Font Size: a A A

A Role Of Collagen In The Pathogenesis Of Gastric Cancer

Posted on:2011-09-16Degree:DoctorType:Dissertation
Country:ChinaCandidate:A Q LiFull Text:PDF
GTID:1114360305458012Subject:Internal Medicine
Abstract/Summary:PDF Full Text Request
Gastric cancer is a common malignant tumor worldwide. It remains the second most common cause of cancer death. Early detection and resection improve the long-term survival, as gastric cancer could be more curable if diagnosed at earlier stage. The prognosis of advanced cancer remains poor due to high recurrence and metastasis. Most patients are asymptomatic in the early stages of GC. With current histopathological diagnostic approaches, only about 15% to 20% of GC patients are eligibly detected in early stages. With no reliable marker or diagnostic model available for the early diagnosis of GC, it is imperative to identify and develop new biomarkers and/or diagnostic model to improve the early detection of GC.The malignancy is evidenced by the presence of the morphological atypia and biological invasive behavior of the cell. Cell atypia is also observed in the inflammation, premalignant lesions, such as metaplasia or dysplasia gastric epithelium, which is one of the contributions to low diagnostic rate of early gastric cancer (EGC). It is widely accepted that atrophy-metaplasia-dysplasia-carcinoma is sequential development of GC carcinogenesis. Atrophic gastritis, intestinal metaplasia and/or dysplasia are believed to be precancerous or predisposing conditions for GC and the majority of GC results from precancerous lesions. The traditional histopathological classification systems of GC, on the basis of cell morphological atypia, are difficult to efficiently predict the risk of sequential development from precancerous lesions. From molecular mechanism of cancer cell invasion, we can search for more molecular information so as to earlier discriminate EGC from premalignant lesions than traditional histopathological level.Molecular mechanisms of neoplastic transformation have been under the thorough investigation. Novel techniques and new genes involved into the gastric cancer development have been indentified and showed promising prospect. For example, Cui and his colleagues developed a microarray-based prewarning system for early detection of gastric cancer and precancerous lesions. Yu found several novel genes associated with early gastric carcinogenesis and development by microarray analysis. Our previous microarray analysis has shown the aberrant expression profile of collagens in GC, suggesting that collagens have a great potential to be used as a diagnostic and prognostic marker in GC and maybe become biomarkers of atypical cell invasion. To identify potential biomarkers or diagnostic model for the early detection of GC, we compared the expression of collagen genes from different group patients'gastric tissues including normal group (chronic superficial gastritis), premalignant lesions, EGC and advanced gastric cancer (AGC). In addition, we investigated the effects of collagen type I that upregulated in gastric cancer tissues on gastric carcinoma cell lines as tumor microenvironment and by siRNA technology to silence COL1A1 gene.Material and Methods1 To identify potential biomarkers or diagnostic model for the early detection of GC, we compared the expression levels of collagen genes among different group patients' gastric tissues including normal group (chronic superficial gastritis), premalignant lesions, EGC and advanced gastric cancer (AGC) by three sequential phase. Gastric tissues samples were obtained from 106 patients who underwent gastrectomy or gastroscopy. Eight collagen genes were further determined by our microarray analysis and related literature. The expression levels of eight genes in clinical tissue samples of stomach were detected by real-time quantitative reverse transcription-PCR (qRT-PCR). The significance of eight collagens mRNA levels among different groups (phaseⅠ:normal, premalignant and GC; phaseⅡ:Premalignant, EGC and AGC; phaseⅢ:normal, premalignant and GC) was determined by ANOVA or Kruskal-Wallis Test where appropriate. Binary logistic regression analysis was performed to predict the probability that a sample was from a GC patient (disease) or not (control), and to propose the possible cancerous predictive formula for early diagnosis of gastric cancer.2 To investigate the role of collagen typeⅠon invasion and metastasis of gastric cancer, we emplyed three gastric cancer cell lines originated from various differentiation stages to explore the effects of collagen typeⅠon gastric cancer cell behavior including cell morphology, cytoskeleton organization, migration and proliferation as tumor microenvironment.In addition, the possible molecular mechanism was explored by detecting the expression levels of related signal protein and gene.3 COL1A1 in BGC-823 cell lines was knocked down by small interference RNA. MTT assay and Transwell migration assay were performed to investigate the effects of COL1A1 silencing on cell proliferation and migration.Results1①PhaseⅠ:66 matched tissue samples from different regions of resected stomach of 22 surgery patients (22 malignant lesions,22 premalignant lesions and 22 normal tissues) were chosen for qRT-PCR analysis by 8 collagen genes. Seven genes were statistically significantly different between pair-matched malignant and premalignant groups (P<0.05). Six of these genes (COL1A1, COL1A2, COL3A1, COL6A3, COL8A1 and COL11A1) were significantly overexpressed and only one (COL4A6) were under-expressed in GC patients as compared with premalignant controls.②PhaseⅡ:33 samples from different patients (11 premalignant lesions,11 EGC and 11 AGC) were chosen for further qRT-PCR analysis by 7 collagen genes. COL11A1 was up-regulated in GC patients than those premalignant group control, especially up-regulated in EGC patients(P< 0.05); COL4A6 were significantly down-regulated in GC patients than those premalignant group control (P< 0.05). Binary logistic regression analysis resulted in a predictive formula with COL11A1, COL8A1 and COL4A6 to be the best grouping as variables that predicted GC with an overall correct classification of 100%.③PhaseⅢ:106 from different patients (20 normal tissues,28 premalignant lesions, 58 GC) were chosen for further qRT-PCR analysis by 7 collagen genes. Four genes were statistically significantly different between premalignant group and GC groups (P<0.05). Three of these genes (COL1A1, COL1A2 and COL11A1) were overexpressed and COL4A6 was under-expressed in GC patients. On the basis of the relative expression values of collagen genes in the tissues of GC patients and premalignant lesions, the model from phase II was validated and Binary logistic regression analysis resulted in a new predictive formula with COL1A1, COL1A2, COL3A1, COL4A6, COL6A3 and COL11A1 to be the best grouping as variables that predicted GC with an overall correct classification of 91.5%.2①Gastric carcinoma cell lines BGC-823 and SGC-7901 showed apparent changes in cell morphology after 24 h collagen type I stimulation. Cells cultured on collagen type I exhibited elongated or enlarged shape with a higher cytoplasm/nucleus ratio as compared with the cells directly cultured on dishes. They groveled on the collagen type I substrate with spreading of pseudopodia and a loss of cell-cell contacts typically observed in mesenchymal cells. Immunofluorescence microscopy demonstrated that cells grown on collagen type I were scattered with a loss of cell-cell contacts but more lamellipodia and filopodia, characteristic of a great motility.②Transwell migration assays showed that BGC-823 and SGC-7901 cells showed eightfold increase in cell migration towards the collagen type I-coated outer-side porous membranes compared with uncoated ones, which indicated collagen type I induced cell migration in the similar way as chemotactic factors, corresponding to cell morphology change and cytoskeleton reorganization.③Related signal pathway study showed that the E-cadherin/catenin adhesion complex associated with the actin cytoskeleton were reduced after collagen type I stimulation, Whereas, the amount of E-cadherin remained constant in NCI-N87 and BGC-823 cells after collagen type I stimulation; the amount ofβ-catenin was slightly reduced in BGC-823 cells or remained constant in NCI-N87 cells after 2 days of collagen type I treatment. These results suggested that collagen type I treatment causes disruption of the E-cadherin/catenin adhesion complex and dissociation from the actin cytoskeleton in gastric carcinoma cells, subsequently resulted in a decrease in cell-cell adhesion. These results maybe associated with the phosphorylation of FAK, tyrosine phosphorylation ofβ-catenin and the dissociation of endogenous PTEN andβ-catenin.④MTT assays showed the proliferation rates of BGC-823 cells and SGC-7901 cells grown on collagen type I were significantly increased compared with the controls on dishes, which was related withβ-catenin nuclear translocation and the activation ofβ-catenin-LEF/TCF pathway.3①Three recombinant plasmids targeting COL1A1 using pSilencerTM 4.1-CMV neo siRNA expression vector were constructed successfully and were transfected into gastric cancer BGC-823 cells.②The expression of COL1A1 mRNA and protein in BGC-823 cells was significantly inhibited by small interfering RNAs (siRNA) transfectants, especially by COL1A1-shRNA-1 transfectant.③MTT assays showed that the proliferation rate of BGC-823 cells with COL1A1-shRNA-1 trasfected was significantly decreased compared with the parental cells and the negative control cells. ④Transwell migration assays showed that the migration capacity of BGC-823 cells with COL1A1-shRNA-1 trasfected was decreased compared with the parental cells and the negative control cells.Conclusions1 COL4A6 and COL11A1 can be potential mRNA markers for early gastric cancer, and COL11A1 was more sensitive than COL4A6.2 The possible cancerous predictive formula was developed and showed remarkerble discriminatory power for the diagnosis of early gastric cancer. This provided new biological diagnose model for clinical diagnosis of early gastric cancer.3 Collagen type I not only induced gastric carcinoma cells scattering and cytoskeleton remodeling, but also prompted cell migration and proliferation,which was related to tyrosine phosphorylation and nuclear translocation ofβ-catenin.4 Gastric cancer BGC-823 cell transfected with COL1A1-shRNA was successfully established, whose COL1A1 expression was clear decreased and resulted in a clear reduction of cell proliferation and migration. This maybe provided potent therapy target for gastric cancer.
Keywords/Search Tags:gastric cancer, precancerous lesion, collagen, gene, biomarker, diagnosis, model, molecular mechanism
PDF Full Text Request
Related items