BACKGROUND: Nearly 15% of acute myeloid leukemia (AML) cases are caused by aberrant expression of AML1-ETO, a fusion protein generated by the t(8;21) chromosomal translocation. Since its discovery, AML1-ETO has served as a prototype to understand how leukemia fusion proteins deregulate transcription to promote leukemogenesis. Another leukemia fusion protein, E2A-Pbx1, generated by the t(1;19) translocation, is involved in acute lymphoblastic leukemias (ALLs). While AML1-ETO and E2A-Pbx1 are structurally unrelated fusion proteins, we have recently shown that a common axis, the ETO/E-protein interaction, is involved in the regulation of both fusion proteins, underscoring the importance of studying protein–protein interactions in elucidating the mechanisms of leukemia fusion proteins.

OBJECTIVE: In this review, we aim to summarize these new developments while also providing a historic overview of the related early studies.

METHODS: A total of 218 publications were reviewed in this article, a majority of which were published after 2004. We also downloaded 3D structures of AML1-ETO domains from Protein Data Bank and provided a systematic summary of their structures.

RESULTS: By reviewing the literature, we summarized early and recent findings on AML1-ETO, including its protein–protein interactions, transcriptional and leukemogenic mechanisms, as well as the recently reported involvement of ETO family corepressors in regulating the function of E2A-Pbx1.

CONCLUSION: While the recent development in genomic and structural studies has clearly demonstrated that the fusion proteins function by directly regulating transcription, a further understanding of the underlying mechanisms, including crosstalk with other transcription factors and cofactors, and the protein–protein interactions in the context of native proteins, may be necessary for the development of highly targeted drugs for leukemia therapy.