key: cord-0793101-fdegv6hf
authors: Plumb, Jonnie; Duprex, W. Paul; Cameron, C. H. Stewart; Richter-Landsberg, Christiane; Talbot, Pierre; McQuaid, Stephen
title: Infection of human oligodendroglioma cells by a recombinant measles virus expressing enhanced green fluorescent protein
date: 2002
journal: J Neurovirol
DOI: 10.1080/135502802317247785
sha: 24b98b0cada166c802aa2556d8428b5b144f7f47
doc_id: 793101
cord_uid: fdegv6hf

One of the hallmarks of the human CNS disease subacute sclerosing panencephalitis (SSPE) is a high level of measles virus (MV) infection of oligodendrocytes. It is therefore surprising that there is only one previous report of MV infection of rat oligodendrocytes in culture and no reports of human oligodendrocyte infection in culture. In an attempt to develop a model system to study MV infection of oligodendrocytes, time-lapse confocal microscopy, immunocytochemistry, and electron microscopy (EM) were used to study infection of the human oligodendroglioma cell line, MO3.13. A rat oligodendrocyte cell line, OLN-93, was also studied as a control. MO3.13 cells were shown to be highly susceptible to MV infection and virus budding was observed from the surface of infected MO3.13 cells by EM. Analysis of the infection in real time and by immunocytochemistry revealed that virus spread occurred by cell-to-cell fusion and was also facilitated by virus transport in cell processes. MO3.13 cells were shown to express CD46, a MV receptor, but were negative for the recently discovered MV receptor, signaling leucocyte activation molecule (SLAM). Immunohistochemical studies on SSPE tissue sections demonstrated that CD46 was also expressed on populations of human oligodendrocytes. SLAM expression was not detected on oligodendrocytes. These studies, which are the first to show MV infection of human oligodendrocytes in culture, show that the cells are highly susceptible to MV infection and this model cell line has been used to further our understanding of MV spread in the CNS.

Measles virus (MV) has been identi ed as the etiological agent of the human CNS infections subacute sclerosing panencephalitis [SSPE] (ter Meulen et al, 1983) and measles inclusion body encephalitis [MIBE] (Agamanolis et al, 1979) . Both diseases are characterize d by infection of neurons and glial cells after incubation periods ranging from months, MIBE, to years, SSPE, following the primary infection. As yet, the site of viral persistence within the body remains unknown (Rima et al, 1995) . Restrictions of viral gene expression have been demonstrated in brain tissues from patients with both diseases affecting the genes encoding the matrix, fusion, and haemagglutinin proteins (Baczko et al, 1986; Cattaneo et al, 1988; Billeter and Cattaneo, 1991;  ter Meulen, 1997) . Similar observations have been made in a rodent model of subacute measles encephalitis suggesting that transcriptional downregulation of MV occurs in the early stages of infection (Schneider-Schaulie s et al, 1989) .

Neuropathologica l and immunopathological studies on autopsy tissue isolated from individuals with SSPE have established that gray matter neurons and their processes are infected at high levels and that many oligodendrocytes are infected in the white matter (Budka et al, 1982; Allen et al, 1996) . Viral antigen is detected within both the nuclei and cytoplasm of oligodendrocytes in SSPE brain tissue (Allen et al, 1996) . Although the infection is widespread in the hemispheric white matter, relatively low numbers of infected oligodendrocytes are observed in the brain stem and spinal cord. Astrocytes are also infected throughout the CNS but to a much lesser degree. The virus is distributed throughout the CNS, from the temporal and frontal cortices, to the medulla, pons, and cervical spinal cord regions. In some cases low numbers of antigen positive neurons have been observed in the cerebellum (Allen et al, 1996) . Such studies have suggested transynaptic spread of the virus in a cephalo-caudal direction. Observations using the differentiated human neuron cell line, NT2, have given support to the hypothesis of transynaptic spread (Lawrence et al, 2000) but as yet the mechanism has not been elucidated. It is also unclear how the neurons or oligodendrocytes initially become infected or how the infection is propagated through the white matter.

Some recent studies on cultured astrocytes and neurons have shown the importance of extended cell processes in cell-to-cel l spread of virus. Cultured human astrocytoma cells have intimately associated extended cell processes that are utilized by MV in the infection of surrounding cells (Duprex et al, 1999) . Studies have also shown that cultured human neurons can become infected via the extended processes that are in contact with more readily infected neuroepithelial cells in mixed cell populations (McQuaid et al, 1998) . Such a mechanism may be due to the presence of virus proteins in the neuroepithelial cell plasma membrane where it contacts the neurons. Furthermore, it has also been shown that viral spread in cultured neurons occurs in the absence of syncytium formation and with minimal extracellular virus production (Lawrence et al, 2000) .

Nearly all of the published work on the susceptibility of CNS cells to MV infection have utilized astrocytic and, to a lesser extent, neuronal cell lines (Miller and Carrigan, 1982; Schneider-Schaulies et al, 1990 McQuaid et al, 1998; Duprex et al, 1999 Duprex et al, , 2000 Lawrence et al, 2000) . Studies on glial cell lines have indicated that, in contrast to nonneural cells, MV transcription can be down-regulated by intrinsic host cell factors, whereas the differentiatio n state of the cells in uences the translation of virus proteins (Schneider-Schaulie s et al, 1993) . Despite these investigations there are very few studies that have utilized either primary cells or oligodendrocyte cell lines to study MV infection in this important cell type. In one previous report, Atkins et al (1991) reported that a nonrodent-adapted strain of Edmonston strain of MV multiplied and produced a cytopathic effect in primary cultures of rat oligodendrocytes (Atkins et al, 1991) . Viral infection in that study was only monitored by cytopathic effects and no viral immunocytochemistry or ultrastructural investigations were undertaken.

In this study, we aimed to establish a model of MV infection of oligodendrocytes and to utilize this model to determine if the virus infection was propagated via cell processes. MO3.13 is an immortal human-human hybrid derived by lectin-enhanced , polyethylene glycol-mediated somatic cell fusion between the thioguanine-resistan t rhabdomyosarcoma mutant RD-TG.6 and primary human oligodendrocytes obtained from cultures of adult temporal lobectomies (Talbot et al, 1993; Ursell et al, 1995) . MO3.13 cells are characterize d as being positive by immunocytochemistry and Western blotting for the oligodendrocyte speci c markers myelin basic protein (MBP) and proteolipid protein (PLP) . The OLN-93 cell line was established from primary cultures of glial cells prepared from the brains of 1-day-old Wistar rats (Richter-Landsberg and Heinrich, 1996) . In some recently published studies (Duprex et al, 1999 (Duprex et al, , 2000 , we have come to appreciate the usefulness of a recombinant MV, which expresses enhanced green uorescence protein (EGFP). This virus was therefore chosen to examine oligodendrocyte cell infection and virus spread from cell-to-cell . The MO3.13 cell line was also analyzed for expression of a receptor used by vaccine strains of MV, CD46 (Naniche et al, 1993 ) and for another recently described MV receptor SLAM (Tatsuo et al, 2000) . In parallel investigations snap-frozen blocks of white matter from an autopsy case of SSPE were examined for the expression of CD46 and SLAM by immunocytochemistry.

At the split ratios used, MO3.13 and OLN-93 cells were highly proliferative and attained conuency within 3 days. Both cell lines expressed markers indicative of oligodendrocyte lineage ( Figure 1A , B). Un xed MO3.13 cells were examined for the presence of CD46 by indirect immuno uoresence. Expression of CD46 was present in localized patches on the cell surface of all cells ( Figure 1C ). MO3.13 cells were negative for SLAM expression ( Figure 1D ). As expected, the B-cell line B95a showed SLAM expression on all cells ( Figure 1D1 ). 

To extend these in vitro observations SSPE tissue sections were immunocytochemically stained for CD46 and SLAM. CD46 was expressed on all cerebral endothelium and on subpopulations of cells in the white matter with the morphological characteristics of oligodendrocytes ( Figure 1E ). Small numbers of neurons in the gray matter were also observed to express CD46. SLAM was absent from cells in the parenchyma of the brain but was expressed on cells within the perivascular in ltrates surrounding blood vessels ( Figure 1F , arrow). SSPE tissue sections were immunohistochemically stained with MAb to measles virus nucleocapsid (N). Based on morphological criteria neurons in the gray matter and oligodendrocytes in the white matter were the cells types predominantly infected with MV from the nine cases of SSPE studied (Allen et al, 1996) . In very rare instances chains of MV-infected interfasicular oligodendrocytes were observed in the white matter ( Figure 1G ). No viral antigen was detectable in processes between such groups of cells. The areas surrounding such chains of cells also contained scattered MV-positive oligodendrocytes (data not shown).

To establish an in vivo model of MV infection of oligodendrocytes , MO3.13 cells were infected with MVeGFP. The cells were readily infectable and up to 90% of cells became infected 48 h postinfection (h.p.i.). EGFP auto uoresence was observed in the cell bodies and ne processes of the infected cells. Infectious virus was recovered from both the supernatant and cell-associate d components. MVeGFP was cultured by ve passages on MO3.13 cells at an MOI of 0.01. Throughout this process, infected cells were evident by GFP uoresence 24 h.p.i. and more than 90% of the cells were infected by 72 to 96 h.p.i. Titres of MVeGFP(verop1) and MVeGFP(MO3.13p5) were obtained in triplicate by TCID50, assay on Vero and MO3.13 cells. During this time, viral titres did not change (Table 1 ), indicating that the virus did not signi cantly adapt to MO3.13 cells. By comparison OLN-93 cells growing on 25-cm 3 asks or glass coverslips could only be infected with MVeGFP at a very low level. EGFP auto uoresence was observed in small clusters of infected OLN-93 cells (approxi- (Duprex et al, 1999) . The punctate nature of the nucleocapsid staining is typical of MV-induced intracytoplasmic inclusion bodies. It is also noteworthy that EGFP auto uoresence can be detected in cells that are negative for viral antigen (Figure 2A , arrow b) con rming previous observations and indicating that utilization of the recombinant virus is a very sensitive means to detect virus-infected cells.

On acetone-xed cells at 48 h.p.i. MV antigens were detected using an SSPE serum, which predominantly detects N and phosphoprotein proteins, visualized by a FITC-conjugated rabbit anti-human secondary antibody. The intermediate lament, vimentin, was used as a counterstain to show both uninfected and infected cells. Vimentin was detected using a MAb visualized by an Alexa 568-conjugated goat antimouse secondary antibody. Once again large accumulations of nucleocapsid were seen in the cytoplasm of infected cells ( Figure 2B inset) . Antigen was also detected in the connecting processes between cells ( Figure 2B , arrows) and in the ne, branching processes of singly infected cells ( Figure 2B , asterisk).

Monolayers of MO3.13 cells growing on 25-cm 3 asks, infected at an MOI of 0.01, were xed and processed for ultrastructural analysis. A high percentage of the cells viewed by EM were infected as evidenced by an abundance of spherical nucleocapsid inclusion bodies dispersed or aggregated throughout the cytoplasm ( Figure 2C , dashed circle). This is consistent with the immunocytochemical data ( Figure 2B , inset). Figure 2D is representative of nucleocapsids in cross-section observed to underlie cell membranes displaying modi cation or thickening (arrows) that are typical of the MV budding structures that form and release mature MV virions. The membranes of MO3.13 cells were examined at greater magni cations to identify areas of membrane thickening or modi cation and evidence of MV budding. Different stages of virus budding were observed along the membranes. In Figure 2E , the photomicrograph demonstrates a typical infected MO3.13 cell with ve mature MV virions (arrows) at the cell surface. The formation of a virus bud begins with thickening of cell membranes, which project outwards from the cell until it pinches off to form a mature MV virion with characteristic surface projections ( Figure 2F, arrows) .

Having established the MO3.13 cell line as a suitable model, we wished to examine the spread of MV in oligodendrocytes. MO3.13 cells were infected at a MOI of 0.01 with MVeGFP and infected areas observed regularly by UV microscopy from 24 h.p.i. Two representative time courses of MVeGFP-infected cells that illustrate both the spread of virus along cell processes and the fusion of virus-infected cell bodies are shown in Figures 3A and B . In Figure 3A , virus spread is mediated via interconnectin g cell processes. At 24.75 h.p.i., EGFP is present in a cell process (arrow) of an infected cell. Within 15 min, GFP is present in the cell body of a neighbouring cell. By 26.5 h.p.i., the infection has progressed to adjacent cells. Figure 3B illustrates an example of fusionmediated spread of MVeGFP. Three neighbouring small syncytia (27 h.p.i., arrows) fuse together over a 3.5-h time period, forming a larger syncytium. By 96 h.p.i., process-and fusion-mediated spread of virus led to infection of the complete cell monolayer ( Figure 3A , insert). Cytopathic effect due to syncytia formation was observed throughout the monolayer.

In the present study, we have demonstrated that MV is capable of infecting the human oligodendrogliom a cell line MO3. 13. This is the rst report of MV infection of a human oligodendrocyte cell line, an important cell type known to be infected in SSPE. The cells are readily infectible by a recombinant MV and infectious virus was produced. By comparison, the rodent oligodendrocyte cell line, OLN-93 displayed a low susceptibility to MVeGFP infection and infectious virus was not released. MO3.13 cells expressed high levels of cell surface CD46 but were negative for the recently described MV receptor SLAM. This is consistent with the observation in SSPE tissue sections where CD46 but not SLAM expression was detected on oligodendrocytes in the white matter. However, the numbers of CD46-positive oligodendrocytes detected by immunohistochemistry was only a small percentage of the total number of oligodendrocytes in the white matter of the brain areas examined. A more detailed description of CD46 and SLAM expression in the normal and MV-infected CNS and peripheral tissues is currently in preparation (McQuaid et al, unpublished results) . It has been shown that not all cells that are susceptible to MV infection express detectable levels of CD46 (Yanagi et al, 1994; Dunster et al, 1995; Horvat et al, 1996) , and SLAM is only constitutively expressed on immature thymocytes, CD45RO high memory T cells, and a proportion of B cells (Sidorenko and Clarke, 1993; Cocks et al, 1995) . The mechanism of MV entry into these cells remains unclear and would seem to indicate that other MV receptors may exist on mammalian cells.

In a previous study, we made use of a human astrocytoma cell line to observe virus spread using a recombinant EGFP-expressing MV virus (Duprex et al, 1999) . Individual infected cells, identi ed by EGFP auto uoresence, were monitored by CSLM and the virus spread was shown to be cell process-mediated with a rapid progression of GFP from cell-to-cell . Utilizing this technique, MO3.13-propagated MVeGFP virus was observed to spread in MO3.13 cell monolayers. Virus propagation through the monolayer was observed to be predominantly a cell processmediated event. The actual mechanism of virus spread from an infected cell process to an adjacent uninfected cell process has not been established in these studies. It is possible that microfusion of cell processes occur with the accompanying passage of viral ribonucleoprotein into an uninfected cell process. However, infected cells were also observed to spread infection by cell body-to-cell body fusion. Adjacent cells become infected and then fuse to form syncytia. These observations have previously been made in MVeGFP-infected murine neuroblastoma cells where virus spreads from cell-to-cel l both by fusion and via cell processes (Duprex et al, 2000) .

MV infection of alpha/beta interferon receptordefective mice expressing human CD46 has been used to suggest that replication is much more efcient in the rodent CNS than the peripheral nervous system, with the virus propagating mostly in the easily accessible ependymal cells (Mrkic et al, 1998) . Viral RNA or antigen was often detected in contiguous cells, suggesting that in the brain of transgenic mice MV propagation may be based largely on lateral cell-to-cel l contacts. This demonstrated both in vivo and in vitro, MV spread in the CNS most likely involves localized fusion events at cell-to-cel l contact points without the requirement for speci c viral receptor(s) (McQuaid et al, 1998; Duprex et al, 1999; Lawrence et al, 2000) . Similarly, in the present study we have shown that human oligodendrogliom a cells in culture spread MV infection both by fusion and along interconnectin g processes.

Previous ultrastructural investigations of MV infected astrocyte and neuronal cell lines have demonstrated viral nucleocapsids dispersed or in small clumps throughout the cytoplasm (Macintyre and Armstrong 1976; McQuaid et al, 1998; Lawrence et al, 2000) . EM observations of astroglial cultures displayed MV in various stages of assembly and extracellular virions were routinely observed. These mature virions have surface projections, thickened envelope membranes, and the nucleocapsid tubules are observed to be dispersed peripherally in a spiral form (Macintyre and Armstrong, 1976) . Electron microscopy of undifferentiated and differentiated NT2 cells showed that viral budding occurred very rarely in undifferentiated cells and was not observed on the surface of differentiate d NT2 cells (McQuaid et al, 1998; Lawrence et al, 2000) . However, nucleocapsids were aligned at the cell membrane of differentiate d NT2 cells, in neuronal processes and at presynaptic neuronal membranes. Ultrastructural analysis of MV-infected MO3.13 cells revealed an abundance of MV nucleocapsid within the cytoplasm, both in isolation and as aggregates . Typical plasma membrane modi cations and various stages of budding (Fleury et al, 1980) were also observed consistent with the productive nature of MV infection from MO3.13 cells.

This pattern of MV nucleocapsid localizatio n mirrored observations made with single-and duallabeled immuno uoresence of MVeGFPp5-infected MO3.13 cells. MV nucleocapsid was observed throughout the cell cytoplasm with dense clumps of virus evident in the perinuclear regions. When duallabeled for MV antigens and EGFP, it was evident that infected MO3.13 cells expressed GFP auto uoresence in the absence of detectable MV antigen. Such observations have led to the conclusion that EGFP expression provides an early indicator of MV infection in vitro (Duprex et al, 1999) .

In the adult CNS, fully differentiate d interfasicular oligodendrocytes occur in the white matter and are characterize d by their many connections to segments (internodes) of myelin sheaths wrapped around axons. The processes, which link the oligodendrocyte cell body to the sheath, are narrow and tortuous (Knobler et al, 1974) . The paranodal regions of oligodendrocyte can also contact other glial cells of the CNS such as astrocyte by gap junctions, indicating functional coupling (Berry et al, 1995) . Gap junctions establish intracellular channels of communication through which ions and small solutes (<1300 D) can pass (Mugnaini, 1982) . However, it is unlikely that viral nucleocapsid could pass from cell-to-cel l via intact gap junctions. Occasionally interfasicular oligodendrocytes may be seen aligned in rows. Where this occurs, the cell membranes of adjacent cells are in intimate contact but lack speci c junctional contacts. In very rare examples, in the white matter from cases of SSPE, chains of interfasicular oligodendrocytes were demonstrated to have detectable levels of MV. However, in such autopsy tissue, xed for light microscopy, it is impossible to determine if virus is present in the very ne oligodendrocyte processes in the white matter. Oligodendrocytes also occur in the gray matter as perineuronal satellite cells. Individual satellite cells are thought to be in close contact with single neurons. In some pathological situations, such cells have been described to have processes extending to myelin sheaths (Ludwin, 1979) . The extent of oligodendrocyte processes in the CNS and the ndings, in the present study, of extensive infection of oligodendrocyte processes in vitro raises the possibility of oligodendrocyte-to-neuro n infection in the CNS as a means of initiating spread of virus throughout the CNS. It is postulated that, once present in neurons, MV can spread transneuronally throughout the CNS (Allen et al, 1996; Lawrence et al, 2000) , possibly by fast axonal transport (Oldstone et al, 1999) .

Like a small number of other viruses, MV has been identi ed as potentially persisting in the CNS (Liebert, 1997) . For example, Borna virus causes CNS disease in several species, and recent studies have suggested a potential role for these viruses in human mental health (Gonzalez-Duni a et al, 1997) . The polyomavirus JC, which infects oligodendrocytes in vivo, is associated with most cases of progressive multifocal leukoencephalopath y, a demyelinating disease of the CNS leading to death within months of rst presentation (Askamit, 1995; Eggers et al, 1999) . However, the speci c cell type or neuroanatomical location(s) of CNS viral persistence remains unknown. The nding that MO3.13 cells can sustain persistent coronavirus infections (Arbour et al, 1999a (Arbour et al, , 1999b raises the possibility that other viruses, such as measles, may persist in oligodendrocytes in vivo.

A logical progression of these experiments will be to use primary oligodendrocyte cultures (McCarthy and De Vellis, 1980; Gates et al, 1985) to analyze virus spread. It would also be important to analyze CD46/SLAM expression on mature primary oligodendrocytes.

Two cells lines of oligodendrocyte lineage, human MO3.13 (Talbot et al, 1993) and rat OLN-93 (Richter-Landsberg and Heinrich, 1996) , were maintained in Dulbecco's Modi ed Eagle Medium (DMEM) supplemented with 10% (v/v) fetal calf serum (FCS), 0.1% penicillin and streptomycin, and 4 mM glutamine in 25-cm 3 asks at 37 ± C with 5% (MO3.13) or 10% (OLN-93) CO 2 added. OLN-93 cells have been shown to express the oligodendrocytespeci c markers, galactocerebroside , MBP, PLP, and myelin-associated glycoprotein (MAG) (Richter-Landsberg and Heinrich, 1996; Strelau and Unsicker, 1999) . When con uent, cells were passaged at a 10:1 (MO3.13) or 20:1 (OLN-93) split ratio into 25-cm 3 asks.

Virus infection of cells MO3.13 and OLN-93 monolayers grown in 25-cm 3 asks were infected with adapted MVeGFP virus that had previously been grown on Vero cells (Duprex et al, 1999) . Cell monolayers (60% con uency) were rinsed with maintenance medium (DMEM supplemented with 2% FCS) and cells were infected with virus at a multiplicity of infection (MOI) of 0.01 and incubated for 1 h at 37 ± C. After this time unadsorbed virus was removed, maintenance medium was added, and cells were incubated at 37 ± C for varying time periods. Cell sheets were monitored by UV microscopy (Leica) daily for the appearance of MVeGFP positive cells. For adaptation of MVeGFP to oligodendrocyte cell lines, heavily infected monolayers (>90%) were freeze-thawed three times. Virus stocks were stored at ¡70 ± C. Virus titres were determined by TCID 50 s (Reed and Muench, 1938) , and the asks were monitored by uorescence microscopy to determine that MVeGFP expression was retained.

Indirect immuno uoresence was used to characterize the cells with commonly used oligodendrocyte markers and for the presence/absence of MV receptors. Cells were grown to con uence on glass coverslips, then xed in ice-cold acetone. Primary antibodies were applied for 1 h at 37 ± C and the coverslips were washed twice for 5 min in PBS. Secondary antibodies, rabbit anti-mouse Alexa 488, or goat antirabbit Alexa 488 (Molecular Probes) were diluted in PBS (1:500) and incubated for 1 h at 37 ± C. Nuclei were then counterstained in propidium iodide (Sigma 2¹g/ml) for 30 s, washed in PBS, and the coverslips mounted using Citi our (Amersham). Primary antibodies used to characterize the cells were polyclonal antibodies to galactocerebrosid e (Chemicon, 1:20), PLP (Chemicon, 1:20) , and MAb to vimentin (Dako 1:50).

A polyclonal anti-CD46 antibody (1:200, Gift from F. Wild, Institut Pasteur de Lyon, France) was used to examine the expression of the MV receptor CD46 on MO3.13 cells. The antibody was applied to un xed cells overnight at 4 ± C. Cells were washed in PBS and xed in ice-cold acetone. To examine expression levels of SLAM on MO3.13 cells, coverslips were xed in 10% formalin and washed in PBS. MAb to SLAM (Kamiya Biomedical Company, 1:1000) was applied overnight at 4 ± C. For CD46 detection, cells were then incubated in Alexa 488 goat anti-rabbit (Molecular Probes 1:500) and for SLAM detection cells were incubated in Alexa 488 rabbit anti-mouse (Molecular Probes 1:500) for 1 h at 37 ± C. Following further washes, nuclei were counterstained with propidium iodide and the coverslips mounted in Citi our. The B-cell line, B95a, was grown on glass coverslips and immunostained as described previously as a positive control for SLAM expression in cultured cells.

For virus detection, cells grown on glass coverslips in Petri dishes to 60% con uency were infected at an MOI of 0.01 as before. Infected cells were incubated for various time periods at 37 ± C. Cells were xed in either ice-cold acetone or 4% paraformaldehyde for 10 min, depending on the antibody to be used. Anti-measles virus N MAb and hyperimmune SSPE serum were used as described previously (Duprex et al, 2000) . Primary antibodies were incubated on the cells for 1 h at 37 ± C. Incubation was followed by two 5-min washes with PBS. The secondary antibodies used for the detection of single-or duallabeling were Alexa 488 rabbit anti-mouse, Alexa 568 goat anti-mouse, and rabbit anti-human FITC (Dako 1:50). These were diluted in PBS and added to the coverslips and incubated for 1 h at 37 ± C. Coverslips dual-labeled with SSPE serum and vimentin were subsequently incubated in rabbit anti-human FITC and Alexa 568 goat anti-mouse. Coverslips were washed twice in PBS and mounted using Citi our.

A Leica TCS/NT confocal scanning laser microscope (CSLM) equipped with a krypton/argon laser was used to examine the samples for uoresence. Alexa 488-labelled samples or MVeGFP auto uoresence was visualized by excitation at 488 nm with a 506-538 band-pass emission lter. Alexa 568labelled samples were imaged by excitation at 568 nm with a 564-596 band-pass emission lter.

Cells were grown to 60% con uence in 25-cm 3 tissue culture asks. Cells were infected at an MOI of 0.01 with MVeGFP. As previously described, an inverted UV microscope was used to monitor the monolayers for the appearance of infected cells (Duprex et al, 1999) . In initial experiments, asks were oriented on the microscope stage and marked to permit the repeated observation of chosen groups of infected cells in the monolayers. Observations were made over a period of 24 h at hourly intervals. In additional experiments singly infected or infected groups of cells were selected and the time-lapse ability of the confocal system was used to acquire Z-series images every 20 min over 48 h (Duprex and Rima, 2001) .

Infected MO3.13 cells were xed and embedded for ultrastructural analysis. Monolayers were xed in 2.5% glutaraldehyde for 90 min at 4 ± C, then rinsed in 0.2 M cacodylate buffer for 30 min at 4 ± C. Cells were then post xed for 1 h at room temperature in 2% osmium tetroxide and rinsed in distilled water. Monolayers were dehydrated in graded alcohols and proplyene oxide was used to detach the cells from the ask. Cells were subsequently embedded in Agar 100 embedding resin as previously described (McCormack et al, 1983) . Semithin sections were stained in toludine blue to identify areas with high numbers of cells and ultrathin sections (90 nm) were cut from these regions. Sections were lifted onto copper EM grids and stained with uranyl acetate and lead citrate. Sections were examined on a Hitachi H-600 transmission electron microscope.

Cryostat sections (12 ¹m) were cut from snap-frozen tissues from a SSPE case and xed in 10% formalin. After blocking endogenous peroxidase in 0.5% H 2 O 2 in methanol for 10 min, sections were incubated in polyclonal anti-CD46 or MAb to SLAM overnight at 4 ± C. Furthermore, selected blocks of predominantly white matter tissue from nine autopsy cases of SSPE were immunohistochemically stained for MV antigens as described previously (Allen et al, 1996) . Bound receptor or viral antibodies were detected using diaminobenzidine or aminoethylcarbazole , as peroxidase substrate. Sections were counterstained with haematoxylin. Photomicroscopy was carried out on a Leitz Aristoplan tted with differential interference contrast.

Immunosuppressive measles encephalitis in a patient with a renal transplant

The signi cance of measles virus antigen and genome distribution in the CNS in SSPE for mechanisms of viral spread and demyelination

Acute and persistent infection of human neural cell lines by human coronavirus OC43

Persistent infection of human oligodendrocytic and neuroglial cell lines by human coronavirus 229E

Progressive multifocal leucoencephalopathy : a review of the pathology and pathogenesis

Multiplication of rubella and measles viruses in primary rat neural cell cultures: relevance to a postulated triggering mechanism for multiple sclerosis

Expression of defective measles virus genes in brain tissues of patients with subacute sclerosing panencephalitis

Nervous system

Molecular biology of defective measles viruses persisting in the human central nervous system

Measles virus antigen in panencephalitis. An immunomorphological study stressing dendritic involvement in SSPE

Biased hypermutation and other genetic changes in defective measles viruses in human brain infections

A novel receptor involved in T-cell activation

Moesin, and not the murine functional homologue (Crry/p65) of human membrane cofactor protein (CD46), is involved in the entry of measles virus (strain Edmonston) into susceptible murine cell lines

Observation of measles virus cell-to-cell spread in astrocytoma cells by using a green uorescent protein-expressing recombinant virus

In vitro and in vivo infection of neural cells by a recombinant measles virus expressing enhanced green uorescent protein

Green-uorescent protein. Measles virus cell to cell spread; real-time visualisations

Quanti cation of JC virus DNA in CSF of patients with HIV associated PML: a longitudinal study

Differences in early ultrastructural aspects of the replication of measles and subacute sclerosing panencephalitis viruses in a cell culture from a human astrocytoma

The pathogenicity of the A7, M9 and L10 strains of Semliki Forest virus for weanling mice and primary mouse brain cell cultures

Borna disease virus and the brain

Transgenic mice expressing human measles virus (MV) receptor CD46 provide cells exhibiting different permissivities to MV infections

Oligodendroglial ensheathment of axons during myelination in the developing rat central nervous system. A serial section electron microscopical study

Measles virus spread between neurons requires cell contact but not CD46 expression, syncytium formation, or extracellular virus production

Measles virus infections of the central nervous system

The perineuronal satellite oligodendrocyte. A role in remyelination

Fine structural changes in human astrocyte carrier lines for measles virus

Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue

The establishment and characterisation of a cell line and mouse xenografts from a human malignant melanoma

A human glial hybrid cell line differentially expressing genes subserving oligodendrocyte and astrocyte phenotype

Measles virus infection and replication in undifferentiated and differentiated human neuronal cells in culture

Reversible repression and activation of measles virus infection in neural cells

Measles virus spread and pathogenesis in genetically modi ed mice

Membrane specializations in neuroglial cells and at neuron-glia contacts

Human membrane cofactor protein (CD46) acts as a cellular receptor for measles virus

Measles virus infection in a transgenic model: virus-induced immunosuppression and central nervous system disease

A simple method for estimating fty percent endpoints

OLN-93: a new permanent oligodendroglia cell line derived from primary rat brain glial cultures

Temporal and geographical distribution of measles virus genotypes

Restriction of measles virus gene expression in acute and subacute encephalitis of Lewis rats

Restricted expression of measles virus in primary rat astroglial cells

Spontaneous and differentiation-dependent regulation of measles virus gene expression in human glial cells

Characterization of a cell surface glycoprotein IPO-3, expressed on activated human B and T lymphocytes

Expression and function in two oligodendroglial cell lines representing distinct stages of oligodendroglial development

Neurotropism of human coronavirus 229E

SLAM (CDw150) is a cellular receptor for measles virus

Molecular and cellular aspects of measles virus persistence in the CNS

Subacute sclerosing panencephalitis

Localization and partial characterization of a 60 kDa citrulline-containing transport form of myelin basic protein from MO3.13 cells and human and white matter

Measles virus infects mouse broblast cell lines, but its multiplication is severely restricted in the absence of CD46