Retinoic acid signaling plays a restrictive role in zebrafish primitive myelopoiesis

Retinoic acid (RA) is known to regulate definitive myelopoiesis but its role in vertebrate primitive myelopoiesis remains unclear. Here we report that zebrafish primitive myelopoiesis is restricted by RA in a dose dependent manner mainly before 11 hpf (hours post fertilization) when anterior hemangioblasts are initiated to form. RA treatment significantly reduces expressions of anterior hemangioblast markers scl, lmo2, gata2 and etsrp in the rostral end of ALPM (anterior lateral plate mesoderm) of the embryos. The result indicates that RA restricts primitive myelopoiesis by suppressing formation of anterior hemangioblasts. Analyses of ALPM formation suggest that the defective primitive myelopoiesis resulting from RA treatment before late gastrulation may be secondary to global loss of cells for ALPM fate whereas the developmental defect resulting from RA treatment during 10-11 hpf should be due to ALPM patterning shift. Overexpressions of scl and lmo2 partially rescue the block of primitive myelopoiesis in the embryos treated with 250 nM RA during 10-11 hpf, suggesting RA acts upstream of scl to control primitive myelopoiesis. However, the RA treatment blocks the increased primitive myelopoiesis caused by overexpressing gata4/6 whereas the abolished primitive myelopoiesis in gata4/5/6 depleted embryos is well rescued by 4-diethylamino-benzaldehyde, a retinal dehydrogenase inhibitor, or partially rescued by knocking down aldh1a2, the major retinal dehydrogenase gene that is responsible for RA synthesis during early development. Consistently, overexpressing gata4/6 inhibits aldh1a2 expression whereas depleting gata4/5/6 increases aldh1a2 expression. The results reveal that RA signaling acts downstream of gata4/5/6 to control primitive myelopoiesis. But, 4-diethylamino-benzaldehyde fails to rescue the defective primitive myelopoiesis in either cloche embryos or lycat morphants. Taken together, our results demonstrate that RA signaling restricts zebrafish primitive myelopoiesis through acting downstream of gata4/5/6, upstream of, or parallel to, cloche, and upstream of scl.     

Knockdown of zebrafish crim1 results in a bent tail phenotype with defects in somite and vascular development

The Crim1 gene encodes a transmembrane protein containing six cysteine-rich repeats similar to those found in the BMP antagonist, chordin (chd). To investigate its physiological role, zebrafish crim1 was cloned and shown to be both maternally and zygotically expressed during zebrafish development in sites including the vasculature, intermediate cell mass, notochord, and otic vesicle. Bent or hooked tails with U-shaped somites were observed in 85% of morphants from 12 hpf. This was accompanied by a loss of muscle pioneer cells. While morpholino knockdown of crim1 showed some evidence of ventralisation, including expansion of the intermediate cell mass (ICM), reduction in head size bent tails and disruption to the somites and notochord, this did not mimic the classically ventralised phenotype, as assessed by the pattern of expression of the dorsal markers chordin, otx2 and the ventral markers eve1, pax2.1, tal1 and gata1 between 75% epiboly and six-somites. From 24 hpf, morphants displayed an expansion of the ventral mesoderm-derived ICM, as evidenced by expansion of tal1, lmo2 and crim1 itself. Analysis of the crim1 morphant phenotype in Tg(fli:EGFP) fish showed a clear reduction in the endothelial cells forming the intersegmental vessels and a loss of the dorsal longitudinal anastomotic vessel (DLAV). Hence, the primary role of zebrafish crim1 is likely to be the regulation of somitic and vascular development.