninvasive and effective method to improve the diagnostic efficiency of ovarian cancer.It is unlikely that a single biomarker will reach a specificity of 99.6%,positive predictive value of 10%,and sensitivity higher than 75% in screening ovarian cancer.Pathological diagnosis is the gold standard;imaging technology can help to detect ovarian cancer.However,they are not suitable for census because of high cost.Tumor markers are glycoprotein,which composed of oligosaccharide chains and chains of amino acid residues by glycosidic bond covalent connection.At present,CA125,HE4,CA724,CA199 are commonly used in the diagnosis of ovarian cancer.CA125 is a kind of ovarian cancer associated antigen which is most widely used,However,it has high false positive rate and serum levels may be in the high range of 50%-60% in symptomatic stage I patients.Human epididymis protein 4(HE4)is a glycoprotein expressed in epithelial ovarian cancer;it doesn't rise in a lot of gynecological benign diseases.It has higher sensitivity and specificity in the diagnosis ofovarian cancerthan CA125(especially in ovarian serous cystadenocarcinoma and endometrioid adenocarcinoma).We need to seek early ovarian cancer urgently in clinical.Therefore,seeking for screening ovarian cancer diagnosis and follow-up of high sensitivity and high specificity of non-invasive biological markers will face significant challenges and explore the new pathogenesis of ovarian cancer may provide a new early warning signals for clinical.Glycosylation is one of the most common post-translational modifications.Nearly 70 percents of the protein in the body is a kind of glycoprotein,and it is one of the most important regulating ways in many cells,membrane and secreted proteins.During progressions of the tumor,glycans play an important role in major pathophysiological events.In recent years,the rapid development of proteomics and glycomics research provides a new opportunity for the study of tumor biology and tumor markers.The characteristic of tumor associated carbohydrate changes has important implications for cancer diagnosis,monitoring and prognosis assessment.With the development of DNA sequencer-assisted fluorophore-assisted carbohydrate electrophoresis(DSA-FACE),highperformance liquid chromatography(HPLC),matrix-assisted laser desorption/ ionization time of flight mass spectrometry(MALDI-TOF-MS)and lectin microarray,which providing us rapid,high sensitive,high throughput and reliable quantitative methods for N-glycan profiling.This study aims to evaluate of ovarian cancer patient's serum,tissue(cancer-cancer side)in order to identify the abnormal glycosylation of pathogenic mechanisms,explore the ovarian cancer occurrence and development mechanisms of biological characteristic changes of the N-glycans of sugar molecules,seek non-invasive diagnostic methods to improve the early diagnosis of ovarian cancer.In this study,we analyzed serum N-linked glycan and identified specific structures of N-glycan in ovarian cancer by DSA-FACE.After serum N-glycan profiling,two mathematic diagnostic models were constructed with statistical analysis and step logistic regression.Risk ovarian malignancy algorithm(ROMA),CA125,HE4 and the two diagnostic models were retrospectively evaluated.By using lectin blot,RT-PCR,we determined the diagnostic value of the N-glycan characteristic changes in the diagnosis of ovarian cancer and the disease monitoring.We recruit 70 cases of ovarian cancer and 111 cases of ovarianbenign tumor,137 cases of normal control,compared with ovarian cancer characteristic of N-linked glycan group spectra change.ROC curve was used to assess the effect of the abundances of some kinds of N-glycans to diagnosis and differential diagnosis.We found that in ovarian cancer group the abundance of Peak1(NGA2F),Peak9(NA3Fb)and Peak12(NA4Fb)was increased,Peak6(NA2F)was decreased.The area under the ROC curve(AUC)of Peak1 was 0.754,corresponding to the best cut-off point of sensitivity and specificity were 57.1% and 83.8%respectively.The AUC of Peak9(NA3Fb)was 0.795,corresponding to the best cut-off point of sensitivity and specificity were 62.9% and 82.7% respectively.The AUC of Peak12(NA4Fb)was 0.830,corresponding to the best cut-off point of sensitivity and specificity were 72.9% and 81.1% respectively.The AUC of Peak6(NA2F)was 0.878,corresponding to the best cut-off point of sensitivity and specificity were 78.4% and 81.4%respectively.The AUC of CA125 was 0.860,corresponding to the best cut-off point of sensitivity and specificity were 82.9% and 69.4% respectively.The AUC of HE4 was 0.879,corresponding to the best cut-off point of sensitivity and specificity were 54.3% and 99.1%respectively.The AUC of Model_Sugar(0.903),Model_Complex(0.963)and ROMA(0.913),were higher than those of CA125 and HE4.Model_Sugar had 90.1%specificity,which was increased 20.7% and 6.3% compared with CA125 and ROMA.The sensitivity of Model_Sugar(82.9%),which was increased 28.6% and 2.9% compared with HE4 and ROMA.The accuracy of Model_Sugar(82.3%)was increased 7.7% and 0.5% compared with CA125 and HE4,the same as that of ROMA.Model_Complex had a sensitivity of88.6%,which was increased 5.7%,34.3% and 8.6% compared with CA125,HE4 and ROMA,respectively.The specificity of Model_Complex(94.6%)was increased 25.2%and 10.8% compared with CA125 and ROMA.The accuracy of Model_Complex(92.3%)was increased 17.7%,10.5% and 10.0% compared with CA125,HE4 and ROMA,respectively.Peak1 was increased in ovarian cancer in postmenopausal group.Thus,further corresponding to Peak1,Peak9 and Peak12,we collected five kinds of lectin blots assays containing Aleuria aurantia lectin(AAL),Lens culinaris agglutinin(LCA),Lens culinaris agglutinin(LTL),Erythrina cristagalli lectin(ECL),Phytohemagglutinin-lymphocyte type(PHA-L).The expression of AAL,ECL and PHA-L in ovarian cancer group was higher in ovarian cancer than that in ovarian benign tumor and healthy controls,and there was no significant difference between benign tumor and healthy group.In addition,we tested 9 paired tumor and adjacent tissue samples and 11 benign tissue samples by lectin blots.The results showed that PHA-L and ECL was higher in ovarian cancer.At the same time,we also tested paired tumor and adjacent tissue as well as 11 benign tissue proteins by RT-PCR,which showed the levels of ?1,6-GlcNAc residues and GNT-V expression increased in tumors than that in adjacent tissues and benign tumor tissues.There was no significant difference between adjacent tissues and benign tumor tissues.There was no significant difference between benign tumor tissues,paired tumor and adjacent tissues of B4GALT1 mRNA.In conclusion,the diagnostic models based on N-glycan markers are valuable,which maybe an useful method to improve the recognition rate of ovarian cancer,especially associated with the noninvasive method.We can conclude that the higher of the content of glycoprotein in abnormal glycosylation in ovarian cancer may be increased from its related transferase expression.The N-glycan combined CA125 and HE4 can significantly improve the effectiveness of the diagnosis of ovarian cancer.