AML1-ETO (AE) is the most frequent fusion gene in human AML. Previously, we and others have demonstrated that the fusion is not able to cause leukemia on its own in experimental murine models, but that it needs collaborative partners. Following the ‘two-hit model’ hypothesis that class I and class II mutations have to collaborate to induce leukemia, we could previously demonstrate that AE (class II) collaborates with FLT3-length mutation (class I) to generate AML in the murine bone marrow transplantation model (BMT model). We now demonstrate that AE can collaborate with mutations of its own class in inducing AML. We focussed our analyses on the TALE homeobox gene family, namely Meis1 and Meis2, as already known in the case of Meis1 to be a potent co-factor for Hox gene associated leukemias and being classified as other homeobox genes as a class II mutation.
First, real-time RT - PCR confirmed that MEIS1 is expressed at high levels in a subgroup of AE positive AML patients. Furthermore, MEIS2 was highly and aberrantly expressed virtually in all AE patients (n=70) compared to normal bone marrow. Expression patterns of both MEIS genes correlated with their promoter methylation in the t(8;21) patients and cell lines. To test the functional relevance of MEIS expression in human AE patients, we analysed collaboration of AE with Meis genes in the BMT model: Meis1 and Meis2 were retrovirally expressed alone or in combination with AE in 5-FU treated primary murine bone marrow and injected into lethally irradiated recipients (AML1-ETO alone, n=10; EGFP control, n=7; AE+Meis1, n=14; AE+Meis2, n=4). None of the mice in the AE as well as in the control group developed disease. In contrast, mice transplanted with BM co-expressing AE+Meis1 and AE+Meis2 developed lethal disease after a median latency of 102 and 255 days respectively. AE+Meis1 induced MPS in three, AML in seven and ALL in three cases, whereas AE+Meis2 induced AML in all cases. Functional relevance of MEIS expression was further confirmed in human AML cell lines as shRNA mediated depletion of MEIS1 as well as MEIS2 impaired cell growth in AE positive AML cell lines and so far one primary AE AML sample. To understand the mechanisms of AE/MEIS collaboration co-immunoprecipitation assays were performed documenting weak binding of Meis1, but strong interaction of Meis2 with AE. ChIP-Seq is currently being performed to test whether DNA binding of MEIS genes alter in collaboration with AE. Taken together, we could demonstrate that AE collaborates with Meis genes in AML. It furthermore shows that a class II mutation can be converted into an overt oncogene by a mutation of its own class.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.