Gastrulation signalling makes the round go long

Abstract

We sought to determine how the Src Family Kinases, Fyn and Yes, are involved in early vertebrate development. Using morpholinos we knocked down Fyn and Yes and found that the resultant phenotype bore striking similarities to Wnt11 and Wnt5 mutant zebrafish. Further analysis proved that Fyn/Yes morphants have defective CE cell movements during gastrulation which required the small GTPase RhoA. However, Fyn and Yes are not linear components of the non-canonical Wnt pathway. Instead they appear to act in parallel converging on or upstream of RhoA. This led to the question what is upstream of Fyn and Yes? Previous research in Xenopus and the mouse has identified the C-terminal Src kinase (Csk) as a key regulator of vertebrate gastrulation. Csk phosphorylates the inhibitory tyrosine residue in the C-terminal tail of SFKs leading to inactivation. Using a morpholino’s, we knocked down Csk during early zebrafish development and found that disruption of Csk results in defective gastrulation cell movements. Importantly, we were able to rescue the Csk morphants by partially knocking down Fyn and Yes, indicating that Csk is directly upstream of these two SFKs. We next sought to determine what is upstream of Csk. The most promising candidate was Shp2, as Shp2 negatively regulates Csk by dephosphorylating the membrane bound adaptor protein PAG1. Csk is now no longer targeted to the plasma membrane and cannot inhibit membrane bound SFKs. We were able to demonstrate that Shp2 was necessary for CE cell movements during zebrafish gastrulation. These results are consistent with data from Xenopus and mouse. In addition, we show that Shp2 is upstream of Fyn and Yes by epigenetic analyses. Moreover, Shp2 knock down was rescued by active RhoA, placing Shp2 in a pathway parallel to non-canonical Wnt signalling. Our data, together with biochemical data, suggest a model which sees Shp2 negatively regulate Csk, the inhibitor of Fyn and Yes which in turn act upstream of RhoA in gastrulation cell movements. Because of Shp2’s involvement in early embryonic development we sought to determine whether the defects associated with human Noonan and LEOPARD syndrome (NS/LS) which are caused by activating and dominant negative mutations in Shp2, respectively, are also caused by defective gastrulation cell movements. Both NS and LS patients present with overlapping symptoms, including short stature, hypertelorism and heart defects. We noticed that some of these features were also present in Shp2 morphants. Here we show that zebrafish expressing either NS or LS mutated Shp2 develop defects that bear striking similarities to NS/LS patients, in that they are shorter, their eyes are spaced wider apart and they develop heart defects. Furthermore, we found that some of these defects were due to defective gastrulation. In conclusion, we identified a novel signalling pathway involving Shp2, Csk, Fyn/Yes and RhoA that acts in parallel to non-canonical Wnt signalling in gastrulation cell movements. Our data suggest that at least some of the symptoms in human NS and LS patients are caused by disruption of the Shp2-Csk-Fyn/Yes-RhoA signalling pathway and result from gastrulation cell movement defects.