Genomic amplification of oncogenes and inactivation of tumor suppressor genes are recognized as critical in the pathogenesis of human cancer. To identify the chromosomal changes associated with hepatocarcinogenesis, we applied genome imbalance map (GIM) analysis about 35 hepatocellular carcinomas. GIM analysis, which is one of genotyping with single nucleotide polymorphism quantitation and detects DNA copy number alterations and loss of heterozygosity (LOH) events simultaneously using high-density single nucleotide polymorphism arrays, showed that amplification of 1q21.3-42.2 (57.1%), 5q23.3-35.2 (25.7%), 6p25.3-12.3 (22.8%), and 8q21.11-24.3 (51.4%) and LOH of 1p36.31-36.21 (20.0%), 4q32.3-35.2 (25.7%), 6q14.1-26 (20.0%), 8p23.2-12 (48.5%), 13q12.3-22.1 (22.8%), 13q21.32-22.1 (22.8%), 16p12.1-24.1 (28.5%), and 17p13.2-12 (54.2%) were significantly associated with hepatocellular carcinoma. Most regions with altered chromosomal regions identified by GIM were also detected in the previous reports using comparative genomic hybridization and comprehensive allelotyping about liver cancer. In order to confirm the accuracy of this method, we demonstrated chromosomal copy number gain by fluorescence in situ hybridization for amplified chromosomal regions detected by GIM, and LOH using allelotyping by microsatellite markers. Furthermore, we performed hierarchical clustering analysis using altered chromosomal regions, suggesting that cyto-band subclusters with higher concordance score has potential interactions of cyto-bands in hepatocarcinogenesis. For the purpose of elucidating how aneuploidies affect transcriptome in cancer cells, we compared to the expression data obtained by oligonucleotide array using the corresponding samples which were investigated by GIM, and found that genome imbalance resulted in the altered expression of numerous genes. Data of ours showed that GIM analysis can detect allelic amplification of cancer specimens with high-resolution compared with comparative genomic hybridization and LOH regions with ease than comprehensive allelotyping analysis.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]