Spontaneously arising human tumours are heterogeneous and often show genomic instability, which makes it difficult to distinguish cancer-causing from bystander mutations. Scott Lowe and colleagues used an integrative oncogenomic approach, comparing gene amplifications in mouse and human hepatocellular carcinomas (HCC), and identified two oncogenes, cIAP1 and YAP , that are co-amplified and cooperate to promote tumorigenesis.

Inactivation of p53 and activation of MYC are two important events in HCC. The authors developed a new model of HCC tumorigenesis by isolating liver progenitor cells from Trp53 −/− mice, expressing Myc by retroviral transduction and then transplanting the Trp53−/−/Myc cells into recipient mice, which led to liver tumour formation. They then conducted genome-wide scanning of the resultant tumours and human HCC samples using an approach called representational oligonucleotide microarray analysis (ROMA), which identifies gene amplifications at high resolution. ROMA identified an amplification in human HCCs that was syntenic to a region amplified in the murine Trp53−/−/Myc HCCs. Both encoded many overlapping genes that might have a role in HCC development.

So, which gene on this amplicon is driving tumorigenesis? The authors narrowed down the list of candidates by looking at the expression of each gene that was present on both amplicons, and found that cIAP1 and YAP mRNA and protein were consistently increased in mouse and human HCCs. They then used their mouse model to evaluate the roles of these two genes in liver tumorigenesis. The expression of cIAP1 or YAP in the Trp53−/−/Myc progenitor cells accelerated tumour onset and progression and increased the tumour burden, whereas reducing the expression of cIAP1 or YAP using short hairpin RNAs partially inhibited tumorigenesis. Furthermore, expression of cIAP1 and YAP together synergistically increased tumour growth.

cIAP1 is an inhibitor of apoptosis (IAP) protein, but the role of IAP proteins in tumorigenesis is controversial. YAP is a transcriptional co-activator that has been shown to promote apoptosis, which is inconsistent with an oncogenic role. However, its Drosophila homologue, Yorkie, functions in the Hippo–Salvador–Warts pathway to control tissue expansion through the transcription of cyclin E and the Drosophila Iap2 gene, which has functional similarities to cIAP1. In addition to identifying the importance of the Hippo pathway in HCC and perhaps other cancers, the work by Lowe and colleagues has provided a new mouse model of HCC and highlighted the potential power of integrating data from appropriate mouse models and oncogenomics.