Figure 12. Working hypotheses for Golgi protein recycling by COPI-dependent and COPI-independent pathways.
(A) The existence of multiple intra-Golgi recycling pathways can explain why inactivating COPI generates hybrid structures that repeatedly gain and lose late Golgi proteins. Green represents early Golgi proteins that recycle in COPI vesicles, and red represents late Golgi proteins that recycle by COPI-independent pathways. Under normal conditions, early Golgi proteins depart while late Golgi proteins arrive, and then late Golgi proteins depart in turn. When COPI is inactivated, the recycling of early Golgi proteins is inhibited, so when late Golgi proteins arrive, a hybrid structure is generated. This hybrid structure can subsequently lose late Golgi proteins by COPI-independent pathways, and then the process begins again. (B) During maturation of the late Golgi, proteins are likely to recycle by several pathways. The thick arrow represents cisternal maturation that occurs during and after conversion to a late Golgi compartment (Daboussi et al., 2012; Day et al., 2013). Peripheral membrane proteins such as Sec7 are recruited to late Golgi cisternae by activated GTPases, and are subsequently released from the membrane by GTP hydrolysis. Some transmembrane proteins such as Vps10 travel to the prevacuolar endosome in clathrin-coated vesicles with the aid of the Gga1 and Gga2 adaptors, and then recycle to newly formed late Golgi cisternae with the aid of cargo scaffolds such as retromer. Other transmembrane proteins such as Kex2 are postulated to recycle from older to younger late Golgi cisternae in clathrin-coated vesicles with the aid of the AP-1 adaptor.