Rapid kinetics of Ca2+-induced fusion of phosphatidylserine/phosphatidylethanolamine vesicles. The effect of bilayer curvature on leakage.

We have employed both small unilamellar vesicles (SUV) and large unilamellar vesicles formed by the reverse phase evaporation technique (REV) to study the initial kinetics of membrane aggregation and fusion. Stopped flow measurements of the calcium-induced changes in the turbidity of SUV and REV, formed from 1:1 (mol/mol) mixtures of bovine phosphatidylserine (PS) and Escherichia coli phosphatidylethanolamine (PE), were used to follow particle aggregation. Simultaneous measurements of the fluorescence resonance energy transfer from N-(7-nitro2,1,3-benzoxadiazol-4-yl) (NBD)-PE to rhodamine (Rho)-PE incorporated into the vesicle bilayers established that 1) both initial aggregation and fusion can be described as a bimolecular process and 2) the rate-limiting step of membrane fusion is aggregation. Thus fusion takes place in the microsecond time domain. Parallel experiments, which simultaneously measured aggregation and the dequenching of encapsulated carboxyfluorescein (CF) in the presence and absence of antifluorescein antibodies in the suspension medium, established that the small unilamellar vesicles rapidly lose their contents of CF as they fuse. On the other hand, the first few cycles of fusion of the large unilamellar vesicles are nonleaky, but leakage develops within 1-2 s as the particles grow in size. Thus the results demonstrate that the SUV are poor models for the study of nonleaky fusion, while the REV must be carefully tested before unambiguous interpretation of fusion assays involving the formation of tight complexes (such as the terbium-dipicolinic assay) can be made. NBD-PE undergoes very rapid, Ca2+-promoted changes in quantum yield which can obscure the resonance energy transfer signals. Thus data from the NBD-PE/Rho-PE energy transfer pair must be carefully scrutinized for artifacts.