Phosphorylation of AML1/Runx1 Is Essential for T Cell Differentiation and Early Hematopoietic Development.

AML1/Runx1 is one of the most frequent targets of gene aberrations associated with human acute myelogenous leukemia and myelodysplastic syndrome. Furthermore, gene-targeting studies in mice demonstrated that AML1 is a critical regulator for T-cell differentiation and early development of definitive hematopoiesis. It has been shown that phosphorylation of AML1 at specific serine/threonine residues (276, 293, 300, 303) controls both transcriptional activity and rate of degradation. However, the biological consequences of phosphorylation with respect to AML1 function are unclear. In this study, we evaluated hematopoietic activities of AML1 mutants which harbor serine/threonine-to-alanine (A) or serine/threonine-to-asparatic acid (D) mutations at these phosphorylation sites using primary culture systems. We first evaluated the biologic effects of AML1 phosphorylation on T-cell differentiation using the fetal liver (FL)/OP9-DL1 coculture system. In this system, AML1-excised FL cells failed to undergo normal T-cell differentiation, which is successfully restored by the reintroduction of wild-type AML1. AML1-4A, which cannot be phosphorylated at any of the four serine/threonine residues, showed a severely impaired capacity to rescue the defective T-cell differentiation of AML1-deficient cells. We also demonstrate that blocking the ERK-mediated phosphorylation of AML1 by the MEK inhibitor significantly suppressed the AML1-induced T-cell differentiation. Next, we assessed the effect of AML1 phosphorylation on early hematopoietic development using para-aortic splanchnopleural (P-Sp) region/OP9 coculture system. In this system, AML1-deficient P-Sp cells failed to produce any hematopoietic cells and this hematopoietic defect can be rescued by retrovirally transferred AML1. AML1–4A could rescue the hematopoietic defect of AML1-deficient P-Sp cells, but the emergence of hematopoietic cells in cultures infected with AML1–4A was reproducibly delayed two or more days compared with those infected with wild-type AML1. Because it was shown that serine 462 is phosphorylated upon phorbol ester treatment together with the four phosphorylation sites, we then generated AML1 mutants in which serine 462 is substituted with alanine (A) or aspartic acid (D) in AML1–4A or AML1–4D respectively (AML1–5A or AML1-5D). Interestingly, introduction of AML1–5A into AML1-deficient P-Sp cells did not generate any hematopoietic cells. Moreover, AML1–5A completely lost its T-cell differentiation activity. In contrast, the phospho-mimic protein, AML1–5D rescued the hematopoietic defects of AML1-deficient cells as efficiently as wild-type AML1. Taken together, these results suggest that phosphorylation of AML1 at the five serine/threonine residues (276, 293, 300, 303 and 462) is essential for T cell differentiation and early hematopoietic development.