Hepatocellular carcinoma (HCC) is a highly malignant disease that is associated with a high cancer morbidity and mortality. Although genome-wide characterizations have uncovered numerous non-random chromosomal alterations in HCC tumors, the affected genes in many aberrant genomic loci remained largely undefined. Since cumulative evidences have shown that copy number imbalances serve as underlying causes in driving aberrant expressions of tumor suppressors and proto-oncogenes, in this study we have undertaken an integrated analysis of high-resolution genomic mapping and expression profiling with an aim to elucidate for the underlying candidate genes in the common aberrant loci of HCC. Using a cDNA array platform, correlative analyses in 20 HCC cell lines highlighted 18 candidates that manifested both copy number gains and gene expression up-regulations (R2>0.8). These included glutamate receptor interacting protein 2 (GRIP2), insulin-like growth factor binding protein 2 (IGFBP2), calcium/calmodulin-dependent protein kinase ID (CAMK1D), RNA binding motif protein, X-linked (RBMX), paternally expressed 10 (PEG10) and WW domain binding protein 4 (WBP4). Of interest was paternally expressed 10 (PEG10) resided within the HCC progression-related chr 7q21 region. Further quantitative PCR and quantitative RT-PCR analyses verified ∼75% of cell lines with concurrent copy number gains and over-expression of PEG10. Moreover, a higher relative copy number of PEG10 was common in HCC cell lines (1.38, 0.02-5.76; median, range) compared to normal livers (0.93, 0.83-1.28) (p=0.036). A similar but more profound increase in PEG10 mRNA expression level in cell lines (302.3, 0.80-5687.0) compared to normal liver controls (0.54, 0.06-30.7) was also observed (p<0.0001). To ascertain the significance of PEG10 over-expression in HCC, we further examined its expression levels in a cohort of primary HCC (n=48) and the corresponding pre-malignant liver tissues adjacent to the tumor (n=40). Results showed an overall significantly higher expression of PEG10 in primary HCC (13.84, 0.01-744.4) compared to adjacent non-tumorous liver tissues (1.60, 0.12-73.86) (p=0.007). Interestingly, a progressive trend of increased PEG10 expression levels was observed from normal livers to the putative pre-malignant adjacent cirrhotic/fibrotic livers, and to HCC tumors, with estimated median values of 0.54, 1.6 and 13.84, respectively (p<0.0001). Taken together, our current study demonstrated the usefulness of integrated genomic mapping and expression profiling analyses in identifying candidate genes in genomic regions. Our results also highlighted PEG10 as a potential candidate that demonstrated progressive up-regulations during HCC development, and that genomic gain may represent one of the mechanisms in the induction of PEG10 over-expression.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]