Dissertation (The Rockefeller University, New York)
- K. V. Holmes
The effect of pH on the membrane-fusion activity of Sendai virus was examined (pH 5.0-9.5) by using, as assays of activity, hemolysis of chicken erythrocytes and the fusion of baby hamster kidney (BHK-21) cells. Exposure of virus to basic pH increased fusion activity; the optimum pH was found to be 9.0. All assays were carried out at pH 7.0, and the virus retained enhanced fusion activity after it was exposed to basic pH and returned to neutral pH. The enhanced fusion activity was correlated with an irreversible conformational change in the fusion protein (F protein) of the virus, as demonstrated by a change in the circular dichroism spectrum of the protein. The fusion protein (F protein) ofparamyxoviruses, which forms spike-like projections on the surface of the virus, is involved in the fusion of the viral membrane with target cell membranes (1-3). This fusion event is responsible for several biological activities of the virus-i.e., virus penetration, cell fusion, and hemolysis. The active F protein consists oftwo disulfide-linked polypeptides, F1 and F2, which are derived from an inactive precursor, Fo, by proteolytic cleavage by a host cell enzyme (4). Cells that lack a protease capable of cleaving FO produce fusion-inactive, noninfective virions (1-3). We have found (5) that isolated F protein reconstituted into membranes with phosphatidylcholine showed fusion activity, if a mechanism for attachment ofthe F protein to cell membranes was provided. This could be provided not only by the receptor-binding protein (virus hemagglutinin and neuraminidase protein; HN protein) of the virus but also by a lectin such as wheat germ agglutinin. Much effort has been directed to the understanding of the mechanism of virus-induced membrane fusion. Phospholipase activity in the virus was excluded in early studies (6). The possibility that fusion involved proteolytic activity ofthe F protein (7) on the target membrane was made unlikely by the finding of Haywood (8) that virions could fuse with liposomes lacking proteins. The relevance of fusion of virus with such vesicles in vitro to fusion of virions with cell membranes was recently established by our finding that virus containing cleaved F protein, but not those with uncleaved FO, could fuse with liposomes containing phospholipids and cholesterol (unpublished experiments). A direct hydrophobic interaction between the F protein and the target membranes was proposed on the basis of (i) the finding of a long hydrophobic amino acid sequence at the NH2 terminus generated by the cleavage and (ii) the fact that this sequence is highly conserved among paramyxoviruses (9-11). In addition, oligopeptides with sequences resembling this NH2terminal sequence were found to specifically inhibit the fusion action of the F protein (11). A fusion mechanism involving a hydrophobic interaction was further supported by our finding that the activating cleavage of the F protein involves a conformational change with exposure of a new hydrophobic region on the protein (12). The membrane-fusion activity of paramyxoviruses occurs over a wide pH range; however, an optimum pH of 8.4-8.5 was found for cell fusion and hemolysis by the paramyxovirus SV5 (13, 14). We have investigated the effect of pH on the membrane-fusing activity of Sendai virus and have confirmed that the fusing activity is enhanced at basic pH; we also found that this elevated activity remains when the virus is returned to neutral pH. An irreversible change in the conformation of the F protein was found by measuring the circular dichroism (CD) of the protein, which correlated with the increased membrane-fusing activity of the virus. MATERIALS AND METHODS Virus. The RU and Z strains of Sendai virus were grown in 10-day-old embryonated chicken eggs, harvested 48 hr after infection, and purified by repeated pelleting as described (5). Virus was suspended in 10 mM phosphate/150 mM NaCi, pH 7.2, and stored at -700C. Viral protein concentration was determined by a modified Lowry procedure (15). Hemolysis Assays. Virus pelleted from stock virus suspensions (2 mg of viral protein per sample) was resuspended in 1 ml of each of the following buffers: 25 mM Na acetate, pH 5.0, 5.5, and 6.0; 25mM Na phosphate, pH 6.5, 7.0,-and 7.5; 25mM Tris-HCl, pH 8.0, 8.5, and 9.0; and 25 mM NaHCO3, pH 9.5. Virus suspensions were kept on ice for 1 hr and then loaded onto discontinuous sucrose gradients consisting of 2 ml of 50% sucrose and 8 ml of20% sucrose in 10mM Na phosphate/150 mM NaCl, pH 7.0 (Pi/NaCI), and centrifuged at 28,000 rpm for 3 hr in a Spinco SW 41 rotor. The virus was collected at the 50%/ 20% sucrose interface and dialyzed against Pi/NaCl overnight; then the hemagglutination titer and viral protein concentration were determined. Chicken erythrocytes, 2-4 days old, were washed twice and resuspended in Pi/NaCl to a final concentration of 0.5%. Virus samples exposed to each pH (1.2 A.g of viral protein in 0.5 ml ofPi/NaCl containing 16 hemagglutinating units) were mixed with 0.5 ml of erythrocytes in Pi/NaCl, pH 7.0, kept on ice for 20 min, and shifted to a 370C water bath. Multiple samples of virus exposed to each pH were prepared, and one set ofsamples was removed from the water bath at each time point, chilled on ice for 5 min, and spun in an Eppendorf centrifuge at 40C for 7 min. The clear supernatants were removed, and their absorbance at 545 nm was read in a Zeiss PM6 Abbreviations: NP protein, virus nucleocapsid protein; F protein, virus fusion protein; HN protein, virus hemagglutinin and neuraminidase protein; BHK-21, baby hamster kidney cells; Pi/NaCl, 10 mM sodium phosphate/150 mM NaCl, pH 7.0; CD, circular dichroism.