Inhibition of sendai virus fusion with phospholipid vesicles and human erythrocyte membranes by hydrophobic peptides

Abstract

Hydrophobic di- and tripeptides which are capable of inhibiting enveloped virus infection of cells are also capable of inhibiting at least three different types of membrane fusion events. Large unilamellar vesicles (LUV) of N-methyl dioleoylphosphatidylethanolamine ( N-methyl DOPE), containing encapsulated 1-aminonaphthalene-3,6,8-trisulfonic acid (ANTS) and/or p-xylene bis(pyridinium bromide) (DPX), were formed by extrusion. Vesicle fusion (contents mixing) and leakage were then monitored with the ANTS/DPX fluorescence assay. Sendai virus fusion with lipid vesicles and Sendai virus fusion with human erythrocyte membranes were measured by following the relief of fluorescence quenching of virus labeled with octadecylrhodamine B chloride (R 18), a lipid mixing assay for fusion. This study found that the effectiveness of the peptides carbobenzoxy- l-Phe- l-Phe (Z- l-Phe- l-Phe), Z- l-Phe, Z-D-Phe, and Z-GIy- l-Phe- l-Phe in inhibiting N-methyl DOPE LUV fusion or fusion of virus with N-methyl DOPE LUV also paralleled their reported ability to block viral infectivity. Furthermore, Z- d-Phe- l-PheGly and Z-GIy- l-Phe inhibited Sendai virus fusion with human erythrocyte membranes with the same relative potency with which they inhibited vesicle-vesicle and virus-vesicle fusion. The evidence suggests a mechanism by which these peptides exert their inhibition of plaque formation by enveloped viruses. This class of inhibitors apparently acts by inhibiting fusion of the viral envelope with the target cell membrane, thereby preventing viral infection. The physical pathway by which these peptides inhibit membrane fusion was investigated. 31 P nuclear magnetic resonance (NMR) of proposed intermediates in the pathway for membrane fusion in LUV revealed that the potent fusion inhibitor Z- d-Phe- l-PheGly selectively altered the structure (or dynamics) of the hypothesized fusion intermediates and that the poor inhibitor Z-GIy- l-Phe did not. One possible interpretation of these 31P NMR results was that the inhibitory peptide stabilized a membrane structure with a large radius of curvature, when the fusion pathway demanded a membrane defect with a small radius of curvature. This hypothesis was tested by determining the influence of an inhibitory and a noninhibitory peptide on the formation of membraneous structures with small radii of curvature, through ultrasonic irradiation of phospholipid dispersions. The inhibitory peptide prevented the formation of membrane structures with small radii of curvature, while the noninhibitory peptide did not prevent the formation of such structures. These data taken collectively suggest that the peptides that inhibit the fusion of enveloped viruses could act, at least in part, by interfering with the formation of intermediates in membrane fusion (possibly punctate membrane defects induced by the virus) that force lipids into structures with small radii of curvature.