Abstract Aneuploidy has long been associated with cancer but whether it is a cause or consequence of oncogenesis remains unknown. Hepatocellular Carcinomas have frequent aberrations in chromosomes 1 and 8, and oncogene addiction has yet to be clearly demonstrated in this tumor type. Therefore, we developed a mouse model to test the hypothesis that chromosome instability was a driver of liver cancer. To induce chromosomal instability in mouse hepatocytes, we generated a floxed allele of the spindle accessory checkpoint protein Mad2 and induced Cre-mediated deletion in hepatocytes by intercrossing with an Albumin-Cre transgenic. Loss of Mad2 in hepatocytes resulted in the development of both benign Hepatocellular Adenomas (HCA) and malignant Hepatocellular Carcinomas (HCC) by 12 mo. Backcrossing onto a p53-deficient background increased the penetrance and speed of disease with nearly all double deficient mice having HCC by 4 mo. Analysis of the tumor genome by array CGH showed recurrent gain and loss of whole chromosomes with a lack of small focal alterations. We showed whole chromosomes were under selective pressure by comparing HCCs with HCAs as well as T cell lymphomas that arose in the Mad2 floxed line crossed to an Lck-Cre transgenic. Comparing HCCs to HCAs showed that the magnitude but not the pattern of chromosomal aberrations is what distinguishes these tumor types. Further genetic analysis of HCCs and HCAs from the same mouse showed half of the tumor pairs shared a common ancestor, suggesting that in this model HCAs can progress to HCCs through continuing chromosomal instability. Comparing the genomes of HCCs and T cell lymphomas that resulted from Mad2-deficiency in these different tissues showed tissue-specific gain and loss patterns. Chromosomes 4 and 12 were specifically lost in HCCs but gained in lymphomas while chromosome 18 was gained in HCCs but lost in lymphomas. In addition, chromosome 15 was frequently gained in both tumors. Combined these findings show that gain and loss of whole chromosomes can be evolutionarily selected, resulting in tumorigenesis. Next, we tested how aneuploidy alters cell phenotype to induce cancer by performing matched array CGH and gene expression arrays on 29 HCCs. Gene dosage significantly altered expression since chromosome ploidy was significantly correlated with mRNA expression from that chromosome. We also identified a set of twelve genes that was recurrently over-expressed in over 80% of HCCs; most of this set was also over-expressed in human HCCs. Notably Igf2 in murine HCC was over-expressed by 50 fold. IGF2 over-expression in human HCCs has been linked to aneuploidy through loss of loss of Chr8p which causes global demethylation and loss of imprinting at the IGF2 locus. Thus, aneuploidy events are also able to affect the expression of specific genes. Since aneuploidy is frequently observed in human HCC, we performed a synteny analysis of human and mouse chromosomes to find regions that are gained or lost in both species. Five regions were found to be significantly gained by aneuploidy in both species. These regions were located on human chromosomes 1q and 8q, and mouse chromosomes 1 and 15, with the most frequently amplified region containing MYC (hChr8q/mChr15). In total, we show that chromosomal instability is sufficient to drive liver cancer through evolutionary selection acting on whole chromosomes, that aneuploidy alters cell phenotype through gene dosage and other mechanisms, and that the genetic and phenotypic similarities between this mouse model and human liver cancer suggest that aneuploidy is also a driver of human cancer. Citation Format: Lee A. Albacker, Floris Foijer, Ying Yue, Darin Takemoto, Stehpanie Davis, Rodney T. Bronson, Peter K. Sorger. Aneuploidy as a driver of liver cancer. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr PR02.
Read full abstract