PU.1 Is a Critical Downstream Target of AML1.

PU.1 is expressed in hematopoietic stem cells (HSC), progenitors and differentiating blood cells except terminally differentiated T cells, erythrocytes and megakaryocytes. PU.1 is required for commitment of HSC to multiple lineages. PU.1 −/− embryos die perinatally and fail to generate myeloid and B cells. We previously reported that a DNase I hypersensitive site located 14 kb upstream of the PU.1 transcription start site (−14 DHS) confers myelomonocytic specific gene expression. Targeted deletion this DHS fragment in mice results in a decrease in PU.1 expression in bone marrow to 20% of wild type levels, subsequently leading to a profound decrease in macrophages and B cells. Within the DHS fragment is a “core” consisting of a distal (296bp) and a proximal (253bp) region, which are highly conserved among different species. The PU.1 promoter by itself cannot direct gene expression in vivo . However, −14 DHS confers to the promoter the ability to direct expression of a reporter gene in granulocytes, monocytes, and B-cells of transgenic mice. The proximal region can itself direct high-level gene expression. The proximal region contains 3 AML1 sites. These results, along with data indicating that PU.1 expression is selectively absent from Aml1 −/− embryos (Okada, et al, Oncogene. 1998), suggested that AML1 is likely to be upstream of PU.1. Electro-mobility gel shift assays and chromatin immunoprecipitation assays confirmed that AML1 binds to all 3 AML1 sites both in vitro and in vivo. Mutation of the 3 AML1 sites dramatically reduced the DHS activity of conferring gene expression. We used real time PCR to quantitatively measure PU.1 expression in both embryonic and adult hematopoiesis. We found that PU.1 expression was completely lost in the 9.5 dpc yolk sac, 10.5 dpc AGM and fetal liver of Aml1 −/− embryos, suggesting that AML1 is required for PU.1 expression during embryonic hematopoiesis. To evaluate the effects of AML1 loss in the adult hematopoiesis, we employed a conditional Aml1 knockout allele in which LoxP flanked Aml1 (Aml1 F/F ) was excised by Mx1 promoter driven Cre expression following injection of pIpC. These mice show that Aml1 is not required for maturation of myeloid lineages in adult mice. However, these mice develop a mild myeloproliferative phenotype characterized by increasing in bone marrow and peripheral blood (PB) neutrophils, a 5 fold increasing in HSC, and 2–3 fold increasing myeloid progenitors. Spleen and liver contain infiltration by myeloid cells. These mice also display a dramatic decrease (~80%) in PB platelets and bone marrow megakaryocytes. Furthermore, there are significant blocks in lymphoid development, including reduced numbers of pre-B, pro-B and mature B cells, as well a block in T cell maturation at the DN2 (CD4 − ;CD8 − ;CD44 + ;CD25 + ) stage. We observed a 70% reduction of PU.1 expression in sorted HSC, progenitors, Gr1 + /Mac1 + and B-cells from these mice relative to control mice. In contrast, upregulation of 3–5 fold expression in Ter119 + , CD41 + , and T cells in these mice compared to controls. Our data shows that PU.1 is a critical target gene of AML1, and AML1 regulates PU.1 in both positive and negative way. We are currently testing the ability of restoration of PU.1 expression to rescue specific defects in Aml1 F/F ; Tg (Mx1-cre) mice, as well as investigating the role of decreased PU.1 expression in human AML in which the function of AML1 is disrupted.