Search

The RET receptor tyrosine kinase plays important roles in regulating cellular proliferation, migration, and survival in the normal development of neural crest derived tissues. However, aberrant activation of RET, through oncogenic mutations or overexpression, can contribute to tumourigenesis, regional invasion, and metastasis of several human cancers. RET is expressed as two main isoforms with unique C-terminal sequences that differ in protein interactions and subcellular trafficking in response to RET activation, and which also have distinct oncogenic potentials. The long isoform, termed RET51, is internalized from the membrane in response to stimulation by its ligand, GDNF, but is known to recycle back to the surface via RAB11 endosomes. However, the mechanisms regulating this process and its cellular effects have not been defined. Here, we show that recycling of RET51 requires a multicomponent complex that includes the endosomal-sorting protein GGA3, which mediates GDNF-dependent slow recycling of RET51 receptors to the plasma membrane. Our data show that the GRB2 adapter associates with RET51 through interactions with its C-terminal sequences, facilitating recruitment of active ARF6 and GGA3 interaction, and that depletion of GGA3 or ARF6 reduced RET51 recycling. Further, GGA3 knockdown accelerated RET51 degradation and also attenuated RET-mediated AKT activation. Finally, we showed that recycling of RET51 to the cell surface through association with GGA3 and ARF6 contributes to RET51-dependent cell motility, migration, and invasion. Our data establish RET recycling as a mechanism coordinating location and duration of RET signals in order to direct cell movement and invasion.

The RET receptor tyrosine kinase is implicated in normal development and cancer. RET is expressed as two isoforms, RET9 and RET51, with unique C-terminal tail sequences that recruit distinct protein complexes to mediate signals. Upon activation, RET isoforms are internalized with distinct kinetics, suggesting differences in regulation. Here, we demonstrate that RET9 and RET51 differ in their abilities to recruit E3 ubiquitin ligases to their unique C-termini. RET51, but not RET9, interacts with, and is ubiquitylated by CBL, which is recruited through interactions with the GRB2 adaptor protein. RET51 internalization was not affected by CBL knockout but was delayed in GRB2-depleted cells. In contrast, RET9 ubiquitylation requires phosphorylation-dependent changes in accessibility of key RET9 C-terminal binding motifs that facilitate interactions with multiple adaptor proteins, including GRB10 and SHANK2, to recruit the NEDD4 ubiquitin ligase. We showed that NEDD4-mediated ubiquitylation is required for RET9 localization to clathrin-coated pits and subsequent internalization. Our data establish differences in the mechanisms of RET9 and RET51 ubiquitylation and internalization that may influence the strength and duration of RET isoform signals and cellular outputs.This article has an associated First Person interview with the first authors of the paper.