Font Size: a A A

Verification And Identification Of Valuable Differentially Expressed Genes From Our Gastric Cancer CDNA Microarry Data

Posted on:2006-07-01Degree:DoctorType:Dissertation
Country:ChinaCandidate:H L HuangFull Text:PDF
GTID:1104360152494767Subject:Elderly digestion
Abstract/Summary:PDF Full Text Request
Background and Aims:Gastric cancer is one of the leading causes of cancer deaths worldwide. Unfortuanately, little is known about the molecular genetics underling the pathoetiology and progression of gastric cancer, so far ,there lacks effective methods to early diagnosis and treatment of gastric cancer.xDNA microarray techonology has provided researchers the unique opportunity to investigate the mechanism during gastric cancer development and progression, this approach has also been an improtant way to identify signifcant biomarkers, but many work has to been done to validate the array-based observations before identification of a useful biomarkers. We previously reported on the different gene expression patterns of gastric cancer vs matching nonneoplastic gastric mucosa by using complementary DNA(cDNA) microarry technology., here we futher investigate whether some of the differentially expressed genes could be of value for diagnosis and prognostic predication for gastric cancer.Methods:1: 8 genes seletcted from our differentialy expressed pattern data plus 3 genes totally 11 genes were analysed by using RT-PCR on 22 fresh surgical samples of gastric tumour tissue and matching noncancerous mucosa. The genes analyzed are S100A2, S100A4, S100A6, AKR1C1, AKR1C2, AKR1C3, AKR1C4, FAM3B, IBP4, CKBB, and SAFB1. Real-time RT-PCR were used to futher verify the different expression levels of three selected S100A2, S100A4, S100A6 genes.2: A gastric cancer TMA(tissue microarry) containg 1020 tissue dots which represented 208 cases was constructed. It was characterized by it's matching design and survival data supported.3: Four genes of interested, namly S100A2, S100A4, S100A6, FAM3B were further validated either by IHC (immunohistochemistry) or in situ hybridization. Detailed analysis of S100A2, S100A4, S100A6 and FAM3B expression with clinicopathologic parameters and survival was performed.4:In order to explore the function role of S100A2 gene in gastric cancer , RNAi silencing of S100A2 was performed and the effective silengcing sequence was determined.Results:1: Of 8 genes seletcted from our differentialy expressed pattern data, 5 genes showed the same differenail expression pattern as shown in cDNA microarray.data, they are S100A6, IBP4, AKR1C2, FAM3B, SAFB1 and CKBB. The RT-PCR's confirmation rate of microarray data is 62.5%. As quantitated by Real-Time RT-PCR, the transcript level of S100A2 in primary cancer lesion was elevated in 80% samples when compared with matching nonneoplastic mucosa and the average up-regulation level was 10.78 fold. 55% cancer lesion showed higher transcript level of S100A4 than their counterpart nonneoplastic mucosa, the average up-regulation level was 2.31 fold. S100A6 transcript level was higher in 74% primary cancer leision with an 2.25 fold up-regulation than the matching nonneoplastic mucosa. After rectified by P 2-microglobulin, the relative expression level of S100A2, S100A4 and S100A6 were 2.83E-04, 6.44E-02, 0.41 respectivly.2: A gastric cancer TMA(tissue microarry) containg 1020 tissue dots which represented 208 cases was successfuly constructed. It is characterized by prognosis information support and matching design.3: As displayed by immunohistochemistry, the positive rate of S100A2,S100A4 and S100A6 each in turn of nonneoplastic mucosa, tumor lesion and metastatic lymph node were 9.6%/32.2%/43.5%, 9.4%/28.1%/32.2%, 34.3%/84.1%/90.9% respectively. Each showed a higher expression level in cancer and metastasis lymph node than their matching nonneoplastic mucosa (P^0.05). In nonneoplastic mucosa, S100A2 overexpression was only related with lymphatic and vascular tumor emboli(P=0.0017).. In primary tumor, S100A2 expression in 0,1, III stage was higher than that of II stage cancer(P=0.05). In metastasis lymph node, higher expression of S100A2 was significantly-associated with younger age (p=0.004). Patients with high S100A2 expression in lymph node showed a poorer prognosis tendency as compared to those has a low expression(P=0.07). S100A4 expression in nonneoplastic mucosa has no relation with any cliniclpathologic patamters. Both in primary cancer and metastasis lymphnode S100A4 expression incresed with deeper invasion (P=0.014, P=0.013.respectively) . An increased S100A4 expression level can also be seen in metastasis lymph node with TNM stage progression.. The S100A4 expression level either in cancer or metastasis lymph node was associated with prognosis, patients with higher expression had a poorer prognosis(P=0.034, P=0.002 respectively). 65.5% of patients showed an increased S100A6 expression in cancer lesion when compared to nonneoplastic counterpart. S100A6 expression in nonneoplastic mucosa was associated with age,patients in older group had an elevated SI006 expression when compared with younger group(P=0.022). In primary cancer, S100A6 overexpression corelated with large tumor size and deeper invasion (P=0.022 and 0.013 respectively). In metastasis lymph node , the S100A6 expression level was difference according to primary cancer location .Lymph node metastasis which originated from distal stomach tumors showed the higest S100A6 positive rate than those from proximal tumor and corpus tumor(P=0.024).There was no association between S100A6 expression level and prognosis. But compared with nonneoplastic , an increased S100A6 expression in tumor leison predicated a poorer prognosis if compared with a decreased S100A6 expression, although the difference was not .significant. According to the Spearman corrlation coefficient analysis, sigificant positive correlations between S100A2 and S100A6, S100A2 and S1004 was found, even the changes of expression between tumor and nontumor mucosa was significantly positively correlated. But there was no correlation between S100A4 expression and S100A6 expression.4: 81.5% patients showed decreased FAM3B expression in tumor when compared with their nonneoplastic counterparts. The mean total gray 0.8 was used ascutoff value to devieded the patient as positive or negative. The positive rate in nonneoplastic mucosa, tumor and metastasis lymph node was 34.3%, 84.1%, 90.9% respectively, the difference of the positive rate was of significance between them(P=0.00). Both in nonneoplastic mucosa and metastasis lymph node the FAM3B expression level had no relation with any clinicopathologic parameters while in tumor, higher FAM3B was related to deeper invasion and TNM stage.progression.(P=0.008 and P=0.049 respectively). The expression level in each of the three component or the expression change status were of no statistical significance to prognosis, but a trend toward a better prognosis were found if there was a higher expression level..5: In Kaplan-Meier analysis, the tumor size, depth of invasion, pathologic grade, lymph node metastasis, TNM stage, and the expression level of SI00A4 in tumor or metastasis lymph node had influence on survival. In multivariate regression analysis, FAM3B expression in nonneoplastic mucosa emerged as a highly significant independent parameter (P=0.011). If analyze a subgroup of the patients with lymph node metastasis, the FAM3B expression in nonneoplastic mucosa still kept in model while TNM stage lost it predicator value ,but tumor size ,S100A4 expression level in lymphnode were also significant independent prognostic variables.6: Through transient transfection of RNAi vector into SGC-7901 cell and analysis of S100A2 expression both at mRNA level and protein level, An effecient S100A2 siRNA sequence was selected.Conclusions:l:We validated our previous cDNA microarry resluts throug RT-PCR and real-time RT-PCR. We think real-time RT-PCR which could quantitate gene copies is a highly sensitive and reproducible technique.2:The gastric cancer TMA we constructed is of great value to analyze the gene expression pattern in situ and it's relation with clinicobiological parameters both from mRNA level and protein level. It proviede a rapid way to search for valuable biomarkers from microarray result.
Keywords/Search Tags:gastric cancer, tissue microarray, immunohistochemistry, in situ hybridization, RT-PCR, real-time RT-PCR, RNA interference, prognosis
PDF Full Text Request
Related items