Maintaining compositional lipid gradients across membranes in animal cells is essential to biological function, but what is the energetic cost to maintain these differences? It has long been recognized that studying the passive movement of lipids in membranes can provide insight into this toll. Confusingly the reported values of inter- and, particularly, intra-lipid transport rates of lipids in membranes show significant differences. To overcome this difficulty, biases introduced by experimental approaches have to be identified. The present study addresses the difference in the reported intramembrane transport rates of dimyristoylphosphatidylcholine (DMPC) on flat solid supports (fast flipping) and in curved free-standing membranes (slow flipping). Two possible scenarios are potentially at play: one is the difference in curvature of the membranes studied and the other the presence (or not) of the support. Using DMPC vesicles and DMPC supported membranes on silica nanoparticles of different radii, we found that an increase in curvature (from a diameter of 30 nm to a diameter of 100 nm) does not change the rates significantly, differing only by factors of order ∼1. Additionally, we found that the exchange rates of DMPC in supported membranes are similar to the ones in vesicles. And as previously reported, we found that the activation energies for exchange on free-standing and supported membranes are similar (84 and 78 kJ/mol, respectively). However, DMPC's flip-flop rates increase significantly when in a supported membrane, surpassing the exchange rates and no longer limiting the exchange process. Although the presence of holes or cracks in supported membranes explains the occurrence of fast lipid flip-flop in many studies, in defect-free supported membranes we find that fast flip-flop is driven by the surface's induced disorder of the bilayer's acyl chain packing as evidenced from their broad melting temperature behavior.

No abstract