Abstract 2371

Missense mutations in the gene encoding hematopoietic transcription factor GATA1 cause congenital anemias and/or thrombocytopenias. Seven such mutations are reported. All of these give rise to amino acid substitutions within the amino terminal zinc finger (NF) of GATA1, producing a range of clinical phenotypes. Thus, V205M, G208R, and D218Y cause severe anemia and thrombocytopenia; G208S, R216Q, and D218G cause thrombocytopenia with minimal anemia; R216W gives rise to thrombocytopenia and congenital erythropoietic porphyria. One of these mutations, R216Q, occurs at the DNA binding interface and alters the ability of GATA1 to recognize a subset of cis motifs in vitro. Other mutations, including V205M, G208S, D218G, and D218Y, occur outside the DNA binding domain of the NF and inhibit interactions with the GATA1 cofactor FOG1 as determined by in vitro binding assays. However, these two mechanisms do not easily explain the broad spectrum of phenotypes associated with the mutations. For example, how do two substitutions of the same residue bring about disparate phenotypes?

We examined the effects of each mutation on erythroid maturation, lineage-specific gene expression, in vivo target gene occupancy, and cofactor recruitment by introducing altered forms of GATA1 into murine GATA1-null proerythroblasts. The V205M, G208R, and D218Y mutations severely impaired erythroid maturation, recapitulating patient phenotypes. The G208S mutation also severely impaired erythroid maturation, causing a more pronounced defect than that expected from the clinical presentation. In contrast, R216Q and D218G produced mild effects in erythroid cells consistent with patient phenotypes. The porphyria-associated mutation R216W also produced relatively subtle effects in erythroid cells. We note that among the mutants, failure to activate gene expression strongly correlated with failure to repress gene expression. ChIP assays revealed that the V205M, G208R, and D218Y mutations impaired GATA1 target site occupancy. This indicates that despite normal DNA binding in vitro, the association with cofactor complexes is required for stable binding to chromatinized target sites in vivo. In contrast, the G208S mutant exhibited relatively normal chromatin occupancy, but reduced recruitment of FOG1 and SCL/Tal1 to GATA1-bound sites at erythroid genes. D218G also perturbed cofactor recruitment without greatly affecting GATA1 binding to its target genes. Notably, this mutation diminished SCL/Tal1 recruitment without significantly altering FOG1 occupancy. This implicates the SCL/Tal1 transcription complex in the pathogenesis of disorders caused by certain GATA1 mutations. Moreover, by uncoupling GATA1 chromatin occupancy and cofactor recruitment, G208S and D218G offer potentially useful tools for unraveling site-specific mechanisms of GATA1-regulated gene expression. Finally, both the R216Q and R216W mutants displayed relatively normal GATA1 chromatin occupancy and FOG1 and SCL/Tal1 recruitment at most sites. R216W presents as porphyria, and selective defects in the regulation of heme biosynthetic genes have yet be uncovered. Given that R216Q presents as thrombocytopenia, defects caused by this mutation may be revealed only in the context of megakaryocytes. Studies using similar rescue assays of a GATA1-null megakaryocyte-erythroid progenitor line are underway and will be discussed.

In concert, our results reveal that in vivo analysis of GATA1 in its native environment provides mechanistic insights not obtainable from in vitro studies. Moreover, they demonstrate the usefulness of gene complementation assays for the dissection of transcription pathways surrounding normal and altered GATA1 to improve our understanding of disease.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution