Abstract
The AML1/ETO fusion gene is caused by chromosomal translocation t(8;21), which is among the most frequent chromosomal rearrangement seen in acute myeloid leukemia (AML). AML1/ETO functions in a dominant negative fashion to directly repress AML1-mediated transcriptional activation, and block the differentiation of myeloid progenitors. Notably, although it is oncogenic, AML1/ETO is not by itself sufficient to cause fully malignant AML, but can do so when there are concurrent oncogenic mutations (Higuchi 2002). This suggests that AML1/ETO requires cooperation with additional genetic alterations that are as yet largely unknown. The BXH-2 mouse strain is characterized by high frequency of spontaneous leukemia caused via insertional mutagenesis by a B-ecotropic murine leukemia virus. The proviral insertion site can then be used as a molecular tag, proviral insertion tag (PIT), to identify mutations/dysregulations of genes involved in leukemogenesis. Indeed, the BXH-2 strain has been utilized as a powerful forward genetic system to identify a number of leukemia-associated genes by our lab and collaborators. In this study, we have attempted to identify AML-related genes and pathways that could cooperate with AML1/ETO expression in development of AML. We have introduced the conditional AML1/ETO allele and Mx1-Cre, whose expression is induced by pIpC, into BXH-2 strain background to generate AEB mice after 4~6 consecutive backcrossing, and monitored daily these mice for sign of disease. Notably, the latency of leukemia in pIpC-injected AEB mice (mostly 5~8 months) is significantly shorter than that in saline-injected AEB control mice (mostly over one year). Immunophenotyping the tumor cells showed positivity for myeloid cell surface markers. Furthermore, leukemogenicity was tested and confirmed by transplanting the tumor cells generated from enlarged lymph nodes into B6C3F1 recipients. These data suggest that these mice developed a fatal, aggressive myeloid leukemia. Southern blot showed, that in these leukemic cells, the recombined allele of AML1/ETO was formed by Cre induction. Induced expression of AML1/ETO was not only detected in peripheral white blood cells but also in tumors. Importantly, these AML1/ETO AML also harbor clonal proviral insertions, indicating that AML1/ETO-initiated leukemia is dramatically accelerated by chronic retroviral infection in our model. To identify AML1/ETO-cooperative leukemia genes, we have isolated several PITs from AML1/ETO-accelerated leukemias. These PITs have also been mapped onto mouse genome using Celera and Ensembl mouse genome databases, and been found to target new candidate genes. Interestingly, we have discovered a new common PIT, which have not been reported in previous work which has identified over 2000 PITs. This common PIT targets a G-protein coupled receptor-like gene, a member of the GPCR family. We believe that these studies will provide important information for better understanding the mechanisms of development of leukemia with t(8;21), and reveal promising candidate genes or pathways serving as the potential therapeutic targets for improving AML treatment.
Author notes
Corresponding author
