Multifaceted Actions of GFI1 and GFI1B in Hematopoietic Stem Cell Self-Renewal and Lineage Commitment

Abstract

Growth factor independence 1 (GFI1) and the closely related protein GFI1B are small nuclear proteins that act as DNA binding transcriptional repressors. Both recognize the same consensus DNA binding motif via their C-terminal zinc finger domains and regulate the expression of their target genes by recruiting chromatin modifiers such as histone deacetylases (HDACs) and demethylases (LSD1) by using an N-terminal SNAG domain that comprises only 20 amino acids. The only region that is different between both proteins is the region that separates the zinc finger domains and the SNAG domain. Both proteins are co-expressed in hematopoietic stem cells (HSCs) and, to some extent, in multipotent progenitors (MPPs), but expression is specified as soon as early progenitors and show signs of lineage bias. While expression of GFI1 is maintained in lymphoid primed multipotent progenitors (LMPPs) that have the potential to differentiate into both myeloid and lymphoid cells, GFI1B expression is no longer detectable in these cells. By contrast, GFI1 expression is lost in megakaryocyte precursors (MKPs) and in megakaryocyte-erythrocyte progenitors (MEPs), which maintain a high level of GFI1B expression. Consequently, GFI1 drives myeloid and lymphoid differentiation and GFI1B drives the development of megakaryocytes, platelets, and erythrocytes. How such complementary cell type- and lineage-specific functions of GFI1 and GFI1B are maintained is still an unresolved question in particular since they share an almost identical structure and very similar biochemical modes of actions. The cell type-specific accessibility of GFI1/1B binding sites may explain the fact that very similar transcription factors can be responsible for very different transcriptional programming. An additional explanation comes from recent data showing that both proteins may have additional non-transcriptional functions. GFI1 interacts with a number of proteins involved in DNA repair and lack of GFI1 renders HSCs highly susceptible to DNA damage-induced death and restricts their proliferation. In contrast, GFI1B binds to proteins of the beta-catenin/Wnt signaling pathway and lack of GFI1B leads to an expansion of HSCs and MKPs, illustrating the different impact that GFI1 or GFI1B has on HSCs. In addition, GFI1 and GFI1B are required for endothelial cells to become the first blood cells during early murine development and are among those transcription factors needed to convert adult endothelial cells or fibroblasts into HSCs. This role of GFI1 and GFI1B bears high significance for the ongoing effort to generate hematopoietic stem and progenitor cells de novo for the autologous treatment of blood disorders such as leukemia and lymphoma.

Keywords: GFI1; GFI1B; hair cells; hematopoietic stem cells; hemogenic epithelium; transdifferentiation.

Figure 1

Structure and function of human GFI1 and GFI1B. (A) Schematic depiction of the structure of the proteins, showing the SNAG suppressor domain, the less characterized intermediate domain involved in protein/protein interactions, and the six zinc finger domains (ZF) localized at the C-terminal end with those three involved in DNA binding shown in green and the three other domains that play a role in interaction with other proteins shown in silver. The two isoforms of Gfi1b are shown with the longer megakaryocyte-specific isoform 1 that has the six zinc fingers and the short erythroid-specific isoform 2 that lacks two zinc fingers due to the fusion of ZF1 and ZF3. (B) Schematic representation of the GFI1 (top) and GFI1B (bottom) complexes with different partners that promote gene silencing by removal of open chromatin signatures and induction of marks that correlate with closed chromatin. WRD, Wnt regulatory domain.

Figure 2

Separate roles of GFI1 and GFI1B during definitive hematopoiesis. Schematic representation of hematopoiesis from the initial production of primordial definitive hematopoietic stem cells (HSCs) from the hemogenic epithelial cells (HEs) found in the aorta-gonad-mesonephros (AGM) through endothelial-to-hematopoietic transition (EHT) during embryonic development until the production of all blood cells in the definitive hematopoietic sites such as fetal liver and adult bone marrow. Cellular compartments in which GFI1 plays a role are indicated as a blue area, whereas the compartment where GFI1B is important are indicated as the yellow areas, highlighting both overlapping and exclusive functions of GFI1 and GFI1B at all stages of hematopoiesis. Factors that contribute to cell fate decisions along with GFI1 or GFI1B are indicated. LMPP, lymphoid-primed multipotent progenitors; CMP, common myeloid progenitors; CLP, common lymphoid progenitors; GMP, granulocyte-monocyte progenitors; MEP, megakaryocyte-erythrocyte progenitors; MkP, megakaryocyte progenitors; Eosino, eosinophils; Neutro, neutrophils; Baso, basophils; DCs, dendritic cells; NK, natural killer cells.

Figure 3

Summary of the strategies using overexpression of GFI1 and/or GFI1B. Shown are the combination of GFI1 and GFI1B with other factors to generate transplantable hematopoietic stem cells (HSCs) through trans-differentiation of non-hematopoietic cells, including induced pluripotent stem (iPS) cells (A), non-hemogenic endothelial cells (ECs) (B), and fibroblasts (C). All these strategies involve an intermediary hemogenic endothelial state either in vivo by means of teratoma formation or in vitro. When known, the efficiency of HSC generation as assessed by measuring chimerism in graft assay in recipient mice is indicated for both short-term (ST) and long-term (LT) engraftment. na, not assessed. References: iPS cells (Tsukada et al., 2017); non-hemogenic endothelial cells (Sandler et al., 2014; Lis et al., 2017; Barcia Duran et al., 2018); fibroblasts (Pereira et al., 2013; Gomes et al., 2018; Daniel et al., 2019).

Figure 4

GFI1 and inner ear hair cells. Strategies to generate induced hair cell-like cells through trans-differentiation of non-hair cells in vitro by forced expression of GFI1 in combination with other factors. The efficiency to produce cells that harbor hair cell characteristics in vitro is indicated along the starting cell type used to achieve trans-differentiation (Costa and Henrique, 2015; Costa et al., 2015; Duran Alonso et al., 2018; Menendez et al., 2020).