We have taken advantage of the ease with which one can effect combined loss-of-function on multiple members of gene families in the zebrafish to examine the role of the ETS transcription factor family during vascular development. Gain-of-function studies in mice have suggested an important role for these factors in endothelial specification and/or differentiation, but loss-of-function studies of individual ETS factors have generally yielded less dramatic defects in vascular development, possibly because of extensively overlapping expression and functional overlap between a number of different ETS factors within the vasculature, including Ets1, Fli1, ERG, and TEL (reviewed in (Sato, 2001
)). Reduction in levels of the ETS-domain protein etsrp
has been reported to cause vascular defects in zebrafish (Sumanas and Lin, 2005
As we show in this study, there are at least three additional ETS factors expressed within the zebrafish vasculature. We sought to determine whether these other factors also contribute to formation of the vasculature and what the consequences are of combined reduction in levels of multiple vascular ETS transcription factors. We show that the four vascular ETS factors function in a combinatorial fashion in the endothelium, by using morpholinos targeting etsrp and three other vascular ETS factors (ets1, fli1, and fli1b. Simultaneous reduction in levels of all four ETS factors by combined morpholino knock-down leads to a near-total loss of endothelial differentiation.
Morpholino injections targeting the four vascular ETS factors show that they are not equivalent, (), with a relative importance etsrp > ets1 > fli1 ≥ fli1b. Reduction of etsrp by morpholino injection causes the most severe defects in trunk circulation and intersegmental vessel sprouting, and etsrp is able to support some vascular differentiation on its own, in animals injected with morpholinos against the other three factors. In contrast, knock-down of fli1 or fli1b causes only mild vascular defects and they have little or no capacity to support vascular development on their own (i. e. when the other ETS factors are knocked down). These differences could reflect differential functional capabilities of the different ETS factors. Despite a high degree of sequence homology, some ETS factors do bind slightly different sites. In addition, ETS factors have been shown to be differentially regulated by post-translational modification and interaction with a large number of accessory factors. Some of the differences in vascular ETS factor requirement that we have observed could also reflect the relative timing or amount of expression of each of the genes. The etsrp gene is expressed strongly in both anterior and posterior expression domains at the 5 somite stage, unlike the other vascular ETS genes at this stage (). Finally, cross-regulation of the ETS factors could also account for differences in the relative requirement for each gene. Loss of etsrp in etsrpy11 mutants results in strong reduction in not only etsrp but also fli1b message levels at 24 hpf (). Despite the substantial quantitative differences in relative importance of the four ETS factors, there do not appear to be obvious qualitative differences in their morpholino vascular phenotypes, and for the strongest vascular phenotype reduction in multiple factors is required. In contrast, our data suggest that etsrp plays a more critical role in maintenance of hematopoietic development. Although gata1 and gata2 expression are normal in etsrpy11 mutants during early somitogenesis (14hpf/10 somites, data not shown), nearly complete loss of the expression of gata1 and strong reduction in gata2 is observed at 24 hpf (). In contrast, injection of morpholinos against all four vascular ETS factors at “medium” dose levels results in little or no loss of expression of these genes despite dramatic effects on vascular markers (). This presumably reflects residual etsrp protein present in animals injected with “medium” doses of the etsrp morpholino.
Form our molecular characterization it is clear that the etsrpy11
mutation represents a null mutant for etsrp, with an early stop codon predicted to terminate the translated peptide early in the protein, well before the essential ETS DNA binding domain. Morpholino knockdown results in a phenocopy of the mutant defect. An earlier report of etsrp morpholino knock-down described a somewhat more severe vascular phenotype, with loss of nearly all endothelial cells (as measured by flk1-GFP transgene expression) at the highest morpholino dose levels (Sumanas and Lin, 2005
). The reasons for this difference are not clear, but it could partly reflect different vascular-specific transgenic lines used to visualize vessels (fli-EGFP in our study vs. flk-GFP in Sumanas et al.), differences in the developmental stages at which marker genes were assayed, or toxicity of high doses of injected morpholinos. In our hands, as noted above, a more complete defect in vascular development requires combined morpholino targeting of multiple ETS factors.
In conclusion, we have shown that ETS factors function combinatorially in the zebrafish vasculature and that the function of these factors is essential for endothelial specification and differentiation. These factors do not act equivalently, however, since the functional requirement for some of them is much greater than for others. Additional work will need to be done to further explore the differences in functional requirement for the different ETS factors, and determine whether any of these differences reflect differing functional capabilities of the proteins.