Abstract
The Runx1 transcription factor is among the most frequently mutated genes in acute myeloid leukemia (AML). In addition to the generation of the RUNX1-RUNX1T1 fusion protein in patients with the translocation (8;21), circa 15% of patients with a normal karyotype harbor missense mutations or frame-shift mutations in the RUNX1 gene, in most cases resulting in either a highly compromised protein lacking DNA-binding activity or a null allele. Recent studies have demonstrated that AML transformation normally occurs at the level of the granulocyte-monocyte-progenitor (GMP), thus we sought to dissect the role of Runx1 at this critical stage of myeloid differentiation by examining the hematopoietic progenitor compartment in conditional Runx1 knockout (KO) mice. Earlier studies have demonstrated increased cell numbers in the stem cell compartment in these mice, but a rigorous assessment of myeloid progenitors has not been performed. Our analysis showed that loss of Runx1 resulted in increased levels of all myeloid progenitors, with a 2.2-fold increase in the absolute GMP numbers. Furthermore, Runx1-deficient GMPs gave rise to 40% more colonies than controls, demonstrating increased self-renewal activity within this population. Notably, in addition to being larger, Runx1-deficient colonies did not exhibit the typical structure of GM colonies, which reflects the differentiation status of the cells composing the colonies. Indeed, retarded or impaired differentiation of Runx1-defcient cells was demonstrated by analysis of cell morphology and cell surface markers, with little or no mature forms observed in colonies cultured for 7 to 10 days. Notably, no significant shift to either the G or M lineages was observed in vitro or in vivo.
To identify Runx1 target genes that impact at the level of the GMP, expression analysis of Runx1 wild-type and KO GMPs was performed. In addition, the transcriptome of primary KO GMPs genetically engineered to conditionally re-express RUNX1 in vivo was determined before or after Runx1 induction. A total of 36 reciprocally regulated genes were identified using stringent criteria to assess expression levels and fold-induction. Notably, many of the Runx1 target genes encode adhesion factors important for retaining HSPC in the stem cell niche, and which are normally down-regulated during differentiation. These genes were up-regulated in Runx1-deficient GMPs and down-regulated after Runx1 induction. Genome wide DNA-binding analysis established Runx1 binding to regulatory regions of target genes and furthermore revealed cooperative binding of Runx1 with several other TFs implicated in complex networks regulating self-renewal and differentiation of early hematopoietic progenitors. In vitro adhesion assays confirmed that loss of Runx1 resulted in increased adhesion to hematopoietic stromal cells.
In summary, we conclude that Runx1 has at least two critical functions in normal myeloid development. In the early stem cell and progenitor compartment, Runx1 represses genes that mediate adhesion to the stem cell niche, thereby facilitating the differentiation program at the expense of self-renewal. On the other hand, during myeloid maturation, Runx1 augments both G and M developmental programs, presumably by facilitating up-regulation of critical granulocytic and monocytic genes. We predict that this former function is of critical importance during leukemogenesis, in which reduced levels of functional Runx1 increases the number of myeloid progenitors with self-renewal potential, which are critical targets of secondary genetic mutations during the clonal evolution of the leukemic clone.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.