Genome-Wide Analysis of Pancreatic Cancer Using Microarray-Based Techniques

Background/Aims: Microarray-based comparative genomic hybridisation (CGH) has allowed high-resolution analysis of DNA copy number alterations across the entire cancer genome. Recent advances in bioinformatics tools enable us to perform a robust and highly sensitive analysis of array CGH data and facilitate the discovery of novel cancer-related genes. Methods: We analysed a total of 29 pancreatic ductal adenocarcinoma (PDAC) samples (6 cell lines and 23 microdissected tissue specimens) using 1-Mb-spaced CGH arrays. The transcript levels of all genes within the identified regions of genetic alterations were then screened using our Pancreatic Expression Database. Results: In addition to 238 high-level amplifications and 35 homozygous deletions, we identified 315 minimal common regions of 'non-random' genetic alterations (115 gains and 200 losses) which were consistently observed across our tumour samples. The small size of these aberrations (median size of 880 kb) contributed to the reduced number of candidate genes included (on average 12 Ensembl-annotated genes). The database has further specified the genes whose expression levels are consistent with their copy number status. Such genes were UQCRB, SQLE, DDEF1, SLA, ERICH1 and DLC1, indicating thatthese may be potential target candidates within regions of aberrations. Conclusion: This study has revealed multiple novel regions that may indicate the locations of oncogenes or tumour suppressor genes in PDAC. Using the database, we provide a list of novel target genes whose altered DNA copy numbers could lead to significant changes in transcript levels in PDAC.