Relevance network between chemosensitivity and transcriptome in human hepatoma cells.

Generally, hepatoma is not a chemosensitive tumor, and the mechanism of resistance to anticancer drugs is not fully elucidated. We aimed to comprehensively evaluate the relationship between chemosensitivity and gene expression profile in human hepatoma cells, by using microarray analysis, and analyze the data by constructing relevance networks. In eight hepatoma cell lines (HLE, HLF, Huh7, Hep3B, PLC/PRF/5, SK-Hep1, Huh6, and HepG2), the baseline expression levels of 2300 genes were measured by cDNA microarray. The concentrations of eight anticancer drugs (nimustine, mitomycin C, cisplatin, carboplatin, doxorubicin, epirubicin, mitoxantrone, and 5-fluorouracil) needed for 50% growth inhibition were examined and used as a measure of chemosensitivity. These data were combined and comprehensive pair-wise correlations between gene expression levels and the 50% growth inhibition values were calculated. Significant correlations with significance were used to construct networks of similarity. Fifty-two relations, including 42 genes, were selected. Among them, nearly 20% were various types of transporters, and most of them negatively correlated with chemosensitivity. Transporter associated with antigen processing 1 was associated with resistance to mitoxantrone, consistent with previous reports. Other transporters were not reported previously to associate with chemosensitivity. Resistance to doxorubicin and its analogue, epirubicin, were positively correlated with topoisomerase II beta expression, whereas it negatively correlated with expression of carboxypeptidases A3 and Z. Response to nimustine was associated with expression of superoxide dismutase 2. Relevance networks identified several negative correlations between gene expression and resistance, which were missed by hierarchical clustering. Our results suggested the necessity of systematically evaluating the transporting systems that may play a major role in resistance in hepatoma. This may provide useful information to modify anticancer drug action in hepatoma.