The adult hematopoietic system is derived from hematopoietic stem cells (HSCs) situated within the bone marrow (BM). According to Waddington’s epigenetic theory, various blood cell types originate from unstable stem/progenitor cells and eventually fall into a stable cell fate development track (Waddington, 1957; Ladewig et al., 2013) producing myeloid and lymphoid cells for immunity, erythrocytes for oxygen and carbon dioxide transport and platelets for coagulation. The process of hematopoietic lineage formation resembles a branching tree structure (Figure 1). Within the human bone marrow, the apex point of this classical branching structure is self-renewing HSCs which are typically characterized by the phenotype CD49f⁺CD90⁺CD45RA^–CD34⁺CD38^–LIN^– (Notta et al., 2011).

Numerous genetic mutations result in hematopoietic disorders with unbalanced lineage output, such as RUNX1 mutations, leading to familial platelet disorder (Preudhomme et al., 2009), IRF8 mutations resulting in mononuclear phagocytes-related human primary immunodeficiencies (Hambleton et al., 2011), and mutations in ELANE or HAX1 resulting in severe congenital neutropenia (Ye et al., 2011). A prime example is GATA2 deficiency syndrome (Hahn et al., 2011; Hsu et al., 2011; Ostergaard et al., 2011; Spinner et al., 2014; Calvo and Hickstein, 2023). GATA2 deficiency syndrome, caused by germline mutations in the hematopoietic transcription factor GATA2, stands out because multiple lineages can be affected and patients often present with monocytopenia, B cell deficiency, NK (natural killer) cell deficiency and Dendritic Cell deficiency (Dickinson et al., 2011; Novakova et al., 2016). Neutropenia also occurs in GATA2 deficiency patients (Pasquet et al., 2013) and inversions of the CD4/CD8 T cell ratio have been reported (Mutsaers et al., 2013; Ganapathi et al., 2015), indicating that GATA2 plays a crucial role as a key component within the BM hematopoietic hierarchy, orchestrating the differentiation and maintenance of diverse hematopoietic cell lineages. Furthermore, GATA2 deficiency syndrome patients have a high predisposition to develop (pediatric) myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML) with a median age of onset of 17 years (Wlodarski et al., 2016; Homan et al., 2021), however, before the onset of malignancy, the disease is also life-threatening due to anemia, bleeding disorders, or immunodeficiency with nontuberculous mycobacterial infections (NTM), fungal infections, and human papillomavirus (HPV) infections (Spinner et al., 2014; Ganapathi et al., 2015; Calvo and Hickstein, 2023). Therefore, it is vital to understand the role of GATA2 in the molecular determinants of hematopoietic cell fate.

A schematic representation of the classical tree-like hematopoiesis model shows formation of the various lineages in human bone marrow. The HSC population forms the apex of this hierarchical model, and differentiates into distinct lineages. Important modulators of the lineage choices are depicted, such as transcription factors, transcription regulators, membrane-bound receptors, cytokines, and epigenetic regulators.

To preserve the hematopoietic system, HSCs are required to self-renew. To preserve the self-renewal capacity of HSCs in the BM microenvironment, a variety of extracellular and intracellular factors must provide support. Extrinsically, different cellular factors, such as stem cell factor (SCF) and thrombopoietin (TPO), organize a coordinated extracellular microenvironment to preserve the self-renewal and maintenance of HSCs (Stoffel et al., 1999; Ema et al., 2000; Fox et al., 2002; Yoshihara et al., 2007; Mendelson and Frenette, 2014; Kokkaliaris et al., 2016). Intrinsically, the self-renewal of HSCs is influenced by multiple transcription factors, including GATA2, GFI1, and EVI1, and epigenetic regulatory molecules, such as TET2 and DNM3TA (Zhu and Emerson, 2002; Hock et al., 2004; Huck et al., 2014; Jeong et al., 2018; Xavier-Ferrucio and Krause, 2018; Aljoufi et al., 2022). GATA2 has various roles in supporting the maintenance of adult HSC characteristics. Complete knockout of Gata2 in mice results in apoptosis of HSCs (Tsai et al., 1994; de Pater et al., 2013; Gao et al., 2013). In proliferating HSCs, Gata2 expression is activated by EVI1 and it was shown that haploinsufficiency of Gata2 impairs cell cycle in mice (Ling et al., 2004; Yuasa et al., 2005). Interestingly, a Gata2 reporter mouse model showed that all HSCs have intermediate levels of Gata2 and that Gata2 is variable in multipotent hematopoietic progenitor cells, suggesting that different levels of Gata2 influence lineage determination (Kaimakis et al., 2016). Interestingly, Gata2 protein levels were observed to be constantly fluctuating in embryonic definitive HSPC formation during the endothelial-to-hematopoietic transition (EHT), indicating that Gata2 expression is a dynamic process in HSPC generation, likely required for normal lineage differentiation. Gata2 heterozygous animals displayed reduced Gata2 protein fluctuations and this may be the underlying cause of the lineage differentiation defects (Eich et al., 2018). Together, this shows that the gene dosage of Gata2 in embryonic and adult HSPCs is crucial for normal lineage differentiation.

As HSCs differentiate into various hematopoietic lineages, they receive extrinsic and intrinsic signals that prompt specialization towards specific blood cell lineages, resulting in the gradual reduction of self-renewal and multi-potency. Extrinsically, cytokines, including Flt3L, SCF, granulocyte colony-stimulating factor (G-CSF), interleukin-1 (IL-1), interleukin-3 (IL-3), interleukin-6 (IL-6), and interleukin-11 (IL-11), coordinate the development of multipotent progenitors (MPPs) from HSCs. SCF expression can be detected in several niche cells, including osteoblasts, endothelial cells and LepR⁺ perivascular stromal cells, suggesting the importance of the microenvironment for HSC maintenance and differentiation (Ding et al., 2012; Zhou et al., 2014; Zhou et al., 2017).

MPPs are heterogeneous with distinct transcriptomic characteristics. Combined single-cell barcoding and transcriptional analysis reported that MPPs in mice could be further defined as MPP1, MPP2, MPP3, and MPP4, which showed different features and lineage bias through cell fate decisions (Rodriguez-Fraticelli et al., 2018). The first lineage priming separates myeloid and lymphoid differentiation from erythroid lineage differentiation (Notta et al., 2016; Belluschi et al., 2018). MPPs are gradually directed to the myeloid and lymphoid lineages (Velten et al., 2017). Upregulation of Rag1, Ikzf1, and Ebf1 in the MPP population will lead to lymphoid bias, while the upregulation of Cebpa and Irf8 will lead to myeloid bias (Wolfler et al., 2010; Pietras et al., 2015; Lenaerts et al., 2022). Although differentiation does not occur in a clear step-wise manner, several progenitors like Lymphoid-Primed Multipotent Progenitors (LMPPs), Common Myeloid Progenitors (CMPs) and Common Lymphoid Progenitors (CLPs) can be recognized and will be discussed as such.

Human primary cells remain a treasured resource and are widely used to understand the molecular processes underlying hematopoietic cell fate. For instance single cell RNA sequencing from primary patient samples showed clear lineage differentiation defects that reflect the defects observed in patients due to increases in Mk/erythroid priming genes such as GATA1 and decreases in myeloid priming genes and lymphoid priming genes such as SPI1 originating in the HSPC population (Wu et al., 2020). Furthermore, recent methylation data from GATA2 deficiency patient samples clearly distinguished symptomatic and asymptomatic patients from healthy donors and highlighted the changes in methylation that underlie leukemia development in these patients (Marin-Bejar et al., 2023).

Unfortunately, due to genetic heterogeneity and the inability to genetically alter these cells, a true comparative study remains impossible with primary cells. Therefore, there will always be a need for model systems to study human disease. Current zebrafish and mouse models of GATA2 deficiency syndrome were only able to partially phenocopy the lineage differentiation defects observed in patients (Dobrzycki et al., 2020; Abdelfattah et al., 2021; Gioacchino et al., 2021a; Avagyan et al., 2021; Gioacchino et al., 2021b; You et al., 2022; Mahony et al., 2023). This could be due to significant disparities between animal models and humans concerning adult size, aging and niche components. The differences may be caused by variations in the quantities of hematopoietic progenitor cells (HPCs) and osteoblasts between these species (Jung et al., 2005; Asada et al., 2015). Differences in population sizes may have impact on the concentration of signaling molecules in human and mouse bone marrow. Furthermore, there are significant differences in cytokine release by the various niche components and differences in cytokine requirements of HSPCs between mouse and human (Scheerlinck, 2014).

To solve these barriers, ex vivo human cell models can be used like induced pluripotent stem cells (iPSCs) and embryonic stem cells (ESCs). In iPSCs and ESCs the embryonic hematopoiesis can be simulated, but up to a limited extent. The three waves of embryonic hematopoiesis, from 1) early primitive hematopoiesis, to 2) definitive progenitor hematopoiesis, to 3) definitive HSC formation can only be partially simulated (Sturgeon et al., 2014; Ng et al., 2016; Dzierzak and Bigas, 2018), and the formation of definitive HSCs has not been documented to date. A clear advantage of these models is the fact that these cell models can easily be genetically altered. Atkins et al. also utilized both human ESCs and iPSCs to understand human primitive embryonic hematopoiesis and further detailed the developmental mechanisms of erythroid-myeloid progenitor and lymphoid lineages (Atkins et al., 2022). iPSCs have been used to investigate lineage development (Carcamo-Orive et al., 2017), of B cell (Zhang et al., 2022), T cell (Wang Z. et al., 2022c), NK cell (Zhu et al., 2020; Woan et al., 2021), erythroid lineage (Xin et al., 2021) and myeloid lineage (Mulero-Navarro et al., 2015), providing insights into the molecular mechanisms of various lineages decisions, cellular maturation and cell function. Besides modelling normal hematopoiesis, pluripotent stem cells have been used to study malignant hematopoiesis. Patient-specific iPSCs have been used to understand the mechanisms of leukemic transformation, as well as screening patient-specific drugs (Turhan et al., 2019; Olofsen et al., 2020; Bigas et al., 2022; Olofsen et al., 2023). It must be noted that iPSC of patients with GATA2 mutations showed only marginal differences. Specifically, GATA2 patient-specific iPSCs exhibited nuanced differentiation phenotypes dependent upon the tissue which the iPSCs were derived from. Hematopoietic maturation was reduced from iPSC where GATA2 was mutated using CRISPR/Cas9. This heterogeneity in differentiation outcomes hampers the investigation of the role of GATA2 in lineage differentiation using this model system (Jung et al., 2018).

In addition to employing human cell models, the utilization of humanized animal models represent a valuable method to investigate the functional role of GATA2 in hematopoietic lineage determination. The most common humanized animal model to study hematopoiesis is the NSG immunodeficient mouse model, which allows us to study the mechanism of hematopoietic lineage determination with human cells in vivo (Adigbli et al., 2020). By xenotransplantation, it is possible to trace the differentiation of HSPCs carrying GATA2 mutations. However, as previously elucidated, it is crucial to consider the impact of microenvironmental components on hematopoiesis. Although the differentiation of human HSPCs can be activated in mouse bone marrow by the expression of human cytokines, the biological difference between mouse and human should be considered and may not represent the best model to study lineage differentiation defects.

Another complication in the study of GATA2 deficiency syndrome is the vast variety between patients. Some patients remain asymptomatic, while others suffer from immune deficiencies and yet others develop myeloid malignancies at an early age. Important considerations are the many different mutation types that are found between these patients, but also the environmental factors that contribute to our health, i.e., secondary injuries like infection. Inflammation has been recognized as driver of leukemogenesis and could contribute to disease progression and the variety observed between these patients (Essers et al., 2009; Rodriguez-Meira et al., 2023). A more likely model for GATA2 deficiency may thus be the current mouse models with an addition of a secondary injury like transplantation or stimulation with LPS, know to induce inflammation (Abdelfattah et al., 2021).

GATA2 has a pivotal role in HSC self-renewal and hematopoietic lineage determination. The precise expression regulation of GATA2 has a profound impact on hematopoietic development, as high expression of GATA2 is required for HSC self-renewal and maintenance, while procedural downregulation is imperative to facilitate downstream lineage differentiation. Recent advances in the field in terms of new animal models to understand the precise role of GATA2 in lineage differentiation can significantly contribute to the development of treatments for the life-threatening cytopenias from which the majority of GATA2 deficiency syndrome patients suffer (Gioacchino, et al., 2021a; Gioacchino et al., 2021b; Mahony et al., 2023). In recent years, the employment of single cell sequencing technologies resulted in remarkable progress in the comprehension of the molecular regulatory processes governing hematopoietic cell fate determination (Ranzoni et al., 2021; Zeller et al., 2023; Zhang et al., 2023). These technological advancements have supported and continue to support the deconstruction of the functions of GATA2 in hematopoiesis, and the pathophysiological mechanisms behind GATA2 deficiency syndrome. The precise mechanism behind GATA2 deficiency-related immunodeficiency, the variation between patients and the progression to myeloid leukemia remains to be elucidated. Furthermore, considering the different genetic backgrounds and inflammatory burden, caution should be exercised when addressing research questions and conclusions between human and animal models. Thus, there is a need to continue the development of animal or human cell-based research models.