Specific Movement of Cell Membranes Fused with HVJ (Sendai Virus)

Transcription

1 Proc. Nat. Acad. Sci. USA Vol. 71, No. 5, pp , May 1974 Specific Movement of Cell Membranes Fused with HVJ (Sendai Virus) (cap formation/projection of cytoplasm/microvilli/vacuoles) YOSHIO OKADA, JEMAN KIM, YUMIKO MAEDA, AND IGNEZ KOSEKI* Department of Animal Virology, Research Institute for Microbial Diseases, Osaka University, Suita, Osaka 565, Japan Communicated by James D. Ebert, February 19, 1974 ABSTRACT When Ehrlich ascites tumor cells fused with UV-inactivated HVJ (Sendai virus) were cultured, the viral envelope antigens, which had been integrated at random into the cell membranes, accumulated at one site on the fused cell. The site showed a specific structure, namely, a projection of the cytoplasm from the spherical fused cell surface with clusters of many long microvillilike protrusions from it. When the fused cells were conjugated with antibody against Ehrlich ascites tumor cells and cultured, the antigen-antibody complex also accumulated on the projection. Associated with the development of the projection and accumulation of viral antigens, vacuoles developed at the base of the projection in the cytoplasm of the fused cells in culture. The development of the projection and vacuoles and the accumulation of viral antigens were all inhibited by glucose, mannose, galactose, sodium azide, and cycloheximide, and these inhibitions were reversible. All these structures were observed in multinucleated cells, but few in mononucleated cells after reaction with HVJ (Sendai virus). We consider that these three phenomena correlate with each other and mqay result from changes irl cellmembrane systems caused by fusion of cells. The cell-fusion phenomenon with hemagglutinating virus of Japan (Sendai virus) (HVJ) (1-5) seems to offer a suitable experimental system for studies on the biological nature of animal-cell membranes (6). Ehrlich ascites tumor cells rapidly agglutinate on addition of HVJ at 00; this agglutination occurs with adsorption of HVJ virions onto the surface of the cells. When cell aggregates are incubated at 370, the viral envelopes react with the cell membranes, the membranes of adjacent cells in cell aggregates fuse together within a few minutes, and spherical fused cells appear within 30 min. The efficiency of cell fusion is a function of the concentration of HVJ added (7). In the course of these reactions, some abnormal conditions for the cells are produced. At an early stage, the cell-membrane structure becomes unstable due to reaction with the virus; decrease in the membrane potential and membrane resistance, with leak of ions through the cell membranes, can be detected electrophysiologically. The instability is reversed with progress of cell fusion (8). At about the same stage of the reaction, fusion of viral envelopes with cell membranes occurs (9); glycoproteins and lipids originating from the envelope are explanted into the cell membranes. Within 30 min of the cellfusion process, the original cell-surface area decreases, the degree of decrease depending on the degree of fusion of the cells; Abbreviations: HVJ, Hemagglutinating virus of Japan, synonym: Sendai virus. * Present address: Seoao de Biologia Celular, Divisdo de Biologia Animal, Instituto Biol6gico de Sao Paulo, SAo Paulo, Brazil the excess membranes are probably treated in some unknown ways in the fused cells, utilizing energy (3). Polykaryocytosis also appears. In the present paper, a specific movement of fused cell membranes is deduced from observations on the movement of viral antigens integrated into the membranes of fused cells. MATERIALS AND METHODS As Ehrlich ascites tumor cells, cells passaged in the abdomen of ddo mice or cells of an established line passaged in vitro were used. Results with the two were similar in the present experiments. The cells were washed and suspended in 9 volumes of balanced salt solution (140 mm NaCl; 54 mm KCl; 0.34 mm Na2HPO4; 0.44 mm KH2PO4; buffered with 10 mm Tris * HCl at ph 7.6) containing 2 mm CaCI2. HVJ, Z strain, was passaged in embryonated eggs, purified by differential centrifugation (2), and then, suspended in balanced salt solution at a concentration of 2000 hemagglutinating units. The virus was inactivated by UV-irradiation before use. A mixture of equal volumes (0.5 ml each) of tumor cells and HVJ was incubated at 00 for 5 min and then the mixture was incubated in a water bath at 370 for 30 min with shaking. After completion of the cell-fusion reaction, cells were washed to remove free virus at 00, and then cultured in minimal essential medium supplemented by 10% calf serum at 37. When inhibitors were used, they were dissolved in the medium. After culture, cells were washed again and exposed to antiserum against HVJ envelope developed in rabbit at 00 for 20 min. Then the cells were washed at 0, and mixed with fluorescein isothiocyanate-labeled anti-rabbit IgG 7S antibody (Hyland Biochemicals). After incubation for 20 min at 00, the cells were washed and observed under a microscope with incident-light fluorescence (Leitz). In the case of FL cells, an established line from human amnion, cells in monolayer culture were dissociated with trypsin and EDTA; after washing, they were resuspended at a concentration of 10% in balanced salt solution-ca medium. Then, an equal volume of 500 hemagglutinating units of UV-inactivated HVJ was added for cell fusion. For electron microscopical examinations, cells were fixed with 2.5% glutaraldehyde in phosphate-buffered saline (ph 7.2) at 40 for 40 min and postfixed in 1% osmium tetraoxide in phosphate-buffered saline at 40 for 1 hr. They were washed and stained with 2% uranyl acetate in ethanol at room temperature for 1 hr, then dehydrated and embedded in Epon. RESULTS AND DISCUSSION Biological activities originating from viral envelopes were associated with Ehrlich ascites tumor cells that had been fused with HVJ. Added erythrocytes adsorbed all over the surface

2 OA,2044 Cell Biology: Okada et al. Proc. Nat. Acad. Sci. USA 71 (1974) (FIG. 1. Legend follows at top of page 2045.)

3 Proc. Nat. Acad. Sci. USA 71 (1974) Movement of Fused Cell Membranes 2045 FIG. 1. Micrographs of Ehrlich ascites tumor cells fused with HVJ, showing'cap formation. (A) Fluorescent images of cells immediately after fusion, seen by the indirect fluorescent technique with antiserum against HVJ envelope. (A-1) Ring-shaped fluorescent images (X200). (A-2) The fluorescent image of a fused cell, when focused on the cell surface (X950). (B) Fused cells after culture for 4 hr, observed by the indirect fluorescent technique with antiserum against HVJ envelope. (B-i) Cap-like fluorescent images of fused cells (X200). (B-2) Phase-contrast micrograph (X600) and (B-3) fluorescent micrograph (X600) of the same fused cell, showing the coincidence in position of the projection and the cap. (C) The fluorescent image of the surface of fused cells after 4 hr in culture, observed by the indirect fluorescent technique with antiserum against HVJ envelope (X950). (C-i) The projection in the upper left portion of the cell surface projects upwards from the surface. (C-2) The cap is in the center of the cell surface. (D) Fluorescent images of fused cell surfaces observed by the indirect fluorescent technique with antiserum against Ehrlich ascites tumor cells (X950). (D-i) The fused cell was cultured for 2 hr, then stained with antibody and examined. Fine fluorescent grains are distributed evenly over all the surface, including the projection (upper left). (D-2) The fused cell, which had been conjugated with rabbit antiserum against tumor cells immediately after fusion, and then cultured for 2 hr and stained with FITC-labeled anti-rabbit IgG antibody. A typical cap structure is seen. of fused cells (hemadsorption), and this adsorption was completely inhibited by treatment of the cells with antiserum against HVJ. Viral neuraminidase and cell-fusion activities were also detectable on the cell surfaces. By use of fluorescent antibody against HVJ envelope, a ring-shaped fluorescence could be seen on the cell surface. The structure of viral envelopes fused with cell membranes was easily demonstrated with ferritin-labeled antibody against HVJ envelope in an electron microscope, The association of viral biological activities with the cell membranes was mainly due to fusion of viral envelopes with the cell membranes (10). When the fused cells were cultured in minimal essential medium supplemented with 10% calf serum at 370, biological activities, such as cell fusion or hemadsorption, disappeared in 2-4 hr without liberation of viral envelopes from the cell surfaces into the medium (10). The structure of the viral envelopes fused with cell membranes was in a dynamic state. With decrease in the biological activities, two characteristic structures developed in the fused cells after culture for 2-4 hr. One was a projection from the spherical surface of the fused cells (Fig. 1, B-2; Fig. 2, A-i and A-2). The other was vacuoles in the cytoplasm (Fig. 2, A and B). Fluorescent microscopy showed that the viral antigens were evenly distributed over the cells immediately after fusion. As shown in Fig. 1, A-i and A-2, very fine fluorescent grains were seen randomly distributed on the cell surface just after fusion, when a microscope with incident-light fluorescence was focused on the cell surfaces at the highest magnification. By focusing on the middle of the cells, a ring-shaped smooth fluorescent image was seen. When fused cells were examined after culture for 2-4 hr, the fine fluorescent grains were seen to have become larger and their distribution was no longer even. The fluorescence accumulated at one site on fused cells while other areas had sparse fluorescence (Fig. 1, B-i). Hence, the strong fluorescent area was seen to be surrounded by weak, patchy fluorescence. This strongly fluorescent area is called the "cap" in this paper. In Fig. 1C, a cap on a fused cell is observed from the top. Strong fluorescence is seen accumulated in the cap area, less at the base of the cap, and very little beyond this. These images give the impression that viral-envelope antigens homogeneously integrated into the initial cell membranes move to the cap area on further culture of the cells. It is interesting that the cap area coincides in position with the projection described above, as shown in Fig. 1, B-2 and B-3. Electron microscopically, many clusters of long microvilli-like protrusions project from the surface of the projection (Fig. 2C). With ferritin-conjugated antibody against HVJ envelope, large accumulations of viral antigens were detected on the projection (10). The cap became detectable on fused cells after culture for 1 hr as an intermediate image of the typical cap and became maximal after 4 hr. After 20 hr, the fluorescence of the cells, including that of the projection, became very weak. Typical caps were also formed on FL cells when a suspension of the cells was fused with HVJ and cultured in standard medium for 2 hr. Caps developed in cultures with minimal essential medium without serum, but not in balanced salt solution medium with 1 mm CaC12, which is a medium containing only nonionic salts used for the cell-fusion reaction. Cap formation was inhibited by 0.5 M glucose, mannose, or galactose, 1 mg/ml of cycloheximide, or 50 mm sodium azide, and the inhibitions were reversible. In the presence of one of these inhibitors, no projection of the cells developed and the fluorescence was observed as a ring-shaped image. But when the medium was replaced by fresh standard medium, the projection developed after 2 hr and showed great accumulation of viral antigens (cap). Cap formation was also ph dependent, appearing at ph but not at ph 6.0. Cap formation may depend on the fluidity of the cell membranes since it was inhibited at 00 or by monosaccharides (11) and is ph dependent, and it may be correlated with an energy-dependent, synthetic reaction(s) since it was inhibited by sodium azide or cyclohexirnide or in balanced salt solution-ca medium. Actinomycin D had no effect on cap formation. In medium containing cytochalasin B (10,g/ml), fused cells developed multiple caps. Cap formation was observed on fused cells, and one cap was formed per fused cell, irrespective of the size of the fused cell. But cap formation was not frequent on mononucleated cells, which had been treated with HVJ and in which viral antigens were integrated into the cell membranes. As shown in a previous paper (2), when cells were treated with HVJ, cell fusion occurred randomly. After the cell-fusion reaction, the cell population consisted of nonfused cells, and bi-, tri-, tetra-,.. nucleated cells formed by random fusion. When fused cell samples were examined, 153 of the 162 cells examined (94%) showed ring-shaped fluorescent images after 0 hr of culture. After 4 hr, 176 of the 196 cells examined (90%) had formed caps. Only 20 cells showed ring-shaped images, of which 15

4 Cell Biolo 2046 I ;I-.f.,.:,, 4 4k. 0 " Ilk-1All-:in"...-'.,,,,,,e 2%. Proc. Nat. Acad. Sci. USA 71 (1974) Okada et al. : -,.'-'.r C'. '4. '.'. 4'. p. Ar.A ( 4 7. I;I it,.; 5ss, "*ha;il Cl-1 (ws -tto-,p.,.o p.)j; (FIG. 2. Legend follows at top of page 2047.)

5 Proc. Nat. Acad. Sci. USA 71 (1974) Movement of Fused Cell Membranes 2047 FIG. 2. Structure of the projection and localization of vacuoles developing on fused cells in culture. (A) Photographs obtained with an interference microscope of fused cells, cultured for 4 hr (X 1000). (A-i) The upper portion of the cell is the projection (p), and many small vacuoles are observed at its base (v). n: nucleus. (A-2) The projection (p) is in the middle of the huge fused cell. The projection projects upwards and many fine protrusions are observed on it. Many vacuoles (v) are seen surrounding its base. (B) Photographs similar to those of A, of a fused cell, cultured for 20 hr (X 1000). (B-i) A large vacuole separates the projection from the area of the nuclei. (B-2) Image obtained by focusing on a little lower portion of the same cell, showing many small granules surrounding the vacuole. (C) Electron micrographs of a projection developing on a fused cell after 4 hr of culture. (C-i) A cluster of microvilli-like structures and many small vacuoles at the base (X9350). (C-2) The cluster of microvilli-like structures at high magnification (X51,000). cells were mononucleated, two were binucleated, and three contained three to five nuclei. Vacuole formation in the cytoplasm was one of the characteristic changes of fused tumor cells on culture for 4 hr. As shown in Fig. 2A, many vacuoles developed in the fused cells. The vacuoles seemed to develop from the base of the projection and became larger and fused together during further culture. In Fig. 2B, a fused cell after culture for 20 hr is also shown in which a large vacuole separates the area containing nuclei from the projection. Small granules, of the size of mitochondria, are seen surrounding the vacuole. The degree of the vacuole formation varied with the size of the fused cells; vacuoles were observed in all cells containing over 10 nuclei within 4 hr but rarely in small fused cells and mononucleated cells. Vacuole formation was inhibited under the conditions inhibiting cap formation described above, but when these inhibitory conditions were replaced by optimum conditions, cap formation started and vacuoles developed. Vacuole formation may result from abnormal conditions in the cells, but their development also seems to be correlated with an energydependent synthetic reaction, and inhibition of the fluidity of cell membranes may inhibit their development. Electron microscopically, the surface structure of the vacuoles, facing to the inside of them, seems to be similar to that of cell-surface membranes. The present findings may be understood by supposing that accumulation of viral antigens on the projection is a biological function of the cell membranes to segregate foreign materials integrated into the membranes in a specific site. However, no projections were observed frequently on unfused cells. As reported previously (12, 13), when cells that had been treated with UV-inactivated HVJ were cultured, mononucleated cells grew as well as untreated cells, but the efficiency of progeny production decreased with increase in the number of nuclei in fused cells. Thus, insertion of viral antigens into the cell itself does not seem to be lethal and may have little effect on functions of the cell membrane. Moreover, it was demonstrable that mononucleated FL cells treated with HVJ divided into daughter cells and that these retained viral antigens on their surfaces. Paramyxoviruses, including HVJ, are well known to cause persistent infection in some kinds of cells. These cells grow well but retain viral antigens on their surfaces. The biological meaning of cap formation and the development of projections are unknown. However, from the present findings it seems that both these phenomena are observed as a result of movement of cell membranes resulting from cell-to-cell fusion. Tumor cells were coated with rabbit antiserum against Ehrlich ascites tumor cells immediately after fusion and cultured for 0 and 2 hr in minimal essential medium containing 10% calf serum. They were then chilled and stained with FITC-labeled antibody against rabbit IgG. The fused cells showed fine fluorescent grains distributed evenly over their surfaces initially, but after culture for 2 hr, the fluorescent grains had mostly accumulated on the projections of the cells and formed a cap as shown in Fig. 1, D-2. The projection was the common site for accumulation of viral antigens and cellsurface antigen-antibody complex. When fused cells were stained with antiserum against the tumor cells and with FITC-labeled antibody against rabbit IgG after culture for 2 hr, the fluorescent grains were seen distributed evenly over all the cell surface, including the projection (Fig. 1, D-1). The present results suggest that there is a correlation between the formation of the projection, the cap, and the vacuoles. We consider that the regulation of cell-membrane systems, including cell-surface membranes and membrane systems in the cytoplasm, becomes disorganized when cells fuse, and that fused cells show a tendency to reorganize the membrane systems. The three phenomena presented here may be correlated with this tendency. Previously, one of the authors (14) observed some specific structures in fused Ehrlich ascites tumor cells. Perhaps there is a correlation between these structures and the structures reported here. This work was supported by Cancer Research Grant and from the Ministry of Education of Japan. 1. Okada, Y., Suzuki, T. & Hosaka, Y. (1957) Med. J. Osaka Univ. 7, Okada, Y. & Tadokoro, J. (1962) J. Exp. Cell Res. 26, Okada, Y., Murayama, F. & Yamada, K. (1966) Virology 27, Okada, Y. & Murayama, F. (1966) Exp. Cell Res. 44, Okada, Y. (1969) Curr. Top. Immunol. Microbiol. 48, Frye, L. D. & Edidin, M. (1970) J. Cell Sci. 7, Okada, Y. & Murayama, F. (1968) Exp. Cell Res. 52, Okada, Y. (1972) in Membrane Research (Academic Press, New York), pp Morgan, C. & Howe, C. (1968) J. Virol. 2, Koseki, I., Maeda, Y., Kim, J. & Okada, Y., unpublished data. 11. Loor, F., Forni, L. & Pernis, B. (1972) Eur. J. Immunol. 2, Yamanaka, T. & Okada, Y. (1966) Biken. J. 9, Yamanaka, T. & Okada, Y. (1968) Exp. Cell Res. 49, Okada, Y. (1962) Exp. Cell Res. 26,