Identification of Three Phases of Hematopoieisis in Zebrafish and Their Differential Requirements for Runx1 and Gata1 Functions.

Primitive hematopoiesis in the zebrafish takes place in the intermediate cell mass (ICM), while definitive hematopoiesis takes place in the kidney. Recently, a new transition site called caudal hematopoietic tissue (CHT), or posterior blood island (PBI) was identified. Using lineage tracing the hematopoietic cells originating from ICM were shown to transit through CHT and eventually populate kidney and thymus. However, the lineage relationship of the cells at these sites and the genetic control of early hematopoiesis in the zebrafish remain to be resolved. Transcription factors Gata1 and Runx1 are required for primitive and definitive hematopoiesis respectively in mammals, and are likely candidates as key hematopoietic regulators in the zebrafish. By ENU mutagenesis and reverse genetic screening, we have generated a zebrafish runx1 mutant line with a truncation mutation, W84X, in the runt homology domain and a hypomorphic gata1 mutant line with a missense mutation, T301K, in the C-terminal zinc finger domain. We used hypomorphic allele in combination with the previously characterized gata1 null mutation, vlad tepes (vlt) to assess the requirements for gata1 during primitive and definitive hematopoiesis. Gel-shift analysis showed that the T301K gata1 protein had reduced binding affinity for DNA as opposed to complete lack of binding by the vlt mutant protein. This reduced activity is sufficient for hematopoieisis since gata1^T301K/T301K embryos had normal circulation at all stages and survived to adulthood, while gata1^vlt/vlt embryos never developed circulation and died around 11–15 days post fertilization (dpf). On the other hand, compound heterozygous gata1^T301K/vlt embryos lacked circulation until 7 dpf, regained circulation around 8–11dpf and survived to adulthood. Analysis of markers for definitive hematopoiesis by in situ hybridzationan and crossing with transgenic Tg(cd41-GFP) fish indicated that definitive hematopoiesis was normal. These data suggest dosage effect of gata1 function during primitive and definitive stages of hematopoiesis, indicating that partial gata1 activity was sufficient for definitive hematopoiesis. Furthermore, we identified two phases of definitive hematopoiesis by characterization of the runx1 truncation mutation. runx1^W84X/W84X embryos had normal circulation until 7dpf, gradually lost circulation around 8–11dpf, stayed bloodless until 20–25dpf and the surviving embryos regained circulation, while majority of them died during the bloodless phase. Approximately twenty percent of runx1^W84X/W84X embryos survived to adulthood. By in situ hybridization, definitive hematopoietic stem cell markers, runx1 and c-myb, were not detectable in the runx1 mutant embryos. However, crossing with transgenic Tg(cd41-GFP) fish showed that cd41+ stem cells of definitive hematopoiesis were retained in the runx1^W84X/W84X embryos and migrated from ICM to CHT and then to kidney as wildtype clutch-mates. In runx1^W84X/W84X mutant Tg(gata1-GFP) and Tg(cd41-GFP) embryos the bloodless phase was accompanied by lack of gata1-GFP+ erythroid cells and cd41-GFP+ circulating thrombocytes, which reappeared after recovery of circulation. These data suggest that there are two phases of definitive hematopoiesis: larval and adult, and that runx1 is absolutely required for the larval stage. In conclusion, we have identified three stages of hematopoiesis in the zebrafish and revealed the differential dosage requirement for gata1 and runx1 during these three stages.