EVALUATION OF VIRAL MEMBRANE-FUSION ASSAYS - COMPARISON OF THE OCTADECYLRHODAMINE DEQUENCHING ASSAY WITH THE PYRENE EXCIMER ASSAY

Membrane fusion, in particular the fusion of enveloped viruses, is often measured with an assay based on octadecylrhodamine (R18) fluorescence dequenching. We have studied the association of R18 with membranes and used the R18 assay to measure virus fusion in model systems and in cultured cells. The results were compared with those of an assay based on the decrease in excimer fluorescence of pyrene-labeled phospholipids. For liposomes made from premixed R18 and phosphatidylcholine (PC), R18 fluorescence quenching was proportional to the concentration of the probe up to about 4 mol %. No quenching was found at very low concentrations of R18. However, various artificial and biological membranes labeled by the addition of R18 from an ethanolic solution showed significant quenching at such low R18 concentrations. Thus, some of the R18 was not randomly distributed but likely was associated with the surface of the membranes in the form of highly quenched clusters or micelles. Moreover, in influenza virus membranes, R18 appeared highly quenched at very low concentrations, indicative of the probe interacting with viral proteins. In contrast, pyrene-labeled PC incorporated in either liposomes or reconstituted viral membranes (virosomes) showed an excimer/monomer fluorescence ratio proportional to the concentration of probe. When intracellular membrane fusion was investigated with R18-labeled influenza virus or Semliki Forest virus (SFV), fluorescence dequenching was observed in the absence of fusion, most likely due to spontaneous probe exchange. On the other hand, both SFV, metabolically labeled with pyrene-phospholipids, and influenza virosomes, containing pyrene-labeled PC, showed a decrease in pyrene excimer fluorescence only under conditions permitting fusion with the endosomal membrane. These data indicate that pyrene-labeled lipids, in contrast to R18, are not susceptible to spontaneous exchange. Fusion of influenza virosomes, containing both R18 and pyrene-PC, with erythrocyte ghosts showed that the highly quenched population of R18 can be transferred to the donor membrane; however, the transfer of R18 occurred more slowly than the transfer of pyrene-lipids. The infectivity of metabolically labeled SFV was unaffected. Thus, although the R18 assay gives satisfactory results in many in vitro systems, the pyrene assay is better suited for a quantitative assessment of viral membrane fusion.