GATA2 +9.5 Enhancer: From Principles of Hematopoiesis to Genetic Diagnosis in Precision Medicine

Abstract

Purpose of review

By establishing mechanisms that deliver oxygen to sustain cells and tissues, fight life-threatening pathogens and harness the immune system to eradicate cancer cells, hematopoietic stem and progenitor cells (HSPCs) are vital in health and disease. The cell biological framework for HSPC generation has been rigorously established, yet recent single-cell transcriptomic analyses have unveiled permutations of the hematopoietic hierarchy that differ considerably from the traditional roadmap. Deploying mutants that disrupt specific steps in hematopoiesis constitutes a powerful strategy for deconvoluting the complex cell biology. It is striking that a single transcription factor, GATA2, is so crucial for HSPC generation and function, and therefore it is instructive to consider mechanisms governing GATA2 expression and activity. This review focuses on an essential GATA2 enhancer (+9.5) and how +9.5 mutants inform basic and clinical/translational science.

Recent findings

+9.5 is essential for HSPC generation and function during development and hematopoietic regeneration. Human +9.5 mutations cause immunodeficiency, myelodysplastic syndrome and acute myeloid leukemia. Qualitatively and quantitatively distinct contributions of +9.5 cis-regulatory elements confer context-dependent enhancer activity. The discovery of +9.5 and its mutant alleles spawned fundamental insights into hematopoiesis, and given its role to suppress blood disease emergence, clinical centers test for mutations in this sequence to diagnose the etiology of enigmatic cytopenias.

Summary

Multi-disciplinary approaches to discover and understand cis-regulatory elements governing expression of key regulators of hematopoiesis unveil biological and mechanistic insights that provide the logic for innovating clinical applications.

INTRODUCTION

The diversification of hematopoietic stem cells (HSCs) into multipotent and lineage-committed progenitors, which generate the essential blood cells, involves complex cellular transitions that have been extensively studied. Applying single-cell transcriptomics to hematopoietic stem and progenitor cells (HSPCs) extracted from their microenvironments has introduced new permutations [1,2] of the classical hematopoietic cell developmental hierarchy [3]. This work has led to a model in which HSCs generate a continuous stream of differentiated progeny, an entirely different paradigm relative to models invoking well-defined cellular intermediates. Though the cell biological complexity of hematopoiesis is formidable, it is striking that a single transcription factor, GATA2, is so crucial for HSPC generation and function. Given this vital role, it is instructive to consider how GATA2 expression and activity are controlled in the distinct cellular contexts of the hematopoietic system.

The transcription factor GATA2 is essential for developmental and regenerative hematopoiesis, including hematopoietic stem cell (HSC) emergence, maintenance of HSC activity, myeloid and myelo-erythroid progenitor cell differentiation and erythroid precursor cell maintenance. Gata2-null mice exhibit impaired multi-lineage hematopoiesis and die at ~E10.5 [4]. GATA2 triggers the endothelial to hematopoietic transition in the mouse embryo and a comparable transition has been modelled with human induced-pluripotent cells (iPSCs) [5–7]. GATA2 depletion reduces hematopoietic stem and progenitor cell levels and function in mice, human cord blood [8] and iPSC [5,9] systems. Hypomorphic Gata2^fGN/fGN mice with Gata2 expression 5-fold lower in bone marrow mononuclear cells versus wild type cells [10,11] have thrombocytopenia, hyperchromic and macrocytic erythrocytes, and upon aging, develop leukocytosis [12].

Considering the vital GATA2 functions in multiple sectors of the hematopoietic hierarchy, there are many opportunities for even small changes in GATA2 levels/activity to derail hematopoietic processes. We proposed that GATA2 levels/activity need to be maintained within a restricted physiological window to establish and maintain context-dependent and fragile GATA2-dependent genetic networks. Thus, high or low GATA2 levels/activities corrupt circuits and networks that sustain steady-state hematopoiesis and enable the hematopoietic system to regenerate in physiological and pathological contexts. Supporting this concept, ectopic GATA2 overexpression in bone marrow suppresses hematopoiesis [13], and high GATA2 mRNA is associated with poor prognosis of AML in adult [14] and pediatric [15] patient cohorts. Human GATA2 mutations, which are loss-of-function or gain-of-function, dependent upon context, cause a disease termed GATA2 deficiency syndrome that is discussed later in this review.

CIS-ELEMENT REQUIREMENTS FOR +9.5 ENHANCER FUNCTION

Using erythroid cells to discover principles of genetics and epigenetics and to understand how GATA factors control erythropoiesis, we identified five conserved GATA2- and GATA1- occupied chromatin sites upstream (−77, −3.9, −2.8, and −1.8 kb relative to the transcription start site) and within an intron (+9.5 kb) of the murine Gata2 locus [16–19]. GATA2 occupancy of its own locus implies positive autoregulation, and GATA1 displacement of GATA2 instigates a GATA switch that represses Gata2 transcription [19–21]. While these GATA1/2-occupancy sites exhibit enhancer attributes, −1.8 [22] or −2.8 [23] deletions revealed they are not essential for Gata2 expression and hematopoiesis, albeit they contribute to Gata2 expression in progenitors, and −1.8 maintains Gata2 repression in maturing erythroblasts [22]. The −3.9 deletion was inconsequential for Gata2 expression and hematopoiesis [24]. However, the −77 and +9.5 deletions revealed their essential functions to support embryogenesis and hematopoiesis. The −77 deletion is embryonic lethal after E15.5, and myelo-erythroid progenitor cell fate is corrupted, despite normal HSC levels [25]. By contrast, +9.5 deletion is lethal at ~E14.5 and abrogates HSC generation in the AGM region [26,27]. The +9.5 also increases Gata2 transcription at other developmental stages and in the adult during hematopoietic regeneration following stress [28]. Based on the extraordinarily important +9.5 activity to control GATA2 levels in HSPCs and erythroid precursors, and its direct role in, and utility for diagnosis of blood diseases, this review focuses principally on +9.5 structure, function and dysfunction.

+9.5 sequence conservation suggests the presence of multiple transcription factor-binding motifs that might function additively, synergistically or redundantly (Figure 1). In addition, certain motifs might operate only in restricted physiological and/or pathological contexts. Considerable progress has been made in testing these models utilizing +9.5 mutant alleles in mice (Figure 2). Studies with strains harboring a mutant E-box-spacer-AGATAA composite element (“composite element”) [26,27] and Ets [28] (Figure 1 and 2) indicate that individual motifs can exert qualitatively and quantitatively distinct activities. While the E-box-8bp spacer-AGATAA composite element structure is highly conserved, the spacer sequence differs among species, with the mouse and human spacers differing by 2 bp. ChIP-seq data revealed GATA2 occupancy at composite elements with 6- to 14-bp spacers, and the 8-bp spacer is overrepresented [29]. The 8-bp spacer allows maximal enhancer activity in a transient transfection assay [30] and complex assembly in vitro is abrogated by E-box or GATA motif mutations [31].

Figure 1:

+9.5 enhancer sequence conservation. (A) Alignment of the +9.5 core. Deviation from the consensus sequence is indicated by red text. (B) Alignment of Gata2 intron 4 among 19 mouse strains. Conserved sequences are indicated in maroon. Insertions indicated by arrowheads and deletions or mismatches with white.

Figure 2:

Mouse Gata2 +9.5 mutant alleles.

The +9.5 cis-elements described represent canonical binding motifs for the hematopoietic transcription factors SCL/TAL1 (E-box: CANNTG) [32,33] and GATA2 [4] or GATA1 [34,35] (WGATAR), as well as Ets transcription factor (GGAW) [36] family members, e.g. ERG [37], FLI1 [38] and ETV2 [39] that also have cell-type-specific expression patterns. GATA2 and SCL/TAL1 commonly co-occupy chromatin sites [40–42], and the coregulator LMO2 can also be present [43]. Deletion of the composite element in fetal liver cells prevents GATA2 and SCL/TAL1 chromatin occupancy [24]. +9.5 enhancer activity requires E-box, GATA, and Ets motifs in cell- and transgenic mouse-based reporter assays [30,40,44], and E-box deletion from a +9.5-containing transgene abolishes expression in mouse embryo endothelium [45]. Forced expression of GATA2, TAL1, LMO2 and ETV2 in iPSCs promotes the endothelial to hematopoietic transition [46], although whether these proteins function through the +9.5 in this context was not described.

Homozygous deletion of the composite element (+9.5^−/−) [26] or the E-box and Ets motifs, while retaining the GATA motif [+9.5(E-box;Ets)^−/−] [28], permits embryonic development beyond the stage at which Gata2^−/− embryos die (E14.5 vs. E10.5) [4]. However, HSC emergence is abrogated in +9.5^−/− aorta-gonad-mesenephros (AGM) [26–28]. Since AGM-derived HSCs populate the fetal liver [47], hematopoietic stem and progenitor cells (HSPCs) are severely depleted in the mutant fetal livers. Deletion of a single +9.5 allele decreases Gata2 expression in fetal liver by ~50% and HSPC levels by ~30% [26]. Although Gata2 expression in +9.5^+/- AGM is not significantly altered, colony forming activity is 2-fold lower [27]; given the small percentage of GATA2-expressing hemogenic endothelial cells in the AGM, one would not expect to see mRNA changes with bulk RNA measurements. Competitive transplantation of heterozygous +9.5^+/- fetal liver or AGM revealed ~3-fold decreases in long-term repopulating activity [26,27]. The unique phenotypes of −77 and +9.5 mutant alleles informed hematopoiesis mechanisms.

Combining different enhancer mutant alleles in compound heterozygous mice provides an innovative strategy to generate unique mouse models with phenotypes distinct from conventional perturbations and elucidate mechanisms underlying enhancer function. The combination of a single +9.5 mutant allele with a single −77 mutant allele, either of which elicit only minor phenotypes by themselves, to yield compound heterozygous mice allows development to a later stage (~E15.5) than +9.5^−/− mutants [48]. This extended developmental window revealed +9.5 activity to regulate GATA2 expression and function in progenitor cells, which was not detected in the +9.5^−/− context in which HSC emergence is ablated. While all myeloid progenitor populations are lower in the compound heterozygous mutants, megakaryocyte erythrocyte progenitors (MEPs) are essentially eliminated, disproportionately relative to other myeloid progenitors [48].

By contrast to the multi-motif mutations of the +9.5^−/− allele [26] and the compound heterozygous mutant described above [48], a homozygous single-nucleotide Ets motif mutant, which models a human GATA2 deficiency syndrome mutation, is not embryonic lethal [28]. In this mutant, HSC emergence in the AGM decreases 2-fold, and HSCs in the fetal liver decrease 4-fold. In the steady-state, these defects do not persist in adults, although multipotent progenitors modestly decrease. However, following 5-flurouracil treatment, +9.5(Ets)^−/− mice are defective in expanding HSPCs, and HSCs are reduced in a competitive transplantation assay [28]. Thus, the Ets motif promotes regeneration after hematopoietic injury. These studies revealed differences between +9.5 function in regenerative versus developmental hematopoiesis, and given the +9.5 cis-element complexity and context-dependent activities, future studies will almost certainly further transform concepts.

MECHANISTIC INSIGHTS DERIVED FROM “+9.5-LIKE” ENHANCERS

Among the reported Gata2 enhancers, only +9.5 contains a conserved composite element. Within the human genome, nearly 9000 composite elements with CATCTG-(N8-N14)-AGATAA permutations exist [29,49]. Limiting the spacer length to 8 reduces the number to 797 in mice, 62 of which are GATA2-occupied; 34 of these were GATA2/SCL/TAL1-co-occupied. Composite elements at Bcl2l1, Dapp1, and Samd14 were functionally validated by gene editing in G1E-ER-GATA1 proerythroblast cells [29]. “+9.5-like” elements at the Kit promoter, Runx1 intron, Smad1 intron, Klf1 promoter, Ebp4.2 promoter, Gata1 promoter and Smad5 promoter exhibit enhancer activities in transfection and transgenic mouse assays but have not been functionally analyzed at their endogenous loci [49]. SCL/TAL1, GATA2, the Ets factor PU.1 and coregulators LMO2 and LDB1 occupy the Runx1 intronic enhancer [50]. As SCL/TAL1 occupancy can occur even without an E-box adjacent to the GATA motif [51], how different configurations of GATA motif-containing cis-elements translate into unique functions remains elusive. Composite elements also reside at select GATA1-occupied chromatin sites [29,41] and are likely to be broadly important in GATA2 and GATA1 contexts.

Based on the Ets motif (GGAW) sequence simplicity [36] and existing mechanistic knowledge, it is impossible to predict whether a particular Ets motif is essential, contributory or inconsequential for enhancer function. Furthermore, the rules governing the specificity of how Ets factor family members occupy chromatin in vivo are unknown. Although comparative genomics involving the integration of diverse datasets, including conservation, ChIP-seq data, natural genetic variation and patient mutations, can be used to stratify sites to unveil important insights, genetic editing is essential to establish important functions at endogenous loci [29].

DEVELOPMENTAL VERSUS REGENERATIVE +9.5 ENHANCER FUNCTIONS

A mechanism that controls hematopoiesis during embryogenesis and in the adult may have mechanistic components that are essential in only one of these contexts. This concept is exemplified by the critical GATA2 activity to control hematopoiesis in multiple contexts, while the essential +9.5 and −77 activities to regulate GATA2 expression and hematopoiesis are context-dependent. It is instructive therefore to compare and contrast +9.5 functions during development and regeneration.

The transcription factors SCL/TAL1 and GATA2 and the coregulator LMO2 are expressed in HSPCs and in certain lineage-committed progenitor cells. In the distinct regulatory milieus, presumably, these factors assemble on composite elements at target gene ensembles in a cell type-specific manner, although many questions remain unanswered regarding mechanisms governing multimeric complex assembly and function [33,49]. GATA2-mediated transcriptional activation is enhanced in a context-dependent manner by multi-site phosphorylation [52,53], and other components of the complex are also phosphorylated [54–58]. In principle, post-translational modifications might impact complex assembly and/or function but the importance of these mechanisms in vivo and whether they operate similarly or distinctly during development and regeneration has not been described.

Differential +9.5 functions in development and regeneration may reflect differences in the levels/activities of its transcription factor components in these contexts. The Ets factor ETV2 is transiently expressed during embryogenesis and silenced once definitive HSPCs are generated [59–61]. Conditional Etv2 deletion using Tie:Cre or Vav:Cre does not impact steady-state hematopoiesis [59]. By contrast, treatment of the conditional mutant mice with polyinosinic:polycytidylic acid (pIpC), which activates interferon signaling, or 5-fluorouracil (5-FU), which kills proliferating cells and activates quiescent HSCs, causes rapid HSC depletion. ETV2 occupies +9.5 in embryonic stem cells and is required for Gata2 induction post-5-FU treatment [28]. It is likely therefore that ETV2 induction by hematopoietic stress contributes to +9.5 regenerative functions. As developmental hematopoiesis is relatively normal in +9.5(Ets)^−/− mice [28], the Ets motif activity exemplifies a context-dependent permutation of the +9.5 mechanism.

PATHOGENIC HUMAN +9.5 ENHANCER MUTATIONS

Considering the essential GATA2 activities during embryogenesis and in the adult, GATA2 dysregulation would be expected to be at the forefront of at least certain blood diseases. Human heterozygous GATA2 germline mutations cause primary immunodeficiency, myelodysplastic syndrome (MDS), acute myeloid leukemia (AML), pulmonary alveolar proteinosis, defects in the vasculature and lymphatic systems and additional complex phenotypes [62–65]. Although the penetrance is incomplete, human genetic analysis with multi-generational pedigrees implicates GATA2 dysregulation as the disease-instigating mechanism. Mutations include heterozygous point mutations, small insertions and deletions within the gene body or large deletions encompassing the gene. Predicted loss-of-function mutations in the C-terminal zinc finger can inhibit DNA binding [66,67] yet retain activity or exhibit hyperactivity, e.g. with the R307W N-finger mutant [67]. GATA2 expression levels are lower in the majority of normal karyotype-AML patients, based on RNA-seq analysis of CD34⁺ cells [68]. GATA2 expression can be lower in MDS/AML patients versus asymptomatic relatives harboring the mutation [69].

Strikingly, GATA2 deficiency syndrome germline mutations can reside in and disrupt the +9.5 [26,44]. These mutations include deletion (c.1017+513del28) [26] or substitution (c.1017+532T>A) [70] within the E-box, or, most frequently, a C>T transition in a 3’ Ets motif (c.1017+572C>T) [44,70–74]. GATA2 expression is reduced from the mutant allele [26,44]. Additional patient mutations have been detected, but based on emerging principles of +9.5 structure/function, their functional consequences are unclear (Figure 3).

Figure 3:

Human +9.5 variants. Red text indicates mutations predicted to be pathogenic. Blue text indicates variants of unknown significance. Mutations were described in ClinVar/ClinGen (https://www.clinicalgenome.org/data-sharing/clinvar/).

GATA2 +9.5 mutant patients exhibit decreased GATA2 mRNA expression, consistent with a GATA2 deficiency [44]. However, since GATA2 coding mutations can exert either loss-of-function or gain-of-function activity, depending upon the mutation and biological and molecular outputs analyzed [7,67,75], the evidence supports a model in which insufficient or excessive GATA2 activity is pathogenic. In certain patient cohorts, more severe aberrations, including frameshifts and deletions, correlate with vascular defects, including lymphedema [65,76,77]. In a European cohort, patients with missense mutations have a high risk of developing leukemia [73]. Missense mutations in the N-or C-terminal zinc fingers differ in disease presentation [78]. Much more evidence involving larger sample cohorts is required to rigorously assess potential genotype-phenotype correlations.

GATA2 deficiency syndrome patients frequently present with recurrent viral, mycobacterial, and fungal infections. Although infections may be secondary to immunodeficiency, the infectious agent may dysregulate cellular processes that control HSPCs. Inflammatory signals, including interferons, IL-1, and G-CSF, induce HSC proliferation while impairing self-renewal [79]. These alterations can skew differentiation, e.g. increase myelopoiesis and suppress erythropoiesis [80], increase granulocytes, macrophages, and dendritic cells and decrease B-cells [81], or favor monocytic differentiation at the expense of other lineages [82]. Metabolic byproducts may also alter HSPC function. Mutations of human genes encoding components of the Fanconi Anemia (FA) pathway, which resolves DNA crosslinks, lead to bone marrow failure. While FA mouse models do not spontaneously develop bone marrow failure, increasing acetaldehyde by mutation of Aldh2 elevates HSC DNA damage [83], inducing bone marrow failure and leukemogenesis. Inactivation of homologous recombination factors BRCA1, BRCA2, or RAD51 also renders cells hypersensitive to acetaldehyde [84]. It will be important to ascertain the contribution of extrinsic and intrinsic mechanisms to disease progression instigated by GATA2 mutation.

The majority of GATA2 deficiency syndrome patients have additional mutations and/or cytogenetic abnormalities. Somatic mutations in ASXL1, KRAS/NRAS, SF3B1, SETBP1 and additional genes have been described [85–87]. Within the +9.5-mutant patient cohort, chromosomal abnormalities are common, including trisomy 1q, trisomy 21 [88], complex cytogenetics including der(Y)t(Y;1)(q11.23;q21) [73], +1, +8, der(1;7)(q10;p10) [71] and, most frequently, monosomy 7 [70,72,88]. Regardless of the mutation, GATA2 deficiency syndrome patients with infections, MDS/AML, and/or CMML can be treated with hematopoietic stem cell transplant (HSCT) [89]. Overall survival declines after disease presentation [90]. Critically, donor GATA2 genotype must be established, as matched related or haploidentical potential donors may carry a GATA2 mutation even if they are asymptomatic [88].

CONCLUSION

As a critical regulator of hematopoiesis, GATA2 levels must be established and maintained through enhancer-dependent mechanisms and almost certainly additional mechanisms remaining to be discovered. These mechanisms are highly context-dependent and cannot be extrapolated from one cellular context to another or a physiological to a pathological state. GATA2 dysregulation creates a predisposition to develop or instigate pathologies including immunodeficiency, bone marrow failure and leukemia. In addition to missense, frameshift and splice site mutations, GATA2 +9.5 germline mutations occur in patients with GATA2 deficiency syndrome. These mutations highlight the importance of the essential +9.5 function to control hematopoiesis and suppress the development of blood diseases. Accordingly, it will be crucial to discover and understand the factors and signals that confer cell type-specific +9.5 activities, identify the full ensemble of GATA2-regulated genes, proteins and small molecules and elucidate how these components constitute cell type-specific regulatory networks and circuits governing stem and progenitor cell genesis and/or function.

Key Points.

• GATA2 is a critical regulator of hematopoiesis in diverse cellular contexts.

• Context-dependent establishment of GATA2 levels/activity is achieved through multiple enhancers.

• GATA2 enhancers consist of cis-regulatory elements that mediate overlapping and distinct functions during development and regeneration.

• Dysregulation of GATA2 expression (decreased or increased) generates a predisposition to or instigates pathogenesis.