Characterization of Function Network of Novel Genes Over-Expressed in Human Hepatocellular Carcinoma

Examining alteration of gene expression pattern during carcinogenesis of HCC by genomics and proteomics approach has opened a new avenue to understand biology of HCC. During the past few years, I have developed bioinformatics tools to examined cDNA microarray databases of human HCC [1]. Several novel genes (HURP, FLJ10468, FLJ10540, RHAMM and FLJ11252) overexpressed in human HCC have been identified and characterized ( [2], [3]-[4]-[5] ). HURP has been shown to play important role in cell cycle G2/M control ([4]). FLJ10468 (Borealin/Dasra B) was identified to form complex with Survivin and Aurora B Kinase as a chromosome passenger protein ([3]). RHAMM has been shown as mammalian specific G2/M regulatory ([4]). FLJ10540 was shown transforming activity in variety of cell types ([5]). FLJ11252 has profound effect on cell migration (unpublished results). Genome wide gene knockdown by RNAi in principle could identify all essential genes for growth control of human hepatoma cells. However, unique growth behavior of human hepatoma cells may depend on interaction of multiple genes. How to select multiple targets of human HCC to attack becomes a challenge. Using system biology analysis such as robustness of network to identify a set of genes essential for integrity of the network becomes a novel approach in cancer research. In this three years project, I will combine experimental and bioinformatic approaches to construct a special protein interaction network for cell cycle G2/M control. Systemic deletion of single or multiple nodes of network will be performed by computer simulation and lentiviral based RNAi knockdown. The integrity of network, cell viability and drug resistance will be examined as the readout of the responses of the system to gene knockdown. Two different experimental approaches will be taken. One is using five genes (HURP, FLJ10468, FLJ11252, RHAMM and FLJ10540) I identified in human HCC as a core member to identify all physical interaction proteins through protein complex isolation, proteomics analysis. The other is to identify minimal cell cycle regulated promoter of 5 core member genes. Then using DNA pull-down and proteomic analysis may identify common transcriptional factors or coactivators associated with these promoters. Experimental identified proteins will then become core member for next stage of analysis. Additional interaction partners of previously identified proteins will be identified by data-mining of protein interaction database of yeast, worm and fly. To construct such special protein interaction network from both experimental and bioinformatic approaches will provide insight about G2/M control of human hepatoma cells. Multiple genes knockdown by computer simulation and RNAi will help us to identify multiple targets for growth control and drug resistance of human HCC cells. The information generated from this proposal will be immediately applied to develop effective therapeutic approach for human HCC in the future.

Project ID:PA9706-0689
External Project ID:NSC96-2311-B182-005-MY3