Developmental Stage-Selective Effect of Somatically Mutated GATA-1 in Down Syndrome AML M7--a Potential Basis for Transient Myeloproliferative Disorder.

Transient myeloproliferative disorder (TMD) is a self-limited disorder of newborns with Down Syndrome (DS). Recently, acquired mutations in GATA-1 have been identified in megakaryoblasts from nearly all DS neonates with TMD, and in virtually all DS patients with the related acute megakaryoblastic leukemia (DS-AMKL, or DS-AML M7). These mutations lead to production of an N-terminus truncated GATA-1 protein (GATA-1s). Here we address the biological properties of GATA-1s. Two alternate hypotheses may account for the temporal nature of TMD and restriction of DS-AMKL to infants. First, GATA-1 somatic mutation under the DS background might occur only during a narrow developmental window and additional mutations are needed to sustain a premalignant clone; alternatively, GATA-1 mutation might occur at any time, but only fetal progenitors are sensitive to effects of GATA-1s. Our findings provide support for the latter possibility. In our study, we have generated ES cells and mice harboring a GATA-1 knockin allele with a truncation at this domain (GATA-1^ΔN). ES cells bearing GATA-1^ΔN exhibit slightly perturbed erythropoiesis, but generate a unique type of large thrombopoietin-dependent megakaryocytes-containing colonies, when differentiated in vitro. Remarkably, adult mutant mice have normal platelet and red blood cell counts, and their corresponding progenitors. During embryonic development, however, mutant embryos are anemic in E12.5 fetal livers, but return to normal shortly thereafter. Mutant fetal livers contain elevated numbers of CD41+ cells throughout embryonic development, especially during mid-gestation. Despite these abnormalities in proliferation, differentiation/maturation of both erythrocytes and megakaryocytes are relatively normal. Strikingly, in in vitro hematopoietic colony assays, cells from yolk sacs and early-mid stage fetal livers give rise to a large number of macroscopic hyperproliferative megakaryocytic colonies (CFU-MKs). This unique type of macroscopic CFU-MK is greatly reduced in number after mid-fetal liver stage and not detectable after birth. Interestingly, wild type mid-gestation fetal liver cells ectopically expressing GATA-1s through retroviral transduction also give rise to similar macroscopic CFU-MK colonies, as do wild type ES cells expressing GATA-1s from a GATA-1 regulatory element, when differentiated in vitro. Overall, our data suggest that the GATA-1 N-terminal domain controls proliferation of megakaryocytes, but has a more limited role in their differentiation. The N-terminal truncation of GATA-1, therefore, uncouples proliferation from differentiation, and does so dominantly. The biological effects of GATA-1s appear restricted to a unique, previously unrecognized yolk sac/fetal liver progenitor, a finding that we propose accounts for the transient nature of TMD and the infantile restriction of DS-AMKL. This observation may also pertain to other infant leukemia syndromes where the target cell may prove to be an embryonic/fetal progenitor rather than an adult myeloid progenitor. Our results thus support a model for the pathogenesis of TMD/DS-AMKL in which trisomy 21 in DS patients provides a “sensitized” background that favors exaggerated expansion of this fetal-type progenitor harboring GATA-1s. This is predicted to lead to TMD, which then evolves into full-blown DS-AMKL following additional, as yet unknown, somatic mutation(s).