1. ¹Department of Genetics, Harvard Medical School, Boston, MA, USA
2. ²Genetics Division, Brigham and Women's Hospital, Boston, MA, USA

Correspondence: Professor P Leder, Department of Genetics, Harvard Medical School, NRB 356B, 77 Avenue Louis Pasteur, Boston, MA 02115, USA. E-mail: leder@genetics.med.harvard.edu

Received 3 July 2007; Revised 30 August 2007; Accepted 6 September 2007; Published online 22 October 2007.

Abstract

Oncogenesis is a progressive process often involving collaboration between various oncogenes and tumor suppressors. To identify those genes that collaborate with oncogenic ras, we took advantage of the Tg.AC transgenic mouse, a line that harbors the v-Ha-ras transgene and spontaneously develops an array of malignant tumors. By crossing Tg.AC mice on an inbred FVB background to other inbred strains, F1 mice were created that could be analysed using genome wide, single nucleotide polymorphism (SNP) screens. Loss of heterozygosity (LOH) in tumors and tumor cell lines marked a somatic event, possibly the inactivation of tumor suppressor gene(s). LOH could also represent DNA damage, a sign of genomic instability in the pretransformed cell. Nonetheless, the screens showed no evidence of such generalized genomic instability. Instead, they revealed a single region of LOH on chromosome 4 that occurred via somatic recombination/gene conversion, generating a region of isoparental disomy. This LOH provided a clue that linked v-Ha-ras to the inactivation of the Ink4a locus in 25 of 32 tumor cell lines. This collaboration is seen regardless of tumor type or genetic background. In contrast, tumors that develop in bitransgenic mice bearing both the v-Ha-ras gene and a heterozygous mutant p53 allele tend to retain the Ink4a locus and instead lose the p53 wild-type allele. This suggests that different strategies can be selected to collaborate with v-Ha-ras in tumorigenesis.