Variation in phenotypes from a Bmp-Gata3 genetic pathway is modulated by Shh signaling

Abstract

We sought to understand how perturbation of signaling pathways and their targets generates variable phenotypes. In humans, GATA3 associates with highly variable defects, such as HDR syndrome, microsomia and choanal atresia. We previously characterized a zebrafish point mutation in gata3 with highly variable craniofacial defects to the posterior palate. This variability could be due to residual Gata3 function, however, we observe the same phenotypic variability in gata3 null mutants. Using hsp:GATA3-GFP transgenics, we demonstrate that Gata3 function is required between 24 and 30 hpf. At this time maxillary neural crest cells fated to generate the palate express gata3. Transplantation experiments show that neural crest cells require Gata3 function for palatal development. Via a candidate approach, we determined if Bmp signaling was upstream of gata3 and if this pathway explained the mutant’s phenotypic variation. Using BRE:d2EGFP transgenics, we demonstrate that maxillary neural crest cells are Bmp responsive by 24 hpf. We find that gata3 expression in maxillary neural crest requires Bmp signaling and that blocking Bmp signaling, in hsp:DN-Bmpr1a-GFP embryos, can phenocopy gata3 mutants. Palatal defects are rescued in hsp:DN-Bmpr1a-GFP;hsp:GATA3-GFP double transgenic embryos, collectively demonstrating that gata3 is downstream of Bmp signaling. However, Bmp attenuation does not alter phenotypic variability in gata3 loss-of-function embryos, implicating a different pathway. Due to phenotypes observed in hypomorphic shha mutants, the Sonic Hedgehog (Shh) pathway was a promising candidate for this pathway. Small molecule activators and inhibitors of the Shh pathway lessen and exacerbate, respectively, the phenotypic severity of gata3 mutants. Importantly, inhibition of Shh can cause gata3 haploinsufficiency, as observed in humans. We find that gata3 mutants in a less expressive genetic background have a compensatory upregulation of Shh signaling. These results demonstrate that the level of Shh signaling can modulate the phenotypes observed in gata3 mutants.

Predicting Modifiers of Genotype-Phenotype Correlations in Craniofacial Development

“…All zebrafish transgenic lines were maintained from stocks derived from the AB wild-type strain. The following transgenic lines were used: Tg(fli1:EGFP) v1 [ 47 ], Tg(sox10:mRFP) vu2 [ 48 ], and Tg(hsp70l;GATA3:EGFP) [ 20 ]. These are referred to as fli1:EGFP , sox10:mRFP , and hs:GATA3-GFP for clarity, respectively.…”

“…The embryos were then returned to fresh embryo media and incubated at 28.5 °C. This heat shock treatment has been demonstrated to rescue the facial defects in gata3 morpholino injected embryos [ 20 ].…”

“…While HDR is quite rare, microsomia is more common, with an incidence rate of at least 1 in every 5600 embryos [ 19 ]. We have shown that the craniofacial phenotypes in gata3 mutant zebrafish are also highly variable and depend upon genetic background [ 18 , 20 ]. The variability in both human and zebrafish GATA3 mutants are likely due to variable disruption of the GATA3 transcriptional network, possibly due to environmental perturbation.…”

Embryonic Nicotine Exposure Disrupts Adult Social Behavior and Craniofacial Development in Zebrafish

“…Zebrafish are recognized as useful models for neurodevelopmental and craniofacial disorders, toxicology, and behavior. Zebrafish embryos have been used to study the effects of various chemical exposures on craniofacial development [ 12 , 13 , 14 , 15 ]. Furthermore, zebrafish are highly social, and so are useful for studying how chemical exposures can alter behavior.…”

The extended analogy of extraembryonic development in insects and amniotes

“…Furthermore, changes in GATA3 expression are associated with changes in BMP and FGF signalling in other vertebrate tissues [88,89], and both signalling pathways are required for correct amnion development in Tribolium [38,90]. BMP signalling has been shown to be functionally important for amnion development in Mus ( [53,91,92], and see below), but not FGF signalling.…”