Role of epsin 1 in synaptic vesicle endocytosis

Abstract

Epsin has been suggested to act as an alternate adaptor in several endocytic pathways. Its role in synaptic vesicle recycling remains, however, unclear. Here, we examined the role of epsin in this process by using the lamprey reticulospinal synapse as a model system. We characterized a lamprey ortholog of epsin 1 and showed that it is accumulated at release sites at rest and also at clathrin-coated pits in the periactive zone during synaptic activity. Disruption of epsin interactions, by presynaptic microinjection of antibodies to either the epsin-N-terminal homology domain (ENTH) or the clathrin/AP2 binding region (CLAP), caused profound loss of vesicles in stimulated synapses. CLAP antibody-injected synapses displayed a massive accumulation of distorted coated structures, including coated vacuoles, whereas in synapses perturbed with ENTH antibodies, very few coated structures were found. In both cases coated pits on the plasma membrane showed a shift to early intermediates (shallow coated pits) and an increase in size. Moreover, in CLAP antibody-injected synapses flat clathrin-coated patches occurred on the plasma membrane. We conclude that epsin is involved in clathrin-mediated synaptic vesicle endocytosis. Our results support a model, based on in vitro studies, suggesting that epsin coordinates curvature generation with coat assembly and further indicating that epsin limits clathrin coat assembly to the size of newly formed vesicles. We propose that these functions of epsin 1 provide an additional mechanism for generation of uniformly sized synaptic vesicles.

Fig. 1.

Characterization of a lamprey epsin 1 ortholog. (A) Domain structure of lamprey epsin. Identity to human epsin 1 for each region is indicated in percentage. Antibodies were raised to residues 42–207 (ENTH) and 262–496 (CLAP) (black lines). (B) Liposome sedimentation assay with wild-type ENTH domain (with and without liposomes) and mutant ENTH domain (with and without liposomes), stained with Coomassie blue. (C) GST-CLAP pull-down with lamprey brain extract. Clathrin and AP2 were detected with clathrin heavy chain and α-adaptin antibodies, buffer, or extract with GST alone. (D) Protein extracts from lamprey brain analyzed by SDS/PAGE and Western blot using ENTH and CLAP antibodies. (E) (Upper) Immunoprecipitation of lamprey brain extract with ENTH and CLAP antibodies. Also shown are control incubations with buffer alone and with control IgGs (2.5% of the pellet was loaded). (Lower) Shown is the remaining signal in the unbound material, control IgG, and 1% of the starting material. (F) Western blot analysis of GST-CLAP pull-downs showing selective inhibition of the binding of clathrin and AP2 by CLAP antibodies but not by control IgGs. (G) Pull-down with the GST-β appendage domain of AP2 from lamprey brain extract. Coomassie staining (Left) and silver staining (Right) are shown. (H) Western blot analysis of bound material vs. start material with AP180 and epsin CLAP antibodies (at locations indicated in G). (I) Effects of epsin CLAP and epsin ENTH antibodies on epsin binding (from lamprey brain extract) to the GST-β appendage domain of AP2. Note the reduced binding of epsin induced by the CLAP antibodies.

Fig. 2.

Epsin is accumulated at synaptic regions and clathrin-coated pits. (A) (Upper) Confocal images from the axonal surface of a giant axon coinjected with Alexa488-labeled ENTH and Alexa546-labeled CLAP antibodies. (Lower) Confocal images from an axon coinjected with Alexa488-labeled VAMP-2 antibodies and Alexa546-labeled CLAP antibodies. (B) Electron micrographs of clathrin-coated pits at the periactive zone of a reticulospinal synapse stimulated with 30 mM K⁺ for 30 min and labeled with CLAP antibodies. Note the gold particles (5 nm) associated with clathrin-coated pits. (Scale bars: A, 1 μm; B, 200 nm.)

Fig. 3.

Microinjection of CLAP antibodies impairs synaptic vesicle recycling. (A) Electron micrograph of a control synapse stimulated at 5 Hz for 30 min. (B) A synapse from an axon microinjected with CLAP antibodies and stimulated at 5 Hz for 30 min. Note the depletion of synaptic vesicles and the occurrence of clathrin-coated structures at cisternae (arrows) and the plasma membrane (open arrows). a, axoplasmic matrix; d, dendrite; sv, synaptic vesicles. (Scale bars: 200 nm.)

Fig. 4.

Microinjection of CLAP antibodies shifts the proportion of coated pits to early stages and increases their size. (A and B) Shallow coated pits from control (A) and CLAP antibody-injected (B). (C and D) Examples of nonconstricted (bucket-shaped) coated pits from control (C) and CLAP antibody-injected axons (D). (E) Flat clathrin-coated patch on the presynaptic membrane (arrows) from an axon microinjected with CLAP antibodies. A coated pit is also present to the left. (Scale bars: 100 nm.) (F) Quantitative analysis showing representation of different stages of coated pits in synapses microinjected with CLAP antibodies as compared with noninjected controls (n = 181 coated pits for CLAP antibody-injected axons, n = 109 coated pits for control synapses; P < 0.001 with χ² test). (G) Length of the invaginated part of shallow coated pits (P < 0.001, Student's t test) and nonconstricted coated pits (P < 0.001, Student's t test; n = 15 coated pits for each condition).

Fig. 5.

Microinjection of CLAP antibodies causes accumulation of enlarged coated structures. Electron micrographs of extended clathrin coats on membrane structures in synapses microinjected with CLAP antibodies. (A) Large coated structures with vacuolar and tubular regions. (B) Large clathrin-coated structure with connection to the plasma membrane. (C) Large clathrin-coated vacuolar structure with pit-like extensions (arrows). (D) Stereoscopic view of the surface of a vacuolar structure captured in the plane of sectioning (250-nm-thick section). Merge of two electron micrographs taken at ±8° is shown. (Scale bars: 200 nm.)

Fig. 6.

Microinjection of ENTH antibodies inhibits synaptic vesicle recycling. (A) Electron micrograph of a synapse stimulated at 5 Hz for 30 min from an axon microinjected with ENTH antibodies. Note the depletion of synaptic vesicles (compare with Fig. 3A), the presence of cisternae, and the absence of clathrin-coated intermediates. (B) Examples of coated structures with distorted morphology found in a different section of the synapse shown in A. a, axoplasmic matrix; d, dendrite; sv, synaptic vescicles. (Scale bars: 200 nm.) (C) Proposed role of epsin 1 in synaptic vesicle endocytosis. Epsin 1 (black) helps to generate membrane curvature, which is stabilized by interactions with the assembling coat (Upper). Interactions of epsin with clathrin (light gray) and AP2 (dark gray) also take part in restricting clathrin assembly to curved regions. After perturbation of epsins interactions with clathrin and AP2 (Lower) the stabilizing effect of epsin on curvature is disturbed. Moreover, clathrin assembly extends beyond the limit of the curved membrane.