Blood and Runx1: got to have it, really need it

Comment on Goyama et al, page 3558, and Sun and Downing, page 3565

Altered levels of AML1 (Runx1) differentially affect myeloid and lymphoid sub-populations.

Two articles in this issue of Blood address the question: How much functional acute my-elogenous leukemia 1 (AML1)/Runx1 (Runt-related) protein is required for normal hemato-poietic cell development, differentiation, and proliferation? The answer seems to be that greater than 50% of a functional AML1 transactivator is required. Defining the role of AML1 (officially known as Runx1, a term widely used by the scientific community but less so by the “leukemia” community) in different hematopoietic precursor compartments contributes greatly to our understanding of how AML1-eight-twenty-one (ETO; or other leukemia-associated mutant forms of AML1/Runx1) participates in human leukemogenesis.

The one paper, by Goyama and colleagues, examines the role that AML1 plays in the development of hematopoietic cells contained in the para-aortic splanchnopleural/aorta-go-nad-mesonephros (P-Sp/AGM) region of the embryo, cells that behave like definitive hematopoietic precursor cells in stromal-based cultures. They reintroduce wild-type (and mutant forms of) AML1 into AML1 null P-Sp/AGM–derived cells and show rescue of these precursor cells; this contrasts with earlier reports that suggest such cells are not formed in the absence of AML1. The trans-activating domains in AML1, and not its repression domains, are required to rescue the in vitro phenotype of cells at this stage of development.

In the other Blood paper, Sun and Downing report that AML1 gene dosage plays a substantive role in regulating the numbers of hematopoietic stem cells (ie, long-term repopulating hematopoietic stem cells [LTR-HSCs]) and progenitor cells (evaluating short-term repopulating hematopoietic stem cells [STR-HSCs], cobblestone area–forming cells [CAFCs], and colony-forming units [CFUs], as well as a modest role in regulating platelet numbers and CD4⁺ T-cell percentages) by carefully analyzing AML1 heterozygous adult mice. Loss of a single AML1 allele decreased the number of LTR-HSCs (by 50%) but increased the number of early myeloid progenitor cell populations. No change in cell cycle parameters or self-renewal was seen, leaving the biologic basis for the competitive repopulating advantage and the increased number of progenitor cells not yet explained.

The translocations that generate leukemia-associated fusion proteins result in haploinsufficiency for both rearranged genes. Therefore, it is critically important to understand the consequences of having less than the full function of AML1 in different cellular subsets that have different intrinsic properties. The concept of a cancer stem cell was first substantiated by the identification of the nonobese diabetic–severe combined immunodeficiency (NOD-SCID) leukemia-initiating cell in John Dick's lab a decade ago.¹  This concept is now being extended to “solid” tumors, although it has been made more complex by the recent demonstration that progenitor cells can acquire self-renewing mechanisms and act “stem cell”–like.²  Different cellular compartments may also have different requirements during embryonic and fetal development than during adulthood. For instance, Ichikawa et al³  found that although AML1 function is essential for the development of definitive hematopoiesis, it is not required for the maintenance of hematopoietic stem cells in the adult. This could relate to expression of other Runx family members, as the Goyama et al paper shows that Runx2 and Runx3, which are expressed in certain hematopoietic cells, can substitute for Runx1 function in P-Sp cells. Rosenbauer et al⁴  have recently shown that having 20% (but not 50%) of PU.1 function can lead to a leukemic phenotype, and further studies will no doubt similarly clarify the hematopoietic consequences of having lower than normal levels of AML1. By defining the relative contributions of loss of gene function and the gains of function seen with the leukemia-associated fusion proteins to leukemogenesis, it will become clearer how to attack malignant hematopoietic cells that may contain less than or more than their normal complement of transcription factor function.