The paradigm of transcriptional regulation of differentiation has been demonstrated with considerable veracity in the hematopoietic system, and Osawa and colleagues (page 2769) add a novel player to the growing list of significant transcription factors.

We first learned about the Gfi family of transcription factors through the identification of Gfi-1 as an integration site of the Moloney murine leukemia virus. Gfi-1 was also shown to participate with c-myc and pim-1 in lymphoma formation in Eμ/myc and Eμ-pim1 transgenic mice, respectively. Gfi-1B was identified by its sequence similarity to Gfi-1, and both proteins display identical domain architecture with 1 SNAG and 6 zinc finger domains.

Osawa and colleagues became interested in Gfi-1B due to its high expression in hematopoietic stem cells. Overexpression ofGfi-1B in CD34+ cells resulted in the formation of EPO-independent erythroblast formation in clonogenic assays, as well as strong induction of Glycophorin A–positive cells in liquid culture. Gfi-1B did not affect mitogenesis in proliferation assays. Structure-function experiments revealed that the zinc finger domains of Gfi-1B are required for formation of Glycophorin A–positive cells. Enforced expression of Gfi-1B resulted in the expression of GATA-2, with no effect on SCL andGATA-1.

In parallel studies, the Orkin laboratory has shown that targeting Gfi-1B results in embryonic lethality due to a profound defect in formation of enucleated erythrocytes (Saleque et al, Genes Dev. 2002;16:301-306). Both erythroid and megakaryocyte development is arrested in Gfi-1B null animals.

Several important issues regarding Gfi-1B remain to be resolved. Evidence is presented within the Osawa study that Gfi-1B can serve as both an activator or a repressor of transcription, depending on the promoter and cell type under study. This report also suggests that Gfi-1B may regulate GATA-2 expression, but this could be due to an accumulation of erythroid cells in their assay. Gfi-1 binds to the SUMO ligase PIAS3, suggesting a role in negative regulation. Clearly, identification of additional Gfi-1/1B targets, identification of potential synergy between Gfi-1/1B and the panoply of hematopoietic transcription factors, and characterization of significant Gfi-1/1B binding interactions will tell us more about the interplay of this interesting family of transcription factors.