An activation-collision mechanism for cholesterol transfer between membranes.

We report the results of experiments which show that cholesterol transfer between membranes cannot proceed by aqueous diffusion, as widely held, but must involve a more complex mechanism. (a) The rate of transfer of [3H]cholesterol from red blood cells was found to vary inversely with the size of the acceptor particle (ghosts, vesicles of ghosts, liposomes, and plasma lipoproteins). (b) The transfer of [3H]cholesterol from red blood cells to ghosts was accelerated by the presence of plasma, even though the plasma competed with the ghosts as an acceptor. (c) The rate of transfer of [3H]cholesterol from red blood cells to ghosts decreased to zero with increasing dilution but was not simply second-order. (d) The cholesterol in retinal rod disc membranes is not at equilibrium with plasma lipoproteins in that disc cholesterol increased when the homogenates were incubated in vitro with plasma. (e) The kinetics of cholesterol transfer cannot be limited by unstirred layer effects since the transfer of lysolecithin in the same system was faster than that of cholesterol by 3 orders of magnitude. The simplest model compatible with all the data suggests a two-step pathway involving a first-order followed by a second-order process. The first step could be a unimolecular activation event, perhaps the movement of the sterol in the donor particle to a more exposed (hydrated) position. In the second step, the activated sterol would be transferred during transient collisions between donor and acceptor particles. When collision is not rate-limiting, the overall process would appear to be simply first-order, hence kinetically indistinguishable from the aqueous diffusion mechanism. The activation-collision model thus not only rationalizes our data but is also consistent with the simpler kinetics previously reported for the transfer of both membrane phospholipids and sterols.