Direct delivery had been plausible, as sorting of GPI-anchored proteins into rafts, and thus away from basolateral proteins, occurs in the trans-Golgi network (TGN) before either departs for plasma membranes. Few or no apical proteins were ever seen in the basolateral membrane. And finally, transport across the cell seemed unlikely: the main method of departure, clathrin-mediated endocytosis, did not transport rafts.
Lippincott-Schwartz recently found that rafts are continually endocytosed by a nonclathrin pathway, overcoming the clathrin objection. And at the basolateral surface, she says, “all of these biochemical experiments would have potentially missed a transient appearance.”
The NIH team found that GPI-YFP and basolateral proteins were segregated as they left the TGN, but nevertheless shared the same tube-shaped carriers. (Others have claimed one cargo type per carrier, but did not prebleach to reduce background.) Those carriers must be paying a visit to the basolateral membrane, as both proteins got stuck intracellularly when fusion to the basolateral membrane was prevented with tannic acid, a cell-impermeable fixative. When tannic acid was applied to the apical side, only the apical cargo was intracellular, presumably after making its transient basolateral stop. Evidence for this trip across the cell came from a tracer and antibody, which were applied to the basolateral side but travelled with GPI-YFP apically.
A nonraft apical protein went straight to the apical side of the cell. Why should the cell bother with the more circuitous route taken by GPI-YFP? Lippincott-Schwartz suggests that raft-associated proteins may help carry other cargos across the cell so they can do their job in the gut or elsewhere. Bacteria, some of which are known to attach to rafts, may even grab on and get a free ride. ▪