Dynamin-independent Endocytosis Research Articles

SNAP25, but Not Syntaxin 1A, Recycles via an ARF6-regulated Pathway in Neuroendocrine Cells

Soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptor (SNARE) proteins mediate cellular membrane fusion events and provide a level of specificity to donor-acceptor membrane interactions. However, the trafficking pathways by which individual SNARE proteins are targeted to specific membrane compartments are not well understood. In neuroendocrine cells, synaptosome-associated protein of 25 kDa (SNAP25) is localized to the plasma membrane where it functions in regulated secretory vesicle exocytosis, but it is also found on intracellular membranes. We identified a dynamic recycling pathway for SNAP25 in PC12 cells through which plasma membrane SNAP25 recycles in approximately 3 h. Approximately 20% of the SNAP25 resides in a perinuclear recycling endosome-trans-Golgi network (TGN) compartment from which it recycles back to the plasma membrane. SNAP25 internalization occurs by constitutive, dynamin-independent endocytosis that is distinct from the dynamin-dependent endocytosis that retrieves secretory vesicle constituents after exocytosis. Endocytosis of SNAP25 is regulated by ADP-ribosylation factor (ARF)6 (through phosphatidylinositol bisphosphate synthesis) and is dependent upon F-actin. SNAP25 endosomes, which exclude the plasma membrane SNARE syntaxin 1A, merge with those derived from clathrin-dependent endocytosis containing endosomal syntaxin 13. Our results characterize a robust ARF6-dependent internalization mechanism that maintains an intracellular pool of SNAP25, which is compatible with possible intracellular roles for SNAP25 in neuroendocrine cells.

Endocytosis and autophagy in cerebral ischemia and excitotoxicity

Endocytosis and autophagy are known to be enhanced in some cases of neuronal death, and there is evidence that they can be directly involved in the death mechanism. We here report that both these trafficking events are enhanced by cerebral ischemia in neonatal (P12) rats, in which the main (cortical) branch of the middle cerebral artery was occluded by cauterization in association with clamping of the common carotid artery on the same side for 1.5 h. We also analyze the endocytic phenomenon pharmacologically in dissociated neuronal cultures exposed to NMDA. The enhanced endocytosis was shown in rats that received an intracerebroventricular injection of 4.4 kD FITC-dextran at the time of carotid unclamping, and were sacrificed 24 h later. There was strong neuronal endocytosis restricted to the ischemic zone, which was specifically labelled by immunocytochemistry for early endosomal antigen 1 (EEA1). In dissociated neuronal cultures a similar NMDA-induced endocytosis of FITC-dextran was observed; it increased with the dose of NMDA and the duration of NMDA treatment, and with a toxic dose the strongest endocytosis was observed microscopically to be in neurons stained with propidium iodide (a marker for dying or dead cells). However, to study specifically the endocytotic mechanism without the complication of cell death, we determined a combination of NMDA concentration and timing that enhanced endocytosis but with minimal cell death, as judged by LDH release, and measured the endocytosis in cellular extracts by fluorescence spectrometry. This revealed two components of dextran internalisation: they were both confirmed to be endocytic by the fact that they did not occur at 4°C, but they had completely different pharmacological profiles. One occurred constitutively (even in the absence of NMDA) and was blocked by inhibitors of classical fluid-phase dynamin-independent endocytosis (e.g. by wortmannin). The other component was induced by NMDA, insensitive to fluid-phase inhibitors, but sensitive to inhibitors of dynamin-mediated endocytosis such as 0.4 M sucrose and also to the JNK pathway inhibitor D-JNKI1, a powerful neuroprotectant. The latter inhibitors had little effect on the constitutive component of endocytosis. Although most dynamin-mediated endocytosis is receptor-mediated, competition experiments showed that the NMDA-induced endocytosis did not depend on a receptor specific for dextran. Enhanced lysosomal activity, presumably autophagy, was shown in neurones within the ischemic region (again at 24 h after carotid unclamping) by acid phosphatase histochemistry and by immunohistochemistry for lysosome-associated membrane protein (LAMP)-1. Cathepsin D was also increased in the ischemic region, but appeared to be in glia, not in neurons. The importance of these results is twofold.

Distinct dynamin-dependent and -independent mechanisms target structurally homologous dopamine receptors to different endocytic membranes.

D1 and D2 dopamine receptors are structurally homologous G protein-coupled receptors that serve distinct physiological functions both in neurons and nonneural cell types. We have observed that these receptors are selectively endocytosed in HEK293 cells by distinct dynamin-dependent and -independent mechanisms. Although these endocytic mechanisms operate with similarly rapid kinetics, they differ in their regulation by agonist and deliver D1 and D2 receptors specifically to different primary endocytic vesicles. After this segregation into different endocytic membranes, both D1 and D2 receptors recycle to the plasma membrane. Similar results are observed in Neuro2A neuroblastoma cells coexpressing both receptors at high levels. These findings establish that "classical" dynamin-dependent and "alternative" dynamin-independent endocytic mechanisms differ in their physiological regulation, sort structurally homologous signaling receptors in the plasma membrane, and mediate distinct early endocytic pathways leading to recycling endosomes. Our results also refute the previous hypothesis that dynamin-independent endocytosis targets G protein-coupled receptors selectively to lysosomes, and they suggest a new role of endocytic sorting mechanisms in physically segregating structurally homologous signaling receptors at the cell surface.