AML1/ETO

Since 1993, we have been studying the AML1-ETO fusion protein, which is found predominantly in the FAB-M2 subtype of acute myelogenous leukemia (AML) and is generated by the t(8;21). AML1 activity is critical for the development of definitive hematopoiesis; and haploinsufficiency of AML1 has been linked with a propensity to develop AML.

The AML1 gene is one of the most frequently targeted genes in acute leukemia and is altered by translocations, deletions, and mutations. The AML1-ETO fusion protein contains the first 177 amino acids of AML1 (including the DNA-binding domain but not the transcriptional activation domain of AML1) and almost the full length of the ETO protein.

Our group first demonstrated that AML1-ETO functions as a transcriptional repressor in 1995, and showed that it dominantly inhibits AML1 function. The role of the ETO domain in causing this repression has recently been clarified; ETO binds the co-repressors N-CoR and mSin3, and recruits histone deacetylase (HDAC) activity, allowing AML1-ETO to repress transcription. AML1-ETO interferes with the function of AML1 (RUNX1, CBFA2), but we have shown that it also represses transcription via protein:protein interactions with other transcription factors, including MEF and Smads. Thus, AML1-ETO can inhibit some TGFb-generated signals, and we are identifying which ones are or are not inhibited.

Murine experiments suggest that AML1-ETO expression may not be sufficient for leukemogenesis. However, the cell background in which these fusion proteins are expressed may be critical to the phenotype observed. We have used retroviral gene transfer to examine the biological effects of AML1-ETO on the behavior of human hematopoietic stem and progenitor cells.

Following transduction of human CD34+ cells, we showed that AML1-ETO expression inhibited committed progenitor (CFU) colony formation but enhanced the growth of stem cells (cobblestone area-forming cells), resulting in a profound survival advantage of transduced over nontransduced cells. AML1-ETO-expressing cells retained progenitor activity and continued to express the CD34 surface antigen throughout the 5-week-period of long-term culture. Thus, AML1-ETO enhances the self-renewal of pluripotent stem cells, the physiological target of many acute myeloid leukemias.

We are now defining which oncogenic signals are needed to cooperate with AML1-ETO and cause human leukemia. We are also using gene chip microarrays to define AML1-ETO target genes in the physiologically relevant human CD34+ cell. Biochemical studies are also underway to define what proteins AML1-ETO interacts with and how these protein complexes regulate the activity of AML1 and AML1-ETO.