MECHANISMS OF PROTEIN TRANSPORT AT THE ER-GOLGI INTERFACE

The Endoplasmic Reticulum represents the first station of the secretory path-way, where proteins destined to the cell surface or to some intracellular organelles are recruited in specific ER subdomains, the ER Exit Sites (ERES), and start to travel into transport carriers to reach the proper final destination. Even though cargoes are usual-ly recruited to ERES by a sequence-dependent mechanism, it is known that other fac-tors contribute to protein export from the ER. Using model fluorescent tail-anchored proteins our group previously demonstrated that the length/hydrophobicity of the transmembrane domain is an important factor determining recruitment to or exclusion from ERES: a protein with a short TMD (FP-17) is excluded from ERES and retained in the ER, while a longer TMD (FP-22) determines enrichment in ERES. In order to clarify the molecular mechanism underlying this TMD-dependent transport, we first compared the transport of an export signal-bearing (VSV-G DxE) membrane protein with our model protein FP22, which lacks an export signal. FP22 and VSV-G accu-mulate together at ERES, but VSVG reaches the plasma membrane more rapidly than FP22. To investigate the basis of this difference, we combined cDNA microinjection to temperature blocks and live-cell imaging approaches that allowed us to analyze the transport at early steps of the secretory pathway at the ER-Golgi interface. At 20°C, a temperature at which only the transport between the ER and the Golgi is allowed, all of the VSVG accumulates in the Golgi, while FP-22 remains distributed between the ER and the Golgi. After bleaching the Golgi fraction of FP22 we observed a rapid, energy-dependent, fluorescence recovery, indicating an efficient ER to Golgi transport even in the absence of the export signal and suggesting that FP22 may be re-cycled between the two compartments. In agreement, a rapid emptying of the Golgi was observed after ER bleaching (accompanied by a fluorescence recovery of the ER fraction). To investigate whether this phenomenon is restricted to our model protein only or it is more general event, we then tested the behavior of a signal-deleted form of VSV-G (VSV-G AxA). Similarly to FP22, VSVG AxA is distributed between the Golgi and the ER at 20°C and Golgi fluorescence rapidly decreases after ER bleach-ing, suggesting a new role of the ER export signal, which is important not only in re-cruiting cargoes at the ERES, but also in preventing their recruitment into futile cy-cles between the Golgi and the ER, which delay their arrival to the cell surface. To further characterize the mechanism of TMD-dependent sorting, we then in-vestigated the role of membrane curvature; our group previously demonstrated that FP-22 is segregated from FP-17 in specific ER subdomains, which are characterized by membrane curvature (ERES and ER tubules). In collaboration with Bruno Goud and Jean-Baptiste Manneville (Institute Curie, Paris), we created highly curved do-mains using membranes composed of a uniform lipid composition (POPC, palmitoyl-oleyl-phosphatidylcholine) or ER lipids extracted from rat liver microsomes, and we analyzed the distribution of our two model proteins in flat and curved domains. Our results indicate that the two proteins are uniformly distributed in curved membranes and strongly suggest that the membrane curvature alone cannot drive the TMD-dependent partitioning of membrane proteins in ERES and ER tubules. Taken together, our data contribute to clarify the role of two fundamental fac-tors influencing the transport of membrane proteins along the secretory pathway that were never investigated before.