A mutational analysis of the endophilin-A N-BAR domain performed in living flies

Endophilin is a cytoplasmic protein with an important function in clathrin-dependent endocytosis at synapses and elsewhere. Endophilin has a BAR (Bin/Amphiphysin/Rvs-homology) domain, which is implicated in the sensing and induction of membrane curvature. Previous structure-function studies of the endophilin-A BAR domain have almost exclusively been made in reduced systems, either in vitro or ex vivo in cultured cells. To extend and complement this work, we have analyzed the role played by the structural features of the endophilin-A BAR domain in Drosophila in vivo.

The study is based on genetic rescue of endophilin-A (endoA) null mutants with wild type or mutated endoA transgenes. We evaluated the viability of the rescuants, the locomotor behavior in adult flies and the neurotransmission at the larval neuromuscular junction. Whereas mutating the endophilin BAR domain clearly affected adult flies, larval endophilin function was surprisingly resistant to mutagenesis. Previous reports have stressed the importance of a central appendage on the convex BAR surface, which forms a hydrophobic ridge able to directly insert into the lipid bilayer. We found that the charge-negative substitution A66D, which targets the hydrophobic ridge and was reported to completely disrupt the ability of endophilin-BAR to tubulate liposomes in vitro, rescued viability and neurotransmission with the same efficiency as wild type endoA transgenes, even in adults. A similar discrepancy was found for the hydrophilic substitutions A63S/A66S and A63S/A66S/M70Q. The A66W mutation, which introduces a bulky hydrophobic side chain and induces massive vesiculation of liposomes in vitro, strongly impeded eye development, even in presence of the endogenous endoA gene. Substantial residual function was observed in larvae rescued with the EndoA(Arf) transgene, which encodes a form of endophilin-A that completely lacks the central appendage. Whereas a mutation (D151P) designed to increase the BAR curvature was functional, another mutation (P143A, DeltaLEN) designed to decrease the curvature was not.

Our results provide novel insight into the structure/function relationship of the endophilin-A BAR domain in vivo, especially with relation to synaptic function.