Wellcome Research Laboratories, Beckenham, Kent, England. Royal Postgraduate Medical School, London, England. (Accepted 27 January I972)

Transcription

1 J. gen. ViroL (I972), I5, Printed in Great Britain Interaction of Sendai (HVJ) Virus with Human Erythrocytes: a Morphological Study of Haemolysis Cell Fusion By K. APOSTOLOV Wellcome Research Laboratories, Beckenham, Kent, England AND JUNE D. ALMEIDA Royal Postgraduate Medical School, London, England (Accepted 27 January I972) SUMMARY The morphology of Sendai virus envelope is described in detail. The interaction of the virus envelope and erythrocytes is examined by negative staining and ultrathin sectioning. In this morphological study haemolysis is seen to progress in the following sequence: attachment of virus to cell membrane; dissolution of cell membrane at the site of attachment; dissolution of spikes opposite the dissolved cell membrane; fusion of the innermost virus (nanogranular) layer with the cell membrane proper; extension of this fused area with annular formation and extrusion of the virus internal components into the lumen of the cell; finally integration of the virus envelope into the cell envelope and detachment of the spikes. This stage could then lead to haemolysis and fusion. Haemolysis could occur through breaks in the unstable virus part of the integrated membrane and cell fusion could occur when this area is stabilized. Cell fusion of two or more cells is achieved by bridging with interspersed virus envelopes. INTRODUCTION Viruses of the paramyxovirus group can cause lysis and fusion of susceptible cells (Okada, 1958). Cell fusion and formation of heterokaryons produced by inactivated Sendai virus is becoming an important tool in biological research (Poste, I97o), but the nature of the processes involved in lysis and fusion of cells by paramyxoviruses is not clear. It is widely suspected that these two phenomena are aspects of one process. The morphological aspect of haemolysis and fusion has been studied by the technique of ultrathin sectioning by several groups of authors. In a study of haemolysis of human O group cells by Sendai virus, Howe & Morgan (1969) observed attachment of the virus on to the cell, followed by a fusion of the virus envelope with the cell membrane, with extrusion of virus internal components inside the erythrocyte. The morphology of cell fusion was studied by Schneeberger & Harris (1966) for fusion of interspecific cells, and by Hosaka & Koshi (1968) for fusion of Ehrlich ascites cells. In the discussion of their results both groups of authors do not postulate active involvement and disintegration of the virus envelope. The present paper will describe the results of an electron microscope study of lysis and fusion of human red cells by Sendai virus, using the techniques of ultrathin sectioning and negative staining. A new hypothesis explaining fusion and haemolysis as a single process will be put forward. 16 VlR 15

2 228 K. APOSTOLOV AND JUNE D. ALMEIDA METHODS Virus. Sendal virus strain M34996 obtained from the Institute for Medical Research, Mill Hill, courtesy of Dr Pereira, was used throughout. The virus was inoculated into the allantoic cavity of lo-day-old chick embryos and incubated for 5 days at 35. Harvested allantoic fluid was clarified by centrifugation at 3ooo rev./min, for IO min. Virus was concentrated and partially purified by differential centrifugation at 3o,ooo rev./min, for 30 rain. and stored at --20 until needed. Red cells. Freshly drawn human group O blood was used throughout. A 20 % red cell suspension was made from 3 times washed blood in IO vol. of veronal buffered saline (VBS). Attachment of virus to red cells. The virus was adjusted to a haemagglutinating titre of about 32,0o0 in 2o ml. VBS at 4 and o'5 ml. of 2o% cell suspension was pipetted in and vigorously shaken. The preparation was kept at 4 for 4 to I8 hr with occasional shaking. Unattached virus was eliminated by washing three times in 2o ml. of chilled VBS, centrifuging at 25ooo rev./min, for IO min. between each wash. The final pellet was used for electron microscopy and also for the study of haemolysis and fusion. Haemolysis and fusion. A water bath set at 37 was used for incubation. For the study of haemolysis, cells with attached virus which had been kept at 4 were incubated for 3 to 15 rain. at 37 and then chilled immediately in iced water. The preparations were washed three times in chilled VBS as before and the pellet was used for electron microscopy. The same steps were followed for fusion, with one important difference. The cells with attached virus maintained at 4 were pelleted at I5OO rev.]min, for 5 min. and after this the undisturbed pellet in VBS was incubated at 37 for 5 to I5 min. This pellet was then broken and washed free of haemoglobin. Electron microscopy. A Philips E.M. 2oo was used in this study, usually at an instrumental magnification of x I6,5oo. Negative staining. Specimens were prepared by mixing equal amounts of the virus-cell suspension and 2 % phosphotungstic acid (PTA) at ph 6. For the study of virus morphology, pelleted virus was resuspended in distilled water to a convenient, i.e. faintly opalescent, density. For the study of virus attachment, pelleted cells were mixed with distilled water and quickly stained with PTA. For the study of haemolysis, washed lysed cells were mixed with distilled water and then stained. Ultrathin sectioning. A standard procedure published elsewhere (Apostolov & Flewett, 1969) was followed. RESULTS Negative staining Morphology of the envelope of Sendal virus By negative staining Sendai virus particles are markedly pleomorphic. Not only the shape, but also the overall thickness and thickness of the constituent layers of the envelope vary from particle to particle. The surface projections on the particles also vary in length, from 5 nm. to 20 nm. (Fig. I, 2). In addition, the particles show considerable variation in the number of spikes present on the surface. There are some with a continuous dense covering, and others with projecting spikeless blobs. The thickness of the layer underneath the spikes also differs from particle to particle.

3 Interaction of Sendai virus and red cells 229 Fig. I. A Sendai virus particle displaying uniform 20 rim. projections and continuous innermost (nanogranular) layer. Particles such as this are not common. Fig. 2. Sendai virus particle attached to red cell membrane. Note the regularity and apparent elonga tion of spikes in contact ~ith the membrane. Interaction of virus particles and cell membranes These phenomena can be studied at higher resolution with this technique. The spikes of attached virus particles are seen to be in close contact with the cell membrane. In the region of contact the spikes are regular and perpendicular to the cell membrane, while the spikes not in contact are less regular and less orderly (Fig. 2). Fig. 3-6 are from preparations where lysis at 37 has proceeded for 2 to 5 min. They are arranged in what seems to be a logical sequence of events in the interaction between virus and cell membrane resulting in haemolysis. Fig. 3 shows the incipient dissolution of cell membrane opposite the spikes of the attached virus particle. Then there is detachment of I6-2

4 230 K. A P O S T O L O V AND JUNE D. A L M E I D A Fig. 3. Initial stages of interaction between Sendai virus and a red blood cell. The cell membrane in the process of dissolution in three areas (arrows). Fragmentation of the membrane is best resolved in the larger area. Fig. 4. The beginning of the fusion of the virus envelope with the cell membrane. Both the innermost layer of the virus and the membrane of the cell show signs of dissolution. spikes, a n d fusion o f the n a n o g r a n u l a r layer with cell m e m b r a n e p r o p e r (Fig. 4, 5). F u s i o n then p r o c e e d s until it is c o m p l e t e - t h a t is, until the n a n o g r a n u l a r layer b e c o m e s an integral p a r t o f the cell m e m b r a n e (Fig. 5, 6). I n this process the internal c o m p o n e n t o f the virus is extruded into the cell while the spikes r e m a i n on the outside. Finally, as Fig. 6 shows, there can be b r e a k s in the virus p a r t o f the i n t e g r a t e d m e m b r a n e with escape o f material. I f haemolysis is allowed to p r o c e e d for l o n g e r p e r i o d s 0 5 rain. o r m o r e ) no fusing

5 Interaction of Sendai virus and red cells 23! Fig. 5. This micrograph shows complete fusion of the innermost virus membrane with the red cell membrane. It should also be noted that the spikes stop at the point of fusion. Fig. 6. A complete entry of a Sendai virus particle into a red blood cell. The innermost layer is fused with the cell membrane, the spikes in the virus part of the membrane are missing, virus internal components are clearly seen inside the red cell. There is a small break in the virus part of the integrated membrane. particles are seen, while most of the internal c o m p o n e n t s of the virus contained within the red cell are shown to be long straight rods. It is i m p o r t a n t to emphasize that, in o u r opinion, the breaks seen in the virus part of the integrated m e m b r a n e are n o t artifacts due to the osmotic effect of distilled water in the process of negative staining. The a d d i t i o n o f distilled water to erythrocytes already lysed by virus failed to increase h a e m o g l o b i n in the supernatant.

6 232 K. A P O S T O L O V AND JUNE D. A L M E I D A Fig. 7. A section o f a virus particle attached to a red cell. The layers o f the virus envelope are clearly seen. Fig. 8. A section of Sendai virus particle after its fusion to a red cell. The virus innermost layer is seen fused to the cell membrane (arrows), the spikes on the outside are absent. Virus internal components are clearly defined inside the cell. Fig. 9. Thin section of fused region between two cells. Two virus particles can be seen bridging the gap between the cells. Fig. ~o. A similar area to Fig. 9 but seen with negative staining. A virus particle is bridging two adjacent membranes, which are showing incipient stages of fusion.

7 Morphology of virus envelope Interaction of Sendai virus and red cells 233 Ultrathin sectioning On ultrathin sections three layers of the virus envelope can be seen. Outermost is the layer of projections, then an intermediate or electron-transparent layer, and finally innermost is the basal or nanogranular layer (Fig. 7, and Apostolov & Flewett, I969). In this respect Sendai, as well as mumps and Newcastle disease virus (unpublished observations), resembles influenza A, B and C (Apostolov & Flewett, 1969). However, there is a significant difference between these two groups of viruses inasmuch as the envelope of influenza A, B and C is uniform in appearance irrespective of the technique used, whereas the envelope of Sendal, NDV and mumps shows variation from particle to particle in the thickness of these layers. Appearance of red cells after virus attachment and after lysis and fusion Our results confirm and extend some of the previous observations of Howe & Morgan (1969). At the attachment site the cell membrane is invaginated and the spikes are perpendicular to it (Fig. 7). After haemolysis has been allowed to proceed the cell membrane is seen to be fused with the virus envelope and the internal component is seen in the lumen of the red cell (Fig. 8). Human red cells that are exposed to virus at 4, pelleted by light centrifugation, and then exposed for 5 rain. at 37, are seen to show lysis and fusion concomitantly. Groups of several cells are fused with apparent loss of haemoglobin. The sites of fusion can be recognized as a darker region containing virus material. Some of the cells are bridged by virus particles in the process of fusing to each of two contiguous cells (Fig. 9). The initial stages of bridging are seen in Fig. IO, obtained by negative staining. DISCUSSION From our results it is clear that the whole of the virus envelope is involved in the interaction with the cell membrane. The process is initiated by the spikes but the actual fusing agent is the innermost or the nanogranular layer. An examination of all the micrographs showing virus particles in the process of fusion to cell membranes showed that they had predominantly long spikes. The failure by different authors to produce precipitous haemolysis by material from disrupted virus particles confirms the importance of an integral virus envelope and innermost layer for the process. In retrospect, the part played by the innermost layer is explained by the fact that it represents converted cell membrane incorporated in the virus envelope. The outcome of interaction between the virus and the cell may be lysis or fusion of two or more cells. These two phenomena can be seen happening simultaneously in Fig. 9. From the morphological point of view, precipitous lysis is leakage of material through holes in the cell membrane. An example is haemoglobin escaping through the IO nm. holes produced by complement in immune haemolysis (Humphrey & Dourmashkin, 1965). Production of holes in the erythrocyte membrane in our system might be effected in two ways. First, fusion of the intact side of a disrupted virus particle, in which Sendai populations abound, may create a' chimney' effect. Secondly, the virus part of the integrated membrane without spikes may be unstable and break with an escape of material. Although the first mode must occur occasionally, our results favour the second one as the normal course of events; these breaks are presumably repaired in actively metabolizing cells with little or no

8 234 K. APOSTOLOV AND JUNE D. ALMEIDA leakage. On the other hand, the integrated membrane may be stabilized or is stable enough, so that no breaks occur. Fusion could occur by the same basic mechanism of simple bridging of two or more cells by the envelope of the virus particles. This mechanism can be seen in those micrographs showing virus material at sites of fusion, and the fact that we could only obtain fusion when the cells with absorbed virus were lightly pelleted, and then exposed to 37. Briefly, in our hypothesis we propose the following sequence of events for virus-induced cell fusion. Virus makes initial contact by means of the surface projections with more than one cell. This leads to dissolution of the cell membrane and loss of virus spikes at the sites of cellular attachment. This stage is followed by fusion of the nanogranular layer of the virus with the cell membrane proper. The interior of the virus particle is now in contact with the interior of one or more cells and there is intrusion of the internal virus components into the cell lumen. The presence of the virus now becomes less obvious and the cell membrane returns to its original outline in the case of attachment to a single cell and close juxtapositioning of adjacent cell membranes in a fusion situation. Finally, spikes disappear from the outer surface of the cell or cells and it is suggested that the innermost or nanogranular layer of the virus is replaced with cell membrane. This stage of replacement of the virus membrane with cell membrane would be an important one in the stabilization of the system - that is, in the prevention of lysis - and would be more likely to occur in actively metabolizing cells. This mechanism directed towards a single cell is the one by which the virus is able to insert infectious nucleic acid into the cell, and it is only fortuitous that the same mechanism occurring simultaneously with two cells leads to fusion. We thank Dr D. J. Bauer for discussions of the present work and checking the manuscript. The excellent technical assistance of Mr J. A. Short and Mrs A. J. Collard is gratefully acknowledged. REFERENCES APOSTOLOV, K. & FLEWETT, T. H. (1969). Further observations on the structure of influenza viruses A and C. Journal of General Virology 4, 365. EIOSAKA Y. & KOSEII, Y. (1968). Electron microscopic study of cell fusion by H.V.J. virions. Virology 34, 419. HOWE, C. & MORGAN, C. (1969). interactions between Sendai virus and human erythrocytes. Journal of Virology 3, 7o. HUMPHREY, J. H. & DOURMASHKIN, R. R. (1965). Electron microscope studies of immune cell lysis. Complement C.LB.A. Foundation Symposium, pp. 175-I86. OI,:ADA, Y. (I958). The fusion of Ehrlich's tumor cells caused by H.V.J. virus in vitro. Biken Journal x, lo3. POSIE, G. (I 97O). Virus-induced polykaryocytosis and the mechanism of cell fusion. Advances in Virus Research x6, 3o3. SCrtNEEBERGER, E. ~. & HARRIS, H. (1966). An ultrastructural study of interspecific cell fusion induced by inactivated Sendai virus. Journal of Cell Science x, 4Ol. (Received 16 November I97I)