Characterization of Measles Viruses in Establishment of Persistent Infections in Human Lymphoid Cell Line

Transcription

1 J. gen. Virol. (I976), 33, I Printed in Great Britain Characterization of Measles Viruses in Establishment of Persistent Infections in Human Lymphoid Cell Line By T. MINAGAWA, T. SAKUMA, S. KUWAJIMA, TOMOKO K. YAMAMOTO AND H. IIDA Department of Microbiology, Hokkaido University School of Medicine, Sapporo o60, Japan (Accepted 16 June I976) SUMMARY Human lymphoid cells (NC-37) were infected with attenuated measles vaccine virus (Schwarz, AIK-C, and CAM-7o strains), subacute sclerosing panencephalitis virus (Mantooth and Halle strains), neurovirulent TYCSA strain, and wild type virus (Edmonston and Toyoshima strains) at an input multiplicity of o-oi. These strains were divided into two groups by their capacity to establish carrier states. CAM-7o, Toyoshima, and Edmonston strains did not set up persistent infections in NC-37 cells, whereas AIK-C strain induced chronic cyclic infection and the Schwarz, TYCSA, Mantooth and Halle strains could set up persistent infections and furthermore two types of persistent infections were recognizable. Cells persistently infected with Schwarz strain contained nucleocapsid structures in both nucleus and cytoplasm, and produced infectious virus of Io 4 to IO 5 p.f.u./ml over IOO days after the inoculation of the virus but the cap-formation of measles antigens on the cell membrane was seldom observed. However, in cells persistently infected with TYCSA strain, nucleocapsid structures were rarely observed in the nucleus, but the cap-formation of measles antigens on the cell membrane was often observed. The titre of carried virus was always higher than the number of cells in the range of lo 6 to IO 7 p.f.u./ml. Mantooth strain was similar to Schwarz strain and Halle strain was similar to TYCSA strain in the properties of their carrier states. These carrier states were stable and the cells grew normally for over one year. INTRODUCTION There are several reports of persistent infoctions by wild type strains of measles virus (Rustigian, i966; Norrby, I967; Minagawa, I97I; Raine et al. I97I; Menna, Collins & Flanagan, I975). These persistent infections would be maintained by virus mutants selected in cultures as compared with the original viruses (Minagawa, I97I; Preble & Youngner, I975). In addition, attenuated measles vaccine virus also established persistent infections such as Schwarz strain in hamster embryonic fibroblasts (Knight, Duff & Rapp, I972) and Schwarz strain in African green monkey brain cells (Bather et al. I973)- Viruses isolated from patients with subacute sclerosing panencephalifis (SSPE), regarded as variants of measles virus, caused persistent infections in human brains (Oyanagi et ak I97I ; Doi et al. I972; ter Meulen, Katz & Muller, 1972). Steele et al. 0976) established a persistent infection with SSPE virus in human prostate adenoma cells. However, Horta-Barbosa et al. (I97I a) and Hamilton, Barbosa & Dubois (1973) reported that SSPE viruses did not possess

2 362 T. MINAGAWA AND OTHERS common properties as attenuated and/or defective viruses and some of them rather resembled wild type measles virus. At present, the difference between attenuated and/or defective viruses and wild type virus is not clear. Schumacher, Albrecht & Tauraso 0972) studied the properties of wild type virulent strains, attenuated vaccine strains and neurotropic strains in their cytopathic effect, plaque formation, growth potential and interferon production, but the common properties of the attenuated virus were not demonstrated. The interactions between measles virus and lymphocytes are important in viral pathogenicity in vivo, since measles virus was strongly associated with lymphoid tissues (Yamanouchi et al. 1973) and isolated from lymph nodes of patients with SSPE (Horta-Barbosa et al. I97I b). Sullivan et al. 0975) reported that measles virus could replicate in human lymphocytes of peripheral blood and furthermore Joseph, Lampert & Oldstone 0975) confirmed the replication of measles virus in B- and T-cells of human lymphocytes. Lymphoblastoid cell lines derived from B- and T-cells of human lymphocytes have been established and employed in various studies. It was of interest whether or not measles virus could replicate in these cell lines. Recently Barry et al. (I976) reported that Edmonston wild type and attenuated viruses were able to replicate and establish persistent infections in both B- and T-lymphoblastoid cells. However, the properties of their carrier cultures were not described in detail. Simultaneously with their work, we established measles-persistent infections in human lymphoblastoid cell lines. The present report describes the establishment of persistent infections in human lymphoid cell line (NC-37) by attenuated measles vaccine virus, neurotropic virus and SSPE virus, and discusses the characterization of various measles virus strains in the establishment and maintenance of persistent infections in NC-37 cells. METHODS Cell cultures. Human lymphoid (NC-37) cell line which had been established in the Pfizer laboratories by Dr E. M. Jensen from the normal peripheral blood of a human donor was supplied by Dr T. Osato of Department of Virology, Cancer Institute of Hokkaido University. NC-37 cell is a B-cell like line and harbours a repressed genome of Epstein-Barr virus (EBV) in each individual cell (Sugawara, Mizuno & Osato, ~972; Reedman & Klein, 1973). NC-37 cells were cultivated in RPMI ~64o medium supplemented with IO ~ (v]v) heatinactivated foetal bovine serum. Vero cells for the assay of measles virus were cultivated in Eagle's minimum essential medium (MEM) supplemented with 4 ~o (v/v) calf serum. These cells were not contaminated with mycoplasma; none was evident by culture or by ultrastructural assessment. Attentuated measles vaccine viruses. The Schwarz strain, vaccine lot no. H3-23 (I.4 Io4 p.f.u./ml) which had been passaged serially in chick embryo cells (Schwarz, I964) was obtained commercially by Takeda Chemical Industries Ltd. The origin of this virus is wild Edmonston strain. The CAM-7o strain, vaccine lot no. 604 (2-8 xo 5 p.f.u.]ml) which had been adapted in chick embryo (Ueda et al. I97O) was obtained commercially from Tanabe Chemical Industries Ltd. The origin of this virus is wild Tanabe strain. The AIK-C strain, vaccine lot no. TV-I2 (3"3 x lo 4 p.f.u./ml) which had been passaged in sheep kidney cells and chick embryo cells (Makino et al. I974) was supplied by Dr S. Makino of Kitasato Institute, Tokyo. The origin of this virus is wild Edmonston strain. At present, these strains are used as vaccine virus in Japan. Subaeute selerosing panencephalitis (SSPE) viruses. The Mantooth strain was isolated from brain biopsy of a patient with SSPE by Horta-Barbosa et al. 0969) and the Halle strain was isolated from lymph node biopsy of a patient with SSPE by Horta-Barbosa et al.

3 Measles persistence in human lymphoid cells 363 (i97i b). These strains which had undergone four passages in Vero cells were kindly supplied by Dr K. Yamanouchi of National Institute of Health, Tokyo. The titres of Mantooth strain and Halle strain were 7"5 x io ~ p.f.u./ml and 9"7 ~o~ p.f.u./ml, respectively. Neurovirulent virus. The TYCSA strain which had been cloned from wild Toyoshima strain by Dr H. Khono of the National Institute of Health and passaged in Vero cells was highly neurovirulent in rodents (Shishido et al. 1973) and composed predominantly of large plaque virus (I.z Io 6 p.f.u./ml). Wild strain viruses. The Edmonston wild strain which had been passaged in human amnion cells was also supplied by Dr K. Yamanouchi (I-O Io 5 p.f.u./ml). The Toyoshima strain was passaged in Vero cells in our laboratory. The titre of the stock virus was 5"5 lo s p.f.u./ml. These wild strain viruses were composed predominantly of small plaque virus. Infection of cell cultures. One ml of NC-37 cells (I to 6 cells) was inoculated with o-2 ml of measles virus at an input multiplicity of o'oi. The mixtures of cells and virus were incubated for I h at the room temperature and then centrifuged at rooo rev/min for 5 rain. The supernatant fluids were discarded and the cell pellets suspended in 5 ml of fresh medium and cultivated in plastic dishes. Every day I ml of cell suspension was harvested and I ml of fresh medium was added for 7 days after the inoculation of the virus, and thereafter every 5 to 7 days 4 ml of culture was harvested and fresh medium was added. The number of live cells and the percentage of haemadsorption-positive cells of the harvested samples were assayed and then the residual samples were frozen at -80 C to assay for total infectivity. Assay of infectious virus. The plaque assay was performed in Vero monolayer cultures with o.i ml of virus inoculum per flask at the appropriate dilution. After virus adsorption by shaking for 3o rain, the monolayers were overlaid with 5 ml of RPMI I64o medium containing I ~ (w/v) Difco agar, o.t 5 ~o (w/v) bicarbonate, 5 ~ (v/v) calf serum, and o.ooz ~(w/v) neutral red. Plaques were counted on the I4th day after the inoculation of the virus. Haemadsorption (Had) assay. Cell (o. 5 ml) suspensions of infected or uninfected NC-37 cells were mixed with o'5 ml of 2 ~oo (V]V) rhesus monkey erythrocytes and incubated at 37 C for 3o min. After that, the mixtures were washed twice with o.m M-phosphate buffered saline solution (PBS) to remove unadsorbed erythrocytes and then observed microscopically. Rhesus monkey erythrocytes were supplied by Dr S. Makino of Kitasato Institute, Tokyo. Fluorescence microscopy of acetone-fixed cells. Cells of carrier cultures were smeared on the glass slide, air-dried and fixed with cold acetone at 4 C for 3o min. They were exposed to fluorescein isothiocyanate (FITC)-labelled gamma globulin of rabbit anti-measles serum obtained commercially from Toshiba Chemical Co. at 37 C for 3o min and washed 3 times with PBS and then observed microscopically. Fluorescence microscopy of unfixed cells. Cells of carrier cultures were washed once with PBS, mixed with FITC-labelled antibody described above at 37 C for 30 min and washed several times with PBS until non-specific labelling with FITC-labelled antibody was removed from uninfected NC-37 cells. After that, labelled cells were mixed with 25 ~ (v/v) glycerin in PBS and observed microscopically. Electron microscopy. Cells of carrier cultures were centrifuged at IOOO rev[min for 5 min. The cell pellet was fixed with 5 ~o (v/v) formalin for ~6 h and post-fixed with 2 ~o (w/v) osmium tetroxide for z h, dehydrated with ethanol and acetone, and embedded in epoxy resin. After polymerization of the epoxy resin, sections were cut with the LKB ultramicrotome and stained with uranyl acetate followed by lead citrate. Electron microscopic

4 364 T, MINAGAWA AND OTHERS (a) ' ' T X. o 6 p..~ -o-.a," en " //..... o " <lt'~, - - * " T ( I I I I I "~ " '' X -0,% <l~ 2 I I I I I I I Days after inoculation of virus Fig. I. The replication of virus and cells in NC-37 ceils infected with measles virus of Edmonston wild strain and Toyoshima strain. NC-37 cells (5 x io ~ cells) in tube were inoculated with Edmonston wild strain (a) and Toyoshima strain (b) at an input multiplicity of o.oi. After adsorption of the virus at room temperature for I h, the mixtures were centrifuged at 1ooo rev/min for 5 rain and then the supernatant fluids were discarded to remove unattached virus and the cell pellets which were resuspended with 5 ml of fresh medium containing 4 ~ (v/v) foetal bovine serum were cultivated in a plastic flask (Falcon 3oi3) at 37 C- At the indicated times, I ml of ceil suspension was harvested to count the number of live cells (O- -O) and the percentage of Had-positive cells (A- -A), and to assay the infectious virus (~--@). observations were made in a Hitachi HS-9 electron microscope. For scanning electron microscopy, the formalin-fixed cells were dehydrated in ethanol and acetone, continued through CO 2 for critical-point drying, coated with gold, and examined on a JSM-SI scanning electron microscope. Cloning of cells in semi-solid agar. Two hundred cells with measles-persistent infections were suspended in RPMI I64o medium containing o'3 ~o Difco agar, zo ~ foetal bovine serum and I o ~ rabbit anti-measles serum, plated on to plastic dishes (45 mm), and incubated at 37 C for I month in a CO2 incubator. After that, colonies were picked up with capillaries and further cultivated in RPMI T640 medium containing Io ~ foetal bovine serum, and then the infectious centres, haemadsorption test and fluorescent antibody technique (FAT) of cells of each clone were carried out.

5 - -- d "G A < <1 (a) Measles persistence in human lymphoid cells i, t t ) u ~o,.,,,.7, \ xx/& t ', & -4 ~ 'a-" N. o I I I I I I P I I ~ ~ - /a ~ e e o..... o ~ ~ I I I! I Days after inoculation of virus Fig. 2. The replication of virus and cells in NC-37 cells infected with attenuated measles virus of AIK-C strain (a) and CAM-7o strain (b) at an input multiplicity of o.oi. (3- -, the number of live cells; A- -A, the percentage of Had-positive cells; O---O, the titre of infectious virus. RESULTS Infection with wild strains NC-37 cells were inoculated with Edmonston wild strain and Toyoshima strain at a multiplicity of o.oi. As shown in Fig. I, virus yield and Had-positive cells reached the maximum on the 3rd day and 6th day after the inoculation of Toyoshima strain and Edmonston strain, respectively. The maximum yields of infectious virus of these viruses were very high, more than ios p.f.u./ml; cell fusion was not observed and all cells were degraded after the maximum yields of the virus had occurred. Although infectious virus and live cells could not be detected 2o days after the inoculation of these viruses, the cultures were continued for 2 months. However, cell growth was never observed. The infections with these viruses at input multiplicities of o.ooi and o.oooi were chased for over 5 months, but persistent infections were never established. 100 so m 365 Infection with attenuated vaccine strains NC-37 cells were inoculated with AIK-C, CAM-7o and Schwarz strains at an input multiplicity of o.oi. As shown in Fig. 2(a), chronic cyclic infection of AIK-C strain in NC-37 cells was observed, but persistent infection was not established. Virus yield and Had-positive cells reached the maximum on the 3oth day after the inoculation of the virus, when fusion of infected cells was seen frequently. After that, fused cells were degraded and the virus yield decreased. When the virus yield reached the minimum, cell growth began again. On the 5oth day after inoculation of the virus, the second peak of infectious virus appeared. Thereafter, the growth of cells ceased. Thus the chronic cyclic infection was characteristic for the infection with AIK-C strain. As compared with other attenuated strains, fusion activity of AIK-C strain appeared to be too strong to establish a persistent

6 366 T. MINAGAWA AND OTHERS ~'l,') ' ' o o '.o-o ~- '-... ~- - [-,,,<~.,,%>.%.~,o- c,- o- o-_ o -~... --o 5 to ~ #2 4'" ~'-,../ \ / - ~ ~ ~._._....., "" -" " " 4 / ~ l ~ i ~ l ~ I ~ I ~ I A I A I ~ I ~ I ~ ~ I ~ I -- A A A -- --~4X-- "A 100 ~gf 3 2 NC-Schwarz 50 -~ O v 2.6 < 6 I r I I I O /)' ~'-. ~ ~'~'--'~'-'--" ~ ~ -0- ' o- ~ -0- ' ~ ~'- ~ "X3"- - o - ~ -oi-- ~ ~ o ~ / e ~ - _ 5 ~\/,"A to.\...,.- -o- 4 ~'~ "ba ~----~'~----~--a----<----'~--~--a--a--~----~----~-- --A---- ~-- A I l NC-TYCSA 50 <1 1 s ZZ I I I I I O Days after inoculation of virus Fig. 3. The replication of virus and cells in NC-37 cells infected with attenuated measles virus of Schwarz strain (a) and neurovirulent virus of TYCSA strain (b) at the input multiplicity of o-oi. O- -O, the number of live cells; A--A, the percentage of Had-positive cells; 0---0, the titre of infectious virus. infection. CAM-7o strain did not induce cell fusion by infection at an input multiplicity of o.oi. As shown in Fig. 2 (b), virus yield and Had-positive cells reached the maximum on the 8th day after the inoculation of the virus, and thereafter infected cells were degraded and virus yield decreased. This culture was further continued and 5o days later cell growth was again observed. These cells, however, were not infected with the virus, since the growing cells were Had- and fluorescent antibody technique (FAT)-negative and did not produce infectious virus, and furthermore were as sensitive to wild type measles virus as the un, infected NC-37 cells. Although the experiments were repeated at a multiplicity of o.om, CAM-7o strain could not set up a persistent infection in NC-37 cells. On the other hand, a persistent infection was readily established after the infection with Schwarz strain. As shown in Fig. 3 (a), infected cells were growing without any detectable c.p.e, of measles virus and on the 15th day after inoculation of the virus the maximum yield of the virus was observed, and too days later a stable carrier state was established. In this system, it was characteristic that the titre of infectious virus in the culture was always lower than the number of live cells in the range of o.i to o.oi p.f.u./cell. Infection with neurovirulent strain Fig. 3 (b) shows that TYCSA strain could set up a persistent infection in NC-37 cells. Virus yield and Had-positive cells reached the maximum on the 5th day after the inoculation of the virus, and then cells decreased but recovered 2o days later, and thereafter the replication of cells and virus were seen constantly. In this system, cell fusion was seen often, but fused cells were not rapidly degraded, and a high titre of virus (more than Io 6 p.f.u.]ml) was produced.

7 Measles persistence in human lymphoid cells 367. (.) i i, i t J, i fi 5 ~,;_.<.,~-7 ~ ~ -.--\ 4 g'-\,,".a--a-a----a---~-~---~-~-,~--a-~----~-- --~-----a--~ [~-~ --o--d 2 i,~, \4, i" NC-Mantooth <1 "--0 = I I I I I! I I I I 0 5o z: = 6. e--'\/'/'-'-"-*~./~ / 5 %. o- -o-~ o/~_ o o-~_ - -O.o". -..o ~c-.. _ - o... o- -.o 4 ~- -~. a.a-- ~-~-a-=-- -a---~-z-- -z-e,-=-~-~ ~... a-- -- <,- ~ - t00.f~ / /.1 3 -c- -o" 2 t,1, + NC-Halle <I l" I I P I I l i I Days alter inoculation of virus Fig. 4. The replication of virus and cells in NC-37 cells infected wilh SSPE virus of Mantooth strain (a) and Halle strain (b) at an input multiplicity of o.oi. --, the number of live cells; &--A, the percentage of Had-positive cells; O--O, the titre of infectious virus. Infection with SSPE viruses As shown in Fig. 4, virus yield and Had-positive cells reached the maximum on the 3rd day after the inoculation of Mantooth strain and Halle strain. After that, because of the severe c.p.e, with these viruses, almost all cells were degraded and intact cells were not detectable, but these cultures were kept further in a CO~ incubator. After 50 days' cultivation, cell growth and infectious virus were observed and almost all the growing cells were shown to be Had-positive. Thereafter, cells of these cultures were growing and releasing infectious virus constantly. The pattern of the replication of cells and virus in the culture persistently infected with Halle strain resembled the carrier state with TYCSA strain, since cell fusion was often observed and virus yield was always higher than the number of cells (Fig. 4b). On the other hand, cells infected with Mantooth strain resembled the carrier state with Schwarz strain, since cell fusion was seldom observed and virus yield was always lower than the number of cells (Fig. 4a). Growth curves of cells of carrier cultures Growth curves of cells of carrier cultures with Schwarz strain (NC-Schwarz), TYCSA strain (NC-TYCSA), Mantooth strain (NC-Mantooth) and Halle strain (NC-Halle) were examined simultaneously. As shown in Fig. 5, cells of NC-Schwarz and NC-Mantooth were growing at the same rate as uninfected NC-37 cells, whereas NC-TYCSA and NC-Halle cells showed a longer lag phase than NC-37 cells. However, such differences were not observed when 5 x lo 5 cells of each carrier culture were plated on to dishes. 50

8 368 65[ 6.0 T. MINAGAWA AND OTHERS I ' I i I i I ' /Qm /Y / E 6 2 i o J / 5.5 i Days Fig. 5. Growth curves of cells of carrier cultures. Cells of carrier cultures of NC-Schwarz (0--0), NC-Mantooth (A--A), NC-TYCSA (11--11), NC-I-Ialle (V--V) and uninfected NC-37 ( - -O) were diluted to Io 5 cells/ml with growth medium, One ml of each cell suspension was seeded on to ~6 plastic dishes (Falcon 35 ram). Every day the number of cells in 2 dishes of each culture was cotmted. Fluorescent antibody staining Fixed and unfixed cells were stained with FITC-labelled anti-measles serum gammaglobulin which had been absorbed with sufficient NC-37 cells. Fig. 6 and 7 show the membrane fluorescent antibody staining of cells. Ring-shaped measles antigens were observed on the cell membrane of all cells of NC-Schwarz and NC-Mantooth (Fig. 6 a and b). Measles antigens of NC-Halle cells were localized on the cell membrane as if cap formation of measles antigens had occurred (Fig. 6c). This capping of measles antigens in almost all NC-TYCSA cells was also observed as shown in Fig. 7. A polar distribution of measles antigens on the cell membrane of acetone-fixed NC-HaUe and NC-TYCSA cells was observed but the localization of measles antigens in nucleus or cytoplasm was not determined by this antibody. Electron microscopic observations As shown in Fig. 8, tubular structures were seen in both nucleus and cytoplasm of NC-Schwarz and NC-Mantooth cells. Tubules in the cytoplasm and in the nucleus appeared to be identical with nucleocapsid structures of granular filament (GF) and smooth filament (SF), respectively, as reported by Oyanagi et al. (~970- We could not detect GF in the nucleus

9 Measles pers&tence in human lymphoid cells 369 Fig. 6. Membrane-fluorescent antibody staining of cells of carrier cultures. Cells of NC-Schwarz (a), NC-Mantooth (b) and NC-Halle (c) were washed once with o'oi M-PBS and stained by the direct methods as described in the text. Note the capping of measles antigens on the cell membrane of NC-Halle cells. Magnification 800.

10 370 T. MINAGAWA AND OTHERS Fig. 7- Membrane-fluorescent antibody staining of NC-TYCSA cells. Magaification x 4o0 (a) and ~6oo (b). Note the capping of measles antigens of almost all cells of NC-TYCSA. and SF in the cytoplasm of NC-Schwarz and NC-Mantooth cells. Furthermore, numerous pleomorphic particles with a diam. of 200 to 250 nm were seen on the cell membrane and in the intercellular space. Fig. 9 shows the fine structures of tubules in the nucleus and in the cytoplasm, and virus particles in the intercellular space. Tubules in the nucleus are hollow structures with periodic striations on the outer surface and their width is uniform, r4 to ~6 nm. The inner diam. of the tubules is 5 to 7 nm (Fig. 9a). Although the tubules in the cytoplasm appeared to have the same structure as those in the nucleus, granular materials were attached to the outer surface and the width of these structures was not uniform, 17 to 25 nm (Fig. 9 b). Virus particles in the intercellular space had surface projections and tubular structures were seen under the envelopes of the particles. The inner diam. of the tubular structures of these particles is the same as the diam. of tubules in the nucleus (Fig. 9 c). These particles are morphologically similar to defective SSPE virus reported by Raine et al. (I973) and Raine, Byington & Johnson (I974). On the other hand, in NC-Halle and NC-TYCSA cells, tubules in the nucleus were seldom observed but those in the cytoplasm and capping of cellular microvilli were observed (Fig. ~oa). Tubular structures (GF) were observed only in the cytoplasm under the cap of microvilli (Fig. iob) and virus particles which were budding from the surface of microvilli were seen (Fig. ~ o c). The cap of microvilli was seen often in NC-TYCSA and NC-Halle cells. In addition, the cell surfaces of NC-Schwarz, NC-TYCSA and uninfected NC-37 cells were observed by scanning electron microscopy. Numerous microvilli diffusely covering the cell surface of NC-37 cells and NC-Schwarz cells were shown in Fig. r I (a) and (b), although the microvilli of NC-Schwarz cells were shorter than those of NC-37 cells. On the other hand, the cell surface of NC- TYCSA cells was characteristic. There were two kinds of cell in NC-TYCSA cultures. One was similar to NC-Schwarz cells and another formed caps of microvilli on the cell surface (Fig. I I c and d). Cells of the latter type were prominent in NC-TYCSA cells. This cap formation of microvilli appeared to be identical with that observed in Fig. 7 and Io. Cloning of cells by the semisolid agar method Only one clone in a total of 49 clones did not produce infectious virus and was negative in the Had test and FAT, while other clones were similar to parent NC-Schwarz cells. Similar negative clones were obtained from NC-TYCSA and NC-Halle cells (Table I). This showed

11 Measles persistence in human lymphoid cells 37I Fig. 8. Electron photomicrographs of NC-37 cells persistently infected with Schwarz strain (a) and Mantooth strain (b). Tubular structures in the nucleus and in the cytoplasm of these cells and virus particles in the intercellular space can be seen.

12 372 T. MINAGAWA AND OTHERS Fig. 9. High magnification of tubular structures in the nucleus (a) and in the cytoplasm (b), and virus particles in the intercellular space (c) of NC-37 cells persistently infected with Schwarz strain. Note the hollow structure with periodic striations on the outer surface (arrow on a) and their uniform width of tubules in the nucleus, the granular structure of tubules (arrow on b) without uniform width in the cytoplasm, and the surface projections on the envelope and cross-sections of tubular structures (arrow on c) under the envelope of the virus particles.

13 Measles persistence in human lymphoid cells 373 Fig. I o. Electron photomicrographs of NC-37 cells persistently infected with TYCSA strain. (a) Cap formation of microvilli on the cell membrane and tubular structures in the cytoplasm under the cap of microvilli are seen (b) High magnification of tubular structures of scattered type in the cytoplasm and (c) particles with alignments between microvilli. 25 vrg 33

14 374 T. MINAGAWA AND OTHERS Fig. l L. Scanning electron photomicrographs of uninfected NC-37 cell (a), NC-Schwarz cell (b) and NC-TYCSA cells (c and d). The surface of the NC-37 cell and the NC-Schwarz cell is covered with the numerous microvilli but the cap formation of microvilli is seen often in NC-TYCSA cells. Magnif.cation x 8ooo. Table T. Summary of the properties of carrier cultures Virus- Cap formation Carrier producing of measles cultures Had(X) FAT (~) c.p.e.* clones antigens* NC-Schwarz leo Ioo - 48/49? - NC-Mantooth I /50 -- NC-TYCSA 99 9o + 46/5o NC-Halle 99 9o + 43/5o + Nucleocapsids* in Nucleus Cytoplasm (+) ++ * --, positive cells were not detectable; +, positive cells were less than 25 ~/oo; -[-'~-, positive cells were about 5o ~; + + +, positive cells were more than 75 ~. t Virus-producing clones per total clones. that almost all infected cells were able to form colonies and produce infectious virus. Negative clones would be obtained by segregation of the virus genome at the division of cells as shown in a previous paper (Minagawa, I97I). Properties of carried virus Plaque sizes of Schwarz strain, Mantooth strain, TYCSA strain and Halle strain were compared with those of viruses produced from cultures of NC-Schwarz, NC-Mantooth,

15 A Measles persistence in human lymphoid cells 375 Fig. 12. Plaque size of parent and carried viruses. Plaque size of carried virus produced from NC-Schwarz (a), NC-Mantooth (b), NC-TYCSA (e) and NC-Halle (d) was compared with that of each parent virus. Table 2. Efficiencies of plaquing of measles viruses at 39 C* Parent virus Carried virus Loglo (p.f.u./ml) Efficiencies Log10 (p.f.u./ml) Efficiencies, of plaquing c--"-oc C' of plaquing c C Virus strains (39/34 C) C (39[34 C) AIK-C 4'00 3'26-0"74 5"53 3'86 - I "67 Schwarz 4" 65 4' o-47 4" 3o 2" 8o - I- 5o Mantooth 5"54 5"36 - o- 18 4"o4 2"69 - I' 35 TYCSA 6"4o 6"3o -o.io 6'53 5'82 -o'7i Halle 5 "79 5 "48 - o'31 5" 68 4" 8 o - o. 88 * Viruses were adsorbed on to Vero monolayers at room temperature for 30 min. Infected cultures were incubated at 34 C and 39 C (+_0"5 C) after they were overlaid with nutrient agar medium. Plaques were counted on the ~4th day after the inoculation of the virus. In all experiments the titres of each virus express mean values of duplicate tests. 25-2

16 T. MINAGAWA AND OTHERS Virus strains Edmonston Toyoshima CAM-7o Table 3. Efficiencies of ptaquing of measles viruses at 39 C* Original virus NC-37 passed virus c % f Logt0 (p.f.u./ml) Efficiencies Log10 (p.f.u./ml) ~--~ of plaquing c ~ 34 C 39 C (39/34 C) 34 C 39 C 5"36 5"I8 -o.i "97 6"I5 6"08 --0" q5 4"94 4"60 --0"34 5'5t 5-II * The methods were as described in Table 2. Efficiencies of plaquing (39/34 C) "45 NC-TYCSA and NC-Halle, respectively. As shown in Fig. 12, the plaque size of the parent strain was not different from that of each carried virus. Each strain had a characteristic marker of plaque size and this marker was kept in virus produced from each carrier culture. It is certain that these carrier cultures were not established by infection with contaminated virus. The efficiencies of plaquing at 39 C of carried viruses, especially viruses from NC-AIK-C, NC-Schwarz and NC-Mantooth cultures, were lower than those of the parent viruses (Table 2), whereas the efficiencies of other strains passed in NC-37 cells were not different from those of the original viruses (Table 3). Since the heat inactivation curves of these carried viruses were not different from those of their parent viruses, temperature sensitive (ts) variants might be selected in the establishment of these carrier cultures. However, ts variant viruses might not play a role in the maintenance of these carrier states, since the carrier cultures were not degraded by shift of cultivation temperature from 37 C to 34 C; rather, attenuated and/or defective viruses might be selected in carrier cultures, especially in NC-Mantooth and NC-Halle cultures, since these carried viruses readily established persistent infections in NC-37 cells in the same way as Schwarz strain. DISCUSSION Human lymphoid (NC-37) cells were inoculated with wild type measles virus, attenuated measles vaccine virus, SSPE virus and neurovirulent virus at an input multiplicity of o-oi. Two wild type measles viruses and one attenuated vaccine virus showed acute and cytolytic infection, but two attenuated viruses, two SSPE viruses and one neurovirulent virus showed chronic or persistent infections in NC-37 cells. NC-37 cells appeared to be so optimal for the replication of measles virus that the maximum yield of wild Edmonston and Toyoshima strains was more than IO s p.f.u./ml. These wild type viruses could never set up persistent infections in NC-37 cells even when the m.o.i, of these viruses was lower than o.ooi. Furthermore, wild strains freshly isolated from patients with natural measles infection did not establish a persistent infection in NC-37 cells at any multiplicity of infection (T. Minagawa & S. Kuwajima, in preparation). Recently, Barry et al. (1976) established persistent infections of Edmonston wild and attenuated strains in human lymphoid cell lines. They inoculated the virus at a very low multiplicity so that attenuated and/or defective variants might be produced and selected for the establishment of persistent infections in the same way as other persistent infections with wild type viruses as described in the introduction. Consequently, it is necessary to compare the carried virus with the original inoculum. On the other hand, Schwarz vaccine strain replicated slowly in NC-37 cells without any detectable c.p.e, and readily established a persistent infection. However, AIK-C vaccine strain induced a chronic cyclic infection but did not establish a persistent infection, and CAM-7o vaccine strain began to grow rapidly and its maximum titre was higher than that

17 Measles persistence in human lymphoid cells 377 of Schwarz strain and all cells were degraded. Thus AIK-C and CAM-7o strains appeared to be different from the Schwarz strain in their virulence in NC-37 cells. The first one-step growth of Mantooth and Halle strains was not different from that of CAM-7o strain and almost all cells were infected and degraded. Similar results were obtained by Castro, Burnstein & Byington (i972) and Hamilton et al. (1973) using a hamster brain-adapted Mantooth strain in Vero cells and Halle strain in BSC-I cells, respectively. However, the attenuated and/or defective virus population may have been selected from the virulent virus population and 5o days later persistent infections were established. In these cases, viruses produced from NC-Mantooth and NC-Halle cultures readily established persistent infections in NC-37 cells in the same way as the Schwarz strain did. In addition, the neurovirulent TYCSA strain readily established a persistent infection in NC-37 cells. This strain was a clone variant of the wild Toyoshima strain and was highly neurovirulent in newborn hamsters (Shishido et al. I973). Neurotropism of measles virus in rodents does not necessarily correlate with the wildness of the virus, but rather the passage history of the virus. Janda, Norrby & Marsyk (I97I) reported that the undiluted passage (UP) variant which contained predominantly DI particles (Hall, Martin & Gould, 1974) was highly neurotropic. UP virus displayed a pronounced autointerference during multiplication in vitro and induced chronic infections in HeLa ceils (Oddo, Flaccomio & Sinatra, 1961). Thus the neurovirulence of TYCSA strain might not conflict with the establishment of a persistent infection in NC-37 cells. Measles virus strains, therefore, were divided into two groups by their capacity to establish carrier states in NC-37 cells. Edmonston wild strain, wild Toyoshima strain and CAM-7o strain could not set up persistent infections in NC-37 cells, whereas Schwarz strain, AIK-C strain, TYCSA strain, Mantooth strain and Halle strain could set up chronic or persistent infections. The latter, furthermore, was differentiated into two types of carrier states, namely NC-Schwarz type and NC-TYCSA type. NC-Mantooth cells belong to NC-Schwarz type and NC-Halle cells belong to NC-TYCSA type. To compare the properties of NC-Schwarz type with those of NC-TYCSA type, the following observations were made: (I) cell fusion was often seen in NC-TYCSA cultures, but never in NC-Schwarz cultures; (2) virus yield from NC-TYCSA cultures was always higher than that from NC-Schwarz cultures, and virus titre in NC-TYCSA was always higher than the number of cells but virus titre in NC-Schwarz was always lower than the number of cells; (3) nucleocapsid structures were seen in both nucleus and cytoplasm of NC-Schwarz cells, but only in the cytoplasm of NC-TYCSA cells; (4) cap formation of microvilli and measles antigens were seldom seen on the cell membrane of NC-Schwarz cells, but often on the cell membrane of NC-TYCSA cells. Considering the results so far obtained, it is speculated that virus in NC-Schwarz and NC-Mantooth cells might be more defective than that in NC-TYCSA and NC-Halle cells. Almost all cells obtained by cloning of cells of these carrier cultures in the presence of antimeasles serum were infected and produced infectious virus, so measles-persistence of these cultures might be maintained by the vertical transmission of virus nucleocapsids. Additionally, in NC-TYCSA and NC-Halle cultures, capping of measles antigens may play a role in the maintenance of persistent infections. A repressed EBV genome which is harboured in each individual NC-37 cell was not affected by infection with measles virus and may play no role in the establishment or maintenance of measles persistence, since Schwarz, TYCSA and Halle strains established persistent infections in the T-cell type lymphoid cell line (CCRF-CEM) which does not contain EBV genome (T. Minagawa, to be published).

18 378 a:. MINAGAWA AND OTHERS Cap formation of microvilli and measles antigens at the cell membrane of NC-TYCSA and NC-Halle cells was observed by immunofluorescent antibody technique and electron microscopy. Ehrnst, Weiner & Norrby (1974), Joseph & Oldstone (I974) and Lampert, Joseph & Oldstone (1975) reported the capping of measles antigens on cells infected with measles virus by the treatment of infected cells with anti-measles serum. However, this cap formation of measles antigens in NC-TYCSA and NC-Halle cells occurred spontaneously without anti-measles serum. TYCSA strain was a large plaque virus and Halle strain contained very large plaque virus and these two viruses induced severe cell fusion in Vero cell cultures. It was speculated that the accumulation of virus antigens on the cell membrane might be prominent in cells infected with either TYCSA strain or Halle strain. Thus the possibility of variation of virus antigenic expression in these cells might be supported by the results reported by Joseph, Perrin & Oldstone (1976). Such virus modification of cell membranes may stimulate the movement of cytoplasmic membranes and induce cap formation of measles antigens in order to clear out the foreign antigens from the cell surface. This phenomenon may play a role in the maintenance of measles-persistence and ensure a high titre of infectious virus produced from NC-TYCSA and NC-Halle cultures. At present there are no markers in vitro of attenuated virus, neurotropic virus or SSPE virus. Considering the results so far obtained, the establishment of persistent infections in NC-37 cells may reveal common properties of attenuated and]or defective virus. Therefore, investigations of the mechanism of persistent infection of measles virus in NC-37 cells may provide the mechanism of attenuation and/or defectiveness of measles virus. We thank Dr T. Osato, Dr K. Abe, Dr K. Yamanouchi and Dr S. Makino for their encouragements and helpful advice. REFERENCES BARRY, D. W., SULLIVAN, J. L., LUCAS, S. J., DUNLAP, R. C. & ALBRECHT, P. (I976). Acute and chronic infection of human lymphoblastoid cell lines with measles virus. Journal of Immunology xi6, BATHER, R., FURESZ, J., FANOK, A. G., GILL, S. D. & YAROSH, W. (I973). Long term infection of diploid African green monkey brain ceils by Schwarz measles vaccine virus. Journal of General Virology 2o, 4oi-4o5. CASTRO, A. E., BURNSTEIN, T. & BYINGTON, D. P. (1972). Properties in cell culture of a hamster brain-adapted subacute sclerosing panencephalitis virus. Journal of General Virology x6, DOI~ Y.~ SANPE, T., NAKAJIMA, M., OKAWA, S., KATHO, T., ITHO, H., SATe, T., OGUCHI, K., KUMANISHI, T. & TSUBAKI, T. (I972). Properties of a cytopathic agent isolated from a patient with subacute sclerosing panencephalitis in Japan. Japanese Journal of Medical Science and Biology 25, EHRNST, A., WEINER, L. & NORRBY, E. (1974). Fluctuation and distribution of measles virus antigens in chronically infected cells. Nature, London 248, HALL, W. W'., MARTIN, S. J. & GOULD, E. (1974). Defective interfering particles produced during the replication of measles virus. Medical Microbiology and Immunology x6o, 155-I64. HAMILTON, R., BARBOSA, L. & DUBOIS, M. (1973). Subacute sclerosing panencephalitis measles virus: study of biological markers. Journal of Virology I2, HORTA-BARBOSA, L., FUCCILLO, D. A., SEVER, J. L. & ZEMAN, W. (I969). Subacute sclerosing panencephalitis: isolation of measles virus from a brain biopsy. Nature, London 22x, 974. HORTA-BARBOSA, L., FUCCILLO, D. A., HAMILTON, R., TRAUB, R., LEY, A. & SEVER, J. L. (I97I a). Some characteristics of SSPE measles virus. Proceedings of the Society for Experimental Biology and Medicine i34, HORTA-BARBOSA, L., HAMILTON, R., WaTrlO, B., FUCClLLO, D. A. & SEVER, S. L. (I971 b). Subacute sclerosing panencephalitis: isolation of suppressed measles virus from lymph node biopsies. Science, New York I73, 84o-841. JANDA, Z., NORRBY, E. & MARUSYK, H. (197I). Nettrotropism of measles virus variants in hamster. Journal of Infectious Diseases I24, JOSEPH, B. S. & OLOSTONE, M. B. A. (I974). Antibody-induced redistribution of measles virus antigens on the cell surface. Journal oflmmunology xx3, I2O5-12o9. JOSEPh, B. S., LA~rPERT, P. W. & OLDSTONE, M. B. A. (I975). Replication and persistence of measles virus in defined subpopulation of human leukocytes. Journal of Virology x6, 1638-I649

19 Measles persistence in human lymphoid cells 379 JOSEPH, B. S., PERPaN, L. H. & OLDSTONE, M. B. A. 0976). Measurement of virus antigens on the surfaces of HeLa cells persistently infected with wild type and vaccine strains of measles virus by radioimmune assay. Journal of General Virology 30, KNIGHT, P., DUFF, g. & RAPe, r. (I97Z). Latency of human measles virus in hamster cells. Journal of Virology Io, 995-IOOl. LAMPERT, P.W., JOSEPI~, B.S. & OLDSTONE, M.B.A. (I975). Antibody-induced capping of measles virus antigens on plasma membrane studied by electron microscopy. Journal of Virology xs, Iz48-I255. MAKINO~ S., SASAKI~ K., NAKAMURA, N.~ NAKAGAWA~ M.~ NAKAJIMA~ S., NAOI, M.~ INOUE, K.~ HASHIMOTO, T., MIZUNOE, K. & KASAHARA, S. (I974). Studies on the modification of the live AIK measles vaccine. II. Development and evaluation of the live AIK-C measles vaccine. The Kitasato Archives of Experimental Medicine 47, 13-2I. lvlenna, J. H., COLLINS, A. R. & FLANAGAN, T. D. (I975). Characterization of an in vitro persistent-state measles virus infection: establishment and virological characterization of the BMG/MV cell line. Infection and Immunity ii, I52-I58. MINAGAWA, T. (I97I). Studies on the persistent infection with measles virus in HeLa cells. I. Clonal analysis of cells of carrier cultures. Japanese Journal of Microbiology 15, I. MINAGAWA, T. (1971). Studies on the persistent infection with measles virus in HeLa cells. II. The properties of carried virus. Japanese Journal of Microbiology 15, NORRBV, E. (I967). A carrier cell line of measles virus in Lu to6 cells. Archivfiir die Gesamte Virusforschung 20, oooo, r. ~., FLACCOMIO, R. & SlNATRA, A. 0961). Giant cell and strand-forming cytopathic effect of measles virus lines conditioned by serial propagation with diluted or concentrated inoculum. Virology 13, 55o-553. OYANAG1, S., TER MEULEN, V., KATZ, M. & I,:OPROWSKI, ri. (1971). Comparison of subacute sclerosing panencephalitis and measles viruses: an electron microscope study. Journal of Virology 7, I76-I87. PREBLE, O. Z. & YOtrNGNER, J.S. (I975). Temperature-sensitive viruses and the etiology of chronic and inapparent infections. Journal oflnfectious Diseases I3r, RAINE, C. S., BYINGTON, O. P. & JOHNSON, K. l'. (I974). Experimental subacute sclcrosing panencephalitis in the hamster; ultrastructure of the chronic disease. Laboratory Investigation 3 x, RAINE, C. S., FELDMAN, L. A., SHEePARD, R. D. & BORNSTEIN, M. B. (1971). Ultrastructural study of long-term measles infection in cultures of hamster dorsal-root ganglion. Journal of Virology 8, RAINE, C. S., EELDMAN, L. a., SHEPPARD, R. O. & BORNSTEIN, M. 13. (1973). Subacute sclerosing panencephalitis virus in cultures of organized central nervous tissue. Laboratory Investigation 28, o. REEDMAN, B. M.,~ KLEIN, 6. (1973)- Cellular localization of an Epstein-Barr virus (EBV)-associated complement-fixing antigen in producer and nonproducer lymphoblastoid cell lines. International Journal of Cancer xr, o. RUSrlGIAN, R. (1966). Persistent infection of cells in culture by measles virus. I. Development and characteristics of HeLa sublines persistently infected with complete virus. Journal of Bacteriology 92, SCHtrMACHER, H.P., ALBRECHT, P. & TAURASO, N. M. (I972). Markers for measles virus. II. Tissue culture properties. Archiv fiir die Gesamte Virusforschung 36, lo. SC~WARZ, a. J. r. (I964). Immunization against measles; development and evaluation of a highly attenuated live measles vaccine. Annales Paediatrics 202, SHISHIDO, A., KATOH, S., KOBUNE, K. & SATO, T. (1973). Growth of measles virus in nervous tissues. I. Neurotropic properties of measles virus in newborn hamsters. Japanese Journal of Medical Science and Biology 26, STEELE, R. W., FUCCILLO, D. A., HENSEN, S. A., VINCET, M. M. & BELLANTI, J. A. (I976). Specific inhibitory factors of cellular immunity in children with subacute sclerosing panencephalitis. The Journal of Pediatrics 88, SUGAWARA, K., MIZUNO, F. & OSATO, T. (1972). Epstein-Barr virus-associated antigens in nonproducing clones of human lymphoblastoid cell lines. Nature New Biology 239, StrLLlVAN, J. L., BARRY, O. W., LUCAS, S. J. & ALBRECI-IT, P. (I975). Measles infection of human mononuclear cells. I. Acute infection of peripheral blood lymphocytes and monocytes. The Journal of Experimental Medicine I42, TER MEtrLEN, V., KATZ, M. & MtrLLE~, D. (1972). Subacute sclerosing panencephalitis: a review. Current Topics in Microbiology and Immunology 57, UEDA, S., TAKAHASHI, M., MINEKAWA, Y., OGINO, T., SUZUKI, N.~ YAMANISHI, K., BABA, K. & OKUNO, Y. (I970). Studies on further attenuated live measles vaccine. I. Adaptation of measles virus to the chorioallantoic membrane of chick embryo and clinical tests on the strain. Biken's Journal x3, 11 I-I I6. YAMANOUCHI, K., CHINO, F., KOBUNE, F., KODAMA, H. & TSURUHASHI, T. (1973). Growth of measles virus in the lymphoid tissues of monkeys. Journal of Infectious Diseases I28, (Received 5 April I976)