A child with dyserythropoietic anemia and megakaryocyte dysplasia due to a novel 5′UTR GATA1s splice mutation

Abstract

We describe a child with dyserythropoietic anemia; thrombocytosis; functional platelet defect and megakaryocyte dysplasia. We show that (i) this constellation of hematopoietic abnormalities was due to a germline mutation within the 5′UTR of GATA1; (ii) the mutation impaired a 5′UTR GATA1 splicing site, promoting production of the shortened GATA1s isoform lacking the N-terminus; (iii) expression of the GATA1 N-terminus is restricted to erythroblasts and megakaryocytes in normal marrow, consistent with the patient’s abnormal erythropoiesis and megakaryopoiesis. Our findings provide insights into the clinically relevant in vivo function of the N-terminal domain of GATA1 in human hematopoiesis.

Introduction

Germline mutations within the X-linked transcription factor GATA1 occur in a spectrum of childhood blood disorders (reviewed in [1]). Alternative splicing produces two GATA1 isoforms [2]: full-length GATA1 (flGATA1) and shortened GATA1 (GATA1s), which lacks the N-terminal domain.

Rare defects known as “GATA1s mutations” [3] reduce synthesis of full-length GATA1 while GATA1s is still produced. Acquired GATA1s mutations drive transient myeloproliferative disorder and AML in Down syndrome patients (reviewed in [4–6]). Germline GATA1s mutations in individuals without trisomy 21 have been associated with Diamond-Blackfan anemia (DBA) [7,8], a syndrome of RBC aplasia and otherwise normal hematopoiesis [9]. DBA-associated ribosomopathy impairs GATA1 synthesis [10], perhaps because the complex structure of the GATA1 5′UTR demands error-free ribosomes for efficient translation. Thus, ribosomopathy has been mechanistically linked to GATA1 and erythropoiesis failure in DBA. However, thrombocytopenia with structural platelet abnormalities and dyserythropoiesis [11] as well as childhood myelodysplastic syndrome (MDS) [12] were reported in other GATA1s families. Since GATA1s patients are rare and clinically heterogeneous, the clinical impact of inherited GATA1s mutations on human hematopoiesis remains to be fully elucidated.

Results

We describe a child with dyserythropoietic anemia, megakaryocyte dysplasia and platelet malfunction due to a disease-causing mutation within the 5′UTR of GATA1. We explored the impact of this novel mutation on the expression of GATA1 splice variants in human bone marrow. Our findings support the role for the N-terminus of GATA1 in in vivo megakaryocyte function and erythropoiesis.

A 4-year old male developed fatigue and pallor secondary to anemia (hemoglobin: 4.4 g/dL). His growth and development were normal with no congenital malformations. His past blood counts revealed progressive macrocytic anemia with reticulocytopenia and persistent fetal hemoglobin first noted at 3 months of age, as well as chronic thrombocytosis (platelets:387-947,000/mm³) and occasional neutropenia with no frequent infections (lowest absolute neutrophil count:495/mm³) (Figure 1A). The folate and vitamin B12 levels were normal. Normal DEB chromosome-breakage test and lymphocyte telomere length excluded Fanconi anemia and dyskeratosis congenita, respectively.

Figure 1. Hypoplastic anemia, megakaryocyte dysplasia and thrombocytosis due to mutation within the 5′UTR of GATA1.

(A) Chronic anemia and thrombocytosis over the course of 5 years. (B) Decreased erythropoiesis and megakaryocyte dysplasia seen on bone marrow aspirate. (C) Immunohistochemistry with CD71 (erythroblast marker) and CD61 (megakaryocyte marker) reveals decreased erythropoiesis and accumulation of megakaryocytes in the patient’s bone marrow compared to a healthy individual. Right panel shows dysplastic megakaryocytes (black arrows) in the patient’s marrow (Wright-Giemsa stain). (D) GATA1 sequencing reveals a novel mutation in the affected child. RT-PCR and Western blotting demonstrate decreased full-length GATA1 transcript (E) and protein (F) in the patient. (G) Schematic representation of GATA1 alternative splicing (only first three exons are shown for simplicity). The GATA1^c-.21A>G mutation produces GATA1s phenotype by disrupting full-length GATA1 splicing.

DBA was initially suspected due to progressive macrocytic anemia beginning in infancy [9]. However, the patient’s bone marrow analysis revealed not only paucity of RBC precursors (Figure 1B–C) and dyserythropoiesis (Supplementary Figure 1), but also prominent megakaryocytosis (Figure 1B–C) with megakaryocyte dysplasia (Figure 1B; Supplementary Figure 2), which is not seen in classic DBA [9]. Accordingly, sequencing of the patient’s nine DBA-associated ribosome genes (RPL11, RPL35a, RPL5, RPS10, RPS17, RPS19, RPS24, RPS26, and RPS7) and deletion-duplication analysis of RPS19, RPL5, RPL11, RPL35A, RPS17 and RPS26 produced normal results. Normal cytogenetics, MDS-FISH and blast count excluded MDS. Thus, we asked whether his anemia mimicking DBA [7,8] but associated with megakaryocyte dysplasia [11] reflected a germline GATA1 defect. Indeed, Sanger sequencing revealed a novel mutation within the 5′UTR of GATA1 (c.-21A>G or c.-19-2A>G) at position 48,791,089 on the X-chromosome (GRCh38.p2 primary assembly), which affects the absolutely conserved A in the -2 position of the splice acceptor site (Figure 1D). Consistent with an X-linked recessive inheritance, mother was an asymptomatic carrier, and all healthy male siblings had wild-type GATA1 (Figure 1D).

We hypothesized the GATA1^c-.21A>G transcript splicing is abnormal as our in silico analysis [13] suggested disruption of a 5′UTR consensus splice site. Accordingly, the GATA1^c-.21A>G mutation decreased in vivo flGATA1 expression (Figure 1E–F), which is consistent with the GATA1s phenotype [7,8,11,12]. The patient’s anemia improved on corticosteroids similar to other GATA1s individuals [7,8,12]. Thus, the GATA1^c.-21A>G mutation produced the GATA1s phenotype through destruction of a splice site within the 5′UTR of GATA1 (Figure 1G).

Past clinical reports have generated conflicting in vivo data regarding the impact of inherited GATA1s mutations on megakaryopoiesis in non-Down syndrome patients [7,8,11]. However, accumulation of dysplastic megakaryocytes in the GATA1^c.-21A>G marrow with peripheral thrombocytosis strongly suggests that flGATA1 regulates human megakaryocyte proliferation and function in vivo in individuals without trisomy 21. Moreover, the GATA1^c.-21A>G platelets displayed subclinical aggregation deficiencies (Supplementary Figure 3), further implicating flGATA1 in physiological megakaryopoiesis. Thus, we methodically examined GATA1 splicing in human marrow, hypothesizing that flGATA1 expression correlates with the role of flGATA1 in distinct lineages. To that end, we developed an immunohistochemistry assay specific for the GATA1 N-terminus (Figure 2A). Full-length GATA1 was absent from the GATA1^c.-21A>G marrow, confirming that the mutation confers GATA1s phenotype (Figure 2B–C). Double-immunohistochemistry detected flGATA1 in healthy megakaryocytes and erythroblasts but not in other hematopoietic lineages (Figure 2C), consistent with the observation that loss of flGATA1 disrupted RBC and platelet production in our patient (Figure 1).

Figure 2. Full-length GATA1 expression is restricted to erythrocyte and megakaryocyte precursors during hematopoiesis.

(A) Antibodies used for immunohistochemistry. Antibody against the N-terminus of GATA1 recognizes only full-length GATA1, while the C-terminal antibody recognizes both GATA1 isoforms (flGATA1 and GATA1s). (B) Loss of flGATA1 expression in the GATA1^c.-21A>G patient’s bone marrow. Note that (i) flGATA1 is not expressed in all hematopoietic cells of a healthy individual, and (ii) GATA1s production is not affected by the GATA1^c.-21A>G mutation. (C) Expression of N-terminal GATA1 domain (flGATA1; blue) is restricted to erythroblasts and megakaryocytes during human hematopoiesis. Appropriate hematopoietic lineage markers (brown) were co-stained as shown. (D) GATA1s mutations cause a range of phenotypes from Diamond-Blackfan anemia (red) to multilineage hematopoietic dysplasia (blue). Novel mutation described in this work is marked with asterisk. The arrow indicates that the long-term risk of MDS in GATA1s patients presenting with pure RBC aplasia remains to be determined as one GATA1s patient initially diagnosed with DBA developed MDS later in childhood [12].

Discussion

GATA1 orchestrates the production of RBCs and platelets [14–16]. Alternative splicing provides an incompletely understood pathway to fine-tune this transcription factor’s activity during hematopoiesis. Full-length GATA1 differs from the short isoform (GATA1s) by the presence of the 83 amino acid-long N-terminus, which activates GATA1-driven erythropoiesis [2,17] and recruits flGATA1 to a subset of megakaryocyte and erythroblast genes [3]. Thus, GATA1 isoforms control partially overlapping but not identical transcriptional modules of erythroblast and megakaryocyte maturation.

Acquired GATA1s mutations are key drivers of transient myeloproliferation and AML in Down syndrome [4–6]. Inherited GATA1s mutations that decrease production of flGATA1 through disrupting exon 2 splice sites [7,8,11] or the initiation codon [12] have been reported in rare non-Down syndrome patients (Figure 2). GATA1s patients uniformly develop hyporegenerative anemia, confirming that the N-terminus of GATA1 is indispensable for erythropoiesis. However, since the reported clinical symptoms in GATA1s individuals range from DBA [7,8] to erythroblast and megakaryocyte dysplasia [11] and DBA progressing to MDS [12], the role of flGATA1 in other hematopoietic cell lines needs clarification.

Striking accumulation of dysplastic megakaryocytes in the GATA1^c.-21A>G patient’s marrow (Figure 1) suggests that loss of flGATA1 unleashes megakaryocyte proliferation in vivo, possibly due to de-repression of the E2F cell-cycle regulator [18]. In further support of this notion, independent ex vivo studies showed that hematopoiesis in GATA1s iPS cells is skewed towards generation of abnormal megakaryocytes at the cost of erythropoiesis [3]. Indeed, expression of the GATA1 N-terminus is restricted to erythroblasts and megakaryocytes (Figure 2), supporting a physiological role of flGATA1 in these hematopoietic lineages. This assay may provide a future screening tool for GATA1s phenotype. Intriguingly, both thrombocytopenia [11] and thrombocytosis (this work) occur in GATA1s individuals, including GATA1s patients diagnosed with DBA [8]. A systematic evaluation of megakaryopoiesis in additional GATA1s patients will determine whether megakaryocyte dysplasia is universally seen in GATA1s individuals. Potential genotype-phenotype correlations will remain unknown until more GATA1s patients are identified, especially since the same GATA1^c.220G>C mutation caused DBA-like phenotype in one family [8] and multilineage dysplasia in another [11].

The GATA1^c.-21A>G platelets are defective (Supplementary Figure 3). This is in agreement with ultrastructural platelet abnormalities described in GATA1^c.220G>C family [11], likely secondary to disrupted cytoskeletal remodeling in flGATA1-deficient megakaryocytes [19]. Future studies will determine whether GATA1s patients develop clinically significant bleeding later in life.

Prominent megakaryocyte abnormalities (this work; [11]) may provide subtle clinical clues to differentiate GATA1s mutations from “classical” DBA caused by ribosomopathy, which is defined as pure red blood cell aplasia [9]. GATA1 sequencing should be considered in males with congenital multi-lineage dysplasia and/or DBA-like clinical presentation. Interestingly, MDS-associated spliceosome mutations globally alter expression of multiple hematopoietic regulators, including GATA1 [20]. Given the potential risk of MDS in GATA1s patients [12], hematopoiesis should be closely monitored in individuals with inborn GATA1s mutations.

Supplementary Material

Supp Material

Supplemental Text 1. This file contains supplemental materials, methods and references, as well as three supplemental figures.

Abbreviations key

CD

Cluster of differentiation

DEB

Diepoxybutane

DBA

Diamond-Blackfan anemia

FISH

Fluorescent in situ hybridization

flGATA1

Full-length GATA1

GATA1

Globin transcription factor 1

GATA1s

Shortened GATA1

iPS cells

Induced pluripotent stem cells

IRB

Institutional Review Board

MDS

Myelodysplastic syndrome

NIH

National Institutes of Health

RBC

Red blood cell(s)

RNA

Ribonucleic acid

RPL

Ribosomal protein L

RPS

Ribosomal protein S

RT-PCR

Reverse transcription-polymerase chain reaction

5′UTR

5′ untranslated region