Construction And Preliminary Application Of Oligonucleotide Microarray Specialized For Pancreatic Adenocarcinoma-Associated Genes

Objective: To investigate the feasibility, reliability and accuracy of construction oligonucleotide microarray specialized for pancreatic adenocarcinoma-associated genes and its preliminary application in screening differential gene expression profiles between normal pancreas and pancreatic carcinoma. Additionally, the differential expression genes obtained from the results of microarray hybridization were furtherly analysised biologically with aiming to provide theoretical basis for the study of pathogenisis and earlier diagnosis of pancreatic cancer.Methods: Pancreatic cancer-associated genes were purposively selected by retrieving authoritive databases such as MEDLINE, Chinese periodical full-text data base (CNKI) and so on, and all the genes suspected to be involved in pancreatic carcinogenesis according to literatures were further verified with Genbank database. Oligonucleotide probes of the selected genes were designed and synthesised. Those target oligonucleotide probes dissolved in spoting liquid were spotted on substrates coated with homobifunctional coupling agent (APS-PDC) by OmniGrid100 spotting robotic hand, and sixteen spots of positive, negative and blank control were arranged to moniter the reliability of the hybridization. Samples of six pancreatic cancer and three normal pancreatic tissues were collected with the approval of the patients. A piece of each tissue sample was immediately frozen upon resection in liquid nitrogen and stored at -80℃until use. Total RNA were extracted from frozen tissues with TRIzol methed according to the manufacturer's protocol, then treated with DnaseⅠ(Rnase free), purified with QIAGEN Rneasy Kit. To assure the quality of obtained total RNA, three tests were performed which included spectrophotometer, agarose gel electrophoresis and Lab on chip. Three total RNA samples obtained from normal pancreatic tissues were mixed and used as a standard normal sample throughout the experiment. Labeled cDNA targets for hybridizations were synthesized by reverse transcription from standard- andcancer-total RNA samples in the presence of Cy5-dCTP and Cy3-dCTP (Amersham Phamacia Biotech), respectively. And then, the labeled samples were purified with QIAquick PCR purification Kit according to the company's protocol. After quantitating, the separately synthesized Cy3- and Cy5-labeled targets were combined, mixed with hybridization solution and denatured at 95℃for 2 minutes, prehybridized at 70℃for 20 minutes, and then added on the chip and covered with glass. After incubating at 42℃for 16-18h in the hybridization chamber, the slides were washed with solutions of 2×SSC/0.1%SDS, 1×SSC/0.1%SDS and 0.5×SSC for 10 minutes respectively, and centrifugated for 5 minutes. Hybridized microarray was scanned by Agilent laser scanner, and then calculated the coefficient of correlation of inter-slides and inner-slides, detection rate and coefficient of variation to assess the quality of the hybridization. The acquired image was analyzed by Imagene3.0 software (BioDiscovery, Inc.). The intensity of each spot at the 2 wavelengths represented the quality of Cy3 and Cy5 respectively. The signal intensity of Cy3 and Cy5 were normalized by Genespring software with Normalize to positive control genes method. Each ratio of Cy3 to Cy5 was computed, and the Ratio and Cluster analysis were carried out. In order to verify the reliability and accuracy of the result of microarray hybridization, three genes, CDC25B,BRAF and TUSC3, were employed to undertake quantitative real-time PCR (Sybrgreen method), and beta-actin gene was taken as internal control. The QRT-PCR reaction was runned with Invitrogen RT-PCR kit according to the protocol, and the product of PCR was calculated by comparative Ct method.Results: The results of microarray hybridization: Ninety-four pancreatic cancer related genes were selected to construct the oligonucleotide microarray after retrieving literature and verifying in Genbank database. After hybridization and scanning, the acquired image has a lower background and a higher signal-noise ratio which has an uniform background, well-distributed space between spots, the center of unit located on the pinel and less noise signal originated from dirt and other pollution. The signal intensity of the positive control spots were well-distributed and conformity, and the signal of the negative control spots were fairly low. The change of the coefficient of variation were very small, and there were higher coefficient of correlation in inter-slides and inner-slides. According to the screening standard, twenty-four genes differentially expressed in pancreatic cancer against normal pancreas were identified, which included seventeen up-regulated and seven down-regulated genes.The results of QRT-PCR: To further verify the reliability of microarray results, three genes, CDC25B, BRAF and TUSC3, were randomly selected to perform quantitative real-time PCR. The concentration of initial templete of the three genes calculated by comparative Ct method were obtained, which represcent the gene expression level in different samples. The average templete amount of CDC25B, BRAF and TUSC3 in normal pancreatic tissues were 0.0006350±0.00021, 0.0010167±0.00025 and 0.0387493±0.00965, whereas in pancreatic cancer they were 0.001988±0.00065, 0.0026062±0.00109 and 0.0206028±0.00401 respectively. Compared with normal pancreatic tissues, the expression of CDC25B and BRAF were increased for 3.13 and 2.56 times respectively(P<0.05), and TUSC3 decreased for 1.88 times in pancreatic cancer (P<0.05). The results of QRT-PCR confirmed the results of microarray hybridization.Conclusion: The construction of oligonucleotide microarray specialized for pancreatic cancer by selecting genes that are nonredundant and preferentially expressed in pancreatic cancer are desirable for its speciaity, flexibility and sensitivity, which can simultaneously detect multiple pancreatic cancer-associated genes. In comparison with normal pancreatic tissues, the gene expressing profile in pancreatic cancer are very different, which can provide valuable information for further studying of the pathgenesis and earlier diagnosis of pancreatic cancer. Additionally, the gene expression of CDC25B,BRAF and TUSC3 are significantly different between pancreatic cancer and normal pancreas, which may play an important role in the pathgenesis of pancreatic cancer, and suggest that they might be potent candidates for diagnostic marker or therapeutic target for further studies.