Defining Candidate Oncogenes Leading to Increased Susceptibility to Leukemia in eed Mutant Mice.

Abstract We previously showed that loss of function of the chromatin regulator polycomb group gene (PcG) eed increases susceptibility to lymphoid tumour development (Sauvageau, M. et al. Blood 104 (11); 213). The human eed locus is in a region of chromosome 11 (11q14.2–22.3) which is associated with recurrent deletions in several reported cases of lymphoproliferative diseases. The persistant expression of the wild-type allele of eed is maintained in lymphomas developing in eed+/− animals. This suggests a haploinsufficient function for this gene in preventing tumorigenesis. We recently exploited an insertional mutagenesis screen to identify new cooperative leukemia-inducing genes in eed mutant mice. 325 retroviral insertion sites (RIS) representing 189 different gene loci have been identified by inverse-PCR from a total of 51 lymphomas (25 eed−/− and 26 wild-type). This screen revealed 18 gene loci known to be cancer related and 21 new common insertion sites (CIS) such as sh2d3c (n=4 tumours), fgfr3 (n=15 tumours) and myo5b (n=7 tumours). More than 97% of these CIS were common to both eed mutant and wild-type animals. We next assessed the expression levels of all the genes found in a 100 Kb region of genomic DNA surrounding the CIS identified in our screen in eed mutant and wild-type tumours as well as in the corresponding non-leukemic tissues. Almost 17% of the 48 genes analyzed, including fgfr2, fgfr3, ttc16 and adam15, are aberrantly expressed in non-transformed eed mutant cells when compared to wild-type controls (levels 5–100+ fold). Importantly, several of these genes are either aberrantly expressed in wild-type and eed mutant tumours or represent true CIS thus indicating their possible contribution to early tumour development in eed mutant mice. Moreover, our results do not incriminate eed as a modifier of gene expression once cells are fully transformed, since the oncogenic complement is similar in eed mutant and wild-type leukemias. Collectively, our results link the haploinsufficient role of eed to the appropriate regulation of genes involved in leukemia development.