Zebrafish gata2 Orthologs Are Necessary but Not Sufficient for Developmental Hematopoiesis

Germline mutations in the GATA2 gene have been identified in patients with a variety of hematological disorders unified by their with predisposition to myelodysplastic syndromes (MDS) and defective immunity. The GATA2 gene encodes a transcription factor that plays a major role in hematopoiesis. However, the precise effects of GATA2 haploinsufficiency in developmental hematopoiesis and the impact of this in MDS predisposition remain unknown. We have developed zebrafish mutants of gata2 to model this and demonstrate that heterozygosity results in defective primitive and definitive granulopoiesis.

Zebrafish carry two orthologues to the mammalian GATA2 gene: gata2a and gata2b . gata2a and gata2b demonstrate distinct expression patterns. By 32 hours post fertilization (hpf), expression of gata2a is seen throughout the dorsal aorta (DA) and cardinal vein, primitive erythrocytes and the nervous system. In contrast, gata2b expression is restricted to the ventral DA (VDA) and branchiomotor neurons. We hypothesised that this division of expression may translate into a requirement for both ohnologs as necessary but not sufficient for normal developmental hematopoiesis.

Homozygous gata2a mutants carrying an 11bp deletion ( gata2a^um27 ) have been previously described. These mutants have abnormal vascular development in the aorta resulting in defective circulation. We analysed developmental hematopoiesis in these mutants. gata2a^um27/um27 and gata2a^{um27 /+} showed normal hemangioblasts, and primitive hematopoiesis before 18 somites. At 22hpf however myeloid transcription factor cebpa showed a dose dependent decrease in expression resulting from gata2a loss, while pu.1 expression remained normal. This suggests that Gata2a may be a direct regulator of cebpa in primitive myeloid cells, or that cebpa expression defines a sub-population of myeloid cells. We then assessed hematopoietic stem cells (HSCs) in gata2a^um27 mutants. gata2a^um27/um27 showed severely decreased or absent HSCs in the caudal hematopoietic tissue (CHT, fetal liver equivalent) at 52hpf, likely due to absent blood flow in the aorta. gata2a^um27/+ mutants also showed a decrease in HSCs at this time point, however this decrease was transient, resolving by 72hpf (A). Further assessment of definitive hematopoiesis showed that gata2^um27/um27 lacked mature granulocytes and thymocytes, in keeping with the observed HSC defect. By contrast myeloid cells in gata2a^um27/+ at 4dpf (assessed by Sudan Black (SB) stain) were unaffected (B,C) . Interestingly lck-expressing thymocytes showed a range of phenotypes, from normal, through reduced and some with a complete absence of thymocytes in one of their thymii.

We next used CRISPR to generate a gata2b mutant line. This line carries a 4bp deletion leading to a truncated protein lacking both zinc fingers ( gata2b^U5008 ). In contrast to gata2a^um27/um27 , gata2b^u5008/5008 have normal circulation and were viable beyond 5dpf. Loss of gata2b by morpholino knockdown results in absence of HSC, however gata2b^5008/+ showed no difference in HSC numbers in the CHT at 60hpf. Myeloid cells were unaffected in gata2b⁵⁰⁰⁸ mutants by SB at 48hpf, however, by 4dpf gata2b⁵⁰⁰⁸ show a dose dependent reduction in myeloid cells (D). This suggests that Gata2b regulates myeloid cell proliferation or differentiation after the emergence of HSCs.

We next studied haematopoiesis in crosses of gata2a^um27/+ to gata2b^u5008/+ . Quantification of granulocytes showed normal numbers at 48hpf in all resulting genotypes. By 4dpf gata2b^5008/+ have a reduced number of SB expressing granulocytes regardless of their gata2a genotype, suggesting that Gata2a is dispensable for definitive myelopoiesis (E). Quantification of HSCs show HSC numbers are unaffected in gata2a^um27/+ ; gata2b^u5008/+ at 48hpf. Assessment of thymocytes by contrast shows a greater than additive effect on thymocyte numbers than in single mutants alone, suggesting that both Gata2a and Gata2b contribute to lymphoid development in an additive manner. Further characterisation of the effects of gata2a^um27/+ ; gata2b^u5008/+ in combination is ongoing.

In conclusion our results support the notion at that haploinsufficient levels of Gata2a and Gata2b are necessary but not sufficient for normal developmental hematopoiesis and that combined heterozygous loss of these genes models the haploinsufficieny observed in patients in GATA2 haploinsufficient diseases.

Figure

View largeDownload slide

Figure

View largeDownload slide

Close modal