UCSF

Autophagy is traditionally described as an autodigestive pathway in which cells perform the orderly degradation and recycling of dysfunctional or unneeded cytosolic components within autophagosomes. However, autophagy also has its less intuitive but biologically distinct roles in secretion. These secretory functions, collectively termed secretory autophagy, range from the secretion of inflammatory cytokines to the biogenesis of extracellular vesicles (EVs). The study of EVs has generated significant interest of late given their ability to evoke intercellular communication and non-cell autonomous activity in both normal homeostatic cell activity and pathological settings. However, these exciting phenotypic applications have somewhat overtaken the spotlight of EV research leading investigations of the specific molecular components involved in EV loading and secretion to fall by the wayside. Further research is needed detailing the underlying mechanistic components of EV secretion if we are to fully understand and take advantage of the exciting capabilities of EVs in health and disease.Recently, our lab described a secretory autophagy pathway termed LC3-dependent EV loading and secretion (LDELS) that facilitates the packaging of cytosolic cargoes, particularly RNA-binding proteins (RBPs), into extracellular vesicles for secretion outside the cell. The LDELS pathway utilizes a pool of LC3, a member of the ATG8 protein family that functions in autophagy substrate selection, localized at single membrane endosomes rather than double membrane autophagosomes. While our previous study focused on the requirement for this pathway in the secretion of soluble RNA-binding proteins, we also observed a collection of proteins containing membrane-spanning regions that relied on this pathway for secretion. In particular, we identified transferrin receptor as a novel substrate of the LDELS pathway. Transferrin receptor is a membrane glycoprotein responsible for cellular iron uptake and has been implicated in diseases such as hematological disorders, neurodegenerative diseases, and cancer. Additionally, it was the first protein identified to be secreted in EVs. Similar to other LDELS targets, TFRC secretion via EVs genetically requires components of the LC3-conjugation machinery but is independent of other ATGs involved in classical autophagosome formation. Interestingly, the secretion of EV-associated TFRC is actually enhanced upon inhibition of classical degradative capabilities. Furthermore, TFRC binds ATG8 orthologs via a cytoplasmic domain LC3-interacting region (LIR) motif, which may facilitate the loading of TFRC into EVs. However, this LIR motif is not solely responsible for the functional secretion of TFRC. This may be explained by the observation that TFRC exists both in the conventionally annotated orientation and in a reversed orientation in EVs, thus LC3 residing within the lumen of EVs may interact with LIRs localized to either the cytoplasmic or extracellular domains of TFRC. Additionally, the packaging and secretion of TFRC into EVs depends on multiple ESCRT pathway components, suggesting the ESCRT pathway is involved in the budding of TFRC into intraluminal vesicles, the precursors of EVs. Lastly, the secretion of TFRC requires the RAB27A, a small GTPase implicated in the docking and release of EVs into the extracellular space. Based on these results, we propose that the LDELS pathway promotes the TFRC incorporation into EVs and its secretion outside the cell.