gata2is Required for the runx1-Independent Hematopoiesis in Zebrafish

The current notion about how hematopoietic stem cells (HSCs) are generated identifies the transcription factor RUNX1 as an essential factor for the emergence of definitive hematopoietic stem cells (HSCs) from the hemogenic endothelium. Consequently, Runx1knockout mice fail to develop definitive hematopoiesis and lack all definitive blood lineages and cannot survive past embryonic day 12. However, even though zebrafish with arunx1stop codon mutation (runx1^W84X/W84X) presented defects in definitive hematopoiesis during embryogenesis, runx1^W84X/W84Xembryos could develop to fertile adults with blood cells of multi-lineages, raising the possibility that HSCs can emerge without RUNX1. In order to determine if a RUNX1-independent mechanism can support the generation of HSCs we have generated three new zebrafish runx1^-/- with engineered deletions of the runx1gene using TALEN and CRISPR-Cas9. Our analysis shows that all three mutants have identical phenotypei.e., failure to develop definitive hematopoiesis during early embryogenesis, with later reemergence of hematopoietic cells and survivalof therunx1 mutants to adulthood, further confirming the existence of a RUNX1-independent mechanism for the emergence of HSCs. In the absence of a functional runx1, a cd41-GFP⁺population of hematopoietic precursors can still be detected in the aorta-gonad-mesonephros (AGM) region and in the hematopoietic tissues of the mutant embryos.

Single cell RNA sequencing of the wild type and mutant HSC/HSPC at embryonic and larval stages confirmed the presence of a population of runx1^{- /-}cd41:GFP^low cells expressing HSC signature genes at 2.5 days post fertilization. At larval stages the runx1^-/-HSCs maintain their ability to generate erythroid and myeloid lineage progenitors but they present a different expression profile compared to the wild type. In order to uncover the compensatory mechanism that drives the repopulation of the hematopoietic compartment in the absence of runx1we identified the molecular signatures that separate the runx1^-/-HSC/HSPCs from the wild type and subsequently focused our attention on the transcription factors differentially expressed in the runx1^-/-HSC/HSPCs. Our analysis shows that the master transcription factor gata2b is strongly upregulated in the runx1^{- /-}HSCs during the recovery of hematopoiesis and it is also upregulated in the kidney marrow of the surviving runx1^-/-adults. Given the key role of GATA2 in the HSC development and maintenance in both mouse and zebrafish, gata2b represented a strong candidate gene with the potential ability to drive the rescue of the runx1^-/-phenotype. Indeed, a loss of function mutation or knock-down of gata2b can significantly reduce or abolish the survivability of the runx1^-/-fish, indicating that gata2bis responsible for rescuing hematopoiesis in the runx1 mutant fish.

Overall our results show that even though runx1 is necessary for the normal emergence of definitive HSCs in the embryos, in the absence of runx1the transcription factor gata2 is able to support definitive hematopoiesis that is sufficient for the embryos to develop to functional adults in the zebrafish. The current notion about how hematopoietic stem cells (HSCs) are generated identifies the transcription factor RUNX1 as an essential factor for the emergence of definitive hematopoietic stem cells (HSCs) from the hemogenic endothelium. Consequently, Runx1knockout mice fail to develop definitive hematopoiesis and lack all definitive blood lineages and cannot survive past embryonic day 12. However, even though zebrafish with arunx1stop codon mutation (runx1^W84X/W84X) presented defects in definitive hematopoiesis during embryogenesis, runx1^W84X/W84Xembryos could develop to fertile adults with blood cells of multi-lineages, raising the possibility that HSCs can emerge without RUNX1. In order to determine if a RUNX1-independent mechanism can support the generation of HSCs we have generated three new zebrafish runx1^-/- with engineered deletions of the runx1gene using TALEN and CRISPR-Cas9. Our analysis shows that all three mutants have identical phenotypei.e., failure to develop definitive hematopoiesis during early embryogenesis, with later reemergence of hematopoietic cells and survivalof therunx1 mutants to adulthood, further confirming the existence of a RUNX1-independent mechanism for the emergence of HSCs. In the absence of a functional runx1, a cd41-GFP⁺population of hematopoietic precursors can still be detected in the aorta-gonad-mesonephros (AGM) region and in the hematopoietic tissues of the mutant embryos.

Single cell RNA sequencing of the wild type and mutant HSC/HSPC at embryonic and larval stages confirmed the presence of a population of runx1^{- /-}cd41:GFP^low cells expressing HSC signature genes at 2.5 days post fertilization. At larval stages the runx1^-/-HSCs maintain their ability to generate erythroid and myeloid lineage progenitors but they present a different expression profile compared to the wild type. In order to uncover the compensatory mechanism that drives the repopulation of the hematopoietic compartment in the absence of runx1we identified the molecular signatures that separate the runx1^-/-HSC/HSPCs from the wild type and subsequently focused our attention on the transcription factors differentially expressed in the runx1^-/-HSC/HSPCs. Our analysis shows that the master transcription factor gata2b is strongly upregulated in the runx1^{- /-}HSCs during the recovery of hematopoiesis and it is also upregulated in the kidney marrow of the surviving runx1^-/-adults. Given the key role of GATA2 in the HSC development and maintenance in both mouse and zebrafish, gata2b represented a strong candidate gene with the potential ability to drive the rescue of the runx1^-/-phenotype. Indeed, a loss of function mutation or knock-down of gata2b can significantly reduce or abolish the survivability of the runx1^-/-fish, indicating that gata2bis responsible for rescuing hematopoiesis in the runx1 mutant fish.

Overall our results show that even though runx1 is necessary for the normal emergence of definitive HSCs in the embryos, in the absence of runx1the transcription factor gata2 is able to support definitive hematopoiesis that is sufficient for the embryos to develop to functional adults in the zebrafish.