No abstract

Manuscript 2Synapses are often far from the soma and independently cope with proteopathic stress induced by intense neuronal activity. However, how presynaptic compartments turnover proteins is poorly understood. We show that the synapse-enriched protein EndophilinA, thus far studied for its role in endocytosis, induces macroautophagy at presynaptic terminals. We find that EndophilinA executes this unexpected function at least partly independent of its role in synaptic vesicle endocytosis. EndophilinA-induced macroautophagy is activated when the kinase LRRK2 phosphorylates the EndophilinA-BAR domain and is blocked in animals where EndophilinA cannot be phosphorylated. EndophilinA INTRODUCTIONNeurons can be metabolically very active, firing at rates of more than 100 Hz.Synaptic proteins and organelles are used and re-used multiple times and accumulate damage as a result of this stress. Furthermore, synapses are often located far away from the cell body and must therefore in part operate independently. This raises the question how synapses maintain protein quality. Given that neurodegeneration is thought to start with subtle synaptic defects before evolving into blunt neuronal death (Burke and O'Malley, 2013;, the mechanisms of synaptic protein homeostasis are likely relevant for the understanding of neurodegenerative disease.Macroautophagy is well-placed to mediate protein turnover, but how the needs of synapses are served by this process has not been well-studied. Cellular signals like stress and amino acid deprivation induce macroautophagy, where cytoplasm is engulfed by double membrane structures before fusion with degradative lysosomes (Mizushima et al., 2011).Autophagosomes have been visualized using fluorescent markers in yeast, Drosophila and mammalian cells and are often observed as Atg8/LC3 positive puncta (Kabeya et al., 2000;Scott et al., 2004). At the stage of initiation, Atg9-positive vesicles fuse into elongated preautophagosomal structures with growing edges (He et al., 2006). These edges are highly curved and harbor lipid packing defects. These edges serve as protein docking sites, attracting specific autophagic factors such as Atg3, Atg14/Barkor and Atg1 that insert into such zones, recognizing specific lipids (phosphatidylinositol-3-phosphate (PI(3)P)) and lipid packing defects (Fan et al., 2011;Nath et al., 2014;Ragusa et al., 2012). The recruitment of these factors then promotes the further steps of autophagosome formation; in particular the E2-like protein Atg3 itself recruits the autophagic marker LC3/Atg8 (Nath et al., 2014). However, how these highly curved edges are formed and maintained is very poorly understood.While autophagy has been mostly analyzed in the soma of cultured cells and yeast, autophagic markers have also been observed away from the soma at neuronal synapses 4 (Hernandez et al., 2012;Maday and Holzbaur, 2014;Williamson et al., 2010) and these markers were shown to be transported along axons (Maday and Holzbaur, 2014). However, how autophagosomes are formed at synapses an...

Exchange of proteins at sorting endosomes is not only critical to numerous signaling pathways but also to receptor-mediated signaling and to pathogen entry into cells; however, how this process is regulated in synaptic vesicle cycling remains unexplored. In this work, we present evidence that loss of function of a single neuronally expressed GTPase activating protein (GAP), Skywalker (Sky) facilitates endosomal trafficking of synaptic vesicles at Drosophila neuromuscular junction boutons, chiefly by controlling Rab35 GTPase activity. Analyses of genetic interactions with the ESCRT machinery as well as chimeric ubiquitinated synaptic vesicle proteins indicate that endosomal trafficking facilitates the replacement of dysfunctional synaptic vesicle components. Consequently, sky mutants harbor a larger readily releasable pool of synaptic vesicles and show a dramatic increase in basal neurotransmitter release. Thus, the trafficking of vesicles via endosomes uncovered using sky mutants provides an elegant mechanism by which neurons may regulate synaptic vesicle rejuvenation and neurotransmitter release.

Presynaptic terminals are metabolically active and accrue damage through continuous vesicle cycling. How synapses locally regulate protein homeostasis is poorly understood. We show that the presynaptic lipid phosphatase synaptojanin is required for macroautophagy, and this role is inhibited by the Parkinson's disease mutation R258Q. Synaptojanin drives synaptic endocytosis by dephosphorylating PI(4,5)P 2 , but this function appears normal in Synaptojanin RQ knock-in flies. Instead, R258Q affects the synaptojanin SAC1 domain that dephosphorylates PI(3)P and PI(3,5)P 2 , two lipids found in autophagosomal membranes. Using advanced imaging, we show that Synaptojanin RQ mutants accumulate the PI(3)P/PI(3,5)P 2 -binding protein Atg18a on nascent synaptic autophagosomes, blocking autophagosome maturation at fly synapses and in neurites of human patient induced pluripotent stem cell-derived neurons. Additionally, we observe neurodegeneration, including dopaminergic neuron loss, in Synaptojanin RQ flies.Thus, synaptojanin is essential for macroautophagy within presynaptic terminals, coupling protein turnover with synaptic vesicle cycling and linking presynaptic-specific autophagy defects to Parkinson's disease.

Synaptic vesicle reformation depends on clathrin, an abundant protein that polymerizes around newly forming vesicles. However, how clathrin is involved in synaptic recycling in vivo remains unresolved. We test clathrin function during synaptic endocytosis using clathrin heavy chain (chc) mutants combined with chc photoinactivation to circumvent early embryonic lethality associated with chc mutations in multicellular organisms. Acute inactivation of chc at stimulated synapses leads to substantial membrane internalization visualized by live dye uptake and electron microscopy. However, chc-inactivated membrane cannot recycle and participate in vesicle release, resulting in a dramatic defect in neurotransmission maintenance during intense synaptic activity. Furthermore, inactivation of chc in the context of other endocytic mutations results in membrane uptake. Our data not only indicate that chc is critical for synaptic vesicle recycling but they also show that in the absence of the protein, bulk retrieval mediates massive synaptic membrane internalization.

Synapses are often far from their cell bodies and must largely independently cope with dysfunctional proteins resulting from synaptic activity and stress. To identify membrane-associated machines that can engulf synaptic targets destined for degradation, we performed a large-scale in vitro liposome-based screen followed by functional studies. We identified a presynaptically enriched chaperone Hsc70-4 that bends membranes based on its ability to oligomerize. This activity promotes endosomal microautophagy and the turnover of specific synaptic proteins. Loss of microautophagy slows down neurotransmission while gain of microautophagy increases neurotransmission. Interestingly, Sgt, a cochaperone of Hsc70-4, is able to switch the activity of Hsc70-4 from synaptic endosomal microautophagy toward chaperone activity. Hence, Hsc70-4 controls rejuvenation of the synaptic protein pool in a dual way: either by refolding proteins together with Sgt, or by targeting them for degradation by facilitating endosomal microautophagy based on its membrane deforming activity.

PI(3,4,5)P3 is a low-abundance lipid thought to play a role in the regulation of synaptic activity; however, the mechanism remains obscure. We have constructed novel split Venus-based probes and used superresolution imaging to localize PI(3,4,5)P3 at Drosophila larval neuromuscular synapses. We find the lipid in membrane domains at neurotransmitter release sites, where it concentrates with Syntaxin1A, a protein essential for vesicle fusion. Reducing PI(3,4,5)P3 availability disperses Syntaxin1A clusters and increasing PI(3,4,5)P3 levels rescues this defect. In artificial giant unilamellar vesicles, PI(3,4,5)P3 also induces Syntaxin1A domain formation and this clustering, in vitro and in vivo, is dependent on positively charged residues in the Syntaxin1A-juxtamembrane domain. Functionally, reduced PI(3,4,5)P3 causes temperature-sensitive paralysis and reduced neurotransmitter release, a phenotype also seen in animals expressing a Syntaxin1A with a mutated juxtamembrane domain. Thus, our data indicate that PI(3,4,5)P3, based on electrostatic interactions, clusters Syntaxin1A at release sites to regulate neurotransmitter release.

Studying gene function in the post-genome era requires methods to localize and inactivate proteins in a standardized fashion in model organisms. While genome-wide gene disruption and over-expression efforts are well on their way to vastly expand the repertoire of Drosophila tools, a complementary method to efficiently and quickly tag proteins expressed under endogenous control does not exist for fruit flies. Here, we describe the development of an efficient procedure to generate protein fusions at either terminus in an endogenous genomic context using recombineering. We demonstrate that the fluorescent protein tagged constructs, expressed under the proper control of regulatory elements, can rescue the respective mutations and enable the detection of proteins in vivo. Furthermore, we also adapted our method for use of the tetracysteine tag that tightly binds the fluorescent membrane-permeable FlAsH ligand. This technology allows us to acutely inactivate any tagged protein expressed under native control using fluorescein-assisted light inactivation and we provide proof of concept by demonstrating that acute loss of clathrin heavy chain function in the fly eye leads to synaptic transmission defects in photoreceptors. Our tagging technology is efficient and versatile, adaptable to any tag desired and paves the way to genome-wide gene tagging in Drosophila.